Environmental and Demographic Change and Rural Violence in Peru: a Case Study of the District of Chuschi, Ayacucho

Environmental and Demographic Change and Rural Violence in Peru: A Case Study of the District of Chuschi, Ayacucho

Athanasios (Tom) Deligiannis

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Political Science University of Toronto

Environmental and Demographic Change and Rural Violence in Peru: A Case Study of the District of Chuschi, Ayacucho

Athanasios (Tom) Deligiannis

Doctor of Philosophy

Department of Political Science University of Toronto

2020 Abstract

Considerable debate exists about whether and how human-induced pressure on the natural environment contributes to violent conflict. This dissertation examines whether human-induced environmental and demographic change helped generate rural violence in the District of Chuschi, in the South-Central Andes of Peru in the latter half of the 20th century, in the years leading up to the Sendero Luminoso insurgency in 1980.

The dissertation’s findings reveal that long-term patterns of resource capture and human-induced environmental and demographic change, combined with prevailing local, regional, and national political-economic currents in Peru over the 20th century to increase inter-community competition for valuable but limited agricultural resources – over range lands, crop lands, and water resources. In particular, these changes conditioned patterns of violence between two neighbouring communities in the District – the Community of Chuschi and the Community of

Quispillacta. Changes aggravated conflictual relationships between Chuschi and Quispillacta over rights to long-contested resources that were essential to increasingly pressured livelihoods in both communities. Change and contest resulted in winners and losers, with implications for social stability years into the future. Quispillacta largely emerged as the victor in these community resource conflicts. However, their victory fostered enduring resentment in Chuschi - ii the traditional administrative centre of the District - and among Chuschi’s associated hamlets.

These grievances spilled out during the Sendero insurgency as some in rival communities falsely accused Quispillacta of being a centre of Sendero militancy, accusations likely stemming from community members who lost out in the inter-community land competition. Quispillacta suffered terribly as a result in the Peruvian military’s counter-insurgency bloodletting, disproportionately more than Chuschi and its associated hamlets – score-settling by counter- insurgency.

Human pressure on the local environment and people’s adaptive choices to these pressures thus helped condition patterns of violence in the decades before the Sendero insurgency, and patterns of violence once the insurgency heated up in the area. Understanding why violence took the shape that it did in Chuschi during the insurgency and over the previous century in various small- scale confrontations requires examining the interface of human-induced environmental change and demographic change, in the relation to socio-economic changes in the region.

iii

Acknowledgments

I am deeply grateful to my family, friends, and colleagues, who have stood by me and supported me for many years as I completed this dissertation. First, I want to thank Tad Homer-Dixon for being the intellectual inspiration for this project, and for not giving up on me as the years passed and my dissertation remained unfinished. Without Tad’s continued support and encouragement, I would never have reached the end. Tad’s research and guidance over the years have pushed me to uncover and make sense of the complex local reality in Peru. I also want to thank the other members of my committee, Steven Bernstein at the University of Toronto, and Simon Dalby at the Balsillie School of International Affairs. Both have been extremely understanding of my progress, and provided tremendously insightful and helpful feedback on my dissertation. Gavin Smith was an early member of my committee who, unfortunately, had to leave the committee as the years dragged on; however, his participation and support was very important, given his extensive work in highland Peru. Kate Neville provided outstanding feedback and suggestions for improvement as the internal reader. Joshua Busby of the University of Texas at Austin made insightful suggestions as the external reader of this dissertation.

Many people assisted me during my fieldwork in Peru. I would like to thank Kimberley Theidon for putting me in touch with Marcela Machaca at ABA, in Huamanga, Ayacucho. Marcela’s help was crucial for connecting me with my outstanding field assistants in the District of Chuschi, Hereberto Nuñez Mejiá and Mario Silverio Huamaní Conde. I owe a deep debt of gratitude to Hereberto and Mario for their assistance in arranging interviews and field assessments in Quispillaccta and Chuschi. I also want to thank Edwin Medina Ramírez, who accompanied us during our fieldwork in the District and assisted me with translation. The fieldwork would have been impossible without Hereberto, Mario, and Edwin. Don Theodosio Flores Galindo graciously allowed me to stay in his house in Llacctahurán during my fieldwork in the District. Don Theodosio’s taciturn presence was intimidating at first; but also inspired me to dig deeper to get to know the reality facing comuneros in the District. I would also like to thank the elected authorities of Chuschi and Quispillaccta for granting permission for my study: Vincente Chaupin Huaycha, who was Alcalde of Chuschi at the time, and the Varayoc of Quispillaccta – the traditional authorities. Over sixty community members in Chuschi and Quispillaccta graciously agreed to interviews during my time in the District, and I am eternally grateful to them for allowing me learn about their lives. In Huamanga, Yovanna Mendieta

Tacuri and Hugo Sarmiento Medina assisted me with research. In Lima, my research assistant Vivien Weiner was outstanding in helping me connect with various Peruvian scholars and NGOs. I would also like to thank Martin Scurrah and Igidio Naveda Felix, who were then with Oxfam America, for their generous assistance and advice. Raul Grados and Luis G. Orellana graciously provided support and hospitality during my trips to Lima.

During my time at UPEACE, in Costa Rica, Dr. Rolain Borel, Dr. Ronnie De Camino, Dr. David Hoffman, and Jan Breitling were outstanding colleagues who helped me understand the Central American agro-ecological reality. In Canada, I’m lucky to have wonderful friends and colleagues who have encouraged and supported me in completing my dissertation: Dr. Moshe Khurgel, Dr. Craig Johnson, Dr. Francine McKenzie, Dr. Jodi Salter, and Dr. Alicia Sliwinski, have been particularly important. Jodi’s encouragement two years ago to join her dissertation boot camp at the University of Guelph was instrumental in getting my dissertation organized and back on track.

Finally, but most importantly, this project would never have been completed without the unconditional love and support of my family – especially my best friend and partner, Kelly Main, and my kids Eva and John. Kelly probably wants me to end this sentence as fast as possible, because then the dissertation will be finally finished. For the sacrifices that she’s made, her name should also be on the front cover. I promise that I’ll never again take my research or my marking on family holidays. Lastly, I want to thank my parents Elias and Evangelia Deligiannis, who showed me what could be achieved with hard work, even if you start with nothing. They were my original introduction into the lives of αγρότες, and I’ve never stopped learning from them.

Table of Contents

Acknowledgments...... iv

Table of Contents ...... vi

List of Tables ...... x

List of Figures ...... xi

List of Appendices ...... xiii

Preface...... xiv

Chapter 1 Introduction ...... 1

Overview of case study region ...... 3

Key findings and plan of dissertation...... 15

Timeline of key events ...... 26

Chapter 2 Literature Review and Research Design: Do environmental scarcities impact violent conflict onset? ...... 29

Introduction ...... 29

Theories of revolution and rural rebellion ...... 33

Research linking environmental change and violent conflict ...... 42

Specific explanations for unrest in Peru...... 48

2.4.1. Pre-land reform unrest ...... 48

2.4.2. Post-Land Reform unrest ...... 55

2.4.3. Debates on the origins of Sendero and the degree of peasant support ...... 59

2.4.4. Comprehensive theories of the origins of Sendero’s revolt ...... 61

2.4.5. Criticizing comprehensive theories: Sendero exceptionalists...... 63

The state of the peasantry ...... 68

2.5.1. Ecological and demographic Influences ...... 68

Research focus and research design ...... 74

2.6.1. Key questions ...... 74

Chapter 3 The Evolution of Environment-Conflict Research: Toward a Livelihood Framework ...... 79

Introduction ...... 79

Environmental change and the role of the state – mediating scarcity, undermining capacity, or actively exploiting ...... 80

Shifting the focus from the state to household livelihoods: A household-livelihood framework for environment-conflict research ...... 82

Conclusion ...... 97

Chapter 4 Historical Structural Changes and Transformations in Rio Pampas Affecting Natural Assets and Livelihoods in Chuschi and Quispillaccta ...... 99

Introduction ...... 99

Natural asset capture and entitlement capture in Western Cangallo – Pre-Conquest to 20th Century ...... 101

4.2.1. Pre-Colonial and Colonial developments ...... 101

4.2.2. Mestizo elite and Catholic Church: impacts in the 19th and 20th Centuries ...... 118

4.2.2.1. Regional Structural Change, Patterns of Elite Capture, and Land Holding Patterns in Cangallo ...... 118

4.2.2.2. Communities in Conflict? Revealing the Impacts of Landed Estates and Mestizo Elites in Western Cangallo ...... 128

4.2.2.3. Elite Capture of Scarce Cultivated Lands in the District of Chuschi .....133

4.2.3. Conclusion ...... 142

Chapter 5 Elite Capture of High-Altitude Pasture Lands in the District of Chuschi ...... 144

Introduction ...... 144

Livelihoods and the high-altitude grazing zones of the district ...... 145

The history of elite capture of high-altitude grazing lands in the District of Chuschi .....152

5.3.1. Community struggles to regain control of captured high-altitude grazing lands.156

5.3.2. Chuschi’s Yaruca dispute on the north-west borders ...... 164

5.3.3. Quispillaccta’s dispute over Hacienda Quicamachay lands on the NE borders ..168

vii

Chapter 6 Controlling the Centre: The Struggle to Control the High Sunni and Puna Zones of the Rio Cachi Basin in the District ...... 178

Introduction ...... 178

Disputes in the 1940s and 1950s to control the centre ...... 179

Factors drawing district members to settle the high Sunni and Puna zones ...... 183

Inter-community clashes in 1960 over control of disputed high Sunni and Puna zones .197

Détente and maneuver – continuing struggles to control disputed zones between Chuschi and Quispillaccta, 1960s and 1970s ...... 201

Peace in the midst of civil war – final settlement of land disputes in the District of Chuschi and the impacts of this settlement in the context of Sendero’s insurgency in the area ...... 204

6.6.1. The 1981 peace agreement between Chuschi and Quispillaccta ...... 204

Conclusion ...... 216

Chapter 7 Supply and Demand Pressures in the District of Chuschi ...... 220

Introduction ...... 220

Demand-induced pressures on livelihoods in the District of Chuschi ...... 221

Demand-induced impacts on livelihood resources in the District of Chuschi ...... 231

7.3.1. Perspectives on the peasant household ...... 233

7.3.2. Trends in declining land availability...... 235

Adaptations and their impacts ...... 243

7.4.1. Migration...... 243

7.4.2. Extensification and intensification of land use ...... 247

Precipitation and temperature changes...... 272

Declining agricultural yields and crop quality ...... 277

Conclusion ...... 287

Chapter 8 Living Between the Sword and the Stone ...... 290

Introduction ...... 290

viii

Violent revenge and the Peruvian Government’s counter-insurgency campaign in 1983...... 291

Explaining violence in the district in the early 1980s in the context of historical land conflicts ...... 300

Chapter 9 Conclusion ...... 308

Qualitative fieldwork pays off – delayed temporal impacts of resource capture and revised understanding of conflict causes ...... 308

Revising theoretical links between environmental change and violent conflict ...... 312

Scarcities can condition violence during violent conflict ...... 316

Scarcities shape structures and processes, rather than trigger conflict ...... 321

Livelihoods should be at the centre of environment-conflict research ...... 323

Postscript ...... 326

Bibliography ...... 333

Appendices ...... 350

List of Tables

Table 1: District of Chuschi Land-Use According to Agro-Ecological Zone………………….136 Table 2: Quispillaccta’s 1960 Population: Main Town and Annexes………………………….187

List of Figures

Figure 1: Map – Department of Ayacucho……………………………………………………...... 2 Figure 2: Map – Province of Cangallo……………………………………………………………3 Figure 3: Map – District of Chuschi………………………………………………………………4 Figure 4: Toronto Group’s Model of the Causal Links between Environmental Scarcity and Violence …………………………………………………………………………………………44 Figure 5: Watters’ Model of the Peasantry ...... ………………….………………………………69 Figure 6: Rural Household Livelihood Framework for Environment-Conflict Research……….90 Figure 7: Peru Population: Pre-Conquest (1520) to 2007…………………………………...….108 Figure 8: Large and Small Landholdings as a Total Percentage of Agricultural Land in Ayacucho’s Provinces, 1961 & 1972…………………………………………………………...124 Figure 9: Department and Provincial Comparison of Percentage of Land Held by Those with Less Than 5H, 1972………………………………………...……………………………...…...125 Figure 10: Map - Province of Cangallo and the Location of the Hacienda Pomacocha….....….126 Figure 11: Map - Agro-ecological Zones in the District of Chuschi….……..…………………135 Figure 12: Map - Major Land Disputes in the District of Chuschi in the 20th Century…...…...145 Figure 13: Map - Rio Cachi Watershed in Vinchos and Chuschi Districts….………....………147 Figure 14: Map - Disputed Territory Between Chuschi and Hacienda Yaruca in NW of District of Chuschi………………….………………………………………………………………..….164 Figure 15: Map - Disputed Territory in Land Conflict Between Quispillaccta and Hacienda Putaje in NE of District…………....……………………………………………………………169 Figure 16: Map - Disputed Zone between Chuschi and Quispillaccta, 1940…….…………….179 Figure 17: Map - Disputed Zone between Chuschi and Quispillaccta, late 1950s to 1980s…...188 Figure 18: Total Cattle in Peru and Ayacucho, 1950-1980…………………………………….192 Figure 19: Map - Current boundaries of Quispillaccta (black) within the District of Chuschi overlaid with historic conflict zones…..………………………………………………………..212 Figure 20: Population of Ayacucho’s Provinces, 1876-1993…………………………………..223 Figure 21: Population of Department of Ayacucho and Province of Cangallo, 1791-2007……224 Figure 22: Population of Province of Cangallo and District of Chuschi, 1876-2007……….….225 Figure 23: Population of Province of Cangallo and District of Chuschi, 1940-2007……….….226 Figure 24: Chuschi and Quispillaccta: Births and Baptisms, 1850-2000………………………228 Figure 25: Cultivated Land (H) per Capita in the District of Chuschi……………………….…240 xi

Figure 26: Altitudinal Shifts in Maize Zone in District of Chuschi……………………………259 Figure 27: Map - Land Use in the District of Chuschi………….……………………………...269 Figure 28: Average Annual Rainfall Chuschi Station and Rio Cachi Basin Station (mm)…….275 Figure 29: Chuschi Station Precipitation, 1963-1981…………………………………………..276 Figure 30: Comparison of Historical and Contemporary Corn Yields in Sacks/Yugada in Quispillaccta and Chuschi According to Interviews……………………………………………281 Figure 31: Historical Potato Yields in Sacks/Yugada in Quispillaccta and Chuschi According to Interviews…………………………………………...…………………………………………..282 Figure 32: Environmental Scarcity-Migration-Conflict Links Worsened Group Conflict Rather than Contributing to Insurgency Onset...... 312

xii

List of Appendices

Appendix 1: Field Measures of Soil Erosion in Quispillaccta………………………………….348 Appendix 2: Population of Peru, Ayacucho, Province of Cangallo, and District of Chuschi….350 Appendix 3: Notes on the Use of Box and Whisker Plots…………………..………………….351 Appendix 4: Map – Department of Ayacucho………………………………………………….354 Appendix 5: Map – District of Chuschi and Surrounding Districts…………………………….355 Appendix 6: Map – High Resolution Topographical Map of District of Chuschi……………...356

xiii

Preface

The question at the heart of this dissertation is not the question that led me to the southern highlands of Peru. I went to Peru hoping to discover whether environmental scarcities and demographic change helped cause the Sendero Luminoso (Shining Path) rebellion. This question anchored my environment-conflict research in the long tradition of research on rural rebellions and revolutions. But field work in the District of Chuschi (D. Chuschi), where the Shining Path initiated the violence in 1980 that would plague Peru for more than a decade, instead led me to a puzzling local story and a different question.1

In early November, 2004, I was in the isolated Quispillaccta hamlet of Huertahuasi, in the D. Chuschi, Ayacucho. I’d been in the district for almost two months, interviewing elderly residents in various hamlets, trying to piece together long-term patterns of livelihood change. Memories were still raw from the violence of the 1980s, so I rarely asked direct questions about events in the years of the Sendero insurgency. Instead, I focused on understanding how local livelihoods had evolved in the decades leading up to 1980, when Sendero initiated its violent challenge to the Peruvian state by burning the ballot boxes in the D. Chuschi for the upcoming national election.

On this day, however, an informant was talking about the manchay tiempo – the ‘times of fear’ during the Sendero insurgency and the government’s counter-insurgency campaign.2 He described the terrible ordeal that he endured at the height of the civil war violence in the district when he was the only survivor of a Peruvian Army massacre of community members from Quispillaccta in early June, 1983. Captured at the end of May in a sweep of Quispillaccta’s high-altitude hamlets by a platoon of Peruvian Army soldiers accompanied by a group of almost one hundred peasants from the neighbouring communities of Chuschi and Cancha Cancha, my informant explained in a calm, measured voice how he and 15 other bound Quispillacctinos from

1 In order to differentiate the community of Chuschi from the District of Chuschi, I will use D. Chuschi to refer to the District of Chuschi. All other usages of Chuschi refer to the community of Chuschi, one of several peasant communities in the District of Chuschi.

2 Interview, Identity withheld, Huertahuasi, Quispillaccta, 5 November, 2004. Throughout this dissertation I have carefully weighed whether or not to identify my local informants. If the information is sensitive or possibly defamatory, I have kept identities private. This mixed approach to obscuring interview identities is common today among scholars working in Peru on sensitive post-Sendero topics. xiv

5 different hamlets were marched to a school in Chuschi’s town centre, where the captives were beaten and held for one week without food and water, before being shipped by helicopter to the nearby Peruvian army base in the District of Totos. There, they were again tortured and questioned to identify Senderista militants in Quispillaccta – torture that included vicious beatings and being hung from the building’s rafters.

That afternoon my informant overheard his guards say that the “Quispillacctinos say goodbye today.” Later that night, he and his fellow captives were roughly led, with hands bound, to a nearby brushy spot, on the edge of a deep ravine, and given shovels to dig two large pits. It was my turn to dig together with Francisco Núñez of Cuchoquesera. On the other side was Moisés Huamaní Ccallocunto. I got tired of digging; my hands hurt and I said to Francisco: ‘Pancho, it’s your turn make the hole. ‘When I came out of the hole, already deep, I put my hands in my pocket and found a handful of coca and a little sugar. In my mouth I felt a sweet flavor. Then I looked up and back in the darkness. While the soldiers were distracted with other prisoners, I decided – in a blink of the eye to escape and I threw myself into the ravine full of thorns and rocks.3

The soldiers noticed his escape and rained gunfire around him. Injured from the fall, and hiding in a large bush, he evaded the bullets and the soldiers’ attempts to find him, and later listened as his fellow captives were killed and buried. When the sounds of the soldiers and their shoveling stopped, he carefully made his way back to Quispillaccta. Years later, he was able to tell his remarkable story to Peru’s Truth and Reconciliation Commission, identifying the mass grave of

3 Quoted in an account of the killing of Moisés Huamaní Ccallocunto, one of the victims of the Sankaypata massacre, as told to the victim’s family by my informant. See Felimón Salvatierra, Honorato Méndez, and Oseas Núñez, La Vida Ya No Era Vida: Un homenaje a la vida y memoria de las víctimas de Allpachaka, Chiara y Quispillaqta (Huamanga: Hivos/Impunity Watch/Paz y Esperanza, 2016), 249. His account published in Salvatierra et al. is consistent with what he told me in 2004. But he adds a few more details in this account compared to what he told me, so I have decided to use these words. xv his fellow captives.4 None of the dead were Sendero militants, as far as my informant knew; only innocent Quispillacctinos, caught up in an orgy of violence.5

My informant was lucky; luckier than the dozens of other community members from Quispillaccta who were murdered by Sendero and military forces during the first years of Sendero’s insurgency. There were several cases in May and June 1983 when small groups of Peruvian Army soldiers accompanied by groups from Chuschi and associated hamlets rampaged across Quispillaccta’s high-altitude barrios, looting and destroying houses as they went, and capturing numerous innocent Quispillacctinos in the wrong place at the wrong time. Quispillaccta, in fact, suffered far more than any other community in the D. Chuschi during the war, both at the hands of Sendero and the Peruvian security forces. According to Gustavo Flórez Salcedo, between 1980 and 1985, 73 Quispillacctinos were murdered or disappeared in the district, compared to only 15 from Chuschi.6 Field work in the D. Chuschi in 2006-07 by Marté Sánchez Villagómez, based on records kept by community authorities, list 140 Quispillacctinos and 16 Chuschinos killed after the government’s counter-insurgency operations began in the district in 1983.7 Quispillaccta’s high-altitude hamlets in the Rio Cachi watershed were particularly hard hit, with 58 murders and disappearances, 3 rapes, and 146 cases of torture

4 The Peruvian Army massacre in Sankaypata is outlined in: Peruvian Truth and Reconciliation Commission Report (TRC), Book 7, Chapter 2, Section 2.3. (http://www.cverdad.org.pe/ifinal/pdf/TOMO%20VII/Casos%20Ilustrativos- UIE/2.3.%20TOTOS%20SANCAYPATA.pdf), Accessed June 26, 2019. Kimberly Theidon provides a careful discussion of Peru’s Comisión de la Verdad y Reconciliación (Truth and Reconciliation Commission) in her outstanding book on violence and reconciliation in Peru, in part based on her experiences as a researcher for the TRC in Ayacucho. See Kimberly Theidon, Intimate Enemies: Violence and Reconciliation in Peru (Philadelphia: University of Pennsylvania Press, 2013). 5 Peru’s TRC report on the massacre notes that those take by the military were on a list of supposed Sendero militants or they were accused by the Chuschino community members accompanying the soldiers on their sweep of Quispillaccta. 6 Gustavo Flórez Salcedo, “Rivalidades comunales y contiendas electorales: micropolítica en las elecciones distritales de Chuschi. El caso de las comunidades campesinas de Chuschi y Quispillaccta,” in Alejandro Diez Hurtado ed., Tensiones y transformaciones en comunidades campesinas, CISEPA-PUCP. Lima, 2012: 234. There is no definitive accounting of the district’s dead, missing, wounded, and tortured during the Sendero insurgency. Different authors provide different figures, though all agree that Quispillaccta suffered far more casualties than Chuschi and its annexes. Figures on rapes and sexual violence are almost certainly under-reported, given the on- going trauma and social stigma around sexual violence in Peru. See Theidon, 2013, chapter 5. 7 Marté Sánchez Villagómez, El Horror Olvidado: Memoria e historia de la violencia política en Ayacucho, Perú (1980 -2000), PhD Diss., Departament d'Antropologia Social i de Prehistòria, Universitat Autònoma de Barcelona, 2015, 230. xvi recorded in the five barrios of Catalinayuq, Cuchoquesera, Pampamarca, Puncupata, and Unión Portrero between 1980-1985.8

Early in the conflict, Quispillaccta gained a reputation as a hotbed for Sendero support in the district. Press reports in the mid-1980s identified two of the ballot box burners as Quispillacctinos, though it now appears that they played relatively minor roles in that attack.9 In interviews, a local evangelical pastor went further and laid the blame more widely on Quispillaccta: “In all honesty, I’m not going to lie … That was [a job done] by the people of Quispillaccta. This is something of which the [police] posts in Cangallo and the police station in [Huamanga] are well aware. That night of the burning of the ballot boxes, there was a man [in charge] who came from another place, and that man acted with men from Quispillaccta; it was them.”10 Such testimony, and comments from Chuschinos have convinced some, like historian Miguel La Serna, that Quispillaccta was a Sendero stronghold, and that Quispillacctinos

8 Salvatierra et al., La Vida, 218. 9 Gustavo Gorriti, The Shining Path: A History of the Millenarian War in Peru (Chapel Hill, NC: The University of North Carolina Press, 1999), 17; Oseas Nuñez Espinoza, Cuchoquesera, http://www.tutiempo.net/Tierra/Pe1ru/Cuchoquesera-PE018555.html (accessed June 24, 2019). Four Sendero militants and the Municipal Registrar were involved in the burning of D. Chuschi’s ballot boxes. Nuñez notes that one of the four was a 16-year old girl named Alejandra. It is not clear if Alejandra was from Quispillaccta. Two older Senderista cadres were not from the district - a teacher at Chuschi’s Ramon Castilla secondary school, Bernardino Azursa, possibly of Huancaíno origin, and a 20 year-old named Victor from Andahuaylas province, in the Department of Apurímac. According to Nuñez Espinoza, Victor was living for some time before the attack in a house in Chuschi, rented from the Meneses family, close to the creek dividing Chuschi and Quispillaccta. He and Alejandra travelled about the district spreading Sendero’s message and organizing popular schools to indoctrinate residents. I have not found any information about the fourth Sendero militant involved that day. The Municipal Registrar, Lorenzo Conde, from Quispillaccta, was forced to open the municipal offices to the Senderistas, according to Nuñez Espinoza. Peru’s TRC notes that the Sendero teacher from Andahuaylas who helped burn the ballot boxes was named, José. Peruvian Truth and Reconciliation Commission Report, Volume 5.1, Section 2.1, 18. La Serna notes, based on an anonymous interview with a then-young Chuschino Senderista, that a Senderista teacher from outside the community organized the first popular schools in the district out of a house rented from a Chuschino butcher, who was also suspected of running a cattle rustling operation. However, there is no mention about whether this Senderista teacher was also involved in burning the ballot boxes in D. Chuschi, and La Serna’s account of where the teacher lived does not corroborate the account of Nuñez Espinoza. Miguel La Serna, The Corner of the Living: Ayacucho on the Eve of the Shining Path Insurgency (Chapel Hill, N.C.: The University of North Carolina Press, 2012), 145-147. Theidon notes that one of those who burned D. Chuschi’s ballot boxes was later Shining Path’s military commander for the Zonal Committee in eastern Cangallo, in Vilcashuamán. Theidon, Intimate Enemies, 328-9. In a newly published account of the ballot box incident, Orin Starn and Miguel La Serna claim that five masked men accosted the Municipal Registrar and burned the election ballots. They identify the Registrar as Florencio Conde. Unfortunately, they offer no footnotes or citations to explain their sources. Orin Starn and Miguel La Serna, The Shining Path: Love, Madness, and Revolution in the Andes (New York: W. W. Norton, 2019), 85-6. 10 Quoted in La Serna, The Corner of the Living, 162. xvii embraced Sendero far more than other communities in the district, using the insurgency as a way to get back at Chuschinos, their historic enemies.11

However, recent interviews with survivors of the dirty war and their families and field work in the district raise serious questions about claims that Quispillaccta was the main source of Sendero support in the district and that their higher death rate in the civil war resulted from Peru’s security forces correctly targeting a community that had more Sendero militants and supporters. Such claims are puzzling because, after extensive research piecing together the centuries-long land conflicts in the district, it appears that Quispillaccta had the most to lose from throwing its support behind Sendero’s insurgency. They were the clear victors in the long- running district land conflicts with Chuschi and its annexes. Inter-community land confrontations between the 1940s and early 1980s and the legal trials that spawned from these conflicts were decisively won by Quispillaccta. When a final peace agreement was signed between Chuschi and Quispillaccta in late 1981, as Sendero was tightening its grip on the D. Chuschi, Quispillaccta consolidated its control over most of the long-contested lands in the Rio Cachi basin and central Kimsa Cruz zone of the district. (See district map, Figure 19, Chapter 6) The current boundaries of the Community of Quispillaccta show it holding a vast area of once- contested high-altitude puna lands in the centre of the D. Chuschi.

