Remote Sensing the Margins of the Gold Trade: Ethnohistorical archaeology and GIS analysis of five gold trade networks in Luzon, Philippines, in the last millennium BP 9781407357065, 9781407357072

Table of contents :
Cover
Title page
Copyright page
Dedication
Acknowledgements
Of Related Interest
Abstract
Contents
List of Figures
List of Abbreviations
1. Introduction: Interdisciplinary and Multiscalar Ethnohistorical Archeology
1.1 ‘Small Worlds’ towards Modeling Evanescent and Permanent Markets
1.2 Interdisciplinary Methods: Historical, Archeological, Ethnological, and Geographical/Spatial, Remote Sensing Studies
1.3 A Multiscalar Approach to Examining Complex Regional Production and Trade Networks
1.4 Overview of Book
1.5 Significance of this Book
2. Background: Northwestern Luzon Production and Trade Systems – An Overview through Ethnohistory and Archeology
2.1 Gold and Other Mining and Production Networks Elsewhere in the World
2.2 Philippine Islands Societies in the Tenth–Sixteenth Centuries and the Seventeenth–Twentieth Centuries Colonial World: A Brief History of Origins and Coastal–Upland Interactions
2.3 Northwestern Luzon Social, Political, and Economic Patterns of the Early Historic and Colonial Periods: An Overview of What We Know from Ethnohistory and Archeology
2.3.1 At the Time of Spanish Contact
2.3.2 Flexibility as Resilience
2.4 Chapter Summary
3. Historically and Archeologically Documented Networks for Gold Working And Coastal Markets in Northwestern Luzon
3.1 The Tonglo Network Linking the Mines of Balatok and Acupan with the Coast of Aringay and Agoo
3.1.1 Aringay and Agoo Evanescent Markets
3.1.2 Tonglo, Gold Bulking Village
3.2 The Gasweling Network Linking Apayao, Locjo, Cabcaben, and Catampan Mines with Calincamasan, Baratao, and Bauang on the Coast
3.2.1 Some Ethnohistorically Documented Settlements in the Gasweling Network
3.2.2 Juan de Salcedo 1572 Account of His Conquest of Northwestern Luzon
3.2.3 Alonso Quirante 1624 Account of His Expeditionary Force to Benguet and the Assaying of Gold Nuggets
3.3 The Lepanto Network – Apayao and Danac linking Lepanto Mines with Tagudin and Purao
3.3.1 Lepanto Mines as the Oldest Worked Mine in the Cordilleras
3.3.2 Purao and Tagudin Coastal Settlements
3.3.3 Danac and Apayao Bulking Stations
3.4 Angaqui Network Linking Patiacan/Minlaoi and Possibly Lepanto Mines with Dumaquaque and Candon in the Coastal Account
3.4.1 Ilocano Sampans Loaded with Gold Headed for China and Porcelain on its Return Trip
3.4.2 Gold from the Ilocos Coastline
3.4.3 Ethnohistorical Accounts of the Slave-for-Gold Exchange and Strategic Trade Alliance Formation
3.4.4 Overland, Riverine, and Maritime Trade Routes of the Angaqui Network
3.4.5 Ethnohistorically and Archeologically Documented Settlements of the Angaqui Network
3.5 The Abra Network as the Epitome of a More Permanent Market System in Northwestern Luzon
3.5.1 Tingguian/Itneg of Abra and Their Accounts on Early Gold
3.5.2 Coastal Port Settlements of Santa Maria, Narvacan, Vigan/Caoayan, Magsingal, Cabugao-Nagsingcaoan, and Calanutian
3.5.3 Bangued, Tayum, Bucao, Lagang-ilang Bulking Stations and the Tineg Jump-off Point to Cagayan Valley in the East
3.6 Chapter Summary
4. Theoretical Trade Models for Multiscalar Archeology: Social Network Analysis, Worlds Systems, Interaction Spheres, and Agency
4.1 Social Network Analysis
4.2 Multiscalar Interaction: Models from World Systems, Social Network Analysis, Interaction Spheres, and Agency Theory
4.2.1 World Systems Theory, implications for archeology
4.2.2 World Systems Analysis
4.2.3 Negotiated Peripherality
4.2.4 Multiscalar Approach
4.2.5 Trade Diaspora, Distance Parity Model
4.2.6 Regional/Multiscalar Studies of Lowland–Upland trade in Philippine and Southeast Asian archeology
4.2.7 Agency-Reciprocity, Utang Na Loob
4.3 Chapter Summary
5. Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis
5.1 Importance of trails and transportation routes in ancient settlements
5.2 Tried and tested approaches to identifying trails
5.3 Geographic Information Systems and Its Utility in Modeling Trails and Transportation Networks
5.4 Geographic Information Systems and Remote Sensing Methods for the Tonglo Region
5.5 Methods used for Gasweling Region
5.6 Methods used for Lepanto Region
5.7 Methods for the Angaqui Region
5.8 Method Used in Abra Region
5.9 Modeling reduccion and Trade Connections through Social Network Analysis
5.10 Chapter Summary
6. Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing
6.1 Tonglo
6.2 Gasweling
6.3 Lepanto
6.4 Angaqui
6.5 Abra
6.6 Abra Northeast Branch SNA
6.6.1 Reduccion and inter-settlement trade – the Northeastern Abra River Branch
6.7 Chapter Summary
7. Conclusion: From Agency to Marginality in the Dynamic Gold Trading of Northwestern Luzon
Bibliography

Citation preview

L E A IN N L IO ON IT D L D IA A ER AT

M

Remote Sensing the Margins of the Gold Trade Ethnohistorical archaeology and GIS analysis of five gold trade networks in Luzon, Philippines, in the last millennium BP Michael Armand P. Canilao B A R I N T E R NAT I O NA L S E R I E S 2 9 8 8

2020

Remote Sensing the Margins of the Gold Trade Ethnohistorical archaeology and GIS analysis of five gold trade networks in Luzon, Philippines, in the last millennium BP Michael Armand P. Canilao B A R I N T E R NAT I O NA L S E R I E S 2 9 8 8

2020

Published in 2020 by BAR Publishing, Oxford BAR International Series 2988 Remote Sensing the Margins of the Gold Trade isbn isbn doi

978 1 4073 5706 5 paperback 978 1 4073 5707 2 e-format

https://doi.org/10.30861/9781407357065

A catalogue record for this book is available from the British Library © Michael Armand P. Canilao 2020 WorldView2 satellite imagery of the Tila-Angaqui trail using the vegetation band combination (Courtesy of the DigitalGlobe Foundation).

cover image

The Author’s moral rights under the 1988 UK Copyright, Designs and Patents Act, are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher. Links to third party websites are provided by BAR Publishing in good faith and for information only. BAR Publishing disclaims any responsibility for the materials contained in any third party website referenced in this work.

BAR titles are available from: BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK email [email protected] phone +44 (0)1865 310431 fax +44 (0)1865 316916 www.barpublishing.com

To my parents Nars and Carl on your 36th Wedding Anniversary. To my brother JC, and my sister Ceres. To my sons Kan and Anond. To my wife Poi.

Acknowledgements I would like to acknowledge the support of the following institutions and individuals: the Digital Globe Foundation, Hexagon Geospatial, University of Illinois at Chicago (UIC) Chancellor’s Research Award through the Graduate College UIC, American Council for Learned Societies (ACLS)/Henry Luce Award through the Department of Anthropology UIC, the Joshua Terry Award through the Department of Anthropology UIC, the Robert Corley Memorial Scholarship through the Liberal Arts and Sciences UIC, the Chicago Consular Corps Award of the UIC Office of International Affairs, the Provincial Government of Ilocos Sur through Hon. Deogracias Victor Savellano, the Field Museum, the University of Michigan Museum of Anthropological Archeology, the National Archives of the Philippines, Archivo General Militar de Madrid, Archivo del Museo Naval Organo de Historia y Cultura Naval Madrid, Mr. Michael G. Price (Jackson, Michigan, United States of America USA), MLI Ingel (Vigan, Ilocos Sur, Philippines), the National Museum of the Philippines, the National Commission on Indigenous Peoples Philippines, the Archeological Studies Program of the University of the Philippines Diliman, the Cordillera Studies Center of the University of the Philippines Baguio, the National Committee in Historical Research, the National Commission on Culture and the Arts Philippines, the Research Assistance Program of the Philippine Social Science Center, PhilGIS.org, the National Mapping and Resource Information Agency and its main agency Department of Environment and Natural Resources (DENR), Ministry of Economy Trade and Industry (METI) and National Aeronautics and Space Administration (NASA) for Advanced Spaceborne Thermal Emission and Reflection Global Digital Elevation Model (ASTER GDEM), Ucinet for Windows: Software for Social Network Analysis (Borgatti, S.P., Everett, M.G. and Freeman, L.C. 2002. Harvard, MA: Analytic Technologies), Mines and Geosciences Bureau of DENR, and the Yanoria’s of Sinait, Ilocos Sur. My heartfelt gratitude to Dr Laura Lee Junker, my PhD advizer at UIC. I would also like to acknowledge my UIC Professors, Dr John Monaghan, Dr Robert ‘Bob’ Hasenstab (my Masters in Environmental and Urban Geography, Comprehensive Examinations Committee Chair), Dr Grace Barretto-Tesoro, Dr Mitch Hendrickson, Dr Vincent LaMotta, and Dr William Parkinson, Dr Cecilia Smith, Dr Sloan Williams, Dr Brian Bauer, Ms. Melanie Kane, Mr. Jamie Kelly, Dr John Terrell, Dr Tarini Bedi, Dr Mark Liechty, Dr Mark Golitko, Dr Joel Palka, Dr Laura Klein, Dr Lisa Niziolek, and Dr Lawrence Keeley. I also want to thank my academic advizers in the University of the Philippines: Dr Victor Paz (University of the Philippines Diliman) and Dr June Prill-Brett (University of the Philippines Baguio).

iv

Of Related Interest Tibetan Silver, Gold and Bronze Objects and the Aesthetics of Animals in the Era before Empire Cross-cultural reverberations on the Tibetan Plateau and soundings from other parts of Eurasia John Vincent Bellezza BAR International Series 2984

Oxford, BAR Publishing, 2020

The ‘Crescent-Shaped Cultural-Communication Belt’: Tong Enzheng’s Model in Retrospect An examination of methodological, theoretical and material concerns of long-distance interactions in East Asia Edited by Anke Hein BAR International Series 2679

Oxford, BAR Publishing, 2014

The Silk Roads of the Northern Tibetan Plateau during the Early Middle Ages (from the Han to Tang Dynasty) Tao Tong BAR International Series 2521

Oxford, BAR Publishing, 2013

Crossing the Straits: Prehistoric Obsidian Source Exploitation in the North Pacific Rim Edited by Yaroslav V. Kuzmin and Michael D. Glascock BAR British Series 2152

Oxford, BAR Publishing, 2010 Identity and Reciprocity in 15th Century Philippines Grace Barretto-Tesoro

BAR British Series 1813

Oxford, BAR Publishing, 2008

For more information, or to purchase these titles, please visit www.barpublishing.com v

Abstract Gold studies on the Indian Ocean–West Philippine Sea/South China Sea (IO–WPS/SCS) world system have tended to focus on global and often homogenous patterns in the fields of archeology and history. However, there is increasing interest in pursuing the gold studies by starting from the putative ‘peripheries’ from where gold nuggets were mined in crude tunnels or panned in streams. This work provides a ‘bottom-up’ view of culturally contingent, historically situated engagement by local Northwestern Luzon island gold miners and traders as a local point of entry into a IO– WPS/SCS world system. The research shows how Igorot societies negotiated their peripherality in the expansive porcelain-for-gold exchange system that was creeping onto their shores. The research looks at how the Igorot miners practiced their agency through their participation in tabutabuans or evanescent market encounter at the coastal trading centers. The book also explores the role of debt in the receding agentive positioning of these Igorot gold producers. The book posits that five gold trade networks in the region present varying degrees of commitment to this articulation with the IO–WPS/SCS World System during the Early Historical Period (EHP) to Historical Period (HP). Most historians working in the same region use the term Early Modern Period (EMP), which significantly corresponds with what most archeologists working in the region call, on the other hand, the EHP. Some of the networks feature a more evanescent market encounter which allows the miners from the uplands to dictate or strongly negotiate their position in the transaction for gold with the lowland coastal polities. This mainly manifests in the liberty these upland miners choose when undertaking gold mining and gold trading activities unbeholden to the coastal trading centers schedule. Other networks on the other hand feature the opposite, which is a more permanent market in full articulation into the trade network, featuring full dependency on the influx of exotic goods from the system, specifically prestige goods such as export ceramics. In this model the upland miners are fully beholden to their coastal trading partners schedule and are forced to work mines to pay for their debts (utang na loob). Meanwhile, other networks feature a quasi-/semi-permanent market, which has a tendency to fluctuate between evanescent market encounters and permanent markets. Historical events such as colonization, inter-tribal wars, and pestilence may have a bearing on the shifts seen in these semi-permanent markets. The models/findings in the book are based on multiscalar and inter/multidisciplinary analyses using regional geographic information systems (GIS) data, high-resolution multispectral satellite remote sensing data, ethnographic data, primary and secondary written historical data, archival maps and images, oral tradition data, and archeological data on the Early Historical Period to Historical Period. Multiscalarity is a specific feature of this book, demonstrated by weaving together micro/single or composite archeological or historical site data into regional GIS data. The strength of the book is its presentation of a palimpsest of data that in its totality is arguably more powerful than its separate component parts. Indeed, this work looks at the forest of methods rather than just the individual method tree.

vi

Contents List of Figures...................................................................................................................................................................... ix List of Abbreviations............................................................................................................................................................ xi 1. Introduction: Interdisciplinary and Multiscalar Ethnohistorical Archeology.......................................................... 1 1.1 ‘Small Worlds’ towards Modeling Evanescent and Permanent Markets................................................................... 1 1.2 Interdisciplinary Methods: Historical, Archeological, Ethnological, and Geographical/Spatial, Remote Sensing Studies......................................................................................................................................... 7 1.3 A Multiscalar Approach to Examining Complex Regional Production and Trade Networks.................................... 7 1.4 Overview of Book...................................................................................................................................................... 7 1.5 Significance of this Book........................................................................................................................................... 8 2. Background: Northwestern Luzon Production and Trade Systems – An Overview through Ethnohistory and Archeology................................................................................................................................... 9 2.1 Gold and Other Mining and Production Networks Elsewhere in the World.............................................................. 9 2.2 Philippine Islands Societies in the Tenth–Sixteenth Centuries and the Seventeenth–Twentieth Centuries Colonial World: A Brief History of Origins and Coastal–Upland Interactions.................................... 10 2.3 Northwestern Luzon Social, Political, and Economic Patterns of the Early Historic and Colonial Periods: An Overview of What We Know from Ethnohistory and Archeology................................................... 11 2.3.1 At the Time of Spanish Contact........................................................................................................................ 15 2.3.2 Flexibility as Resilience................................................................................................................................... 16 2.4 Chapter Summary.................................................................................................................................................... 17 3. Historically and Archeologically Documented Networks for Gold Working And Coastal Markets in Northwestern Luzon........................................................................................................................................... 19 3.1 The Tonglo Network Linking the Mines of Balatok and Acupan with the Coast of Aringay and Agoo................. 19 3.1.1 Aringay and Agoo Evanescent Markets........................................................................................................... 19 3.1.2 Tonglo, Gold Bulking Village.......................................................................................................................... 21 3.2 The Gasweling Network Linking Apayao, Locjo, Cabcaben, and Catampan Mines with Calincamasan, Baratao, and Bauang on the Coast........................................................................................................................ 21 3.2.1 Some Ethnohistorically Documented Settlements in the Gasweling Network................................................ 21 3.2.2 Juan de Salcedo 1572 Account of His Conquest of Northwestern Luzon....................................................... 25 3.2.3 Alonso Quirante 1624 Account of His Expeditionary Force to Benguet and the Assaying of Gold Nuggets................................................................................................................................................... 26 3.3 The Lepanto Network – Apayao and Danac linking Lepanto Mines with Tagudin and Purao................................ 29 3.3.1 Lepanto Mines as the Oldest Worked Mine in the Cordilleras........................................................................ 29 3.3.2 Purao and Tagudin Coastal Settlements........................................................................................................... 29 3.3.3 Danac and Apayao Bulking Stations................................................................................................................ 32 3.4 Angaqui Network Linking Patiacan/Minlaoi and Possibly Lepanto Mines with Dumaquaque and Candon in the Coastal Account............................................................................................................................ 32 3.4.1 Ilocano Sampans Loaded with Gold Headed for China and Porcelain on its Return Trip............................... 32 3.4.2 Gold from the Ilocos Coastline........................................................................................................................ 37 3.4.3 Ethnohistorical Accounts of the Slave-for-Gold Exchange and Strategic Trade Alliance Formation............ 37 3.4.4 Overland, Riverine, and Maritime Trade Routes of the Angaqui Network...................................................... 38 3.4.5 Ethnohistorically and Archeologically Documented Settlements of the Angaqui Network............................ 39 3.5 The Abra Network as the Epitome of a More Permanent Market System in Northwestern Luzon......................... 50 3.5.1 Tingguian/Itneg of Abra and Their Accounts on Early Gold........................................................................... 50 3.5.2 Coastal Port Settlements of Santa Maria, Narvacan, Vigan/Caoayan, Magsingal, CabugaoNagsingcaoan, and Calanutian......................................................................................................................... 54 3.5.3 Bangued, Tayum, Bucao, Lagang-ilang Bulking Stations and the Tineg Jump-off Point to Cagayan Valley in the East.............................................................................................................................. 57 3.6 Chapter Summary.................................................................................................................................................... 57

vii

Remote Sensing the Margins of the Gold Trade 4. Theoretical Trade Models for Multiscalar Archeology: Social Network Analysis, Worlds Systems, Interaction Spheres, and Agency........................................................................................................................... 65 4.1 Social Network Analysis.......................................................................................................................................... 65 4.2 Multiscalar Interaction: Models from World Systems, Social Network Analysis, Interaction Spheres, and Agency Theory............................................................................................................................................... 66 4.2.1 World Systems Theory, implications for archeology....................................................................................... 67 4.2.2 World Systems Analysis................................................................................................................................... 68 4.2.3 Negotiated Peripherality................................................................................................................................... 69 4.2.4 Multiscalar Approach....................................................................................................................................... 70 4.2.5 Trade Diaspora, Distance Parity Model........................................................................................................... 70 4.2.6 Regional/Multiscalar Studies of Lowland–Upland trade in Philippine and Southeast Asian archeology....... 71 4.2.7 Agency-Reciprocity, Utang Na Loob............................................................................................................... 72 4.3 Chapter Summary.................................................................................................................................................... 72 5. Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis............ 75 5.1 Importance of trails and transportation routes in ancient settlements...................................................................... 75 5.2 Tried and tested approaches to identifying trails..................................................................................................... 77 5.3 Geographic Information Systems and Its Utility in Modeling Trails and Transportation Networks....................... 80 5.4 Geographic Information Systems and Remote Sensing Methods for the Tonglo Region....................................... 81 5.5 Methods used for Gasweling Region....................................................................................................................... 83 5.6 Methods used for Lepanto Region........................................................................................................................... 83 5.7 Methods for the Angaqui Region............................................................................................................................. 86 5.8 Method Used in Abra Region................................................................................................................................... 86 5.9 Modeling reduccion and Trade Connections through Social Network Analysis..................................................... 89 5.10 Chapter Summary.................................................................................................................................................. 90 6. Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing................................................................................................................................................ 93 6.1 Tonglo...................................................................................................................................................................... 93 6.2 Gasweling................................................................................................................................................................. 93 6.3 Lepanto..................................................................................................................................................................... 94 6.4 Angaqui.................................................................................................................................................................... 95 6.5 Abra.......................................................................................................................................................................... 99 6.6 Abra Northeast Branch SNA.................................................................................................................................. 105 6.6.1 Reduccion and inter-settlement trade – the Northeastern Abra River Branch............................................... 105 6.7 Chapter Summary.................................................................................................................................................. 105 7. Conclusion: From Agency to Marginality in the Dynamic Gold Trading of Northwestern Luzon.................... 107 Bibliography......................................................................................................................................................................111 The digital appendix referenced in the text is available to download from: www.barpublishing.com/additionaldownloads.html

viii

List of Figures Figure 1.1. Five gold trade networks in Northwestern Luzon.............................................................................................. 2 Figure 1.2. Diagram showing multidisciplinary datasets...................................................................................................... 3 Figure 1.3. Interpretive Framework used in the book........................................................................................................... 3 Figure 1.4. Map of change through time highlighting Angaqui, Abra networks in particular to illustrate shift from evanescent to permanent market (EHP tenth to sixteenth century)...................................................................................... 4 Figure 1.5. Relevant oral tradition, written historical or archeological data sources used in delineating Northwestern Luzon settlements on the book basemap (Early Historical Period tenth to sixteenth century and Historical Period seventeenth to early twentieth century)................................................................................................................................ 5 Figure 1.6. Map showing data on gold and tradeware quantities or gold and tradeware culture found on relevant settlements of Northwestern Luzon...................................................................................................................................... 6 Figure 3.1. Composite Map showing Tonglo gold trade network in Northwestern Luzon................................................ 22 Figure 3.2. Tonglo in Mount Calugong, westerly view from Tongshu locality in Mount Santo Tomas............................ 23 Figure 3.3. Composite map. Note Gasweling gold trade network in Northwestern Luzon................................................ 24 Figure 3.4. National Mapping and Resource Information Authority 1:250000 map showing Tabio Gold atop Apayao mines reprojected to WGS 84 UTM Zone 51Q.................................................................................................................. 26 Figure 3.5. Map showing Juan de Salcedo’s 1572 seaborne conquest of Northwest Luzon.............................................. 27 Figure 3.6. Route of the Expeditionary force of Alonso Quirante to the gold mines in 1623............................................ 28 Figure 3.7. Composite map. Note Lepanto gold trade network. Note that this network features an evanescent market situation at Apayao.................................................................................................................................................. 30 Figure 3.8. Composite map. Note Lepanto gold trade network. Note that this network also features a permanent market situation at Danac different from the Apayao......................................................................................................... 31 Figure 3.9. Danac site near Amburayan River using WV2 NIR variation band combination............................................ 32 Figure 3.10. Danac coffin tradition..................................................................................................................................... 33 Figure 3.11. Apayao site (using WV2 NIR variation band combination)........................................................................... 34 Figure 3.12. Apayao coffin tradition (Ilocos Sur archeology Project II, 2012).................................................................. 34 Figure 3.13. Angaqui trade network in Northwestern Luzon............................................................................................. 35 Figure 3.14. From Ilocos Sur archeology Project season one (ISAP-1) files..................................................................... 36 Figure 3.15. Larger bamboo raft in the lower Abra River, turn of the twentieth century................................................... 39 Figure 3.16. Guard structure along the trail somewhere in the Angaqui network with Bontok warriors, turn of the twentieth century........................................................................................................................................................... 40 Figure 3.17. Trellis along the trail somewhere in the Angaqui network with Bulol posts, turn of the twentieth century................................................................................................................................................................. 40 Figure 3.18. Foreground the Town of Angaqui, background Tirad Pass/Tila Peak Locality.............................................. 41 Figure 3.19. Unknown Cartographer(1794). Ydea Aproximada del Teritorio entre Cagayan e Ylocos............................. 42 Figure 3.20. Tauguaz, Rafael (Cartographer). (1891). Croquis del Distrito Politico Militar de Tiagan y Rancherias de Ilocos Sur..................................................................................................................................................... 43 Figure 3.21. Unknown Cartographer. (undated) Croquis de la comprension de Balidbid y parte de la de Sta. Lucia....... 44 Figure 3.22. Pennarubia, Don Esteban (commissioned map). (1868). Croquis de la provincia de Abra........................... 45 Figure 3.23. Artist illustration of the Patiacan secondary burial coffin.............................................................................. 46 ix

Remote Sensing the Margins of the Gold Trade Figure 3.24. Archeological features in Minlaoi open site (Using WV2 soils and constructions combination).................. 47 Figure 3.25. Watch tower foundation stones on the crest of Minlaoi Hill (Ilocos Sur archeology Project II, 2012)......... 48 Figure 3.26. Watch tower somehere in the Angaqui network, turn of the twentieth century............................................. 49 Figure 3.28. Note Abra trade network in Northwestern Luzon.......................................................................................... 51 Figure 3.29. Composite photo of Abra Network with Abra River in foreground............................................................... 52 Figure 3.30. Composite archival photo of raft plying the river.......................................................................................... 53 Figure 3.31. Composite archival photos of smithies somewhere in the Abra network...................................................... 55 Figure 3.32. Pandan in Vigan-Caoayan arrow shows where a biray vessel is buried under brackish sediments............... 56 Figure 3.33. Arrow showing Calanutian site, which was excavated during the Ilocos Sur Archeological Project of 2011 and 2012................................................................................................................................................................. 58 Figure 3.34. Calanutian Site Map....................................................................................................................................... 58 Figure 3.35. Various artifacts excavated in Calanutian....................................................................................................... 59 Figure 3.36. Arrow showing sites in Nagsingcaoan, Cabugao........................................................................................... 60 Figure 3.37. Easterly view from the Nagsingcaoan gateway.............................................................................................. 60 Figure 3.38. Encircled in yellow is an ijang in Refaro, San Juan where tradeware was found on the surface.................. 61 Figure 3.39. Clearing in Refaro with tradeware surface finds............................................................................................ 61 Figure 3.40. Mapa Geografico del Centro de la Abra en la Provincia de Ylocos............................................................... 62 Figure 3.41. Plano Topografico que comprehende una parte de la Provincia de Ylocos Sur, en la que se hallan situados los Distritos Militares de Tinguianes e Ygorrote.................................................................................................. 63 Figure 5.1. Map (1837) showing 1829 route of Guillermo Galvey.................................................................................... 82 Figure 5.6. Trail spectral signature enhanced with WV2 modified infrared combination 732........................................... 84 Figure 5.7. Note example of a ridge trail............................................................................................................................ 85 Figure 5.10. Data fusion of WorldView-2 indices (workflow) using ERDAS Imagine 2016............................................ 85 Figure 5.11. Least cost path algorithm (workflow) using ArcMap 10.4............................................................................. 86 Figure 5.26. Trail connecting Cabugao coastal settlement with Malague in Cagayan valley............................................ 87 Figure 5.27. Guillermo Galvey route passing from Bangued to Tayum to Tineg then to Malague.................................... 87 Figure 5.37. Baluarte/Tamag Hills Southwest of Vigan encircled in yellow...................................................................... 88 Figure 5.38. Bantay Hills Northeast of Vigan encircled in yellow..................................................................................... 89 Figure 5.39. Arrow shows ijang facing the interior which is the site of the Church of Our Lady of Assumption built 1765............................................................................................................................................................................ 90 Figure 5.43. SNA similarities model using UCINET software. Matrix of least cost path distances in between the two models.................................................................................................................................................................... 91 Figure 5.44. SNA dissimilarities model using UCINET software. Matrix of least cost path distances in between the two models.................................................................................................................................................................... 92 Figure 6.4. Bauang as seen from the Banangan ridge in Sablan on a clear day................................................................. 94 Figure 6.5. Close-up of Bauang as seen from the Banangan ridge in Sablan on a clear day............................................. 95 Figure 6.18. Two candidate ijangs north and south of San Fernando cove........................................................................ 96 Figure 6.20. Data-fused and image-enhanced indices showing the possible lantaka emplacement................................... 97 Figure 6.22. Area around Galan mines north of the Tongayan ridge.................................................................................. 98 Figure 6.27. Note that Quinali settlement is on the banks of Balasian River (Rio Quinali). Pennarubia, Don Esteban (Commissioned Map). (1868)............................................................................................................................................. 99 Figure 6.32. Note the spectral signature of the vegetation associated with this trail segment......................................... 100 x

List of Figures Figure 6.33. Note the spectral signature of the vegetation associated with this trail segment......................................... 101 Figure 6.34. Note trough and 15 degree dog-leg turn in this trail segment...................................................................... 102 Figure 6.35. Guard structure along the trail approaching Bontok settlement in the Angaqui network with Bontok warriors, turn of the twentieth century.............................................................................................................................. 103 Figure 6.38. Arrows show the expansion upriver............................................................................................................. 104

xi

List of Abbreviations ASP

Archeological Studies Program (University of the Philippines)

ASTER GDEM

Advanced Spaceborne Thermal Emission and Reflection, Global Digital Elevation Model

BAS

Benguet Archeological Surveys

CAR

Cordillera Administrative Region (Philippines)

CPM

Commandancia Politico Militar (Spanish period military province)

DEM

Digital Elevation Model

DENR

Department of Environment and Natural Resources (Philippines)

EHP

in Philippines, Early Historical Period (tenth to sixteenth century)

ERDAS

Earth Resources Data Analysis System (Hexagon Geospatial, USA)

ESRI

Environmental Systems Research Institute (USA)

FM

Field Museum (USA)

GDEx

Global Data Explorer

GIS

Geographic Information Systems

HP

in Philippines, Historical Period (seventeenth to twentieth century)

IO–WPS/SCS

Indian Ocean–West Philippine Sea/South China Sea

ISAP

Ilocos Sur Archeology Project

LP DAAC

Land Processes Distributed Active Archive Center

METI

Ministry of Economy Trade and Industry (Japan)

MGB

Mines and Geosciences Bureau (Philippines)

NAMRIA

National Mapping Resource and Information Authority (Philippines)

NAP

National Archives of the Philippines

NASA

National Aeronautics and Space Administration

NM

National Museum (Philippines)

NWL

Northwestern Luzon

PGIS

Provincial Government of Ilocos Sur

SNA

Social Network Analysis

UIC

University of Illinois at Chicago

USGS

United States Geological Survey

UTM

Universal Transverse Mercator

xiii

1 Introduction: Interdisciplinary and Multiscalar Ethnohistorical Archeology 1.1 ‘Small Worlds’ towards Modeling Evanescent and Permanent Markets

precisely because of the need to facilitate the movement of goods upland to lowland and vice versa. What is further argued is that the nodes to these transport networks lend a hand in determining if upland miner groups participated in either evanescent markets or permanent markets. Evanescent markets feature more land-based trails that cut into forests or follow mountain ridge trails on the way to the coast. Permanent markets instead have an added component of riverine transport networks that facilitate bulk movement of goods.

Using a bottom-up approach, this book focuses on a known gold source or production periphery (or margin) relative to the Indian Ocean–West Philippine Sea/South China Sea (IO–WPS/SCS) gold trade emporium – Northwestern Luzon Island, Philippines (Fig. 1.1) and its gold deposits in the Grand Cordillera Mountains. The work begins with the indigenous groups who started the chain of interactions that eventually resulted in gold ending up in production centers within the IO–WPS/SCS. The work, being both multidisciplinary and multiscalar, operated at various scales of analyses, weaving together available archival, historical, ethnographic, and archeological data on individual sites within larger networks and regions (Fig. 1.2). Modeling through geographic information systems (GIS) was particularly helpful in establishing a baseline predictive model that was used as a guide in remote sensing. The book, however, has identified several instances wherein these predictive models do not match reality. Historical models utilizing archival maps, written historical records, or oral tradition records, however, often do correspond strongly with the remote sensing analysis which is presented below. The GIS methods included basemap-georeferencing of ethnohistorical and archeological sites, visibility and viewshed analysis, least cost path analysis, suitability modeling, as well as other spatial analyst tools.

The location of bulking villages within these transport networks also allows us to differentiate evanescent from permanent market encounters (Fig. 1.3, Fig. 1.4). The bulking centers with surveilling capacity of the coast that facilitate ship-spotting allow goldminers to plan to meet the merchant ships and deal with them directly without going through middlepersons. A non-surveilling situation, on the other hand, is a feature of a permanent market since the important factor would be proximity to a riverine system that can facilitate movement of bulked goods through rafts. A third variable that also leads to a strong determination of a market situation is the presence of a trade diaspora, with the argument that coastal peoples occupying bulking villages indicate the presence of middleperson groups who facilitate the approval of credit that allows miners to get items in advance and be able to pay in the future. A trade diaspora from the highlands, on the other hand, will ensure that miners will not deal with middlepersons and be able to plan to meet trade vessels on the coast upon spotting their movement on the horizon.

This book examines regional-scale interactions between upland extraction points, small-scale tribal gold miners and the lowland middlemen, and foreign maritime merchants in evanescent coastal markets to more permanent ones. It compares five regional coastal–upland trade systems in Northwest Luzon (Tonglo, Gasweling, Lepanto, Angaqui, and Abra regions) through historical, archeological, and ethnographic data, to examine how small, upland tribal groups mined gold and made complicated decisions on the best strategies for bulking and storing the gold in easily accessible areas along what were calculated as easily traversed routes to the coast in order to reach coastal markets and maritime traders faster than upland gold miner competitors. Based on trade and exchange commitment to the IO–WPS/SCS trade emporium, Northwestern Luzon settlements can functionally be classified into coastal settlements, bulking villages, or gold mines. It is argued in this book that these three settlement locales are articulated

It is the identification of either evanescent or permanent market situation that will ultimately provide a glimpse into the receding agentive position of the gold miners, receding resiliency especially with the advent of reduccion or Spanish assimilation, and finally an increase or prevalence of debt finance or ‘utang na loob.’ The book is a palimpsest of the tenth to sixteenth century Early Historical Period and the seventeenth to early twentieth century Historical Period, with chronology facilitated by the multifaceted data (historical, archeological, ethnographic) employed in the book (Fig. 1.5 and Fig. 1.6). Most historical and archeological work on the IO–WPS/ SCS trade has tended to focus on the ‘big picture’ of largescale trade powers from early to mid-second millennium AD (the East Asian empires, South Asian empires, mainland 1

Remote Sensing the Margins of the Gold Trade

Figure 1.1. Five gold trade networks in Northwestern Luzon (ASTER GDEM is a product of METI and NASA). Note that data frame grid throughout the book is UTM WGS 1984.

2

Introduction

Figure 1.2. Diagram showing multidisciplinary datasets. Note that regional, multiscalar analysis is facilitated by GIS and Remote Sensing.

Figure 1.3. Interpretive Framework used in the book.

commodity exchange by powerful states and empires within the South China Sea–Indian Ocean trade spheres (reinforced through the lens of ‘histories’ disseminated by literate foreign traders from distant state centers), instead examining the ‘small worlds’ of chains of producers and consumers in this complexly linked oceanic trade world. This approach is viable especially if these producers and consumers have available written and oral histories. In Northwestern Luzon, the various indigenous groups have recorded oral traditions as well as primary and secondary historical documents that contain information on gold extraction and working. Some of these data are in depth,

SE Asian states, and European colonizers) (see Hall et al. 2010, Stein 1999), with little emphasis on the multitude of small-scale producers and traders of the Southeast maritime sphere, where a wide range of economic and social transactions occur that keep these trade networks viable. Some of the known gold sources in Southeast Asia for instance include islands in the Philippine and Indonesian archipelago, the Malay peninsula, and parts of Burma. It is only recently that some archeologists and historians have moved away from emphasizing centric control of 3

Remote Sensing the Margins of the Gold Trade

Figure 1.4. Map of change through time highlighting Angaqui, Abra networks in particular to illustrate shift from evanescent to permanent market (EHP tenth to sixteenth century). The top map shows thirteenth century prior to the rapid expansion of IO–WPS/SCS trade with Luzon in the fourteenth to fifteenth century, which is depicted in the lower map. Note that due to increases in trade interaction the bulking stations in Abra valley are established on the route leading upriver to the Quinali and Lepanto mines. Also worth noting is the move of Angaqui from the mountain peak with surveilling capacity for Candon to the river banks of the Abra with non-surveilling capacity for neither Candon nor Vigan, arguably to facilitate bulked goods raft transport.

providing a window into social political and economic aspects of this gold-producing lifeway.

to the dialogue on multi-ethnic trade by focusing on mined gold as a critical upland product associated with social constructions of wealth and power in the client societies, but there is a significant potential to expand the work to other commodities, for example involving raw material extraction (e.g., archeologically retrievable tropical forest plant and animal products, including ivory, spices, hardwoods, etc.), which will provide a fuller picture of forms of engagement and power relations at multiple scales.

Meanwhile, important new works have begun to emerge in the archeology of commodity production at various scales and diverse strategies of trade in both Mainland and Island Southeast Asia (see Junker 2018, Junker and Smith 2017, Hendrickson et al. 2017). In Junker (2018) and Junker and Smith (2017), the foragers in Island Southeast Asia and their exchange relationship with sedentary agriculturalists is viewed as a complex and agentive strategy of resilience. Research by Hendrickson et al. (2017) also emphasizes agency of the small-scale forest savvy Kuay groups of Cambodia who supported the Khmer Empire. While this book contributes to the dialogue on multi-ethnic trade by focusing on gold as a commodity, there is a nascent interest in once-peripheralized and unstudied small-scale societies who were key players in the Southeast Asia– East Asian–South Asian maritime and land-based trade sphere, as will be discussed in subsequent chapters. It should also be noted that this book attempts to contribute

The book is a unique contribution to the study of gold production in Southeast Asia because it intensively studies the connections between chains of producers and consumers that eventually reached out of the Philippine archipelago and fuel; by the sixteenth century, a truly world system focused heavily on the gold economy. As noted above, this book investigates questions of agency vs. marginality of small-scale gold miners in Northwestern Luzon between the tenth century and the twentieth century. That is, the book privileges previously disenfranchised groups in this 4

Introduction

Figure 1.5. Relevant oral tradition, written historical or archeological data sources used in delineating Northwestern Luzon settlements on the book basemap (Early Historical Period tenth to sixteenth century and Historical Period seventeenth to early twentieth century).

5

Remote Sensing the Margins of the Gold Trade

Figure 1.6. Map showing data on gold and tradeware quantities or gold and tradeware culture found on relevant settlements of Northwestern Luzon. Most of these sources pertain to the Early Historical Period tenth to sixteenth century.

6

Introduction these varieties of identities transcended time, going back to precontact periods, but quite akin to analyzing a genealogical tree, one is able to reconstruct the relationships and connections of these more contemporary groups. This aspect of the work was facilitated through the analysis of 8,875 square kilometers of high-resolution and multispectral WorldView-2 (pansharpened at 0.46 meter resolution) and WorldView-3 (pansharpened at 0.31 meter resolution) satellite imagery of these five networks in Northwestern Luzon, provided by the DigitalGlobe Foundation. GIS predictive modeling was also made possible by Advanced Spaceborne Thermal Emission and Reflection Radar Global Digital Elevation Model (ASTER GDEM) from the Ministry of Economy Trade and Industry of Japan and the National Aeronautics and Space Administration of the USA. Open access shapefiles for educational use were also sourced from PhilGIS.org. Satellite imagery enhancement of the raw satellite imagery was also carried out using ERDAS Imagine version 2016 software (Hexagon Geospatial, Madison, AL, USA)and Geographic Information Systems (GIS) analysis through ArcMAP versions 10.4, 10.5 (ESRI, Redlands, CA, USA) software also made possible the predictive modeling of trails and settlement locations within the five networks of Northwestern Luzon. But it should be emphasized that for crucial interlocutors of this regional, geographic data as well as ethnographic, historical, and archeological data were available for specific sites within Northwestern Luzon.

analysis. The local and regional (multiscalar) analysis employed in the book utilized advanced remote sensing methods to analyze five gold trade networks in the Island. The book may be the first to use archeometrical methods, specifically GIS and satellite remote sensing, in West Philippine Sea/South China Sea regional gold studies, which is rich with historical and ethnographic data but in terms of theorizing gold production and distribution, has remained speculative in nature. 1.2 Interdisciplinary Methods: Historical, Archeological, Ethnological, and Geographical/ Spatial, Remote Sensing Studies This book recognizes the diversity of the Northwestern Luzon gold data from centuries of various multi-cultural perspectives (e.g., written by Spanish, Filipino, foreign Asian traders including Chinese records) in the form of primary and secondary documents and oral accounts with information on production and trade systems, as well as recent archeological works on both single site and regional levels, ethnological studies (involving relatively recent oral histories and material culture through museum collections and archival documentation), and reconstructions of interactive economic land use and commodity transport systems through Geographic Information Systems data, in the analyses of these production and trade systems. This plethora of sources then necessarily requires multidisciplinary approaches that seek to avoid ‘siloing’ the knowledge from the various disciplines (History, archeology, Geography [GIS], and Cultural Anthropology). It is posited in this book that to truly understand larger systems of human interaction at various scales, one has to use a multidisciplinary approach.

1.4 Overview of Book The following chapter two will provide an overview of the historical, ethnographic, and archeological data available regarding gold trade in Northwestern Luzon, and how these various media can illuminate and expand our understanding of a complex and critical local system of production and trade that fueled the larger IO–WPS/SCS ‘world trade system’ of the late first millennium to the late second millennium AD. This is crucial in setting the stage for the theory as well as the method in this book. This chapter’s multidisciplinary approach to material evidence and multi-vocal sources of written and oral histories over the span of the second millennium CE paints a complex picture of both the recorded memory and materiality of gold mining, its social and political configurations, its economic ramifications on local and regional scales, and the strategies used by mining and merchant communities over time and space in Northwestern Luzon.

1.3 A Multiscalar Approach to Examining Complex Regional Production and Trade Networks It is vital to examine social and economic relations in NW Luzon within a regional level of analysis rather than single sites, since interaction spheres in prehistory and varying historical periods involve a plethora of actors and groups who dynamically shape the gold production and trade. While much of the archeology in the Philippines and in Southeast Asia in general remains focused on single sites rather than the larger regions in which people interact as social units with distinct ethnic/linguistically bounded identities, larger economic spheres are formed for material benefit, and distinct ethnic/culture ‘edges’ often form on the landscape where conflict or friction occurs around social norms and exchange.

Chapter three provides a background on the methods directly used in the analysis including the various GIS tools as well as the methods of enhancing high-resolution multispectral imagery. In terms of GIS methods, this book deploys least cost path analysis, visibility and viewshed analyses, weighted overlay technique, and historical map georeferencing, among others. In terms of satellite imagery both WorldView-2 with 8 multispectral bands featuring 1.85 meter resolution and one panchromatic band at 0.46 meter resolution and WorldView-3 with

This book focuses on the gold working indigenous groups within the Northwestern Luzon region. Accordingly, the gold extracting and gold working groups are lumped together within the cultural identities of Ilocano, Tagalog, Igorot, Pangasinan, etc. There are, however, more distinct subgroupings, which include Ibaloi, Tingguian, Itneg, Kankanaey, Bago, Bontok, Ifugao, etc. It is important to make the point that this book is not arguing that 7

Remote Sensing the Margins of the Gold Trade eight multispectral bands featuring 1.24 meter resolution and one panchromatic band at 0.31 meter resolution were utilized. After image enhancement or pas sharpening WorldView-2 resolution is 0.46 meter while WorldView-3 resolution is 0.31 meter. Finally, a section of this chapter features Social Network Analysis of the Northwest branch of the Abra riverine trade network.

Northwestern Luzon will be the first step to a continuing effort to broaden the reach of this research, noting the other areas with gold sources in South and East Asia like Sumatra and Borneo and gold centers like Southern China, South India, Vietnam, Thailand, and Singapore, among others. As part of what is hoped to be a postdoctoral project, this research will blossom into collaborative and multidisciplinary projects among gold specialists in South and East Asia. The researcher will align his postdoctoral studies pursuing these international partnerships true to the multiscalar characteristic of the book. The research is also multidisciplinary in as much as it uses a conjunction of methods from oral tradition studies, written ethnohistory, remote sensing, and GIS.

Chapter four provides a review of various models – such as World System Theory, modified forms of World System Analysis, a new emphasis on the ‘periphery’ or intersection of multiple interacting political and economic centers, multiscalar approaches to socio-economic interactions, and formalized ‘social network analysis’ – as theoretical representations of relationships at a regional level that have some utility in both predicting and analyzing on-the-ground spatial patterning of regionalscale archeological sites associated with gold production systems in the interior and their articulation with lowland markets on the coast feeding into larger oceanic maritime trade within and beyond the Philippine archipelago. The chapter also attempts to integrate these larger scale prototypes of interactions with structural theories of agency and reciprocity at a smaller scale, focused heavily on historically and ethnographically documented elements of social relationships that appear to persist over time in the region.

Studies have shown that gold was still being exploited by small-scale indigenous miners of the Cordillera Mountains of Luzon who resisted reduccion (assimilation) into the cabeceras (colonial villages) well into the Spanish period. This is directly in contrast to other areas around the world where the mining of precious minerals was centrally organized as a state affair for the colonial government (i.e., Peru, Mexico). For several centuries, this gold was traded in evanescent to more permanent markets in the coastal settlements and frequent clients included Chinese and Japanese merchants. Evanescent markets allowed the small-scale miners to bypass middlepersons and deal directly with foreign merchants. On that basis, these innovative miners may have been a real driving force with agency in a way that we are not used to thinking about with regards to mining labor.

Chapter five provides the analysis on the results of the methods and its direct implications on the data of chapter two, historical, ethnographic, and archeological data, and chapter three. Here we begin to see the differentiated articulation of the local gold trade system among the five trade networks. Chapter six provides a discussion of the implications of the findings based on the analysis. Finally, chapter seven will provide the conclusions, which will wrap up the important findings of the book, arguing that various commitments to the overseas gold for porcelain trade reflect the variable agentive decision making of the Igorot gold miners. 1.5 Significance of this Book A major contribution of this research is the focus on the use of anthropological theory to understanding local smallscale gold miners who were articulated to a broader global world system. The book focuses on the small worlds of disenfranchised mining groups who used to pursue a latitude of agency before some of these groups became fully interfaced with the IO–WPS/SCS trade system. The research in Northwestern Luzon will stand as a model on how to redress archeological inquiry starting from the margins or peripheries before moving to the centers. The impact may also reverberate across the South and East Asian region as scientists interested in the historical and archeological research of gold will see the immense utility of combining high-resolution remote sensing techniques as well as GIS spatial analysis in answering fundamental questions about the gold trade. Further, the analysis in 8

2 Background: Northwestern Luzon Production and Trade Systems – An Overview through Ethnohistory and Archeology By the tenth century and most likely earlier, the northwestern coast of Luzon Island in the Philippines was known by Chinese explorers and traders, merchants from other nearby island archipelagos, and distant polity centers in both Mainland and Island Southeast Asia. Northwestern Luzon attracted gold traders from within and outside the archipelago, as attested in Chinese trade records and voyager accounts, as well as formalized Chinese import tallies from ‘tributary’ trade mission reports in which gold appears to have been a significant draw for foreign traders even before European contact and conquest. The emphasis in this chapter is a discussion of the multi-vocal historical sources, ethnographic sources, and archeological data relevant to reconstructing the locations, actors/groups (e.g., interior mining specialists, traders, lowland middlemen, foreign merchants), and production and transactional activities that constituted the gold mining and exchange systems in this particular region of Northwestern Luzon. However, the extended discussion of these various sources for reconstructing aspects of a millennium of specialized gold mining and trade in Northwestern Luzon will be briefly prefaced by an overview of what we know generally about the ‘social world’ of the larger Philippine archipelago in the time period of interest.

Historical Period in the Philippines. One can almost argue that the use of EMP was with historians whereas EHP was with archeologists. There are comprehensive works that deal with the EMP in Southeast Asia (Andaya, B. W., & Andaya, L. Y. (2015). Reid, A. (2000). Andaya, L. Y., & Andaya, B. W. (1995). Reid, A. (Ed.). (1993). Andaya and Andaya (2015), for instance, argue that the general acceptance of Early Modern Period in Europe and America also influenced other regions with newer periodizations coming up like early modern China and early modern Japan (6). The former coincides with the Ming and early Ching dynasties of China, which are parallel to the Early Historical Period in the Philippines. But how do we deal with the developmental evolutionary dichotomization of traditional versus modern that seems to be in the bandwagon of using the word modern? In another work Andaya and Andaya (1997) seek a break from the traditional notion that modernity implies Western culture but that in fact historical data in Southeast Asia shows that there is a purported modern spirit in earlier times, beyond recent history (391). Their work allows us to break away from an almost unchallengeable and historically determined notion that modern is intrinsically connected to Euro-American (ibid).

2.1 Gold and Other Mining and Production Networks Elsewhere in the World

At this juncture let us look at cross comparative materials that deal with upland–lowland interactions in Southeast Asia. Scott (2009) presents an excellent work showing the difficulty of assimilating upland locales into state societies. While he focuses on the mountain backbone of mainland southeast Asia known as the Zomia (13), some parallelisms can be drawn to that of Luzon Island and its very own Cordillera mountain ranges. He makes a spot-on assessment when he argues that ‘specialization by altitude and niche within the hills leads to scattering… yet longdistance travel, marriage alliances similar subsistence patterns, and cultural continuity help foster coherent identities across considerable distances’ (2009: 18). This is seen in the case of the Ibaloi and the Kankanaey who will be more broadly termed as specialized gold miners yet quite cosmopolitan as well, because of their interactions with the coastal settlements, by consanguinity or affinity (i.e., trade diasporas). Other references that talk about mainland Southeast Asia upland groups and their autonomy include Kersten (2003).

There have been several studies of gold and other mining and production networks elsewhere in the world (Jyotishi, Lahiri-Dutt, Sivramkrishna 2018; Grätz 2004; Graulau 2001; Miller, Desai, Lee-Thorp 2000). Networks more generally speaking have also been studied specifically in the Americas (Hofman, Mol, Hoogland, Rojas 2014; Erickson 2011). Jyotishi, Lahiri-Dutt, and Sivramkrishna, for instance, apply a diachronic perspective that looks at both formal and informal categories of mining both in historical and contemporary contexts in the Nilgiri-Wayanad region in Southern India (2018: 235). Their work also features primary historical sources covering a span of over two centuries (2018: 242-252). Gratz (2004), on the other hand, focuses on the gold trading network of Northern Benin West Africa. Gratz’ synchronic study tackles the logistics of proliferation of an informal gold trade network (146). Gralau (2001), on the other hand, looks at the political economy and gender dynamics of peasant mining production in Amazon rainforest of Brazil. This synchronic study looks at contemporary peasant small-scale mining tapping on a critical contextual bottom-up approach (71-72).

Demandt (2015) demonstrated that aside from the gold ore, gold ornaments in mainland Southeast Asia were also important prestige goods connected to economic and political power in early trading communities of the region (305). Gold transferred hands through a putative interregional coastal–inland prestige goods network (323). His work is important because its shows that while

There are several works that feature the Early Modern Period (EMP), which appears to coincide with the Early 9

Remote Sensing the Margins of the Gold Trade ‘forager’ groups (Junker 2002b). These forager groups are almost certainly connected to initial Pleistocene inhabitants of the archipelago based on archeological evidence and paleoanthropological studies, particularly their association with clearly dated ‘Upper Paleolithic’ Pleistocene sites with human remains and artifacts in many areas of the Philippine archipelago (Junker 2002) and recent DNA data that places them with other Late Pleistocene Asian foragers on mainland Southeast Asia, areas of island Southeast Asia outside the Philippines, East Asia, and Australia (Barker and Richards 2012). Some of the peoples referred to as ‘tribal’ societies who lived (and still live) in the interior occupied short-term shifting settlements or permanent settlements, based on economies combining swidden agriculture, while others practiced intensive, permanent, and sometimes irrigated pond agriculture, likely entered the archipelago around 6,000-4,000 years ago) in a population expansion of maritime-focused Austronesian-speaking peoples from mainland Southeast Asia (Soares et al. 2015, Spriggs 2011) into the archipelago and beyond (as probable ancestors of the earliest Micronesian and Polynesian societies) (Soares et al. 2015, Spriggs 2011). These latter groups, significantly biologically and linguistically distinct from the earliest aboriginal populations, filled both maritime/ coastal-focused and inland-focused economic niches.

gold was mined mainly for the overseas market, even the Southeast Asian locals placed value on this yellow metal as part of their adornment (dressing up, p. 306). An important point that was raised by Demandt is that more gold objects reached Southeast Asia during the Early Historical Period, coming over the established trade routes. Arguably, some significant amount of gold would have come from the Philippine Islands, notably Northwestern Luzon. In terms of cross comparative mainland southeast Asian trade networks, Bellina (2003) demonstrated through statistical analysis of characterization data on trade beads that there were two periods of interaction in Southeast Asia with the earlier, Period One, manifesting high quality almost made-to-order beads manufactured using Indian technologies, which correspond to a well-established exchange relationship dating back to the first half of the first millennium (2003: 291). The next period corresponds to the first millennium and sees the opening of manufacturing centers in Southeast Asia, this time producing beads of medium to mediocre quality (2003: 293). While the first period appears to show a strict patronage and restricted access to high value prestige goods, the second period appears to coincide with a purported mass production of this formerly restricted access now open access prestige good. Arguably, this is a process captured in the microcosmos of Northwestern Luzon – the manifestation of the establishment of permanent market centers in Northwestern Luzon.

By the Sung dynasty in China (AD 950-1279) or earlier, long distance trade of prestige goods with China and numerous established Southeast Asian maritime polities was already established in the archipelago (Junker 1999, 1998). Chinese textual references to the Philippines from the tenth century onwards are significant in describing what were almost certainly large-scale, socially stratified, and politically centralized societies, referred to as ‘chiefdoms’ by Junker (1999, 1998), in coastal areas and along interior rivers. Spanish documents (primarily from the European contact period in the sixteenth century and Spanish colonial period) must be parsed or deconstructed from Spanish cultural constructions and biases, just as Chinese reactions to Philippine political and social organization were developed through a lens of culturally-constructed views of their place in the then-known world (Junker 1998). Other scholars have suggested that Junker’s model, focused on cross-cultural comparisons of ‘middle range’ (neither relatively egalitarian or state) societies – with some evidence of permanent political centralization and hierarchies, significant forms of institutionalized social stratification, strong differentials in both material wealth and control of resources – is too focused on a processual model that does not emphasize the historically-contingent, unique aspects of structure in these coastal trading polities (Barretto-Tesoro 2008). Whatever specific model of socio-political structure for the tenth century and later periods is espoused by historians and archeologists, it is clear that a multidisciplinary and multiscalar regional approach is necessary for analyzing these networks of social and economic interaction.

South Asia historian Himanshu Prahba Ray (1989) has long argued that South Asia India has had interaction with kaleidoscopic varieties of Southeast Asian societies from as early as the beginning of the first millennium (42). While he has focused on inscriptions and literature. It is important to state that several of the mining communities in the Philippines have attributed mining techniques if not to the Chinese, then to the Indians. In terms of cross comparative data, it is important to note the work of Chieh-Fu Cheng, which looks at the archeology of trails in the Batongguan Region of Taiwan (2019). It is important to note that the region is quite similar to the Cordilleras of Northwestern Luzon in terms of its high altitude and steep terrain, among others. Cheng used field study and comprehensive archival research as his methods. He concludes that the configuration of trails reflects practices of colonization (2019: ix). 2.2 Philippine Islands Societies in the Tenth–Sixteenth Centuries and the Seventeenth–Twentieth Centuries Colonial World: A Brief History of Origins and Coastal–Upland Interactions Historical sources consistently report that by the time of European contact and likely earlier, politically complex and socially stratified societies inhabited coastlines and lowland river valleys in the Philippine archipelago, while the mountainous interior of most of the island was occupied by smaller-scale societies, including small-scale

Lowland–upland trade between coastal and interior societies was likely of long duration well before the tenth 10

Background Hirth and Rockhill 1911; Vega 1609, Carillo 1756 in Scott 1988: 217-242). The island itself has a high profile in early Chinese records, being called Liu-hsin, Luzong, Lu-sung, and Sung-tsai (Scott 1984: 71; Laufer 1908: 250; Pires 1513-1787 [1971]). By the Ming dynasty, China has recognized the wealth of Luzon because of the gold it possesses (Wu Hsio Pien 1575 in Laufer 1908: 257). Luzon also appears in various names in Spanish documents including Lucoes (Pires 1513-1787 [1971]) and Lusson (Rada 1574: 224). Portuguese merchant Tome Pires (1513-1787 [1971]) states that Borneans go to Luzon for countless gold. Vega narrates that 22 carats of gold were being mined by the Igorot and brought down to emporiums in Pangasinan and he claims this has taken place for many centuries (1609: 22). These emporiums are also called evanescent market encounters (Allegre 1998: 148), which are analogous to traveling bazaars (Braudel 1980: 107) in the Mediterranean. These markets emerge right at the shore near the incoming merchant vessel, and the shops close upon the departure of the same vessel. The inhabitants of Luzon apparently were enculturated into the intricacies of gold trade early in life as evidenced by children learning the system of weights related to gold (Anonymous 1586: 381). They could even identify the source of the gold based on its hue and color (Rada 1574: 224). According to Wilson, the knowledge of gold extraction, such as timbering the shafts, use of specialized tools, and other practices were no doubt stimulated and increased by the Chinese presence as traders in these coastal market communities (1932: 1).

century written records of lowland–upland exchanges associated with coastal polities, including finds of (preeighth century) exotic glass and semiprecious stone beads, decorated earthenware, and bronze and gold dated roughly between the first millennium BC and first millennium AD in ‘Metal Age’ sites in the Philippines (Junker and Smith 2017, Junker 1999, 1998). This attests to some form of trade reaching the islands from mainland Southeast Asia, western island Southeast Asia, and likely indirectly China and India during this essential prehistoric period prior to written records (Junker and Smith 2017). Unfortunately, pre-tenth century ‘Metal Age’ sites have not generally been studied systematically in terms of a regional approach, but instead as individual sites in various regions of the archipelago, so it is very difficult to project regional settlement patterns and possible contexts of exchange relationships in the Metal Age and earlier periods. In general, while the trade linkages between interiors and coasts have been widely accepted, few historians, anthropologists, and archeologists have gone beyond the analysis of coastal centers and their oceanic trade strategies to examine the small-scale communities in the interior that were linked by production, consumption, and trade to the more visible polity centers and ports in the historical records. In addition, although there may be similarities between islands and regions in terms of social configurations, it should still be noted that such social landscapes have diverse, economically specialized, and culturally and linguistically distinct populations that alternatively trade, cooperate, compete, and war with one another.

Among the relevant secondary sources are works by historians and Cordillera scholars (i.e., Scott 1988, 1984, 1974; Tenrengren 1963: 555; Keesing 1962: 48-49, 304). Tenrengren suggests that gold trade with China, Borneo, and Japan has taken place for many centuries (1963: 558). The Chinese and Japanese merchants came to the coastal maritime trading centers of Luzon like Agoo (formerly port de jopon) to get gold (Keesing 1962: 13; Wernstedt and Spencer 1967: 250). Keesing adds that the imports included cloth, ceramic ware, semiprecious beads, and worked and unworked metals, while the exports included gold, deerskins, carabao horn, beeswax, and fibers (1962: 13). In another source we find out that the gold was transported to the smithies of Canton in southern China (Chan 1978: 52). Jesuit priest Alonso Sanchez, who ministered in the Philippines, observed a great amount of gold being cast in the form of small boats, weighing half a pound each when he went to Canton in the sixteenth century, as noted above as the advent of Spanish colonization of the Philippines (Chan 1978: 52).

2.3 Northwestern Luzon Social, Political, and Economic Patterns of the Early Historic and Colonial Periods: An Overview of What We Know from Ethnohistory and Archeology This section reviews ethnohistorical and archeological data presently available for the area of Northwestern Luzon not only regarding upland/interior groups but also the lowland/coastal groups. Like many Philippine islands, these two groups of economically interacting peoples are situated in a volcanic archipelago island which abruptly rises from the West Philippine Sea/South China Sea and has a varied upland–lowland topography. In fact, when viewed orthogonally, the lowland peoples inhabit a restrictive north to south belt with an average width of 5 kilometers (west to east). What follows immediately inland is the grand Cordillera Central at a massive width of anywhere between 50 and 70 kilometers. The two worlds are interconnected not only in terms of geography but also in terms of ancestry (Azurin 1991, Himes 1998) and should be studied as a whole unit. Therefore, this chapter will review ethnohistorical and archeological data on the lowland coastal groups before moving on to the upland groups.

Based on the ethnohistoric record thus far discussed there are some indications that a core and periphery relationship was starting to take shape in the Early Historical Period (tenth to sixteenth century) Northwestern Luzon. What is particularly interesting in these early records is the emphasis on the essential economic relations of the lowland coastal leaders with the upland groups. Perhaps brought about by the relative proximity of the coastal

There are several secondary and primary sources that allude to the pre-Spanish antiquity and scale of production of gold in Northwestern Luzon (Chao Ju-kua 1225 in 11

Remote Sensing the Margins of the Gold Trade further exploration in the book in lieu of world systems theory and its criticisms discussed in chapter three. Such necessities included exotic food items (p. 54), iron (p. 52), Chinese silk, porcelain, timber, and foodstuffs. This will be investigated further in subsequent chapters examining the relationships through World Systems Analysis (WSA) and Negotiated Peripherality. Chan expounds that most Chinese merchants who sailed to the Philippines for trade were from the Chang-chou province in southern China or the trade city of Canton (1978: 52) because of the comparative closeness of these ports, but Chan’s view of the scope of the trade partners in Northwestern Luzon may be limited in comparison to other historians.

areas of Northwestern Luzon to the massive mountains of the Cordilleras (Wernstedt and Spencer 1967: 328-331), there was constant flow of information and goods between the upland and lowland areas. An anonymous sixteenth century account states that the lowlands ‘people enjoy it [gold] for they have more communication with the miners than anyone else’ (Unsigned 1586: 382). Beyer (1979: 119) argues that even the trade between the Chinese and the coastal peoples was guided by reciprocity and credit, with Chinese and other traders distributing their goods to native traders in credit, allowing reciprocity at a later date. These creditors remained in port until the native traders came back to the coast with the promised products, including gold. Chao Ju-kua, a Chinese chronicler in the thirteenth century, provides a detailed account of this exchange (Chao Ju-kua 1225, English translation in Hirth and Rockhill 1911: 162), noting that the period of waiting can last as short as a few days to as long as a few months. The early Spanish accounts describe similar exchange practices with Quirante, noting that the Igorots have significant debt/credit with the coastal villages as a result of this trade for gold (Quirante 1624: 279). It can be surmised that the items traded included prestige goods like porcelain as well as livestock like cattle and pigs. As a side note, this credit system may be an early manifestation of the world system encroaching on the uplands with the lowland coastal leaders benefitting from the arrangement. This is quite akin to the fur trade in North America at contact with Europeans (Kardulias 2002). This process was observed through archeological and ethnohistorical evidence in the Tanjay Region on Negros Island, where forager/ agriculturalist relations may have started as mutualistic, but may have become coercive once the hunter-gatherers became linked to the lowland political economy and by extension the larger Southeast Asian world system (Junker 2002a: 240). Chao Ju-kua (1225), in this early thirteenth century account, clearly describes a system similar to the later trade strategies described for Northwestern Luzon: that all goods from the evanescent market encounter are taken in credit and those in debt come back with payment after several days. Quirante (1624), on the other hand, indicates that the Igorots in the highlands were in so much debt to their lowland counterparts that there seems to be stagnation in their general life conditions and they could not reach a higher level of prosperity. As discussed previously, Barretto-Tesoro proposes the concept of ‘utang na loob’ (Barretto-Tesoro 2008: ii) as a possible mechanism that guides this lowland and upland relationship, which goes beyond economistic credit.

This research investigates further the likelihood that Chinese and Japanese trade diasporas (Stein 1999) may have been around in Northwestern Luzon and how these pockets of foreigners became an apparatus for the trade system. In particular, further analysis on the diaspora assertions are raised by other recent historical analyses, particularly Wernstedt and Spencer (1967: 119) and Mateo (2004: 122). H. Otley Beyer, who integrated archeological evidence and historical references in numerous midtwentieth century publications, dates increased trading and settlement by Chinese in the Philippines as occurring between the twelfth and thirteenth centuries (1979: 121, 1987: 123). He also notes that fifty or more Japanese were recorded as trader-residents, whether short term or long term, by Spanish conquerors of Manila in the midsixteenth century. In terms of later times during the Spanish period, some Chinese and Japanese became integrated into the colonial apparatus as forced labor or polos. This was the case in Mankayan in the northern Benguet Province because of the industrial mining undertaken at Sociedad Minero-Metalurgica Cantabro Filipino (SMMCF) in 1890 (Bagamaspad and Hamada-Pawid 1985: 177-178). Ethnohistorian Felix Keesing points out (1962: 109) that up to 600 Chinese were relocated and employed by the Spaniards at the SMMCF in Mankayan. There was a veritable division of labor for this industrial scale effort, wherein Igorot men carried the mineral ore from the mines to the foundry, the Chinese operated the foundry, and Igorot women helped with the washing of the metal (109). It should be pointed out that copper and gold mining often go in hand in hand, since some Cordillera deposits contain both metals – porphyry copper and gold deposits. The preceding paragraphs emphasize generally the strategic economic importance of Northwestern Luzon as attested by historical records from the early second millennium AD to the post-conquest Spanish Period. However, specific geographic places are also identifiable as Early Historical Period settlements associated with gold production, making it possible to identify gold producing and marketing sites through archeological analysis. Mateo’s (2004) work on the Ilocos region, when taken together with Newson’s work (2011) on demography in the sixteenth century, is an important guide to locating sixteenth century settlements that can be translated into GIS spatial data for the analysis. These two works in tandem

In terms of the relationship between the Chinese traders and the coastal Northwestern Luzon traders, Chan (1978) incongruously argues that the relationship was superficial and there was no influence exerted on the natives (p. 53). However, he contradicts himself later when he argues that early Filipinos became dependent on the commodities brought in from China at low prices for their necessities so frequently that even the Spaniards who came in the sixteenth century had come to rely on Chinese products (p. 54). This assertion of dependency is a target for 12

Background the other hand, was active from the fourteenth to fifteenth century based on the Late Yuan and Ming Chinese tradeware (ibid). Barrio Balingasay in Bolinao Pangasinan meanwhile is a particularly interesting Early Historical Period site that featured gold teeth pegging aside from twelfth to fifteenth century Chinese ceramics (Legaspi 1974: 7-9). The Magsingal church complex also sits atop a pre-Spanish contact settlement where hundreds of Chinese and mainland Southeast Asia ceramics have been excavated by private individuals. Similar to the Bolinao site, the Calanutian site (twelfth to sixteenth century) also featured gold tooth pegging among the burials, and in one approximately 12-year-old female burial there was even a flat gold sheet mouth cover. Aside from the gold, the site also features Indian carnelian beads, Indo-Malaysian beads, Thai ceramics and elephant ivories, and Chinese, Khmer, and a large amount of Annamese ceramics (Canilao 2015: 31). An interesting prospect would be to conduct sourcing on these gold materials to determine if they were traded with the upland mining settlements. Gold sourcing can become complicated, however, if rework was done on the material, which would have eliminated the trace elements necessary for analysis. Recent archeological surveys also revealed pre-Spanish contact surface finds in Caoayan, Cabugao, and San Juan, all in Ilocos Sur (Canilao 2012, 2011). Beyer mentions Santa in Ilocos Sur specifically as a site where graves with Sung ceramics were exposed during road construction activities (1947: 218). Beyer also mentions finding Late Tang to Early Sung period ceramics in middens around Lake Paoay in Ilocos Norte (1949: 215).

with early ethnographies, early travel accounts, early pronouncing dictionaries, turn of the twentieth century photo plates, and Spanish period maps (i.e., Urrutia 1759 in Scheerer 1975; Galvey 1842 in Scheerer 1905; Semper [1861] 1975; Meyer [1883] 1975; Barton 1930; Gironiere 1854) provide leads to the exact location of settlements and the trails that connected them. As further explained in the methodology sections of later chapters, Mateo (2004: 85) enumerates the villages based on the principle that ‘areas apportioned or allotted as encomienda were the [original pre-Spanish] population centers.’ Based on the sixteenth century encomienda list the relevant coastal settlements are as follows: Alinguey (Aringay), Baratao (Mateo identifies as Agoo or Bauang but Newson submits it is San Fernando), Tagudin, Dumaquaque (Santa Lucia), Candon, Narvacan, Vigan, Bantay-Bantaguey, Panay (Magsingal), Cabugao, Sinait, Barao (Badoc), Bacarra, Ballecilio (Bangui). Within the river valleys were Purao (Balaoan), Vintar, Danglas, Ylagua (Laoag), Bonsan (San Nicholas), and Cacabayan (Paoay-Batac) (2004: 85). Paz (2005, 2006) also successfully implemented this logic of using sixteenth century records to identify probable Late Pre-Hispanic/Early Spanish sites when his excavations in old church grounds in Mindoro Oriental hit on deposits related to pre-Spanish contact settlements. Demographic data assembled by Newson (2011) have also been a valuable source for the book because her work tracks the first contact of Spanish conquistador Juan de Salcedo with the inhabitants of coastal Northwestern Luzon, including Pangasinan and Ilocos. Her work presents information on settlements, populations, and more importantly amount of gold readily available from the settlements that were exacted as tribute in this early period of Spanish conquest (2011: 179-189), as well as demographic estimates on the upland settlements of Northern Luzon (pp. 218-248). Newson’s book also provides tables and maps that show where initial contact with Spaniards took place (180-182, 196, 214, 226-227, 246-248).

There are also some upland archeological discoveries that can be mentioned (Acabado 2009; Canilao 2008, 2015; Caballero 1996; Bodner 1986, Prill-Brett 2000; Maher 1974, 1980), which will be discussed in more detail below. What is apparent in these various sources is that the Ibaloi and other upland groups seem to have pursued a broad spectrum of socio-economic activities including smallscale gold mining and panning in the Early Historical Period. In fact, Wilson observed that the upland people engaged in a broad spectrum of economic strategies where on one hand they are indefatigable prospectors but at the same time they also hunt, fish, harvest root crops (taro, camote), harvest rice in the paddies, gather wood, and raise livestock (1932: 4). In between these activities they can investigate outcrops, slides, and cuts and assay gold-bearing veins (ibid). At this juncture we can relate this to the risk-reducing social strategies that people use to mitigate extraction of patchy resources (Rockman and Steele 2003: 17).

At this juncture we turn to some of the archeological works in the Northwestern Luzon lowlands that provide a material record of relevant sites with datable import ceramics (largely Chinese, but also Thai or Vietnamese) which have been well studied stylistically and chronologically by archeologists and art historians. What these data show is that porcelain trade was already well established in the coastal areas and these exotic goods are commonly found in burials and to a lesser degree in a habitation context. Some sites also have high-carat gold as orifice covers for interred individuals or as tooth peggings. The lowland coastal sites include Bolinao in Pangasinan, Calincamasan and Aludaid in San Juan (formerly Baratao), La Union, Magsingal in Ilocos Sur, and Calanutian in Sinait, Ilocos Sur. Tidalgo investigated the Marzon site in Calincamasan, La Union, and he concluded that the fluorescence of the site was between the twelfth century and the Spanish contact period based on the Late Sung to Late Ching period tradeware ceramics (1979: 11). The Aludaid site, on

Some archeological efforts have been conducted in the uplands of Northwestern Luzon. Caballero carried out an archeological survey on the ridge of Dalicno and her team found evidence of habitation and cave burial sites (1996: 180). They also found a stone grinder or gaidan and porcelain shards inside a cave (1996: 180), and they concluded that the site was used to ground, mill, and smelt gold (ibid). A systematic survey conducted in Chuyo and 13

Remote Sensing the Margins of the Gold Trade 18540 Photographer Dean C Worcester). They used the sheer force of cold water to crack the already-heated rock face (Bagamaspad 1985: 73). These workings can be archeologically identified because they leave large cuts or slides on the rock like the Pelican Slide in the Suyoc mine (Wilson 1932: 15-17). The eroding rocks are then processed through the traditional placer mining methods or gravitational gold separation techniques.

Tonglo sites also recovered some fourteenth century Ming period ceramic sherds, stoneware, earthenware, cattle molars, and marine bivalve shells (Canilao 2015, 2008). Prill-Brett (2015: 251-252) reports an adult jar burial that was akin to Early Metal Age burials in the lowlands dated to around first millennium AD. The burial also included traces of millet (ibid). Bodner’s excavations in Tukukan, Mountain Province show that settlements were already established by the seventh century with limited trading with distant lowland areas involving exotic or local manufactured goods (1984). In contrast, Acabado’s (2009) excavations in Ifugao show that the complex rice terracing technique began in 1585 (at Spanish contact), but earlier work by Maher (like Bodner) shows dates as early as the seventh century for rice production (1980: 235) with evidence of some limited trading with lowland groups as well. Therefore, controversy regarding upland group identities, agricultural systems, and upland–lowland specialization and trade continues to be debated.

Based on oral tradition sources that recount these traditional methods as far back as the pre-Spanish fifteenth century, the gold are bulked into cakes (Bagamaspad 1985: 73) and are then traded into the lowlands. As early as the sixteenth century, Mirandaola argues that there are three types of gold cakes based on varying carats; these are bizlin, malubay, linguinguin (Mirandaola 1574: 224). Keesing (1962: 18), who examines Spanish accounts, also argues that the purest form is from [sic perhaps traded in] Lingayen, Pangasinan in Northwestern Luzon and that the source of this gold is the tributaries of the Agno river in Benguet. The sixteenth century account of Mirandaola also states that gold is also brought down to Baratao, Turrey (Purao), Alinguey, and Danglas (1574: 223). Buzeta also identifies the destinations as Agoo, Lingayen, Aringay, San Fernando, Santo Tomas, and Puerto de Japon (1850-51).

At this juncture in the discussion, we will look at the mining technology of the upland peoples. The mining technology was simple and non-destructive to the environment. Their enculturation and socialization early in life makes them experts in gold craft including appraising and sometimes the adulteration of gold with other metals (Canilao 2015: 169, Sande 1576, Tenrengren 1963, Anonymous 1586). The working of placer mines involves the extended family with each member receiving their fair share and the head receiving a slightly larger amount of the gold for supervising the effort (Canilao 2015: 169, Lopez 1992, Perez 1902). They fashion out permanent and semi-permanent sluice boxes that strain the gold from the high water of the rainy season, then during the dry season they demolish the sluice boxes and recover the gold nuggets and particles that are trapped in natural riffles in the rock at the bottom of the stream (Wilson 1932: 15-16). As a side note, we also see how flexible their social-economic strategies are, as in the case when they adapt new foodstuffs or crops that are to their advantage. This is seen for instance in the shift from taro to rice. Keesing (1962) argues that the ethnohistorical record shows that rice was not introduced in Benguet until the nineteenth century (p. 65). He suggests that yams and camotes were more common prior to this period (p. 65). As will be discussed below, these upland groups also diversified into cattle herding and tobacco planting during Spanish colonial times.

The most comprehensive colonial ethnography of the Igorot gold miners (including the Ibaloi subgroup) was written by Alonzo Quirante (1674: 262-304). His account of his 1624 expedition gives us an idea of the socioeconomic system of the Igorots. The places he visited in Benguet are areas that are associated with both the Ibaloi (most notably Antamok) and the Kankanaey. He notes that the settlements are established on the peaks of mountains and he specifically states that this is so that they can have a commanding view of trails/paths (269). In the context of the evanescent market encounter, these vista points may have been positioned to also monitor ship arrivals on the coasts (Canilao 2017). During nighttime these peaks are also illuminated by campfires (273). Their settlements are unpretentious and simple and can be abandoned upon a moment’s notice (for example, during Spanish punitive expeditions) and simply built somewhere else (ibid). Curiously, Quirante states that the chiefs are not differentiated from the rest of the people (ibid) except in possessing more cattle crania in their houses. Note that in the translation of Quirante’s work, Blair and Robertson use the word ‘chief’ in describing local leaders. However, based on the ethnography of Quirante, the type of sociopolitical structure he may have been dealing with is more of a ‘tribe’ than a ‘chiefdom,’ in terms of evolutionary anthropological classification, but this will be a question addressed throughout the book. Quirante (1624) adds that there are many ‘chiefs’ in the purported nation so that one chief exists for every kin group of every ten or twelve houses (p. 273), but his perception of forms of leadership are undoubtedly colored by Spanish encounters elsewhere in the world during the sixteenth and seventeenth centuries. However, his account does recognize differences between the upland and lowlands. Burial practices are viewed as

In terms of lode mining technology, due to the smallscale, family-based to community-based (Caballero 1996, Wilson 1932) nature of the activity, their ancient toolkit includes digging tools, hardwood pick axes tipped with iron traded from the lowlands, wooden shovels, baskets to carry the ore, and bamboo or wood piston blowers with a flap hide for check valve (Wilson 1932: 9-11). To bore through a solid rock with gold veins they would line up hollowed-out pine tree logs strewn together with carabao hide bulkheads, then divert cold mountain stream water to areas near the mines (as shown in a photo titled ‘Turn of the 20th century waterworks’ © Field Museum CSA 14

Background markedly different from the practices seen in the lowlands (p. 275). The Igorots also live on yams and camotes as staple, and when it is time to move, they just pull these plants up by the roots (ibid) and bury them somewhere else. Interestingly, Quirante (1624) also speaks of a semimobile group of Igorots in their midst who move from village to village (p. 275). It is possible that these groups are the remontados or vagamundos who were avoiding Spanish colonization in the lowlands (to be explained in more detail below). The upland groups were also clannish and warring according to Quirante (1624: 275-276), but during times of peace, they participated in the evanescent market encounters in the coastal lowlands to exchange gold for certain animals or cattle (p. 279).

‘mountain thieves’ because they were viewed as outlaws who refused to submit to Spanish authority (p. 84). Keesing (1962) alludes to the practice of remontado in explaining demographic upheavals on the coastlines as being due to extensive movement of people into the interior and in the specific case of Sinait, Ilocos Sur they crossed the foothills into the Abra area (p. 125). Indeed, the geographic and topographic characteristics of the Cordillera served the purposes of the remontados. Mateo also introduces the class of people who moved from town to town as tribute collectors arrived, which she calles vagamundos (p. 90). In the case of the Ilocos populations Mateo admits there is indeed remontado based on written accounts, but it is difficult to estimate the exact numbers involved in these movements (2004: 195).

In Northwestern Luzon we see resistive people escaping to either wide river valleys nestled within Cordillera foothills (i.e., Abra, La Union, Ilocos Sur) or within the cliffs and elevated valleys of the Cordillera Mountains (i.e., Benguet, Mountain Province, upland Ilocos Sur). Several works have argued that the Cordilleras remained for the most part free from direct control during Spanish colonization and were branded as heathens or infieles by the Spaniards (Tolentino 1994, Keesing 1962, Scott 1988).

There seems to have been waves of socio-economic shifts that can be gleaned among the Ibaloi as they negotiated their historically changing landscape. This was a shift of subsistence emphasis from goldmining to cattle herding, and finally intensive wet rice agriculture (Canilao 2011: 20), but again all this is within a broad spectrum of subsistence strategies. Cattle herding played an important role in the shifting economies of the Igorots, notably in the sixteenth century (Canilao 2011, Tapang 1985, Carino 1985). Cattle introduction into Benguet was an offshoot of gold trade in the lowlands. We can also look at the oral tradition of the Ibaloi as a source of information. Bagamaspad and Hamada Pawid present a concise picture of transformations undertaken by the Ibaloi when they argue that ‘gold laid the foundation in earliest remembered times of a Benguet society [and] its trade in the lowlands brought in the cow as early as 1600s’ (p. 231). According to Bagamaspad and Hamada-Pawid (1985) the cattle consolidated and solidified the position of the elite baknang/kadangyan by the 1700s (p. 137). This in turn concentrated control of gold in the hands of even fewer baknang (1985: 138). The conduct of feasts called peshits was institutionalized by the baknang who had full control on the frequency and lavishness of the event (p. 139). The cattle and other livestock were increasingly being used in these prestige feasts (Canilao 2011: 67), and Quirante’s 1624 account already established the early presence of cattle in the highlands (p. 279). By 1829, Guillermo Galvey’s account shows that not only cattle but also carabaos and horses had become common in Benguet (Scheerer 1905: 176). According to Bagamaspad and Hamada-Pawid (1985: 231), there were 25,000 cattle, 2000 ponies, and 1000 horses by 1908. Tapang also argues that distinctive, clearcut social stratification emerged among the Ibaloi with the introduction of cattle herding (Tapang 1985). It is argued that the increased ability to carry out peshit feasts (butchering of cattle for distribution) further solidified the position of the baknang or elite. It is interesting to make a cross-cultural comparison in the case of the impact of Spanish colonization of San Augustin de Tucson in the Spanish-held North American ‘West’ of the seventeenth and eighteenth centuries (Pavao-Zuckerman and LaMotta

2.3.1 At the Time of Spanish Contact Indeed, even during the Spanish period there was illicit/ unregulated trade between upland communities and the lowland communities, which the Spaniards could hardly regulate (Mateo 2004: 216, also Jesus 1980: 178). The unregulated trade snaked through the same gold trails between the now Christianized towns and the upland infieles (ibid). This trade was noted between Igorots and Tingguians and Christianized settlements of Tagudin, Bangar, Vigan, Narvacan, and Candon (p. 217). Interestingly, Mateo notes that Candon had intimate trading ties with the Bontok, who maintained a centuriesold trade network with them (ibid) even through Spanish times. Mateo notes that the main trade item is still gold during the Spanish period. In fact, iron implements were sometimes melted down to cast scales specifically for the gold trade (p. 218). Also, during this period, minted silver was already being used as a form of currency (p. 219). Mateo states that in exchange for gold, Ilocanos sold cotton, pigs, carabaos, dogs, and horses. She highlights the fact that cattle were only a recent introduction and immediately became a popular trade item for prestige feasts or cañao (p. 219). Interestingly Mateo also adds that slaves were also a common item traded and they were used to labor in gold mines (ibid). She notes narratives of either lowlanders captured for labor or upland slaves captured in exchange for animals (p. 220). According to Mateo (2004), one form of resistance to Spanish contact was to become remontado or renegades, which she describes as those who escape to the mountains (p. 83). The literal Spanish translation is ‘to mount again’ or ‘to take to the woods’ (ibid). Another term that became synonymous with remontados is ladrones monteses or 15

Remote Sensing the Margins of the Gold Trade the Spanish contact period wherein they still kept the gold trade going, but at the same time ventured into cattle herding, intensive wet agriculture, and dry cultivation of tobacco which was very profitable.

2007). Pavao-Zuckerman and LaMotta (2007) use both archeology (zoo archeology) and ethnohistory in their research on economic changes in the region. In this example, we see how livestock changes, specifically the rejection, partial integration, then full integration of cattle reflected indigenous response to missionization and eventual engagement within regional economic networks, notably mining (p. 241).

2.3.2 Flexibility as Resilience An overarching theme of my book is that flexibility in economic and social strategies is key to resilience over both the short term and the long term. As noted in the previous sections, the historical literature definitively shows that gold sourced upland was being traded by Northwestern Luzon coastal peoples for centuries before Spanish contact, and there are clear forms of evidence of occupation in many areas of the uplands that were inhabited in the pre-Spanish period, including archeological sites dated to before the sixteenth century.

Early on, mining itself exerted some demand for livestock like pigs, which were offered in the entrances of gold mines. During a cañao ritual, the blood is sprinkled on the entrance of the mines together with a pinch of gold dust. This is to obtain better fortune in finding gold (Wilson 1932: 24). If successful in finding gold, ritual killings of several pig are carried out for deities and anitos as well (Caballero 1996: 170). The entry of the cattle in the Cordilleras is also reflected in the shifting burial traditions. This can be seen in the study of different burial coffin traditions in Patiacan, Apayao, and Danac sites within the highlands of Ilocos Sur (Canilao 2015). Of the three, the Apayao site (dated 1888-1889) reflected class stratification, with a couple occupying a rock shelter interred inside two massive carabao motif log coffins (2012: 64-65). In contrast, the Patiacan coffins, which were hollowed-out logs, appeared to be 1700s based on oral history (ibid). The coffins in Danac, which are hollowed cylindrical logs, seem to be a direct offshoot of the jar burial practice which is found all over Luzon and associated with the Metal Age. Among the three examples, intensification is evident in Apayao and the increased value of cattle is evident in the motif of the baknang/kadangyan ‘elite’ burial (ibid).

In Bodner’s Tukukan excavations, she has an early cultural horizon in the seventh century (1986: 421) which show permanent villages already being established in Bontok (p. 422). There is also evidence of a limited trading system, as well as specializations like spinning, weaving, earthenware ceramics, and lithics (ibid). Bodner argues that it was in the fifteenth century that trading systems in Bontok became intensified (p. 465), a conclusion based on the entry of imported porcelain and exotic earthenware ceramics, even as the local earthenware production became specialized (pp. 421-422, 465). A century later with the advent of the Spaniards, rapid agricultural changes and social transformations swept the Bontok area (p. 466). Extensive migrations were also logged during this period, possibly or even likely associated with Spanish control in coastal areas, disrupted traditional political formations, and access to social capital and wealth. Interestingly the pre-sixteenth century sites did not show rice domestication but instead a pre-rice agricultural economy (p. 466), which had evidence of millet, taro, and other crops. Bodner argues that social stratification (rise of the kadangyan) was brought about by agricultural intensification, specifically the adoption of pond field agriculture for taro (not rice) (1986: 473). The Lubuk site, for instance, has a datable pond field to 1410-1450 (1986: 415).

The introduction of tobacco may have opened up a whole new dimension of both resistance and opportunity for the upland groups. According to Jesus (1980: 178), the smuggling of tobacco by the infieles in the Cordillera Mountains was partly instigated by the increased demand for Philippine tobacco in the 1840s (p. 178). Some of the smuggled tobacco was channeled to British Singapore (ibid). In some cases, tobacco itself became the medium of currency in the illicit barter exchange between the upland and lowland, making this cash crop highly lucrative (1980: 179). The gravity of the tobacco problem is illustrated in the saga of Tonglo village. Tonglo started out as a bulking village serving the gold trade (Vivar 1755 trans. Scheerer 1975, 11-12) from the mining settlement to the coastal trading centers of Aringay and Agoo. A tobacco punitive expedition was unleashed on the Tonglo residents in 1759 because the Alcalde Mayor of Pangasinan was outraged by the high profits made by the natives out of the illegal tobacco plantations they had operated in the slopes of the village (Canilao 2012: 61).

Maher’s (1980) work in Ifugao, accordingly, can be somewhat reconciled with Acabado’s work (2009). Maher’s excavation dates settlements and also limited trade to as early as the seventh century (1980: 235) and Maher also obtained a thermoluminescence date in one potsherd from Kiyanngan village and received a date of AD 1130. Acabado’s (2009) work on the other hand shows that the megalithic terracing of the slopes to accommodate wet agriculture began in the late sixteenth century. Finally, as stated earlier, the archeological survey of early settlements also revealed pre-sixteenth century tradeware in mining contexts (Caballero 1996; Canilao 2015, 2008), again supporting the scenario that mining activities involving gold extraction, significant settlements with intensive agriculture based on either rice or other crops, and interaction with coastal traders who brought

In terms of wet agricultural intensification Acabado (2009) and Bodner (1986) both show that this increased after contact, probably due to more in-migration into the uplands as a result of Spanish colonization (ibid), a trend that will be further discussed below. Thus, we are afforded a picture here of how the upland groups were navigating 16

Background in exotic trade porcelains were features of the social and cultural landscapes. It should be noted, however, that it is quite difficult to extrapolate from a single site to a whole region. The research represented in this book is purposely large scale in scope, and GIS modeling, combined with some regional archeological evidence, is a stronger basis for examining the complexities of resources, landscapes, social formations and interactions, and resilience/ flexibility in the face of change. 2.4 Chapter Summary Through this chapter, the book demonstrated that the production and consumption of gold in Northwestern Luzon dates to the Early Historical Period as seen in archeological, historical, and oral tradition data. This section of the book also shows that Luzon was already an early trade destination based on some early Chinese writings dated before the sixteenth century. These preSpanish writings are significant in establishing that early interaction likely spurred evanescent market encounters on the coast that eventually transformed into permanent market encounters. It is important to emphasize that the main items exchanged were most likely gold for tradeware, specifically exotic porcelain. The next chapter will focus on theoretical spatial models in archeology associated with various scales of human interactions that illuminate how these interior tribal societies may have navigated various layered trade spheres and calculated strategies for maximum social capital within their society and minimum vulnerability in these trade transactions.

17

3 Historically and Archeologically Documented Networks for Gold Working And Coastal Markets in Northwestern Luzon slaves, and other valuables. These five networks are identified in distinct river drainages connecting Spanish Period (and in some cases likely prehistoric) coastal ports connected by valleys and river courses to mining areas typically 40 or more kilometers from the coast, with ‘bulking’ sites situated further down trails for coastal transport. Specific historical sources (including both written and oral sources), along with archeological data, are used to paint a picture of the rationale of interior gold miners to bulk their gold at specific locations, to time their descent to coastal ports for gold trade, and to compete effectively with other communities associated with upland gold products in exchange for lowland resources.

This chapter focuses on the five networks or regions of gold trade in Northwestern Luzon, noting the available historical, ethnographic, and archeological information relevant to each case. This serves as interlocutor to the remote sensing and GIS analysis conducted in this book. Some of the information presented is in block quote form to give the reader the opportunity to assess the interpretation of the information. In the previous chapter, the discussion focused on various theoretical models used by archeologists and geographers to exemplify how societies and groups of varying size and complexity, sharing a physical landscape, might structure specific strategies of social interaction, specialization, and resource exchange, and in some cases threaten coercion or hostility, with the aim of leveraging their unique access to certain resources. In this chapter, there is a shift from abstract and generalized archetypes of trade geographies to an analysis of the historical, ethnographic, and archeological evidence for pre-Hispanic Period and Spanish colonial period gold mining, processing, transport, and trade in Northeastern Luzon, involving strategic planning by small upland tribal communities to optimize their competitive advantage in providing mined gold and forest products to coastal merchants.