These facts raise an important question about whether the nature of violence during the civil war in the district was influenced by the outcome of long-running land conflicts between communities? Anthropologist Billie Jean Isbell hinted at this in 2005, when she wrote that, “Quispillaccta was the target of such strong reprisals due to the old litigation that it maintained with its neighbors because of disputes over land boundaries and not because it supported Sendero Luminoso more.”12

This puzzle about how violence during the civil war was linked to long-running resource conflicts in the district lies at the heart of this dissertation. The thesis traces the origin and

11 La Serna, The Corner of the Living, 163. 12 Isbell, my translation, Spanish translation of To Defend Ourselves, Billie Jean Isbell, Para defendernos: ecología y ritual en un pueblo andino, vol. 6 (Cuzco: Centro de Estudios Regionales Andinos Bartolomé de las Casas, 2005). Isbell’s comments on Quispillacta’s attitudes toward Sendero are more definitive in the this later Spanish translation of her book, compared to the 1985 English edition. xviii development of the district’s land conflicts and the role played by environmental scarcities and demographic change in the district over many decades, in the context of evolving economic, political, and social currents in the region more generally, to help us understand how and why violence happened in the latter twentieth century, and why one community suffered to a much greater extent than other communities in the D. Chuschi during the early years of the counter- insurgency operations in 1983 and 1984. While the dissertation does offer some suggestions about the original question driving my research in Peru, the main findings from archival and field work speak less to the causes of Sendero’s support and more to how environmental scarcities and demographic change can help to condition patterns of violence during civil wars. We cannot understand why Quispillaccta suffered more in the anti-Sendero dirty war, or why Quispillaccta was labelled as a hot bed of Sendero support, unless we understand the long history of conflict over land, water, and grazing opportunities in the D. Chuschi between different communities and Quispillaccta’s ultimate victory in most of those conflicts. These findings thus offer new insights about the role of environmental scarcities and demographic change in violent conflict – not their role in violent conflict generation, but instead their role in explaining “the logic of violence in civil war.”13 This is a burgeoning area of civil war research over the past fifteen years, but not a theme linked by environmental conflict research to date.

So, the logic of the dissertation’s layout reflects my attempt to grapple with the challenge of starting down one research path, formulating a livelihood framework to gain insights on the environmental change-revolution/insurgency pathway, but then getting into the field and finding evidence that human environmental change – in the context of many other key factors and influences – instead helped to condition patterns of violence between communities in and around the district, rather than contributing to the rise of Sendero’s insurgency in this area.

13 Stathis Kalyvas, The Logic of Violence in Civil War (Cambridge, Cambridge U.P., 2004). xix

Chapter 1 Introduction

Considerable debate going back to the 1990s exists among scholars and policymakers about whether and how human-induced pressure on the natural environment contributes to violent conflict. Recent research about the impact of climate change on social stability has furthered the intensity of debates about environmental change-conflict linkages. No consensus exists among scholars about key causes, key causal mechanisms, or the best methodologies to move the field forward.14 Over the past twenty years, most of the findings on these questions have emerged from qualitative, desk case-studies and large-scale statistical studies. Qualitative research on the linkages between environmental change and violent conflict, however, has stalled since the burst of work in the mid-1990s.

Most scholars agree that one strategy for moving the field forward and critically building on the work to date requires detailed field studies of environment-conflict linkages, sensitive to the local variability and complexities of socio-economic and political relationships around human- induced environmental change.15 Such studies can deepen our understanding of whether human- induced environmental change plays a role in generating violent social conflict, and if so, how these linkages may operate.16

This rationale forms the background to this dissertation’s examination of whether human- induced environmental and demographic change in the D. Chuschi played a role in generating the Sendero uprising in the South-Central Andes of Peru in the latter half of the 20th century.

14 Tobias Ide, “Why do conflicts over scarce renewable resources turn violent? A qualitative comparative analysis,” Global Environmental Change 33 (2015): 61-70. 15 Richard A. Matthew and Geoffrey D. Dabelko, “Environment, Population, and Conflict: Suggesting a Few Steps Forward,” Environmental Change & Security Project Report, Issue 6 (Washington D.C., The Woodrow Wilson Center, Summer 2000), 100; Tobias Ide and Jurgen Scheffran, “On climate, conflict and cumulation: suggestions for integrative cumulation of knowledge in the research on climate change and violent conflict,” Global Change, Peace & Security,” 26, no.3 (2014): 277. 16 Case studies are highly effective in identifying causal mechanisms linking environmental change and conflict. See Daniel M. Schwartz, Tom Deligiannis, and Thomas Homer-Dixon, “The Environment and Violent Conflict,” in Environmental Conflict, Paul F. Diehl and Nils Petter Gleditsch, eds. (Boulder, CO: Westview Press, 2001), 280-88. This is an area that some scholars have recently argued needs more attention in future research. See Halvard Buhaug, “Climate-conflict research: some reflections on the way forward,” Climatic Change 6 (May/June 2015): 271.

1 2

The D. Chuschi is an important case.17 Residents are acutely dependent upon natural resources for their livelihood security. Equally important, the D. Chuschi was the location of the first armed actions by the Sendero Figure 1: Department of Ayacucho Luminoso uprising in 1980, where they burned the ballot boxes for the upcoming presidential elections. In the early years of Sendero’s insurgency in Peru, the D. Chuschi and the wider province of Cangallo where it is situated, suffered disproportionately along with several other provinces in the Department of Ayacucho from high levels of civilian casualties as a result of violence between the Sendero militants and the Peruvian military’s counter-insurgency campaign. The first years of Sendero’s insurgency serve as the end point for this dissertation, as I trace back through detailed field work, over 60 ethnographic field

17 This case is not a “crucial” case, in social science methodological terms. Instead, this case fits more closely to a “most-likely” case, according to case study design discussed by Alexander George and Andrew Bennett. Crucial cases are rare, according to Harry Eckstein. Instead, the D. Chuschi case provides a good fit with a most-likely case, according to George and Bennett. “In a most-likely case, the independent variables posited by a theory are at values that strongly posit an outcome or posit an extreme outcome.” Alexander L. George and Andrew Bennett, Case Studies and Theory Development in the Social Sciences (Cambridge, MA: Harvard U.P., 2005), 170. Given the reliance of district residents on natural resources for household survival and given the tremendous demographic change that the area went through in the twentieth century, the D. Chuschi seems to provide an excellent single case to examine the applicability of Homer-Dixon’s theory linking environmental scarcities to violent conflict. As George and Bennett note, “Single cases serve the purpose of theory testing particularly well if they are “most- likely,” “least-likely,” or “crucial” cases.” George and Bennett, Case Studies, 111. Crucial cases and their limitations are discussed in Gary King, Robert O. Keohane, and Sidney Verba, Designing Social Inquiry: Scientific Inquiry in Qualitative Research (Princeton, NJ: Princeton U. P., 1994), 209-11.

3 interviews with elderly residents, and through archival research in Peru and North America, the role that human-induced environmental change and demographic change played in violence in the D. of Chuschi.18 This dissertation employs a process-tracing methodology throughout to help make sense of these relationships across space and time.19

Overview of case study region

The following study focuses on assessing the local impact of scarcities on peasant livelihoods and local patterns of violence by conducting a case study in the D. Chuschi, Cangallo province, Department of Ayacucho.20 Covering Figure 2: Province of Cangallo about 1916 km2, Cangallo is currently divided into 6 districts with a total population of about 35,000 people.21 Within the province, two clearly defined zones can be identified. The western zone, encompassing the districts of Totos and Paras, is isolated by geography and Source: Municipalidad Provincial De Cangallo. Mejoramiento y Ampliación del Servicio de Agua Para Riego en 07 comunidades del the absence of road connections from the distrito de los Morochucos – Cangallo – Ayacucho. Cangallo, Ayacucho, Octubre 2012.

18 Most of my interviews were with men in Chuschi and Quispillaccta. I have only one formal interview with a female head of a household in the district. I am acutely aware of the gender limitations of my fieldwork in the District, a region where gendered notions of authority and power are still a daily reality. On a few occasions, wives joined their husbands during my interviews. Other times, I could tell that women in the household were listening to our interview from nearby. Sometimes the female member of the household would shout out a comment, in response to something her husband had just said to me. In retrospect, I regret not making a greater effort to speak to women in the district, or use a female assistant in my field research. As much as possible, I have tried to supplement my fieldwork interviews with the voices of local women, particularly widowed survivors of the dirty war in the early 1980s. 19 Andrew Bennett and Jeffrey T. Checkel, "Process Tracing: From Philosophical Roots to Best Practices," in Process Tracing: From Metaphor to Analytical Tool, ed. Andrew Bennett and Jeffrey T. Checkel (Cambridge: Cambridge University Press, 2015).

20 Peruvian “departments” are analogous to Canadian provinces or US states, while Peruvian “provinces” are roughly analogous to Canadian municipalities. Provinces are composed of several districts in Peru. Each district has a designated district town centre, and numerous associated communities and hamlets. In recent years, Peru has begun a regionalization process to combine departments into regions. However, no regions have yet to be created, and departments remain the main administrative jurisdiction below the national level. Confusingly, departments are now led by regional governments. 21 In 1984, a portion of the eastern part of province of Cangallo bordering on the department of Apurimac was split off to create the new province of Vilcashuamán.

4 provincial capital, and is instead bound more closely with the province of Huamanga to the north.22 The eastern zone includes the area surrounding the provincial capital (also named Cangallo), and is bisected by road networks going west along the Rio Pampas valley toward the D. Chuschi and east to the province of Vilcashuamán, while another road runs north-south to the provinces of Huamanga and Victor Fajardo. Cangallo has traditionally been composed of marginal peasant communities engaged primarily in subsistence or semi-subsistence farming and high-altitude livestock production. There were few large haciendas (large estates), or fundos (small estates) in Cangallo.23

Town Centre of Community of Chuschi

Town Centre of Community of Quispillaccta

Figure 3: District of Chuschi

Source: Perú, Departamento de Ayacucho, Plan Para el Uso Sostenible de los Recursos Naturales Altoandinos Y Prevención de Los Efectos Climatológicos Adversos del Distrito de Chuschi del Departamento de Ayacucho, N.P., 2006: 105.

22 CVR, Informe, Volume 4, 39-40. 23 CVR, Informe, Volume 4, 40 & 18.

The D. Chuschi lies roughly in the centre of province in an area dominated by high mountain grasslands and deep river valleys. Spread over 49,000H, about 60% of the D. Chuschi’s land area consists of high-altitude pasture lands above 3900m, where agriculture is risky and limited to a few hardy tubers and grains. A range of mountains runs through the centre of the D. Chuschi, with the northern half of the district draining into the Rio Cachi watershed that flows north through the District of Vinchos to the arid flat lands around the capital city of Huamanga24 in northern Ayacucho. The southern half of the D. Chuschi sits on the northern bank of the Pampas River valley, which runs west to east, plunging from high puna grasslands in the department of Huancavelica to the west at heights of over 4500m to about 2000m east of the D. Chuschi. Cultivated agriculture has historically been more common in the southern half of the district, below 3900m, with crops such as maize, tubers, beans, wheat, barley, and fruit trees dominating. According to the most recent census, the district has 8,321 residents, living in several dozen population centres of various sizes.25 There are three recognized peasant communities in the D. Chuschi: Chuschi, Quispillaccta, and Cancha Cancha. Chuschi has been the main population centre in the immediate area since the colonial era, and was designated the district headquarters when the district was created in 1857. Both Chuschi and Quispillaccta have a historical town centre in the southern end of the district, fronting the Pampas River. Each also has various associated annexes and hamlets located in different sectors of the district.

Ayacucho has historically been one of the poorest departments in all of Peru, with some of the worst development indicators in the country. Throughout history, Ayacucho has seen repeated patterns of conquest, exploitation, rebellion, and marginalization. Incan Conquest of the tribes in Ayacucho took place long before the Spanish arrived in Peru, profoundly transforming the communities and cultures living in the area. Spanish invaders entered the area in 1539, founding the capital of Huamanga on the northern plains and dividing the lands and peoples of the area into Spanish-controlled estates. Spanish Conquest set in motion hundreds of years of exploitation, suffering, and dislocation for Ayacuchanos. Diseases brought by the Spanish

24 The capital city of the Department of Ayacucho is sometimes called Ayacucho and sometimes Huamanga. To avoid confusion, I refer to the department capital as Huamanga throughout the dissertation. 25 Instituto Nacional de Estadística e Informática (INEI), Directorio Nacional de Centros Poblados, Tomo 2, Censos Nacionales 2017: XII de Población, VII de Vivienda, y III de Comunidades Indígenas, Lima: INEI, Setiembre 2018, 410.

6 conquerors helped to decimate the residents of the area, dropping populations 50-75%, as outlined below in section 4.2.1. But isolated highland areas like Ayacucho suffered less than other more heavily populated areas like coastal Peru. The Battle of Ayacucho in 1824 led to the defeat Spanish Royalist armies in Peru and marked the beginning of Peru’s independence. Spanish and Mestizo exploitation of Ayacucho continued throughout this period, only partly impacted by Peru’s war with Chile fifty years later, which led to widespread destruction of Ayacucho by invading Chilean forces.

Though constantly fought over because of its relatively strategic position between Lima and Cuzco to the east, Ayacucho remained a relative backwater for most of Peru’s post-Conquest history. The D. Chuschi’s isolated location in mountainous central Cangallo was even further removed from Peru’s economic and political power centres. Compared to the rest of the department, this part of Ayacucho – including the other central provinces of Vilcashuamán, Victor Fajardo, and Huancasancos – has been politically and economically isolated throughout history. Geography has contributed to uneven patterns of development throughout Ayacucho’s history, with the northern provinces establishing stronger commercial ties to the central Sierra and Lima through road connections to the north, while the provinces of southern Ayacucho have traditionally maintained stronger commercial ties with the coast and the large urban centres in southern Peru than with the rest of the department.26 Economic power in Ayacucho has long rested with the northern-most provinces in the department – La Mar, Huanta, and Huamanga (the location of the department capital, Huamanga). In 1959, these three provinces had 205 haciendas in operation, representing over 80% of the total haciendas in the department.27 By contrast, in central Ayacucho, there were only 18 haciendas in the early 1960s, and all were located in the provinces of Cangallo and Vilcashuamán. The lack of good road connections and the relative absence of large amounts of good agricultural land because of the steep geography

26 CVR, Informe, Volume 4, 19. The southern part of Ayacucho is home to the provinces of Lucanas, Parinacochas, Sucre, and Paucar del Sara Sara – the latter two created in 1986 by breaking up the former. Here, the CVR notes, communities and haciendas were primarily engaged in livestock ranching for the coastal zone of Ica, Arequipa (Peru’s 2nd largest city), and Lima. See CVR, Informe, Volume 4, 18-19. 27CVR, Informe, Volume 4, 18.

7 have thus further marginalized isolated districts in central Ayacucho like Chuschi, in this already marginalized part of the Peruvian Sierra.28

Household agricultural production in the D. Chuschi has mostly been for subsistence purposes, though some households always traded crops to fill needs. In the latter half of the twentieth century, some households sold small portions of their cultivated production for cash – mainly tubers and garlic. Most households also raise livestock, selling them or their products as needed for cash. Household economic activity around livestock production increased in the latter twentieth century.

Travelling today to the D. Chuschi from Ayacucho’s capital Huamanga typically means enduring a long, bone-jarring trip on a local combi, as I did in 2004. Combi’s are the small Toyota vans that serve as local transportation throughout the Peruvian Sierra. Outfitted with bench seats in the back, ours had room for twelve, but typically left with about twenty people jammed in every available nook and cranny. Those departing Huamanga – the capital of the department – to return to Quispillaccta and Chuschi were bringing back supplies – sacks of fertilizer, seed, household goods, fresh baking, and even farm animals like ducklings or chickens, sitting in crates on the roof or in the aisles of the van.

The road to Chuschi and Quispillaccta takes 5-6 hours along twisting, bumpy roads, heading south from Huamanga. The city of Huamanga sits in a broad, flat valley, situated in the northern half of the department. The climate is dry, semi-arid. One wonders how anyone can make a living in the near desert-like conditions. But this is September, and the dry season is just ending. Vegetation is sparse, though. Eucalyptus trees are interspersed with low native shrubs where the land is not in cultivation. In the driest areas, prickly pear cactus surrounds the roads and fields providing an annual harvest out of the parched land. Since the planting season is only just starting the land is brown from the dry months of June, July, and August. September’s hesitant rains come sparingly in the area immediately around Huamanga. High mountains loom in every direction – to the east, a ridge of mountains runs north into the province of Huanta; beyond these

28 CVR, Informe, Volume 4, 18-9. The only exception to the relative lack of economic dynamism in central Ayacucho was the eastern part of the central provinces, an area of humid, high-altitude tropical valleys near the border with Apurimac. With greater agricultural versatility, this area proved to be economically more dynamic through the production of coffee, cocoa, fruit, and coca leaf – the precursor for cocaine.

8 eastern mountains the land sharply descends to the Amazonian lowlands. To the south and west, another wall of mountains looms in the distance. This western cordillera of the Andes catches moisture on the western slopes, leaving the eastern slopes and plains around Huamanga semi- arid. The ridges to the south west, in the direction of the province of Cangallo – the location of Quispillaccta and Chuschi – climb to well over 5000m. Water from those peaks, flowing through the D. Chuschi, was tapped in the mid-1980s by the Rio Cachi irrigation project to bring drinking and irrigation water to Huamanga and the surrounding plains, much to the consternation of district residents who had no say in the project and suffer from their own water shortages.

In 1959, the first road opened from Huamanga to Chuschi wound out of the Department capital to the south, through the inter-montane valleys to the broad high plains of Pampa Cangallo, before turning west alongside the Rio Pampas (Pampas River) toward the D. Chuschi. The western districts of Cangallo – Chuschi, Totos, and Paras – lie in a land of thrusting mountain peaks, sliced through like a knife on their southern borders by the deep Rio Pampas river valley. The steep hills offer only narrow slivers of valley bottomland that is heavily cultivated by farmers, forcing the inhabitants to cultivate steep hillsides which, in places, have a slope of more than 20 degrees. The northern parts of these districts are high montane punas - grasslands of varying quality and steepness, with cattle, sheep, and camelids raised on lands sometimes at dizzying heights, up to 5000m.

Back when there was only one road to Chuschi, you entered Chuschi from the southeast, passing by the town square before you crossed the bridge over a small creek into the main square of Quispillaccta. Today, however, that route is less traveled. A newer dirt road has cut more than an hour off the trip, and enters the area from the neighbouring District of Vinchos to the northeast, through the Rio Cachi watershed, passing through some of the upper barrios of Quispillaccta and Chuschi, climbing over the high ridge above the main towns – which reaches more than 4400m at this point –before descending into the valley that divides the towns of Chuschi and Quispillaccta. While the newer route is shorter, even four hours of winding, sometimes jarring roads takes its toll on the passengers in the combi.

Descending into the valley holding the town centres of Chuschi and Quispillaccta, we enter into a broad, horseshoe-shaped valley that runs perpendicular to the Rio Pampas river valley due south of us. The river is invisible almost 2 km below. The road skips back and forth over the

9 current boundaries of the two communities; ownership of the valley is now roughly shared between both. In fact, we pass directly through areas which 40 years ago were hotly contested by the two communities. It was here in the upper reaches of the valley, and in the puna above – the area known as Kimsa Cruz – that, in the first days of May 1960, a huge clash took place between Chuschi and Quispillaccta. For three days, thousands of comuneros (community members) from both communities battled with clubs, slings, rocks, and even a few guns. In the end, dozens on both sides were injured, while three Quispillacctinos were shot dead, and two were seriously injured by bullet wounds. One unlucky Chuschino was later caught and beaten to death. The decades-long conflict between the two communities had again drawn blood.

Today, however, relations between Chuschi and Quispillaccta are peaceful. There is no anxiety as we make our way down into the valley, past the isolated hamlets of Yuraqcruz – a barrio of Quispillaccta, past the small chacras (fields) planted with broad beans, potatoes, wheat, and, a little lower down, as we pass into the quichwa zone – traditionally, the corn-growing area below 3300m – which sits both above and below the town centres of Chuschi and Quispillaccta.29 Spread out on either side of the small brook which divides the valley sit the main population centres for Chuschi and Quispillaccta, straddling the valley at about 2800m.

The town centres of Chuschi (foreground) and Quispillaccta (background behind line of trees. The tower of Chuschi’s colonial-era Catholic church is visible in front of the town square. A small brook divides the town centres, located in a small valley in the south of the district. The brook flows south in the Rio Pampas, to the left. Source: http://chuschi.com/

Situated around the main squares of both towns are a Catholic Church and the offices of the community authorities. Chuschi also houses the offices of the district officials and the District

29 Billie Jean Isbell, To Defend Ourselves: Ecology and Ritual in an Andean Village, (Austin, TX: University of Texas Press, 1978): 51 & 55.

Alcalde (mayor). The majority of the small stores in the two communities are scattered close to Chuschi’s main plaza, as is the district medical post.

The two population centres are situated like many Andean communities in the Sierra, anchored in quichwa zone to allow easier access to both the sallqa or puna zone above 4000m where alpacas, llamas, sheep, and cattle have traditionally been kept, and the mayapatan, or river bottom, around 2300m – the area specializing in corn and fruit production like apples, peaches, oranges, and prickly pear. Peasant household economies in the Andes have traditionally employed a variety of altitude zones for agricultural production – from small parcels on the bottoms of valleys for maize production, right up the side of hills to the high-altitude puna where they raise camelids and sheep. This practice is known as Andean verticality.30 Locating the main population centres in the quichwa zone facilitated verticality, and puts the community within the culturally important maize production zone. But their current location is not entirely due to rational agricultural decision-making, as is discussed below in chapter 4.1 in more detail.