3.1 The Tonglo Network Linking the Mines of Balatok and Acupan with the Coast of Aringay and Agoo 3.1.1 Aringay and Agoo Evanescent Markets Tabu-tabuan or tiangge (Allegre 1998: 148), trans. ‘evanescent market encounters’, in contrast to ‘permanent markets’ that were centrally organized, were common occurrences on the shores of the Philippine archipelago. The literal translation of tabu is bucket, which is used to measure merchandize during trade (i.e., grains, edible mollusks). While in more contemporary Philippine markets we see tin cans being used for this same function, it is probable that the original tabu were worked bailer or melo shells (Melo melo sp.)(Canilao 2017a). These structured but ephemeral market interactions were likely necessitated by the fact that gold producers and their gold-seeking clients were part of a chain of multi-ethnic participants who independently met in what Karl Polanyi (1963) refers to as ‘ports of trade’. There are other authors who have looked at ports of trade in Asia (Hall 2011, Shaffer 1996, Chaudhuri 1985).

Key to the market encounter between upland gold miners, lowland coastal-centered buyers and middlemen, and foreign maritime merchants coming from far-flung ports within the IO–WPS/SCS orbit is the concept of an ‘evanescent’ (i.e., ephemeral or transitory) market known as tabu-tabuan/tiangge. Because gold sources are found in the rugged interior uplands a significant distance from the coast, small-scale, low technology gold mining was historically carried out by small communities of tribal miners who extracted the gold in various ways, bulked it, and transported it to the coast as efficiently and as quickly as possible to meet gold-seeking merchants at coastal ports before competing upland mining groups could set up an evanescent market and acquire exclusive access to the transaction. Historical sources indicate that gold bulking stations that aggregated gold from the many small extraction sites extractions were often located near upland vista points where the miners could view an approaching trading ship many kilometers from the coast and quickly ready the gold for rapid transport down efficiently sited trails to the evanescent market.

One enduring characteristic of these ephemeral/transient markets is that these spontaneously emerge on the shores adjacent to where a merchant ships has dropped anchor for convenience of unloading heavy wares (including ceramics, bronze gongs, ivory, and other exotic goods) (i.e., see Chao Ju-kua, twelfth–thirteenth century account in Hirth and Rockhill 1911: 160). In addition, the market is ideally set up near riverine delta locations so that transshipped goods using smaller boats and rafts from the interiors are accessible. During the tabu-tabuan, stalls are quickly set up at the shore with different sellers who placed their island products out on display (Allegre 1998). Based on some early Chinese writings, an evanescent encounter can be as short as a few days, whereas a permanent market encounter can stretch into months on end depending on the return of locals who have taken items in credit (again,

This chapter reviews the oral tradition, written historical, and archeological evidence for five trade networks connecting interior small-scale gold miners with coastal markets on the northwestern coast of Luzon Island where gold could be exchanged for ceramics, bronze gongs, 19

Remote Sensing the Margins of the Gold Trade Here we are introduced to an important offshoot of these market encounters: credit, with delayed reciprocity. This Spanish account jives with the Chinese account of Chao Ju-kua (in Hirth and Rockhill 1911: 159-162) centuries earlier, in which the exchange of porcelain and other imported goods were significant enticements to provide raw gold and forest products to continue this exchange and that credit greatly facilitates these transactions. In the book, it is argued that the increasing role of credit or debt finance in these EHP transactions correlates with the shift from more evanescent towards more permanent markets in NWL. The Rada quote above is important because it alludes to a transitional phase from more evanescent to semi-permanent markets with creditors being the foreign merchant vessels. Because of the added requirement to be able to extract the gold for the waiting creditors, having efficient trails which facilitate quick trips from coasts to interiors is of maximum advantage. It is argued that soon the coastal settlements eventually became the creditors forging trade connections and patronage with the interior groups.

see Chao Ju-kua, twelfth–thirteenth century in Hirth and Rockhill 1911: 159-162)(Canilao 2017a). As noted in chapter two, Ibaloi small-scale miners definitely participated in the coastal evanescent market encounter, as a number of historians (Savellano 2009, Mateo 2004, Scott 1988, Bagamaspad and HamadaPawid 1985) have gleaned from their reading of various Spanish accounts. The historian Mateo elaborates that ‘[w] ith their gold and forest products the Igorots would trek via land trails and waterways to Ilocos and Pangasinan to the west and Cagayan to the east’ (2004: 56). The use of land trails meant that they would have to use the most efficient path from the bulking villages to the coast, and the use of waterways also meant that they would be able to transport more goods using flat riverine rafts or riverine boats (Canilao 2017a). Access to gold, as a marketable commodity that was highly valued as raw material for luxury items for local lowland Philippine elites and for export into broader oceanic markets, brought lowland provisions such as foodstuffs and salt, and more importantly prestige goods from overseas such as porcelain that might not otherwise have been accessible for the highland Ibaloi. Miguel de Legaspi, a sixteenth century Spaniard, indicated in his writings that the native interior populations do not get more than what they need from the ground in their mining activities, or more than the amount necessary to have access to the basic necessities, because this will anger their gold-god named Balitok:

As mentioned above, the use of efficient land trails that allowed for quick movements from bulking villages to coast is quite important in terms of economic competition. The Least Cost Path function in GIS can facilitate this type of predictive modeling. Ship visibility from a significant distance in the interior also appears to be a very important factor for upland miners to participate in an evanescent market. Visibility is enhanced by a strong knowledge of landscape in the interior, that was likely important in periods well before European merchants and administrators landed on the island, creating a mental map for surveilling encounters with outsiders that could be as likely to be dangerous as potential allies and traders. It is likely that these interior societies practiced some form of predictive modeling of coastal-upland vistas and effective routes towards the shore, allowing them to maximize their knowledge of activities on the coast, looking for both possible nefarious encounters and opportunities to outcompete with other groups in gaining valuable lowland trade items through sophisticated knowledge of the landscape. The visibility and viewshed functions in GIS simulate this type of predictive modeling: what can be seen at various points on the landscape, and how surveillance sites allow sophisticated calculations for determining the placement of gold bulking centers and the most expedient routes to the coast.

Because they say that their god orders them not to take out gold, except on the arrival of foreign vessels. (Legaspi 1569: 57) Writing in 1756, Spanish Augustinian priest Manuel Carillo states: The custom of the Filipinos is to go to the placer or mine and get exactly what is necessary when they want to cover the cost of the tribute or their basic needs, and do not worry about anymore. (273) Another description indicates that Ibaloi gold miners themselves were directly dealing with the foreign merchants without a middleperson (Canilao 2017a). However, the passage below indicates the early beginnings of credit in these transactions.

A caveat, however, is that these GIS modeling techniques only provide a baseline predictive model, which provides a stage to discuss variability in the real world. The book has consistently seen that the historical model rather than the baseline predictive model matches the remote sensing data on the ground. Further, it should also be emphasized that it is the nodes that connect the transportation network that will facilitate determination if a particular market is more evanescent or permanent in orientation, thus providing some degree of relative chronology.

No one of these Indians [sic] has more than a very little gold. For if they get a couple of pairs of earrings and a couple of pairs of bracelets and a pair of anklets for the feet, they do not look for any more, they do not strive to hoard it… If perchance any merchant junk comes [to trade] they buy all the goods in credit, after which, in order to pay, all the village assembles and goes to the mines, in order that no one will dare to touch them, and accordingly get the gold that they have to give. (Rada 1569: 224-225)

At this juncture, we will examine the landscape features associated with gold production, including the strategic 20

Documented Networks Gold mined from the Acupan area, which receives preliminary processing in Tonglo, is eventually transported via hiking trails and rafting routes until it gets to the coastal settlements like Aringay and Agoo to be loaded onto maritime vessels. Aringay, for instance, is commonly identified as one of the oldest coastal settlements based on both written historical sources and oral tradition sources (Prill-Brett 1998, Bagamaspad and Hamada-Pawid). Sixteenth century Spanish chronicler Andres Mirandaola writes that Alinguey or Aringay was a source of gold in Luzon (1574: 223), and in fact Spanish conquistador Martin de Goiti also exacted gold from this village in the sixteenth century (Newson 2009, Canilao 2017a).

placement of gold bulking centers for rapid downhill transit, the placement of surveilling features (i.e., features with clear vistas towards coastal trading towns and arriving ships), and the judicious routes promoting rapid transit to the coast, examining this purposefully constructed social landscape by beginning with the Tonglo network. 3.1.2 Tonglo, Gold Bulking Village Early Cordillera scholar Otto Scheerer identifies Tonglo as one of the early Ibaloi villages that started out as a gold trading center before Spanish contact in the sixteenth century, and eventually albeit, briefly a mission center in the eighteenth century (1975: 198). Scheerer cites the work of Pedro Vivar an Augustinian priest who established the mission center,

3.2 The Gasweling Network Linking Apayao, Locjo, Cabcaben, and Catampan Mines with Calincamasan, Baratao, and Bauang on the Coast

Tong[l]o is the largest of the Rancherias known to me; it consists of 220 souls of good friendly people of the most sincere nature encountered by me in this country. The town is situated four leguas east of Cava [Aringay] in a gully formed by a steep mountain, and in an environment that lacks even a small stretch of level space… Hither is brought all the gold that comes dirty from the mines; they refine it somewhat and take it down to the Ilocos for sale, bringing up in return so many buffaloes and cattle. (Vivar, as quoted in Scheerer 1975, 184)

3.2.1 Some Ethnohistorically Documented Settlements in the Gasweling Network In this section we look at the archeologically (i.e., Canilao 2011, Tidalgo 1979) and ethnohistorically (i.e., Newson 2011, Bagamaspad and Hamada-Pawid 1985) delineated coastal sites of Bauang, Baratao/Atuley (in present-day San Fernando, La Union), Calincamasan/Aludaid (in presentday San Juan-Bacnotan, La Union), the bulking station Gasweling/Darew (in present-day Kapangan, Benguet), and the mining villages of Cabcaben, Catampan, Apayao/ Galan, and Locjo (in present-day Tublay, Kapangan and Kabayan, Benguet) (Fig. 3.3). Trade in this area appears to have flourished during the Early Historical Period up to the Historical Period (tenth to early twentieth century). The Gasweling site (also called Darew, see Bagamaspad and Hamada-Pawid 1985: 42) is identified as an Ibaloi pregenerational settlement.

Indeed, the account shows that Tonglo was a gold bulking village. In principle any bulking village would occupy a halfway area between mines and the coastal settlements. Here gold receives preliminary processing such as getting bulked into ingots or cakes ready for the overseas transport. Outside Luzon, bulking stations have also been noted in other gold producing islands in Southeast Asia like Sumatra and Borneo (see Dobbins 1983 and Harrison 1970). Because of the importance of commemorating Tonglo for its stakeholders, several Cordillera scholars have attempted to pinpoint its location (Prill-Brett, Salinas-Ramos, and Follosco 1998, Reyes-Boquiren 1997, Bagamaspad and Hamada-Pawid 1985, Scott 1974, Scheerer 1975). Using a triangulation of early historical records with systematic archeological survey, Tonglo was found on the slopes of Mount Calugong, Barangay Taloy Norte, Tuba Municipality, Benguet (Fig 3.1, 3.2) (Canilao 2012: 58-78).

Based on the oral tradition accounts of the Ibaloi gold miners, the Gasweling settlement was similar to Tonglo in that it was located on a strategic mountain top with an uninterrupted view of the coastal settlements: Darew [Gasweling] is located on the top of the mountain west of Kapangan, commanding a vantage view of the coastline and the fertile valley to the east. (Bagamaspad and Hamada-Pawid 1985: 42) Bagamaspad and Hamada-Pawid explain that Darew (Gasweling) is distinctly remembered as one of the earliest settlements of the Ibaloi:

Often described as ‘dirty,’ the gold ore brought to Tonglo may have principally been mined from Acupan as described in published oral tradition sources (Scheerer 1975: 202). Acupan, specifically the mines of Balatok (from balitok lit. trans. gold and god), is also the closest to Tonglo and is one of the ancient mined areas in the region. Augustinian Priest Father Antolin O.P. also included Acupan as an early gold mine in his map dated 1789 (see Scott 1988). Acupan is also in the same gold-rich river valley as Antamok called the Virac valley. Antamo[k] mines was the gold mining site whose gold deposits were assayed by Spanish officer Capt. Alonso Quirante (1624: 283) (Canilao 2017a).

Primacy in Southern Benguet is attributed to the twin settlements of Darew-Palaypay in Tublay; ChuyoTonglo in Tuba; and Imbose-Amlimay in Kabayan by informants of pre-generational accounts. (1985: 28) Settlements like Palaypay in Kapangan and Imbose even farther east in Kabayan, however, are typical of a location with emphasis on river valley agriculture rather than gold mining and panning as a way of life (Canilao 2011). Early 21

Remote Sensing the Margins of the Gold Trade

Figure 3.1. Composite Map showing Tonglo gold trade network in Northwestern Luzon. Note that this network features an evanescent market situation (ASTER GDEM is a product of METI and NASA).

migration into Benguet was recounted and recorded as follows:

movement was into fertile valleys of the southern Cordilleras through three main drainage systems: the southwest Aringay and Galiano rivers into the Mangitkiran Range to Chuyo (Bakakeng) and Tonglo (Tili) in Tuba; the northwest. Amburayan river into one of the tributaries at Darew (Gasw[e]ling) and Palaypay (Pungayan) in Kapangan; and the mighty Agno River

It all began, ‘Nonta… bayag…’ very long ago… in twin settlements identified by informants of the municipal histories of Tuba, Tublay, and Kabayan. From the western coastal plains of northeastern Luzon, 22

Documented Networks

Figure 3.2. Tonglo in Mount Calugong, westerly view from Tongshu locality in Mount Santo Tomas (photo by author 2012) (Canilao 2017a).

from Lingayen, Pangasinan first into Baloy in Itogon and further inland into Imbose (Pacso) in Kabayan and Amlimay in Buguias. (Bagamaspad and HamadaPawid 1985: 42)

village wise men or a council of elders. (Bagamaspad and Hamada-Pawid 1985: 42) They further state…

The strategic-defensive context of Gasweling village is provided in the Tublay Municipality accounts:

They had access to the lowland plains extending from Tagudin to Lingayen in the west, Ahin and Bayombong to the east, and Loo to the north. Within this general area, they traded, intermarried, forged alliances and came to recognize as kin those who lived in the more populous settlements of Tagudin, Agoo, Tubao, Ambangonan (Pugo) Lingayen, Dagupan, Binalonan, Tayug, Safid (San Manuel), Imogen, Ituy, Tinok, and Ahin.

High cliff walls of the adjoining Amburayan River provided a natural deterrent to unfriendly incursions from the coast as well as busol attacks from the north. It was also naturally shielded from epidemics brought in by later migrants. (Bagamaspad and Hamada-Pawid 1985: 42) Again, this bulking station was strategically located to have a panoptic view of the coast. This passage above also points out its defensive feature, also found in other bulking stations like Tonglo and Apayao (Canilao 2017a and 2017b). What is notable in the accounts is the prosperity of the villagers brought about by the gold trade:

In ordinary times, the people mined gold at Cabcaben (Bingaongao, Bongdolan, Tublay) and along the Tuel river to Bagong in Sablan. They tended rice fields in Binalonan and cultivated salt beds in Lingayen. (1985: 42) The interior and coastal settlements may have developed connections to Gasweling later in the Historical Period when more periodic and eventually a more permanent markets had already been established. In modeling the

For many years, the people lived in peace and prosperity rendering obedience to the leadership of warriors and the rich men who governed with the concurrence of 23

Remote Sensing the Margins of the Gold Trade

Figure 3.3. Composite map. Note Gasweling gold trade network in Northwestern Luzon. Note that this network features an evanescent market situation (ASTER GDEM is a product of METI and NASA).

evanescent market, we factor in Euclidean distance to identify the closer coastal settlements, the logic being that ship spotting and expedient upland to coast transit are crucial elements in quickly reaching coastal traders. Thus, the settlements of Baratao, Bauang, and Calincamasan are important sites to model the early gold trade (Canilao 2017c).

Bagamaspad and Hamada-Pawid’s informants also talk about a twin settlement that arose adjacent to Gasweling: Palaypay was the center of extensive trade carried on by individuals accompanied by heavily armed escorts through non-ally lowland settlements and to the coast. Gold was exchanged for brass, gongs, porcelain jars, 24

Documented Networks Gasweling network goes on:

clothing, pig-iron, blankets, agricultural products and salt among others. Slaves were also procured in exchange for the yellow metal. (1985: 42-45)

Besides the above-mentioned places [they mention Atok, Ambuclao… area] Catampan and Bileng-Bilis in Kapangan had also gold mines that people worked through lode mining. Placer mining was also done in the Amburayan River. (1985: 73)

Bagamaspad and Hamada-Pawid explain in detail the social network of the lowland and the upland villages, even connections to ‘Princess Urduja’ who figures prominently in Pangasinan lore:

Catampan is a sitio in Barangay Balakbak, Kapangan, while Bileng-Bilis itself is a Barangay in Kapangan (Canilao 2017c).

The extent of inter-settlement alliances is climaxed in the memory of Tublay informants with the reign of Deboxah, Princess Urduja, in Pangasinan. She is acknowledged as the granddaughter of Udayan, an outstanding warrior of Darew. Her death signaled continuous decline of kinship and alliance between highland and lowland settlements. (1985: 45)

Father Francisco Antolin O.P. plotted the Apayao mines as well as Locjo mines in his 1789 map. Victor Ananayo georeferenced this map in William Henry Scott’s (1988) book Notices of the Pagan Igorot in the Interior of the Island of Manila/Francisco Antolin 1745-1796. Apayao mines in the georeferenced Antolin map overlap with Tabio mines in a 1:250000 scale topographic map from the National Mapping and Resources Information Authority (Fig. 3.4). Santo Nino mines appear also within the Tabio ridge and Scott states that Galan mines is just (a stone’s throw) north of the ridge where Fort Santiago was established by Antonio Carreno de Valdes in 1623 (1974: 34). He confirms that the Fort is located at the Santo Nino Mines which is in the vicinity of present-day Barangay Ambassador (Scott 1974: 30). This allows us to plot Galan mines just north of present-day Barangay Ambassador.

Gasweling appears to be the crucial settlement and bulking village within this network which has networking ties with coastal settlements. The bulking village of Gasweling was digitized as being at the peak of Mount Salat in Barangay Gasweling, Benguet (Canilao 2017b). Archeological data reaching back to the twelfth century through the sixteenth century Spanish contact period was unearthed in Calincamasan/Aludaid. Some Late Sung to Late Ching period Chinese tradeware ceramics, for instance, were excavated specifically from Calincamasan (Tidalgo 1979: 11). The Aludaid site, on the other hand, appears to have flourished from the fourteenth to fifteenth century, based on the presence of Late Yuan and Ming Chinese tradeware (ibid).

At this juncture, we look at two historical accounts that provide some locational cues to important historical settlements. The first is the Juan de Salcedo account of his coastal conquest of Luzon. Later we also look at Alonso Quirante’s 1623 (published 1624) account closely, to get a sense of the landscape, which his expeditionary force traversed en route to the Galan Mines to assay gold (Canilao 2017c).

Historical demographer Linda Newson (2011: 180) shows that Baratao had 910 houses when it made contact with Martin de Goiti’s expeditionary forces in 1572-73. Baratao is the old name of present-day San Fernando and indeed Newsom plots this location as the coastal cove area of present-day San Fernando, a few kilometers north of Bauang (ibid). According to Newson, Baratao, together with Turney, Aringay, and Dingras, was identified to be one of the mines rich in gold (2011: 188). Baratao and Bauang were also surmised to be in Pangasinan (2011: 190). Pangasinan historian Rosario Cortes also lists Baratao and Bauang as existing at the arrival of the Spaniards (as pericontact settlements) (1974: 26).

3.2.2 Juan de Salcedo 1572 Account of His Conquest of Northwestern Luzon Early Filipino historian Isabelo delos Reyes writes in his work History of Ilocos Volume 2 that Salcedo in his 1572 conquest of Luzon sacked several villages along the coastline of present-day Zambales, Pangasinan, La Union, Ilocos, and Cagayan (Fig. 3.5). As for the Pangasinan to La Union area, Reyes narrates:

Although ethnohistorian Felix Keesing places Baratao in Bauang (1962: 54), the Juan de Salcedo Luzon conquest account is more supportive of Baratao being in San Fernando based on remote sensing through WorldView-2 satellite imagery. Nevertheless, Bauang is also modeled as a coastal settlement, since all sources agree that it is one of the earliest in the area as well as it is strategically located at the mouth of the Bauang/Naguilian River (Canilao 2017c).

On the night of the following day, Salcedo arrived at another town named Atuley, nestled on a very steep rock defended by the natives safely from above. When Salcedo saw that it was inaccessible from the place where they had landed, he divided his men into groups of six soldiers each and ordered them to look for a gorge through which they could safely climb up. They found a narrow pass and together with six of his men, Salcedo managed to reach the top with some effort, without being noticed by the natives. When the latter finally saw the Spaniards, they fled the town in confusion. Followed by his men, Salcedo entered the town and

Bagamaspad and Hamada-Pawid already pointed out that Cabcaben or Bingaongao and Bongdolan, in Tublay existed as ancient mines (1985: 42). The list in the 25

Remote Sensing the Margins of the Gold Trade

Figure 3.4. National Mapping and Resource Information Authority 1:250000 map showing Tabio Gold atop Apayao mines reprojected to WGS 84 UTM Zone 51Q (Canilao 2017c).

saw that it was the most beautiful that he had seen so far in the Philippines with its well-ordered streets and houses. The Spaniards found out from two natives that the town was the capital of the district, and that in the interior were many barrios and towns. Salcedo ordered them to tell their companions that he only wished to be their friend. After two hours, the same natives returned with many of their prinicipalia. The Spaniards received them warmly, thereupon the natives asked if they could call all their companions so that they could make a peace treaty on the following day; they asked the Spaniards to take their boats out in the meantime. (Reyes 1890 trans. Imson 2014: 139) [emphasis mine].

habitation sites (villages), Lacsina concludes based on archeological data that the ijangs themselves are primarily used for human habitation (2009). Excavation in Savidug ijang has revealed some Song and Yuan ceramics with the recovery stratigraphic layer radiocarbon dated to 760 plus/ minus 190 uncalibrated BP (Bellwood and Dizon 2013: 52) (Canilao 2017c). 3.2.3 Alonso Quirante 1624 Account of His Expeditionary Force to Benguet and the Assaying of Gold Nuggets Alonso Quirante’s Expedition to the Mines of the Igorottes (1624) shows that Galan is within the same ridge formation where Apayao mines, Santo Nino mines, and Tabio mines are located (Fig. 3.6). A reconstruction of his expedition is presented here using the historical accounts and overlaying the digitized waypoints on a digital elevation model of the area. The geographic setting is closely revealed in the accounts:

What is quite important in the account is the description of an elevated rock hill village with an orderly layout. This natural rock formation-turned-human settlement is a hillfort or an ijang (Canilao 2017c). Ijang is equivalent to the word ili/idi plus local affix (pers. comm. Grace Barretto-Tesoro 04 November 2018) among the Ilocanos who inhabit the coast of Northwestern Luzon (Lacsiana 2009: 69).

Sunday, the twenty-fifth, I went [to] pass the night on account of the convenient supply of water, at Rio Frio (i.e., Cold River) with my said men, marching through

Although Hornedo (2000) argues that ijangs are temporary 26

Documented Networks

Figure 3.5. Map showing Juan de Salcedo’s 1572 seaborne conquest of Northwest Luzon (Canilao 2017c).

27

Remote Sensing the Margins of the Gold Trade

Figure 3.6. Route of the Expeditionary force of Alonso Quirante to the gold mines in 1623 (ASTER GDEM is a product of METI and NASA) (Canilao 2017c).

the extremely hot sun for one and one half leguas. Next day, Monday, February twenty-six, about one o’clock, I reached the new mines called Galan by their natives, located about three leguas from Rio Frio. As the Ygolotes had learned of our approach, or had seen us about to set fire to some houses – about two hundred which they had located in various places about the said mines and hill – they sought shelter without leaving anything except some small heaps of metal which they were digging in order to work.

On one of the five elevations which I have said that the Ygolotes worked, namely, the said new one called Galan (it is being the chief one, as I have said), I camped, and built the Fort Santiago, under whose advocacy (i.e., Santiago or St James) they say it was before… The mouths of the mines are in the northern part [of the ridge], about a stone’s throw from the said fort, and the mine discovered extends from above downward in the manner of a horizontal vein or a shell for the distance of a musket-shot from northwest to southeast, and then twist about for another equal distance to the direction that looks toward northwest and west, until it disappears into the depths of a ravine or watercourse where there is but little sun. That is not the case with the one that extends northwest and southeast, for it is flooded with sunlight most of the day. When I reached that place the Ygolotes were working the said mines through many mouths or passages that they had opened, following the metal of one large vein, from which they were taking out the ore that was softest and easiest to dig, although it contained blue iron pyrites that contain antimony.

Next day, the twenty-seventh, having reconnoitered the said place, and having seen that it was suitable and secure, and that within a stone’s throw on the same elevation were the mines and veins of most importance that are yet known to exist among the said Ygolotes, according to the information given by men who already had experience of them before, I determined to establish a camp and fortify myself in them. (Quirante 1624: 266-267). William Henry Scott argues that Rio Frio is one of the headwaters of the Bauang/Naguilian River (1974: 34) (again, Fig. 3.6). Continuing with the account, Alonso Quirante provides more specifics about Tabio ridge:

Having investigated and examined the above mentioned, I judged it best to open a trial place or mouth high up, and in the middle of all the mine works that the said 28

Documented Networks Ygolotes were carrying on, in order to get all the body of the metal from the top which is more than one braza wide, and from the crust of the earth. On the fifth of March following, we began to open it, and, following the opening for ten estados, we encountered the said mines that the Ygolotes were working, by which our field of work was enlarged much more on the level, at the sides, and vertically; and we continued to get metal for assaying (Quirante 1624: 280-282) [emphasis mine].

Monforte led expeditionary force of 100 Spanish troops and 2,000 local infantrymen to Lepanto mines coming from the Tagudin coastal village (Scott 1974, also cited in Bagamaspad and Hamada-Pawid 1985: 146). He explored seven entrances to gold lode mines (ibid). Bagamaspad and Hamada-Pawid (1985) argue that regular trading relations was already established in the area by the 1600s, In Bakun, traders bartered Chinese jars from Tagudin and blankets from the Ilocos for gold from Mankayan. Gold and mining activities in Mankayan branched off into intensive trade activities. With gold mined from the labon [lode mines], the people of Panat bought hogs, carabaos, rice, salt, and clothing from Tagudin through Cervantes. The gold was also made into earrings called dinampulay and head beads called benge which were sold to the Bontocs in exchange for blankets for the dead. (p. 76)

The description of Galan matches the whole mountain ridge that includes the boundary of present-day municipalities of Atok and Tublay within the Barangays of Daklan, Caponga, Caliking, Ambassador, Ba-ayan, and Acop’s. This ridge is also the source of the headwaters of the Naguilian River and is labeled as ‘Tabio gold’ in 1:250000 scale topographic map. The topographic map identifies the whole formation as Mount Toyongan which separates Agno from the Amburayan-Naguilian water basin. Analysis of highresolution WorldView-2 imagery in the discussion section will also allow us to look at the area in Galan that was assayed by Quirante for gold (Canilao 2017c).

The trade routes also went through Cervantes, immediately north of the mines. The town, within the river basin of the mighty Abra River, indeed later on became a hub from Lepanto to the historical port of Vigan. Cervantes as a bulking village facilitated organizing the goods bound downstream. It was also articulated to coastal settlements directly to the west like Tagudin with this route becoming the present-day road known as Bessang Pass (Canilao 2017b).

3.3 The Lepanto Network – Apayao and Danac linking Lepanto Mines with Tagudin and Purao 3.3.1 Lepanto Mines as the Oldest Worked Mine in the Cordilleras

3.3.2 Purao and Tagudin Coastal Settlements The two ancient coastal settlements west of Lepanto include Purao and Tagudin. Frey Juan de Medina generally talks about gold found in Ilocos having been bartered by people from the interiors:

In this section we look at the Lepanto gold district (Fig. 3.7) and its adjacent coastal area to the West. The area includes the tri-boundary of the present-day provinces of La Union, Ilocos Sur, and Benguet. Regarding the ethnohistory of the region, Keesing states that,

A great quantity of gold has been obtained from the province (Ilocos), not that the province yields it but the Igorots bring it down from the mountains[. W] hen peaceful they bring down gold, which they extract from their mines[; ]and they exchange it for cattle, which those along the coast own[. ]They also take away blankets, which the people of Ilocos make of excellent quality… (Medina 1893 in Bagamaspad and HamadaPawid 1985: 279)

[o]ld trading trails exist[ed] between the coast and interior[,] the most important are those linking the mountain communities with Balaoan [Purao] and Tagudin to the South, and with Candon to the North. (Keesing 1962: 92). Lepanto in Mankayan, Benguet is commonly identified as the earliest site for traditional lode and placer mining (Bagamaspad and Hamada-Pawid 1985: 146, 205, 227). It is also the place of origin for the Kankanaey ethnolinguistic group (Bagamaspad and Hamada-Pawid 1985: 39). Gold prospector Laurence Wilson states that the site for industrial mining used to be mined in ancient times (1932: 9). Pelican Slide of the Suyoc gold vein, the largest mined vein in the area (pp. 3, 15), features a long-eroded slide, half a kilometer wide, that is testament to gold mining in the area. There is even an origin legend of Suyoc gold, believed to have been mined first by natives from the historical settlement of Cervantes in the headwaters of the Abra River (p. 27) (Canilao 2017b).

The popular color scheme for this Iloko blanket included a two-color combination playing on black, dark blue, white, yellow, and red pigments (Bagamaspad and HamadaPawid 1985: 78-79). As a side note, it is important to point out that historian Grace Mateo notes a settlementorganizing principle in which ‘areas apportioned or allotted as encomienda were the [original, pre-Spanish] population centers’ (2004: 85). Using this principle, we can examine Spanish historical maps and identify Purao (Balaoan), Vintar, Danglas, Ylagua (Laoag), Bonsan (San Nicholas), and Cacabayan (Paoay-Batac) as good candidates for original, pre-Spanish settlements (2004: 85). Purao and Tagudin are indeed the two coastal towns that are the closest to Lepanto. In fact, Purao (Balaoan) and

The earliest written record of mining in the area dates back to 1667 when Spanish Admiral Pedro Duran de 29

Remote Sensing the Margins of the Gold Trade

Figure 3.7. Composite map. Note Lepanto gold trade network. Note that this network features an evanescent market situation at Apayao (ASTER GDEM is a product of METI and NASA).

Tagudin also appear in oral tradition accounts published by Bagamaspad and Hamada Pawid’s map of circa 1500s settlements (1985: 27) (Canilao 2017b).

mentioned in early Spanish records when conquistador Juan de Salcedo encountered the inhabitants of this village in 1572 (Keesing 1962: 96). Using the Salcedo account, Purao is located a bit inland, following the Amburayan River, which is the second largest river in Northwestern Luzon, next to the Abra River (ibid). Salcedo and his troops had to traverse some ground en route to the village,

Tagudin was the staging area for Admiral Monforte’s expedition into Lepanto. He used an existing trail at the time that led to the mines. Purao on the other hand was 30

Documented Networks

Figure 3.8. Composite map. Note Lepanto gold trade network. Note that this network also features a permanent market situation at Danac different from the Apayao (ASTER GDEM is a product of METI and NASA).

and in their absence the troops guarding their vessels were attacked by some village warriors. It should be noted that the contemporary Balaoan town is around 5 kilometers inland and at least 3 kilometers from the Amburayan River. In terms of historical demography, the population of these ancient villages can be surmised from the amount

of tribute both Purao and Tagudin, along with Candon, surrendered to the Spaniards. Based on the 1591 account of Encomiendas, Purao had 2,000 tributes from a population of 8,000, Tagudin 900 tributes from a population of 3,600, and Candon 900 tributes from a population of 3,600 (Dasmarinas 1591: 105) (Canilao 2017b). 31

Remote Sensing the Margins of the Gold Trade 3.3.3 Danac and Apayao Bulking Stations

during what he termed the protohistoric period, now also called the Early Historical period (2003: 1,20).

Some systematic archeological surveys in 2012 were conducted on two of the bulking villages that were articulated to the Lepanto-to-coast network. These are the two ancient settlements of Danac and Apayao (Canilao 2012). Danac dates back to the pre-Spanish period, featuring a unique burial tradition characterized by hollowed-out log cylindrical coffins made for mummified remains (ibid) (Figs. 3.9, 3.10). Apayao, on the other hand, features some anthropomorphic cattle carabao motif coffins (Fig. 3.11, 3.12) similar to those documented by Dean Worcester during the turn of the twentieth century (see CSA22590 photographer DC Worcester © The Field Museum). These coffin traditions show the connection between the coffin traditions of the Danac and Apayao settlement with Igorot settlements further inland (Canilao 2017b).

At a regional level, Mateo argues that Ilocos (Fig. 3.13) was part of an international trade network that may have extended to Melaka, India, and China (2004: 42-43). Mateo begins to discuss how this periphery or margin in the gold trade is in fact articulated to a bigger world system overseas through China: A 1574 estimate claims that thirty to forty Chinese sampans came to the country annually and during prosperous years even as many as fifty. (2004: 41) Curator Berthold Laufer, who specialized in Asian anthropology at the Field Museum at its inception in the early twentieth century, then claimed that the Chinese– Philippine trade was in place by the twelfth century (1908: 154). But H. Otley Beyer, the father of Philippine anthropology, pushes this date even further back to the tenth century:

3.4 Angaqui Network Linking Patiacan/Minlaoi and Possibly Lepanto Mines with Dumaquaque and Candon in the Coastal Account

[R]esearches have shown that the first considerable trade between the Islands and China was due to the hard and venturesome Arab traders who, being ousted from the South and Central China coast by the antiforeign uprising shortly after the middle 9th c, found a new way to secure Chinese and Korean productions for

3.4.1 Ilocano Sampans Loaded with Gold Headed for China and Porcelain on its Return Trip Archeologist Eusebio Dizon argues that the Philippines participated in a regional maritime exchange network

Figure 3.9. Danac site near Amburayan River using WV2 NIR variation band combination (Courtesy of the DigitalGlobe Foundation) (Canilao 2017b).

32

Documented Networks China (978: 52). He cites an account by Jesuit priest Alonso Sanchez who observed gold from the Philippines being cast in the form of small boats, weighing half a pound each during his trip to Canton in the sixteenth century (Chan 1978: 52). The so-called ‘periphery’ was also sending its own boats to the center apparently because Ilocanos eventually adapted their own maritime vessel sampans, referred to locally as the biray. Remains of this vessel type were seen in Pandan, Caoayan, Ilocos Sur during a systematic riverine archeological survey (Canilao 2015) (Fig. 3.14). The vessel and the span of its operations is alluded to in the Ilocano oral tradition of Biag ni Lam-ang, Dua pay ti sasakayan a balitok a bulawan nga agbubunag ti pinggan idiay ili a Kasanglayan, ta nalpas met a naikamang iti ari idiay Puanpuan. Isu ti partesko ken kabagian idiay ili a Kasanglayan, isu ti napanna nagbiahian tay barangayko a sampan; pamayak adda itan, ta napan nagkarga ti pinggan. (Yabes 1958: 38) [emphasis mine].

Figure 3.10. Danac coffin tradition (Eduardo Bersamira, Commissioned Work, Ilocos Sur archeology Project II, 2012) (Canilao 2017b).

the Western market by tracing a new route northward from the Malacca straits – passing through Bornean, Philippine, and Formosan waters to southern Japan and Korea. Some Western and Southeast Asia trade goods were carried by these voyagers into the Philippines and elsewhere on the northern trip – whereas Chinese ceramics and other products were distributed through the Philippines on the southern voyage. This trade seems probably to have begun about 890 AD, or shortly thereafter and continued until at least about the middle of the 12th c. After the middle 10th c, however, the Arab ships began again to trade with the Central and South China ports of Chuangchow and Canton – and many of the ships seem to have gone southward from the first of these ports through the Philippines and Bornean waters rather than along the Indo-China coast. From the late century onward the Arab ships pursued both routes and the first actual recorded mention of the Philippines in Chinese written history (as so far available), is the arrival of an Arab ship at Canton with a load of native goods from Mindoro in the year 982 AD. (Beyer 1979: 115)

I have two gold tradeships plying between here and the Chinese country trading in porcelain. I have commercial connections with the king of Puanpuan, My relative in that Chinese land. My boat sampan has gone there on a voyage, and probably now it has returned with its cargo of porcelain. In this passage we see that gold was exchanged mainly for porcelain, but according to Junker, Between the tenth and fourteenth centuries Philippine chiefs in a limited number of polities strategically located along the western littoral [includes Northwest Luzon, Ilocos] dominated a low-volume trade for Chinese porcelains and a wide variety of other foreign luxury goods (glass beads, silks, bronze objects). (1999: 19)

The Arab ship, or dhows, mentioned in the passage would indeed sail through Northwestern Luzon before crossing the Batanes channel, headed for Formosa (present-day Taiwan) (Canilao 2017d).

The fifteenth and sixteenth centuries saw the intensification of this trade (ibid). This is evidenced by archeological units in burial sites in Northwestern Luzon which reveal whole ceramic pieces as well as sherds (possibly from habitation sites) provenienced to Annam/ Dai Viet (present-day Vietnam), Siam (present-day

Chinese Studies expert Albert Chan argues that Philippine gold actually ended up in smithies of Canton in Southern 33

Remote Sensing the Margins of the Gold Trade

Figure 3.11. Apayao site (using WV2 NIR variation band combination) (Courtesy of the DigitalGlobe Foundation) (Canilao 2017b).

earnings just to purchase a single jar from Candon (ibid) (Canilao 2017d). 3.4.2 Gold from the Ilocos Coastline The Ilocanos appear to have emerged as the middlepersons in the gold exchange: Gold and forest products were not obtained in the lowland, but through trade with the Igorots. The Ilocanos served as middlemen, buying gold from the Igorots and selling it to foreign traders. For the gold and the forest products of the Igorots, the Ilocanos bartered rice, cotton, livestock and salt. There are no figures to indicate the volume of trade between these two groups, but their commercial relations appear to be significant since trade became the focal point of Ilocano-Igorot relations. (2004: 45)

Figure 3.12. Apayao coffin tradition (Ilocos Sur archeology Project II, 2012) (Eduardo Bersamira, Commissioned Work, Ilocos Sur archeology Project II, 2012) (Canilao 2017b).

Several sources indicate that the Ilocano settlements transacted business with not only Chinese but also Japanese merchants (Caballero 1996: 29, Keesing 1962: 97). An unsigned Relacion (1586) also shows that the Ilocanos mediated these transactions by dealing with the upland miners to their east in the following passage:

Thailand), Khmer (present-day Cambodia), and China (i.e., see Canilao 2015, Tidalgo 1979, Legazpi 1974). German traveler and ethnographer Hans Meyer also observed stoneware dragon jars in Candon, Ilocos Sur intended to be sold in the upland areas in Mankayan, Lepanto mines (1975a: 117). Meyer himself was perplexed when his personal Igorot interpreter wasted all of his month’s

They [Ilocos people] are husbandmen and possess very large fields. Consequently, it is a land abounding in rice 34

Documented Networks

Figure 3.13. Angaqui trade network in Northwestern Luzon. Note that the shift of this network Angaqui node from Tila peak locality to the Abra Riverbanks reflects the encroachment of the Abra network to this network (ASTER GDEM is a product of METI and NASA).

35

Remote Sensing the Margins of the Gold Trade

Figure 3.14. From Ilocos Sur archeology Project season one (ISAP-1) files. Main photo: Biray remains (sampan) at Pandan, Ilocos Sur (Photo by author). Inset: the beached biray circa 1980s (Ilocos Sur archeology Project I, 2011 sourced from Mayor Germelina S. Goulart) (Canilao 2017d).

and cotton. There is also considerable gold, for the chief mines of these islands are situated in the mountains of that province. These people [Ilocos people] enjoy it for they have more communication with the miners than anyone else. (unsigned 1586: 382)

sixteenth century Spanish accounts as the wealthiest in the archipelago suffered the general post-Spanish Conquest decline in gold production after the sixteenth century (Scott: 1974: 13). This so-called decline, however, is increasingly debunked, by historians, noting that several of the interior mining ethnolinguistic groups merely ventured into other diverse lifeways like cattle herding and upland agriculture, lessening the effort put into gold mining and panning (Canilao 2017d).

That they were merely middlepersons to the transactions can be surmised by the fact that the actual extraction was carried out by the Igorots and not by them. Yet the gold came to be known as Ilocos gold in the eyes of foreign merchants. One Spanish Governor, General Francisco de Sande, even states:

Mateo in her dissertation (2004) notes in detail the transformation of the relationship between the Ilocanos and Igorots from trading partners to enemies due to Spanish colonization in the region. Despite this grim assessment of the state of affairs between these two related groups, Scott, on the other hand, notes that 320 years of lowland commercial cooperation between lowland Pangasinan and Ilocos with the Igorots was not interrupted during the Spanish period:

In this island there is much gold among the natives in jewelry, and they trade it. There are a great many reports of gold mines, and because they say that the best are in the province of Ilocos. (Sande 1576: 88) Cordillera historian Willian Henry Scott (1970: 700) also notes that another Spanish Governor, Antonio de Morga, in his Sucesos de las Islas Filipinas (1609: 183), claims that Igorots mined the gold but it was the Ilocanos who refined and distributed it. In follow-up work, he concludes that the Igorot gold fields that were regularly referred to in

Lowland merchants travelled around buying up carnelian beads to sell them at a peso a piece; Igorot G-strings were woven on Ilocano looms in the eighteenth century as in the twentieth; Igorot miners could refresh 36

Documented Networks themselves with lowland basi and molasses cakes; and a slave trade in both directions continued into the American regime. (Scott 1970: 713)

coastal settlements peoples and even the foreign merchants themselves. Increasing inter-village alliances facilitated a smoother flow of products from upstream to downstream and vice versa. Bagamaspad and Hamada-Pawid (1985) explicate this phenomenon of building strong inter-group relations:

It is important to highlight the slave trade and to add to that some changes did occur in the upland–lowland relationships based on historical evidences provided by Scott (1971), like the increased instances where upland miners deceived their newly hispanicized trade partners on the coasts in their gold exchanges. In addition, by the eighteenth century the Igorots themselves were melting (refining) the gold into cakes because ‘ordinarily they sell it in sheets or ingots so there is opportunity for deception and alloying it with silver or red copper’ (Antolin 1789, translated by Scott 1971: 123). Scott, in citing Guillaume La Gentile (1779-1781), states that this illicit and deceptive trade for gold was participated in by lowland coastal friars and village leaders or alcaldes in exchange for silver dollars (Scott 1974: 141). In one account it appears that the Spanish Governor of Pangasinan owed some money to Igorot debtos in 1855 to the tune of 117,000 pesos, which caused complications in carrying out punitive expeditions against the Igorots debtors (1970: 713) (Canilao 2017d).

The trade routes brought the trader from Banget-Pico to Sabdang, present-day Sablan, also earlier called Disdis, to Galiano into Aringay then San Fernando or from Betdi, Tublay to Pico into Kafagway to Chuyo and down to Agoo through portions of present-day Marcos Highway. The practice of trade lasted days and developed still another dimension to the practice of arranged marriages, kaising where a trader maintained more than one family over a dispersed area. Such a relationship was both a formation of a nuclear family as well as the establishment of an alliance between the trader and his kin and the kin and his neighbors of his wife in settlements along the trade routes. (p. 77) [emphasis mine] Bagamaspad and Hamada-Pawid vividly demonstrate the advantage of this form of social networking when they argue that the gold miners and panners from more interior settlements became more dependent on the newly created elite and rich baknangs (also called kadangyans) amongst them in the highlands, whose wealth was built upon control of lowland trade items sourced from their well-placed kin on the lowlands and cattle grazing in their upland estancias (1985: 138). More importantly by the 1880s, these baknangs or kadangyans either controlled or owned the Cordillera iron, copper, and gold mines according to Hans Meyer (1975a: 126). The items they brought in from the lowlands in exchange for gold fueled a consumption economy among the Igorot gold producers (Bagamaspad and Hamada-Pawid 1985: 138) (Canilao 2017d).

3.4.3 Ethnohistorical Accounts of the Slave-for-Gold Exchange and Strategic Trade Alliance Formation Ibaloi oral tradition accounts of slave trading were recorded by Bagamaspad and Hamada-Pawid (1985): Slave trade introduced the baga-en into Ibaloy society. Literally meaning ‘one who is sent’, the baga-en concept is closely associated with that of Busul or enemy. For while earliest remembered slaves were spoils of intertribal warfare or purchases from head takers who were potential allies or enemies, slave trading soon became a part of the gold trade. (p. 130)

3.4.4 Overland, Riverine, and Maritime Trade Routes of the Angaqui Network

They cite Kabayan municipal history in terms of the barter value of slaves:

They argue that at the nadir of the slave-for-gold trade in the 1800s, an Ibaloi family possessed up to five slaves (ibid) (Canilao 2017d).

Historical records show that the Angaqui network trade was carried out through overland trails, river rafts, and finally ocean-crossing merchant sampans (see Azurin 1991, Mateo 2004). Indeed, in contrast to the other main drainage systems of Northwestern Luzon, the Grand Abra River is characterized as broad and open (Meyer 1975a: 108). This has ramifications in terms of riverine transport networks. Mateo (2004), for example, explains the logistics of the Igorot–Ilocano linkage through the geography of rivers:

Of direct relevance to the book is the observation that, although the upland gold miners/producers are traditionally assumed to be an exploited group and literally peripheral to controlling the trade with respect to the world systems theory (WST literature), there appears to be some strategic maneuverings on the part of the miners/producers themselves to ensure that their stakes in the trade are optimized when at the negotiation table with

Despite their contrasting geographical terrains, Ilocos and the Cordillera are geographically and ecologically linked. The three largest and most important rivers in the Ilocos region – the Amburayan, Abra, and Laoag Rivers – serve as a vital link between the lowland Ilocos and the upland Igorots since these rivers crisscross the Cordillera and Ilocos […] but it is the Abra River which has traditionally been the major highway linking Ilocos

[Slaves were] exchanged for gold, coins, animals and ceremonial blankets. A slave would be worth a kolebao, death blanket; a pig, molmol; or a piece of gold. These people would be most commonly between the ages of 6 to 16 years. (p. 137)

37

Remote Sensing the Margins of the Gold Trade Today these horses are quite ubiquitous in the Abra Valley. There is an account that describes a trip through the trails of this area astride a horse. This account is by Roy Barton in his chapter High Trails in Headhunterdom, which describes the trip coming from Cervantes settlement of the upper Abra River:

Sur and Cordillera. It boasts an overall length of fiftyfive miles from Lepanto on the thickly forested western slope of the Cordillera to the coast. From Lepanto it flows along a Northerly course via Agnet River near Lagangilang then moves to a southerly direction. Numerous tributaries flow into the Abra so that by the time it reaches the Abra valley, which is the only wide and fertile flatland in the province it becomes a river of notable size. It eventually empties into the South China Sea through the Banaoan gap, a V-shaped gorge near Santa, Ilocos Sur. (2004: 46) (emphasis mine)

We expected to reach Bontok that night. The rain grew heavier. The trail was deep with mud at places – dangerous. There was danger that our ponies, struggling, would flounder off the side of the trail, in which case, several seconds might be required to reach the earth again. (1930: 24)

The raft becomes the central mode of transportation:

Trampling action by horse hooves would have hastened the formation of troughs within the trading trails. Several archival photo plates from the Dean C. Worcester collection show some of these turn of the century trails somewhere in the Angaqui network (Canilao 2017d) (Figs. 3.16, 3.17).

During the rainy season, the Igorots loaded their cargoes on rafts or boats and went down to Abra and Amburayan Rivers to trade with the Ilocanos... Of the three rivers, the Abra River appears to be the most traversed route by the Igorots judging from the amount of references to it. It could also be because Abra River descends to Vigan, the capital of Ilocos. It appears to be the chief avenue of trade and communication prior to the construction of the Abra-Ilocos Sur highway in the twentieth century. Despite being narrow and rapidflowing, it could apparently accommodate freightage by bamboo rafts as far as forty eight kilometers between coast and the interior. (2004: 48-49) (emphasis mine)

3.4.5 Ethnohistorically and Archeologically Documented Settlements of the Angaqui Network At this juncture, I identify some of the relevant settlements and trails articulated within the Angaqui network (Fig. 3.18). It appears that the organization of coastal settlements in Ilocos had astounded the first Spaniards who entered the area in the sixteenth century. In general, when talking about the Southern Ilocos area which includes Purao, Tagudin, and Dumaquaque, Keesing argues:

Generally speaking, two types of raft traversed the three different sections of the Abra River. Deeper and wider parts in the lower and middle Abra were probably navigated by a longer and wider raft (Fig. 3.15) while the shallow, narrow but fast-flowing upper Abra that drains the gold rich areas of Northern Benguet, Upland Ilocos, and Western Mountain Province would probably have been more accessible using smaller and sleeker rafts. Mateo illustrates raft travels using the 1819 source by Francisco Alban OP:

It seems fair to assume from the 1591 and 1612 figures that the people along this section of the coast, later to be called Southern Ilocanos, number around the latter part of the sixteenth century about 15,000. By demographic perspective this number is unusually high density of settlement, and the Augustinians established more centers to serve it than was done in any other parts of Ilocos comparable in size. (1962: 99)

At certain points, the raft had to be dragged upstream by the locals with ropes while walking along the banks of the river, wading through shallow water and clinging to rocks or swimming and pushing the raft. It took fourteen hours to go upstream from Vigan to the Abra Valley, but only five hours going down river. (2004: 49)

The better layout of the towns in Northwestern Luzon, specifically the La Union–Ilocos coastline, has received much praise as illustrated in one passage by Spanish chronicler Ortega that talks about Salcedo’s conquest through Luzon when he attempted to circumnavigate Luzon:

Apparently multiple modes of transportation were needed to complete the upland–lowland travel:

They reported that the population there was large and that there were many good settlements close to one another and that they were better governed than in the other parts. (Ortega 1572: 257)

Tributary streams – Tineg to the Northeast and the Binongan, Malanas, Ba-ay, Bucloc, and Ikmin to the east and southeast – together with a series of pathways and horse trails linked the eastern flank of the Abra Valley to the interior Cordillera settlements of Lepanto Kalinga and Apayao. Besides the Abra River, trails crossed westward from the Abra Valley to the Narvacan coast and the coastal points. Similarly, trails descended from the north Abra into Ilocos Norte. Rivers and trails were the routes used by Igorot traders in carrying their goods to Ilocos. (2004: 49.

These are the coastal settlements, but as for the interior settlements, an expedition targeting Lepanto mines led by Don Antonio Hernandez provides a list of settlements that appear to be the bulking stations: On February 3, 1850 an expeditionary force of 70 troops and 250 cargadores led by an engineer Don Antonio Hernandez, together with Commander Manuel 38

Documented Networks

Figure 3.15. Larger bamboo raft in the lower Abra River, turn of the twentieth century (© The Field Museum, CSA 28583. Photographer: Fay-Cooper Cole) (Canilao 2017d).

Maria; however, belongs to the Abra network further north because, like Narvacan a few kilometers north of this settlement, these coastal settlements were destinations of traders who chose the overland route, mainly by horse and sometimes by trek from the Bangued area rather than the raft route which brings them to Vigan–Caoayan during the Early Historical Period. According to Perez, the growth of Santa Maria was seen side by side with the older village of Candon when the Commandacia Politico Militar (CPM) district of Lepanto was created in 1852; thus ‘Santa Maria and Candon acquired great importance as points of departure for the interior of the mountains’ (Perez 1988: 23). Keesing argues that by the nineteenth century Candon was recognized not only as a major trading center but also a base for mission work (Keesing 1962: 108). Mateo also states that:

Coballes, left Manila for the Ilocos coast. From Candon, Ilocos Sur they passed through Salcedo, Concepcion, Angaqui, and Cervantes and reached Mankayan on the 24th. (Bagamaspad Hamada-Pawid 1985: 172) Missions in Tiagan, Angaqui, Cayan, and Mankayan were established in 1869 and were an offshoot of the establishment of the Cantabro-Filipina of Mankayan (Bagamaspad and Hamada-Pawid 1985: 174) (Canilao 2017d). Three Spanish period maps (Figs. 3.19, 3.20, 3.21) available in the National Archives of the Philippines (NAP) show parts of the Angaqui network. Angaqui (Quirino) and Cervantes were crucial bulking villages in the historical period. Don Esteban Pennarubia’s map dated 1869 in Archivo Museo Naval, Organo de Historia y Cultura Naval in Madrid, Spain also shows the area in the late nineteenth century (Fig. 3.22). Perhaps worth emphasis is the centrality of Angaqui in this network, which appears to be serving the Minlaoi to Candon and Dumaquaque network, but also the Lepanto to Abra network, and Abra to Vigan network (Canilao 2017d).

Candon was another Igorot frequented town, but its location, halfway between Narvacan and Tagudin, and its distance from the Abra and Amburayan, would appear to have disincentives for traders. But Candon, which was created in 1591 [hispanicized], had a long history of intimate ties with the Igorots of Bontoc (in present-day Mountain Province), that were established by the intermarriage of a scion of its most prominent baknang family, the Abayas, with the daughter of the founder of Tetepan, a village in Western Bontoc. (2004: 217) (emphasis mine)

Dumaquaque and Candon are the main coastal settlements but it should be noted that in the two National Archives of the Philippines maps (again see Figs. 3.19, 3.20), Santa Maria is also shown as a critical settlement. Santa 39

Remote Sensing the Margins of the Gold Trade

Figure 3.16. Guard structure along the trail somewhere in the Angaqui network with Bontok warriors, turn of the twentieth century (© The Field Museum, CSA 18562. Photographer: Dean C. Worcester) (Canilao 2017d).

Figure 3.17. Trellis along the trail somewhere in the Angaqui network with Bulol posts, turn of the twentieth century (© The Field Museum, CSA 18525. Photographer: Dean C. Worcester) (Canilao 2017d).

40

Documented Networks

Figure 3.18. Foreground the Town of Angaqui, background Tirad Pass/Tila Peak Locality (© The Field Museum, CSA 22411. Photographer: Dean C. Worcester) (Canilao 2017d).

Dumaquaque is present-day Santa Lucia based on more recent secondary historical researches (Newson 2009: 181, Mateo 2004: 43. 67). The earlier work of Keesing (1962: 98), however, identifies this as Tagudin. Early historian Isabelo de los Reyes, however, argues that the gold from this settlement was the target of merchant ships coming from China and Japan (1890, trans. 2014: 140). Going back to Keesing, he argues that Manuel Buzeta’s 185051 census shows that gold nuggets were panned in the waterways of Santa Lucia (1962: 107). The said waterway is in fact the Buaya River located south of Santa Lucia whose headwaters are on the westerly slope of the three mounts, Mount Monserat, Mount Ampayao, and Mount Tila/Tirad (Mount Estileta). This mountain chain forms the natural barrier that separates the Abra basin from the coast (Canilao 2017d).

which is based on Martin de Goiti’s second expedition to Ilocos in 1572-1573, show that from Dumaquaque, 385 tahels of gold was exacted, whereas from Candon it was 120.5 tahels (2009: 180). For reference, one tahel of gold is equivalent to 39.537 grams of gold (Peralta-Imson annotation in Reyes 2007: 57). Therefore, Dumaquaque village gave a tribute of 15,221.745 grams or 15.2 kilos or 33.5 pounds of gold and Candon a tribute of 4,764.209 grams or 4.7 kilos or 10.3 pounds of gold. Patiacan and Mount Quinali appear to be the source of gold for these mines but as mentioned above, the Lepanto gold district also became articulated through the Cervantes settlement (Canilao 2017d). A secondary burial tradition was recorded in Patiacan during the second season of the Ilocos Sur Archeology Project (Fig. 3.23). Some of these coffins were as old as 200 years and as late as 100 years, since present-day residents can recall the people interred as ancestors two to three generations back in time (Canilao 2015). Patiacan is literally translated as ‘viewing or spotting.’ Indeed, a GIS viewshed analysis from the Minlaoi hilltop nearby with watch tower vestiges shows it has good visibility of the Abra River as well as Angaqui. Community history of Patiacan was recorded in 1911 and it appears that the contemporary-day settlers in Patiacan believe that ‘the

Juan de Salcedo’s 1572 conquest includes an account that he captured a village chief called Silata in Dumaquaque. Silata was able to bring a delegation to meet Salcedo and to offer around 100 ounces or 2,834.95 grams (2.8 kilos or 6.25 pounds) of gold (Reyes 1890, trans. 2014: 141). In terms of population, Keesing cites a 1591 Account of Encomiendas stating that Dumaquaque and Candon coastal settlements each individually had a population of 3600 with 900 tributes in gold (1962: 98). Newson’s data, 41

Remote Sensing the Margins of the Gold Trade

Figure 3.19. Unknown Cartographer(1794). Ydea Aproximada del Teritorio entre Cagayan e Ylocos (Courtesy of the National Archives of the Philippines) (Canilao 2017d).

42

Documented Networks

Figure 3.20. Tauguaz, Rafael (Cartographer). (1891). Croquis del Distrito Politico Militar de Tiagan y Rancherias de Ilocos Sur (Courtesy of the National Archives of the Philippines) (Canilao 2017d).

43

Remote Sensing the Margins of the Gold Trade

Figure 3.21. Unknown Cartographer. (undated) Croquis de la comprension de Balidbid y parte de la de Sta. Lucia (Courtesy of the National Archives of the Philippines).

44

Documented Networks

Figure 3.22. Pennarubia, Don Esteban (commissioned map). (1868). Croquis de la provincia de Abra (España. Ministerio de Defensa Archivo del Museo Naval 78–27. Filipinas. Croquis de la Provincia de Abra. 1868 al 1874) (Canilao 2017d).

45

Remote Sensing the Margins of the Gold Trade ancestors of Patiacung barrio trace back to 1793, when they went to a place called Balaoa, west of Santa Lucia in Ilocos Sur’ but were forced to migrate further inland because of the Christians in that town (Keesing 1962: 113). Balaoa may just have a different spelling but must have been Buwaya, the river south of Santa Lucia that is panned for gold. The migration account may have been part of the remontados or those who fled to mountainous interiors to escape proselytizing efforts in the lowlands or who may have been drawn in by the prospect of small-scale mining in Mount Quinali (Canilao 2017d).

from those elevated ridges two beautiful, enchanting and magnificent panoramas are to be seen, both most delightful. The first is the great precipice that frightens those not accustomed to hiking in such heights and causes vertigo, a precipice which it is necessary to descend to the village of Angaqui. After resting, the inexperienced traveler contemplates the picturesque villages scattered along the ridges of the Cordillera spurs; and in the bottom of the valley, the wide Abra bathing various little islands [river islands] and there in the distance a labyrinth of higher and higher ridges and ranges sometimes covered with undulating clouds.

The Minlaoi site (Fig. 3.24) lay contiguous to the settlement of Patiacan and in 2012 both test pit excavations and surface surveys were carried out at this ancient mining site. Several earthenware sherds were recovered together with some tradeware sherds on the surface, and within crevices of huge limestone boulders towards the west of the site. Sherds of fluted earthenware vessels were also encountered in test pit excavations. Foundation stones of watch towers were also recorded at the top of the hill, overlooking Abra River valley (Fig. 3.25, compare with Fig. 3.26). Traces of ancient mountain terracing is also evident on the slopes. Lode mining slides associated with both small-scale and more industrial scale mining are also seen using the high-resolution and multispectral satellite imagery (Fig. 3.27).

Behind us we have left the other panorama which offers a more delightful view, if you like, for from there the wide tempestuous China Sea is spread-out, washing the beaches of both Ilocos and La Union, so far is the view from those eminent heights on clear days. Along that coast and on the mountain slopes, the civilised villages are scattered with their churches and convents, the plentiful irrigation ditches shining as if they were ribbons of silver fertilizing once sterile fields, all the zealous work of the regular clergy who not only planned them but also contributed their funds to their construction for the benefit of their parishioners. Take the town of Santa Lucia, which, grateful to their priest Fray Exequiel Lanzagorta, who gave the town

Minlaoi and the greater Patiacan settlements are parts of Mount Quinali, which literally translates to ‘dug up/mined.’ Accounts also show that the settlement is not Minlaoi but Quinali (Perez 1902: 105). The Pennarubia map of 1868 also marks Quinali at the banks of the Balasian River (Rio Quinali) in the area of Patiacan–Minlaoi, northeast of Angaqui. To explain this anomaly, Perez argues that the location has been moving all the time because of conflict between the Bago inhabitants and the Bontocs further east (ibid) (Canilao 2017d). Quirino was designated as the new name of Angaqui in 1964 through Republic Act 4035. Historical sources indicate that this bulking station has direct trading relations with the coastal settlements of Santa Lucia and Candon. Father Angel Perez, for instance, gave a succinct account of this settlement and as a side note Perez’s work has been commended as one of the important ethnographic works on Cordillera by anthropologist June Prill-Brett (1978: 48). Perez was an Augustinian priest assigned in Kayan from 1886 until the close of the Spanish period in the Philippines (Prill-Brett 1978: 48). He provides a description of the portion of the trail, its rest stop, and the picturesque view it offers of Angaqui to the east and the coastal towns including Dumaquaque/Santa Lucia, Candon, Santa Maria, and Narvacan to the west: When entering the district, either by Tila Pass, or Tirad as the real Ilocanos more correctly say since it means a long sharp point, and the said mountain peaks with that shape, or through that of Tobalina,

Figure 3.23. Artist illustration of the Patiacan secondary burial coffin (Eduardo Bersamira, Commissioned Work, Ilocos Sur archeology Project II, 2012) (Canilao 2017d).

46

Documented Networks

Figure 3.24. Archeological features in Minlaoi open site (Using WV2 soils and constructions combination) (Courtesy of the DigitalGlobe Foundation) (Canilao 2017d).

abundant irrigation for all their fields (1810), has been celebrating his death anniversary every year with a sung mass. (1902: 87)

Angaqui is the capital of the group and the mission and is located in a pleasant and agreeable place on the slopes or first approaches to Tila and Tobalina… Its many creeks and cascades water spacious fields in the district, from which not only is palay harvested for the citizens, but a surplus, and they export it to towns of Santa Maria and Candon in exchange for the clot[h] which the Ilocanos bring up to sell from the towns mentioned. (p.89) [emphasis mine]

In connection to this account, Hans Meyer traveling in 1880 also describes this rest stop at the trail to Tila peak, adding that this towers behind (west of) Angaqui, and was at that time the only direct route from Ilocos Sur to Lepanto: There the whole coastal province of Ilocos Sur was spread out below us. Beyond the nearest low mountain range, the green plain laying gleaning with the China Sea shimmering beyond it. Thalatta, thalatta! To the north the eye had an unobstructed view all the way to the Province of Abra, and to the south, Cape Namacpagan was darkening in the twilight. A fresh wind was blowing up from the sea. We rested, the porters cooked their meal of rice, and we started the descent with new vigor. In three hours we were down in Lingey. (1975b: 76)

The location of Angaqui was triangulated to Legleg as it appears in a 1:50000 National Mapping and Resource Information Agency topographic map (1989). This is also the present-day location of the Poblacion. Legleg is the term that pertains to gold-rich river sludge/silt/clay or slime: Oral accounts from Mankayan tell that there once was a woman in Panat who went to the river to take a bath. She used Legleg, a type of clay to soap her hair. After bathing she discovered particles of gold clinging to her hair. The woman went home to tell her neighbors about her discovery. The neighbors, hearing this, went to find the source of the gold ingrained in the clay. Thus started the knowledge of gold and its subsequent mining through the labon. (Bagamaspad and HamadaPawid 1985: 67)

As will be shown in the next chapter that remote sensing through WorldView-2 satellite imagery shows trail segments that stand out particularly because of its troughs and associated secondary vegetation. Perez (1902) talks about the central location of Angaqui: 47

Remote Sensing the Margins of the Gold Trade

Figure 3.25. Watch tower foundation stones on the crest of Minlaoi Hill (Ilocos Sur archeology Project II, 2012) (Canilao 2017d).

In a redrafted map published in Bagamaspad and HamadaPawid (1985: 9) based on the 1899 map of the 8th Army Corps of Engineers and an 1890 map of Comandancia Politiko Militar (CPM) Amburayan, Angaqui also appears due east of Gregorio del Pillar or Concepcion, with Quinali or Patiacan located east, southeast (1985: 9)(Canilao 2017d).

Pawid 1985: 76, 77, 78, 83). Bagamaspad and HamadaPawid also state that Cervantes was central between Lepanto, Mankayan, and Candon (1985: 76). Cervantes formerly Mantamang also became more prominent as a Christian town by 1870 because of the operations of the Cantabro-Filipina Mankayan to its south (Bagamaspad and Hamada-Pawid 1985 174, 177). A legend about Suyoc gold recorded by early American gold prospector Laurence Wilson in 1932 states:

At this point, we delve more into the settlement of Cervantes, which stands out in the Benguet oral tradition accounts as part of the string of pre-generational settlements occupied by the Ibaloi in a north to south movement:

The gold in Suyoc, as in other districts, first stood in the shape of a great high tree which reached the heavens. When it fell it was buried and some of the branches (veins) have been found and much gold taken from them. The main trunk has not been found and remain buried very deep[…] Gold was first discovered many generations ago at Suyoc by a band of boys from near Cervantes who were out on a fishing trip. (p. 29)

Ampontoc, Dec-can, Panat and Bagongan were the early settlement areas. Due to an early epidemic, the people in these early settlements moved out from Panat to Lap-angan, Palasa-an, Payen, Esmay and Bakun; from Deccan to Sesecan and Baguyos; from Ampontoc to Comillas, Cervantes and Comay. Others returned to Namiligan. (Bagamaspad and Hamada Pawid 1985: 28-30)

Cervantes served as the capital of the newly combined Commandancia Politiko Militar (CPM) of Tiagan and Lepanto (Keesing 1962: 110). According to Perez this town is strategic and central to other settlements in Northwestern Luzon:

Oral traditions commonly describe Cervantes as a trading center dating to as early as the 1600s. It is a conduit to Mankayan and its gold deposits (Bagamaspad and Hamada 48

Documented Networks 3.5 The Abra Network as the Epitome of a More Permanent Market System in Northwestern Luzon Culture contact between the Philippine archipelago and its Southeast Asian neighbors like China, Japan, India, and as far as Persia may have started in the ninth to tenth century and intensified between the twelfth and fifteenth centuries (Dizon 2004: 1). Products including interior raw materials most notably gold was bartered in exchange for exotic prestige goods (Junker 1999). Berthold Laufer pushes back this date even earlier to the Sung dynasty (tenth to thirteenth century) (Laufer in Cole 1922: 17) (Canilao 2018a). Spanish Governor General Antonio de Morga argues that Vigan is one of the major port locations in the Abra Network (Fig. 3.28. 3.29), On the coasts of Pangasinan, Ylocos, and Cagayan there are some ports and bars, where ships can enter and remain, such as the harbor in Marihuma, the port El frayle, that of Bolinao, the bar of Pangasinan, that of [V]igan, the bar of Camalayuga, at the mouth of the Tajo River (which goes up two leguas to the chief settlement of Cagayan) besides other rivers, bars, harbors, and shelters of less account for smaller vessels throughout the coast of this island. (Morga 1609: 109) Felix Keesing on the other hand argues that at its nadir, Vigan became quite prosperous because of the gold trade (1962: 126); this, however, steadily declined at Spanish contact. In fact a Spanish chronicler laments,

Figure 3.26. Watch tower somehere in the Angaqui network, turn of the twentieth century (© The Field Museum, CSA 22186. Photographer: Dean C. Worcester) (Canilao 2017d).

The town of Fernandina [old name of Vigan] in the province of Ilocos has proved to be so unhealthy a region, that from being the richest town in these islands, it has now only a few inhabitants with no organized cabildo or government. (Unsigned 15821583: 203)

The main road leaves the town from Angaqui in a northerly direction; that which heads for the district of Benguet takes a southerly route via Mancayan; that for Bontoc goes east; finally, that of Amburayan to the west (1902: 108)

Francisco de Sande describes the seafaring tradition of the Tingguian/Itneg which allowed this Philippine group to cross the South China Sea:

Because of this strategic location, Perez describes the populations as heterogenous, a melting pot of not only Ilocanos, Abrenos, and Igorots but even Chinese (p.109). Later in the eighteenth century Cervantes became an intensive center for cattle trade. Hans Meyer writing in the 1880s describes the trail connecting Angaqui with Cervantes as:

They do not understand any kind of work, unless it be to do something actually necessary – such as to build their houses, which are made of stakes after their fashion; to fish, according to their method; to row, and perform the duties of the sailors; and to cultivate the land. The mountaineers make iron lance points, daggers, and certain small tools used in transplanting rice. They are very anxious to possess artillery, of which they cast a little, although but poorly. (Sande 1576: 69) (author’s emphasis)

[…] heavily travelled. Igorots who had purchased cattle and provisions from the market in the coastal province of Ilocos Sur moving homeward, Chinese taking caravans of trade goods into the mountains, and mail runners and patrols of the Guardia Civil, all were encountered there often. (1975b: 75)

One important information is the fact that Tingguian/ Itnegs are good rowers. Rowing implies sea navigation since riverine navigation (i.e., in the Abra River) is carried out on rafts and involves shoving forward using bamboo

Using the information above, indeed Cervantes appears to have been increasingly important as Spanish annexation marched into the area. 49

Remote Sensing the Margins of the Gold Trade poles as well as ropes for pulling alongside the river banks (Fig. 3.30). In another account, Antonio de Morga talks about the sea-crossing biray that was the vessel rowed by Tingguian/Itnegs and their Ilocano kin:

jar owned by the Tingguian of Ilocos Norte. A small jar at San Quintin, Abra was said to be the child of this union and partook of many qualities of their parents. (Cole 1912: 12)

… they time their rowing to the accompaniment of some who sing in their language refrains by which they understand whether to hasten or retard their rowing. (Morga 1609: 82)

Though this jar was sourced from the coast via trade from an early period and handed down to the owner Cabildo of Domayco as a heirloom, a putative oral history already takes root: Magsawi, my jar, when it was not yet broken talked softly, but now its lines are broken, and the low tones are insufficient for us to understand. The jar was not made where the Chinese are, but belongs to the spirits of Kabonian, because my father and grandfather, from whom I inherited it, said that in the first times they (the Tingguian) hunted Magsawi on the mountains and the wooded hills. My ancestors thought that their dog has brought a deer to bay (which he was catching), and they hurried to assist it. They saw the jar and tried to catch it but were unable; sometimes it disappeared, sometimes it appeared again to the wooded hill on their way to town. Then they heard a voice speaking words which they understood, but they could see no man. The words it spoke were: ‘you secure a pig, a sow without young, and take its blood, so that you may catch the jar which your dog pursued.’ They obeyed and went to secure the blood. The dog again brought to bay the jar which belonged to Kabonian (a spirit). They plainly saw the jar go through a hole in the rock which is a cave, and there it was cornered so that they captured the jar which is Magsawi, which I inherited. (Cabildo of Domayco interviewed by Fay Cooper Cole 1912: 12-13)

He talks about the function of these rowing songs: A common device among barbarous and semi-civilized peoples [sic] and even among boatmen in general. These songs often contain many interesting and important bits of history, as well as of legendary lore. (Morga 1609: 82) It should be emphasized that the Tingguian/Itneg who migrated towards the Abra River valley in the interior indeed share with their Ilocano kin traditions for seafaring (Canilao 2018a). 3.5.1 Tingguian/Itneg of Abra and Their Accounts on Early Gold It is argued that gold was both a backbone to trade as well as an aesthetic marker of status among the Tingguian/ Itneg. Felix Keesing argues that the Tingguian/Itneg exhibits more complexity compared to its neighbors in the interiors: Leadership is provided by wealthy headmen or aristocrats. Ceremonial wealth in the form of Chinese jars, copper gongs, and precious beads vies with rice fields and livestock in defining the status of being rich. Ritual and festival life is highly elaborated, with female shamans or mediums taking important responsi­ bilities in religious affairs. (1962: 121) (author’s emphasis)

This jar’s oral history has become well established among its possessors that the jar is perceived to have ‘never been made by the Chinese but by the spirits’ (Laufer postscript in Cole 1912: 18). Porcelain plates also play a symbolic role among the Tingguian/Itneg. Cole was able to document the summoning of spirits through the plates by mediums (1912: 15). These plates, however, are never sold during the lifetime of the medium, but only upon death (1912: 15). Cole describes how the porcelain is used to induce a trance:

The prevalence of Chinese ceramics among the Tingguian/ Itneg was noted by Cole who argues that sampans from China as well as birays from Luzon may have regularly crisscrossed the West Philippine Sea/South China Sea to carry out trade (Cole 1912). He further argues that status was a function of the accumulation of jars (1912: 12) (Canilao 2018a). Having been handed down through generations, Cole adds that these exotic prestige goods may have taken life (agency) of their own as ‘they began to gather to themselves stories of wondrous origins and deed, until to-day certain jars have reputations which extend far beyond the limits of the tribes by which they may be owned’ (Cole 1912: 12). This is to the point that each jar has taken a name (p. 13) like Magsawi, the jar from the Abra Tingguian/Itneg:

When about to call a spirit into her body, the medium sets herself in front of the spirit mat, and covering her face with her hands, she trembles violently, meanwhile chanting or wailing songs in which she bids the spirits to come and possess her. From time to time she pauses and holding a plate on the fingertips of her left hand, she strikes it with a string of sea shells or a bit of lead, in order that the bell-like sound may attract attention of the spirits. Suddenly a spirit takes possession of her body and then as a human the superior being talks with the mortals. (Cole 1912: 15)

It was credited with the ability to talk; sometimes went on long journeys by itself; and was married to a female

Trade ceramics function not only as status markers and spiritual vessels but according to Cole they were also 50

Documented Networks

Figure 3.28. Note Abra trade network in Northwestern Luzon (ASTER GDEM is a product of METI and NASA).

given as covenants to settle inter-tribal and inter-village disputes and feuds, at times even used to compensate the kin of a person whose head was taken by a hunting party (1912: 15). For instance, Cole mentioned that a village was able to reach a peace settlement with another village by offering eleven jars in exchange for the eleven

heads taken during the inter-tribal war (ibid) (Canilao 2018a). It can be argued that the volume of trade ceramics that entered the Abra network may hold as testament to the amount of gold traded in exchange; it should be noted that 51

Remote Sensing the Margins of the Gold Trade

Figure 3.29. Composite photo of Abra Network with Abra River in foreground. Top left, northerly view from San Quintin with Abra River on foreground, Langiden with Bulagao mountains on background. Top right, northeasterly view from Bangued with Abra River on foreground, Dolores-San Juan-La-Paz-Danglas area in the background. Bottom left, Southeasterly View from Bangued with Abra River on foreground, Tayum on the background, Lagang-ilang-Salapaddan mountains over the horizon. Bottom right, easterly view from Tayum with Lagang-ilang-Licuan-Baay-Lacub-Malibcong area on the background (photos by author 2018) (Canilao 2018a).

gold was also a central metal in the social and spiritual lives of the Tingguian/Itneg. Some primary and secondary historical works document how the Tingguian/Itneg used gold to display status. Paul Frost Gironiere was a French traveler and talks about this vividly:

The use of gold and jewels seems to have been common in the old times; the latter are seldom seen in the district today, but the use of bits of gold in the various ceremonies is still common, while earrings of gold and copper are among the most prized possession of the women. Placer mining is well known to the Igorots of the south [Angaqui and Lepanto], who melt and cast the metal into various ornaments. (1915: 21)

Their heads were ornamented with pearls, coral beads, and pieces of gold, twisted among their hair; the upper parts of their hands were painted in blue; their wrists adorned with interwoven bracelets, spangled with glass beads – these bracelets reached the elbow, and formed a kind of half plaited sleeve. (Gironiere 1854: 108-109) (author’s emphasis)

The mines in the interior, including the Abra mines and also the Minlaoi, Quinali, and Lepanto mines, were the source of the gold (Cole 1912: 241). In another work (Cole 1915) describes how feasting has become intertwined with gold. When invitations to ceremonies or celebrations are sent out to guests, they are in the form of betelnuts covered with gold oiled over (1915: 18-19, 24, 31), a form of magic. According to Cole this is still seen today when gold is sent out as a gift to a friend in another town as a form of invitation (1915: 24) (Canilao 2018a).

Cole adds that gold headbands were worn by men to keep their long hairs in place, sometimes each hair is also adorned with golden beads, and ears and finger rings were also made of gold (Cole 1915: 9). Even spirits were said to be surrounded with articles of gold apart from the ubiquitous trade ceramics in their spirit houses (ibid). Cole argues that:

Cole discusses in detail the development of a prestige 52

Documented Networks

Figure 3.30. Composite archival photo of raft plying the river: top photo shows navigation upstream Abra River via pulling ropes (© The Field Museum, CSA 24619. Photographer: D.C. Worcester), bottom photo shows navigation downstream using poles-shooting the rapids (© The Field Museum, CSA 24697. Photographer: D.C. Worcester) (Canilao 2018a).

53

Remote Sensing the Margins of the Gold Trade 3.5.2 Coastal Port Settlements of Santa Maria, Narvacan, Vigan/Caoayan, Magsingal, CabugaoNagsingcaoan, and Calanutian

economy as gold was traded overseas for porcelain jars from China and Annam (Vietnam): It is evident that outside influences of great importance were introduced at a period later than the time when the Chinese first began to trade along the coasts of the Philippines for the prised jars, which play such an important role in mythology, are not to be identified as those of native make but are ancient Chinese vessels dating back at least to the fourteenth and perhaps even to the tenth century. (Cole 1915: 31)[author’s emphasis]

In contrast to the four other regions of gold trade, the Abra network seems to have supported a more permanent market encounter in the coast specifically at the Vigan– Caoayan–Bantay area. This may have been the case as afforded by the Abra River in facilitating bigger bulk exchanges between coast and the bulking stations and the interior gold mining settlements of the Angaqui and Lepanto networks by virtue of its wide and deep rivers. Settlements upriver in Lagang-ilang, Bangued-Tayum, and Bucao appear to have been cogs in a wheel that facilitated higher bulk exchanges towards the coast. Larger riverine rafts plying the lower and middle Abra facilitated the riverine transport towards Vigan–Caoayan–Bantay. The biray, which is a local version of the vessel called sampan, may have also sailed regularly to Dai Viet and Southern China (Fig. 3.32 and refer to Fig. 3.14) (Canilao 2018a).

The Tingguian/Itneg trade network may have spanned as far as the Indian Ocean with agate and carnelian beads being exchanged for gold. Cole states that the Tingguian/ Itneg ‘glass, porcelain, and agate beads, which are second only to the jars in importance, are exceedingly old’ (Cole 1915: 31) (Canilao 2018a). Cole also narrates in detail how gold is intertwined with the Tingguian/Itneg spiritual world. In a Tingguian/Itneg ritual ceremony called Ngorong-or for a person seriously ill with stomach trouble gold is mixed with pig liver and brain and buried beside the center post of the dwelling (Cole 1912: 326). In another ritual gold together with agate is placed around the neck and legs of a sacrificial pig (Cole 1912: 351). According to Cole, when the Tingguian/ Itneg reenact good spirits Iwaginan and Gimbagon, they hold in their hands pieces of gold (356). Cole concludes that:

Despite the existence of this more formal market system which probably was at its zenith in the sixteenth century, alternative smuggling routes and trails that facilitated shipment and movement of gold in lesser amounts still existed. This may be the case with the trails leading to Santa Maria, Narvaca, Cabugao–San Juan, Magsingal, and Sinait/Calanutian from the Abra interiors (Canilao 2018a). Newson summarized in a table the tribute exacted by Juan de Salcedo and Martin de Goiti in their third expedition to Luzon in 1572 (2009: 182). The table lists the number of houses, number of villages, and number of tahels of gold exacted among other pertinent information. Here the valley of Vigan, which probably includes in the cluster Caoayan and Baluarte area, has 12 villages, 1,015 houses, with 615 tahels of gold. Bantay and Bantaguey were listed with 3 villages, 400 houses, and 136 taels of gold (ibid).

for the most part, the life, customs, and beliefs which appear in our reconstruction of ‘the first times’ agree closely with present conditions; certain things which seem formerly of prime importance – such as the sending of a betel-nut covered with gold to invited guests to a festival or ceremony – appear to have their echo in present conditions. The betel nut which played such a momentous part in the old times still holds its place in the rituals of the many ceremonies, although it is not now much used in daily life. (1915: 31) According to Salvador-Amores, photo-documentation serves as proxy to visual or written records of the indigenous groups, which are [in some instances] unavailable (2016: 56). Cole and Worcester’s archival photos indeed show some aspects of gold working and use as well as the ceremonial and spiritual importance of ceramics/jars for the Tingguian/Itneg. Photos of smithies (Fig. 3.31) are found in these photographic collections.

The settlements of Luzondan, Calanutian or Calanutuan, Magsingal, and Sanguian has 14 to 15 villages, 1005 houses, and 341 taels of gold. Sinay or Sinait has 9 to 10 villages with 950 houses and 535 taels of gold exacted. For reference, the Spanish–Philippines tahel is equivalent to 39.537 grams (see Reyes 2007: 57), placing the Sinait tribute at 21,152.295 grams of gold (21.15 kilos of gold). Felix Keesing, on the other hand, presents 1591 figures that states that Bantay and Bantaguey had 1000 tributes from 4000 people, hen Vigan at 800 tributes from 3200 people (1962: 123-124) (Canilao 2018a).

The components of these smithies include a piston blower, various cobbles for crushing ore or hammering, as well as crucibles and other clay conduits. These implements are used not only in iron working but also gold working (Canilao 2018a).

Owing to the richness and fertility of Sinait, Juan de Salcedo made Sinait one of his encomiendas (Foronda and Foronda 1972: 30). The term sinnait is literally translated as ‘to challenge one another’ (1972: 63) (Canilao 2018a). 54

Documented Networks

Figure 3.31. Composite archival photos of smithies somewhere in the Abra network: note bamboo piston forge and other implements (top photo © The Field Museum, CSA 29144. Photographer: Fay-Cooper Cole; bottom photo © The Field Museum, CSA 21979. Photographer: D.C. Worcester) (Canilao 2018a).

55

Remote Sensing the Margins of the Gold Trade Sinait was part of a cluster of settlements that included Badoc to the north and Calanutian to its south. Calanutian is one of the early villages that figures in the oral tradition Biag ni Lam-ang. Calanutian is said to be the hometown of Ines Kanoyan, who is the wife of Lam-ang (Yabes 1958: 309). At contact with the Juan de Salcedo and Martin de Goiti expedition in 1574 there were 950 houses in Sinait, 100 of which were in Calanutian (Canilao 2018a).

and grave provisioning were also recovered from the site (Fig. 3.35) (Canilao 2018a). A 1591 Spanish holdings document states that Cabugao– San Juan was a major source for tributes (Keesing 1962: 123-124) and that though these are separate towns in contemporary times, it appears that they may have been part of another cluster of settlements that included the Salomague port (Keesing 1962: 133), the Nagsingcaoan gateway location, and the Refaro ijang (Canilao 2018a, 2018b).

It appears that Sinait/Calanutian stands out among the competitors of Vigan, based on the amount of gold exacted at Spanish contact. The amount of gold available in Calanutian was at a level comparable to Vigan. This anomaly may be explained by the possibility that this particular settlement was getting their steady supply of gold through the overland smuggling trails (Canilao 2018a).

The Cabugao trail that crosses through the Cordillera Mountains all the way to the Cagayan valley may have terminated at Nagsingcaoan which served as its gateway location (Figs. 3.36, 3.37). This was the conduit for interior products moving to the Sinait, Cabugao, San Juan, and Magsingal area since Vigan has a permanent control of the Abra River. The pre-sixteenth century wealth and prosperity of Calanutian may be attributed to interior products passing through the gateway (Canilao 2018a).

The Calanutian site (Figs. 3.33, 3.34) has been dated to as early as the twelfth century. There are whole pieces as well as sherds of Vietnamese tradeware ceramics dating to the fifteenth century; twelfth century Chinese coins were also excavated. Some carnelian and Asian beads, various Asian ceramics, and gold in the context of dental modification

Archeological transect surveys were carried out in 2012 at a hill lying south of the Barangay center adjacent to the

Figure 3.32. Pandan in Vigan-Caoayan arrow shows where a biray vessel (based on local oral histories, the boat may be between 80 to 150 years old) is buried under brackish sediments (Using WV3 bathymetry combination, green-blue-coastal blue) (DigitalGlobe Foundation) (Canilao 2018a).

56

Documented Networks Bucao, and Lagang-ilang are examples of bulking station that have full interface in the gold trade plying the Abra River. While the Tineg settlement farther in the interior is regarded as the origin place of the Itneg group (itneg means ‘of Tineg’), some sources were found that indicate this settlement served more as a jump-off point for trade and communication linking the Cagayan valley to the east. It is the starting point for a route cutting through the Cordillera Mountains, specifically linking up with the settlement of Malague, which is also called Malaueg which is presentday Rizal in Cagayan Province). Tineg is also articulated to this route because of it is endowed with gold deposits based on the georeferenced mineral resources map.

present cemetery area. The survey recovered some surface finds including earthenware and trade ceramic sherds. The transect survey at Sitio Naguilian at the entry point to the Cordillera trail also revealed some surface finds. Naguilian literally translates as ‘a place where the village was established.’ This site is adjacent to the Cabugao river and serves as the gate towards the Abra mountains to the east. A transect survey was conducted with some earthenware and trade ceramic sherds surface finds being collected at the site (Canilao 2018a). The Nagsincaoan Kakadiran Reba Elementary School became the target for test pit excavation because of surface finds, including a pipe and some ceramic sherds. The hill southeast of the school site also had a cranial fragment on the surface. Three 1 x 1 meter test pits were excavated up to 1.5 to 2 meters below surface to investigate the layers underneath the surface artifacts. No artifacts were excavated in the underlying sediments, indicating that the surface artifacts may have been in a disturbed context (perhaps from nearby areas) (Canilao 2018a).

3.6 Chapter Summary This chapter has shown that the settlements that were delineated in the book as significant to the development of gold production as a component of burgeoning coastal– inland trade were identified through a triangulation of oral tradition, written historical, and archeological data. The chapter is also crucial in laying out the characteristics that determine whether a network features a more evanescent market or a more permanent market. The former will have coastal surveilling capacity that allows ship spotting, enabling the gold miners to participate in an evanescent market encounter. On the other hand, the latter will not feature coastal surveilling but will be tied to a riverine system that facilitates a bigger bulk of gold and other goods to be transported following the logic of economies of scale. It is further argued that the vivid accounts in this chapter are crucial to understanding who occupies these bulking villages. As argued in the book, a trade diaspora at the bulking stations involving mining groups correlates to a more evanescent market-based situation, whereas a trade diaspora from the coastal villages feature a more permanent market situation. The agency of the gold mining producers appears to recede as they begin being trapped within a web of trade intermediaries and they increasingly find themselves indebted (utang na loob) to these coastal middlepersons. Their agency literally vanishes as they find themselves being traded as slaves alongside the luxury items from overseas that they fancied in the first place. With this environmental, historical, and material background of the region’s historical engagement in gold mining and upland–coastal networks of exchange, the next chapter proceeds to a discussion of GIS and Remote Sensing methods that were used in the book.

The Valyador site in Refaro, San Juan (Fig. 3.38, 3.39) seems to be associated with the Cabugao site and appears to be an Ilian or ijang or hillfort refuge/defense. Scatters of earthenware and trade ceramic sherds were documented during an archeological transect survey in 2011. To follow up in this preliminary result, three 1 x 1 meter test pits were excavated the following season but with no buried artifacts (Canilao 2018a, 2018b). 3.5.3 Bangued, Tayum, Bucao, Lagang-ilang Bulking Stations and the Tineg Jump-off Point to Cagayan Valley in the East Several Spanish period maps provide the locations of settlements in Abra with some Spanish period maps showing trails connecting them (Figs. 3.40, 3.41). In this book some GIS georeferencing methods were used to manipulate the maps to improve its scalar accuracy. Abra, the valley itself, is the convergence point of different river systems including the Tineg, Danglas (Suut), Parsuguan, Malanas, Baay, and the massive Abra River that drains the boundaries of Ilocos Sur, Abra, and Mountain Province. The ground is quite fertile owing to this drainage system, in stark contrast to the hills of the province that are dry to arid because they are located on the leeward side of the Bulagao chain of mountains, which forms a north–south wall and receives little precipitation. The fertile river valleys support wet rice agriculture, but more importantly in the context of the gold trade, the river systems facilitated bulk exchanges upriver and downriver. The rivers are also important because these drain some lode deposits in the interior, thus making placer mining possible (Canilao 2018a). As expounded in detail in the next chapter, the Abra network bulking stations differ from the four other networks because the settlements of Bangued/Tayum, 57

Remote Sensing the Margins of the Gold Trade

Figure 3.33. Arrow showing Calanutian site, which was excavated during the Ilocos Sur Archeological Project of 2011 and 2012 (Using WV3 soils and construction combination, red-blue-yellow) (DigitalGlobe Foundation) (Canilao 2018a).

Figure 3.34. Calanutian Site Map (Eduardo Bersamira, Commissioned Work, Ilocos Sur Archeology Project II, 2012) (Canilao 2018a).

58

Documented Networks

Figure 3.35. Various artifacts excavated in Calanutian: upper left, Gold and evidence for tooth pegging; lower left, Chinese coins dated to twelfth century; lower right, various Asian trade beads; center piece is an Illustration of Context 259 a 12-14 old female buried with grave goods including gold, beads, tradeware, and earthenware (Ilocos Sur Archeology Project II, 2012) (artifact photos by author 2011, 2012) compare artifacts with ornaments seen in archival photo showing a Danglas Tingguian girl from elite class with arm beads interspersed with solid gold as well as four centuries-old coins inserted in neck beads (after © The Field Museum, CSA 19296. Photographer: D.C. Worcester) (Canilao 2018a).