Land in all the communities in the district is owned collectively by the community, but has been parceled out to community members on a usufruct basis. These lands can be passed down to family members through inheritance or sold to other community members; however, land cannot be sold to anyone outside of the community. The community can seize abandoned land for redistribution, or it may decide to seize land for community development projects like road construction.

In recent decades, social, economic, environmental, and demographic change, have somewhat altered the traditional settlement pattern in Chuschi and Quispillaccta. While most of the population of Chuschi continues to reside in the Matriz (main town), much of Quispillaccta’s population has disbursed to the other 11 barrios (hamlets/neighbourhoods) spread all over the community territory, particularly those hamlets located in the northern Rio Cachi watershed. Larger barrios in Quispillaccta like Unión Portrero, Puncupata, and Catalinayuq today look like

30 See Enrique Mayer, The Articulated Peasant: Household Economies in the Andes (Boulder, CO: Westview Press, 2002): Chapter 8; John V. Murra, “ ‘El Archipielago Vertical’ Revisited,” Chapter 1 in Shozo Masuda, Izumi Shimada, and Craig Morris eds., Andean Ecology and Civilization: An Interdisciplinary Perspective on Andean Ecological Complementarity (n.p.: University of Tokyo Press, 1985): 3-13 and Murra, “The Limits and Limitations of the “Vertical Archipelago” in the Andes,” Chapter 2 in Andean Ecology and Civilization, 15-20; and Brush, “Diversity and Change in Andean Agriculture,” 273-74.

11 small towns, with stores, schools, and sports fields. Some families in Quispillaccta maintain a family house in the village centre, but only use it when they have business in the area. This decades-long process of intra-communal migration is one of the key parts of the story below in chapters 5 and 6.

Campesinos in this area have always been on the move. Like most Andean peasants, their land is divided into small parcels spread all over the community territory. Chuschi’s settlement pattern used to be the norm in Quispillaccta – campesinos lived in the main town most of the time and traveled between their various plots to tend to their crops or to their corrals in the puna to graze their animals on the community pasture lands in the high puna. While they may stay for a day or two in distant parts of the community while tending to those distant parcels or animals, their main place of residence was in town. Seasonal changes also governed these movements, with fields being sown when rains arrived in September to November, and being harvested from March to May. Livestock were traditionally moved around to different pastures as necessary, brought to lower cultivated fields to feed on crop residues when harvests were completed. Increasingly, however, the upper puna pastures are being divided and parceled into individual walled holdings. Dry season might bring travel to distant towns or agricultural zones for wage work, or a greater focus on simple commodity production in and around the district. Change is a constant the lives of local comuneros, even if at first glance their lives appear traditional and certain.

Until the mid-1970s, the political structure of the D. Chuschi was a dual system, made up of indigenous peasants and their traditional authority structures on one side, and the mestizo- dominated bureaucratic governmental system on the other.31 Chuschi is not just a peasant community, it is also a municipal district of the province of Cangallo, and thus part of the administrative structure of the Peruvian government – district, province, and department – with its own associated district institutions.32 The municipal (district) mayor is the top bureaucratic authority in the district. Working with a small committee of assistants, the mayor is responsible

31 Isbell, To Defend Ourselves, 83. 32 Brush, Mountain, Field, and Family, 59.

12 for village affairs, budget and record duties, village improvements, and keeping order.33 The mayor’s authority is greater than the other district authorities, the governor and lieutenant governor. While the district mayor is generally elected, the positions of governor and lieutenant governor are appointed by provincial officials. The governor and lieutenant governor are primarily responsible for law and order, tax collection in the district, and maintaining communication with the prefect in the provincial capital, the town of Cangallo.34 Bureaucratic hierarchies in the district also traditionally included officials of the Catholic Church, including the priest and various officials tasked to assist the priest, most of whom were assigned from the traditional indigenous authority structure described below.35 To peasants, the bureaucratic authority structure in the D. Chuschi represented foreign domination in the midst of their community’s social structure.36

Peasant communities also have their own bureaucratic leadership structure. Before Agrarian Reform in 1969, the village bureaucracy was led by the junta comunal, who were the guardians of communal lands. The junta comunal was led by an elected president and his deputy, the personero, who is “the legal representative of the village in all land disputes,” and assisted by a six-man committee who were elected with the personero for four years.37 With Agrarian Reform, the junta comunal was abolished and an administrative council (Consejo de Administración) and a vigilance council (Consejo de Vigilancia) were created instead, each led by the elected president.38 In 1987, the Peruvian General Law on Peasant Communities was passed which formalized the Directiva Comunal as the main body of village bureaucracy. Consisting of a president, vice-president, and a minimum of four directors, the directiva comunal is elected for two years by a vote in the community assembly.39 The president of the directiva comunal is the key village leader in the communities of Chuschi, Quispillaccta, Cancha Cancha,

33 Isbell, To Defend Ourselves, 89. 34 Isbell, To Defend Ourselves, 88-89; Brush, Mountain, Field, and Family, 60. 35 Isbell, To Defend Ourselves, 88-89. 36 Isbell, To Defend Ourselves, 89. 37 Isbell, To Defend Ourselves, 84. 38 González, Unveiling Secrets of War, 119. 39 Ley General De Comunidades Campesinas, Peru, No. 24656, 14 April, 1987.

13 and other recognized communities in the district. They represent the interests of community members and lead community administrative and development matters.

Each peasant community also has its own traditional indigenous authority structures that operates in parallel to the state’s bureaucratic structures, serving both to buttress the roles and activities of the bureaucratic authorities, but also to maintain indigenous social structures and cohesion in communities in the district. According to Isbell, the indigenous civil- religious prestige authority system – or varayoq system - in Chuschi has served to reinforce “the structural principles of the comuneros’ conceptualization of space and ecology,” maintain mechanisms of reciprocal aid and community service, and thereby ensure the cohesion and survival of the village.40 Functionally, the traditional authority structures Varayoq of Quispillaccta with their staffs of authority, 1969. played crucial roles in the religious cargo Billie Jean Isbell Andean Collection, Cornell University Library system associated with Catholic fiestas, and worked closely with bureaucratic authority structures to maintain moral and civil order in the district. Some traditional authority members also worked directly with the district governor and lieutenant governor to patrol the village and keep the peace, while others were tasked with duties to secure agricultural fields and livestock herds, or serve as functionaries for various community and district tasks.41 While most Peruvian peasant communities had such traditional authority structures, they have been on the decline over the past fifty years. By the time of Agrarian Reform, the varayoq system had already disappeared in many communities in Peru or was

40 Isbell, To Defend Ourselves, 84-5. See also Mayer, The Articulated Peasant, 125-27 41 Isbell, To Defend Ourselves, 88-94.

14 greatly diminished, with positions left unfilled or abolished.42 Communities in the D. of Chuschi had largely preserved the varayoq system when anthropologists first entered the area in the late 1960s.43 While some positions have been abolished over the decades, they continue struggle to maintain their indigenous authority system up to the present day, as described in the Postscript.

The dual nature of the political structures in the D. of Chuschi stem from a similar dual social system in the district. Identity in Peru is complex and partly socially constructed, and this is also reflected in identity classifications of social structures in the D. Chuschi. According to Isbell, mestizos (vecinos) are outsiders or are emulating outsiders, in contrast to local indigenous peasants. Community members (comuneros) in Chuschi, ascribed identity – whether someone was a comunero or a vecino - based on a person’s participation in communal affairs: “the comuneros, or communal members of the village, who participate in the [traditional indigenous authority] hierarchy, wear traditional dress, and speak Quechua; versus the vecinos, or qalas (literally, peeled or naked ones), who are Spanish speaking, western dressed, foreign nonparticipants in communal life.”44 Until the early 1970s, mestizo vecinos or qalas were usually the shop keepers, district bureaucratic officials, priests, school teachers, and their relatives. Their greater facility with the Spanish language meant that they tended to dominate the bureaucratic positions in the district until the late 1960s and early 1970s. Comuneros who migrate to urban areas, adopt vecinos lifestyles and refuse to participate in communal life could also lose their comuneros status, in the eyes of community members at that time, according to Isbell.45 Social identity became increasingly fluid in the latter 20th century, increasingly altering by the 1960s and 1970s social distinctions in many highland communities described by scholars earlier in the twentieth century. A thorough historical account of the origins and activities of mestizo, misti, qala, and gamonales in the D. Chuschi remains to be written. However, what is clear from recent historical work is that up to the early 1970s they were a frequent source of

42 Mitchell notes that the varayoq system had been abolished in Quinoa, Ayacucho in the late 1960s, while Mayer reports that it was severely handicapped with many positions left unfilled in the late 1960s in the village of Tangor, Department of Pasco, in central Peru. Mitchell, Peasants on the Edge, 163-4; Mayer, The Articulated Peasant, 106. 43 Isbell, To Defend Ourselves, 84-97. 44 Isbell, To Defend Ourselves, 67.

45 Isbell, To Defend Ourselves, 70-3.

15 abuse and controversy for indigenous community members throughout the district, with the impacts being felt particularly strongly on Chuschi and its associated annexes.46

Key findings and plan of dissertation

The findings of this dissertation speak less to the role of human pressure on the environment in causing Sendero’s rural rebellion and instead show how human pressure on the environment helped to cause particular patterns of rural violence in the D. Chuschi. Long-term patterns of resource capture and human-induced environmental and demographic change, combined with prevailing local, regional, and national political-economic currents in Peru over the 20th century (along with some international influences) helped to increase inter-community competition for valuable but scarce agricultural resources – over range land, crop land, and water resources. In particular, this dissertation is a micro-study of how these changes conditioned patterns of violence between two neighbouring communities in the district – the communities of Chuschi and Quispillaccta. Household livelihood changes pushed conflictual relationships both vertically within Chuschi – in conflicts between elites and peasants – but also horizontally, between the peasant communities of Chuschi and Quispillaccta, over rights to contested resources essential to increasingly pressured livelihoods in both communities. Change and contest resulted in winners and losers, with implications for social stability many years into the future.

Quispillaccta emerged as the victor in these local community battles. However, their victory fostered enduring resentment in Chuschi – the traditional administrative centre of the district – and among Chuschi’s associated annexes. These grievances spilled out during the Sendero years as Quispillaccta was branded as a centre of Sendero militancy, partly as a result of accusations from those who lost out in the inter-community land competition. Quispillaccta suffered heavily as a result in the Peruvian military’s counter-insurgency bloodletting, disproportionately more than Chuschi and its annexes. Quispillaccta faced retribution by counter-insurgency because of their land conflict successes.

In this sense, this study’s findings cannot link human-induced environmental change directly to Sendero’s growth in the area; however, human pressure on the local environment and people’s

46 La Serna, The Corner of the Living, chapter 3; and chapters 4 and 5 below.

16 adaptive choices to these pressures did help to condition the patterns of violence in the decades before the Sendero insurgency, and patterns of violence once the insurgency heated up in the area. This thesis is but one strand of a complex local story behind the violence in the district; we are still in the early stages of attempting to piece together how and why violent conflict broke out in the D. Chuschi during Sendero’s insurgency. Certainly, other parts of the story remain to be explored by scholars. But we cannot understand why violence took the shape that it did in this area during the insurgency, and over the previous century in various small-scale confrontations, unless we examine the interface of human-induced environmental change and demographic change, in the context of wider socio-economic developments in the region.

The story in Peru is complicated, this dissertation concludes, because these local competitive dynamics were also happening concurrently with wider regional patterns in Peru, where livelihoods affected by environmental and demographic change in Cangallo and other areas of the South-Central Andes were probably also contributing to dynamics that helped spur the Sendero insurgency. Human-induced environmental and demographic change probably had multi-level influences, with considerable local level variability. However, the field work in this dissertation found only indirect evidence in the D. Chuschi for environmental change and demographic change contributing to the rise of Sendero’s insurgency.

This dissertation adds to the prevailing understanding of how environmental change can contribute to violent conflict by deepening the sophistication of linkages proposed by research to date. It is rooted in detailed field work, and grounded upon a foundation of the literature on rural rebellions, civil war, agro-ecological livelihoods in the Andes, and ethnographic literature on indigenous Peruvian communities. The dissertation also critically evaluates past qualitative research on environmental change-conflict linkages and finds important deficiencies in the level of analysis used by much of the qualitative research to date. The dissertation proposes a new livelihood framework for local-scale qualitative environmental-conflict research. This framework forms the foundation for the study of human-induced pressure on the local environment in the D. Chuschi.

Research on household adaptation to environmental scarcities and other livelihood pressures in the D. Chuschi leads us to revise thinking about the most appropriate level of analysis of environment-conflict research and about prevailing hypotheses about linear processes from

17 environmental scarcities to violent conflict outcomes. Much of the research to date on environment-conflict linkages has chosen to use a state-level of analysis when attempting to determine if environmental scarcities can help cause violent conflict. This dissertation demonstrates that such an approach is misguided, because scarcities impact households and groups at the local level, and their local responses to these pressures are often lost if the focus is on the state level. Situating the household livelihood model at the centre of the causal analysis of the impacts of environmental scarcities allows for a more complex understanding of the impact of scarcities in light of other factors noted by scholars, like political-economic processes, and an appreciation of how household capacities mediate social impacts. Environment-conflict research that posits linear processes from environmental scarcities to social effects and then to violent conflict is simplistic because households adapt and cope to stresses like environmental stresses or economic transformation. Certainly, Homer-Dixon’s arguments that environmental scarcities can constrain agricultural and economic productivity are correct,47 but there are likely few instances of a direct impact from environmental scarcities to social effects as he suggests. The impacts of scarcities defy simple, linear predictions of social effects. Complex adaptive and coping responses of households modelled in this dissertation point to many possible outcomes, and instead suggest that we need to pay attention to changes in household vulnerability and risk and to constraints on adaptation and coping, if we want to model likely social effects of pressures like environmental scarcities and their possible violent conflict implications.

The dissertation is organized across nine chapters. The literature review in Chapter 2 situates the study’s original research question in qualitative research on environmental change and conflict, with the theoretical starting point located in classic research on rural rebellions/revolts - a subset of the larger research field of revolution studies, and in recent research on the causes of civil war. A review of qualitative research on environment-conflict linkages suggests that the impact of human-induced environmental change may result in particular material social effects that make certain types of violent conflict more likely. This work echoes earlier work done by scholars of rural rebellions and revolutions about the role played by grievances (or motivations) and opportunities for rebellion. These arguments have also ignited debates among qualitative researchers about whether political-economic capture of important resources is instead largely

47 Homer-Dixon, Environment, 80-85.

18 responsible for the negative social effects of environmental change on rural residents. Quantitative research on environment-conflict linkages has failed to settle the question of the role or importance of human pressure on the natural environment in causing violent sub-state conflict and rebellions. In each of these bodies of research, debates remain over the impact and importance of material changes to the well-being of residents who participate in violent uprisings. Debates about the importance of such factors also emerge from a review of empirical research on the causes of rural violence in Peru in the latter half of the twentieth century. The chapter concludes by outlining the key areas of investigation for the field work chapters.

Chapter 3 expands critiques of key qualitative research on environmental change-conflict links to make the case for local studies grounded in detailed field work.48 It argues that the level of analysis of most of the qualitative research to date on environmental change-conflict linkages has been done from an inappropriate state-level of analysis. The chapter then develops a detailed livelihood framework for qualitative research on local-level environment-conflict linkages. Continued debates about findings have also sparked questions about the best methodological approaches to study environment-conflict linkages. While many qualitative and quantitative desk studies have been done in the past twenty years, few detailed field studies have been done to explore whether and how human-environmental pressures contribute to violent conflict. This chapter argues that a local, micro case study examining the role of human-induced environmental change and demographic change in the D. Chuschi could be beneficial to understand whether and how such impacts operated before and during the Sendero insurgency.

The livelihood framework developed in this chapter serves as the foundation for the organization of the fieldwork findings from Chuschi and Quispillaccta examined in chapters 4-7. While this project initially aimed to use the livelihood framework as a direct research guide for fieldwork in Chuschi and Quispillaccta, in the field I realized the incredible difficulty of undertaking such detailed household studies. The data requirements are considerable, and it is almost impossible to do such studies backwards – as a historical reconstruction, given the sparse historical livelihood data for the D. Chuschi and highland Peru generally. So, in many ways, this

48 The core of the chapter was published in: Tom Deligiannis, “The Evolution of Environment-Conflict Research: Toward a Livelihood Framework,” Global Environmental Politics 12, no.1 (Feb. 2012): 78-100.

19 framework serves as an aspirational guide for other scholars and a heuristic framework to help organize the fieldwork findings in Chuschi and Quispillaccta, given the available data.

Chapter 4 is the first of four field work chapters which discuss the findings of my research in the D. Chuschi. These chapters map out the scope of conflicts over these community lands and border lands, consolidating field interviews, secondary accounts, and archival research into a comprehensive account of the impacts of land and water tenure conflicts from before Spanish Conquest to the late 20th century. The field work findings are interpreted through the household livelihood lens to map and understand how long-term changes in the agricultural livelihoods of district residents evolved over several hundred years, with a particular focus in the latter half of the twentieth century. The dissertation largely focuses on changes and conflicts over land resources; however, water resources in the district relate to land conflicts in several important ways and are thus examined along with changes in the supply, demand, and distribution of land resources in the field work chapters.

Chapter 4 begins with short snap-shot evaluation of peasant livelihoods in the D. Chuschi in the mid-twentieth century in order to highlight the key household asset categories and livelihood activities of a typical district household. Changes in household livelihoods and changes in specific assets, entitlements, and capabilities over time will be discussed in depth in each fieldwork chapter. However, the field work chapters pay particular attention to long-term changes in the natural asset category and entitlement changes to those assets, while also integrating findings about changes and influences from other asset categories.

Chapter 4’s examination of natural asset resource changes and tenure capture and change in Chuschi and Quispillaccta from pre-Conquest to the 20th century focuses on the capture of cultivated assets in the district – land to grow crops for household survival. Chapter 5 examines elite capture of high-altitude livestock grazing lands in the northern half of the district, and efforts by communities in the district to regain control of these lands in the 20th century. Gradually increasing permanent settlement of disputed lands by members of rival communities in the district – moving from the Colonial-era settlements in the south of the district to the disputed northern lands of the district – was a key process that aggravated group conflict in the district over several decades in the twentieth century. Chapter 5 focuses on efforts by communities to claim lands in the northern border area of the district, while Chapter 6 focuses on

20 clashes to control disputed livestock grazing lands and water resources in the centre of the district. While land clashes on the borders of the district often involved neighbouring elite estates or neighbouring peasant communities, the clash to control the centre of the district was primarily between the community of Chuschi and its allied hamlets and the community of Quispillaccta.

Chapter 6 concludes by examining the historic peace settlement between Chuschi and Quispillaccta, concluded in late 1982, just as Sendero’s insurgency in the area was growing in strength. The implications of the settlement conclude chapter 6, both in the context of debates among scholars about the degree of peasant support for the insurgency and in the context of nature of violence in the district during the first years of the insurgency.

Chapter 7 complements the analysis in chapters 4-6 by outlining demand and supply changes in livelihood resources that increasingly pressured households in the 20th century, because of diminishing asset holdings and increasing challenges in agricultural yields. These changes helped to push many district residents to settle in disputed lands in the north-central zone of the district. The chapter first uses archival population data from church records and government archives to chart population change in the D. Chuschi. These figures are contextualized with data showing provincial, department, and national population trends from Conquest to the late 20th century. The role of population factors has been highly controversial in environment- conflict research to date.49 So, this chapter carefully reconstructs and assesses demographic trends in the District. The data shows that population change in the district began to take off in the early to mid-twentieth century, at the same time as the general population of Peru returned to pre-Conquest levels around the middle of the 20th century. The D. Chuschi thus slightly lagged behind national level population trends. The chapter links local population growth to increasing pressure among households for arable cultivated and pasture lands in the mid-20th century in the district, using household interviews to reconstruct average land asset holdings among community members in Chuschi and Quispillaccta. Interviews also chart the extent of temporary or semi-

49 See Betsy Hartmann, "Will the Circle be Unbroken? A Critique of the Project on Environment, Population, and Security," in Violent Environments, ed. Nancy Peluso and Michael Watts (Ithaca, NY: Cornell University Press, 2001), 32-69.

21 permanent out-migration from the district, trends that became increasingly common in the second half of the 20th century. Like internal district migration, out-migration from the D. Chuschi was a livelihood diversification choice for many households, helping to ease increasing pressures on agricultural production and satisfying increasing consumer and educational desires.

Finally, the chapter assesses how demand pressures on local cultivated resources – mostly located around the traditional Colonial-era settlements in the south of the district – stressed yields and increased the supply scarcity of agricultural production. This is challenging to assess because of the tremendous variation in local production conditions; however, data obtained from interviews and supplemented by archival survey data from the late 1960s appears to show yield declines in key household crops in this area in the three decades before Sendero’s uprising. The introduction of green revolution agricultural technology may have helped increase yields for some in the 1960s; however, local informants indicate that these gains did not last.

Combined demand and supply scarcity changes described in chapter 7 increased pressure on the livelihood activities of community members in the traditional, colonial-era settlement areas of the district at the same time as the communities in the district were struggling to regain control or consolidate control of contested lands in the northern and central zones of the district. These combined pressures on livelihood resources increased the stakes for inter-community competition over control of contested cultivated and pastures lands in the area. In the D. Chuschi, changes in the availability of household assets thus operated in the context of changing national and regional structural conditions to lead small-holder farmers in Quispillaccta and Chuschi to adjust household livelihood strategies in ways that contributed to inter-community conflict. One of the most important household assets is the natural asset category, comprising land and water necessary for growing crops and raising livestock. For farmers whose livelihoods largely depend upon what they can produce, the availability or scarcity of land has been crucially important in moderating household livelihood strategies. Until well into the twentieth century, there were few livelihood alternatives for residents in Chuschi and Quispillaccta other than small-holder farming and herding. Patterns of change over time in land assets and the creation of new and competing entitlements for assets in the north and centre of the district thus provide insights into crucial mechanisms of livelihood change over time in the two communities, and the social conflicts these changes helped to spawn.

When conflicts were settled in the late 1970s and early 1980s over the Rio Cachi watershed lands in the district, decades of efforts to regain control of seized lands in the Rio Cachi basin came to an end. Though every community in the district sought to expand their control of Sunni and Puna lands in this zone, Quispillaccta was the clear winner of these struggles, significantly expanding its presence in the upper Rio Cachi basin from the late 1800s to the 1980s. However, there were clearly winners and losers in the long fight to regain high-altitude grazing lands. Communities members in Chuschi, Cancha Cancha, Putaje, and Ccochapampa lost out to Quispillaccta in the land struggles. When their efforts to gain control of these lands were abandoned or settled in the late 1970s and early 1980s, most of their claims remained unfulfilled and their aspirations unrealized. Sendero’s insurgency and the Peruvian military’s counter- insurgency in the district in the early 1980s appeared to undermine the collective will to continue the land struggles, leaving the enmity with Quispillaccta among some over their decades-long efforts to gain control of the disputed lands unresolved. Similarly, there were also losers on Quispillaccta’s side who may have been angered by settlements reached with Chuschi and its associated hamlets.

Chapter 8 examines how centuries of competition and sometimes violent conflict between Quispillaccta and its district neighbours came to a head in the early 1980s. As Sendero Luminoso began to openly operate in the district in 1981-81, killing local thieves and deviants and driving out or eliminating local government officials, land invasions and intercommunity clashes in the district worsened. Community leaders, tired of decades of competition and expensive land litigation, recognized that the time was ripe for a new approach to finally settle the district’s disputed land boundaries. In a few months during late 1982, they conclusively settled disputed borders and finalized an agreement that settled the long-running land conflict between Chuschi and Quispillaccta. No sooner was the agreement finalized, however, when the Peruvian government’s aggressive counter-insurgency operations began in the district to find and destroy Sendero militants. These events and a series of deadly operations in May and June 1983 by the Peruvian Army are outlined in Chapter 8. For the first time, the bloody events of 1983 are reconstructed in a coherent narrative, illustrating that the differential violence perpetrated against Quispillaccta was likely directed toward the community in retribution for their earlier land competition successes.