59

Remote Sensing the Margins of the Gold Trade

Figure 3.36. Arrow showing sites in Nagsingcaoan, Cabugao (Using WV3 soils and construction combination, red-blueyellow) (DigitalGlobe Foundation) (Canilao 2018a).

Figure 3.37. Easterly view from the Nagsingcaoan gateway, trail segments were remote sensed on the ridge in the center (photo by author 2012) (Canilao 2018a).

60

Documented Networks

Figure 3.38. Encircled in yellow is an ijang in Refaro, San Juan where tradeware was found on the surface (Using WV3 soils and construction combination, red-blue-yellow) (DigitalGlobe Foundation) (Canilao 2018a).

Figure 3.39. Clearing in Refaro with tradeware surface finds (photo by author 2012) (Canilao 2018a).

61

Remote Sensing the Margins of the Gold Trade

Figure 3.40. Mapa Geografico del Centro de la Abra en la Provincia de Ylocos. Unknown Cartographer, Undated (Courtesy of the National Archives of the Philippines) (Canilao 2018a).

62

Documented Networks

Figure 3.41. Plano Topografico que comprehende una parte de la Provincia de Ylocos Sur, en la que se hallan situados los Distritos Militares de Tinguianes e Ygorrote. Unknown cartographer, Undated (Courtesy of the National Archives of the Philippines) (Canilao 2018a).

63

4 Theoretical Trade Models for Multiscalar Archeology: Social Network Analysis, Worlds Systems, Interaction Spheres, and Agency century (Miksic 2011), interlinked to a maritime trade network that spanned the South China Sea (SCS) with further connections to the Indian Ocean (Chaudhuri 2014). Studies of the gold trade, however, have tended to focus on global-scale and often homogenous patterns and processes in the fields of archeology, history, and geography (Hall et al. 2010, Stein 1999, Abu-Lughod 1991). Most of the work done has focused on the putative state ‘centers’ of gold production like Java (Indonesia), Champa (Vietnam) (Miksic 2011), Hoi-An (Vietnam) (Ky-Phuong 2008), Cambodia (Bunker 2008), and China (Bunker 1993). This section will start from the ‘peripheries’ from where gold nuggets were mined (Canilao 2019). This producerside approach has been applied in the study of other commodities in other parts of Southeast Asia (i.e., Junker and Smith 2017, Hendrickson et al. 2017, Junker 2013)

In examining the structure of modes of interaction and trade between Philippine ethnic groups and communities in the critical period between around the tenth and seventeenth centuries, prior to strong direct influence by European traders, conquerors, and colonial powers, there have been a number of theoretical models developed by historians and archeologists serving as paradigms or conceptual frameworks for understanding various types and scales of power relationships that shaped forms of resource extraction, processing, transport, and exchange in island Southeast Asia and within the larger IO–WPS/ SCS trade of this period. This chapter review aspects of various theoretical models that have recently been used by historians and archeologists to better understand forms of transaction at various scales (individual producers and traders, collective groups, groups articulated differentially into hierarchical political and economic structures, ‘state’ or ‘imperial’ structures of oversight). World systems analysis was initially mainly the purview of historians, ethnographers, and ethnohistorians, as well as archeologists examining materialized power structures often over longer periods and generally in well-documented statelevel societies. Having embraced this notion of analyzing forms of social interaction (usually on the grand scale) of societies in connection through trade (and other forms of interaction), more recent scholars have turned to smallscale groups articulated on the margins who often are deemed as having little ‘presence’ in historical records and have received limited archeological study.

The settlements within the Abra Gold Trade Network are probably one of the more complex group of Luzon polities during the Early Historical Period (tenth to sixteenth centuries). The polities within the Abra network tapped on the massive Abra River system to facilitate the bulking and transshipment of goods (most notably gold and ceramics) upstream and downstream through rafts and through intervening segments of land trails. The network is a gold trade periphery in a system that articulated the Indian Ocean and South China Sea (IO– SCS) (Canilao 2019, 2018). It can be argued that the gold that historical documents refer to as Luzon gold may have mainly been sourced from the gold-rich headwaters of the Abra River in the Cordillera Mountains. At its peak, this trading system features the complete transformation of an evanescent market to a permanent market wherein debt relations and a more large-scale exploitation of gold has led to the catalysis of this area as a trade periphery in the IO–SCS world system (Canilao 2018a). While Spanish colonial officials knew about the presence of gold deposits in certain interior niches of the Philippine archipelago, they were unable to fully exploit the resource. It was only in the first half of the twentieth century that systematic mining of the rare metal began to be undertaken under the auspices of the American colonial government (Author 2011). Works have also tracked the more contemporary issues surrounding mining and its stakeholders following recent legislation to resuscitate Philippine mining ( i.e., the revitalization of Artisinal and Small-Scale Mining [ASM], see Verbrugge 2015, Camba 2015).

Chapter two provided an overview of the structure of modes of interaction and trade between Philippine ethnic groups and communities in the critical period between around the tenth century and seventeenth centuries prior to strong direct influence by European traders, conquerors, and colonial powers in the archipelago recorded through ethnohistory and archeology. There are a number of interactional models developed in the current chapter – World Systems, Social Network Analysis, Interaction Spheres, and Actor-Based Agency Theory – that provide insights into the various ‘scales’ at which researchers can view the socio-economic interactions between individuals and groups. Following this general presentation of these spatialized conceptions of multi-layered social interaction, some examples of archeological work in the Philippines and elsewhere in Southeast Asia emphasizing multiscalar studies of intersocietal production and trade will be briefly discussed. 4.1 Social Network Analysis

The Abra Gold Trade Network features a valley inhabited by the Tingguian/Itneg ethnolinguistic group who have rich recorded oral traditions about their gold way of life. One notable Philippine ethnographer and anthropologist even

An elaborate gold culture was already locally developed in the Philippine archipelago beginning in the tenth 65

Remote Sensing the Margins of the Gold Trade Thus, we see how SNA has grown particularly powerful as a tool in evolutionary biology that looks at the diversity and variability of adaptation by animal species. In archeological research we also see the analytical power with archeological data, such as capacity for multiscalar analysis from micro to global, its potential to relate people and archeological evidence, and its ability to mediate social and physical space (Knapett 2013: 4, 7, Terrell 2013). Golitko and Feinman used SNA to show that Mesoamerican economy was dynamic and generally not highly centralized over time (2014: 206). Terrell (2013), on the other hand, has put forward some elementary questions that archeologist have to resolve when using SNA (2013: 20). So,me of the other applications of SNA is seen in the work of Gorenflo and Bell (1991) who, for instance, introduced the use of graph theory, locational models, and simulation to gain insights into aspects of regional organization (p. 95). Another archeologist, Santley (1991), used network analysis to study Central Mexico in the absence of material evidence for actual route of travel. He also used graph theory to investigate Aztec regional economic organization (p. 208). In his work, he concluded that the roadway network was dendritically organized, which impacted the status of settlements in the regional settlement hierarchy (ibid). Although the approach is innovative, it seems to be dependent on the high resolution of archeology within a region, or in the case of my book, total coverage of terrestrial settlement sites. In this book, I have included a preliminary SNA component due to the access to high-resolution terrestrial data in my study region (specifically the northeast branch of Abra River system), as well as ethnohistorical data of this area that identifies the locations of numerous settlements in the five regions of analysis (Canilao 2019).

boldly claims that ‘wealth per capita increases sharply as one goes towards the coast, with many rich men among the Tingguian/Itneg and Ilocanos’ (Eggan 1941: 13-16). It should be stated that the Ilocano linguistic group who directly inhabit the narrow coastline sandwiched between the South China Sea and the Cordilleras appears to be the colonized kin of the Tingguian/Itneg. Of direct importance to this paper is also the fact that this river valley became the battlefront in a battle of ideologies between the free natives and the assimilating Spanish colonial machinery. This river valley became the setting for colonialisminduced sibling rivalry as can be discerned from the distribution and categorization of the settlements of the area (Canilao 2019). GIS analysis Remote Sensing and social network analysis (SNA) can work hand in hand in a case study of the environmental and social variables of the northeastern branch of the Abra riverine trade network (approximately 50 x 50 square kilometers area). The GIS component used predictive modeling and plotted least cost pathways between associated villages. The WorldView-3 satellite imagery facilitated the remote sensing of the terrain. Finally, SNA provided some framework for understanding how the environmental variable relates to trade and exchange relations as well as how the historical variable (reduccion) relates to the categorization of the settlements, which is in turn reflective of an environmental principle (Canilao 2019). Network analysis is simultaneously a method and a theory that studies the spatial relational structure of complex phenomenon, wherein location matters in affecting the properties and behavior of actors. Modern SNA systematically describes social structure and its behavioral outcomes drawing heavily from mathematical graph theory and is a product of a structural-functional paradigm in American anthropological and sociological tradition (i.e., Knappett 2013, Terrell 1976). Today ecology, genetics, and biogeography, along with archeology and anthropology, combine older geographic models from the 1970s with the newer theoretical advances that explore information flow, genetic variation, and social dynamics (i.e., Bearman et al. 2004, Golitko and Feinman 2015, Kaspar and Voelkl 2009, Kerth et al. 2011, Lusseau 2003, Terrell 2013, Vance et al. 2009).

The discussion above illustrated the multidisciplinary power of SNA as both theory and method, which this book will tap into in analyzing the northeastern branch of the Abra River network. Although Snijders points out ‘difficulties of collecting network data, which are multiplied when a researcher wishes to collect them longitudinally’ (2014: 2), I would argue that the growing ethnohistorical archeology and GIS data of Northwestern Luzon would be a good case study that provides a time scale of a millennium (Canilao 2019). 4.2 Multiscalar Interaction: Models from World Systems, Social Network Analysis, Interaction Spheres, and Agency Theory

Kaspar and Voelkl for instance has used SNA for reassessing the simplified classification schemes of primate systems. They were able to demonstrate the utility of ‘a set of network measures that are useful to describe primate social organization using data from 70 groups from 30 different species’ (2008: 343). They found considerable variability (diversity) in terms of group structuring among primate groups (ibid). In terms of elephant variability, Vance et al. were able to model elephant social behavior using bilinear fixed effects including kinship based on changes in the environment (Vance et al. 2009: 273). In terms of dolphins, Lusseau was able to model the self-organizing principles among bottlenose dolphins specifically when it comes to information connectivity (2003: s186).

The World Systems Theory (henceforth, WST) as originally formulated by Immanuel Wallerstein (2004) was influenced by the dependency theory of Andre Gundre Frank and was originally used as an analytical tool to look at the historical experiences and present-day conditions of formerly colonized and now underdeveloped nationstates (1970: 4). The age of colonization and discovery that culminated with the purported long sixteenth century was the turning point towards the economically and developmentally unequal, modern world system 66

Theoretical Trade Models for Multiscalar Archeology the purported systemic logic that guides WST. Some of the usual components for this system logic are mode of production, mode of accumulation, state formation, cycles of centralization and decentralization, and modes of social integration continuationists and transformationists (pp. 127-128). In the context of the ‘New archeology’ and its call for processual studies, WST seemed to be an appropriate tool in light of this paradigm’s emphasis on evolutionary trajectories. Parkinson and Galaty in particular refer to social evolution as one of the processes that the New archeology studies of interaction have tried to address (2009: 6). WST became an attractive tool in identifying catalyzing factors because it seemed to be able to integrate both social evolution (but often in large chronological periods) and somewhat syncretic interaction between social groups (but often in a large aggregate form) (ibid).

(Wallerstein 2004: x). Because of its critique of the global capitalist status quo, the concept became a vogue in the discipline of economics, development studies including development anthropology. Note that WST only endeavored to explain the modern-day system, attributed to the great transformation that began 500 years ago (Kohl 1987: 2, Polyani 1957). Initially intended to be of explanatory value for diachronic and synchronic studies that dealt with the sixteenth century to the present, WST was deemed to require significant adjustment when projected to systems in earlier, pre-modern periods. 4.2.1 World Systems Theory, implications for archeology From the beginning, WST was appealing to archeology because of its macro-scale analysis and processual applications. But because of the criticisms of WST used in archeology (expounded below) early on, Kohl has called for alteration of the original formulation, specifically for its application in archeology (1987: 29). Kohl provided his justification for the use of WST in premodern times (1987: 4), by arguing: 1) WST presents long-range historical data that can be used by processual archeologists who have been accused of being ahistorical, 2) there is, in fact, a continuity from ancient to modern times and Wallerstein’s dichotomy is misplaced, and 3) WST’s emphasis on structure and hegemonic relations among and between societies counters the connotation of just accidental diffusion (1987: 4-5). Kohl specifically vouches for the merits of WST’s application in prehistoric archeology because it forces archeologists to look at the works of macro-historians and historical sociologists, instead of being devoid of theory with narrow historical views (1987: 5). He also justifies the use of WST by arguing that even contacts in prehistory were based on economic factors that were not equitable to all groups and individuals involved in ancient economies (Kohl 1987: 29). He suggests that where there is evidence for meaningful contacts like trade and exchange among societies in the past, a ‘world’ framework must automatically be utilized (1987: 29). He admits, however, that potential misuse of the model is applied ‘literally to earlier social formations’ (ibid), intimating that social scale, along with structural and historical factors, must be considered (ibid).

WST, or its refined version World System Analysis (WSA), is seemingly relevant to this book because the period of accelerating trade interaction in the IO–WPS/ SCS intersects with the Early Historical Period from the tenth to sixteenth centuries and the post-Spanish contact period from seventeenth to early twentieth centuries in the Philippine archipelago. As discussed in chapter two, there is voluminous historical and archeological evidence of intense and heavily structured forms of trade at various scales highlighted in this book – i.e., the vast Indian Ocean trade connected the Asian, African, and European continents and Southeast Asian islands significantly earlier than the tenth to sixteenth centuries’ height of trade and the European colonial period of trade from the seventeenth century onwards; geographically smaller regions in the Philippine islands jockeyed for trade primacy at coastal ports; and the numerous groups in the Philippine island interiors developed strategies to provide desired upland products (including gold) as one end node of this complex and multi-leveled system of exchange. As examined in later chapters, groups in interior Northwestern Luzon already had significant relations with Chinese traders, certainly by the beginning of the second millennium, through a specialized niche of valuable upland resource extraction, particularly gold, and at a macro level they constituted a source point or end node to the IO–WPS/ SCS system.

In the context of archeological and historical studies, Hall and Chase-Dunn also concluded that the world systems perspective is appropriate when it tackles evolution (1993: 124). But in order to apply this to pre-modern cases, archeologists must address three fundamental debates concerning WST’s application to pre-modern societies, namely: 1) defining world system, 2) delineating spatial boundaries of world systems, and 3) resolving the problem of systems logic (ibid). The first debate was prompted by the existence of stateless and classless societies that were still articulated to a world system in the past (Hall and Chase-Dunn 1993: 125). The second debate, on the other hand, concerns those who embrace diffusionist perspectives and those who focus on particularist culture history (p. 126). Finally, the third debate is identifying

However, there seems to be no end to the cycle of reinventing WST and spawning new questions about its utility, with Stein calling for its total rejection and arguing that it is fundamentally flawed. Stein (1999a; 1999b) contends that a common misuse of the WST is when archeologists inappropriately equate any material evidence (i.e., artifacts, especially exotic) of a network of connections to a type of world system (1999a: 4). The gravity of this error is magnified when archeologists conclude that secondary states form by virtue of being in the orbit of pristine states quite akin to some sort of diffusion of social complexity (Stein 1999a: 3). In its extreme, the pitfall is when one makes the inappropriate conclusion that some kind of colonization and forced 67

Remote Sensing the Margins of the Gold Trade At this juncture, I present alternatives to WST, which ranges in a scale from reinvented/redressed alternatives to the WST to total replacement of the WST model.

annexation/assimilation took place in societies in the past. According to Stein this is brought about by a fundamental flaw in WST itself; namely: 1) that the core dominates the periphery by means of military, technology, ideological, or organizational superiority, 2) the assumption that the core controls the trade, and 3) that the core itself supplants the political economy of the periphery (1999a: 4). To drive home the point, he hits on the work of Algaze (1993) on the analysis of the Ubaid and Uruk world system in Mesopotamia as too economistic deterministic (Stein 1999a: 5). Stein presents his own analysis of this network, avoiding the purported pitfall of Algaze in ignoring noneconomic factors (Stein 1999a: 82-116).

4.2.2 World Systems Analysis As mentioned above, there are already calls to look for an in-between position or middle ground position in the WST debate. This position eventually catalyzed into World Systems Analysis which incorporated the notion of negotiated peripherality (Kardulias 1999) or contested peripherality (Allen 1996) to address the agency of the periphery. Kardulias and Hall answered the call for altering WST and offered the modified/reformulated version as the reinvented World Systems Analysis (henceforth, WSA) (2008: 573). One of the hallmarks of this new WSA in archeology is the recognition of self-contained worlds whose boundaries can be drawn based on ‘steep gradients of density of interactions’ (p. 574). Parsing forms of interaction, Kardulias and Hall enumerated four types of networks that can be perceived in the archeological record, namely: 1) bulk goods network, 2) politicalmilitary networks, 3) luxury or prestige goods networks, and 4) information networks (p. 574). This approach that separates different types of interaction that are not strictly economic will be reviewed below. It is important to note that, within the systems of these networks, the author argues that there are built-in cyclical processes that result in ‘pulsation, expansion, contraction and the mere existence already presupposes the underlying system’ (Kardulias and Hall 2008: 575). They also introduce the concept of variable forms of incorporation of the periphery (ibid). In terms of modes of incorporation, the continuum can range from a scale measured from normative interaction to a relationship of subordination (ibid). Those societies who have the tendency to stay within the more egalitarian end of the scale were called either marginal (Sheratt and Sheratt 1991) or hinterland (Frank and Gills 1993).

At this juncture, the discussion homes in on one of the underlying aspects of WST that has sparked debate among archeologists: the unit of analysis. Echoing what Kohl (1987) and Wallerstein (2004) had noted, Frank argues that the misuse of WST is when the concept, in its totality, is applied to worlds before the modern capitalist world system existed (1999: 277). He poses several questions and caveats that need to be addressed before applying this approach, including the need to define the social unit of analysis or the boundary of the world system (p. 292). Frank also argues that the archeologist must be aware of the world system, including the relatedness at different scales and levels of the constituent member sites (1999: 283). He adds that ‘it is ignoring the internal dynamic of the whole world system that puts our partial analyses at risk of missing an essential element of the explanation’ (293). In response to this, Kardulias asserts that the unit of analysis is the world system itself and everything within it that can only be understood in the context of the world system (2007: 6). Frank responds by evaluating the work of archeologists, citing the works of Wells (1999), Kardulias (1999), Peregrine (1999), Jeske (1999), and Alexander (1999) that have managed to elude the problem of not having knowledge of the entire interactional system. Meanwhile, Feinman (1999) attempted to grapple with the problem of defining the unit of analysis, whereas Stein (1999b) dismisses it completely (Frank 1999: 293). Feinman invites archeologists to not be discouraged by missing evidence, but instead adhere to the saying that the absence of evidence is no evidence of absence in the conduct of archeology (ibid). Another contentious aspect of WST that Frank addresses is the issue of ethnicity introduced by Stein (1999b). Frank is dismissive of the idea of ethnicity in the prehistoric record, arguing that the concept is circumstantial and not essentialist (1999: 287). In fact, Frank (1999) echoes Hall who insisted that most world systems were multi-ethnic (Frank 1999: 287), throwing the ethnicity question back to Stein, asking ‘How can one look inward?’ (1999: 287). In this book, I engage the notion of ethnicity in multiscalar terms because the Early Historical Period (tenth to sixteenth century) in Northwestern Luzon has some ethnohistorical, linguistic, as well as archeological evidence for distinct ethnicities (i.e., Ibaloi, Kankanaey, Bago) of gold miners, traders, and coastal peoples in the trade network, as well as umbrella regional groups (i.e., Ilocano, Igorot).

It is important to state that WSA, as formulated by Kardulias and Hall (2008), appears to be more amenable if we are interested with purported peripheries, or if we want to engage a ‘bottom to top’ (or ‘periphery’ to ‘core’) approach rather than privileging the apparently larger socio-political entities involved in interactions. Through WSA, we begin to appreciate the fact that so-called ‘peripheral’ people may have bargained their positions in the system through their control of key resources and extensive local knowledge (Kardulias and Hall 2008: 577). Negotiation is also apparent when indigenous groups in the peripheries select or reject outside elements based on their advantage (Kardulias and Hall 2008: Hall 1986, 1989, Kardulias 2007, Morris 1999). In this book, I explore the applicability of the WSA model in Northwestern Luzon, specifically the notion of negotiated peripherality of the pre-Spanish contact of indigenous groups during the Early Historical Period (tenth to sixteenth century). At this juncture, we will now look at case studies of WSA and assess its applicability to the gold trade in 68

Theoretical Trade Models for Multiscalar Archeology theory, at the same time paying attention to the conscious human actors within a society (pp. 2, 6). Morris argues that this seems to be the case when the Greeks placed emphasis on their own deposits of iron rather than imports like gold, ivory, and bronze (Morris 1999: 4). He points out that during the period between 1050 and 850 BC there appears to have been disagreement and competition between communities over embracing the new exotica (p. 5). Morris contends that in the case of Greek communities, some embraced the influences from the Levant while others resisted it (p. 6).

Northwestern Luzon during the Early Historical Period and examine critiques of the models authored by Frank (1999) and others. 4.2.3 Negotiated Peripherality Kardulias (2002) studies the post-contact Native American market for fur in North America and argues that usually the impetus for change starts from the core moving to the peripheries, but there are also circumstances, as in the case of the Native Americans and their fur trade with Europeans, wherein the periphery outlined the terms of their incorporation (Kardulias 2002: 1). He outlines how Native Americans altered their production strategies in order to participate in the fur trade (ibid), arguing that economic rewards increased the focus on trapping beavers within a system of grades (pp. 5-6). The Native Americans collected the furs and transported the items to collection points or trading posts (p. 6). Though there was direct trade with Europeans, the fur trade also gave rise to the Native American middleman (p. 7). In the case of Northwestern Luzon, discussed in subsequent chapters, the book queries whether the rise of the middlepersons and the establishment of bulking stations present some parallels to the Native American fur trade. In the case of the fur trade, Kardulias concluded that there were four effects of increasing negotiation of peripherality: 1) dependence on trade goods, 2) territorial expansion, 3) changes in social structure (female labor), and 4) effects on resource (2002: 8-11).

By way of critique, Frank points out that the work of Morris seemed like longish history because the dates in his work are not presented chronologically (1999: 282). Frank also accuses Morris of contradicting himself when he distances himself from grand sweeps in his small-scale study but upon closer look the study is actually riddled with chronological issues that catalyze into a grand sweep (pp. 282-283). Finally, Frank also criticizes Morris for not explaining in detail his assertion that the actors had knowledge to negotiate their peripherality (1999: 282). A study of Sheratt and Sheratt (1991) in the Mediterranean is interesting because of their focus on long distance exchange and its repercussions for local elites. They note that the sequences of incorporation move from contact to semi-transfer then finally to full linkage (p. 357). They argue that ‘participation in long-distance exchanges gave local elites opportunity to absorb some of the values and practices of the core areas, generating their own raw material requirements and adding a further diversity of finished products to the system’ (p. 356). Because of the active search for high value materials by peripheral societies and cultures, demand for luxuries preceded production of commodities (ibid). Because sailing conditions relied on winds, currents, landfalls, and good harbors, Mediterranean shipping traffic was tramping along routes with multiple transactions (p. 357). Sheltering points developed, like Troy, where vessels anchored until the arrival of favorable winds (ibid). There seems to be parallelism in the conditions of maritime trade in the cases of the Aegean and Southeast Asian seas, congruent with Braudel’s (1980: 107) traveling bazaar where trade is traipsing from port to port. A parallelism to one of the main components of the Northwestern Luzon trade is the critical concept of the evanescent market encounter where ad hoc coastal trade occurs (Allegre 1998: 148). Sheratt and Sheratt’s (1991) is significant in offering a comparative case to Northern Luzon trade features, with ‘traveling bazaars’ and evanescent market encounters requiring complex decision-making.

In another study, Kardulias (2007) used burial evidence to investigate the semi-peripherality of Malloura Valley in Cyprus in the bulked goods and prestige goods network within a large-scale political-military network (p. 63). Kardulias argues that cultural changes in coastal societies did not immediately impact areas like Malloura (2007: 60). He believes that this is an indication that inhabitants of the valley were beginning to negotiate their peripheral status in the Mediterranean World System. He focuses on the locations of extraction zones, but in the absence of settlement evidence he relies on the use of valley resources instead as correlates for his argument (2007: 60). He focuses on the pattern of resource extraction like embedded procurement (Binford 1980) and syncretism in religion to support this argument. The notion of embedded procurement that was observed in the Malloura is interesting for drawing analogies from my work. A common origin narrative in the oral traditions of the upland groups is a story that they were primarily hunting deer and placer mining on the rivers was incidental (Bagamaspad and Hamada-Pawid 1985). By way of critique, Frank (1999: 281) has suggested that Kardulias should invest more time and effort in locating his system rather than confounding his work with multiple levels. On that note, he also suggests that Kardulias expand his analysis to include Egypt, Anatolia, and Syria and identify other regions like that Aegean that were part of the periphery (ibid).

Therefore, in the context of emergent local elites within a negotiated periphery, Parkinson and Galaty point out that the role of emergent elite as participants in long distance exchange networks becomes integrated with the generalizing tenets of world systems frameworks (2009: 23). The concept is seen as cross cutting the dichotomy of agent and structure, effectively rectifying scales of

Another work on the Aegean by Morris (1999) argues that small-scale empirical studies contribute to generalizing 69

Remote Sensing the Margins of the Gold Trade and developmental change (Stein 1999: 3). He points out that an overemphasis on external dynamics tends to blur the internal dynamics. The macroeconomic orientation of WST/WSA tends to miss any pertinent indicators that the periphery practices its own agency. What if the periphery can also call the shots in regional interactions with other polities at various power positions (1999: 3-4)? It may also be possible that the periphery undergoes a process of negotiation, so that there is really no imposition from the core (Kardulias and Hall 2008: 573).

analyses (ibid). At this juncture we will look at some works that attempt to engage the question of the emergent elite, local power, and agency within the WSA framework. 4.2.4 Multiscalar Approach Another redress to initial simplicity in the WSA is Feinman’s formula, which attempts to look at multiscalar approaches that integrate dual processual theory (Blanton et al. 1996). Feinman (1999) argues for the utility of a macroscale WST perspective, but he qualifies that this should be conducted at various scales simultaneously from households through macro-regions (p. 53). Feinman was aware of the debates of scale or analytical units and luxury goods/preciosities when he dealt with the Valley of Oaxaca as it transitioned from Classical to Postclassical periods (AD 700-900) (1999: 54). He argues for a Mesoamerican world system during the Classical period (1999: 58), while the transition to Postclassical ushered in a fragmented political landscape where there is a minimization of local core–periphery hierarchies (1999: 57). While there seems to be less emphasis on collective public architecture in the Classical period, there seems to be more emphasis on individual tomb accumulation with general artifactual diversity found in the tombs seemingly reflective of broader-scale processes (1999: 56). He notes in his work that the role of prestige exchanges varied in systemic importance from Classical to Postclassical periods, with more emphasis on wealth exchanges occurring in the latter period (pp. 58-59). He also argues that rulership was more collective in the Classical Period, as opposed to a more network-based rulership in the Postclassic Period. Frank (1999: 288) lauds Feinman for his multiscalar approach, but he also reminds archeologists that being too macrocosmic will not easily explain the utility and meaning of items dug up from a site, nor will a too microcosmic approach because of the need for larger context (Frank 1999: 288).

Stein (1999: 44-64) presents both the Trade Diaspora and Distance Parity models of interaction as an alternative framework to WST/WSA. According to Stein, these models recognize inter-societal interaction and its role in individual polity and interregional systems (1999: 45). Its basic principles include the following: 1) recognition of the importance of interregional interaction, 2) focus on the local level, de-emphasis on organizational systemic structure and emphasis on organizational dynamics, and 3) two-track approach looking at cross-cultural applicable mechanisms and culture-specific aspects (ibid). Stein argues that there are social factors in both core and periphery that affect the system endogenously (p. 45). He points out that middlepersons also travel between regions or relocate to a foreign community specifically to facilitate trade (p. 46). He calls this a trade diaspora, which he qualifies as a cohesive group who specialize in the trade of certain commodit(ies) and embed themselves in a host community without assimilating the latter’s culture. Members of this diaspora group lay claim to resource turfs and maintain proper trade values/etiquette within their organization (pp. 47-48). He argues that there is also a veritable diasporato-host relationship continuum beginning with marginal status, followed by social autonomy, and in extreme cases, diaspora domination of host community (p. 49). The Chinese diaspora within Asian host communities was used here as an example (p. 50). At this juncture, it is important to note that in the Philippines, there are several accounts of the pre-Spanish contact Chinese and Japanese diaspora in Northwestern Luzon (see Mateo 2004: 122, Beyer 1979: 119, Chan 1978: 53, Wernstedt and Spencer 1967: 119, Keesing 1962: 13, Wilson 1932: 1), and these will be dealt with later in the book. If Northwestern Luzon is the periphery, it appears that the gold was bound for Canton in China. This is evidenced by the fact that the gold brought in by trade ships that plied Island Southeast Asia were cast into forms in Canton (Chan 1978: 52). Later in the book, I will discuss the possible application of the Trade Diaspora model in my specific case study.

In the context of my book, I choose to examine multiple scales of analysis: at individual sites representing aspects of gold production and trade, within larger regions including coastal and mountainous interiors where differing groups specialized and competed in raw gold extraction and trade, and larger connections with the oceanic trade beyond the Philippine islands. I examine how coastal and the upland groups negotiated their peripherality using multiple scales, similar to Junker’s (1999, 2013) work in the TanjayBais region of Negros Island in the Philippines which demonstrates the utility of a multiscalar approach to the archeological study of complex social landscapes. My work echoes the call for more long-term, regional-scale settlement archeology in early historical Southeast Asia trade that deals with complex and multi-layered social networks.

Finally, to complement the Trade Diaspora model, Stein presents the Distance Parity Model, which shows that exchange moves from ‘asymmetric’ in favor of core, to ‘symmetric’ in favor of both core and periphery as the distance increases, simply because of the decay in hegemonic influence of the core (1999: 62-64). In other words, the balance of power is subject to levels of technology and transport possibilities (p. 64).

4.2.5 Trade Diaspora, Distance Parity Model Stein, on the other hand, argues that WST/WSA is too tradeoriented when it assumes that a purported core dominates a purported periphery by controlling its political economy 70

Theoretical Trade Models for Multiscalar Archeology 4.2.6 Regional/Multiscalar Studies of Lowland–Upland trade in Philippine and Southeast Asian archeology

In the case of Northwestern Luzon, there seems to be a diversity of groups and socio-political complexity as well. In terms of a cross-comparison, in Tanjay the archeological evidence shows that ecological specialization extends for at least a millennium and it is argued that this is part of a long-term cultural adaptation (p. 239). Based on the work in Tanjay, the archeological record seems to indicate that the upland–lowland interaction increased between the fifteenth and sixteenth centuries as the trade interaction between the lowland Philippine chief with overseas merchants increased (p. 239). The archeological correlates of intensified upland–lowland trade include the establishment of secondary centers upriver with clustered hunter-gatherer camps. In the case of Northwestern Luzon, this book will investigate the bulking stations that were located halfway between gold mines and the coast. Junker (2002a) argues that the hunter-gatherer agriculturalist relations started as mutualistic but became coercive over time with increasing linkage to the lowland political economy and beyond (p. 240). As mentioned earlier in chapter two, Quirante notes that the Igorots owed so much credit to the lowlands because of the gold trade.

As noted in chapter two, most archeology in the Philippines focuses on individual sites rather than engaging in regional-level survey and multi-site excavations that reveal complex settlement patterns over larger regions that can provide a window into the interlinked ‘small worlds’ of harvesting resources, producing local products, and exchanging goods with both local and foreign traders. There are a handful examples of still-rare regional studies focusing on the social and economic interactions between distinct communities that are multi-ethnic and have specialized roles in these regional-scale, complex social and economic networks in the Philippines. Green (2010) argues that in the context of the Philippines, both upland and lowland groups should be studied together when looking at their socio-politically complex landscape as they navigated Spanish colonization (p. 541). Junker (1999), and more recently Junker and Smith (2017), suggests that ‘historical records allow researchers to specifically identify the small scale groups who produced archeological sites as long term “foragers” in the margins of lowland polities, while in other cases it is indeterminate whether the groups providing forest products are seasonally mobile foragers or sedentary upland swidden farmers who engage in forest product commerce’ (p. 16). In the case of upland Northwestern Luzon, we have some primary and secondary ethnohistorical sources that we can use as leads to get to the socio-political structure of the upland. Junker and Smith (2017) note that upland and lowland interaction includes highly visible lowland-originating trade goods such as foreign porcelain and stoneware, glass beads, and bronze ornaments and other metal goods at large interior sites that show political and economic connections to polity centers through these prestige goods (p.17). In this book, similar types of archeological correlates (particularly datable porcelain, but also other chronologically known artifacts such as beads, earthenware pottery, and metal ornaments with distinct ‘styles’), together with ecofacts (such as preserved charred seeds of plants in hearths and upland-associated animal remains) will be analyzed in future work centered on these Northwestern Luzon interaction spheres.

Junker (1996) argues that when it comes to huntergatherer and agriculturalists relations, there has been two models used; namely: 1) the isolate model and 2) the interdependent model (p. 389). Junker argues that the latter may be the case for the upland, interior groups of Tanjay (ibid). Using regional distribution of manufactured goods, foodstuff, and tool assemblage as archeological evidence she argues that specialization and exchange extends to at least a millennium (1996: 405, 2002). As counter-argument to the ethnographers who insist that hunter-gatherer groups did not change until recently, Junker (2002b) argues that the evidence supports a significantly developed economic specialization vis-avis ecological zones but economic choices are fluid and dynamic, arguing that ‘[…] once interior trade systems became linked to an external market and enmeshed in lowland chiefly strategies for wealth accumulation, there was likely a shift in trade commodities’ (pp. 379-380). This entails more luxury goods flowing upriver and more exportable forest products flowing downriver (p. 380). Green’s work (2010) in the same region uses geoarcheological methods to show that it is important to study both the lowland and the upland as part of an integrated landscape of socio-economic interaction (p. 541). She was able to demonstrate an increase in upland– lowland interaction towards the late Early Historic Period or in terms of Tanjay chronology, the transition from Santiago to Osmena phases (eleventh to sixteenth century). She notes the importance of a river-based upland–lowland trade in the intensifying political economy in Tanjay (pp. 539-540), likely increased due to the demand for indigenous exports such as interior forest products including hardwoods, rattan, metal ores, spice-yielding plants and trees, resins, and waxes, to maintain expanding foreign trade (ibid). Her study documented an increase in forager and swidden farmer sites on the margins of the

Junker (2002a) carried out a longitudinal research in the Tanjay Region in Negros that aimed to look at the long-term diachronic perspective on agriculturalist and hunter-gatherer interaction (p. 219). Based on the study, ‘ecological specialization and exchange of foodstuffs and manufactured goods’ (p. 239) between lowland and upland groups is not a new phenomenon as suggested by ethnographers. According to Junker (2002a) the uplands in the Philippines by the sixteenth century was inhabited by an ‘amalgam of ethnically and linguistically distinct groups’ (p. 203) and ecologically specialized groups, including hunter-gatherers, ‘tribal’ communities with permanent or shifting agriculture, and emergent ranked societies with intensive agriculture on coastal plains. 71

Remote Sensing the Margins of the Gold Trade lowlands and she uses the distribution of exotic material traded in exchange to track this movement (p. 540). Furthermore, she suggests that the zones occupied by the ancestors of the Bukidnon and Magahat groups were the areas for the expansion along zones occupied by ancestors of the contemporary Ata (p. 540).

Quirante that the Igorots have so much debt/credit with the coastal villages because of the trade for gold (Quirante 1624: 279). As an important side note, it appears that this credit system may be an early manifestation of the system encroaching on the uplands with the lowland costal chiefs benefitting from the arrangement.

Finally, Smith (2014: 232) studied the Malangwa watershed of Negros as it navigated the transition to the early Spanish colonial period. She interpreted the burgeoning of nine settlements in the Early Historical Period to 34 sites in the early colonial period as the result of intensification and the movement of indigenous people fleeing Spanish occupation (ibid). In her study, she found out that imported ceramic density for the 34 early colonial period sites remained similar to the pre-contact levels (p. 230), unlike Tanjay which was rapidly colonized by Spaniards coveting the fertile Tanjay River plain to establish sugar cane haciendas. Using geostatistics she demonstrated that high density sites for ceramics were spatially separated in what might be settlement sites, but these clusters were surrounded with low density sites (ibid). Unlike the collapse of Tanjay settlement patterns in the sixteen to seventeenth century, Smith argues that the pattern of concentrated villages or ‘centers’ with trade ceramics is similar to pre-contact distributions (ibid). Thus, her study is integral in providing us with a picture of how groups in the Malangwa watershed navigated colonial times, on one hand intensifying their social interaction through porcelain exchange, while on the other hand in-migration of people avoiding Spanish rule from nearby islands who integrated into the community (p. 235). She argues that the traditional modes of power for the people of the Malangwa watershed persisted into the early colonial period and there was even an increase in the availability of prestige goods (p. 235). Her unique contribution to the remontado literature is that she presents a case wherein the remontados do not retreat to the mountains of the volcanic islands; instead, they carve out their refuge in the same lowland landscape setting even under Spanish domination.

The main issue with some of the work already done in Northern Luzon (e.g., debates between Maher 1980, 1974, Bodner 1986, Acabado 2009, 2012, 2017, etc.) is that it tends to be single-site focused. In the Ifugao case, it has created a long-debated issue of the dating of intensive rice agricultural in the region as the core topic. This work stands apart in that it employs a regional-scale perspective to examine the complexities of economic and social interactions within five distinct gold trading networks/ regions spanning the pre-Hispanic and Spanish period eras, using archeological survey and excavations combined with GIS modeling and multi-vocal historical accounts to examine local strategies of trade and interaction in the region. It should be noted that Southeast Asia focused archeologists are beginning to direct archeological and historical work on social and economic interactions between large-scale complex societies and adjacent ‘tribal agriculturalist’ or ‘forager-trader’ groups in other areas of Southeast Asia. One example is the work of Mitch Hendrickson et al. (2017) whose interest is in specialized roles of ‘tribal’ or ‘forager’ populations in supplying resources to ‘state’ and ‘empires’ in mainland Southeast Asia, specifically Cambodia. Laure Dussubieux and Berenice Bellina also carried out work in Malaysia and Thailand, focusing on glass beads, noting that the diffusion of these artifacts were directed towards the hinterland and in the West Philippine Sea/South China Sea towards other communities involved in the trade (2018: 34). Other examples are seen in the work of Higham et al. (2011) and Higham (2013), who examined the concept of Southeast Asia’s friction zones, where foraging and farming peoples come into contact (Junker and Smith 2017: 5). Barker’s (2013: 357) work in Borneo focused on the Niah Cave area and also includes significant evidence of these kinds of upland– lowland trade systems, calling attention to the inaccuracy of acculturation and colonization in encapsulating the dynamism of interactions between foragers and farmers.

4.2.7 Agency-Reciprocity, Utang Na Loob In the case of the Philippine archipelago we can also mention the use of a ‘reciprocity as agency’ perspective as advocated by Grace Barretto-Tesoro (2008). BarrettoTesoro’s work looks at the differential impacts of external trade on different societies in the Philippines. She uses an agency perspective to unravel the concept of reciprocity which she argues influences identities and statuses (2008: ii). Her conception of utang na loob, which she translates as debt from inside (indebtedness), allows us a vista on credit relations in the Early Historical Period. This type of debt is not a legal debt but essentially a social or moral debt (p. 40). Barretto-Tesoro alludes to the negative aspect of utang na loob wherein ‘persons of influence oblige others to pay beyond their capabilities that they become indebted or bound to their benefactors for an indefinite period even for generations’ (40). This view of exchange becomes interesting in light of the account by Alonso

4.3 Chapter Summary This chapter of the book explored the theoretical handles that are useful for the book including negotiated peripherality, interaction spheres, trade diasporas regional/ multiscalar studies, and agency-reciprocity. As will be seen in the rest of the chapters, the local Igorot miners negotiated their participation in the IO–WPS/SCS trade emporium through their commitment to either evanescent markets or permanent markets, where the former offered a latitude of agency whereas the latter had limited agency. The receding agency appears to be linked to the solidification of the coastal middlepersons as lenders who 72

Theoretical Trade Models for Multiscalar Archeology offer credit to these primary producers. This debt or credit also cuts to the soul (kaluluwa) of a person as shown in the notion that debt is from the inside (utang na loob). If unpaid debts would result in persons becoming slave subjects traded along with tradeware and other luxury items from overseas. The next chapter talks more about the putative slave-for-gold exchange in Northwestern Luzon as an offshoot of a more permanent market situation.

73

5 Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis Snead, Erickson, and Darling (2009) enumerated some important studies of trails around the world (pp. 2-18). Hyslop’s work on the Inka road system brings to the fore the importance of a long-term, regional archeology, systematic survey, and the use of historical and ethnographic documents (Hyslop 1984, 1991). Another compendium publication based on studies in the New World argued that trails, paths, and roads should be viewed as ‘built environment’ that is noteworthy for description, classification, analysis, and interpretation (Trombold 1991), just as any artifact recorded by an archeologist. Ferguson, Berlin, and Kuwanwisiwma (2009) meanwhile used trail analysis to argue that ancient trails may integrate pedestrian trails, pack animal routes, wagon roads, and modern unpaved roads and have multifaceted utility and meaning, which is the case for the Hopi (p. 34). In the Hopi transportation network, trails and roads overlaid one another over time and have changing function and meaning (p. 40). The researchers cite the modern case of Arizona State Route 264 (p. 26). They add that pilgrimage and ceremonial initiations may be related to trail formation (p. 38). On that note, gold trails in Northwestern Luzon have sometimes been described as headhunting trails because young initiates take their trophy heads along these trails.

In this chapter we look at the remote sensing and GIS spatial analyses tools for identifying transportation and probable transportation routes in Northwestern Luzon. To begin with, the book will discuss cross-regional examples of studies that look at transportation routes as well as approaches to unraveling ancient trails. The chapter will also show how an aerial perspective fed into Geographic Information System (GIS) analysis is also a good approach. The chapter will then discuss each of the five networks in detail with an emphasis on what methods and approaches worked for each setting. The predictive modeling is ceteris paribus (all things being equal); however, I emphasize that the digressions from the predictive model was the focus of the book because these may be attributed to cultural factors. 5.1 Importance of trails and transportation routes in ancient settlements According to Snead, Erickson, and Darling (2009) there is growing appreciation of the importance of trails, paths, roads, tracks, trackways, and related phenomena in the archeology of landscapes associated with efficient transit for social, economic, and political purposes (p. 1). These authors argue for the utility of a landscape approach in studying trails, what they call landscape of movement (p. 3). This approach considers pattern, scale, context, and association, incorporating the trails as features themselves (ibid). The primacy of these interests arises because such features structure and reflect significant elements of human life that vary according to culturally specific norms and practices in specific societies (Snead, Erickson, and Darling 2009: 2).

The theoretical aspect of studying trails has also seen some advances in terms of inferring aspects of political structure, mobility, trade, and economic control. In the New World, one interesting study done in Central Mexico attempted to document patterns of political integration and mobilization based on a materialist mode of transport cost (Hassig 1985, 1991). Meanwhile, other studies of trails were used to infer significant levels of hegemony and control, or lack thereof (Chase and Chase 2001, Ambrosino et al. 2003). Trails also became a window to a notion of political self-sufficiency (Snead, Erickson, and Darling 2009: 14). The importance of trails is also highlighted when they are viewed as structures of movements that provide an archeological window into political economy by reconstructing the movement of people, livestock, and material goods through a landscape (Snead, Erickson, and Darling 2009: 13). Trails reflect the flow of tribute, trade, and other interactions between societies that make political systems work. As noted earlier in this book, Junker has argued the importance of studying maritime and river-based trade in Southeast Asia through longterm regional settlement archeology (Junker 2013: 11), since it is only a regional approach that can reveal both patterns in ‘within region’ trade and the larger connections of extra-regional trade. In this book, I examine how trails articulated these coastal and upland settlements in the Early Historical Period (tenth to sixteenth century) through the Spanish contact period (seventeenth to twentieth century).