The implications of the D. Chuschi case study findings for research on the causes of rural violence and research on the linkages between human-induced environmental change and violent conflict are outlined in Chapter 9, the concluding chapter of the dissertation. The dissertation’s local, field-based approach to examining environmental change-conflict linkages has satisfied many of the suggestions of scholars like Tobias Ide, who have promoted field-based qualitative research as a way of advancing our understanding of debates about environmental change-violent conflict linkages, including debates about climate change and violent conflict.50

One key finding of Chapter 9 is that the resource disputes and their outcomes conditioned important patterns of violence in the D. Chuschi during the first years of Sendero’s uprising, and particularly during the Peruvian military’s counter-insurgency campaign. Quispillaccta, largely the victor of the inter-community conflict over land and water resources, suffered disproportionately in the military counter-insurgency campaign after being characterized as a hotbed of Sendero support. However, given their success in competing with their neighbours over land disputes, the logic of Quispillaccta being the key centre for Sendero support in the district – a common conclusion among scholars and Peruvian experts – is questionable. As a community, they had the most to lose by throwing their support behind Sendero, following their success in consolidating land in the district. While some Quispillacctinos certainly supported Sendero, the same was true for every community in the area. The most logical explanation for Quispillaccta being labelled as a centre for Sendero support and suffering disproportionately in the government’s counter-insurgency war is that grievances stemming from their success in the intercommunity resource competitions led some in neighbouring communities to claim that Quispillaccta was a Sendero stronghold in order to seek revenge for their land conflict successes – score-settling by counter-insurgency.

This conclusion does not speak directly to the question of whether scarcities in livelihood resources directly led to or contributed to the Sendero uprising, which was the original question guiding this thesis. The thesis does finds that scarcities aggravated group conflict between communities in the district in the decades leading up to Sendero’s insurgency. The historical

50 Tobias Ide, “Research methods for exploring the links between climate change and conflict,” WIREs Climate Change 8, no.3 (May/June 2017): 1-14. See also the sources in note 13.

24 account of land conflict in the D. Chuschi affirms Homer-Dixon’s conclusion that “environmental scarcities can aggravate divisions or segmentation among ethnic, religious, and linguistic groups.”51 Clearly, as grazing and cultivated land became scarcer in the district and competition for these resources with neighbouring communities became more pronounced, enmity between communities increased. Violent clashes and the casualties from these conflicts only increased the fear, mistrust, and bitterness toward neighbouring communities. These impacts did help cause households in the district to “turn inward and to focus on narrow survival strategies” for their communities, as Homer-Dixon predicted.52 Hardening attitudes toward each other resulted in fewer interactions between members of different communities in the 1960s and 1970s.

More importantly, however, this dissertation argues that we cannot understand how patterns of violence developed during the insurgency unless local changes in livelihood assets are assessed over the long term. The violence perpetrated during the insurgency and particularly after 1982, when the Peruvian government began an aggressive counter-insurgency campaign in the area, was many times more deadly to district residents than group conflicts over land resources in the decades before Sendero’s insurgency. This violence disproportionately impacted Quispillaccta. Conflicts over environmental resources thus impacted the process and direction of violence in the civil war in the district once it began, with tragic consequences, particularly for Quispillacctinos. So, environmental scarcities were crucially important for shaping patterns of violence during the civil war, in addition to influencing processes that led to violent conflict onset. This finding is new in environment-conflict research and confirms work by Stathis Kalyvas on the sources and nature of violence in civil war, where people caught up in a war take advantage of the situation and denounce rivals “to settle private and local conflicts whose relation to the grand causes of the war or the goals of the belligerents is often tenuous.”53 Denunciations to the Peruvian military in May 1983 by some in Chuschi and associated hamlets, stemming from decades of heated local land conflict, unleashed indiscriminate counter- insurgency violence against Quispillacctinos. From the military’s perspective, this violence was

51 Thomas F. Homer-Dixon, Environment, scarcity, and violence (Princeton University Press, 1999)., 96. 52 Homer-Dixon, Environment, 96. 53 Kalyvas, Logic, 364.

25 necessary in order to turn them against Sendero and toward collaboration with Peru’s security services, precisely as Kalyvas’ theory of indiscriminate and discriminate violence suggests.54

A second key finding of this dissertation speaks to debates among environment-conflict scholars about the causal significance of environmental scarcities and whether they can act as violent conflict triggers.55 Environmental scarcities in the D. Chuschi did not act to trigger violent conflict; instead, various political and policy changes in Peru in the 19th and 20th century triggered local violent conflicts in the district, particularly in the three decades leading up to Sendero’s insurgency. Policy changes on land reform correlated to peasant activism and conflict over land resources going back to the early 19th century, decades before scarcities began to acutely impact household livelihoods in the district. The causal significance of environmental scarcities is instead as a deep structural factor, altering causal pathways and impacting livelihood constraints and opportunities available to households. While some might characterize environmental scarcities as root causes, this would be an overly narrow description of their causal impact because scarcities also exhibit dynamic pressures56, depending on the degree of scarcities and the adaptation and coping strategies employed by households. The structural impacts of environmental scarcities are not “static and invariate over time,”57 as some scholars suggest, but instead change and transform in relation to the degree of human-pressure on the resources. Their causal structural role in the D. Chuschi case is analogous to tectonic plates, as Homer-Dixon has suggested.58 Their precise causal role is likely contextually specific and difficult to generalize, however, and it is conceivable that environmental scarcities could act to trigger conflicts in other contexts.

54 Kalyvas, Logic, 150. 55 Ide, “Why do conflicts over scarce renewable resources turn violent?” 62-3. 56 See Ben Wisner, Piers Blaikie; Terry Cannon & Ian Davis, At Risk. Natural Hazards, People’s Vulnerability and Disasters, 2nd Edition (London: Routledge, 2005), 48. 57 Ide, “Why do conflicts over scarce renewable resources turn violent?” 62. 58 Homer-Dixon, Environment, 18.

Timeline of key events

1532 – Spanish conquest of Peru begins

1567 – Earliest recorded evidence of dispute between Quispillaccta and Chuschi, refers to disputes Canas Indians and the Aymaraes.

1824 – Spanish Royalist forces defeated at Battle of Ayacucho, securing independence of Peru and South America.

1828 – Limited agrarian reform law allows indigenous communities to obtain property titles for lands they occupied and opened up opportunity to buy land on private market.

1840 – Quispillacta begins process to buy Hacienda Santa Catalina, in the area of Catalinayocc in Rio Cachi zone to hold communal herd. By 1890 the Catholic Church took control of this herd.

1857 – District of Chuschi created.

1879-1883 War of Pacific between Chile and Peru leads to destruction and abandonment of some Rio Cachi haciendas.

1920-23 – Wave of peasant rebellions and resistance in highlands against abusive elites in wake of ups and downs of wool market following WW1.

Early 1920s – Chuschi-Hacienda Yaruca litigation begins over lands seized in NW of district.

16 January, 1923 – First record of Quispillacta-Manuel Ruiz dispute (owner of the hacienda Putacca) in NE of district.

27 May, 1941 - Chuschi officially recognized as a peasant community.

1941 – Quispillaccta-Chuschi legal clash over lands in centre and north-east corner of district.

1942 – Quicamachay estate in north-east zone of district bought by Quispillacta from hacienda owners. These lands source of dispute with Putacca hacendados and Cancha Cancha comuneros.

29 November, 1944 - Quispillacta recognized as a peasant community.

May, 1960 – Massive conflict between Quispillacta and Chuschi in Kimsa-Cruz area leaves 3 Quispillacctinos dead and dozens injured on both sides.

March 1962 Quispillacctinos kill 1 Chuschino in revenge for 1960 deaths.

1968 – Presidential coup by General Juan Velasco Alvarado and reform minded military officers seizes power in Peru.

1969 – Agrarian Reform law instituted, leading to elimination of private haciendas in Peru.

1972 – Chuschi gets control of all church lands and animals.

Aug. 1975 – Pres. Gen. Velasco deposed by Gen. Francisco Morales Bermúdez in conservative counter- coup in Peru.

1979 – Land conflict between Quispillaccta and Chuschi in centre and NE of district heats up.

May 17, 1980 – Senderistas enter town hall in Chuschi and burn ballots being stored for national elections.

May 18, 1980 – Military rules ends as democratic elections return Fernando Belaúnde Terry as President of Peru.

September 1980 – Sendero representatives return to D. Chuschi and demand a meeting with the school teachers; deepening of their conscious raising and moralizing attempts by Sendero Luminoso.

March-early Oct. 1981 – repeated clashes between Chuschi and Quispillaccta over lands in centre and NE of district.

16 October, 1981 – Act of Agreement finalized between authorities of Quispillacta and Chuschi, to meet to solve the problem of borders. Efforts to fix the borders begin from this date until 17th Feb., 1982 when the borders were finally settled.

October 1981 – SL close the district government and abolish district political posts.

November 1981 – SL militants return to district without hoods or masks and held a meeting with all the schoolteachers and formed the Laborers and Workers Class Movement (MOTC) and a Popular Committee.

November 1981 – all local officials resigned their posts in a general assembly and signed a document to this effect.

17th Feb., 1982 – settlement between Chuschi and Quispillacta completed. Clause 13 says that, in the event of a breach by one of the parties, they will have to pay 20,000,000 soles.

July 1, 1982 – Masked Sendero militants blow up the post office in Cancha Cancha and captured the governor of Chuschi, Bernadino Chipana, wanting to kill him. But he is spared after the community rejected their call to sanction his killing.

August 2nd, 1982 – Allpachaca research station NE of district ransacked and fields communally planted.

31 December, 1982 – Declaration of Emergency situation in Ayacucho and the installation of the Military Political Command. Peruvian military takes over counter-insurgency against Sendero.

December 20, 1982 – combined forces from the army, PIP, and Civil Guard arrive by truck in Chuschi, beginning security forces’ counter-insurgency campaign in the area.

Feb. 1983 – Chuschi flies the white flag from municipal building and asks that a military post be established. A Civil Guard post was set up afterward.

April 1983 - Peruvian army base set up in D. Totos, west of D. Chuschi.

April, May, June 1983 – Massacres by Sendero and Peru’s security forces kill dozens in district.

Chapter 2 Literature Review and Research Design: Do environmental scarcities impact violent conflict onset? Introduction

Over the past 50 years, Peru’s Southern and Central Sierra has been the scene of repeated protests, small-scale conflicts, and insurgencies. Land invasions, protests, and strikes by rural peasants began in the late 1950s and spread through the area in the early 1960s. Many were led or inspired by the growing peasant unions, peasant syndicates, and peasant federations that sprang up in the area around the same time. These actions continued into the 1970s, even after the Peruvian government undertook widespread land reform in the area. Violent confrontations sometimes accompanied peasant actions: violence between peasants and landlords; violence between peasants and hacienda workers; and even violent confrontations between peasant communities. Hundreds of peasants were also killed and injured as a result of the repression that often followed as local elites, the Peruvian police, or the Peruvian military attempted to suppress peasant militancy.

Although violent conflicts in Peru’s Sierra go back centuries, the peasant activity in highland Peru during the 1960s constituted the largest peasant mobilization in recent Latin American history.59 Capitalizing on the apparent signs of peasant activism in the highlands and inspired by the example of Cuba’s recent revolution, guerrilla insurgencies sprang up in several highland departments in the 1960s following the increased peasant unrest. Quickly and decisively crushed by the Peruvian military, the insurgencies proved to be a pale challenge to Peruvian government. A far greater danger emerged from the southern Sierra in the late 1970s and early 1980s, however, when the Sendero Luminoso initiated a widespread violent insurgency to overthrow the Peruvian state – a challenge that would eventually lead to the deaths of tens of thousands of Peruvians, cost the country billions in damages, and take more than a decade to bring under control.60

59 R. F. Watters, Poverty and Peasantry in Peru’s Southern Andes, 1963-90 (Pittsburg, PA: University of Pittsburg Press, 1994): 245. 6060 Comisión de la Verdad y Reconciliación, Informe final: Perú, 1980-2000 (Lima: Comisión de la Verdad y Reconciliación, Universidad Nacional Mayor de San Marcos, Pontificia Universidad Católica del Perú, 2004),

29 30

Although peasant support for Sendero ultimately proved to be locally contingent and short-lived in many parts of the Sierra, the insurgents were initially able to tap into a significant reservoir of on-going peasant dissatisfaction on issues such as land reform, elite and market exploitation, and state neglect. These were, in many cases, the same issues that had been stirring unrest in the highlands since at least the late 1950s. In fact, one can argue that the unrest in Peru’s highlands in the 1960s and the early 1970s, and Sendero’s insurgency in 1980 represented an historically contiguous period of varying rural protest, conflict, insurgency and mobilization in the area. When viewed as a quarter century of efforts to agitate for change, these developments raise the question of origins and underlying conditions that led to rural unrest and rebellion in Peru between the late 1950s and the early 1980s?

In the context of more than twenty-five years of research on the links between environmental and demographic change and violent conflict, this project examines whether environmental stress and demographic change contributed to the widespread rural unrest, peasant mobilization, and insurgency in Peru in the latter half of the twentieth century. Research on rural unrest in Peru has not adequately explored how environmental stress and demographic change have contributed to these events.61 Greater scholarly attention about how environmental scarcity interacted with other factors responsible for rural conflict and peasant mobilization in the Southern Sierra in the latter half of the twentieth century can enrich our understanding of why those events took place. A focus on the impact of environmental scarcity in the Southern Sierra can offer important revisions to the dominant explanations offered by scholars about rural change in the Peruvian highlands. This study also speaks to wider debates among scholars about the role of environmental and demographic change in causing social conflict, debates that have regained renewed urgency in light of the accelerating impacts of climate change.

The impact of environmental stress and demographic factors on rural change in Peru during the latter half of the twentieth century has not received thorough treatment to date. While the structural causes of poverty and inequality have long been studied by scholars, this study

General Conclusions, 316. A general history of this era can be found in: Deborah Poole and Gerardo Rénique, Peru: Time of Fear (London: Latin American Bureau, 1992). 61 In this dissertation I use Homer-Dixon’s term ‘environmental scarcity’ interchangeably with environmental stress and demographic change. Environmental scarcity is defined below in section 2.3.

31 integrates those insights with the impacts of environmental and demographic stress to investigate its possible impact on rural unrest, popular mobilization, and insurgency during decades of transformation in the southern highlands of Peru through a micro-study of long-term change in the D. Chuschi, Ayacucho.62 The impacts of supply, demand, and structural changes in household assets over the twentieth century can offer unique insights on the question of the causes of rural rebellion and conflict in Peru.

This chapter examines the role of environmental stress and population growth in the rural transformation of highland Peru and its possible linkages to violent rebellion and conflict in the context of three broad bodies of research. At a theoretical level, we first need to have an understanding of the debates about the causes of rebellions and rural revolts – that subset of social science research on civil wars that focuses on understanding the causes of revolutions – and the role, if any, of environmental scarcities in those theories.63 Research on civil wars is vast, so scholars categorize different types of civil wars, particularly with the growth of statistical studies of civil war in the latter twentieth century. Some argue that we can distinguish between revolutionary and secessionist civil wars, and between ethnic versus ideological conflicts, both to better compare events and to help sort out “different possible causal pathways to civil war onset,

62 Blaikie and Brookfield note the common failing of most of the literature on Latin American agrarian issues to adequately assess environmental issues. “There is a very large literature on land tenure and agrarian structures as impediments to productivity and causes of inequality, but in so far as the environment is considered at all in most of this literature, it is only as a passive background to human interaction. The degree to which this is so is quite remarkable. It runs through almost the whole of the vast literature on agrarian issues in Latin America, for example. Even in the series of reports prepared in the 1960s for the Inter-American Committee for Agricultural Development (CIDA), a major outcome of the Punta del Este conference convened in the aftermath of the Cuban revolution to find means of resolving agrarian discontent without revolution, environmental factors are statically and briefly described as a basis for the real information on land use and land tenure … In consequence we have a substantial body of very insightful literature on agrarian problems of transition closely related to political and economic theory, and little until recently on the social, economic and political aspects of environmental transition.” Piers Blaikie and Harold Brookfield, Land Degradation and Society (New York: Routledge,1994): xvii-xviii. 63 David Armitage, “Civil War and Revolution,” Agora 44 (2), 2009: 19-20; Kalyvas, Logic, 19. Armitage and Kalyvas convincingly argue that revolutions are a sub-set of civil wars. For Armitage, groups fighting in civil wars with major societal transformation goals can be characterized as revolutionary. See also David Armitage, Civil wars: a history in ideas (New York: Alfred A. Knopf, 2017). Kalyvas also agrees that revolutions are a subset of civil wars, though he distinguishes between small scale rural revolts or rural rebellions and “[l]arge-scale insurgencies with a predominantly rural base” that are sustained long-enough and have enough organization to challenge the state. In his definition, small-scale rural rebellions would not qualify as civil war. However, if we are interested in civil war onset, the causes of these smaller-scale events can give insights into causes of larger civil wars.

32 duration, and outcome as well as post-conflict peace duration.”64 Unrest in Peru has been categorized as a class-based revolutionary event.65 On the surface, this seems fair because revolts from the late 1950s to the 1980s ranged from limited land invasions to movements with transformative revolutionary goals like Sendero Luminoso.66 However, the degree to which race and identity played a role in these events has been under-explored by scholars.67 So, we must be sensitive to identity dimensions in our analysis, even while grounding the study primarily in the theoretical scholarship on revolutions. These insights are also integrated with a brief review of research linking environmental change to violent conflict in order to explore whether the impact of environmental scarcities can refine these explanations. Thirdly, the chapter provides a clear understanding about how existing research on twentieth century unrest in Peru integrates environmental scarcities into their overall explanations, and whether other factors are instead emphasized as key in those explanations. Scholars from a variety of disciplines have generated insights about why rural conflict and peasant mobilization occurred in highland Peru over the past half century. This chapter’s overview of debates about the causes of rural rebellion in Peru and wider scholarly debates about the impacts of environmental and demographic change on social conflict thus provides the foundation for our detailed examination of events in the D. of Chuschi in the remainder of the dissertation.

64 T. David Mason, Sara McLaughlin Mitchell, and Alyssa K. Prorok, “What Do We Know about Civil Wars? Introduction and Overview,” in T. David Mason and Sara McLaughlin Mitchell eds., What Do We Know about Civil Wars? (New York: Rowman & Littlefield, 2016), 2. Plotted in a four by four table, their categorization would yield four possible types of conflict events: ethnic revolutionary events, ethnic secessionist events, class-based revolutionary events, and class-based secessionist events. 65 Mason et al., “What,” 4. 66 Jack A. Goldstone, "Toward a Fourth Generation of Revolutionary Theory," Annual Review of Political Science 4 (2001): 142. 67 Much of the research examining rebellion in Peru in the latter twentieth century was written when class-based analysis was one of the dominant lenses for explaining civil unrest.

Theories of revolution and rural rebellion

Studies of rural rebellions or peasant revolts can be distinguished from general theories of revolutions. General theories of revolution apply to “great revolutions”68 – such as the French revolution of 1789 – events which Theda Skocpol argued resulted in “rapid, basic transformations of a society’s state and class structures … accompanied and in part carried through by [mass]-based revolts from below.”69 Skocpol argued that revolution occurs when peasant insurrection takes place in the context of a collapsing state – a collapse that comes about because of external military pressure and (or) landlord resistance to agrarian reform. Skocpol’s macro-structural analysis of revolutions in France, Russia, and China70 capped what sociologist Jack Goldstone has termed the “third generation” of revolutionary analysis – a body of scholarly theorizing about revolutions written in the 1960s and 1970s that moved beyond prevailing explanations relying on broad single factors like ‘modernization’ or ‘relative deprivation’ to emphasize instead “a conjunction of multiple conflicts involving states, elites, and the lower classes.”71 Largely an American academic enterprise, much of this work was initially inspired by the concerns at that time with revolutionary movements in South East Asia, an area deemed of primary importance to U.S. national security policy. By 1979, when Skocpol’s book “States and Social Revolutions” was published, research on revolutions was moving in new directions. Her work thus appeared to represent the high-point of scholarly consensus about the structural determinants of revolution.

Research on revolution splintered into many directions in the 1980s. Some scholars sought to deepen Skocpol’s findings through more fine-grained structural analyses, while others departed from her approach to instead emphasize theories focused on agency, contingency, or ideological factors.72 The application of Skocpol’s structural approach to a wider range of cases and the apparent inadequacy of structural approaches in explaining some of these events led to the

68 Goldstone, “Toward,” 142. 69 Quoted in Timothy P. Wickham-Crowley, “Structural Theories of Revolution,” in John Foran ed., Theorizing Revolutions (London: Routledge, 1997): 46-7. 70 Theda Skocpol, States and Social Revolutions: A Comparative Analysis of France, Russia, and China (Cambridge: Cambridge U. P., 1979). 71 Goldstone, “Toward,” 140. 72 The essays in Foran’s volume highlight the diverging paths taken by revolution scholars.

34 blossoming of alternative theoretical approaches. The new theoretical approaches could be differentiated according to the outcome of the revolution, the actors involved, the success or failure of revolutionary activity, and the scope of revolutionary activity. Iran’s Islamic revolution and the fall of communism in Eastern Europe similarly stimulated another broad stream of scholarly theorizing on revolutionary movements around issues of “guiding ideology.”73 Scholars dissatisfied with Skocpol’s “third generation approach” also called “for greater attention to conscious agency, to the role of ideology and culture in shaping revolutionary mobilization and objectives, and to contingency in the course and outcome of revolutions.”74

The inadequacies of Skocpol’s structural approach was also highlighted by theorists who recognized a convergence of thinking between revolution theory and theoretical work on social movements. These scholars realized “that many of the processes underlying revolutions – e.g. mass mobilization, ideological conflicts, confrontation with authorities – have been well studied in the analysis of social movements.”75 This led to the creation of a new literature of “contentious politics” that has worked to combine insights from both fields.76 The implications of this work, according to Goldstone, are both a revised definition of revolutions and the notion that the complexity of revolutions necessitates a more modular approach to future research.77 Goldstone argues that a basic consensus now exists among revolution scholars around three elements that seem to determine regime stability: first, “those factors that affect the strength of the state” (or what some term the opportunity structure for revolt); second, “competition among elites”; and third, popular living standards” (a source grievances).78 As a result, Goldstone

73 Goldstone, “Toward,” 142-3. 74 Goldstone “Toward,” 141. Following the Arab Spring revolts in 2010-11, social movement scholars added a focus on information technologies, like social media, for facilitating revolts or repression in the region. See Philip N. Howard and Muzammil Hussain, Democracy’s Fourth Wave? Digital Media and the Arab Spring (New York, NY: Oxford University Press, 2013). 75 Goldstone, “Toward,” 142. 76 Goldstone, “Toward,” 142. 77 Goldstone, “Toward,” 142. Goldstone offers a broader and more contemporary definition of revolution based on work from the past 20 years. Revolutions are “an effort to transform the political institutions and the justifications for political authority in a society, accompanied by formal or informal mass mobilization and noninstitutionalized actions that undermine existing authorities.” At their core they consist of a common set of elements: “(a) efforts to change the political regime that draw on a competing vision (or visions) of a just order, (b) a notable degree of informal or formal mass mobilization, and (c) efforts to force change through noninstitutionalized actions such as mass demonstrations, protests, strikes, or violence.” 78 Goldstone, “Toward,” 166-7.

35 believes that future work on revolutions need separate models of the “conditions of state failure, the conditions of particular kinds and magnitudes of mobilization, and the determinants of various ranges of revolutionary outcomes, each of which may be the result of contingent outcomes of prior stages in the revolution’s unfolding.”79 Goldstone’s recent survey suggests scholars can narrow the theoretical focus on the various stages and necessary elements on the path to revolution.

Rather than exploring the entire process of revolutionary activity in Peru, recent scholarship suggests that one could study the distinct stages along the path to revolution or the necessary elements which cause revolutions. It would thus make sense, for example, to study separately the origin and nature of the conditions out of which revolutionary movements arose in Peru (an element of what Goldstone would term as the conditions of state failure) from the factors responsible for the emergence and early growth of revolutionary movements in the Peruvian highlands (an element of revolutionary mobilization).

This study seeks to examine whether environmental scarcities influence the origins and nature of the conditions that create revolutionary movements, including factors affecting the grievances or motivations behind revolutionary movements. I do not explore the range of factors that fostered the mobilization and expansion of revolutionary activity, nor the determinants of the range of possible outcomes for the revolution. As a result, a good deal of theoretical work on revolutions and contentious politics is outside the scope of this project and not reviewed in this chapter. The core focus of this micro-study is instead assessing whether environmental change-conflict linkages can help us understand the origins and nature of the conditions that created revolutionary movements in Peru.

Particularly relevant to this study are those theories that focus on rural revolutionary events with limited goals, and the conditions which foster these uprisings. Today, scholars generally accept that “oppositional movements that either do not aim to take power (such as peasant or worker protests) or focus on a particular region or subpopulation are usually called rebellions (if violent) or protests (if predominantly peaceful).”80 Theories purporting to explain rural rebellions, rural

79 Goldstone, “Toward,” 174. 80 Goldstone, “Toward,” 142-3.

36 revolts, rural protests, agrarian revolts, agrarian revolutions, agrarian protests, etc., offer useful insights for the present study of Peru.81 These theoretical models combine a focus on the motivations of particular actors – those in the rural sector like peasants – with an appreciation of the typically limited goals of their revolutionary activities. Insights from this work informs empirical studies of peasant unrest in the latter twentieth century in Peru, discussed below in section 2.4.

Several scholars of peasant rebellion have focused on capitalist market transformation of peasant livelihoods and its role in causing peasant unrest. They debate, however, the impacts of these changes on different sectors of the peasantry and the role of changes in processes of rebellion. Accounts generally emphasize the impacts of grievances and/or motivations behind peasant dissatisfaction and the consequences of changes in opportunity structures during capitalist transformation that enable rebellion. This work is broadly consistent with more recent research on civil wars that debates the importance of motivations behind rebellion (greed or grievance), and the opportunity changes that facilitate revolt.82 Particularly important in these debates is the extent to which the material changes brought about by capitalist transformation and the impacts of altered social relations among peasants and elites helped to facilitate peasant revolts. Eric Wolf argued that peasant dissatisfaction in several twentieth century peasant revolts stemmed from the impacts of the global spread and diffusion of capitalism, which commodified land and labour, severed prevailing social pacts between landlords and peasants leading to increased peasant suffering, and increasingly alienated peasants from their work and their fellow peasants.83 Jeffrey Paige’s study of agrarian revolt in Peru similarly argued that the capitalist transformation of peasant and landlord relationships into wage relationships resulted in explosive tensions because the economic conflicts inherent in capitalist relationships transformed landlord-

81 To avoid confusion, however, I will adopt the term “rural rebellion” for consistency in this proposal unless referring to specific theories. 82 This recent research on civil wars is surveyed in Joseph K. Young, "Antecedents to Civil War Onset: Greed, Grievance, and State Repression," in What Do We Know About Civil Wars? Ed., T David Mason and Sara McLaughlin Mitchell (New York: Rowman & Littlefield, 2016), 38-39. 83 Eric R. Wolf, Peasant Wars of the Twentieth Century (New York: Harper and Row, 1969), 279-80.

37 peasant disputes to political conflicts.84 Tight political control, repression, and restricted political rights were at the heart of peasant grievances in such a system, and spurred mobilization of disaffected wage earning peasants.85

Somewhat related to the class-based explanation of grievances behind rural rebellion outlined by Wolf and Paige, James Scott believes that the issue of peasant subsistence survival is central to peasant grievances in cases of rural unrest in Southeast Asian history.86 The violation of a peasant’s subsistence ethic and the relational obligations between landlords and peasants are at the heart of his moral economy theory. The moral economy of peasants is “their notion of economic justice and their working definition of exploitation – their view of which claims on their product were tolerable and which intolerable.”87 Scott’s argument is thus similar to Wolf, who noted that capitalist market penetration disturbed the social-economic equilibrium between peasants and elites that governed the degree of elite confiscation of peasant surpluses.88 Scott believes that capitalist economic and political transformation during the colonial era in Southeast Asia systematically violated “the peasant’s vision of social equity,” leading to intolerable impoverishment and a basic threat to their subsistence survival.89 While the actual path from grievances to revolution depends upon a variety of additional variables, the rage that spawned peasant rebellion in this region stemmed from the impact of violations of peasants’ moral economy, according to Scott. 90

The moral economy relationship has questionable relevance in Peru, however, given the racism and Spanish conquest of indigenous peoples and their lands in Peru. Historically, there was

84 Jeffery Paige, Agrarian Revolution (New York: Free Press, 1975), 21-24. “The fundamental causal variable in this theory,” Paige writes, “is the relationship of both cultivators and noncultivators to the factors of agricultural production as indicated by their principal source of income.” The political behavior of both cultivating classes and noncultivating elites is largely determined by the source of the majority of their income – whether from landed property, wages, or from profits derived from market or industrial activities. Paige, Agrarian, 10. 85 Paige, Agrarian, 23. 86 James C. Scott, The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia (New Haven: Yale U.P., 1976): 3. 87 Scott, Moral, 3-4. 88 Wolf, Peasant Wars, 281. 89 Scott, Moral, 3-4. 90 Scott, Moral, 4.

38 little, if any, legitimacy to rule conferred by peasants on landed elites in highland Peru. Grievances borne of conquest, exploitation, and expropriation have always been a feature of the relationship between peasants and Mestizo elites, as discussed below in chapters 4 and 5. Racism against “Indian” peasants fueled centuries of abuse, constantly reaffirming the oppressive system governing peasants and undermining any idea of a legitimate moral bargain between peasants and landlords that could have been violated by capitalist transformation. Since the late-1800s, peasants increasingly agitated to overturn the abuses of Conquest, making the notion of the existence of a moral bargain in Peru an illusion.

Although the process of capitalist transformation was crucial in leading to the formation of peasant grievances in the studies discussed above, each emphasizes differing causal mechanisms. Material impoverishment is far more important according to the work of Wolf and Scott, compared to Paige. Paige’s conclusions generally speak to the processes and mechanisms that inhibit or promote social movement mobilization and enable revolts to grow. Various factors can affect the likelihood of revolts growing, according to Paige, such as overly limited goals, external support for political organization, national political changes that can impact the opportunity structure of revolts, the influence of strong political parties, and national economic development changes.91 He says little about factors that cause social grievances to form in the first place, aside from zero-sum struggles over wage levels. Wolf agrees with Paige, however, that grievances are insufficient to mobilize peasants to act without changes in the opportunity structure for revolt and mobilization: “Poor peasants and landless laborers … are unlikely to pursue the course of rebellion, unless they are able to rely on some external power to challenge the power which constrains them.” This tilt to the causal importance of mobilization and opportunity structure changes as key enablers of rural rebellions is consistent with the focus on opportunities in recent debates about whether greed versus grievance factors better explain civil war onset.92

The dislocation and growing economic inequality that frequently accompany capitalist transformation of subsistence agricultural societies suggest a mechanism linking the grievances

91 Paige, Agrarian, 345-48. 92 Young, “Antecedents of Civil War Onset,” 37.

39 outlined by Wolf and Scott to the willingness of peasants to mobilize in rebellion in agricultural wage economies described by Paige. Recent research argues that vertical inequality is inconclusively linked to violent conflict.93 However, scholars have found more support for the impact of perceived or actual horizontal inequality for causing violent conflict.94 Horizontal inequality results from “differences in access and opportunities across culturally deﬁned (or constructed) groups based on identities such as ethnicity, region, and religion.”95 This evidence suggests that relative deprivation mechanisms operate in the face of horizontal inequality.96 Relative deprivation may instead be the mechanism at the heart of Scott’s moral economy argument, either as inequalities between peasants and landlords increase, or as capitalist transformation increases peasant differentiation.97 Group identity definitions are inherently slippery, however, especially in countries like Peru where identity construction based on economic wealth seems to blur the line between vertical and horizontal inequality.

The focus on peasant grievances as an important causal factor in explaining peasant rebellions has been challenged by some scholars. Skocpol claims that peasants are always aggrieved; thus, their grievances fail to explain why a revolution occurs when it does. As a result, any theory focusing on peasant grievances “tries to turn a constant feature of the peasant condition into an explanatory variable.”98 Instead, she “proposes a theory of peasant insurrections that combines the capacity to rebel with the opportunity to rebel.”99 The capacity of peasants to rebel arises from their “internal structural solidarity and their external structural autonomy from elite control.”100 The opportunities for peasant rebellion increase once the repressive apparatus of state control over peasants is removed or weakened as a result of state collapse from international pressures and state fiscal paralysis stemming from entrenched agricultural elites

93 United Nations and World Bank. Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict (Washington, DC: World Bank, 2018): 110. 94 UN, Pathways, 111-112. 95 UN, Pathways, 111. 96 James B. Rule, Theories of Civil Violence (Berkeley: University of California Press, 1988): 200-223; UN, Pathways, 111. 97 I am grateful to Thomas Homer-Dixon for pointing out this point. 98 Skocpol, States, 114-15. 99 Wickham-Crowley, “Structural,” 53. 100 Wickham-Crowley, “Structural,” 53.

40 blocking necessary reform.101 In Skocpol’s theory, Timothy Wickham-Crowley writes, the “weakening of repression, above all of state repression, takes the place of the ‘increased grievances’ of relative deprivation and moral economy theories in helping to understand the timing and actual appearance of uprisings.”102

To be fair, Scott’s focus on the subsistence crisis that developed along with capitalist transformation implies that grievances were not static, as Skocpol suggests. The material changes in peasant well-being that accompanies a subsistence crisis suggests that a tipping point was reached that facilitated peasant mobilization and opposition. While Scott’s argument largely focuses on the impact of capitalist economic transformation to cause a peasant subsistence crisis, by focusing on household survival his theory is sensitive to the role of environmental and demographic stress. Scott mentions that maintaining adequate subsistence livelihoods for the peasantry was complicated by land shortage and population demands – although these factors are not systematically treated in his work. He is similarly aware that “the physical setting of certain areas subjected their inhabitants to fluctuations in yield of such amplitude that, even without the claims of elites, their survival is tenuous.”103 Thus, in spite of the very different context in Peru’s rural rebellions, Scott’s research suggests a role for severe disruptions in peasant subsistence livelihoods and the exogenous impacts of demographic and environmental stress in enabling rural rebellions.

Careful consideration of debates in the rural rebellion research about whether grievances and motivation are as important as changes in opportunities to cause rebellion lead to the conclusion that both factors are important if we are to make sense of how the sources of peasant dissatisfaction contribute to rural rebellion. Research on civil wars and revolutions accept such multi-factor theories and emphasize the importance of the state of the peasantry, factors that

101 Wickham-Crowley, “Structural,” 57. 102 Wickham-Crowley, “Structural,” 54. 103 Scott, Moral, 197. “An uncompromising ecology alone may be enough to spark a great amount of unrest,” Scott writes, “but when it is joined with a dissident intelligentsia based in the region the combination will be far more volatile.” Scott, Moral, 198, footnote 9. He similarly notes that the distribution of water supplies in rain deficit areas is likely to be a source of significant dissention. See Scott, Moral, 198, footnote 10.

41 facilitated peasant mobilization, and structural changes that opened opportunities for rebellion.104 Context matters, however; differing empirical realities may demonstrate different causal mechanisms generating rural rebellions. The consequences of specific agrarian structures noted by Paige and the disruptions to the moral economy of the peasantry noted by Scott should instead be seen as “functional alternatives” expressed in different revolutionary situations, according to Wickham-Crowley.105 Some change in the well-being of the peasantry was an essential element for revolt in many cases.106 Peasants “respond not only to violations of their moral economy, but also to intrusions on their physical economy. That is, we might expect peasants to respond to changes in their relations to the land itself, in particular to their gain or loss of lands due to various social or political measures taken by governments and individuals,” especially when those changes affect historical or contemporary expectations of well-being and progress.107 Processes of rural change like land grabbing or agrarian reform can thus be crucially important for assessing the likelihood of rural revolts because they directly impact peasant well-being.108 However, we could also widen our focus to the range of factors that were responsible for the

104 Young, “Antecedents of Civil War Onset,” 41; Timothy P. Wickham-Crowley, Guerrillas & Revolution in Latin America: A Comparative Study of Insurgents and Regimes Since 1956 (Princeton, Princeton U.P., 1993): 93; Kurt Schock, "A Conjunctural Model of Political Conflict: The Impact of Political Opportunities on the Relationship between Economic Inequality and Violent Political Conflict," Journal of Conflict resolution 40, no. 1 (1996), 99- 100. Wickham-Crowley’s rich multifactor explanation of the causes of revolution in Latin America highlights a variety of factors that influence the likelihood and success of revolutionary mobilization, structural changes that impact the opportunity structure for successful revolt like the impact of elite alliances or external support, and factors that speak to the state of aggrieved groups. Taken together, he argues that “… revolutions came to power in Latin America from 1956 to 1990 only when a rural-based guerrilla movement secured strong peasant support in the countryside and achieved substantial levels of military strength; if that movement also faced a patrimonial praetorian regime (a.k.a. mafiacracy), then it was structurally pressured to seek, and succeeded in securing a cross-class alliance against the patrimonial dictator who, lacking the social bases of support to resist such an alliance, in the end fell to a national resistance; under such conditions the United States tended to withdraw support from the dictatorship because of the symbolic and social pressures exerted by the constitutionalist and electoral symbols under which the revolutionaries and their more moderate allies united.” Wickham-Crowley, Guerrillas, 320. 105 Wickham-Crowley, Guerrillas, 320. 106 Wickham-Crowley, Guerrillas, 118-119. 107 Wickham-Crowley, Guerrillas, 118. Italics in original. He goes on to suggest that peasant assessments of their relative deprivation compared to other groups is a key factor in peasant willingness to rebel: “Just as in Scott’s case, the peasants are always measuring their status at any point in time against some kind of ‘reference group’ or measuring stick, usually their own historically specific experience. Where they feel that their lot is worsened relative to that reference group, we are likely to find radical sentiment, just as Scott found when the peasant moral economy began to decay.” 108 Wickham-Crowley, Guerrillas, 119.

42 outcome of agrarian change if these factors helped materially impact peasant well-being, including processes of environmental and demographic stress.

Research linking environmental change and violent conflict

Research linking environmental change to violent conflict compliments insights from rural rebellion studies by deepening our understanding of processes and factors that impact peasant well-being, and thus possibly play a role grievance (or motivation) formation and opportunity changes.109 Concerns about the security implications of human-induced environmental change have a long and contentious history.110 A central question facing students of security in the post- Cold War period has been whether the definition of security should be defined more broadly to incorporate factors that have not traditionally been understood as determinants of security.111 An important thread of this debate examines the merit of the concept of “environmental security,” which implies a fundamental connection between the pursuit of environmental goals and the pursuit of security.112 Partly because this discussion seems ultimately unresolvable, several researchers focused on the narrower and more tractable question of whether a causal link exists between environmental change and violent conflict. 113

In the 1990s, a number of scholars examining this relationship chose to focus on those areas where both the local environmental relationships were crucial for people’s survival, and the opportunities and capabilities to forestall negative implications were weakest – in the world’s

109 Portions of the following section were first published in Tom Deligiannis, "The Evolution of Qualitative Environment-Conflict Research: Moving Towards Consensus," in Environmental Security: Approaches and Issues, ed. Rita Floyd and Richard Matthew (New York: Routledge, 2013). 110 Simon Dalby, Security and Environmental Change (Cambridge: Polity Press, 2009), chapter 1. 111 Environmental security and environmental sources of conflict remain contested by some prominent peace and conflict scholars. As recently as 2015, Nils Petter Gleditsch questioned the wisdom of expanding traditional notions of security to include the environment, and argued that “environmental stress may be more appropriately viewed as a symptom that something has gone wrong than as a cause of the world’s ills.” Nils Petter Gleditsch, "Climate Change, Environmental Stress, and Conflict," in Managing Conflict in a World Adrift, ed. Chester A Crocker, Fen Osler Hampson, and Pamela Aall (Washington, DC: United Institute of Peace Press/Centre for International Governance Innovation, 2015), 149 & 161. 112 Geoffrey D. Dabelko, "Tactical Victories and Strategic Losses: The Evolution of Environmental Security" (Ph.D. University of Maryland, 2003). 113 Nina Graeger, "Environmental Security?" Journal of Peace Research 33, no. 1 (1996). This focus combined conventional and unconventional determinants of security and largely sidestepped the debate over the merits of the concept of environmental security.

43 poorest, developing states.114 People who are heavily reliant on natural resources for their survival– particularly renewable resources like land, water, and forests – and who are limited in their ability to sustainably manage these resources are particularly at risk of the impacts of human-induced environmental transformation. Today, almost half of the people on the planet rely upon local natural resources for a large part of their well-being.115 Those living in developing countries are particularly tied to their local natural resources and thus vulnerable to human-induced pressure on these resources. Investigating the material impact of changes in these key resources is thus highly relevant.

During the 1990s, qualitative research projects in Canada, led by Thomas Homer-Dixon at the University of Toronto (Toronto Group), and in Switzerland, led by Guenter Baechler (Bern- Zurich Group), set out to conduct a series of qualitative case studies on environmental change- conflict linkages in the 1990s. Each hypothesized that human pressure on natural resource endowments could affect the material well-being of developing societies and increase the risk of conflict.

The Toronto Group’s research suggests that environmental scarcities indirectly help to generate various forms of civil conflict, like insurgencies, group-conflict, coup d’etats, etc. 116 Their research did not support a link between human-induced environmental and demographic scarcities and inter-state conflict. Homer-Dixon defines environmental scarcity as a tripartite variable—a composite of three factors: degradation or depletion of resources (supply-induced scarcity), increased demand for resources due to population growth or increased per capita consumption (demand-induced scarcity), and changes in access to resources due to skewed distributions of resources among social groups (structural scarcity).117 These sources of scarcity

114 Homer-Dixon, Environment, 4-5. 115 Millennium Ecosystem Assessment, Eco-Systems and Human Well-Being: Synthesis (Washington DC: Island Press, 2005), 49. The report is available at: . 116 Homer-Dixon, Environment, 177. 117 Homer-Dixon, Environment, 47-52. Environmental scarcity is a composite term that includes environmental stress and demographic change. According to Homer-Dixon, environmental scarcity is a tripartite concept consisting of supply-induced scarcities, demand-induced scarcities, and structural scarcities. He argues that “ecologists and environmentalists often focus on environmental change, a term that refers only to a human-induced decline in the quantity or quality of a renewable resource – that is, to worsening supply-induced scarcity.” Equally important, however, are demand-induced scarcities caused by the effects of population growth or increasing

44 can operate independently but often interact with important effects. Homer-Dixon hypothesized that environmental scarcities influence the incidence of violent civil conflict through a series of intermediate social effects, like constrained economic productivity, intra or inter-state migration, the creation and aggravation of group tensions and divisions, and the weakening of institutions and the state’s capacity to respond to public needs and effectively deliver public goods. (See Figure 4)

Figure 4: Toronto Group’s Model of the Causal Links between Environmental Scarcity and Violence

Source: Homer-Dixon 1999, 134.

As well, scarcities often interact in particularly important ways to cause resource capture and ecological marginalization.

resource use, which reduces a resource's per capita availability by dividing it among more and more people. Thirdly, “scarcity is often caused by a severe imbalance in the distribution of wealth and power that results in some groups in a society getting disproportionately larger slices of the resource pie, whereas, others get slices that are too small to sustain their livelihoods. Such unequal distribution – or what [Homer-Dixon] call[s] structural scarcity – is a key factor in virtually every case of scarcity contributing to conflict.” Homer-Dixon, Environment, 48 and 15.

Resource capture occurs when the degradation and depletion of a renewable resource (a decrease in supply) interacts with population growth (an increase in demand) to encourage powerful groups within a society to shift resource access (that is, to change the resource’s distribution) in their favor. These groups tighten their grip on the increasingly scarce resource and use this control to boost their wealth and power. Resource scarcity intensifies scarcity for poorer and weaker groups in society.118

Ecological marginalization is often interlinked with resource capture and often a consequence of resource capture. Ecological marginalization occurs when unequal resource access (skewed distribution) combines with population growth (an increase in demand) to cause long-term migration of people to ecologically fragile regions such as steep upland slopes, areas at risk of desertification, tropical rain forests, and low-quality public lands within urban areas. High population densities in these regions, combined with a lack of knowledge and capital to protect the local ecosystem, cause severe resource degradation (a decrease in supply).119

In all cases, Homer-Dixon and his colleagues emphasized that scarcities never act alone to cause conflict, but instead interact with a wide range of contextual factors, operating across multiple levels and multiple scales.120 Thus, according to this model, environmental scarcity is an indirect cause of intrastate conflict.121

Günther Baechler’s Zürich-based Project on Environment and Conflict (Bern-Zürich Group)122 examined a much broader selection of case studies, but came to similar conclusions as the Toronto Group in the end. While sharing a similar concern with the Toronto Group about the impact of environmental change on the material well-being of people in developing countries123,

118 Schwartz, Deligiannis, and Homer-Dixon, Environment and Violent Conflict, 275. 119 Schwartz, Deligiannis, and Homer-Dixon, Environment and Violent Conflict, 275; Homer-Dixon, Environment, 177. 120 Homer-Dixon, Environment, 104 – 106 and 169 – 176. 121 Schwartz, Deligiannis, and Homer-Dixon, Environment and Violent Conflict, 295-7; and Homer-Dixon, Environment, 177. 122 The Bern Zürich Group’s was initially known as the Environment and Conflicts Project (ENCOP). I refer to their work as the Bern-Zürich Group because there were additional projects after ENCOP. 123 Günther Baechler, Violence Through Environmental Discrimination: Causes, Rwanda Arena, and Conflict Model (London: Kluwer Academic Publishing, 1999), xi.

Baechler’s focus on the transformation of human-environment relationships as a starting point of analysis, results in a much broader independent variable than Homer-Dixon’s focus on environmental scarcities. Though environmental transformation encompasses both negative and positive consequences, the Bern-Zurich Group’s focus is the negative consequences of human- induced environmental transformation. It can frequently lead to “environmental discrimination,” which “occurs when distinct actors - based on their international position and/or their social, ethnic, linguistic, religious, or regional identity experience inequality through systematically restricted access to natural capital (productive renewable resources) relative to other actors."124 Baechler takes a similar multi-causal approach to explaining how human pressure on the natural environment can help to cause conflict. Environmental transformation combines with various factors to result in different types of sub-state conflict, such as ethnopolitical conflicts, centre- periphery conflicts, migration conflicts, or in international environmental conflicts.125

Almost three decades after the beginning of research on environment and conflict, little, if any, consensus exists on research priorities, theories, or findings.126 Some scholars have strongly criticized the approaches of Homer-Dixon and Baechler,127 while others have suggested hypotheses that claim to improve the explanatory power of their models.128 Still others have continued their approach, adding conceptual clarity and theoretical sophistication to the frameworks laid out by those initial projects.129 The environmental conflict research agenda has also splintered into a number of research sub-programs, generally following qualitative or

124 Baechler, Violence Through Environmental Discrimination, 87. 125 Baechler, Violence Through Environmental Discrimination, 87. See Fig. 6.1, 180. 126Geoffrey D. Dabelko, "Tactical Victories and Strategic Losses: The Evolution of Environmental Security" (Ph.D. University of Maryland, 2003), 66 and 109; Ide, “Why do conflicts over scarce renewable resources turn violent?” 61-70; Deligiannis, "The Evolution of Qualitative Environment-Conflict Research: Moving Towards Consensus." 127 Marc A Levy, "Is the Environment a National Security Issue?," International Security 20, no. 2 (1995); and Hartmann, "Will the Circle be Unbroken?". 128 Jack A Goldstone, "Toward a Fourth Generation of Revolutionary Theory," Annual Review of Political Science 4, no. 1 (2001), 139-187; Nils Petter Gleditsch and Paul F Diehl, Environmental Conflict (Westview Press Incorporated, 2001); and Indra De Soysa, "Paradise is a Bazaar? Greed, Creed, and Governance in Civil War, 1989- 99," Journal of Peace Research 39, no. 4 (2002), 395-416. 129 Colin H Kahl, States, Scarcity, and Civil Strife in the Developing World (Princeton: Princeton University Press, 2008); Leif Ohlsson, Livelihood Conflicts: Linking Poverty and Environment as Causes of Conflict (Swedish International Development Cooperation Agency (SIDA), 2000); and Deligiannis, "The Evolution of Qualitative Environment-Conflict Research: Moving Towards Consensus."