In the context of a regional-scaled analysis, it is important to be able to identify trail segments in order to study the structural configuration of the linkages that constitute the larger and holistic network (Ferguson et al. 2009: 40). Also, in the context of interaction spheres, it is important to pay attention to trail networks that ultimately carry within its deposits archeological remains that will help us understand interactions between and among groups (Ferguson, Berlin, and Kuwanwisiwma 2009: 41). Another role of trail studies is to document the ‘historical grounding for local ambitions’ in the case of systems of interactions that involve political hierarchies (Snead 2009: 58). This is of direct relevance because, as argued in the theoretical section, this book examines how local elite authority becomes legitimized through the long distance trade of exotics (Parkinson and Galaty 2009: 25-26). But it should be emphasized that the jumping-off point are the interior peoples who are the primary actors moving resources through the trails until they reach the larger entities on the coast. 75

Remote Sensing the Margins of the Gold Trade argues that the road system satisfied the requirements of the Khmer elite by providing a relay system during the reigns of Suryavarman II and Jayavarman II (2011: 445, 454). For the Khmer state, a key element in expansion and maintenance of territories was the road system and the communication system that is a significant system within it (Hendrickson 2011: 444).

Timothy Earle has argued that the spatial distribution of roads and the inferred scale of road systems document the complexity of scale of social and political networks and boundaries (1987: 8, also 1991). Because we are dealing with chiefdoms in the coastal areas who were controlling the trade as gateway communities (Earle 1987: 296), it is important to see how the trails are articulated to the advantage or disadvantage of the coastal trading centers and the upland groups. Based on a cross-cultural, comparative approach, Earle was able to create a heuristic typology of route movements which he envisioned as a working typology and cultural evolutionary model to explain variation in levels of social complexity and political reach (2009: 253-255). He further describes trails as regional and long-distance routes whose function is to facilitate trade over great distances, linking different local groups and separated resource areas (2009: 256). It should be stated, however, that the theoretical dichotomy between informal trails and formal roads is not clear-cut because of the existence of ‘shadow zones’ (Trumbold 1991: 3). An example would be when segments of informal trails are eventually incorporated into a modern road.

One question that emerges in this book is: were the highland trails articulated in such a way as to efficiently facilitate the extraction of gold from the upland mines? In the case of the Khmer, using GIS analysis, Hendrickson argues that a recurring feature of the Northwest, Northeast, and East roads is that they provided Angkor with access to critical economic resources (2011: 453). He also notes the presence of rest house temples (ibid) of gods, which protected the roads. Interestingly there are also bulols/anito (gods) that are placed along the trails of Northwestern Luzon as if to protect the travelers (Turn of the 20th century ‘trail photo’ © Field Museum CSA 18525 Photographer Dean C Worcester). As part of Hendrickson’s methods, he plotted the regional distribution of salt, iron, copper/tin, and gold in relation to the Khmer road system (Hendrickson 2011: 449). In the case of precious metals, he used modern reports that include references to ancient mining (ibid). He noted the value of the East Road as a link to the copper and more notably iron deposits of Preah Khan and Phnom Dek (Hendrickson 2011: 453). Preah Khan appears to be a frontier town used to monitor iron production in Phnom Dek (Hendrickson 2011: 454, Hendrickson et al. 2013: 3). In Northwestern Luzon, the case seems to be the other way around where bulking villages are used by upland mining groups to monitor the coastal trade centers and the arrival of vessels. Another area of interest regarding the spatial movement of materials along roads is the notion of buffer zones and its impact on the valuation of traded materials.

Trails and transportation routes are indeed important to study because they show us how societies efficiently organized space and movement according to their advantage. Trumbold argues in an evolutionary perspective that a crucial task of the archeologist is determining how political integration is reflected in the conversion of more informal routes to formal roads (1991: 3-4). Along this lines, Snead argues that the increasing formalization of the process of travel reflects evolving attempts to control or subvert other ways of travel; thus this may be a counterargument assumption that trails evolve from convenience or efficiency (2009: 43). Thus, these scholars argue that landscapes are structured with multiple narratives that we may not see if we resort to using cost minimization models that reduce informal movements into ‘insignificance, assuming that the people are culture-free’ (43-44). As will be shown below, this book is a cognizance of this risk and will only use the GIS-generated cost minimization function to delineate corridors for evaluation rather than assume the model as the actual trails themselves. These corridors will then be subjected to ‘image alarm function’ which is a tool in the GIS software that will illuminate trail segments.

Other scholars’ works are deemed relevant to this research methodologically because they were conducted in similar semi-tropical environments. In Panama, two roads linked terminal cities across the isthmus linking the Spanish colonial period markets of the Pacific and the Atlantic (Mendizabal and Theodossopoulos 2012: 102). The authors conducted research involving two distinct roads that emerged in Panama in the sixteenth century; namely: 1) the Camino Real, which is a more direct route used in gold and silver transport that cuts through jungles and were only partially paved, and 2) Camino de Cruces, which is a longer more comfortable route that winds through mixed terrestrial and fluvial areas along the Chagres River (ibid). The Camino Real measured between 1.2 and 1.5 meters in width (p. 104), which is about the same as the Northwestern Luzon trails (for instance, see Turn of the 20th century trail photos CSA 18519, 18525, 18562 Photographer Dean C Worcester).

At this juncture we will look at some of the advances in the study of trails and transportation routes in Southeast Asia. Snead, Erickson, and Darling (2009: 6) note that the studies of road networks in the Khmer region in present-day Cambodia continues to expand (i.e., Hendrickson 2008, 2013). Hendrickson uses transport geography to study the Khmer transport system for precious metals, notably iron. Because he deals with the transport mechanisms of precious metals (Hendrickson 2011: 445), albeit in a state level society, his methods are relevant to the Northwestern Luzon case in which I will be looking at the process of transporting precious metal ores from mines to bulking stations and bulking stations to coastal trading centers and how local elites appropriated the trails. Hendrickson

Strassnig (2010); on the other hand, focused his archeological survey on the Camino Real in Panama, 76

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis documenting cobbled sections, ventas (rest stations), and artifact scatters (p. 159, 165, 166). He was able to implement a survey on the trail following colonial reports closely. He notes the difficulties involved in his work, including the ‘complicated topography, the impediments of the tropical climate and vegetation, and the relative remoteness of the project area’ (p. 165). This is interesting because, based on oral tradition accounts of Northwestern Luzon, some of the ancient trails were appropriated later on by Spanish punitive expeditions (Prill-Brett and Salinas-Ramos 1998: 17). As for the difficulties mentioned by Strassnig, the case in Northwestern Luzon is slightly more promising because the tropical vegetation is limited to patches of zones between the coastal lowland and Cordillera foothills. The mountainous areas are mostly covered with tiger grass except for patches of pine trees.

exploration and experience, and knowledge incorporated into social practice, interaction, and lore (2003: 17). Rockman argues that documentary evidence shows that social and ritual environmental information stored in oral traditions will take between 200 and 400 years or more before it becomes useful for landscape learning process studies (2003: 17). As will be expounded below, oral traditions and sometimes the folklore of the peoples of Northwestern Luzon are important sources for the landscape learning process. In the case of the upland groups, Bodner’s archeological research has shown that the Bontoks have been there from as early as the seventh century. Lowland coastal communities on the other hand were occupied several millennia before those in the upland (Bellwood 2005, Solheim II 1981). Rockman (2003: 9) presents the ‘point and arrow pattern’ as one way in which colonization takes place in the landscape. This pattern is important because it allows the reconstruction of colonization when combined with appropriate dating and analytic tools (ibid). One of the applications of this methods was the analysis of American gold rush and Oregon trail migrations (p. 10). When looking at settlement sites in a newly colonized area, Rockman makes a caveat that, based on the point and arrow pattern, sites that are close to one another in terms of distance may have been occupied during different phases of colonization (2003: 10). In direct connection to this book, we have point and arrow maps of the early migrations into Benguet showing people’s migration from lowland coasts to upland, which is based on oral traditions (Bagamaspad and Hamada-Pawid 1985, Prill-Brett and Salinas-Ramos 1998). This researcher has argued that one of the major impetuses for the upland migration is the extraction of gold (Canilao 2011). Interestingly, ethnohistorical (i.e., Keesing 1962) and linguistic evidences (Himes 1998) show that lowland groups are linked to upland groups, as is the case between Pangasinan and Ibaloi as well as Ilocano and Itneg relations.

5.2 Tried and tested approaches to identifying trails In general, this book will adhere to the three-pronged approach advocated by Ferguson, Berlin, and Kuwanwisiwma (2009). The three-pronged approach utilizes aerial interpretation, ground verification, and ethnohistoric research for identifying trail segments (pp. 26-27). This process begins with individual segments before branching out into the entire trail network (ibid). Overall, the oral tradition record and written historical record give us an idea of the function of sites and what they connect. This has to be tested through archeological methods because of some disconnects when the two datasets are compared. On the one hand, oral tradition accounts mention gold mining as an important impetus to settlement locations, but when it comes to actual site reckoning, the informants point to areas nestled in fertile river valleys conducive to wet rice agriculture. As for the written historical records some maps indeed situate the settlements very close to mines. We may be able to resolve this issue using regional settlement analysis and trails analysis through GIS and systematic archeological survey.

At this juncture, it is important to point out the importance of an emic approach to understanding wayfinding as in the case of people migrating possibly for resource extraction. Snead calls for the interpretation of trails using the structures of culture and the structures of movement, or what he terms contextual experience (2009: 44). Thus part of understanding trails is approaching the process through the cultural knowledge or the perspective of the local people (ibid). The emic view becomes valuable when, for example, Snead argues that moving through a landscape is a form of ritual engagement and that trails are akin to liminal zones replete with the unexpected, danger, victory, or demise (1999: 45-46). Barton for instance calls the trails connecting Mountain Province and coastal Ilocos Sur in Northwestern Luzon as ‘head-hunting trails’ because ‘head-hunting expeditions in these regions are nearly always by small parties who attack for ambush’ (1930: 31)

At this juncture we will look at how trails can be identified archeologically. As will be shown below, the challenge is to understand the formation processes of trails as well as the processes of landscape learning and wayfinding. As interlocutors to this exercise, the researcher also seeks relevant ethnohistorical accounts, archeological findings, and even historical linguistics data. This task begins with the ability to understand the origin of trails in terms of humans wayfinding behavior and trail formation literature. Golledge (2003: 25) defines wayfinding as the ‘ability to determine a route, learn it, and retrace or reverse it from memory’ (25). This is a longitudinal process which entails learning the landscape. Rockman (2003: 12) presents the landscape learning process as a means whereby we see consistent processes of learning in colonization. Colonization here means entering a landscape for the very first time and is almost synonymous with migration. According to Rockman individuals gather information from two sources; namely, direct individual

We now turn to how trails become engraved on the landscape using the viewpoint of behavioral archeology 77

Remote Sensing the Margins of the Gold Trade and less direct. In terms of economies of scale, this longer mileage is offset by transportation technology, which is directly related to volume hauled. But of direct relevance to this book, they argue that segments of some pedestrian trails later develop into roads or are obliterated by roads (2009: 26, 40). In the case of Northwestern Luzon, some of the lowland to upland trails have become incorporated into farm to market roads, which are gravel roads, tire paths, or fully cemented roads used by jeepneys (extended cab vehicles) and trucks to haul vegetables. Ferguson, Berlin, and Kuwanwisiwma (2009) state that the repeated use of trails will eventually result in a depression on the landscape(p. 27) and will be associated with ancient villages, artifact scatters, and other evidences. In the case of the Sandia Mesa Trail in the Southwest, Snead states that ancient trails like the Sandia Mesa Trail network have resulted in some deep ruts in the landscape (2009: 50-51). Ferguson, Berlin, and Kuwanwisiwma qualify this observation, noting that if the trail is low-lying it will be subject to erosional, alluvial, or depositional processes that will obscure trail segments (2009: 28, 40).

specifically the formation processes (LaMotta 2012, Schiffer 1987). To prepare for the archeological analysis of trails, we will now look at some of the cultural and natural formation processes involved in the life-history of a trail. LaMotta has argued that understanding the material patterning in the archeological record through cultural deposition resultant from cultural processes is an important toolkit for an archeologist (2012: 86). He enumerates the different components of the archeological record as cultural materials including artifacts, ecofacts, sediments, and features (p. 70). In the case of trails in Northwestern Luzon, this researcher shows that trails in fact include all of these elements. On the one hand, there will expectedly be more sediments and features associated with the trails, but some amount of artifact and ecofact scatters may also be encountered in guard towers, watch towers, or rest stops along the trails. In terms of watch towers there are some photo plates dated to the early twentieth century that may be used as references (Turn of the 20th century ‘watch tower photos’ © Field Museum CSA 18562, 22186, 18519 Photographer Dean C Worcester). LaMotta argues that this archeological record is in fact an aggregate of traces that has been subjected to modifications by natural processes (2012: 67). With the right tools, the archeologist can study the deposits of remains associated with trails and bulk gold movement to gain insight into past human behavior and social life if such traces are preserved and sufficiently diagnostic (ibid).

At this juncture we will look at some methods proposed in the archeological study of trails. Ferguson, Berlin, and Kuwanwisiwma present a very interesting distinction of a trail from the traditional notion of an archeological site, arguing where trails are documented there may not necessarily be any manifestation of archeological materials other than the trail itself. Thus they suggest that the term ‘documentation site’ be used solely to provenience trail locations (2009: 31). They add that trail features including trail width, depth, and azimuth are used to differentiate foot trails from pack trails, and wagon roads (p. 31). Ferguson, Berlin, and Kuwanwisiwma (2009) also talk of putative guiding principles that can be used in identifying foot travel routes (trails) that can easily differentiate ancient trails from contemporary roads and naturally eroded linear expressions in the landscape (27-28). First, by way of aerial photographs (and by extension satellite imagery), trails and footpaths tend to be linear over considerable distances without any major route diversions due to minor topographic obstacles (p. 27). The second important principle they provide is that discontinuous segments of the trail will remain visible in the present as a form of fragmented preservation (ibid). Third, they state that because of heavy and long-term trampling there will be troughing (p. 28, also Sheets and Sever 1991: 69) wherein trails form depressed surfaces. Finally, they call attention to possible parallel routes (Ferguson, Berlin, and Kuwanwisiwma 2009: 28, Robertson 1983, Obenauf 1991) where trails perform as transportation corridors where people travel along a linear corridor or zone rather than walk or ride single file, in a single track. In the case of Northwestern Luzon it is quite possible that the rainy season would result in this phenomenon of parallel tracks, as the natives carve new routes parallel to older trails that have become waterlogged.

Schiffer (1987) points out that sharp gradients on the ground is caused by trampling and this is one in the many ways the patterned movement of people leave traces on the ground (126). Trampling can result in greatly compacted sediments visible on the surface (p. 205). If there were artifacts strewn along the trails, trampling action in tandem with the penetrability of the sediment can result in downward displacement (p. 127), as in the case of some long objects becoming displaced vertically for 7 to 8 cm into the deposit (p. 270) or any object obtaining a horizontal displacement of up to 85 cm (p. 127). According to Schiffer, trampling by people, beasts, or machines will crush, fragment, and abrade an objects based on its mechanical properties (1987: 22) and will eventually reduce an artifact to a stable size. This regularity lends itself to predictive modeling, which documents its effects and identifies formation processes for specific deposits (1987: 22). Schiffer adds that compaction is positively correlated to the amount of foot traffic by people and animals (p. 290), a relationship which may also work in tandem with the permeability of the soil. Permeability is determined by soil moisture, initial bulk density, and soil texture (p. 129). Ferguson, Berlin, and Kuwanwisiwma (2009) discuss a developmental history of trails wherein they begin as narrow pedestrian trails that run straight through rugged terrain, then transform into pack trails with the introduction of horses or other beasts of burden, then finally formal roads for wheeled vehicles (25-26). What is interesting is that as routes widen to accommodate pack animals or wheeled vehicles, the routes become longer

Ferguson, Berlin, and Kuwanwisiwma present thirteen criteria for identifying trail segments on the ground 78

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis 1975: 20). In the case of the lowland settlements, folklore is also an important source of information. For instance, the folklore Biag ni Lam-ang, or ‘Life of Lam-ang,’ talks about Calanutian village, which is one of the coastal trading centers in Northwestern Luzon that had direct trading relations with Southern China (Canilao 2015: 27, Eugenio 2001: 3-21).

(2009 30-31). The criteria include; ‘(1) pedestrian trails are wider that stock trails but narrower than pack trails, wagon trails, and two track roads, (2) pedestrian trails can have abrupt dog-leg turns rather than broad sweeping turns, (3) pedestrian trails will follow a straight course without deviation from topographic obstacles, (4) pedestrian trails have parallel segments [detours, repairs, rainy season traps], (5) pedestrian trails will be a series of discontinuous segments, (6) without erosion or reuse ancient trail segments will be discontinuous, (7) segments may be linear depressions with or without adjacent berms, (8) summit notches may be visible in a hilly relief, (9) troughing in heavily used trails, (10) linear drainage, soil, or vegetation pattern, (11) trails will be enhanced by artifact scatters adjacent to a trail, shrines or markers (i.e., cairns) along route, (12) historic documentation [maps, travel accounts, oral tradition], (13) ethnographic information’ (Ferguson, Berlin, and Kuwanwisiwma 2009: 30-31). These thirteen criteria are used in this book and in fact significant ethnohistorical information is readily available for the research.

There are also oral tradition accounts that are used in this book. For instance, Bagamaspad and HamadaPawid gathered the family histories of the thirteen municipalities of Benguet and combined these data with historical information (1985). Their sources have very vivid recollections about the trade between lowland and uplands. For instance they narrate, ‘[i]n Bakun, traders bartered Chinese jars in Tagudin and blankets from Ilocos for gold in Mankayan[,] with gold mined through labon [placer] the people of Panat bought hogs, carabaos, rice, salt, and clothing from Tagudin through Cervantes’ (1985: 76).They also have vivid descriptions of the actual trails in their narration that ‘[t]rade routes followed the Agno river down to Ambuclao then ascending to Pacso, ascending to Batan, then following the ridge of the Halsema-Mountain Highway to Betdi, descending into the lowlands also through Banget-Pico and Chuyo’ (1985: 77). These accounts also cover the Spanish contact period; for instance, Bagamaspad and Hamada-Pawid state that the construction of trails increased at Spanish contact with some having forced labor or polos (1985: 177). The increased trail construction in Mankayan for example was related to the increased mining activities of the Sociedad Minero-Metalurgica Cantabro Filipino in 1890 (ibid). Prill-Brett and Salinas-Ramos also talk about the trails based on accounts of their non-elite informants, ‘… originating from coastal Aringay, passing through Galiano-Ampusa-Betwang, Yagyagan, Pinalyog then into Pugis’ (1998: 17).

Once again, the ethnohistorical record was used for ‘leads’ to carry out regional settlement archeology with GIS and ground verification. A good portion of the ethnohistorical data that were used in this research are based on travelers’ accounts. Apple (1965: 2), who studied Hawaiian chiefdom trails also emphasizes that travelers’ records are important because they give us an idea of the technology of transport and trail location, as well as the method of construction. In the case of Northwestern Luzon uplands, there are several travelers’ account that are tapped in this book (i.e., Galvey 1842, Semper 1861 [1975], Meyer 1883, Barton 1930, Gironiere 1854). These accounts are rich with descriptions of trails and what villages they connected in the highlands. Some of the accounts have already proven useful in this researcher’s pilot study. In locating Aringay’s coastal trading center, Tonglo bulking village, and Acupan/ Balatok mining settlement, some early travel accounts are used as clues. In Meyer’s 1883 account he writes that he traveled by boat from Manila to San Fernando, La Union, then transferred to carabao carts that brought them to Aringay. He narrates, ‘then the road continued on between the huts of a few pueblos to Aringay, where the trail begins for the mountainous hinterland of the district of Benguet’ (Meyer 1883 in Filipino Book Guild 1975b: 50). During Galvey’s conquest of Benguet he provides a description of Tonglo, writing that ‘[f]inally at 1 in the afternoon we reached the bottom of a ravine, where we found a river called the “Cagaling,” which is the same stream that flows past Aringay and has its source on Mount Tonglo’ (Galvey 1842 in Scheerer 1905: 175). Semper also writes about the Acupan/Balatok mining settlement in 1861, saying that ‘[f] or several days the trail led me along the bed of the Agno valley as well as along the tops of mountains at 4,000 to 5,000 feet through various unimportant Rancherias to Acupan[,] this village is situated on the western bank of Agno, which is very wide here, at an altitude of 5,070 feet on the back of a mountain crest extending north from Mount Alan or Alut’ (Semper 1861 in Filipino Book Guild

Before proceeding to the GIS methodology to be used in this book, it should be emphasized that identifying and locating the different settlements in Northwestern Luzon including mining settlements, bulking villages, and coastal trading centers is of primary importance specifically because created GIS shapefiles are then subjected to various GIS analyses based on some ethnohistorically known aspects of the mining-associated natural and cultural landscapes at particular points in time. Although this is dealt with in detail previously, in summary the primary historical sources include early ethnographies, travel accounts, and early pronouncing dictionaries. Turn of the twentieth century photo plates and Spanish maps are also used to plot the locations in the book. The published oral traditions (Bagamaspad and Hamada Pawid 1985, Prill Brett and Salinas Ramos 1998) of the indigenous groups of upland Northwestern Luzon will have also been of immense utility in locating sites. In terms of secondary sources, there are also a number of scholars who have looked at various sources and compiled a list of settlement locations including maps (i.e., Mateo 2004, Newson 2011, Scott 1988). 79

Remote Sensing the Margins of the Gold Trade of incorporation, etc.) can be accessed through regional settlement analysis using GIS and satellite imagery. According to Snead, Erickson, and Darling one of the more important functions of GIS in archeology is that it provides predictive modeling of trails, paths, and roads using characteristics of the topography as inputs when physical evidence for these features is lacking (ibid). They cite some of the significant works (i.e., Madry and Crumley 1990, Madry and Rakos 1996) that adapted cost pathways analysis for archeology (2009: 11). Of particular interest is the work of Kantner on Chaco Roads of Lobo Mesa, New Mexico where he tested the orientation of the trails against a rational pattern of travel (Snead, Erickson, and Darling 2009: 12, citing Kantner 1997). Kantner found out that the actual features were a poor fit to the model, allowing him to determine that these trails were more than just functional routes (Snead, Erickson, and Darling 2009: 12). Bell, Wilson, and Wickam (2002) were also able to use GIS methodology for reconstructing routes of communication between ancient sites in Sangro Valley, Abruzzo, Italy. Madry and Rakos were able to prove the immense utility of using GIS specifically for line-of-sight and cost-surface, and for optimum corridor selection modules in understanding the location of Arroux River Valley hillforts and known roads (1996: 104). Their ‘results show a strong correlation of these routes with the visibility from the hilltop defenses’ (ibid). The authors also argued for the value of an emerging new generation of least cost path GIS analysis (i.e., Harris 2000, Llobera 2000), where functional protocols will be used for more theoretical orientations (2009: 12).

Mateo (2004), for instance, enumerates the Northwestern Luzon coastal trading centers: Alinguey (Aringay), Baratao (she states as Agoo or Bauang but Newson submits is San Fernando), Tagudin, Dumaquaque (Santa Lucia), Candon, Narvacan, Vigan, Bantay-Bantaguey, Panay (Magsingal), Cabugao, Sinait, Barao (Badoc), Bacarra, Ballecilio (Bangui); and within river valleys were Purao (Balaoan), Vintar, Danglas, Ylagua (Laoag), Bonsan (San Nicholas), Cacabayan (Paoay-Batac) (p. 85). While site locations are inferred from historical records, these points and surrounding landscapes will involve assigning exact shapefile points on the GIS map. Of direct relevance to this challenge is the methodology employed by Victor Paz in his archeological projects on Mindoro Island (2005, 2006). He was able to show that Early Historical Period settlements can be identified in the landscape based on where Early Spanish contact period churches were erected. This was confirmed in the cases of three churches; namely, Bancuro, Bongabong, and Bulalacao (2006: 12). In addition, in this section excavation data will be revisited, including human remains, porcelain, trade beads, and gold from the researcher’s 2011-2012 project in upland and lowland Ilocos Sur province (Canilao 2012a, 2011) as well as survey data including earthenware, stoneware, tradeware sherds, and ecofacts from 2008 to 2012 survey visits in different parts of Benguet (Canilao 2012b, 2008). I also review Northwestern Luzon Early Historical Period archeological sites that were subject previously to archeological investigations (Bodner 1986, Tidalgo 1979, Caballero 1996, Legaspi 1974, Maher 1980, 1974). As a general approach to Cordillera ‘blufflands’ archeology (Canilao 2013), I utilize systematic surface archeology with focused test pit excavations on a case by case basis, correlating surface and subsurface data.

Satellite images have been successfully used to identify trails in some regions. Satellite images have been used to remotely sense ancient trading trails in Chaco Canyon (Sever and Wagner 2011). Sever and Wagner started with the objective of examining the relationship between surface and subsurface archeological features and how this affected the bandwidth on multispectral sensors (Sever and Wagner 2011: 50). Using bands 1, 2, and 3 of the Thermal Infrared Multispectral Scanner, the prehistoric roadways as well as major walled structures and agricultural fields could be seen (2011: 51). Satellite images have also been used in the analysis of ancient trails in the Near East (Kouchoukos 2001). My book is ground-breaking in a sense that it uses the newly available high-resolution imagery for Luzon Island in the Philippines (i.e., WorldView-2, WorldView-3) to delineate trails in this mountainous, volcanic island terrain where trails or pathways can be much narrower with gold trade trails probably measuring one to three meters wide (for instance, see Turn of the 20th century trail photos CSA 18519, 18525, 18562 Photographer Dean C Worcester).

5.3 Geographic Information Systems and Its Utility in Modeling Trails and Transportation Networks At this juncture, we will look at the GIS methodology that will be used in this book. Trumboldt argues that an aerial perspective is the earliest and most effective way of discovering and verifying trails on the ground (1991: 7). Ferguson, Berlin, and Kuwanwisiwma enumerate the features of a trail that will become visible using aerial analysis: (1) linear depressions, or swales after shadow enhancement, (2) anomalous drainage features also after shadow enhancement, (3) soil discoloration, (4) moisture marks, and (5) anomalous vegetation patterns, notably linear patterns caused by type, density, or height compared to ground (2009: 28). Snead, Erickson, and Darling (2009) state that remote sensing, notably satellite imagery, has revolutionized the study of landscape of movement, particularly the analysis of trails (p. 11). When satellite imagery analysis is combined with GIS modeling the research can already integrate interpretive models into the discussion of landscape uses (ibid). In the case of this book it is argued that the models discussed in the theoretical section (i.e., world system analysis, negotiated peripherally, continuum

The work of Sheets and Server (1991) is significant because they used remote sensing aerial photography and a Thermal Infrared Multispectral Scanner in tandem with ground verification in locating prehistoric footpaths in the tropical forests of Costa Rica. The remotely sensed footpaths exhibited some crop mark features associated 80

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis in terms of energetics and time when transporting goods down the mountains.

with the prehistoric trails in some pasture lands and forest areas as well (p. 62). They add, ‘some of the paths can be seen and followed in the field, as linear depressions, particularly where they transverse slopes steeper than about 3 to 4 degrees’ (1991: 62). In the same project they suggest that laser profiler LIDAR will be able to detect dips in the terrain that is caused by footpaths (p. 63). Later they conclude that the prospects for looking for ancient footpaths in tropical moist environments is possible using aerial photography (63). Again, I would argue that topographic and ecological features in Northwestern Luzon are slightly more conducive to trails analysis, especially the zones between mining areas and bulking villages in the highlands. Only small patches of deciduous forests are found in the narrow foothill corridor so ground features like trails are more visible.

The resulting least cost path model was then overlain on image-enhanced high-resolution (pansharpened at 40 cm), multispectral (8 bands) WorldView-2 satellite coverage of the study area. Image enhancement techniques were done to prepare the 1000-square-kilometer satellite scenes using Brovey transform (Panchal and Thakker 2015). But the analysis does not end here, as the researcher then implemented a buffer function which delineated a corridor along the least cost path line. It is within this corridor that the image alarm function was used to illuminate trail segments based on the spectral signature of a known trail. Based on the case study, there are segments of the trail that coincide with the least cost path model, and this is the subject for ground verification. Writ large, however, the book includes several regional-scale analysis in northern Benguet, Mountain Province, Abra, Ilocos Sur, La Union, and Pangasinan (provinces in Northwestern Luzon), allowing comparisons between gold producer strategies in different regions with different topographies and tactical concerns.

At this juncture, it should be noted that the researcher will be following the same GIS protocols deployed in the pilot study of the Aringay–Tonglo–Balatok/Antamok site (Canilao 2017). In this pilot study the researcher conducted viewshed analysis from Tonglo bulking village towards the coast to determine if the location of Tonglo allows for visual spotting of arriving merchant vessels by upland producers. Because of its elevation at 1165 meters above sea level and its Euclidean distance of 20 kilometers from Aringay, the analysis shows that an observer in Tonglo is strategically situated to spot merchant trading vessels measuring an average of 15 m in hull length.

After all remote sensing techniques have been exhausted, the grand finale of a trail research is the culminating ground verification that follows. Trumbold (1991) has quite rightly argued that ‘discussions regarding theory, analytical techniques and network configuration are quite useless unless the networks can be known empirically’ (p. 7). The final step is to conduct ground verification of the trail segments that have been identified. The ground truthing is done to check these trails if they contain artifacts and ecofacts that are associated with the gold trade from the historical periods of interest. In certain areas where the terrain is not permissive, the researcher may utilize a professional drone to fly over the terrain and document the site. This is planned to be implemented as post-doctoral research.

A Least Cost Path model was also implemented to analyze relatively rapid descent pathways to the coast. The first step was to calculate the slope values for every 30-squaremeter area using the ASTER digital elevation model. The resulting slope table was then converted to hillshade for visual effect. The raster has an attribute table per 30 square meters of terrain with values corresponding to slope which correspond to effortless abrupt elevation changes which correspond to less effort and are thus optimal. The resulting raster is then processed using the cost distance tool. This tool calculates cost of effort for every 30-squaremeter area coming from a single point or source in the terrain. The cost distance was calculated for both Acupan/ Balatok and Tonglo. The resulting cost distance raster was then processed using the cost back link tool. This tool will provide cost-effective directional routes from one point or source in the raster. Again, the cost back link was calculated for both Acupan/Balatok and Tonglo. Finally, the resulting raster was subjected to the culminating least cost path analysis where the intersection of the two datasets was to provide the most efficient route from the source to a target destination in the raster. The cost path was calculated for Acupan/Balatok to Tonglo, then from Tonglo to Aringay. The least cost path from the highlands to the coast shows a path that snakes through the rugged Southwestern Cordillera terrain. Once again the resulting route is the most optimal route in terms of time and effort. The route features the most gradual transition in terms of elevation (versus abrupt change). This gradual transition is likely to have been preferred by upland gold producers

5.4 Geographic Information Systems and Remote Sensing Methods for the Tonglo Region Two models were created for the Balatok–Tonglo–Aringay gold trading trail. The ethnohistorical model is based on the 1829 account of Don Guillermo Galvey, who was the first to suppress the upland Ibaloi into submission to the Spanish colony (Fig. 5.1) (Scheerer 1905: 173-178). The least cost path model on the other hand is based on ASTER global DEM. What is immediately apparent is the disconnect between the ethnohistorical model and the GIS model. What can be stated immediately is that the trail based on ethnohistoric accounts seems to be directed towards Pugo, which is a bulk breaking point in the upland trip in the Aringay River system. The reason for this disjunction between predictive models based on ethnohistory and actual results will be discussed later in the book and in the conclusion section. A third GIS model that was created is the visibility model 81

Remote Sensing the Margins of the Gold Trade

Figure 5.1. Map (1837) showing 1829 route of Guillermo Galvey (Courtesy of Archivo General Militar de Madrid) (Canilao 2017a).

from Tonglo gold bulking village (1,161 meters above sea level) to Aringay coastal settlement which is 16 kilometers to the west. This model was to test the contention that the gold miners were sighting the vessel arrivals at the coast in order to time their arrivals for the ensuing evanescent market encounter. The analysis shows that Tonglo village as a gold bulking center has very good visibility of Aringay. A three-dimensional rendering of the view was also recreated from Tonglo (Fig. 5.2, Digital Appendix). Indeed, the Tonglo observer would be able to spot vessels traveling at 1 to 2 knots en route to the Aringay coastal maritime trading center (Fig. 5.3, Digital Appendix). Timing is important for the tabu-tabuan/tiangge or evanescent market to become successful. For the Ibaloi gold miners who have direct access to the gold lode veins in Southwestern Cordillera, they have to come down from the mountains as soon as they spot vessels already at or approaching Aringay; thus the Tonglo gold bulking center needed to be situated strategically to sight the merchant ships. Interestingly, the amount of time it takes a trading party with bulked gold to undertake pedestrian transport from Tonglo to Aringay is around half a day based on Guillermo Galvey’s account (see Scheerer 1905: 173-180). Now, at this juncture, I will discuss the implementation of image enhancement on WorldView-2 satellite imagery to remotely sense trail segments (Canilao 2017a).

103001001F156900, 10300100218C3A00). The three satellite scenes had to be prepared through the pansharpening tool of Erdas Imagine 2016 software. The multispectral layer was subset into band 1 (coastal) alone, bands 2 to 6 (blue, green, yellow, red, red edge) stacked, and bands 7 and 8 (NIR1, NIR2) stacked. Band 1 and bands 7 and 8 were resampled to 50 cm. Multispectral bands were first resolution merged with the panchromatic band and later with the three resampled bands 1, 7, and 8. The pansharpened scenes were then subjected to unsupervised classification using four indices namely: WV2 New iron index, WV2 built-up index, WV2 soil index, and WV2 improved vegetation index. Data fusion (2006: 62) was then carried out whereby the four datasets were layer stacked together and viewed in a 2-dimensional red– green–blue color composite. The ethnohistorical model and the GIS predictive model were then overlain on the image-enhanced WorldView-2 scenes and various band combinations were tested to see if trail segment spectral signatures would be enhanced. Three particularly useful combinations include the following: mining operations (yellow–NIR1–red edge), modified false color infrared (NIR1–green–blue), and soils and constructions (red– blue–yellow) (Canilao 2017a). A closer examination of least cost path routes shows that the model lay adjacent and parallel to present-day trails along the route from Balatok mines to Tonglo bulking village. This is especially true for segments within the Balatok to Tonglo route found in 51Q 244369mE 1810261mN (Fig. 5.4, Digital Appendix) and 51Q 247895mE 1810109mN (Fig. 5.5, Digital Appendix). Scenes containing these

A 1,000-square-kilometer area covering the Balatok– Tonglo–Aringay trail was provided by DigitalGlobe Foundation. These were culled from three highresolution, panchromatic, and multispectral WorldView-2 satellite scenes (Catalog IDS; 103001002D7C2F00, 82

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis enhancing WorldView-2 imagery (Panchal and Thakker 2015) (Canilao 2017c).

trail segments were enhanced using the mining band combination. As for the Tonglo bulking village to Aringay coastal settlement model, there are no approximate trails both past and contemporary. A closer examination of the ethnohistorical model, however, reveals that there are some past trail segments that scale the peak of the same mountain that Don Guillermo Galvey used in his 1829 expedition. The WorldView-2 modified infrared combination also enhanced the spectral signature of a trail located in 51Q 229500Me 1809445Mn. Here we can clearly see how the trail has become associated with secondary growth vegetation that grows in a uniform pattern (Fig. 5.6 compare to Fig. 5.7) (Canilao 2017a). These trail segments were enhanced using several iterations of the data-fused WV2 New iron index, WV2 built-up index, WV2 soil index, and WV2 improved vegetation index. The most enhancement seems to be seen in the built-up index, improved vegetation index, and improved vegetation index combination (BU i-IV i-IV i). Two segments of the trails were remotely sensed located at 51Q 230821mE 1812529mN (Fig. 5.8, Digital Appendix) and 51Q 229146mE 1806792mN (Fig. 5.9, Digital Appendix).

The resulting pansharpened image was subsequently layer stacked back to the resampled layer 1 and layers 7 and 8. The resulting pansharpened image was then converted to WV2 built-up, soil, vegetation, and iron indices separately. The four indices were then layer stacked following the theory of data fusion (Kvamme 2006) (Fig. 5.10). Topographic maps were finally mosaiced together and georeferenced using ERDAS Imagine 2016 software (Hexagon Geospatial, Madison, AL, USA). ArcMap 10.4 (ESRI, Redlands, CA, USA) Spatial Analysts tools were used on the digital elevation model of the network specifically 1) visibility and 2) least cost path modeling. These are based on algorithms created by Environmental Systems Research Institute (ESRI). For the least cost path model the source destination was set at Gasweling settlement, then a cost distance algorithm was run together with a backlink algorithm. The two resulting rasters were then used in running a cost path allocation model with source coming from the various mines and coastal settlements. These models were then superimposed on various band combination enhancements of WorldView-2 imagery for remote sensing of archeological features. For the visibility analysis, ten ships were plotted at the sea adjacent to the coastal settlement (Fig. 5.11). Using the frequency option, visibility from ten ships was run within an outer radius of 50000 meters and starting angle of 40 and 110 degrees. All three coastal settlements were subjected to visibility analysis (Canilao 2017c).

5.5 Methods used for Gasweling Region A 1,000-sq.-km WorldView-2 satellite archived imagery covering the Gasweling network was provided by the DigitalGlobe Foundation. ASTER global Digital Elevation Model was then acquired from the Global Data Explorer (GDEx) of the Land Processes (LP) Distributed Active Archive Center (DAAC) and George Mason University’s Center for Spatial Information Science and Systems. National Mapping and Resources Information Authority (NAMRIA) topographical maps both in 1:50000 and 1:250000 scale were also acquired. Finally, shapefiles were obtained from the open access PhilGIS website for educational use including 1) country shapefile and 2) river shapefiles (Canilao 2017c).

5.6 Methods used for Lepanto Region The researcher was awarded an imagery grant of 1,000 square kilometers of archived WorldView-2 satellite imagery of the Lepanto gold district and Ilocos coastline from the Digital Globe Foundation. The WorldView-2 images are in unprocessed (not image-enhanced) multispectral and panchromatic form. The area came in four scenes following an irregular polygon shape (Catalog IDS; 1030010032CE2500, 103001002E6E8000, 103001000E9FCA00, 10300100218C3A00). The researcher also downloaded a shapefile of rivers from the PhilGIS website. It should be noted that this shapefile only delineated major river systems and not the smaller streams and creeks. The researcher also downloaded the relevant digital elevation model from the LP-DAAC GDEx website of the USGS. This geoTIFF file was converted into a raster that was projected in UTM zone 51. Heads-up digitizing of Lepanto, Purao, Tagudin, Danac, and Apayao was done following maps in historical sources mentioned above and the location of contemporary townsites (Canilao 2017b).

The WorldView-2 satellite image came in both 2-meters multispectral and 0.5-meter panchromatic form. The scenes were first converted from .tiff to .img format using ERDAS Imagine 2016 (Hexagon Geospatial, Madison, AL, USA). The scenes were then clipped to the area of interest (AOI). There are several techniques that can be used to enhance satellite imagery, with the three main types of enhancements being spatial, radiometric, and spectral enhancement. Here we use a type of spatial enhancement that employs edge detection technique by pansharpening (resolution merge) multispectral bands with the panchromatic image. The multispectral scene was subset into layer 1 alone and layers 2 to 6 stacked and layers 7 and 8 stacked. Both layer 1 and layers 7 and 8 stacked were resampled to 0.50 cm following the affine option and using the bilinear method. The 0.5-meter panchromatic image was then resolution merged with the 2-meter resolution bands 2 to 6 using the Brovey transform method. The Brovey transform method has been proven to be a good option when image-

Suitability modeling was undertaken for the Lepanto region (Figs. 5.12-5.25, Digital Appendix). This was done in this particular network because of the unique feature that both an evanescent market and a permanent market situation are evidenced in the nodal articulation of this 83

Remote Sensing the Margins of the Gold Trade

Figure 5.6. Trail spectral signature enhanced with WV2 modified infrared combination 732. Note the vegetation associated with the trail (WorldView-2 images courtesy of Digital Globe Foundation) (Canilao 2017a).

network. Using ArcGIS 10.4 (ESRI, Redlands, CA, USA) the researcher ran slope and hydrologic modeling on the DEM. The resulting raster was then used to delineate drainage basins. Owing to their relatively short area the drainage basin leading to the two coastal villages was merged into one, thus resulting in a Lepanto basin and a combined Purao and Tagudin basin. Accumulated cost was calculated from three locations and it was seen to coincide with the ridgeline between the two drainage basins. This raster will be used later in the book where a least cost path predictive model of the trading trail will be compared to ethnohistorical models. The next step was to digitize the ridge line between the two basins. A Thiessen polygon

was also implemented in the study area. The next step was to digitize the Thiessen mid-line between the mines and coastal villages. The next step was to run multiple ring buffers on the three variables: digitized river shapefile (water source), the digitized ridge line shapefile, and the digitized Thiessen line shapefile (Canilao 2017b). Ten ships were also plotted on the horizon for each of the two coastal villages. The ships were distributed following a hexagonal pattern near the coastal settlements. The GIS visibility function was undertaken, and the resulting color ramp indicated areas where an uninterrupted view of all ten ships was available to the viewer (Tagudin coastal). 84

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis Reclassification was done on the value attributes of the datasets following a 1 to 5 ranking scale with 5 being the highest. Finally weighted overlay was conducted, where all variables were given equally distributed weights of 33 per cent. The next step was to overlay the resulting suitability rasters on WorldView-2 imagery that were pansharpened using the Brovey transform method in ERDAS Imagine 2016 software (Hexagon Geospatial, Madison, AL, USA). The areas with scale values of 5 (highest) were given preliminary investigation. The first round included Tagudin visibility, distance to ridge line, and distance to water. The second round included Tagudin visibility, distance to Thiessen line and distance to water. The third round included Purao visibility, distance to ridge line, and distance to water. The fourth round included Purao visibility, distance to Thiessen line, and distance to water. A fifth round was included where distance to water and Thiessen line were the only variables equally weighted at 50 per cent each (Canilao 2017b). 5.7 Methods for the Angaqui Region Image enhancement was implemented on archived WorldView-2 satellite imagery courtesy of the DigitalGlobe Foundation. The WorldView-2 satellite image came in both 2-meter multispectral and 0.5-meter panchromatic form. The scenes were first converted from .tiff to .img format using ERDAS Imagine 2016 (Hexagon Geospatial, Madison, AL, USA). The multispectral scene was subset into layer 1 alone and layers 2 to 6 stacked and layers 7 and 8 stacked. Both layer 1 and layers 7 and 8 stacked were resampled to 0.50 cm following the affine option and using the bilinear method. The 0.5-meter panchromatic image was then resolution merged with the 2-meter

Figure 5.7. Note example of a ridge trail (© The Field Museum, CSA 21976. Photographer: D.C. Worcester) (Canilao 2018a).

Figure 5.10. Data fusion of WorldView-2 indices (workflow) using ERDAS Imagine 2016 (Hexagon Geospatial, Madison, AL, USA) (WV2 imagery courtesy of DigitalGlobe Foundation) (Canilao 2017c).

85

Remote Sensing the Margins of the Gold Trade

Figure 5.11. Least cost path algorithm (workflow) using ArcMap 10.4 (ESRI, Redlands, CA, USA) (ASTER GDEM is a product of METI and NASA) (Canilao 2017c).

resolution bands 2 to 6 using the Brovey transform method (Panchal and Thakker 2015). The resulting pansharpened image was then layer stacked back to the resampled layer 1 and layers 7 and 8. Topographic maps both 1:250000 and 1:50000 were acquired from the NAMRIA and were also georeferenced and mosaiced using ArcMap 10.5 (ESRI, Redlands, CA, USA). This in turn was used for comparison with Spanish historical maps. Trails were delineated on the base map based on the historical maps.

map was a relatively small-scale map and needed some georeferencing. After georeferencing this mineral resource map, the next step was to digitize these explored and developed mines in GIS and to also overlay these on the WorldView-3 satellite imagery in preparation for remote sensing (Canilao 2018a). The corresponding ASTER global digital elevation model (ASTER GDEM product of METI and NASA) was downloaded using the Global Data Explorer tool (United Stated Geological Survey, Department of the Interior). This raster dataset was used in running least cost path predictive modeling from the digitized settlements. The output raster map was then digitized into trail polylines that were overlain on the digital elevation model (Fig. 5.28, Digital Appendix). Remote sensing was then carried out on the WorldView-3 satellite Imagery using the digitized features as guides (Canilao 2018a).

5.8 Method Used in Abra Region At this juncture, we focus on the GIS and remote sensing method that is a complementary method for ethnohistorical archeology. The first step was to create a working basemap for the Abra network. This was carried out using ERDAS Imagine 2016 (Hexagon Geospatial, Madison, AL, USA) by mosaicking and georeferencing 1:50,000 topographic maps from the National Mapping and Resource Information Authority website. The next step was to process the five available historical maps to derive information on the Abra settlements as well as its trail connections (i.e., Figs. 5.26, 5.27).