47 quantitative research approaches.130 One strand of qualitative research, including projects led by Homer-Dixon, Baechler, and some political ecology scholars, explicitly examine links between conflict and the use of those natural resources directly necessary for survival.131 A second strand of more recent qualitative research examines the links between the commercial exploitation of valuable commodity resources—usually, high-value, non-renewable resources—and conflict.132 Quantitative research on natural resources and conflict also has developed, focusing primarily on evaluating claims linking valuable commodity resources and violent conflict.133 Currently, most attention among both qualitative and quantitative research focuses on extractive sector disputes in poor countries over “abundant” or high-value resources and on quantitative tests of environment-conflict linkages, with somewhat less—but still notable attention—given to links between rapid demographic change and conflict.134

Qualitative research on environmental change-conflict linkages has penetrated policy discussions and provided a theoretical foundation to recent research examining climate change and security linkages.135 However, study of the original questions addressed by Homer-Dixon and Baechler’s

130 Distinctions among strands of environment-conflict research are not clear-cut and frequent sources of contention among scholars. The author acknowledges significant overlap in approaches between qualitative researchers, and among qualitative and quantitative researchers, on the causal relationship between natural resources and violent conflict. For example, distinctions are somewhat artificial between qualitative researchers about whether resources are commercially exploited for sale in markets (sometimes global or regional markets) or whether resource use takes place for household survival. In many cases, both dynamics operate, and one type of exploitation impacts the other, and the subsequent forms of conflict that may develop. 131See Nancy Lee Peluso and Michael Watts, eds., Violent Environments (Ithaca, NY: Cornell University Press, 2001) on the political ecology of environmental conflict. Durham provides an early analysis linking human-induced environmental change and violent conflict. William H Durham, Scarcity and Survival in Central America: Ecological Origins of the Soccer War (Stanford University Press, 1979). I discuss disputes and similarities between the work of the Toronto Group and political ecologists in, Deligiannis, “The Evolution of Qualitative Environment- Conflict Research,” 40-44. 132 Philippe Le Billon, "Resources and Armed Conflicts," The Adelphi Papers 45, no. 373 (2005), 29-49. Le Billon’s arguments are expanded in, Philippe Le Billon, Wars of Plunder: Conflict, Profits and the Politics of Resources (London: Hurst & Company, 2012). 133 Le Billon, “Resources.” Also see James Ron, "Paradigm in distress? Primary Commodities and Civil War," Journal of Conflict Resolution 49, no. 4 (2005), 443-50. I examine disputes around greed and grievance debates in environment-conflict research in, Deligiannis, “The Evolution of Qualitative Environment-Conflict Research,” 51- 53. 134 Research linking high-value commodity resources and conflict, quantitative studies on environment-conflict linkages, and demographic security research are beyond the scope of this review. 135 Wissenschaftlicher Beirat der Bundesregierung Globale Umweltveränderungen (German Advisory Council on Global Change), Climate Change as a Security Risk, Routledge (London, 2009). Global Humanitarian Forum, The Anatomy of a Silent Crisis, Global Humanitarian Forum (Geneva, 2009).

48 projects—that is, of the particular connections between environmental change or scarcity and conflict—has progressed little. Basic ontological, epistemological, and methodological disagreements and, in some cases, harsh polemics appear to have paralyzed research.136 This line of research is largely moribund, with little agreement on fresh questions to move inquiry forward. However, the key conclusions of Homer-Dixon’s research that scarcities can constrain economic productivity, stimulate migration, and worsen group conflict compliments the rural rebellion research discussed above, suggesting specific causal mechanisms that can underpin peasant grievance formation. The following section explores these theoretical hypotheses in the context of empirical research on the causes of rural unrest in Peru in the latter half of the twentieth century.

Specific explanations for unrest in Peru 2.4.1. Pre-land reform unrest

Empirical studies of unrest in Peru echo some of the theoretical conclusions of research on rural rebellions and environmental change-conflict linkages by stressing the impact of changes in highland Peru for mobilizing peasants against landlords and the Peruvian state. Much of the research on rural revolts and revolutions in Peru from the late 1950s to the 1980s is temporally split, with analysts either focusing on peasant unrest in the pre-land reform period in the late 1950s and early 1960s, or on the consequences of Peru’s 1969 agrarian reform and the factors leading to the outbreak of the Sendero Luminoso insurgency in the early 1980s. Although this is a somewhat artificial division, we can usefully divide the following discussion about the factors highlighted by scholars to explain rural unrest in Peru into two broad sections discussing explanations for the pre-agrarian reform and post-agrarian reform periods. Work on the first wave of peasant unrest in the late 1950s and 1960s has emphasized a range of causal factors that altered peasant livelihoods to increase peasant grievances and enable peasant mobilization and unrest.

A recent survey of Peruvian history by Peter Klaren points to the uneven patterns of social and economic transformation in highland Peru in the 1940s and 1950s as key to explaining peasant

136 Thomas F. Homer-Dixon, Nancy Peluso, and Michael Watts, "Exchange: On Violent Environments," Environmental Change and Security Project Report 9 (2003), 89-96.

49 unrest in the Sierra. In the years following WWII, export-led growth propelled Peru’s economy forward, particularly on the strength of the agro-export and mining sectors.137 While Lima and other areas on Peru’s coast were the earliest and biggest beneficiaries of the expanding economy, its modernizing impact gradually percolated into the highlands during the late 1940s and 1950s. Those areas of the Sierra with ready access to the coast, such as the central highland departments of Junin and Cerro de Pasco, were the first to see these benefits as increases in inter-zonal trade and capital inflows led to a “commercial awakening” – a “rapid increase in the movement of money, goods, and people” during the 1950s and 1960s.”138 However, the impact of the commercial awakening in the highlands “was not geographically or socially uniform; it was confined mostly to the Sierra regions with easy access to Lima and the central coast and to a rural bourgeoisie made up of merchants, artisans, bureaucrats, small-to medium-sized landowners, and other inhabitants of small towns and provincial cities.”139 Growth was particularly concentrated in the department and provincial capitals of the central Sierra, which tended to become centres of capitalist transformation compared to the largely traditional agricultural areas surrounding the highland towns. Over time, those urban centres in the midst of capitalist transformation served as the vehicle for greater market penetration into the lives of nearby peasants.

Uneven patterns of market penetration in the Sierra combined with several other significant trends at the time like declining terms of trade for agricultural products to significantly affect the livelihoods of peasants and set the stage for rural unrest, according to Klaren. Economic studies support Klaren’s analysis. Export-led growth in Peru during the 1950s and 1960s led to a general rise in prices throughout the country at the same time as rural incomes stagnated or fell as a result of the combined impact of unfair government pricing policies that held food prices artificially low for the benefit of urban residents, and rising imports of cheaply produced foreign foods.140 Those in the rural sector who produced for the domestic market were gradually being

137 Peter Flindell Klaren, Peru: Society and Nationhood in the Andes (New York: Oxford U.P., 2000): 302-5. 138 Klaren, Peru, 309. Quoting Richard Webb, Government Policy and the Distribution of Income in Peru, 1963- 1973 (Cambridge, MA: Harvard U.P., 1977): 27. 139 Klaren, Peru, 309. 140 Pedro-Pablo Kuczynski, Peruvian Democracy Under Economic Stress: An Account of the Belaunde Administration, 1963-1968 (Princeton, New Jersey: Princeton University Press, 1977): 10-11.

50 squeezed. As the cost of living increased around them, their crops brought an increasingly diminished return for their labor. Klaren argues that “the growing inequality and social differentiation that characterized the highland peasant population as capitalism and the internal market advanced in its customarily variegated and uneven way, set the stage for the upsurge in rural unrest and peasant mobilization that suddenly erupted during the late 1950s and that would peak in the middle of the next decade.”141 Klaren is thus suggesting that rising inequality and increasing social differentiation among highland peasants was increasing pressure on peasant livelihoods, leading to rising grievances or creating new grievances for peasants. These changes enabled conditions for unrest in the highlands.

While capitalist penetration into the Sierra increasingly affected the lives of peasants, trends were developing to slowly dissolve the power of traditional elites in the Sierra, shifting the opportunity structure and enabling peasant mobilization. Paige’s research on the impacts of a growing agro-export economy explain how economic transformation weakened highland landowning elites much more severely compared to coastal estate owners, enabling more effective peasant mobilization in the highlands. Economically weak highland estate owners relied on coercive political and legal control to protect rights to land, and to maintain an exploitative grip on highland peasants. 142 With limited prospects for mechanization or increases in productivity in the Sierra and declining terms of trade for agricultural products, increases in earnings from land holdings could only come at the expense of landowners or peasant cultivators. On Peru’s coast, by contrast, politically and economically powerful sugar and cotton estate owners were able to maintain enough political influence with the central government in the late 1950s and early 1960s to retain coercive control over their land.143 Capitalist penetration in the Sierra and the slow withdrawal of central government support and assistance for the economically weak highland land owners thus opened the door for peasants to re-take stolen lands with little fear expulsion or retribution, while coastal plantation elites were able to forestall similar worker mobilization.144 Klaren also points out that “the long-standing quid pro quo

141 Klaren, Peru, 311. 142 Paige, Agrarian, 339-40. 143 Paige, Agrarian, 340. 144 “Dependence on land [among the upper class]”, Paige writes, “led to economic weakness and political vulnerability, to servile labor and resistance to worker organization, and to static production and zero-sum conflict.

51 between the state and the gamonal [traditional abusive rural elite] class to maintain order in the interior was undermined by the ever-expanding reach of the government by means of new roads and agents sent into the remote corners of the country.”145 At the same time, population changes were helping to shift the political focus of the country’s ruling elite. Since 1940, Peru’s population had risen dramatically, growing 43% by 1960. Legions of Peruvians began to migrate from the Sierra to urban centres like Lima and Arequipa (Peru’s second largest city), swelling the urban population by 3.7% per year, compared to 1.2% in rural areas.146 Increasingly, politicians sought policies to stem the tide of migration from the highlands, and to respond to the demands of growing urban constituencies. Their policies would come at the expense of highland elites.

Outbreaks of peasant unrest in the Sierra in the late 1950s and early 1960s helped to undermine the dominant position of highland landowners as a consensus began to form among the country’s ruling elites in favor of some sort of land reform and redistribution. Anti-land reform forces – including some very powerful government cabinet members – increasingly found themselves isolated as the pace and scope of peasant unrest spread throughout the central and southern highlands, and peasants succeeded in organizing themselves into regional agrarian syndicates, thereby increasingly shifting the tide of Peruvian opinion in favor of immediate land reform.

Many of the same factors transforming Peru and undermining the power of rural elites in this period were also facilitating the mobilization of highland peasants. Paige argues that the rise of the agro-export economy in the Sierra provided economic incentives for highland communities in the central Sierra like San Pedro de Cajas and the peasant farmers of La Convencion valley in

Both in the Sierra and on the coast these political characteristics of the upper class led to political conflict over the ownership of landed property and resulted in a series of peasant land invasions. The difference between the coastal and Sierran haciendas also illustrates that within the category of commercial haciendas the degree of dependence on land versus industrial capital exerts a crucial effect on both the political and the economic behavior of the agrarian upper class. The greater mechanization and commercialization of the coastal cotton estates led to a more influential land-owning class, less use of coercion in labor recruitment, and greater productivity. These economic characteristics in turn led to less political vulnerability, greater tolerance of worker organizations, and a greater ability to concede a share of agricultural income to workers. As a result, conflict in coastal estates was focused not only on the question of land, which concerned the small holders and peasant communities in the valleys, but also on wages, which were the principal concern of the resident laborers of the estates.” Paige, Agrarian, 341-2. 145 Klaren, Peru, 311-12. 146 Klaren, Peru, 316.

52 the southern Sierra to organize and reclaim lost lands from weakened land owners.147 By contrast, peasants relatively untouched by the effects of agro-export penetration of the Sierra, like those in the traditional small-holding community of Hualcan or the peons of the hacienda Vicos, did not mobilize to seize lost lands or agitate for an end to abusive land owner exploitation, according to Paige. While many peasants wanted to end exploitation at the hands of landed elites, they could not overcome what Paige describes as the inhibiting effect on peasant collective action of the risk-adverse, isolated, and internally divided nature of traditional Peruvian peasant communities.148

Other scholars contend, however, that Paige’s focus on the role of the agro-export economy over-simplifies the developments in the highlands that enabled peasant mobilization. Researchers from Cornell University and Peru’s Instituto de Estudios Peruanos (IEP) discarded the “myth of the passive peasant”, which characterizes traditional peasant communities in this period as “naturally tradition-bound, fatalistic, and inclined to resist change.149 In the rapidly modernizing central valley and in southern Sierra population centres like Cuzco they instead found that peasants were not passive agents reacting to structural changes around them, but were actively engaged in transforming their Sierra reality to take advantage of structural changes that weakened traditional elites, even as market pressures described by Klaren squeezed their

147 Paige, Agrarian, 208. Paige writes: “It was the export economy that … made peasant land invasions both organizationally possible and economically profitable. The changing national political climate finally provided an opening for direct collective action, but it was only those communities which had been transformed by the export economy which were in a position to take advantage of the political change.” Paige, Agrarian, 209. 148 Summing up his analysis of traditional subsistence communities, Paige concludes that “Hualcan shares all the political and economic characteristics which made collective political action by the serfs of hacienda Vicos unlikely or even irrational. Its backward economy provides a slim margin of survival which could easily be threatened by either technical innovations in agriculture or changes in the political structure of the community. Precarious agriculture and tiny dispersed plots create the same kind of economic conservatism that was found at Vicos. The internal stratification of the community, particularly the division between the bottom lands and the hillside, creates an atmosphere of individualistic economic competition which inhibits cooperative economic or political action. The administrative structure of the varayoc (community leaders), the teniente gobernador (local official appointed by the metizo officials), and the hacienda mayorales (hacienda lieutenants) ensure that conservative organizations will co- opt or suppress independent peasant leadership. The limited financial resources of the community make it unlikely that union organizers could be supported from dues or other contributions of community members. Hualcan, like most of the traditional subsistence communities of the Sierra, had little reason to participate in the comunero movements of the 1960s and, like almost all of them, remained uninvolved.” Paige, Agrarian, 195. 149 The results of this project are summarized and synthesized in, William Foote Whyte and Giorgio Alberti, Power, Politics and Progress: Social Change in Rural Peru (New York: Elsevier Scientific Publishing Co., 1976).

53 livelihoods.150 The Cornell/IEP findings are corroborated by Howard Handelman’s study of peasant unrest in the Sierra. Handelman found that there were few traditional peasant communities left in the Sierra by the early 1960s, and that the impacts of modernization were seeping into even the most remote corners of the highlands.151 The movement of people and ideas around the highlands was also facilitating change. Klaren, for example, points to the increasing back-and-forth migration of peasants from the Sierra to the coast spreading new ideas and forms of political organizing, the rising penetration of improved forms of communication like radio into the far-corners of the Sierra, and the “subversive” ripples of the Cuban revolution as important for facilitating peasant mobilization.152

Handelman found that the integration of peasant movements into the national political system was crucial for their success in the central Sierra, while their lack of integration was important for their repression by the state in the southern Sierra. The first wave of peasant mobilization reached its peak shortly after the inauguration of President Fernando Belaunde Terry in 1963, and centred around cholo153 and mestizo communities in the central and northern Sierra – in places like San Pedro de Cajas in Junin, and the communities in the mining centre of Cerro de Pasco. Handelman argued that these communities were more modernized than towns in the

150 Whyte and Alberti, Power, 201-2. “In the course of [our] experience,” Whyte and Alberti write, “we abandoned the modernization framework and came to focus on structural change. Discarding the myth of the passive peasant, we came to see peasants as men and women in motion in response to forces they generated among themselves as well as in response to external forces. Community studies were no longer our central concern, and we came to view villages simply as convenient locations for the study of structural change at the microlevel. No longer did we see space and time information simply as background for a community study. In order to explain local events, we had to identify the external forces that impacted upon the community. We now see local events as being the results of the interaction between local conditions and external forces. According to this view, history should take us beyond understanding and enable us to discover those external conditions most relevant in the determination of local events.” Whyte and Alberti, Power, 5-6. 151 Howard Handelman, Struggle in the Andes: Peasant Political Mobilization in Peru (Austin, TX: The University of Texas Press, 1975) 191. 152 Klaren, Peru, 314. Many of these changes were noted by Handelman; however, Handelman believed them to be a product of agro-export growth in the highlands, including a higher proportion of bilingual community members, a higher degree of education, a greater proportion of people who had traveled outside of their village to large urban centers, greater integration into the capitalist economy, the use of modern agricultural techniques, etc. 153 Handelman argues that a “cholo” is “a person of Indian origin who lives among mestizos … and has been partially integrated into the white Spanish-speaking culture of the highlands,” while a mestizo is “a person whose ancestry is at least partially Indian but who has acculturated fully into white Spanish culture.” Handelman, Struggle, 278 and 281. For many Peruvians, cholo came to be seen as a derogatory term, illustrating the severe racism and discrimination in Peruvian society toward highland residents. The question of how to identify highland Peruvians varies depending upon the author and the area studied. This issue is discussed further in chapter 4 below.

54 southern Sierra, better plugged into the national political system, and engaged around “specific economic grievances” – the “scarcity of pasture land, expansion of large livestock estates, and diminishing opportunities for employment in the mines” of the area.154 In the end, they tended to be more successful in consolidating control over the lands that they seized, with the result that there was little violence during first wave. The second wave of peasant unrest was concentrated in the communities of the southern Sierra – especially in La Convencion valley and nearby provinces in the department of Cuzco – involving more traditional, isolated peasants who were far less engaged with national political organizations and relied on the organizational leadership of more ideologically radical peasant federations, rather than local village leaders. These communities mobilized as a result of “longstanding cultural and socioeconomic repression of Indian comunidades (communities).”155 This wave of unrest also took place in an area where there have been frequent peasant rebellions and a history of extremely violent repression at the hands of local elites and representatives of the state. To many of Lima’s political and military elites, the urgency of staving off a Cuban-style revolution proved to be a powerful justification for ignoring the protests of the increasingly marginal highland landowners in favor of some sort of land reform in the central Sierra.156 But as the leftist peasant movement in the south escalated following Belaunde Terry’s election, the old habits of violent retribution against peasants reappeared. Now the Peruvian government was more inclined to label unrest in Cuzco the result of a communist conspiracy, opening the door to violent repression by the military and police of peasant movements and the later insurgency in the area. Consequently, some of the worst massacres of peasants in the 1960s occurred during the unrest in the southern Sierra.157

Peasant mobilization in the highlands in the pre-1969 period, according to this survey, was thus broadly enabled by the commercial awakening in the 1950s and 1960s in the area. Export- agriculture may have kick-started these developments in Peru after World War II, but economic growth soon developed a particular dynamic of its own in the highlands, and was accompanied by myriad developments that facilitated mobilization of peasants, while opening up the

154 Handelman, Struggle, 113. 155 Handelman, Struggle, 113. 156 Klaren, Peru, 314-5. 157 Handelman, Struggle, 116-121. For information on the insurgencies of 1965, see footnote 117.

55 opportunity for revolt by undermining the power and control of highland elites. Importantly, however, trends differed in different parts of the highlands, depending upon a range of factors like integration with national economic and political movements, external ideological influences, and the degree of modern state penetration. Rising inequality and peasant differentiation also stressed many peasant livelihoods by the 1960s, contributing to new grievances among many in the highlands, and setting the stage for continued unrest in the years to come.

Environmental scarcities feature marginally in the first wave of research on unrest in the 1950s and 1960s. Land was certainly a main source of contention for many of the peasant mobilizations; however, there is little discussion of how or when these lands were seized by highland elites. We are left to speculate that land invasions sought to right some long-distant resource capture, making them relatively static grievances. Supply and demand pressures were not examined in detail by scholars, except for passing references to the stresses resulting from population growth. The Cornell/IEP study noted that population growth was contributing to demand-induced pressures on household land assets: “as the population continued to grow in the rural Sierra, sons divided the land of their fathers so that their work obligations remained the same and yet each family could make less out of its own lands.”158 On the whole, however, changing demographic trends are only generally described in Peru and its sub-regions, appearing mainly to aggravate the impact of precipitating factors listed above.159

2.4.2. Post-Land Reform unrest

Peasant unrest, rural mobilizations, and insurgencies did not end with the Peruvian government’s repression in the south-central Andes in the mid-1960s. In fact, land invasions continued sporadically for the next two decades, and deadly insurgencies sprang up in the 1980s – more than a decade after the MIR and ELN insurgencies in the southern Sierra were crushed by the Peruvian military.160 In the 1980s, the Sendero Luminoso (Shining Path) – and to a lesser extent,

158 Whyte and Alberti, Power, 202. 159 Whyte and Alberti, Power, 17-18. 160 The Movimiento de la Izquierda Revolucionaria (MIR) and the Ejercito de Liberacion Nacional (ELN) led the short-lived insurgencies in Peru’s southern highland departments of Ayacucho and Cuzco in 1965 and 1966. Both groups were led by non-peasant intellectuals inspired by the Cuban revolution’s foco model, where “a small vanguard of guerrillas [attempt to] gain the support of peasants in an isolated area and from there, initiate a successful revolutionary war.” Klaren, Peru, 329. Unfortunately, their insurgency was ill conceived, and they made

56 the Movimiento Revolucionario Tupac Amaru (MRTA) – initiated a civil war in Peru which eventually claimed almost 70,000 deaths by 2000, according to Peru’s Truth and Reconciliation Commission.161 Scholarly explanations about post-land reform unrest have tended to concentrate on efforts to understand the origins and causes of the Sendero insurgency, with less focus on the general structural sources of peasant grievances.

What is surprising about the insurgencies of the 1980s, however, is that they came after Peru had undertaken one of the most comprehensive land reform programs in all of Latin America. Most scholars now agree that the disappointing effects of the 1969 Agrarian Reform Law were crucially important in fomenting unrest in Peru, although they disagree about its importance relative to other factors. To many, land appeared to be a scarce commodity for many small- holding peasants in the Sierra before the 1969 agrarian reform, with large landed estates holding most of the land. Huge inequalities between elite landowners and peasants were believed to be the norm, according to researchers. Studies based on Peru’s 1961 census claimed that landholdings greater than 500 acres accounted for no more than 0.4% of all holdings, but represented almost 76% of surveyed holdings.162 By contrast, peasants holding less than 5 hectares of land accounted for more than 82% of the holdings surveyed.163 Later studies of Peru’s pre-Agrarian Reform landholdings concluded that highland land-holding domination by large estates was, in fact, a myth. A small group of estates between 2 and 50 hectares and a large group of minifundio family holdings under 2 hectares actually controlled 80% of Sierra farmland.164

little progress bonding with the peasants of the southern Sierra. Both guerrilla groups were eventually wiped out by the Peruvian military, along with approximately 8000 peasants. See Klaren, Peru, 328-330; Leon G. Campbell, “The Historiography of the Peruvian Guerrilla Movement, 1960-1965,” Latin American Research Review, Volume 8, No. 1 (Spring 1973): 45-70; and the interesting book by the surviving leader of MIR, Hector Bejar, Peru 1965: Notes on a Guerrilla Experience (New York: Monthly Review Press, 1970). 161 Comisión de la Verdad y Reconciliación, Informe final: Perú, 1980-2000 (Lima: Comisión de la Verdad y Reconciliación, Universidad Nacional Mayor de San Marcos, Pontificia Universidad Católica del Perú, 2004), General Conclusions, 316. The MRTA is the group made famous by their take-over of the Japanese embassy in Lima in December, 1996. 162 Solon Barraclough ed., Agrarian Structure in Latin America: A Resume of The CIDA Land Tenure Studies of Argentina, Brazil, Chile, Colombia, Ecuador, Guatemala, and Peru (Lexington, MA: Lexington Books, 1973): 252. 163 Barraclough, Agrarian, Table 11-1, 253. 164 Enrique Mayer, Ugly Stories of the Peruvian Agrarian Reform (Durham, N.C.: Duke University Press, 2009): 15-16. Enrique Mayer notes that an analysis in 1981 by José María Caballero in 1981 found that the degree of inequality was not as severe in the early 1960s as the CIDA figures quoted above made it out to be.

Pre-Agrarian Reform highland landholding reality aside, however, the perception driving calls for land reform in the 1960s was that poor peasants were suffering because landlords controlled most of the land in the sierra, a perception buttressed by the fact that most peasant households had very little land. In the early 1960s, land per capita in Peru was the lowest of all South American countries.165 In the southern Sierra, with more than 80% of the economically active population engaged in agriculture, the average family farm in the 1950s was about 0.9ha, divided into many small plots.166 Given the general scarcity of good agricultural land in the highlands and the fact that there were few large estates to expropriate, land reform as a solution was probably bound to disappoint many peasants.167

The first attempts at land reform actually occurred with the passage in 1962 of a law that attempted to diffuse the unrest in the La Convencion and Lares valleys of Cuzco. Hopes were raised at that time by the reform promises of President Belaunde-Terry, who appointed a land reform commission soon after taking office. However, Belaunde’s land reform law was ultimately a disappointment, watered down by the powerful landowners that dominated the commission helping to draft the law.168 In the end, Law 15037, as it was known, essentially excluded the prosperous coastal estates from expropriation, and merely legitimized those highland land occupations that had already taken place.169 “By 1969, about 4 percent of the land had been redistributed among 1 percent of the rural families.”170 Although limited in its

165 Cynthia McClintock, Revolutionary Movements in Latin America: El Salvador’s FMLN and Peru’s Shining Path (Washington D.C.: United States Institute of Peace Press, 1998): 167. 166 Watters, Poverty, 40-1. 167 Mayer, Ugly Stories, 16. Mayer notes that the “highlands faced an absolute shortage of land. The redistribution of land would therefore not solve many problems, since there was not that much of it to expropriate...” Land holdings by different holding size in Cangallo and the D. of Chuschi are examined in greater detail below in sections 4.2.2.1 and 7.3.2. 168 A. Eugene Havens, Susana Lastarria-Cornhiel, and Gerardo Otero, “Class Struggle and the Agrarian Reform Process,” in David Booth and Bernardo Sorj eds., Military Reformism and Social Classes: The Peruvian Experience, 1968-80 (London: The Macmillan Press Ltd, 1983): 22; Colin Harding, “Land Reform and Social Conflict in Peru,” in Abraham F. Lowenthal ed., The Peruvian Experiment: Continuity and Change under Military Rule (Princeton, NJ: Princeton U.P., 1975): 233-35; and Mayer, Ugly Stories, 18-19. 169 Susan Eckstein, "Revolution and Redistribution in Latin America," in The Peruvian Experiment Reconsidered, Cynthia McClintock and Abraham F. Lowenthal eds., (Princeton, NJ: Princeton U.P., 1983): 362. 170 Eckstein, “Revolution,” 362. Between 1964 and 1969, Harding notes, 14,631 families received a total of 375, 000 hectares. Harding, “Land,” 234. Mayer notes that Belaunde’s land reform law did eliminate serfdom in Peru – the yanaconaje and colonato system, where a family was granted a plot of land on an estate in exchange for obligatory work for the landowner. Mayer, Ugly Stories, 18.

58 application, the law was successful in stoking expectations among peasants for reform and in arousing fear among landowners, who quickly stopped all agricultural investment and proceeded to decapitalize their holdings by selling land and liquidating holdings where possible in anticipation of future reform attempts.171 As a result, agriculture in Peru fell into sharp decline, with per capita food production stagnating in the 1960s.172

While Belaunde lacked the political will to undertake land reform, the unrest of the 1960s crystallized opinion among influential officers in Peru’s armed forces in favor of wide-ranging and comprehensive land reform. The army’s encounters with Peru’s “backward” and exploited peasants in the southern Sierra during its 1960s counter-insurgency operations helped to convince many officers that radical transformation was necessary to modernize Peru’s agrarian sector.173 In fact, since the 1950s, Peruvian military officers had had broad ambitions. They increasingly came to see themselves as a “permanent vehicle for modernization” in Peru.174 As economic crisis began to envelop the Belaunde administration in 1968, Peru’s military stepped back into its traditional interventionist role and overthrew the democratic government. However, this time the officers in charge were preparing to embark on a revolution from above in an effort to remake and modernize Peru, and thereby forestall a communist revolution.175

The Peruvian military’s “revolution from above” was an attempt by the military and high-level state officials to initiate and guide social transformation with little or no mass participation.176 A key part of that effort involved the 1969 Agrarian Reform Law, where the military sought to “gain popular support, destroy oligarchic domination, control conflict and rural discontent, improve income distribution, stop massive migration to the cities, and create a stable agrarian

171 Eckstein, “Revolution,” 362. Needless to say, such actions no doubt only further exacerbated tensions between landowners and peasants who claimed the land. 172 Klaren, Peru, 331. Agricultural decline also accelerated because of a greater willingness by the Peruvian government to allow imports of foreign food products, which undercut domestic producers. This is discussed below in chapter 6. 173 Klaren, Peru, 338. 174 Quoted in Daniel M. Masterson, “Caudillismo and Institutional Change: Manuel Odria and the Peruvian Armed Forces, 1948-1956,” in Linda Alexander Rodriguez ed., Rank and Privilege: The Military and Society in Latin America (Wilmington, Delaware: Scholarly Resources Inc., 1994): 150. 175 In addition to the sources cited above, a bibliography of the extensive literature on Peru’s military “revolution” of 1968 can be found in Klaren, Peru, 468-470. 176 Eckstein, “Revolution,” 348.

59 sector for an expanding internal market.”177 However, during the “implementation of its reform program, the state’s involvement grew and – paradoxically – became increasingly out of touch with popular (and rural) demands.”178 In spite of this, most scholars agree that Peru’s agrarian reform ranks with Mexico and Bolivia as one of the most comprehensive in Latin America. “By 1980, after twelve years of military rule, 8.6 million hectares of land and 2.2 million head of livestock (representing 39 percent of the land and nearly 8 percent of the livestock) had been granted to 390,684 peasants.”179 In a few short years the traditional oligarchy and landed elite, with their large coastal estates and highland haciendas, ceased to exist in Peru.

Agrarian reform, it seemed, had eliminated one of the key sources of peasant grievances behind the mobilizations and unrest of the 1960s. However, the agrarian reform effort had important implications on the likelihood of insurgency in Peru in the next two decades. These are discussed in the context of the following overview of scholarly thinking about the origins and causes of the Sendero insurgency.

2.4.3. Debates on the origins of Sendero and the degree of peasant support

Efforts to explain rural unrest and mobilization in the post-land reform period remain highly contentious. Peasant mobilization and unrest in Peru’s Sierra in the 1950s and 1960s led scholars to search for the various underlying causes of peasant unrest, settling on a combination of both macro causal factors and micro processes of change and mobilization in rural areas. However, this has not occurred to the same extent with scholarly investigations of the origins of the Shining Path. The incredible violence and breadth of the Sendero uprising has somewhat mesmerized scholarly attention in a narrower quest to discover just what makes Sendero tick. Like drivers passing a traffic accident who are unable to take their eyes off the carnage, many scholars have been drawn to attempt to explain how a band of homegrown revolutionaries could be so ruthless in their attempt to topple Peru’s government that they would flay peasants alive,

177 Christine Hunefeldt, "The Rural Landscape and Changing Political Awareness: Enterprises, Agrarian Producers, and Peasant Communities, 1969-1994," in The Peruvian Labyrinth: Polity, Society, Economy, Maxwell A. Cameron and Philip Mauceri eds., (University Park, Pennsylvania: The Pennsylvania State University Press, 1997): 109. 178 Hunefeldt, “Rural,” 109. 179 Hunefeldt, “Rural,” 110-11.

60 and assassinate popular shantytown leaders? Among Sendero’s rivals and critics in Peru, as historian Steve Stern suggests, the “desire to draw moral distance encouraged depictions of Sendero as a freakish evil force outside the main contours of Peruvian social and political history – more an invention of evil masterminds and an expression, perhaps, of the particular regional milieu than a logical culmination or byproduct of Peruvian history.”180 The few – largely American – efforts that sought to explain Sendero by defining the general structural conditions out of which Sendero emerged have been rejected by many scholars, giving way instead to a kind of analytical “exceptionalism” in explaining the insurgency – a discourse with a particular focus on the unique characteristics of Sendero as an organization that seeks to explain the anomalies of its existence: a brutally violent Maoist insurgency breaking out in 1980 just as Peru returned to democratic government after more than a decade of military rule; an insurgency that gradually appeared to grow more powerful and deadly through the 1980s as other Marxist movements around the world were dissolving.

On the surface, “Sendero exceptionalism” makes sense in the context of the actions of the rest of Peru’s radical leftist parties in the late 1970s and early 1980s. The return to democratic elections in 1980 brought out virtually the entire left of Peru’s political spectrum to actively campaign for the presidency, with many prominent radical leftists publicly renouncing armed struggle and pledging to join the democratic process. Among the leading presidential contenders on the ‘formerly’ radical left were prominent peasant leaders from the Sierra movements of the early 1960s, including Genaro Ledesma, who led the first peasant land invasions in Cerro de Pasco, and Hugo Blanco, who gained world-wide notoriety as leader of the peasant movement in La Convención valley. Blanco, whose ‘Tierra o Muerte’181 slogan came to symbolize the mobilizations of the early 1960s, would go on to win the most votes among the far-left parties in the 1980 election (although together the far-left still managed only 14% of the total vote). Nevertheless, peasants and their leaders had a place on the national stage in the 1980 elections, and the left appeared ascendant in Peruvian politics. A far left coalition won Lima’s municipal

180 Steve J. Stern, “Beyond Enigma: An Agenda for Interpreting Shining Path and Peru, 1980-1995,” in Steve J. Stern, ed., Shining and Other Paths: War and Society in Peru, 1980-1995 (Durham, N.C: Duke U.P., 1995): 2. 181 Translated: ‘land or death’. See Hugo Blanco, Land or Death: The Peasant Struggle in Peru (New York: Pathfinder Press, 1972). For an interesting recent interview with Blanco, see: https://www.sudamericarural.org/index.php/noticias/que-pasa/11-peru/3577-peru-entrevista-a-hugo-blanco-tierra-o- muerte, accessed 12 July, 2019.

61 election in 1983, while Alan Garcia led the newly invigorated leftists of the American Popular Revolutionary Alliance (APRA) – Peru’s oldest socialist party – to capture the presidency in 1985.182 So, Sendero’s decision at this precise moment in history to begin the armed struggle reinforced the appearance among many analysts that this was a movement operating outside of history – that Sendero was “rowing against the stream of history, and actually making progress against the current.”183 Such thinking encouraged analysts to explore whether there was something unique driving Sendero Luminoso apart from the range of factors influencing the rest of Peru’s left. It led many analysts to exclusively focus on Sendero’s organizational structure and mobilization, and to ignore or downplay the impact of grievances and changes in opportunities for rebellion.

The consequence of viewing the insurrection as exceptional has been an overly narrow focus in much of the research trying to make sense of Sendero. This exceptionalist discourse artificially limits the search for the root causes of Sendero’s growth – it allows analysts to side-step the issue of whether there were background factors that set the stage for Sendero’s uprising, including whether these background factors were having a detrimental impact on peasants. This discourse also downplays the causal power of grievances or structural factors in causing the insurgency – as the following section on debates about Sendero’s origins and peasant support for Sendero discusses. The exceptionalist interpretation of the rise of Sendero could explain away the need to explore the sources and impacts of grievances in Peru, including whether environmental stress and demographic change was a factor contributing to the insurrection’s outbreak, and it severs the historical and analytical link to the unrest and mobilizations of the 1950s and 1960s.

2.4.4. Comprehensive theories of the origins of Sendero’s revolt

A few years after Sendero began armed struggle in Peru, American political scientists Cynthia McClintock and David Scott Palmer each released influential studies that sought to explain the origins of Sendero. Both argued that a combination of factors including national structural changes and particular contextual developments affecting peasants in Peru, especially in

182 Stern, “Beyond,” 3. 183 Gorriti, Shining, xvi.

Ayacucho, helped to lay the groundwork for Sendero to emerge and fostered support among peasants for violent insurrection. Their attempts to explain Sendero in the context of theories and causes of revolutions opened a rancorous debate among scholars about Sendero’s origins and the degree of peasant support for the insurgency. McClintock’s analysis pointed to the impacts of an ineffectual agrarian reform for highland peasants, a particularly severe economic decline in the 1970s, a subsistence crisis at the same time among rural peasants and the urban poor, the presence of a well-trained, well-organized, and cohesively led revolutionary organization in the form of Sendero, and the declining capacity of the Peruvian state as a result of the economic crisis in the country.184 She argues that support for Sendero in the early 1980s was “substantial among southern highland peasants,” among residents of the town of Huamanga, and in the department of Ayacucho generally.185 By 1991, one year before the capture of much of its leadership, she estimates that popular support for Sendero had grown to about 15 percent nationally.186 McClintock concludes that Sendero’s insurgency supports the hypotheses of several scholars of revolution: she agrees with Scott’s focus on subsistence crises; she concurs with Paige’s concern with economic transformation; and she accepts Wickham-Crowley’s emphasis on the importance of insurgent leadership and organizational capacity, etc. Together, these factors contributed to both the development of grievances among peasants and provided opportunities for violent expressions of these grievances against the Peruvian state.187

184 Cynthia McClintock, “Why Peasants Rebel: The Case of Peru’s Sendero Luminoso,” World Politics, 37(1), 1984: 48-84. See Cynthia McClintock, Revolutionary, for her latest expression of this argument. Note, however, that she offers very little discussion in this book about whether or not Sendero was a “peasant rebellion” – no doubt because of the ferocity of criticism by anthropologists of Peru like Deborah Poole and Gerardo Renique, as discussed below in section 2.4.5. McClintock’s book is specifically focused on identifying the cause of the “expansion” of the Sendero. As she notes, “the reasons for the emergence, expansion, and outcome of the revolutionary challenge may be different;” McClintock, Revolutionary, 10, and 327, footnote 50. 185 McClintock, Revolutionary, 78. 186 McClintock, Revolutionary, 78. 187 See McClintock, Revolutionary, 159-161. McClintock places the issue of a subsistence crisis within a wider context of an economic crisis in the southern Sierra in the years preceding the Sendero insurgency. “I do not argue that economic decline was the only factor in the expansion of the Shining Path,” McClintock writes. “Peru’s economic plunge was a spark igniting dry political timbers; it provided unprecedented opportunity to a shrewd revolutionary organization and provoked new problems for a state whose legitimacy was limited in any case … [T]he argument here is that … the economic plunge in Peru was extraordinarily deep and severely affected not only peasants but also university-educated aspirants to the middle class; especially given that the correlations between the onset of misery and the emergence of Sendero are strong and that economic conditions are frequently cited by Sendero militants as a key reason for their joining the movement, it is appropriate to conclude that the plunge was the triggering variable in Peru’s revolutionary equation.” McClintock, Revolutionary, 161-2. In the on-going debates about the causes of revolutions and rural revolts – ranging from state centric views, society views, agency/structure

David Scott Palmer agreed with much of McClintock’s analysis, arguing various socio-economic and political changes at the national level combined with poverty, isolation, and the failure of reformist initiatives at the local level in the southern highlands of Ayacucho to lead to Sendero’s revolt.188 To Palmer, Sendero’s rise cannot be separated from the realities of Ayacucho at that time. He highlights several factors that impacted rural and highland livelihoods and contextual factors that facilitated the “growth and orientation” of Sendero.189 Rather than emphasizing the impact of a subsistence crisis as McClintock does, however, Palmer believes that the declining capacity of the Peruvian state to deliver on highland resident’s increasing expectations in the 1960s and 1970s resulted in a loss of legitimacy for the Peruvian state, thereby translating grievances among highlanders into mobilized Sendero revolutionaries.190

2.4.5. Criticizing comprehensive theories: Sendero exceptionalists

Critics of McClintock and Palmer, however, have vigorously disputed the extent to which the Sendero insurgency was, in fact, a rural rebellion with wide and active support of campesinos.

perspectives, cultural arguments, ethnicity views, grievance arguments, or views about material structure – McClintock’s emphasis on economic decline as a key factor surely ranks as one of her most important contributions to the field. 188 David Scott Palmer “Rebellion in Rural Peru: The Origins and Evolution of Sendero Luminoso,” Comparative Politics, No. 18, (January 1986): 127-146. In the late 1960s, Palmer served as a young Peace Corp volunteer in Ayacucho and taught at the Universidad Nacional de San Cristobal de Huamanga in Huamanga during the time that Abimael Guzman, the leader of the Shining Path, worked at the university. Palmer was both a witness and an early target of Sendero – watching first-hand as Guzman built a strong Sendero presence at Huamanga University, only to be forced out of Ayacucho later with the rest of the Peace Corp by Guzman and his Sendero cadres. Guzman’s fond memories of this incident can be found in, “‘Exclusive’ comments by Abimael Guzman,” World Affairs, 156(1), Summer 1993: 52-6. 189 David Scott Palmer, "Introduction: History, Politics, and Shining Path in Peru," in Shining Path of Peru, David Scott Palmer, ed., 2nd Edition (New York: St. Martin's, 1994): 30. Palmer expands on context factors, noting the impacts of “a provincial university at a historic moment; a poor, overwhelmingly Indian region isolated in the Andes; an inappropriately conceived and applied agrarian reform in an area of great need; expanding educational opportunities and stagnant employment prospects; the opportunity to train in China during the Cultural Revolution; and a succession of civilian governments at the center unwilling and/or unable to respond appropriately to the needs of the population at the periphery (geographical, economic, social, cultural, and political). Shining Path and its people have been very much affected by such contextual factors. These have served more to prepare the organization and its leadership to seize the initiative, however, than to explain why Maoist revolution started in Peru, or why it started where it did.” 190 Palmer “Rebellion,” 143, footnote 1. Palmer explicitly says that he finds James Davies’ J-curve theory of revolution more persuasive as an explanation of the causes of Sendero’s revolt than McClintock’s focus on a subsistence crisis among peasants. Davies’ arguments and the relative deprivation theories that his work spawned are examined in James B. Rule, Theories of Civil Violence (Berkeley: University of California Press, 1988), 200- 223.

Anthropologists have spearheaded the attack against this ‘American school of Senderologists’— with Deborah Poole and Gerardo Renique leading the charge. They argue that McClintock

fails to examine carefully what might be meant by the term ‘support’ in the Peruvian case and that she instead assumes the de facto existence of a generalised and historically non-specific peasant support network … Rather than considering the specific political structuring and raison d’être of the PCP-SL [Sendero Luminoso] as a political military organization, McClintock explicitly structures her questions to address the literature on comparative peasant and agrarian rebellions and revolutions. The PCP-SL’s relevance to this literature, and therefore its status as a ‘peasant rebellion’, is thereby presented as an a priori assumption rather than as a historical or sociological problem which must be first proven, and then addressed.191

Palmer is also guilty of transforming Sendero from an insurgency with its main base of support in the city of Huamanga, to one based on rural support, according to critics. While Palmer “admits the mestizo and vanguard nature of Sendero as a political party, and the importance of teachers,” the critics note, they dispute his claims of “a ‘logical’ drifting of Sendero towards a sympathetic organic relationship with the ‘peripheral’ peasantry.”192 Poole and Renique instead concludes that “the parties to the violence [had] at best only tenuous connections to the indigenous peasantries whom anthropologists and historians had taken to be the principal actors in Andean rural history and culture.” In fact, “Andean peasants and their community and political organizations have been the principal victims of both Sendero’s war and the counterinsurgency campaigns waged by the Peruvian armed forces,” according to these critics.193

By narrowing the analytical focus to what one might describe as Sendero’s unique identity, Poole and Renique raise the important question about whether or not the rebellion had anything to do with those who were not actual members of Sendero Luminoso. If Sendero’s uprising could be solely explained by the nature of the organization and its members, then perhaps it is

191 Deborah Poole and Gerardo Renique, “The New Chroniclers of Peru: US Scholars and the ‘Shining Path’ of Peasant Rebellion,” Bulletin of Latin American Research, 10(2), 1991: 140. 192 Poole and Renique, “New,” 157. 193 Deborah Poole, “Anthropological Perspectives on Violence and Culture – a View from the Peruvian High Provinces,” in Deborah Poole, ed., Unruly Order: Violence, Power, and Cultural Identity in the High Provinces of Southern Peru (Boulder, CO: Westview Press, 1994): 4.

65 irrelevant to even consider whether peasants supported Sendero or whether urban dwellers supported Sendero. Following from this, then, it would be irrelevant to try to explain the Sendero insurgency by focusing on the subsistence condition of the peasantry, or the impacts of economic transformation in the highlands and cities, unless those factors somehow enabled the mobilization of the guerrilla organization itself. Theories of peasant rebellion would then largely cease to be of any relevance in trying to make sense of the events in Peru.

The foundations of this perspective rest on the anthropological research of Carlos Ivan Degregori and journalist Gustavo Gorriti. In denying linkages to and origins within the peasantry, the emphasis in this group of explanations about the rise of Sendero instead shifts to the particular organizational and ideological elements of Sendero which enabled them to emerge from competing leftist groups in the city of Huamanga in the late 1960s and the early 1970s. In spite of limited support at this time in Ayacucho, Degregori argues that Sendero’s “ideological rigidity and organic cohesiveness” enabled it to become “a sort of dwarf star – the kind in which matter gets so compressed it acquires a great specific weight disproportionate to its size.”194 Sendero rejected a leading role for the masses “in favor of the leading role of the party; the party decide[d] everything.”195 Regional economic crisis, a backward semi-feudal department still controlled by provincial elites, a general yearning among people in Ayacucho for progress stifled by the central government’s indifference to developing the region – all these factors played a part in forming the trajectory of the insurgency, but they were clearly of secondary importance to the unique characteristics of the Sendero movement, according to Degregori.196 Instead of a broad- based peasant rebellion, Sendero’s insurgency was begun by a party made up of “teachers and university professors and students with little influence among the regional peasantry.” It expanded through the incorporation of Ayacucho’s youth – “a significant number of rural youth with secondary school education, or in some instances no more than a primary-school education, who swelled the party ranks and constituted the most active sector of Shining Path’s rural

194 Carlos Ivan Degregori, “Return to the Past,” in Palmer, ed., The Shining Path of Peru, 53. “For that reason nobody detected [Sendero] in 1980,” Degregori writes, “nobody noticed that despite its small size, it had the power to affect decisively the Peruvian political scene of the 1980s.” Gorriti’s book is almost exclusively focused on providing a detailed narrative of the ideological rigidity and organic cohesiveness of Sendero, and linking these elements to its violent activities in the 1980s. 195 Degregori, “Return,” 55. 196 Degregori, “Return,” 55-57.

‘generated organizations.’”197 Their receptivity to Sendero’s message did not come from grievances based on poverty or material hardship. Instead, “power seduced these secondary students.”198 This view has come to dominate informed opinion about the origins of Sendero to such an extent – especially among Peruvian experts – that discussions about other potentially important factors explaining the rise of Sendero, or whether peasants may have been initially receptive to its message, are marginalized. Degregori’s research has thus served to anchor the exceptionalist arguments about Sendero.

In more recent writing, Palmer now accepts much of Degregori’s arguments, emphasizing the particular causal role of Sendero’s strong leadership and its ideological and organizational strength in Ayacucho as key to its success. In Sendero’s case, revolution is leadership, defining reality, setting the terms of the combat and pushing for a response that justifies and legitimates the definition rather than the other way around. Shining Path defines the setting in ways that make revolution the only possible outcome, and then wages revolution on its own terms. Thus, the insurgency itself becomes the independent variable, not dependent on social, economic, or external factors.199

However, Palmer does not push the exceptionalist argument so far as to ignore other “contextual” factors to explain the rise of Sendero’s insurgency. While he concludes that “Sendero is derived from the university, not from the peasantry,” he does acknowledge that a convergence of interests existed between Sendero’s militants and elements of the peasantry and urban poor in the southern Sierra.200 Those peasants who did support Sendero made a rational decision that going along with the Shining Path offered them their best option at the time. “Peasants individually and collectively are rational actors who operate in a context in which their options are extremely limited and in which they tend to be the subordinates in most role relationships. They will accept and work with whoever seems to provide them the best available

197 Carlos Ivan Degregori, “Harvesting Storms: Peasant Rondas and the Defeat of Sendero Luminoso in Ayacucho,” in Steve Stern ed., Shining and Other Paths, 129. 198 Degregori, “Harvesting,” 130. 199 Palmer, "Introduction,” 30. Emphasis in original. 200 David Scott Palmer, “Conclusion: The View from the Windows,” in Shining Path of Peru, David Scott Palmer ed., 2nd edition (New York: St. Martin’s Press, 1994): 262.

67 options at the time.”201 Thus, while McClintock’s characterization of Sendero as a peasant movement naturally leads her to study those grievances factors that might contribute to peasant radicalization, Palmer’s recognition of the importance of differing identities between Sendero militants, peasants, and the urban poor suggests we must be very careful to find the particular set of causes that motivated the different actors in the insurgency. A convergence of interests by peasants not invested in Sendero but willing to tolerate their actions against the state could be important for explaining the growth and spread of the insurgency. McClintock has come to a similar position in her later writing.202 Peru’s economic crisis in the 1970s exacerbated the highland subsistence crisis among peasants and destroyed the economic aspirations and well- being of many in the country, while undermining, at the same time, the ability of the government to deliver on people’s needs. To McClintock, these structural factors provided an opportunity for Sendero to expand its support among those disaffected by these changes, including many aspiring urban professionals only one generation removed from their peasant roots.203 The factors that helped to create a subsistence and economic crisis in Peru are thus key for understanding the growth, expansion, and possible success of Sendero’s revolt.