That we know for a fact that old trails and routes in Abra can be retraced is something mentioned in some historical sources. For instance, Keesing states that Until the development of modern roading, a day’s journey by raft or horseback was required to get from the coastal area through this gorge [Banauang Gap] into the open river flats of the Abra valley. (1962: 119)

At this juncture, it is important to state that unlike the other case studies in Northwestern Luzon there are little to no historical records and accounts that indicate where early gold mining in Abra was carried out. What is available is a small section of the Mineral Resources Map of the Cordillera Autonomous Region (Mineral Lands and Topographic Survey Section, Mine Management Division, Mines and Geosciences Bureau), which includes a plot of explored and developed mines in Abra. The

Later he explains that Trails lead by way of some of their headwater streams into Lepanto, Kalinga, and Apayao. They also cross from the middle Abra valley westward to the coast at 86

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis Narvacan and other points, and form North Abra into Ilocos Norte. (119) With regards to the Cabugao/Calanutian–Tineg–Malague trail a total of three trail segments were remote sensed (Figs. 5.29-5.31, Digital Appendix). With regards to the Bangued/Tayum–Tineg–Malague trail, which was followed by Guillermo Galvey, a total of two were remote sensed (Figs. 5.32, 5.33, Digital Appendix). With regards to the routes from Bangued/Tayum, Bucao to Narvacan, a total of three were remote sensed (Figs. 5.34-5.36, Digital Appendix). Of interest at this point are some potential ijangs/ili or hillfort defenses of the Abra network. It is interesting to note that two ijangs may have protected the main settlement of Vigan–Caoayan. Southwest of Vigan is Balaurte Hills/ Tamag Hills (Fig. 5.37), whereas northeast of Vigan is the Bantay Hill (Fig. 5.38). That the name itself is Baluarte is interesting because its literal translation is ‘place of refuge/safe haven.’ I would argue that the baluarte function, just like an ijang, is to fulfill both refuge and defense requirements. Bantay, on the other hand, function as an ijang that protected the villagers of Vigan–Caoayan– Bantay during the encounter with Juan de Salcedo and his troops (Keesing 1962: 122-123).

Figure 5.26. Trail connecting Cabugao coastal settlement with Malague in Cagayan valley, which is approximately 95 kilometers in terms of Euclidean distance; illustrated in a portion of Mapa que dumuestra la situacion de la Provincia de Cagayan. Unknown Cartographer, Undated (Courtesy of the National Archives of the Philippines) (Canilao 2018a).

But what is more notable based on a close reading of the historical accounts is that Bantay is more of a naval stronghold. With the Vigan settlement located across the Govantes–Mestizo River from Bantay, Isabelo delos Reyes states that Bantay is an old pueblo that had facilities

Figure 5.27. Guillermo Galvey route passing from Bangued to Tayum to Tineg then to Malague highlighted in orange line (Courtesy of Archivo General Militar de Madrid) (Canilao 2018a).

87

Remote Sensing the Margins of the Gold Trade 5.9 Modeling reduccion and Trade Connections through Social Network Analysis

to shelter sampans or birays and other smaller boats (142). Bantay came replete with a small para-naval fleet as can be gleaned by incidental intelligence (Scott 1982) on the account of Isabelo delos Reyes (2014). Reyes mentions that the naval force of Salcedo was intercepted by two native vessels for identification. When they were identified as hostiles, warning was sent out and so warriors were already standing at alert, waiting at the Govantes–Mestizo river banks as the force of Salcedo approached Vigan– Bantay (Reyes 2014: 141-142) (Canilao 2018a, 2018b).

At this juncture we explore how environmental variables in a landscape can account for resulting social connections and relationships. We are using the northeastern branch of the Abra Network as a case study. To carry out the analysis I digitized the environmental features of this quadrant of Abra specifically and delineated the ridgelines which outline the water basins. An 1885 map of the area was also digitized showing the varying degrees of colonization of the towns in the area. The analysis shows that the ability to remain free from colonization was enhanced by the environmental and geographic contexts of the towns vis-avis the drainage systems of the area. This hinterland ijang baluarte concept was explored in Canilao (2019) using remote sensing.

In the Cabugao–San Juan area, the Refaro hill may have served as an ijang. This particular ijang may have been a habitation site as well since some ceramic sherds including earthenware and tradeware pieces were seen in the area during the survey of 2011. That ijangs with ceramics can be evidence for habitation was established in the Batanes island group in the works of Bellwood and Dizon (2013) as well as Lacsina (2009). Finally, also worth further exploration would be the case of the Santa Maria settlement. It is interesting to note that the current ijang or Baluarte (Spanish period church) is facing the flank towards the mountains. There are, however, two hills that may have served as coastal ijangs towards the west of Santa Maria worthy of closer investigation (Fig. 5.39) (Canilao 2018a, 2018b).

The first step is to prepare the 1885 map of Abra entitled Mapa Geografico del centro de la Abra en la Provincia de Ylocos (Fig. 5.40, Digital Appendix), paying particular attention to the legend which shows that the towns have been classified into a) pueblo de christianos (Christian village); b) pueblo de ynfieles de quarenta tributos en Adelante (village of infidels with 40 tributes or more); c) ynfieles de veinte Adelante (20 tributes or more); and

Figure 5.37. Baluarte/Tamag Hills Southwest of Vigan encircled in yellow (Using WV3 soils and construction combination, red-blue-yellow) (Courtesy of the DigitalGlobe Foundation) (Canilao 2018a).

88

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis

Figure 5.38. Bantay Hills Northeast of Vigan encircled in yellow; note Govantes-Mestizo River between Vigan and Bantay where birays and smaller vessels could dock according to Isabelo delos Reyes (Using WV3 soils and construction combination, red-blue-yellow) (Courtesy of the DigitalGlobe Foundation) (Canilao 2018a).

d) Ynfieles de ocho en Adelante (8 tributes or more). The focus will be the northwestern quadrant of the map. Thus, close scrutiny is given to the identified villages. To proceed with georeferencing the villages, the first step is to re-project 1:50000 scale topographic maps of the area that are based on aerial photography in the 1960s to 1970s. Most of the villages in the 1885 map correspond with the village names in the topographic map; thus it was easy to plot points for the analysis. There are, however, some villages whose 1885 names are not plotted in the map; thus there is a need for closer scrutiny. In some case these names are just spelled differently, probably reflecting some Spanish influence. In other cases there is a need to look into barangay or municipal center histories in order to find out the original names of these towns (Canilao 2019).

The digitized village location points are then overlain on a processed digital elevation model of the area. The DEM is converted into a slope raster (Fig. 5.41, Digital Appendix), then least cost path analysis is run from each village to other villages (Fig. 5.42, Digital Appendix) that are within the same basin. In terms of the directionality, the flow direction is followed (Canilao 2019). All the resulting values in the least cost path analysis are then encoded in a UCINET matrix in preparation for social network analysis. The layout was customized to follow the scaling/ordination option and was run in both similarities, which shows strength of ties (Fig. 5.43), and dissimilarities (Fig. 5.44), which shows distances suboptions with adjustment to nearest Euclidean distance unselected (Canilao 2019).

A digital elevation model of the area is also downloaded for use in the GIS analysis. This DEM is reprojected to UTM WGS 84 to match other layers used in the analysis. A slope analysis is run on ArcMap 10.5, which converts each 30 by 30 m land area to its slope value. This is then used to run hydrology analyses that would delineate ridgelines and drainage basins in the area in connection to flow accumulation of water systems from ridges to the sea (Canilao 2019).

The similarities model seems to show intra-village connections and relationship that come into play to facilitate trade and exchange. The model shows Padangitan and Lapaz as conduits in the network and in fact they are the bulking villages located at the intersection of the main Abra River system and its smaller northeastern branch. The model also shows two subnetworks of the northeastern branch with the first inclusive of Lacub, Bumalicong, 89

Remote Sensing the Margins of the Gold Trade

Figure 5.39. Arrow shows ijang facing the interior which is the site of the Church of Our Lady of Assumption built 1765. Encircled are the possible coastal ijangs for further exploration (Using WV3 soils and construction combination, red-blueyellow) (Courtesy of the DigitalGlobe Foundation) (Canilao 2018a).

5.10 Chapter Summary

Paganao, and Pacqued, while the second group includes Caganayan, Anayan, Agsimao, Tineg, and Callago. Again the model plays out well in terms of taking into account the flow accumulation model generated for the Abra basin. This Abra case would be a good case where an environmental variable relates to a social variable namely trade and exchange. Thus, the river transportation system via rafts and its bulking stations would be the organizing principle in terms of trade with a movement upriver showing transition from trade hubs to extraction areas (Canilao 2019).

This chapter of the book has demonstrated how GIS analysis in tandem with remote sensing methods has woven separate patches of information derived from written historical accounts, oral tradition accounts, and archeological works into a palimpsest that has facilitated the macro-scale analysis at the region level. After presenting individual site data it is ultimately the regional analysis that will allow us to talk about the shifts from evanescent to permanent markets. The next chapter is a conversation made at the regional scale and will feature a discussion on how these five networks that form this larger Northwestern Luzon network ultimately coalesce as the region’s articulation to the IO–WPS/SCS trade emporium becomes solidified.

As for the dissimilarities model, this seems to jive with the march of reduccion in the area. And this is supported by the 1885 historical map. It can be seen in the map that the satellite settlements Atip, Callago, Agsimao, Tineg, Anayan, Lacub, Bumalicong Caganayan, Paganao, and Padangitan are classified as villages with the least number of tributes (which is eight), whereas Patqued, Lapaz, and Lagayanhave have at least forty tributes (Canilao 2019).

90

Methods in Identifying Transportation and Probable Transportation Routes through Spatial Analysis

Figure 5.43. SNA similarities model using UCINET software. Matrix of least cost path distances in between the two models (UCINET software attribution: Borgatt, S.P. Everett M.G., and Freeman, L.C. 2002. Ucinet 6 for Windows: Software for Social Network Analysis Harvard MA: Analytic Technologies).

91

Remote Sensing the Margins of the Gold Trade

Figure 5.44. SNA dissimilarities model using UCINET software. Matrix of least cost path distances in between the two models (UCINET software attribution: Borgatt, S.P. Everett M.G., and Freeman, L.C. 2002. Ucinet 6 for Windows: Software for Social Network Analysis Harvard MA: Analytic Technologies).

92

6 Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing miners’ most optimal path, which offered least effort when bringing mined gold nuggets to the bulking village through pedestrian transport. It should be noted that looking at the high-resolution imagery, there are no proximal presentday trails that coincide with this route. This is not to say, however, that the route did not exist but maybe that a future second order analysis using a drone for even better resolution may provide additional data. Segments of past and present-day trails, however, were illuminated along the ethnohistorical model that is based on Don Guillermo Galvey’s account of the trail he used going to Benguet, which originated from the settlement of Pugo. In fact, the settlement would qualify as a bulk breaking village where bulked products could be stacked together then transferred into shallow draft canoes or bamboo rafts. Such riverine vessels would be able to navigate downriver to the Aringay settlement which is the terminus of this river system. Here we see that economies of scale would have been the organizing principle rather than simply cost minimization effort. The Balatok–Aringay trail may have initially followed an earlier cost-minimized model, which was able to accommodate single file, pedestrian travel, but which later on transformed into a combination of pedestrian transport and raft travel possibly to accommodate higher bulk exchanges (Canilao 2017a).

Here, we examine larger landscapes along which the gold extraction, production, and trade occurs, using GIS spatial analysis, to provide insights into how the upland producers engage in decision-making concerning the placement of extraction and production sites for gold and the villages that support gold extraction. Problem-solving involves: (1) siting extraction and production areas for moving quickly along efficient routes to the coast for exchange, (2) determining when to make a coastal journey (i.e., through having strategic ‘siting’ points), and (3) maximizing time economy in getting to coastal ports for maximum strategic advantage in competition with other producers (i.e., developing efficient and safe routes). While historical, ethnographic, and archeological evidence are crucial in understanding how these various factors associated with competing successfully in gold production might have been weighed by these gold-mining communities, GIS is the critical modeling tool that makes it possible to poinpoint landscape factors that may have been considered in the sophisticated calculus of these long-term tribal gold producers. 6.1 Tonglo Being the first network to be the target of the five series case studies in this book, the GIS analysis aided by satellite remote sensing and ethnohistorical research of Tonglo has provided a visualization of the spatial geography and temporal changes within the early Luzon gold trade that took place in this network. A triangulation of these data shows that Ibaloi miners in Tonglo village occupied strategically situated locales to spot arriving merchant vessels at Aringay and were able to undertake pedestrian transport downhill in half a day to participate in the ensuing tabu-tabuan/tiangge or evanescent market that springs to life at the shore near the vessel. The Mirandaola account of gold in Aringay that was quoted above is probably gold bulked in Tonglo and sourced from Acupan. Social network connections were also established as intermarriages between Tonglo and the coastal and mine settlements were correlated with historic evidence. The historical account also describes the subsistence miner as only taking what they needed from the mines in order to gain access to lowland necessities through barter. From the perspective of the Spanish colonial government this was interpreted as a sign of sloth. On the other hand, the historic and GIS analyses allows a more probable scenario where Ibaloi miners are viewed as free agents at liberty to decide how much gold they wanted to bring down to Aringay. That is, the Ibaloi miners had significant agency in these gold transactions (Canilao 2017a).

6.2 Gasweling The first round of visibility analysis (Figs. 6.1-6.3, Digital Appendix) shows that Gasweling is strategically located to have a vantage view of all three coastal settlements of Calincamasan, Baratao, and Bauang. Quite similar to the Tonglo situation, the Gasweling settlement is suitable for viewing ship arrivals (Figs. 6.4, 6.5), which then allows the Igorot miners to estimate hiking schedules to be able to participate in the evanescent markets (Canilao 2017c). Least cost path predictive modeling (Figs. 6.6-6.12, Digital Appendix) has also delineated trail models that connected the mines with the Gasweling bulking station before the leg en route to the coastal settlements. The trails that pass through the Kapangan flats were overlain on the digital elevation model of the area (Fig. 6.13, Digital Appendix). The least cost path model shows a constriction zone adjacent to the Bauang/Naguilian River, on the road to Calincamasan and Baratao (Fig. 6.14, Digital Appendix). Closer examination of the predictive models superimposed on enhanced WV2 imagery show that this area of the trail is adjacent to present-day trails. The same efficient route from A to B would have been used in the past as in the present (Figs. 6.15-6.17, Digital Appendix) (Canilao 2017c). A closer analysis of enhanced imagery of Baratao reveals two areas that are possible ijangs: a culturally modified

The GIS predictive model (least cost path route) from Balatok mines to Tonglo bulking village traces the Ibaloi 93

Remote Sensing the Margins of the Gold Trade

Figure 6.4. Bauang as seen from the Banangan ridge in Sablan on a clear day (Photo by author 2017) (Canilao 2017c).

index (band 2) – WV2 soils index (band 3) proved to be useful in delineating the trails and the plot clearings (Fig. 6.20) (see Canilao 2017c).

rock formation that possibly functioned as sighting point for approaching ships on the coast and an elevated place for firing lantakas or native cannons on possible hostile ships. The rock described in the Juan de Salcedo account may either be the Poro Point, which guards the southern flank of the San Fernando cove, or the flat elevated rock formation (hill) at Barangay Urbiztondo north of San Fernando with parts extending into Dalumpinas Oeste and Dalumpinas Este barangays (Fig. 6.18). Imagery of the Urbiztondo rock hill indeed reveals a network of trails connecting old stilt house clearings/plots (Fig. 6.19). This may be the ‘orderly’ village on top of a rock that the Juan de Salcedo account mentions. In fact this is the prime candidate when compared to other rock hill formation in the area – between Bauang and San Juan La Union. There is also a platform north northwest in the rock that is ideally situated for lantaka emplacement (artillery, traditional cannon). It should be stated that although Juan de Salcedo was the first conquistador to undertake a conquest in Luzon, his soldiers were surprised when they were met with suppressive lantaka fire (canon fire) from the side of the natives (Scott 1982) alongside rudimentary missile weapons (Keeley 1996) like spears and arrows. For this analysis the WV2 Soils and Constructions band combination (enhancement) as well as the data-fused combination WV2 built-up index (band 1) –WV2 soils

Analysis of the imagery of the Mount Toyongan ridge also shows the area called Galan in the historical literature, which Capitan Alonzo Quirante was able to assay in 1624. This area is within the northern face of the Tabio formation and a closer look in the area shows evidence of gold placer and lode mining slides similar to old slides in the Lepanto network. The WV2 Soils and Constructions and WV2 Mining Operations band combinations (enhancements) were particularly useful in this analysis (Figs. 6.21, 6.22) (Canilao 2017c). 6.3 Lepanto In terms of the Lepanto network, the suitability analysis that combined results from visibility analysis, river and ridgeline/drainage basin, and Thiessen midline buffered zoning shows suitable or even ideal locations to situate the bulking stations for optimal communication interaction with the coastal settlements (Figs. 6.23-6.25, Digital Appendix). One location that had good visibility of Tagudin was found to exhibit some characteristics conducive to a bulking station settlement location based on the landscape 94

Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing

Figure 6.5. Close-up of Bauang as seen from the Banangan ridge in Sablan on a clear day (Photo by author 2017) (Canilao 2017c).

features. This is the area of Barangay (village) Sinacbat, Bakun, Benguet with mountain terracing and swidden farming evident in its slopes. This site is a prime candidate for ground truthing in a second order analysis. In the case of the Thiessen midline for the Apayao bulking village site, this has a good visibility of Tagudin. In the case of the Danac bulking village, however, the closest elevated area offering visibility of Purao is approximately two kilometers north of the settlement. In determining the most conducive bulking station location, the method applied in the Lepanto network involved factoring in both a social variable and an environmental variable that were within reach. The social variable was through the creation of Thiessen polygons around mining and coastal settlements, whereas the environmental variable delineated naturally occurring ridgelines that separated drainage basins. The first approach, however, disregards the physical features of the area and focuses mainly on the centrality within a social landscape (i.e., within a trade periphery or margin) (Canilao 2017b).

with capacity for transshipment to rafts because it is situated at the banks of the Amburayan River but with no direct visibility of Purao (Canilao 2017b). The Lepanto network analysis featured a very straight­ forward weight overlay GIS technique on three variables including visibility of ships, distance to waterways, and distance to either the Thiessen line or ridgeline. In this particular case, the GIS modeling is just a first step in analysis of possible transportation routes to buyers and additional ethnohistorical accounts pertaining to the bulking villages will still be sought out in the future (Canilao 2017b). 6.4 Angaqui A Viewshed analysis was carried out in Angaqui network and the results show that Minlaoi in Mount Quinali has a very good visibility of Angaqui on the banks of the Grand Abra River (Fig. 6.26, Digital Appendix). This is important because it offers the ability to observe arrivals of rafts and middlepersons to this bulking center, which has become a node in the riverine transport system (Fig. 6.27). Indeed remnants of a watch tower were found in the area during the second season of the Ilocos. This is as if the evanescent market has moved from coast upriver to Angaqui. It should be noted that Angaqui is also a waypoint of Lepanto later

The analysis show that the Apayao and Danac sites conform to a social (Thiessen model) rather than an environmental (ridgeline) model. Apayao appears to be a gold bulking village situated to have a coastal surveilling of Tagudin, while Danac may have been a bulking village 95

Remote Sensing the Margins of the Gold Trade

Figure 6.18. Two candidate ijangs north and south of San Fernando cove. WV2 soils and constructions band combination (Courtesy of the DigitalGlobe Foundation) (Canilao 2017c).

as bulked goods were being moved by raft. Interestingly, there is a turn of the twentieth century example of a watch tower documented in the Angaqui network by D.C. Worcester (again see, Fig. 3.26). This watch tower would have been very useful in surveilling movements within the Angaqui settlement and environs (Canilao 2017d).

distributed in the Angaqui network in tandem with a network of guard posts along trails; these structures would have formed a robust defense network. Again, the accounts of chronic warfare were noted in the settlement history of Patiacan, sometimes spelled Patiacung (Canilao 2017d).

It should be noted that the Angaqui settlement, given its relatively wide and deep segment of the Abra River, would have been accessible to longer Abra rafts; such vessels can be spotted from the watch tower in Minlaoi. Using Google Earth TM imagery (2016) ruler function it can be noted that the lower and middle Abra River whose total length for the main branch is approximately 88 kilometers has an average bank-to-bank measurement averaging about 70 to 80 meters across with one of the narrower passages of the middle Abra near Luba measuring 25 meters across. Therefore, these rafts may have been shoved upriver or in some parts pulled by ropes originating from the Vigan area or the Abra bend near Tayum and Bangued until it reaches Angaqui (or even Cervantes on high water), and cargo can then be transshipped using the smaller rafts.

Angaqui was a major bulking station given its strategic location along the grand Abra River where heavier loads could be transshipped by bigger rafts. However, it may have functioned simply as a coastal bulking station with an outpost at Tila peak in earlier periods. In terms of Euclidean distance, it is the closest to the Tila peak trail, which may have served as a satellite site and rest stop, which could have served as a ship-surveilling point for the coastal settlements of Candon and Dumaquaque. Again, this feature was described by Frey Antolin Perez during his pedestrian transport through the Tila trail in the late nineteenth century and Hans Meyer noted it in his travels as well. A viewshed analysis (Fig. 6.28, Digital Appendix) was implemented to assess the ship-surveilling utility of Tila and indeed both the Candon and Dumaquaque settlements and its adjacent coast are visible (Canilao 2017d).

In addition to the raft-surveilling utility of the watch tower it also served as a defensive outpost and several are strategically

Aside from the visibility-based analyses, least cost path GIS predictive modeling of trails (Figs. 6.29-6.30, Digital 96

Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing

Figure 6.20. Data-fused and image-enhanced indices showing the possible lantaka emplacement (WV2 built-up index–WV2 soils index–WV2 soils index) (Courtesy of the DigitalGlobe Foundation) (Canilao 2017c).

97

Remote Sensing the Margins of the Gold Trade

Figure 6.22. Area around Galan mines north of the Tongayan ridge. Note placer mining features. WV2 mining operations band combination (Courtesy of the DigitalGlobe Foundation) (Canilao 2017c).

Appendix) connecting the mining settlements with the bulking stations and eventually the coastal settlements was also carried out. Because of the availability of several archival maps for this area, these computer-generated trails were compared to digitized trails based on historical maps from the eighteenth and nineteenth centuries. There are areas along the predictive model that coincide with the ethnohistorical trails. The best and probably most important example is the segment that goes over the Tila ridge between Angaqui and the coastal settlements. This particular area of interest was examined closer using highresolution WorldView-2 satellite imagery and indeed the area shows several old trail segments (Fig. 6.31, Digital Appendix). As argued earlier in the methods chapter, Ferguson, Berlin, and Kuwanwisiwma note that ancient trails will show discontinuous segments that will ‘remain visible in the present as a form of fragmented preservation’ (2009: 27). Three discontinuous trail segments stand out in the remote sensing analysis using the vegetation combination of WorldView-2 satellite. Similar to the case study in Tonglo, Network 1, grasses and low-lying patches of shrubs have grown over the trail and this particular vegetation has a unique spectral profile that results in highlighting the trail itself (Figs. 6.32, 6.33). This remote sensing result was also seen in the Tonglo network, but

using the vegetation combination (Canilao 2017a: 375) (Canilao 2017d). The trail segments in the Angaqui network also exhibit heavy and long-term trampling and troughing due to pedestrian and possibly horse traffic which formed depressed surfaces (Ferguson, Berlin, and Kuwanwisiwma 2009: 28, also see Sheets and Sever 1991: 69). This is more obviously seen in 51Q 248500mE 1897000mN, which also exhibits ‘an abrupt dog-leg turn rather than a broad sweeping turn’ (Ferguson, Berlin, and Kuwanwisiwma 2009: 30-31) (Fig. 6.34), another diagnostic characteristic of ancient pedestrian trampling. For comparison we can look at an archival photo that was taken at the turn of the twentieth century by D. C. Worcester, somewhere within this network (Fig. 6.35). A segment of the historic Angaqui trails articulates Patiacan-Minlaoi in Mount Quinali based on a 1794 map. Analysis of the satellite imagery overlain with the ethnohistorical trail indeed reveals a trail snaking back and forth as it scales the slopes of Mount Quinali (Fig. 6.36, Digital Appendix). This is the trail used by Richard von Drasche during his trip in the 1870s traveling from Besao to Angaqui (1975: 44). Closer remote sensing analysis using 98

Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing

Figure 6.27. Note that Quinali settlement is on the banks of Balasian River (Rio Quinali). Pennarubia, Don Esteban (Commissioned Map). (1868). (España. Ministerio de Defensa Archivo del Museo Naval 78–27. Filipinas. Croquis de la Provincia de Abra. 1868 al 1874) (Canilao 2017d).

products could be moved albeit in lower bulk exchanges. It is important to emphasize once again the amount of gold exacted from this area during the first contact with Spanish conquistadors, which was quite comparable to what was available in Vigan–Caoayan–Bantay (Canilao 2018a).

image-enhanced WorldView-2 imagery of the environs of Minlaoi also reveals an older trail parallel to the newer and wider road used by the gold prospecting company in the area (Fig. 6.37, Digital Appendix). The image enhanced soils and construction combination also show the various archeological features encountered during the Ilocos Sur archeological Project 2012 (Canilao 2017d).

At this juncture we look at ethnohistorical research in terms of the origins of the Tingguian/Itneg of Abra. Keesing argues that a minimal theoretical case or model in Northern Luzon involves an initial migration by a lowland group into an adjacent mountain area (Keesing 1962: 342). Two examples include the Pangasinan–Ibaloi relationship and the Ilocano–Tingguian/Itneg relationship (Keesing 1962: 342). In the case of the former, one of the major impetuses was gold extraction by placer and lode mining, following the Agno River (Canilao 2011). As for the latter case, the multiscalar and multidisciplinary approach to gold production and trade in this book makes it clear that in addition to gold extraction, direct control of the gold trade by the Tingguian/Itneg, whose common ancestors with Ilocanos inhabited the coasts, may have accounted for the major migrations in subsequent periods (Canilao 2015: 10, Keesing 1962: 121, 342, Cole 1945: 149). Attempts to control Abra River waypoints, its gateway locations, its choke points, as well as jumping-off points to the other side of the Cordillera Mountain–Cagayan Valley were a major impetus for interior migration. Later or subsequent

6.5 Abra The most distinct property of the Abra network is that its river system has the capacity to allow higher bulk exchanges. Vigan–Caoayan–Bantay appears to have maintained a higher-order position in the gold trade among its neighboring coastal ‘peer’ settlements to its north like Calanutian, Cabugao, Refaro, and Magsingal and settlements to its south like Dumaquaque, Candon, and Narvacan. Over time, however, these rival peer settlements may have usurped some of the trade by utilizing an intervening network of trails and footpaths that make a beeline through the Abra foothills and onto the coast, bypassing the riverway. These rival trading partners would have intercepted seaborne merchants bound for Vigan to get them to trade with them, instead of Vigan merchants. This allowed them to continuously challenge the trade dominance of Vigan–Caoayan–Bantay by maintaining these alternative routes through which 99

Remote Sensing the Margins of the Gold Trade

Figure 6.32. Note the spectral signature of the vegetation associated with this trail segment (Using WV2 vegetation combination) (Courtesy of the DigitalGlobe Foundation) (Canilao 2017d).

100

Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing

Figure 6.33. Note the spectral signature of the vegetation associated with this trail segment (Using WV2 vegetation combination) (Courtesy of the DigitalGlobe Foundation) (Canilao 2017d).

101

Remote Sensing the Margins of the Gold Trade

Figure 6.34. Note trough and 15 degree dog-leg turn in this trail segment (Using WV2 vegetation combination) (Courtesy of the DigitalGlobe Foundation) (Canilao 2017d).

102

Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing

Figure 6.35. Guard structure along the trail approaching Bontok settlement in the Angaqui network with Bontok warriors, turn of the twentieth century (© The Field Museum, CSA 22186. Photographer: Dean C. Worcester) (Canilao 2017d).

migrations into Abra, on the other hand, may be tied to the earlier emergence of coastal maritime trading centers as evanescent markets transformed into semi-permanent markets, and eventually permanent markets on the coast, notably Vigan–Caoayan–Bantay, Narvacan, Cabugao, and Calanutian (Canilao 2018a).

assimilation, coastal Ilocanos were already hispanicized, with residuals of the indigenous traditions surviving. Anthropologist Fred Eggan, for instance, notes the striking similarity between Tingguian/Itneg and Ilocano villages which, when compared to those in the interiors, are ‘much larger, with clusters of barrios round a nucleus from where the lakay rules’ (Eggan 1941: 13-16). He notes that ‘wealth per capita also increases sharply as one goes towards the coast, with many rich men among the Tingguian/Itneg and Ilocanos’ (ibid)[emphasis mine] He also notes that both groups believe in the god named Kabunian. For the Tingguian/Itneg, Kadaklan is the name of the supreme deity while for the Ilocanos this supernatural deity is named Apo-Dios. All of these appear to have syncretized into the Christian deity during the Spanish period (Eggan 1941: 13-16)( Canilao 2018a).

This is not to discount, however, that movement may have also fluctuated back and forth. There is one particular historical account for the town of Patoc (currently Pennarubia) in Abra where there are claims that their founders were migrants from Tamag hills next to Vigan on the coast. According to Grace Mateo, these are Tingguian/ Itnegs who migrated to Pennarubia Abra to serve as trade intermediaries between their Cordillera kin and the Ilocano, as well as foreign traders (2004: 194) [emphasis mine]. These movements may be reflective of an agentive drive on the part of the coastal peoples to move to the interior in order to fully control strategic ‘choke’ points or bulking stations of the gold trade (Fig. 6.38). This is opposite to what was observed in earlier networks where the miners themselves wanted to have a freehand in the transactions (Canilao 2018a). Spanish colonization may have hastened the inward migration into Abra as coastal districts logged increases in the incidents of remontados or mountain escapees to the mountains. It is probably by the late nineteenth century, and by virtue of reduccion or

In any transaction, middlepersons are bound to have economic advantage by possessing capital they are ready to lend out. It is argued that this is particularly a skillset developed by the Abra Tingguian/Itnegs. Sande identifies usury as the source of wealth: They have quantities of honey and wax, and trade these commodities with the lowlanders… They are all usurers, lending money for interest and go even 103

Remote Sensing the Margins of the Gold Trade

Figure 6.38. Arrows show the expansion upriver. This expansion resulted to Angaqui becoming fully articulated as a node in the riverine transport of bulked goods. (ASTER GDEM is a product of METI and NASA)

to the point if making slaves of their debtors, which is the usual method of obtaining slaves. Another way is through their wars, whether just or unjust, Those who are driven on their coast by storms are made slaves by the inhabitants of the land. They are so mercenary that they even make slaves of their own brothers, through usury. (Sande 1576: 69) (author’s emphasis)

The overarching government system; the ecclesiastical imprint not only in spiritual matters but also technological improvement, literacy, and other secular fields; the making over of indigenous leadership into aristocracy of minor officialdom and wealth, money and commerce, the rise of urban centers, and, by no means least, the making of common cause in antiSpanish movements. (1962: 326)

Based on the above quote, middleperson skills in the gold exchange has many offshoots, including adeptness to full time gold lending (Canilao 2018a).

Fay-Cooper Cole argues that based on the ceramic trade ware data of Cole provenienced by Berthold Laufer (Cole 1912) huge quantities of gold poured through the area. Cole especially mentions a post Spanish contact decline in gold culture in Abra, unlike the volume of trade logged in the centuries before (Cole 1915). Good catalogues of heirloom trade ceramics are available in the area as well as those in burial contexts, mostly a product of Fay-CooperCole’s research (1912). Indeed, the amount of gold used in the barter exchange to acquire these status symbols would indicate how much access to gold the Tingguian/Itnegs had. Once again, we compare the wealth and prosperity of the coastal people as well as the neighboring Abra Tingguian/Itnegs at Spanish contact and post Spanish contact (see Newson 2009, Eggan 1941). Several Cole and Worcester archival photos in the Field Museum also show aspects of gold working as well as the centrality of ceramics/jars in Tingguian/Itneg rituals even then during the turn of the twentieth century (for an example see CSA 28754 Cole archival photos in the Field Museum) (Canilao 2018a).

Later migrations by coastal peoples to interiors allowed them to reunite with relatives who escaped to the Abra interiors at an earlier date. This movement is not unidirectional, however, and it should be noted that return trips to the coast are also plausible according to Newson (2009) and Keesing (1962), as these populations engage in trade practices in the coast. Newson, for instance, argues that: As contacts with the interior increased, others may have drifted down to the coast voluntarily, perhaps encouraged by kinship ties and trading contacts that predated Spanish arrival (emphasis mine) (Newson 2009: 195) But the main question remains – why the difference between the Ilocanos and Tingguian/Itnegs? According to Keesing, the major consolidating force in differentiating between the Ilocanos and Tingguian/Itneg is post-Spanish contact:

104

Discussing the Results of Ethnohistorical Archeology and How it Relates to the GIS Analysis and Remote Sensing 6.6 Abra Northeast Branch SNA

the river system which eventually subsumed the Angaqui and Lepanto gold trade networks by the sixteenth century. Subject to the availability of high-resolution satellite imagery as well as archival map data, the analyses can also be conducted in other known gold trade systems in the Philippine archipelago.

6.6.1 Reduccion and inter-settlement trade – the Northeastern Abra River Branch Finally, SNA of the northeast branch shows that the environmental setting and the geopolitics of access allowed these villages to become intertwined in the riverine trade system during the early historical period while at the same time facilitating some movements into some economic niches or havens of geography (Smith 2014) from the ongoing colonization project that was spreading upstream of the Abra River during the historical period. A putative hierarchy of settlements developed in this branch and became a function of reduccion. Based on the 1885 map this can be ascertained to have been based on tributes per village.

6.7 Chapter Summary This chapter has demonstrated how a permanent market encroaches on erstwhile separate or dispersed evanescent markets. This is encapsulated with the rise of Abra network as it moved further upriver. While the Lepanto and Angaqui networks initially featured bulking villages with gold miner diasporas, coastal visibility, and footpaths and trails as nodes that allowed the participation in coastal evanescent encounters in Purao, Tagudin, Dumaquaque, Candon, and Narvacan, the encroachment of the Abra resulted to the reconfiguration of the nodes towards accommodating bulked goods movement via a riverine transport system. Instead of a strictly east to west nodal articulation, the Abra permanent market after successfully interfacing with the Angaqui and Lepanto gold mines followed the question mark shape of the Abra River, which moves south to north then east to west. This market also resulted in newer bulking station nodes straddling the riverbanks and featuring trade diasporas of Tingguians and Ilocanos from the coast.

The similarities model seems to show intra-village connections and relationship that come into play to facilitate trade and exchange. The model shows Padangitan and Lapaz as conduits in the network; in fact, they are the bulking villages located at the intersection of the main Abra River system and its smaller northeastern branch. The model also shows two subnetworks of the northeastern branch with the first inclusive of Lacub, Bumalicong, Paganao, and Pacqued, while the second group includes Caganayan, Anayan, Agsimao, Tineg, and Callago. Again, the model plays out well in terms of taking into account the flow accumulation model generated for the Abra basin. This Abra case would be a good case wherein an environmental variable relates to a social variable namely trade and exchange. Thus, the river transportation system via rafts and its bulking stations would be the organizing principle in terms of trade with a movement upriver showing transition from trade hubs to extraction areas.

The next chapter will present the main, closing arguments of the book and will also point out the contribution of this work in terms of anthropological knowledge. The next chapter will also point at the limitations of the book and plans to address these in what is envisioned to be continuing longitudinal research.

As for the dissimilarities model, this seems to jive with the march of reduccion in the area. And this is supported by the 1885 historical map. It can be seen in the map that the satellite settlements Atip, Callago, Agsimao, Tineg, Anayan, Lacub, Bumalicong Caganayan, Paganao, and Padangitan are classified as villages with the least number of tributes, which is eight, whereas Patqued, Lapaz, and Lagayan have at least forty tributes. The graphical results of the case study seem to model the relative autonomy of interior Philippine settlements vis-a-vis the Spanish colonial government. The spatial relationships may reveal how indigenous Philippine communities contested and limited the nature of colonization in the archipelago (Mawson 2019). In fact, this niche of the Grand Abra River system may qualify as what Smith has called a haven of geography (Smith 2014), which are packets of resistance that contest Spanish domination in the archipelago. While the current study was only focused on the northeastern branch, subsequent studies will also be undertaken to pursue analysis of the southern branches of 105

7 Conclusion: From Agency to Marginality in the Dynamic Gold Trading of Northwestern Luzon was extended further upriver to the village of Cervantes, which also became an ethnic melting pot like Angaqui.

The book has unraveled some important aspects of the Northwestern Luzon gold trade, paying particular attention to the small-scale mining and panning Igorot communities and how they established trade connections and networks with coastal settlements that served as jumping-off points to the overseas trade in the Indian Ocean and South China Sea. What is increasingly apparent is the fluidity in the level of agency enjoyed by these interior producers in planning strategies of extracting and transporting gold to coastal markets which connected larger networks of consumers in the IO–WPS/SCS trading system. However, a decreasing trend of local control of gold extraction and transport seems to be connected to how they navigated the multifaceted interfaces from evanescent to more permanent market establishments on the coast over time. In later centuries, there appears to be less agency in negotiating positions in the expanding coastal permanent markets, especially because of the introduction of debt finance as a basis for transactions rather than straightforward barter exchange. Debt from inside or utang na loob is moral and social in character that gold miners are forced into slavery due to nonpayment. The Alozon Quirante account shows that several of the upland gold miners were already in debt to the lowland populations as of the seventeenth century. Oral tradition accounts of Benguet are also rich with entries about the increase in slavery in the highlands linked to the encroaching gold trade.

Several transshipment configurations of the gold trade were also recorded in the five networks and appear to be directly related to the volume of bulked goods flowing in and out. In the case of the Tonglo network, the Igorot had an option of hiking down to either Agoo or Aringay. The historical route shows that the route has eventually articulated the riverine village of Pugo in more recent times. It is argued that this was because transshipment was accomplished in this settlement. In terms of lowland goods moving to the interior, this constituted a bulk breaking point. In the case of the Gasweling network, the gold miners had the agency to decide whether to undertake a pedestrian transport down to Bauang, Baratao, or Calincamasan based on their surveillance of these coastal ports for potential merchant vessels and their ability to quickly launch pedestrian transport of gold products to these coastal settlements. In the case of the Lepanto network they had the option of taking routes to either Tagudin or Purao. The bulking village of Danac, however, seems to suggest a scenario where the network connection was semi-permanent, whereas the bulking village of Apayao supported a scenario involving a continuing evanescent market connection. One important locational principle when determining whether an evanescent, semi-permanent, or permanent network connection existed is the presence of coastal surveillance or visibility as modeled through GIS. All evanescent markets tended to have good uninterrupted surveillance or visibility of the coast. Danac in the Lepanto network had no surveillance or visibility of the coast but was accessible by virtue of the Amburayan River, thus allowing bigger bulk product traffic. Apayao, on the other hand, had good coastal surveillance or visibility but could only accommodate smaller amounts of bulk exchanges.

Nevertheless, many Igorot gold miners and panners in the interiors maintained access to not just one but several coastal markets for gold well into the twentieth century. As to the decision as to where to proceed with the gold, a crucial factor appears to have been the development of midway bulking stations, and economic benefit depended on who controlled these intermediary stations as jumping-off points to the coast. Bulking villages tended to be controlled by coastal diasporas if they were interfaced to a permanent market while upland diasporas tended to control bulking stations that fed into evanescent markets.

The reconstructed Angaqui network illustrates several aspects concerning how the location of bulking villages and articulation to markets change as a result of increased demand for gold. First, especially at Spanish contact in the sixteenth century, the scale of transport of bulked goods that passed through Angaqui became significantly greater due to the merger with the Abra Riverine transport system. This merger was intended to precipitate an eventual merger with the Lepanto network further south. The merger with Angaqui ensured that gold from Quinali would end up in the coastal settlements of the Abra network, whereas the merger with Lepanto ensured that Mankayan gold would also end up in the Abra network coastal settlements, rather than other more adjacent lowland ports. These bulked gold shipments would have traveled in high volumes using the

By the sixteenth century at Spanish contact, Tonglo, Gasweling, and the Apayao conduit (Lepanto) appear to have been articulated to a more evanescent type of market. The Danac conduit (Lepanto) and Angaqui, on the other hand, appear to have been articulated to a more semipermanent market. The sixteenth century Abra network, in contrast to the two groups above, is the epitome of a more permanent market that spread far and wide in this major river valley of Luzon to the point of reconfiguring the network connections of its neighboring networks to the south, Angaqui and Lepanto. During the Spanish Historical Period the reach of the Abra riverine network 107

Remote Sensing the Margins of the Gold Trade points along the Abra River. Such development of these bulking centers would have usurped some of the gold trade that used to flow towards other peer coastal settlements like Purao, Tagudin, Dumaquaque, and Candon. The density of Vigan settlements were recorded in the sixteenth century (see Newson 2009), indicating that once the permanent market was established in Vigan it was no longer important to have surveilling visibility of the coast, but instead what was of strategic importance was to establish bulking stations along the river adept in offering services and support like raft repairs and supplies to traders traversing up and down the river (Canilao 2017d).

multiple raft options available along the Abra River. The Abra River itself drains the gold-rich Mankayan area of Lepanto, and place-names along the river allude to gold panning. While the Angaqui settlement may have become a riverine bulking station usurped by the Abra network by the sixteenth century, resulting in more permanent markets taking root in the area, it used to have its very own trade connections with Candon and Dumaquaque by virtue of its Tila sub-settlement which had surveilling capacity for viewing the coast. This earlier articulation was vividly recorded in several early historical accounts. The positioning of the settlements therefore indicates that lower bulk exchanges with Dumaquaque and Candon were possible using the old trails passing through Tila, but larger bulk exchanges with coastal settlements of Vigan– Narvacan–Santa Maria would have been facilitated by the Abra River.

Being a part of an expansive permanent market established on the coast, gold processing in the Abra network appears to be more dispersed throughout settlements in this wide river valley compared to the other networks to its south, notably the Angaqui and Lepanto networks. It is important to emphasize that gold mining and panning was also possible within the Abra riverine valley, being endowed with several gold deposits. In terms of functionality, these reserves were not extracted on a large scale and was only partially extracted during the Spanish period. The Ilocano diaspora who migrated into the valley may have moved in primarily to serve as middlepersons in the gold exchange and to be able to offer credit to the Igorot miners and panners (i.e., Igorots of Quinali and Lepanto). The Abra River valley flats functioned mainly as bulking stations or conduits for the gold trade sourced from these headwater mines. Indeed the prosperity of the valley that was noted by several ethnographers would have been achieved by controlling the flow of the gold trade from Minlaoi and Lepanto. Cole, however, has lamented that the gold culture of the Tingguian may have been at its peak during the early periods (Early Historical Period) but had receded at Spanish contact (Historical Period) (Canilao 2018a).

Indeed, as market interactions became more permanent, rather than evanescent and triggered by ship arrivals, trips by the Igorots became more regular between mines and coasts and the bulking villages in between (Canilao, 2017b: 631–632). Again, historical demographic data tend to support this scenario, given the high populations of Vigan and Narvacan at contact in the sixteenth century based on conquistador Martin de Goiti’s account (Newson 2009: 182). According to Newson, towns in Ilocos numbered between 200 and 400 but Vigan alone had 800 houses with 3000 houses nearby (probably in the satellite settlements of Bantay, Caoayan, and Santa) (2009: 179). Narvacan was more accessible via pedestrian transport and/or horse rides from the Abra River bend in the Tayum–Bangued area. Vigan, on the other hand, was easily accessed through long, wide bamboo rafts through the Banauang Gap, part of the Bulagawan mountain chain. Hans Meyer mentions that he rode a horse on his trip from Narvacan to San Quentin, and Bangued in the 1880s en route to the location of the Guinaang tribe (1975b: 78). Historian Grace Mateo discusses these horse trails in her dissertation (2004) (Canilao 2017d).

Abra settlements like Tayum, Bucao, Bangued, and Lagang-ilang also did not have surveilling that would have allowed visibility of ships on the coast, but were fully integrated into the high bulk exchanges within the Abra River system. Vigan–Caoayan–Bantay was the terminus of this upland–lowland trade system and was the jumpingoff point for foreign gold processing centers. The Abra valley became fully integrated into this more permanent coastal market system because the coastal peoples migrated to control the trade network, with the objective of establishing some form of monopoly on the gold extracted from the headwaters. While some settlements interfaced as waypoints en route to Vigan, it should be noted that some of the settlements on the northwesterly quadrant of the river became smuggling conduits from where direct connections to Vigan’s coastal settlement competitors to the north like Calanutian, Cabugao, and Magsingal were made. This is the possibly the reason we see the fluctuation of stand-by gold in these settlements as seen in tribute records after several expeditions in the sixteenth century. Based on the 1885 map of Abra these settlement conduits include Bandi, Danglas, Dagat, and Labtablang, among others. The settlement of Tineg, however, stands

In terms of bulking villages in the Angaqui and Abra networks the emergence of Cervantes seems to be an example of these types of sites in the outer reaches of the Abra network as alliances and migrations facilitated the expansion in a race to control the gold-rich Lepanto mines. Cervantes together with Angaqui settlements became waypoints in the movement of gold from Lepanto to the coast. Within the Angaqui network, the Dumaquaque/ Candon to Angaqui–Minlaoi/Quinali conduits predate the Vigan–Abra–Angaqui–Cervantes–Lepanto conduits. Within the Lepanto network an earlier Purao/Tagudin– Apayao/Danac–Lepanto conduit may have coincided with the Dumaquaque/Candon to Angaqui–Minlaoi/Quinali route before operations became fully riverine and raft based (Canilao 2017d). Increasingly over time we see the establishment of Vigan as a more permanent market, in tandem with the emergence of bulk breaking points/way 108

Conclusion out among the Abra valley settlements because it was principally a strategic jumping-off point before the long pedestrian transport that cuts through the mossy to deciduous forests of Cordillera Mountains all the way to the Malague settlement in the Cagayan valley (Malague) ( Canilao 2018a).

what is important to point out in the book is the agentive decision-making at many levels and among many interest groups, including adaptations to agentive demands and decisions of others in the network.

At this juncture it is important to note that limits in time and resources has prevented the ground-truthing of some of the footpaths and trails identified in the book. This is being reserved for a second order analysis that will collate all this remote sensed data and will launch an archeological exploration. Accessibility is a major factor to consider in this future endeavor since these remote-sensed areas are only accessible through hikes. In some cases, there may be a need to cut a trail to reach these places. As a contingency measure, a second order analysis will be aided with a quadrocopter drone. This will complement the already available satellite imagery and will provide a more updated view of the terrain. In concluding, it is argued that the contribution of this book to both the archeology and history of Northwestern Luzon’s gold production as part of the IO–WPS/SCS trade is its multidisciplinary approach, its focus on small worlds of producers and consumers and agentive decision-making, and its regional multiscalar approach using GIS and remote sensing. The book was able to tap on multiple sources of data including historical data, ethnographic data, written historical data, oral tradition data, archeological data, GIS data, and high-resolution multispectral satellite data used in remote sensing. These data were then compiled and fed as variables in a spatial form of analysis writ large and writ small. The book studied the complex systems of linkages between producers along a chain which included specialist miners and gold-bulkers, middleperson traders, merchants, and consumers. While several studies on trade has focused on elite in the centers of trade systems (i.e., port settlements) where the raw products are partially processed for export overseas, this study is one of the few that has looked at the activities of the miners and panners of gold who constantly made decisions regarding the various aspects of the trade of gold: when and where to extract gold; when and where to bulk gold; when, where, and how to transport gold; when, where, and with whom to trade this gold, be it to the coastal middlepersons or the foreign merchants themselves. It is also emphasized that in terms of the gold industry activities discussed in this book participants were also playing multiple roles and they negotiated their positions with respect to the foreign merchants and the coastal middleperson. By looking at five networks in Northwestern Luzon, the book argues for a trend for more of the agentive flexibility of the miners and panners to be curtailed as markets move from evanescent to semi-permanent and finally to permanent. Nevertheless, despite temporal and spatial changes in the reach and scale of this market over time 109

Bibliography Apple, Russell. Trails Form Steppingstones to Kerbstones. Honolulu: Bishop Museum Press, 1965.

Acabado, Stephen. “The Archeology of Pericolonialism: Responses of the Unconquered to Spanish Conquest and Colonialism in Ifugao, Philippines.” International Journal of Historical Archeology 21 (2017): 1-26.

Azurin, Arnold Molina. Beddeng: Exploring the IlocanoIgorot Confluence. Manila: Museo ng Kalinangang Pilipino, Sentrong Pangkultura ng Pilipinas, 1991.

______________. “A Bayesian Approach to Dating Agricultural Terraces: A Case from the Philippines.” Antiquity 83 (2009): 801-14.

Bagamaspad, Anavic & Zenaida Hamada-Pawid. A People’s History of Benguet. Baguio City: Baguio Printing and Publishing Corporation, 1985.

______________. “Taro before Rice Terraces: Implications of Radiocarbon Determinations, Ethnohistoric Reconstructions, and Ethnography in Dating the Ifugao Terraces.” Senri Ethnology Studies 78 (2012): 285-305.

Barker, G. “The Niah Caves, the ‘Human Revolution’, and Foraging/Farming Transitions in Island Southeast Asia.” In Rainforest Foraging and Farming in Island Southeast Asia: The Archeology of the Niah Caves, Sarawak, edited by G. Barker, 341-66. Cambridge: McDonald Institute for Archeological Research, 2013.

Alexander, Rani. “The Emerging World System and Colonial Yucatan: The Archeology of Core-Periphery Integration, 1780-1847.” In World Systems Theory in Practice, edited by Nick Kardulias, 103-24. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999.

Barretto-Tesoro, Grace. Identity and Reciprocity in 15th C Philippines, International Series 1813. Oxford: BAR Publishing, 2008.

Algaze, Guillermo. The Uruk World System: The Dynamics of Expansion of Early Mesopotamian Civilization. Chicago and London: University of Chicago Press, 1993.

Barton, R. F. The Halfway Sun: Life among the Headhunters of the Philippines. New York: Brewer and Warren Inc., 1930.

Allegre, Edilberto, ed. Meeting at the Marketplace. Edited by Eusebio Dizon Gabriel Casal, Wilfredo Ronquillo, and Cecilio Salcedo. 7 vols. Vol. 2, Kasaysayan 2: The Story of the Filipino People. New York and Hong Kong: Asia Publishing Company Limited (Reader’s Digest), 1998.

Bearman, Peter. Peter S., James Moody, and Katherine Stovel. “Chains of Affection: The Structure of Adolescent Romantic and Sexual Networks.” American Journal of Sociology 110 (2004): 44-91.

Allen, Mitchell. “Contested Peripheries: Philistia in the Neo-Assyrian World System. Phd Dissertation.” University of California Los Angeles, 1996.

Bell, Tyler, Andrew Wilson, Andrew Wickam. “Tracking the Samnites: Landscape and Communications Routes in the Sangro Valley, Italy.” American Journal of Archeology 106, no. 2 (2002): 69-86.

Andaya, B. W. “Historicising “Modernity” in Southeast Asia.” Journal of the Economic and Social History of the Orient 40, no. 4 (1997): 391-409.

Bellina, B. . “Beads, Social Change and Interaction between India and South-East Asia.” Antiquity 77, no. 296 (2003): 285-97.

___________, and Andaya, L. Y. . A History of Early Modern Southeast Asia, 1400-1830: Cambridge University Press, 2015.

Bellwood, Peter. The First Farmers; the Origins of Agricultural Communities. Australia: ANU, 2005.

Andaya, L. Y., and Andaya, B. W. . “Southeast Asia in the Early Modern Period; Twenty-Five Years On.” Journal of Southeast Asian Studies 26, no. 1 (1995): 92-98.

_____________. and Eusebio Dizon. 4000 Years of Migration and Cultural Exchange (Terra Australis 40). Canberra: Australia National University Press, 2013.

Anderson, D.G. The Savannah River Chiefdoms: Prehistoric Change in the Late Prehistoric Southeast. Tuscaloosa: University of Alabama Press, 1994.

Beyer, H. Otley. “Early History of Philippine Relations with Foreign Countries Especially China.” In Readings in Philippine Prehistory, 112-27. Manila: Filipino Book Guild, 1979.

Antolin O.P., Francisco. “Notices of the Pagan Igorots in 1789 – Part Two (Translated by William H. Scott).” Asian Folklore 30, no. 2 (1971): 27-132. 111

Remote Sensing the Margins of the Gold Trade ______________. “Bulked Gold, Riverine Trade, Hiking Trails, and Worldview2 Satellite Remote Sensing in Northwestern Luzon: The Angaqui Network.” Journal of Archeological Science: Reports 17, no. 2018 (2017d): 689-711.

_____________. “Outline Review of Philippine Archeology by Islands and Provinces.” Philippine Journal of Science 77, no. 3&4 (1949): 205-390. Binford, Lewis. “Willow Smoke and Dogs’ Tails: HunterGatherer Settlement Systems and Archeological Site Formation.” American Antiquity 45 (1980): 4-20.

______________. “Calanutian: Archeology of a PreSpanish Contact Settlement in Ilocos Sur 12th to 16th C (Site Report).” edited by Michael Armand P. Canilao. Quezon City: Archeological Studies Program University of the Philippines Diliman, 2011b.

Bodner, Connie. “On the Evolution of Agriculture in Central Bontoc.” University of Missouri, 1986.

______________. “Determining the Suitability of Gold Bulking Villages Articulated to the Lepanto Gold District of Northwestern Luzon with the Aid of Worldview2 Satellite Imagery.” Journal of Archeological Science: Reports 14, no. 2017 (2017b): 620-37.

Borgatti, S.P., Everett, M.G. and Freeman, L.C. “Ucinet for Windows: Software for Social Network Analysis.” Harvard, Massachusetts: Analytic Technologies, 2002. Braudel, Fernand. The Mediterranean and the Mediterranean World in the Age of Philip Ii Volume 1. Berkeley: University of California Press, 1972.

______________. “Early Historical Gold Trade Networks in Northwestern Luzon, as Reconstructed from Ethnohistorical Accounts, Worldview2 Satellite Remote Sensing and Gis Predictive Modelling: The Gasweling Case.” Journal of Archeological Science: Reports 16c, no. 2017 (2017c): 127-48.

Buzeta, Manuel. Diccionario Geografico-EstadicoHistorico, De Las Islas Filipinas. Vol. 2 volumes. Madrid: Imprenta de D. Jose C. de la Pena, 1850-51. Caballero, Evelyn J. Gold from the Gods: Traditional Small Scale Miners in the Philippines. Quezon City: Giraffe Books, 1996.

______________. “Gold Trade and Evanescent Market Encounters of Northwestern Luzon Island 10th to Early 20th C.” In 8th World Archeology Conference. Doshisha University, Kyoto, Japan, 2016.

Caldwell, Joseph. “Interaction Spheres in Prehistory.” In Hopewellian Studies: Scientific Papers, edited by Joseph Caldwell and Robert Hall. Springfield, Illinois: State of Illinois, 1964.

______________. Mountains and Sea: Coastal, Riverine, and Upland Archeology of Ilocos Sur. Espana: University of Santo Tomas Publishing House, 2015.

Canilao, Michael “Charting the Abra Gold Trade Network of Northwestern Luzon Using Ethnohistorical Archeology and Worldview-3 Satellite Imagery “ Journal of Archeological Science-Reports 19 (2018a): 357-96.

______________. Of Gold, Spanish Conquistadors, and Ibaloi Generational Memory. Baguio City: Cordillera Studies Center University of the Philippines Baguio, 2011c.

______________. “Isles of Garrisons: Remote Sensing Ijangs in Northern Philippines (Sixteenth-Century).” Geojournal – Spatially Integrated Social Sciences and Humanities https:// doi.org/10.1007/s10708-018-9921-0 (2018b).

______________. “The Search for Tonglo: In Pursuit of a Legendary Ibaloi Gold Trading Village.” Philippine Quarterly of Culture and Society 40 (2012): 58-78.

______________. “Lost in the Retelling: Washed-out Balitok (Gold) in Ibaloi Generational Memory.” Hukay international refereed journal of the UP-ASP 15 (2010): 21-29.

______________. “Understanding the Peopling of Benguet through Archeology, Ethnohistory, and Oral Tradition.” University of the Philippines Diliman, 2008.

______________. “Proposal for a Surface Archeology Methodology in Bluff Lands Arcaheology (the Cordillera Region of Luzon).” Hukay international refereed journal of the UP-ASP 15 (2013): 82-89.

______________. “Worldview2 Satellite Imagery in Remote Sensing a Past Gold Trading Trail in Luzon: Testing Ethnohistory-Based and GIS-Based Methods.” Journal of Archeological Science: Reports 12 (2017a): 363-77.

______________. “Social Network Analysis of the Northeast Branch of the Grand Abra River Network Luzon, Philippines, in the Nineteenth Century.” Geojournal – Spatially Integrated Social Sciences and Humanities DOI 10.1007/s10708-019-10028-y (2019).

Carillo, Manuel. “A Brief Account of the Missions of the Four Tribes Called Igorots, Tinguians, Apayaos, and Adanes, Recently Established in the Philippine Islands, in the Mountains of the Provinces of Ilocos and Pangasinan, by the Fathers of Our Patron Saint Augustine, of the Province of the Most Holy Name of Jesus.” In Notices of the Pagan Igorots in the Interior of the Island of Manila, edited by trans. by William Henry Scott. Manila: UST Press, 1756.

______________. “Transformations in Ibaloi Gold Extraction from the Protohistoric Period to the Current Era (14th to 20th C).” Aghamtao 20 (2011a): 20-39. 112

Bibliography Cariño, Jessica. “The Cariños and Baguio-Benguet History.” In Csc Monographs. Baguio City: Cordillera Studies Center, 1985. Chan, Albert. “Chinese-Philippine Relationship in the Late 16th C to 1603.” Philippine Studies 26 (1978): 51-82.

Dizon, Eusebio. “The Role of the Philippines as an Entrepot During the 12th-15th Century’s Chinese and Southeast Asian Trade Network.” In Chinese Export Ceramics and Maritime Trade 12th to 15th c. Chinese Civilisation Centre, City University of Hong Kong, 2004.

Chase-Dunn, Christopher and Peter Grimes. “World Systems Analysis.” Annual Revie of Sociology 21 (1995): 387-417.

_____________. “Underwater and Maritime Archeology in the Philippines.” Philippine Quarterly of Culture and Society 31 (2003): 1-25.

Chaudhuri, K.N. . Trade and Civilization in the Indian Ocean from the Rise of Islam to 1750. London: Cambridge University Press, 1985.

Drasche, Richard von. “The Military Districts of Benguet (1876).” In German Travellers on the Cordillera (18601890), 35-45. Manila: Filipino Book Guild, 1975.

Cheng, Chieh-fu. “The Archeology of Trails in the Batongguan Region of Taiwan: Landscape, People, and Mobility from the Early Seventeenth Century to the Present.” Boston University, 2019.

Dussubieux, Laure and Berenice Bellina. “Glass Ornament Production and Trade Polities in the Upper-Thai Peninsula During the Early Iron Age.” Archeological Research in Asia 13 (2018): 25-36.

Cole, Fay-Cooper. “Chinese Pottery in the Philippines with Postscript by Berthold Laufer.” Field Museum of Natural History (Publication 162) XII, no. 1-278 (1912).

Earle, Timothy. “Chiefdoms in Archeological and Ethnohistorical Perspective.” Annual Review of Anthropology 16 (1987): 279-308.

_______________. “Traditions of the Tinguian: A Study in Philippine Folklore.” Field Museum of Natural History (Publication 180) XIV, no. 1 (1915): 1-202.

____________. How Chiefs Come to Power: The Political Economy in Prehistory. Stanford California: Stanford University Press, 1997.

Cortes, Rosario Mendoza. Pangasinan, 1572-1800. Quezon City: University of the Philippines Press, 1974.

____________. “Paths and Roads in EVolutionary Perspective.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Trumbold, 10-16. Cambridge: Cambridge University Press, 1991.

Dampier, William. “Dampier in the Philippines.” In The Philippine Islands 1493-1898 Vol. 39, edited by Emma Blair & James Robertson, 18-121. Mandaluyong: Cachos Hermanos Inc., 1569.

____________. “Prehistoric Economics and the Archeology of Exchange.” In Contexts for Prehistoric Exchange, edited by Timothy Earle and J.E. Erickson, 1-12. New York: Academic Press, 1982.

Darling, J. Andrew. “O’odham Trails and the Archeology of Space.” In Landscapes of Movement: Trails Paths and Roads in Anthropological Perspective, edited by James Clark Erickson Snead, and J. Andrew Darling, 61-83. Philadelphia: University of Pennsylvania Museum of Anthropology and Archeology, 2009.

____________. “Routes through the Landscape: A Comparative Approach.” In Landscapes of Movement: Trails Paths and Roads in Anthropological Perspective, edited by James Clark Erickson Snead, and J. Andrew Darling, 253-70. Philadelphia: University of Pennsylvania Museum of Anthropology and Archeology, 2009.

Dasmarinas., Gomes Peres. “Account of the Encomiendas of the Philipinas Islands.” In The Philippine Islands 14931898, edited by Emma Blair & James Robertson, 96-141. Mandaluyong: Cachos Hermanos Inc., 1591.

____________ and Jonathan Ericson. “Exchange System in Archeological Perspective.” In Exchange System in Prehistory, edited by Timothy Earle and J.E. Erickson, 1-12. New York: Academic Press, 1977.

Demandt, M. H. . “Early Gold Ornaments of Southeast Asia: Production, Trade, and Consumption.” Asian Perspectives (2015): 305-30.

Erickson, Clark. “Agency, Causeways, Canals, and the Landscapes of Everyday Life in the Bolivian Amazon.” In Landscapes of Movement: Trails Paths and Roads in Anthropological Perspective, edited by James Clark Erickson Snead, and J. Andrew Darling, 1-19. Philadelphia: University of Pennsylvania Museum of Anthropology and Archeology, 2009.

DigitalGlobe. “World View-3 Data Sheet.” edited by DigitalGlobe. Longmont Colorado USA, London UK, Singapore, 2014. __________. “World View 2 Data Sheet.” edited by DigitalGlobe. Longmont Colorado USA, London UK, Singapore, 2013. 113

Remote Sensing the Margins of the Gold Trade Graulau, Jeannette. “Peasant Mining Production as a Development Strategy: The Case of Women in Gold Mining in the Brazilian Amazon.” European Review of Latin American and Caribbean Studies/Revista Europea de Estudios Latinoamericanos y del Caribe 71, no. October (2001): 71-106.

Feinman, Gary. “The Changing Structure of Macroregional Mesoamerica: The Classic-Postclassic Transition in the Valley of Oaxaca.” In World Systems Theory in Practice, edited by Nick Kardulias, 53-62. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999. Ferguson, T.J. G. Lennis Berlin and Leig Kuwanwisisma. “Kukhepya: Searching for Hopi Trails.” In Landscapes of Movement: Trails Paths and Roads in Anthropological Perspective, edited by James Clark Erickson Snead, and J. Andrew Darling, 20-41. Philadelphia: University of Pennsylvania Museum of Anthropology and Arcaheology, 2009.

Green, Debra Kay. “Geo Archeology of Prehispanic Agricultural Landscapes in the Bais-Tanjay Basin Negros Oriental-Dissertation.” University of Illinois at Chicago, 2010. Hall, Thomas. “Incorporation in the World-System: Toward a Critique.” American Sociological Review 51 (1986): 390-402.

Foronda, Marcelino A. and Juan Foronda. Samtoy: Essays on Iloko History and Culture. Manila: United Publishing Corporation, 1972.

___________. Social Change in the Southwest, 13501880. Lawrence, KS: University Press of Kansas, 1989.

Frank, Andre Gunder. “Abuses and Uses of World Systems Theory in Archeology.” In World Systems Theory in Practice, edited by Nick Kardulias, 275-96. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999.

___________. and Christopher Chase-Dunn. “The WorldSystems Perspective and Archeology: Forward into the Past.” Journal of Archeological Research 1, no. 2 (1993): 121-43.

_________________. “The Development of Underdevelopment.” In Imperialism and Underdevelopment, edited by Robert F. Rhodes. New York: Monthly Review Press, 1970.

___________, Nick Kardulias, Christopher Chase-Dunn. “World Systems Analysis and Archeology: Continuing the Dialogue.” Journal of Archeological Research 1, no. 2 (2010): 121-43.

_________________ and Barry Gills, ed. The World System: Five Hundred Years of Five Thousand? London and New York: Routledge, 1993.

Harris, Trevor. “Session 2 Discussion: Moving Gis” Exploring Movement within Prehistoric Cultural Landscapes Using GIS.” In Beyond the Map: Archeology and Spatial Technologies, edited by Gary Lock, 116-23. Amsterdam: IOS Press, 2000.

Galvey, Guillermo. “Account of Spanish Expedition to Benguet in the Year 1829.” Ethnological Survey Publications The Nabaloi Dialect 2, no. 2 and 3 (1829). Gironiere, Paul Frost. Twenty Years in the Philippines. Manila: Filipino Book Guild, 1854.

Hassig, Ross. “Roads, Routes, and Ties That Bind.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Trumbold, 17-27. Cambridge: Cambridge University Press, 1991.

Golitko, Mark and Gary Feinman. “Procurement and Distribution of Pre-Hispanic Mesoamerican Obsidian 900 BC-AD 1520: A Social Network Analysis.” Archeological Method and Theory 22 (2015): 206-47.

___________. Trade, Tribute, and Transportations: The Sixteenth Century Political Economy of the Valley of Mexico. Norman: University of Oklahoma Press, 1985. Hendrickson, Mitch. “New Evidence of Brown Glazed Kilns Along the East Road from Angkor.” Bulletin of the Indo-Pacific Prehistory Association 28 (2008): 52-56.

Golledge, Reginald. “Human Wayfinding and Cognitive Maps.” In Colonization of Unfamiliar Landscapes: The Archeology of Adaptation, edited by Macy and James Steele Rockman, 25-43. London and New York: Routledge, 2003.

________________, Stephanie Leroy, Quan Hua, Phon Kaseka, Voeun Vuthy. “Smelting in the Shadow of the Iron Mountain: Preliminary Field Investigation of the Industrial Landscape around Phnom Dek, Cambodia ( Ninth to Twentieth Centuries A.D.).” Asian Perspectives 56, no. 1 (2017): 55-91.

Gorenflo, Larry and Thomas Bell. “Network Analysis and the Study of Past Regional Organization.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Tromboldt, 80-98. Cambridge: Cambridge University Press, 1991.

________________. “A Transport Geographic Perspective on Travel and Communication in Angkorian Southeast Asia 9th to 15th C.” World Archeology 43, no. 3 (2011): 444-57.

Gratz, Tito. “Gold Trading Networks and the Creation of Trust: A Case Study from Northern Benin.” Africa: Journal of the International African Institute 74, no. 2 (2004): 146-72. 114

Bibliography ________________, Quan Hua, Thomas Pryce “Using in-Slag Charcoal as Indicator of Terminal Iron Production within the Angkorain Period (10th to 13th C Ce) Centre of Preah Khan of Kompong Svay, Cambodia.” Radiocarbon 55, no. 1 (2013): 1-17.

___________. “Hunter-Gatherer Landscapes and Lowland Trade in the Prehispanic Philippines.” World Archeology 27, no. 3 (1996): 389-410. ___________. “Integrating History and Archeology in the Study of Contact Period Philippine Chiefdoms.” International Journal of Historical Archeology 2, no. 4 (1998): 291-320.

Higham, Charles. “Hunter-Gatherers in Southeast Asia: From Prehistory to the Present.” Human Biology 85, no. 1-3 (2013): 21-44.

___________. “Long Term Change and Short Term Shifting in the Economy of Philippine Forager-Traders.” In Beyond Foraging and Collecting Evolutionary Change in Hunter Gatherer Settlement Systems, edited by Ben Fitzhugh and Junko Habu, 339-86. Cambridge: Cambridge University Press, 2002b.

_____________, X. Guangmao, L. Qiang. “The Prehistory of a Friction Zone: First Farmers and Hunter-Gatherers in Southeast Asia.” Antiquity 85 (2011): 529-43. Himes, Ronald S. “The Southern Cordillera Group of Philippine Languages.” Oceanic Linguistics 37, no. 1 (1998): 120-77.

___________and Larissa Smith. “Farmer and Forager Interactions in Southeast Asia.” In Handbook for East and Southeast Asian Archeology, edited by P. Lape J. Habu, J. Olsen, J. Zhichun: Springer DOI: 10.1007/978-1-49396521-2_36, 2017.

Hirth, Friedrich & W.W. Rockhill (trans.). Chan Ju KuaHis Work on the Chinese and Arab Trade in the Twelfth and Thirteenth Centuries Entitled Chu Fan Chi. St. Petersburg1911.

___________. Raiding, Trading, and Feasting: The Political Economy of Philippine Chiefdoms. Honolulu: University of Hawaii Press, 1999.

Hofman, C., Mol, A., Hoogland, M., & Valcárcel Rojas, R. “Stage of Encounters: Migration, Mobility and Interaction in the Pre-Colonial and Early Colonial Caribbean.” World Archeology 46, no. 4 (2014): 590-609.

Jyotishi Amelendu, Kuntala Lahiri-Dutt, and and Sashi Sivramkrishna. “Historical Trajectory of Goldmining in the Nilgiri–Wayanad Region of India.” In Between the Plough and the Pick: Informal, Artisanal and Small-Scale Mining in the Contemporary World, edited by Kuntala Lahiri-Dutt: ANU Press, 2018.

Hornedo, Florentino. The Glitter of Gold in Batanes Cultural Heritage. Espana: University of Santo Tomas Publishing House, 2000. Hyslop, John. The Inka Road System. New York: Academic, 1984.

Kardulias, Nick (Ed.). . World System Theory in Practice: Leadership, Production, and Exchange. New York and Oxford: Rowland and Littlefield Publishers Inc., 1999.

__________. “Observations About Research on Prehistoric Roads in South America.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Trumbold, 28-33. Cambridge: Cambridge University Press, 1991.

____________, and Thomas D. Hall. “ Archeology and World-Systems Analysis.” World Archeology 40, no. 4 (2008): 572-83.

Jeske, Robert. “World Systems Theory, Core-Periphery Interactions and Elite Economic Exchange in Mississippian Society.” In World Systems Theory in Practice, edited by Nick Kardulias, 203-22. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999.

____________. “Negotiation and Incorporation on the Margins of World-Systems: Examples from Cyprus and North America.” Journal of World-Systems Research 13, no. 1 (2007): 55-82. ____________. “Negotiation in a Contested Periphery: Indians in the Fur Trade.” In Meeting on the Political Economy of World Systems (May 4). Riverside California, 2002.

Jesus, Ed de. The Tobacco Monopoly in the Philippines: Bureaucratic Enterprise and Social Change 1766-1880. Quezon City: Ateneo de Manila University Press, 1980.

Keeley, Lawrence. War before Civilization. New York and Oxford: Oxford University Press, 1996.

Junker, Laura. “Conflictive Trade Along the Margins: The Archeology of a Precolonial Philippine Chiefdom (Orig. German).” Antike Welte 5, no. 13 (2013): 1-11.

Keesing, Felix M. The Ethnohistory of Northern Luzon. Stanford: Stanford University Press, 1962.

___________. “Economic Specialization and Inter-Ethnic Trade between Foragers and Farmers in the Prehispanic Philippines.” In Forager-Traders in South and Southeast Asia, edited by Kathleen Morrison and Laura Junker, 20341. Cambridge: Cambridge University Press, 2002a.

Kersten, Carool. Strange Events in the Kingdoms of Cambodia and Laos, 1635-1644. Bangkok: White Lotus Press, 2003. 115

Remote Sensing the Margins of the Gold Trade Lopez, Salvador P. Isles of Gold: A History of Mining in the Philippines. Singapore: Oxford University Press, 1992.

Kerth, Gerald, Nichola Perony, and Frank Schweitzer. “Bats Are Able to Maintain Long-Term Social Relationships Despite the High Fission-Fusion Dynamics of Their Groups.” Proceedings of the Royal Society of London B: Biological Sciences 278 (2011): 2761-67.

Lusseau, David. “The Emergent Properties of a Dolphin Social Network.” Proceedings of the Royal Society of London Biological Sciences 270 (2003): S186-S88.

Knappett, Carl. “Introduction: Why Networks? .” In Network Analysis in archeology, edited by C. Knappett. Oxford: Oxford University Press, 2013.

Madry, Scott and Carole Crumley. “An Application of Remote Sensing and Gis in a Regional Archeological Survey.” In Interpreting Space: Gis and Archeology, edited by S. Green and E. Zubrow K. Allen, 364-81. London: Taylor and Francis, 1990.

Kohl, Philip, ed. The Use and Abuse of World Systems Theory: The Case of the Pristine West Asian State. Edited by Michael Schiffer. Vol. 11, Advance in Archeological Method and Theory. Tucson, Arizona: Academic Press, Inc., 1987. Kouchoukos, N. “Satellite Images and the Representation of near Eastern Landscapes.” Near Eastern Archeology 64, no. 1/2 (2001): 80-91.

___________ and Lynn Rakos. “Line-of-Sight and Cost Surface Analysis for Regional Research in the Arroux River Valley.” In New Methods, Old Problems: Geographic Information Systems in Modern Archeological Research, edited by H.D.G. Maschner, 104-26. Carbondale IL: Center for Archeological Investigations, 1996.

Kvamme, Kenneth L. “Integrating Multidimensional Geophysical Data.” Archeological Prospection 13 (2006): 57-72.

Maher, Robert. “ Archeological Investigations at Burnay District of Southeastern Ifugao, Philippines.” Asia Perspectives 24, no. 2 (1980).

Lacsina, Ligaya. “The Ijang Site of Batanes Northern Philippines: An Archeological Study of the Fortified Hilltop Settlements.” University of the Philippines, 2009.

____________ “ Archeological Investigations in Central Ifugao.” Asian Perspectives XVI (1974). Mateo, Grace Estela. “A History of Ilocos: A Story of the Regionalization of Spanish Colonialism.” Dissertation, University of Hawaii, 2004.

LaMotta, Vincent. “Behavioral Archeology.” In Archeological Theory Today, edited by Ian Hodder, 62-92. Cambridge: Polity Press, 2012.

Medina, Juan de. “History of the Augustinian Order in the Filipinas Islands.” In The Philippine Islands 14931898, edited by Emma Blair & James Robertson, 119-297. Mandaluyong: Cachos Hermanos Inc, 1630.

Laufer, Berthold. “The Relations of the Chinese to the Philippines.” Smithsonian Miscellaneous Collections L, no. IV (1908): 248-83.

Mendizabal, Tomas and Dimitrios Theodossopoulos. “The Embrera, Tourism and Indigenous Archeology: “Rediscovering” the Past in Eastern Panama.” Memorias: Revisita digital de historia y arquelogia desde el caribe colombiano 9, no. 18 (2012): 88-114.

Legaspi, Avelino. Bolinao: 14th to 15th Century Burial Site. Manila: National Museum, 1974. Legaspi, Migue de. “Relation of the Filipinas Islands.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 54-61. Mandaluyong: Cachos Hermanos Inc., 1569.

Meyer, Hans. “The Igorots (1890).” In German Travellers on the Cordillera (1860-1890), 104-28. Manila: Filipino Book Guild, 1975.

Llobera, Marcos. “Understanding Movement: A Pilot Model Towards the Sociology of Movement.” In Beyond the Map: Archeology and Spatial Technologies, edited by Gary Lock, 65-84. Amsterdam: IOS Press, 2000.

__________. “A Trip to the Igorots (1890).” In German Travellers on the Cordillera (1860-1890), 46-103. Manila: Filipino Book Guild, 1975.

Loarca, Miguel de. “Relacion De Las Yslas Filipino.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 34-187. Mandaluyong: Cachos Hermanos Inc, 1582.

Miller Duncan, Nirdev Desai, Julia Lee-Thorp “Indigenous Gold Mining in Southern Africa: A Review. Goodwin Series, Vol. 8, African Naissance: The Limpopo Valley 1000 Years Ago.” South African Archeological Society Goodwin Series 8, no. Dec (2000): 91-99.

Lock, Gary and Trevor Harris. “Enhancing Predictive Archeological Modelling: Integrating Location, Landscape, and Culture.” In Gis and Archeological Site and Location Modelling, edited by M. Mehrer. London and New York: Taylor and Francis, 2001.

Mirandaola, Andres de. “Letter to Felipe Ii.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 33-43. Mandaluyong: Cachos Hermanos Inc., 1574. 116

Bibliography Cuta, Barangay Anilao, Bongabong, Oriental Mindoro). Quezon City: Archeological Studies Program-University of the Philippines Diliman, 2005.

Morga, Antonio de. “Sucesos De Las Islas Filipinas.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 27-209. Mandaluyong: Cachos Hermanos INc, 1609.

Peregrine, Peter. “Legitimation Crises on Prehistoric Worlds.” In World Systems Theory in Practice, edited by Nick Kardulias, 37-52. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999.

Morris, Ian. “Negotiated Peripherality in Iron Age Greece: Accepting and Resisting the East.” In World Systems Theory in Practice, edited by Nick Kardulias, 63-84. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999.

Perez, Fray Angel. Igorots: Geographic and Ethnographic Study of Some Districts of Northern Luzon 1902 (Translated by Enriquetta Fox, Birgit Keith, Anthony Lauria Jr. And William Henry Scott). Baguio City: Cordillera Studies Center, 1988.

Newson, Linda A. Conquest and Pestilence in the Early Spanish Philippines. Honolulu: UNiversity of Hawaii Press, 2009.

Pires, Tome. Travel Accounts of the Islands 1513-1787. Manila: Filipino Book Guild, 1971.

Obenauf, Margaret S. “Photo-interpretations of Chacoan Roads.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Trumbold, 28-33. Cambridge: Cambridge University Press, 1991.

Polyani, K. The Great Transformation. Boston: Beacon, 1957.

Oka, Rahul and Chapurukha Kusimba. “ Archeology of Trading Systems Part 1: Towards a New Trade Synthesis.” Journal of Archeological Research 16, no. 4 (2008): 339-95.

Prill-Brett, June. An Ethno-Archeological Report on Adult Jar Burial Practice in the Central Cordillera Northern Philippines. Vol. 30, Csc Working Papers. Baguio City: Cordillera Studies Center, 2000.

Ortega O.S.A., Francisco de. “Letter to the Viceroy of Nueva Espana.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 256-72. Mandaluyong: Cachos Hermanos Inc., 1573.

_____________ and Tala Salinas-Ramos. Tuba Ethnohistory. Baguio City: Cordillera Studies Center, 1998.

O’Shea, John. “A River Runs through It: Landscape and the Evolution of Bornze Age Networks in the Carpathian Basin.” Journal of World Prehistory 24 (2011): 161-74.

_____________. “The Development of Studies in Cordillera Ethnographic Research.” Aghamtao 1 (1978): 45-70.

Panchal, Shailesh, Rahesh Thakker. “Implementation and Comparative Quantitative Assessment of Different Multispectral Image Pan-sharpening Approaches.” Signal and Image Processing: an International Journal 6, no. 5 (2015): 35-48.

Quirante, Alonso Martin. “Expedition to the Mines of the Igorottes.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 262-304. Mandaluyong: Cachos Hermanos Inc., 1624.

Parkinson, William and Michael Galaty. “Introduction: Interaction and Ancient Society.” In Archaic State Interaction: the Early Mediterranean in the Bornze Age, edited by William Parkinson and Michael Galaty, 3-28. Santa Fe: School for Advanced Research, 2009.

Rada, Martin de. “Letter to the Viceroy of Nueva Espana, Martin Enriquez Dated June 15, 1574.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 286-94. Mandaluyong: Cachos Hermanos Inc, 1574.

Pavao-Zuckerman, Barnet and Vincent LaMotta. “Missionization and Economic Change in the Pimeria Alta: The Zoo Archeology of San Agustin De Tucson.” International Journal of Historical Archeology 11 (2007): 241-68.

Ray, H. P. . “Early Maritime Contacts between South and Southeast Asia.” Journal of Southeast Asian Studies 20, no. 1 (1989): 42-54. Redfield, Robert. Folk Culture of Yucatan. Chicago: University of Chicago Press, 1941.

Paz, Victor J. Advancing History and Heritage through the Study of Excavated Early Spanish Structures in Oriental Mindoro: Report on the Bulalacao Initiative. Quezon City: Archeological Studies Program-University of the Philippines Diliman, 2006.

_____________. “The Social Organization of Tradition.” The Far Eastern Quarterly XV, no. 1 (1955): 13-21. Reid, A. Charting the Shape of Early Modern Southeast Asia: Silkworm books, 2000.

__________. Ulat Tungkol Sa Isang Pag-Aaral PangKasaysayan Sa Sitio Cuta, Barangay Anilao, Bongabong, Oriental Mindoro (Report on the Historical Study of Sitio

______. Southeast Asia in the Early Modern Era: Trade, Power, and Belief: Cornell University Press, 1993. 117

Remote Sensing the Margins of the Gold Trade Reyes, Isabelo de los. History of Ilocos Volume 2 (Translated by Maria Elinora Peralta-Imson). Quezon City: University of the Philippines Press, 2014 (1890).

_________________. Notices of the Pagan Igorot in the Interior of the Island of Manila/Francisco Antolin 17451796. Manila: University of Santo Tomas Press, 1988.

_________________. The Tinguian/Isabelo De Los Reyes (1864-1938); Translated from the Spanish by Ma. Elinora Peralta-Imson. Baguio City: Cordillera Studies Center, 2007.

_________________. Prehistoric Source Materials for the Study of Philippine Prehistory. Quezon City: New Day Publishers, 1984. Semper, Carl. “Trip through the Northern Provinces of Benguet.” In German Travellers on the Cordillera (18601890), edited by Filipino Book Guild, 1-16. Manila: Regal Printing Company, 1975.

Robertson, Benjamin. “Other New World Roads and Trails.” In Chaco Roads Project, Phase I; a Reappraisal of Prehistoric Roads in the San Juan Basin, edited by C. Kincaid. Albuquerque: US Department of the Interior, Bureau of Land Management, 1983.

Sever, Thomas and David Wagner. “Analysis of Prehistoric Roadways in Chaco Canyons Using Remotely Sensed Data.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Trumbold, 42-52. Cambridge: Cambridge University Press, 1991.

Rockman, Marcy. “Knowledge and Learning in the Archeology of Colonization.” In Colonization of Unfamiliar Landscapes: The Archeology of Adaptation, edited by Macy and James Steele Rockman, 3-24. London and New York: Routledge, 2003.

Sheets, Payson and Thomas Sever. “Prehistoric Footprints in Costa Rica: Transportation and Communication in a Tropical Rainforest.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Trumbold, 53-65. Cambridge: Cambridge University Press, 1991.

Sande, Francisco de. “Letter to Felipe Ii.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 21-98. Mandaluyong: Cachos Hermanos Inc, 1576.

Sheratt, Andrew and Susan Sheratt. “From Luxuries to Commodities: The Nature of Meditteranean Bronze Age Trading Systems.” In Bronze Age in the Meditteranean: Studies in Meditteranean Archeology, edited by N.H. Gale, 351-86. Jonsered: Paul Astrom Forlag, 1991.

Santley, Robert. “The Structure of the Aztec Transport Network.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Tromboldt, 198-210. Cambridge: Cambridge University Press, 1991.

Smith, Cecilia Alaina. “Haven Geographies and the Indigenous Prestige Economies of Spanish Colonial Philippines.” University of Illinois at Chicago, 2014.

Savellano, Deogracias. Ilocos Sur: An Illustrated History. Cabugao: Sanicua Publications, 2009. Scheerer, Otto. The Nabaloi Dialect. Manila: Manila Bureau of Public Printing, 1905.

Snead, James. “Trails of Tradition: Movement, Meaning and Place.” In Landscapes of Movement: Trails Paths and Roads in Anthropological Perspective, edited by James Clark Erickson Snead, and J. Andrew Darling, 42-60. Philadelphia: University of Pennsylvania Museum of Anthropology and Archeology, 2009.

____________. “On Baguio’s Past.” In German Travellers on the Cordillera (1860-1890), edited by Filipino Book Guild, 173-218. Manila: Regal Printing Company, 1975. Schiffer, Brian. Formation Processes of the Archeological Record. Albuquerque: University of New Mexico Press, 1987.

___________ Clark Erickson, and J. Andrew Darling, ed. Landscapes of Movement: Trails Paths and Roads in Anthropological Perspective. Philadelphia: University of Pennsylvania Museum of Anthropology and Archeology, 2009.

Scott, James C. . The Art of Not Being Governed: Yale University Press, 2009.

_________________. The Discovery of the Igorots. Quezon city: New Day Publishers, 1974.

___________ Clark Erickson, and J. Andrew Darling. “Making Human Space: The Archeology of Trails, Paths, and Roads.” In Landscapes of Movement: Trails Paths and Roads in Anthropological Perspective, edited by James Clark Erickson Snead, and J. Andrew Darling, 1-19. Philadelphia: University of Pennsylvania Museum of Anthropology and Archeology, 2009.

_________________. “Igorot Responses to Spanish Aims: 1576-1896.” Philippine Studies 18, no. 4 (1970): 695-717.

Snijders, Tom AB, ed. Network Dynamics. Edited by J. S. a. P. J. Carrington, The Sage Handbook of Social Network Analysis. London: Sage Publications, 2014.

Scott, William Henry. Cracks in the Parchment Curtain and Other Essays in Philippine History. Quezon City: New Day Publishers, 1982.

118

Bibliography Soares, Pedro, Jean Trejaut, Teresa Rito, Bruno Cavadas, Catherine Hill, Ken Khing Eng, Maru Mormina, Andreia Brandao, Ross Fraser, Tse-Yi Wang, Jun-Hun Loo, Christopher Snell, Tsa-Ming Ko, Antonio Amorim, Maria Pala, Vincent Macaulay, David Bulbeck, James Wilson, Leonor Gusmao, Luisa Pereira, Stephen Oppenheimer, Marie Lin, Martin Richards. “Resolving the Ancestry of Austronesian-Speaking Populations.” Human Genetics doi:10.1007/s00439-015-1620-z (2015).

Trumbold, Charles, ed. Ancient Road Networks and Settlement Hierarchies in the New World. Cambridge: Cambridge University Press, 1991.

Solheim II, Wilhelm G. “Philippine Prehistory.” In The People and the Art of the Philippines, edited by Casino Eric, George Elis, Wilhelm Solheim II, 17-83. Los Angeles: Museum of Cultural History, University of California, 1981.

Unknown. “Ydea Aproximada Del Teritorio Entre Cagayan E Ylocos.” 1794.

_______________. “An Introduction to the Study of Ancient New World Road Networks.” In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles Trumbold, 1-9. Cambridge: Cambridge University Press, 1991.

Unsigned. “Relation of the Philippine Islands.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson. Mandaluyong: Cachos Hermanos Inc, 1586.

Spriggs, Matthew. “Archeology and the Austronesian Expansion: Where Are We Now? .” Antiquity 85 (2011): 510-28.

Unsigned. “Report on the Offices Saleable in the Philippines.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 202-05. Mandaluyong: Cachos Hermanos Inc, 1582-1583.

Stein, Gil. Rethinking World Systems: Diasporas, Colonies, and Interaction in Uruk Mesopotamia. Tucson: University of Arizona Press, 1999.

Vance, Eric. Elizabeth Archie and Cynthia Moss. “Social Networks in African Elephants.” Computational and Mathematical Organization Theory 15 (2009): 273-93.

________. “Rethinking World Systems: Power, Distance, and Diasporas in the Dynamics of Regional Interaction.” In World Systems Theory in Practice, edited by Nick Kardulias, 275-96. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999.

Vega, Juan Manuel dela. “Expeditions to the Province of Tuy.” In The Philippine Islands 1493-1898, edited by Emma Blair & James Robertson, 281-326. Mandaluyong: Cachos Hermanos Inc, 1609.

Strassnig, Christian. “Rediscovering the Camino Real of Panama: Archeology and Heritage Tourism Potentials.” Journal of Latin American Geography 9, no. 2 (2010): 159-68.

Wallerstein, Immanuel. World Systems Analysis: An Introduction. Durkham and London: Duke University Press, 2004.

Tapang, Bienvenido. “Innovation and Social Change: The Ibaloy Cattle Enterprise in Benguet.” In Csc Monograph. Baguio: Cordillera Studies Center, 1985.

Wells, Peter. “Production within and Beyond Imperial Boundaries: Goods, Exchange, and Power in Roman Europe.” In World Systems Theory in Practice, edited by Nick Kardulias, 85-102. Lanham, Boulder, New York, Oxford: Rowman and Littlefield Publishers, 1999.

Tenrengren, Felix. “A Historical Review of Gold in the Philippines.” Philippine Journal of Science 92, no. 4 (1963).

Wernstedt, Frederick and Joseph Earl Spencer. The Philippine Island World: A Physical, Cultural, and Regional Geography. Berkeley and Los Angeles: University of California Press, 1967.

Terrell, John. “Island Biogeography and Man in Melanesia.” Archeology and Physical Anthropology in Oceania 11 (1976): 1-17.

Wescott, Konnie. “One Step Beyond: Adaptive Sampling and Analysis Techniques to Increase the Value of Predictive Models.” In Gis and Archeological Site and Location Modelling, edited by M. Mehrer. London and New York: Taylor and Francis, 2001.

__________. “Social Network Analysis and the Practice of History.” In Networks Analysis in Archeology, edited by C. Knappett. Oxford: Oxford University Press, 2013. Tidalgo, Jesus. Calincamasan and Aludaid: Contact Peirod Sites in La Union Northwestern Philippines. Manila: National Museum, 1979.

Wilson, Lawrence Lee. Igorot Mining Methods. Manila: published by author, 1932.

Tolentino, Delfin Jr. Resistance and Revolution in the Cordillera. Baguio: University of the Philippines College Baguio, 1994.

Yabes, Leopoldo. “The Ilocano Epic: A Critical Study of the Life of Lam-Ang.” Philippines Social Science and Humanities Review 23, no. 2-4 (1958): 283-338. 119