Anthropological reactions to the work of McClintock and Palmer object to their use of sweeping political theories and processes to link the peasants and the state of the peasantry in the Southern Sierra with the Sendero insurgency. True to their disciplines, they stress the importance of regional and local contexts for Sendero-peasant interactions, emphasizing that accurate analysis of peasant support for Sendero must attempt to distinguish between “active support and passive support.”204 However, questions about whether or not peasants were receptive to Sendero, how receptive they were, how different strata of the peasantry responded to Sendero, whether they actively or passively assisted them, and why peasants may have supported Sendero in some way, are all highly contentious among scholars.205 There is, in fact, evidence suggesting that scholars

201 Palmer, “Conclusion,” 266. 202 McClintock, Revolutionary, 11. 203 McClintock, Revolutionary, 14-5. 204 Poole and Renique, “New,” 169. 205 These are very challenging issues to resolve, according to one well known scholar of Peru. “The question … about degree of support by various peasant strata is extremely difficult to answer. Even the question of initial "support" versus initial "tolerance" is difficult to disentangle (even if you control for timing and regional context of Sendero's initial appearance in a given locale in a war context), since the effect of kin relations with young people

68 such as Degregori, Poole, and Renique perhaps overstate both the degree of separation of Sendero from the peasantry, while also understating the degree of sympathy and tolerance among peasants for Sendero in the early years of the insurgency in the southern Sierra.206 Consequently, the question about what was actually going on in the lives of peasants in the southern Sierra to make them receptive, or at least tolerant, of Sendero’s efforts to organize an armed insurrection remains open, and is one of the key issues explored below in section 8.3.

The state of the peasantry 2.5.1. Ecological and demographic Influences

There is, in fact, research that supports the work of McClintock and Palmer and offers additional insights into the question of the state of the peasantry in the southern Sierra. This work suggests that it is premature to cast aside McClintock’s focus on a crisis in peasant livelihoods in the years leading up to the Sendero insurgency. William Mitchell’s detailed study of the Ayacucho district of Quinoa from the late 1960s to the 1980s, for example, found that ecological constraints combined with population pressure and economic change to force increasing numbers of Quinoa’s peasants to either migrate in search of non-agricultural sources of income or live with ever-increasing poverty and destitution.207 These factors combined, he concluded, to create “a context in which many young people [had] little to lose in violently opposing the established order.”208 In the late 1960s, Antonio Diaz Martinez similarly noted the poverty and desperation of Ayacucho’s peasantry, who chose in large numbers to search for a living elsewhere than

sympathetic to Sendero, along with the effect of intimidation, makes it difficult at times to distinguish support from tolerance.” Steve J. Stern, “Environmental Scarcities and Conflict in Peru,” Personal e-mail to Tom Deligiannis, 30 July 1999. 206 Orin Starn, "New Literature on Peru's Sendero Luminoso," in Latin American Research Review 27, 2 (1992): 215 and 223; Ronald H. Berg, "Sendero Luminoso and the Peasantry of Andahuaylas," Journal of Interamerican Studies and World Affairs 28, 4 (Winter 1986-87): 92; Ton and Gianotten de Wit, Vera, "The Center's Multiple Failures," in Shining Path of Peru, David Scott Palmer, ed., 2nd Edition (New York: St. Martin's, 1994): 70-72. See also, Orin Starn, “Missing the Revolution: Anthropologists and the War in Peru,” Cultural Anthropology 6(1), 1991: 63-91. 207 William P. Mitchell, Peasants on the Edge: Crop, Cult, and Crisis in the Andes (Austin, TX: University of Texas Press, 1991): 22-24. Quinoa is located in the province of Huamanga. 208 Mitchell, Peasants, 24.

69 attempt to scrape a living from lands that were “eroded, poorly irrigated, extremely divided … tired and deforested.”209

Similar insights led another anthropologist, R.F. Watters, to argue that over-population and ecological stress form part of the basic constraints that have hindered Peru’s peasants from improving their livelihoods over the past 50 years, along with class and ethnic barriers, the short-comings of traditional agricultural technologies, unequal terms of trade with urban areas, and unfavorable macro-economic policies. For Watters, these constraints form a “Jack-in- the-box” model of Peru’s peasantry. Only the “most innovative and enterprising peasants can hope to improve their situation … by breaking out of the confining box of [these] limiting circumstances.”210 The implication of Watters’ Figure 5: Watters’ Model of model and of Mitchell’s insights are that the Peasantry ecological stress and demographic change increasingly challenged peasant livelihoods in the decades leading up to Sendero uprising, possibly dashing expectation of improved livelihoods. Assessing changes in environmental scarcity is thus important for assessing the state of the peasantry in the southern Sierra.

Several scholars also point to the impact of changes in agricultural tenure and rural structure from the military government’s agrarian reform program as an important cause of rural unrest – in Peru generally, and in the southern Sierra in particular. In some areas, the impacts of agrarian reform also interacted with economic transformation and modernizing influences to alter rural

209 Quoted in Starn, "Missing," 80. “The coast, the mines of Cerro de Pasco, the jungle of Apurimac serve them as an escape from the poverty of the land,” Diaz Martinez wrote, giving them some temporary or permanent work and a bit of economic income. After the planting, they go to these centers of work and then return for the harvests, bringing with them a few clothes and a little money saved for the family that stayed to take care of the house and the fields. Others emigrate for good, taking their family with them and leaving their small plot to a relative. Sometimes they come back for the fiestas, or don't come back ever.” 210 Watters, Poverty, 323.

70 livelihoods and open new fissures in rural society. Agrarian reform may have increased peasant differentiation in the Sierra, creating new groups of winners while embittering many other peasants against both the state and those who benefited from the reforms.

The process of peasant differentiation was, in fact, taking place in many parts of the Sierra long before the military government’s agrarian reform as capitalist economic transformation penetrated the highlands and increased inequalities between peasants. As Dianna Deere’s research in the north-central Sierra Department of Cajamarca illustrates, rural class structures began to change significantly with the agricultural transformation of the 1950s and 1960s that accompanied the establishment of a dairy industry.211 Communal land holding structures gave way to private holdings. Many hacienda owners transformed themselves into private dairy producers, working the best valley land holdings intensively while selling off their marginal hillside parcels to peasants and former hacienda peons. As a result, “the independent peasant sector grew in population and proportion of land owned but also in internal differentiation.”212 A new group of middle-level peasants were increasingly important. Many earned higher wages by participating in the dairy boom, and increased their personal land holdings by buying land. However, inequalities also increased among the peasant sector as land holdings and income levels among the poorest groups either remained stagnant or declined by the early 1970s. The landless or near landless smallholders (with holdings of barely over 1 hectare) accounted for the bulk of the farming population in Deere’s study area, but their incomes in the early 1970s “were lower than the minimum estimated income from the poorest peon households on the haciendas in 1917.”213

The increasing inequalities in incomes and land-holdings that accompanied private reform in Cajamarca was a typical consequence of the process of peasant differentiation taking place in other parts Peru. Peasant differentiation in the southern Sierra similarly increased inequalities in wealth between sectors of the peasantry. However, in many parts of the southern Sierra, scholars argue that it was agrarian reform and the somewhat later penetration of capitalist economic

211 Carmen Diana Deere and Alain de Janvry, “Demographic and Social Differentiation Among Northern Peruvian Peasants,” Journal of Peasant Studies 8(3), 1981: 364-5. 212 Enrique Mayer, “Patterns of Violence in the Andes,” Latin American Research Review 29(2), 1994: 151-3. 213 Mayer, “Patterns,”155-6.

71 relationships that drove the process of peasant differentiation. These trends appear to have interacted with the impacts of limited ecological resources to amplify existing inequalities in peasant livelihoods.

In Ayacucho, competition for land among the peasant communities has long been fierce, partly because of the poor quality of land in the department.214 With limited quantities of good land, few large haciendas existed in Ayacucho before the military’s agrarian reform. In all of Peru, the government expropriated 1,493 haciendas with a total area of over 7.5 million hectares during the 1970s and 1980s. Government officials, however, had long known that the need for land far outstripped available land. As far back as 1966, the CIDA report pointed out that there was not enough land to fulfill land reform goals, and that only a fraction of those who needed land would be able to receive land under comprehensive land reform.215 This turned out to be true in Ayacucho where the total expropriated during the agrarian reform period amounted to only about 325,000H – about 22% of total agricultural land in the department, benefiting about 18,000 rural households.216 Consequently, the little land that was available for redistribution with agrarian reform was bitterly contested. The dislocations of agrarian reform also provoked many peasant communities to renew bitter feuds with each other over long-standing boundary issues.217 As well, in parts of Ayacucho, anthropologist Billie Jean Isbell reports, agrarian reform opened the door for a new class of mestizo or aspiring mestizos to seize the financial rewards of the reform process, and slip into old patterns of exploitation of the peasant majority. “The same relations of exploitation,” she writes, “were reproduced by the new class of rich peasants, the new state bureaucrats who replaced the old officials, and on ex-haciendas by the administrators of the new

214 de Wit and Gianotten, " Center's," 66. 215 Mayer, Ugly Stories, 19. In the highlands, where need far outstripped land availability, CIDA economists predicted that only 4% of those needing sufficient land would be satisfied by land reform, benefitting about 150,000 families, but leaving 700,000 families unfulfilled. In the end about twice that number of families in the highlands received land from agrarian reform. 216 Comisión de la Verdad y Reconciliacón: Informe Final (Truth and Reconciliation Commission: Final Report), Tomo 4 (Volume 4), August 2003, Lima: 19-20. Accessed online, 21 September, 2003 . Hereafter referred to as (CVR). Total area of agricultural land is taken from Degragori, Ayacucho, 179. 217 CRV, Inform, 66-67.

72 state cooperatives” that were set up on the old hacienda lands by the agrarian reform officials.218 This group was often best placed to take advantage of the opportunities that arose with land reform, and they tended to make up a majority of the newly prospering class of merchants and shop keepers that arose in many towns in Ayacucho.219 At their worst, they were recreating patterns of abusive relationships with poorer peasants in the style of the old gamonal mestizo elites.

The process of land distribution that accompanied agrarian reform was also stimulating differentiation among peasants in the southern Sierra, as R. F. Watters demonstrates in his study of Chilca, in the Cuzco district of Anta. Land distribution was highly controversial among the peasants of the community. Community leaders and the administrators of the cooperatives set up by the government’s agrarian reform program used their positions to get more land out of the reform process than other community members.

A poor man alleged that Aurelio, the President and a former administrator of the co-operative (already one of the richest men in the community), had given help to widows in distress in return for their making their land over to him at death, a practice which was also observed in Qolquepata. The powerful people, it was said, allied with ‘the good people’ and not with the poor … it was widely accepted that land distribution had been carried out in an unequal way.220

These inequalities were reflected in Watters’ survey of incomes between 1964 and 1979, where the distinctions between rich, middle and poor peasants had become “more pronounced and the gaps between groups wider.”221 Agrarian reform in Chilca was also accompanied by increasing capitalist agricultural penetration, as more and more peasants began producing for the region’s agricultural markets. As with land distribution, the richer sectors of the peasantry were best placed to take advantage of the opening markets, further contributing to the widening inequalities in the community.

218 Billie Jean Isbell, “Shining Path and Peasant Responses in Rural Ayacucho,” in Palmer ed., Shining Path of Peru, 2nd Edition (New York: St. Martin's Press, 1994): 80-1. 219 Isbell, “Shining Path and Peasant Responses,” 82. 220 Watters, Poverty, 309. 221 Watters, Poverty, 135.

Various scholars agree that these changes and the unfulfilled expectations of agrarian reform stoked bitterness among much of the peasantry in the southern Sierra. The experiences of the peasants in the Andahuaylas region of the Department of Apurimac are similarly illustrative. In the early 1970s, delays in implementation of the agrarian reform were jeopardizing efforts to eliminate the area’s haciendas. Comuneros and hacienda peons grew increasingly angry as they watched former hacendados (hacienda owners) take advantage of the confusion of pending agrarian reforms to decapitalize their estates, selling or taking with them anything of value.222 Their anger grew as the agrarian reform authorities frustrated their desire to parcel out the haciendas to the peons and land-hungry comuneros who invaded many of the disputed estates in 1974. Instead, Peru’s military brutally evicted the peasants and arrested many of their leaders.223 As in other parts of the Sierra, the state rejected parcelization of the former hacienda lands in favor of massive cooperatives that were often run by outside bureaucrats or corrupt local officials. The frequent mismanagement of these cooperatives further alienated many peasants because some of the cooperative managers used their positions of power to increase their personal land holdings or to steal from the cooperative’s coffers.224 Lastly, as in Ayacucho and Cuzco, the expansion of capitalist agricultural activities that accompanied the agrarian reform was frequently manipulated, to the detriment of the majority of peasants, by a small number of middle-income and rich peasants. This further aggravated existing social tensions as differences in wealth became increasingly polarized.225

The consequence of a botched and derailed agrarian reform in many areas of the southern Sierra like Andahuaylas was that it did little to improve livelihoods of many peasants, alienating much of the peasantry and turning many of them against the state, according to several scholars.226 In some cases, the failures of agrarian reform also convinced peasant leaders that the impossibility of looking to the state for deep and lasting change meant that revolutionary war the only option

222 Berg, "Sendero,” 172; and Florencia E. Mallon, "Chronicle of a Path Foretold? Velasco's Revolution, Vanguardia Revolutionaria, and "Shining Omens" in the Indigenous Communities of Andahuaylas," in Shining and Other Paths: War and Society in Peru, 1980-1995, Steve J. Stern ed. (Durham, NC: Duke University Press, 1998): 95; Mayer, Ugly Stories, 24. 223 See Malon, “Chronicle,” 97-109. 224 Berg, “Sendero,” 172-3. 225 Berg, “Sendero,” 173-4. 226 Mayer, “Patterns,” 153; Isbell, “Shining Path and Peasant Responses,” 82; Berg, “Sendero,” 174.

74 left. As Andahuaylas peasant leader Lino Quintanilla concluded in the months following the military crackdown,

[T]he peasantry’s agrarian problem and the problems afflicting the Peruvian people will only be resolved through revolutionary war, destroying the power of the semi-colonial bourgeois State and constructing another power on the ruins … 227

In the end, agrarian reform appears to have done little to ease the agricultural pressure on the peasantry in the southern Sierra. Although a significant amount of land was distributed through agrarian reform, “the vast majority of peasants in highland Peru still [had] too little land to make a living.”228 Investigating the local consequences of economic change and Agrarian Reform, and its impact on peasant livelihoods in light of changes in the 20th century, appears to be crucial for understanding why some highland peasants were willing to support or tolerate Sendero in the early 1980s. Impacts from environmental scarcities may have combined with these processes to aggravate grievances or motivations among some peasants to support challenger groups against the Peruvian state like Sendero.

Research focus and research design

2.6.1. Key questions

To explore whether human pressure on the natural environment combined with factors and processes discussed above to help cause Sendero’s uprising, this thesis examines whether environmental and demographic change played a role in generating Peru’s rural unrest and the type of causal role played by environmental scarcities. In no way does this thesis assume that environmental scarcity and demographic change played a necessary or sufficient role in causing rural unrest in Peru’s Sierra. Working within a framework of complex causation, scarcities instead act as an INUS condition to interact with other conditions.229 The theoretical literature

227 Quoted in Mallon, “Chronicle,” 112. 228 Mayer, “Patterns,” 155; Noting the inadequacies of Agrarian Reform to solve the land problem in Peru, William Stein writes: “It was an agrarian “reform” which built the foundation for the terrors of the 1980s and 1990s.” William W. Stein, Deconstructing Development Discourse in Peru: A Meta-Ethnography of the Modernity Project at Vicos (Lanham, MD: University Press of America, 2003), 471-472. 229 My methodological explanation for the causal role of environmental scarcities is founded on the INUS approach to understanding complex causality. Due to space constraints I have not provided a full explanation of my framework for understanding causality. Instead, this is explored in detail, along with additional methodological

75 on rural revolts surveyed in this chapter suggests that scarcities and demographic change may have played a role in affecting the subsistence security of the peasantry, thereby helping to motivate peasants to support rural insurgencies as revolution theorists suggest, particularly in the context of differential economic transformation of the highlands which led to some peasants getting richer, while others were left behind. Since environmental scarcities and demographic change, in combination with other political-economic factors, can affect the popular living standards of peasants, they are conceivably part of the set of INUS conditions that gave rise to this period of revolutionary change in Peru. Unfortunately, revolution theorists offer few specific insights into the causal role of environmental scarcities. Homer-Dixon and others who have explored environment-conflict linkages have suggested specific ways in which environmental stress and demographic change may cause violent conflicts such as insurgencies and group conflict. These theoretical hypotheses suggest specific avenues for researching rural unrest in Peru’s southern Sierra that are reflected in the study below, examining the impact of supply and demand scarcities and the impact of structural scarcities in the D. of Chuschi.

The question of productivity constraints is of central importance in this dissertation, attempting to analyze of the impact of resource constraints on peasant subsistence capabilities in our local case study. Land availability in the highlands has been a long-time grievance of peasants. Peasant access to productive resources like land is intimately connected to the question of the state of the peasantry. The reasons for the scarcity of productive resources in the Sierra vary considerably depending on the specific region studied. In some cases, as noted above in section 2.5, long-term trends like modernization and the incorporation of the peasantry into the modern capitalist economy appear to have greatly affected scarcities for the peasantry as rural elites took advantage of the changes to seize peasant lands. As well, observers have argued that these trends increased differentiation among the peasantry and widened wealth differentials among sectors as certain groups of peasants benefited disproportionately from others, and gained more access to production resources. These are enduring trends in Andean history, Steve Stern has suggested, having contributed to peasant unrest and mobilization for decades.230 Some

concerns of critics of the Toronto Group’s approach, in the co-authored paper with Daniel Schwartz and Thomas Homer-Dixon. See Schwartz, Deligiannis, and Homer-Dixon, “The Environment and Violent Conflict.” 230 Steve J. Stern, “New Approaches to the Study of Peasant Rebellion and Consciousness: Implications of the Andean Experience,” in Steve J. Stern ed., Resistance, Rebellion, and Consciousness in the Andean Peasant World:

76 scholars have suggested that peasants lacked access to adequate land resources to make a living during this period.231 Land reform in Peru’s Sierra does not appear to have alleviated this issue in many places, as noted above in section 2.5.1. In fact, Agrarian Reform appears to have provoked new grievances around land, adding another layer to the accumulated grievances surrounding access to land resources. Obviously, the highly variable nature of local productive resources in Peru means that the sources of productivity constraints will be carefully disentangled in this local study.

A related issue that has received little attention from scholars is the social-physical aspects of land scarcity, like the degradation of land resources in the Sierra as a result of erosion, salinization, or unsustainable use. Unfortunately, there is generally little data available on the issue of soil management and soil loss in Peru’s Sierra.232 This study attempts what few scholars have attempted – to explore the relationship between social-physical aspects of scarcity in the Sierra and the question of peasant subsistence crises and rural unrest at the local level. These supply-side changes to environmental resources will also be examined together with the impacts, if any, of demand pressures from consumption and population change. The issue of population growth in the Sierra over the past half century has been noted above by a number of scholars as a key aspect of rural change in the Sierra. But its precise role is controversial and relatively under- explored in local studies in Peru. Scholars acknowledge that the population of the Sierra recovered to its pre-Conquest levels in the years following WWII, but with reduced access to pre-Conquest land holdings. Some like William Mitchell have suggested that growing highland populations and inheritance practices were also reducing per capita land availability, and contributing to the large shift from rural to urban areas across Peru, patterns that he documented in the District of Quinoa in northern Ayacucho.233

18th to 20th Centuries (Madison, Wisconsin: The University of Wisconsin Press, 1987): 5. Stern wisely councils that “political analysis of agrarian movements requires explicit attention to internal differentiation among the peasantry.” 231 Mitchell, Peasants. 232 Stephen B. Brush, “Diversity and Change in Andean Agriculture,” in Lands at Risk in the Third World: Local- Level Perspectives, Peter D. Little, Michael M Horowitz, with A. Endre Nyerges eds., (Boulder, CO: Westview Press, 1987. 233 Mitchell, Peasants.

Complicating the exploration of the population issue is the fact that rural populations have fallen significantly in many parts of the Sierra over the past 50 to 60 years as increasing numbers of Peruvians have migrated to urban areas in search of employment, educational opportunities, etc.234 In fact, the rural population “problem” in the eyes of some Peruvian experts stems from the lack of people in rural areas, which some claim is now hindering agricultural production and leading to land degradation.235 Sendero’s insurgency in the south-central Sierra exacerbated the exodus of peasants to urban areas as they sought sanctuary from violence.236 Many peasants have also engaged in temporary or seasonal migration from rural to urban areas during this period, finding ways to link livelihoods in their rural communities to their lives in urban centres.237 With these complex trends in mind, this study attempts to make sense of the impacts of the longer-term processes and interactions that preceded the current reality in Peru’s highlands, especially the key period when rural population growth began to level off and decline in the southern Sierra. Departments in Peru’s southern Sierra like Ayacucho have had some the country’s highest rates of rural-urban migration over the past half century.238 Like the issue of land availability, however, the question of population growth appears to vary regionally in the Sierra, so this thesis focuses on specific local developments in the D. Chuschi.

Lastly, the literature review and case rationale in chapter 1 and 2 suggest that any study of the impact of environmental scarcity on peasant livelihoods remain flexible on the issue of temporal and analytical scale. Although the conventional wisdom for research design suggests stating the level of analysis used in the study, the overview above suggests that any complete causal explanation for rural unrest in Peru will encompass a complex interaction of macro and micro

234 Jane Collins, Unseasonal Migrations: The Effects of Rural Labor Scarcity in Peru (Princeton, NJ: Princeton U.P., 1989). 235 See Moshe Inbar and Carlos A. Llerena, “Erosion Processes in High Mountain Agricultural Terraces in Peru,” Mountain Research and Development, 29(1), Feb. 2000: 72-79. 236 Between 1981 and 1993, Ayacucho was the only department in Peru to register a negative rate of population growth (-0.2%). CVR, Inform, Tomo 4, 17. 237 See Karston Paerregaard, Linking Separate Worlds: Urban Migrants and Rural Lives in Peru. Berg (Oxford: Berg, 1997); and Andrea Milan and Raul Ho, "Livelihood and Migration Patterns at Different Altitudes in the Central Highlands of Peru," Climate and Development 6, no. 1 (2014), 69-76. 238 Mayer, "Patterns of Violence in the Andes," 155-6. The recent Truth and Reconciliation Commission Final Report notes that Ayacuchanos formed the second largest migrant community in Lima’s shanty-towns, according to Peru’s 1981 census. Of course, the reasons for rural-urban migration cannot be solely attributed to the impact of environmental scarcities. See CVR, Inform, Volume 4, 17.

78 influences, ranging across different levels of analysis. Paige and Klaren, for example, note the effects of international markets on the development of Peru’s agro-export industry, and the resultant impact on peasants in the Sierra. At various times in the 19th and 20th century, increasing global demand for wool stimulated elite capture of grazing lands from southern Sierra communities for export markets. Similarly, Deere and Watters note the regional influence of newly emerging dairy markets on peasant differentiation in Cajamarca and Cuzco. While this study will explore environmental scarcities at the district and provincial level in Ayacucho, the study also is sensitive to those causal impacts on scarcities that stretch to regional, national, or international levels. As well, while the dissertation is primarily interested in events in Peru’s Sierra from the 1940s to the 1980s, the review above suggests that the study remain temporally flexible in discerning causal influences. In some cases, with demographic change or land scarcity in the Sierra, the time-scales for causal mechanisms may operate over many decades. Clearly, we cannot rule out the long view when disentangling the impacts of environmental scarcities. Stern similarly notes the importance of long-term time frames when studying Andean unrest, suggesting that we “must look at multiple time frames simultaneously – relatively short time frames (“conjunctural” and “episodic”) to understand the recent changes that make rebellion or insurrection more likely and possible, and to appreciate dynamic changes that emerge during the course of violent conflicts; and longer time frames spanning centuries to understand the historic injustices, memories, and strategies that shape goals, consciousness, and tactics of rebels.”239 While the particular focus of the dissertation revolves around the unrest that shook the Sierra from the 1950s to the 1980s, this study follows the causal influences to their source where necessary.

239 Stern, “New Approaches,” 11. Elaborating on the importance of long-term frames of reference in studying Andean peasant rebellion, Stern notes that “the precise definition of the relevant long-term frame of reference will depend on the particular case at hand, but it should at least include the period considered relevant in the rebels’ own historical memory, and the period during which the last enduring strategy of “resistant adaptation” was developed. It is difficult to imagine a time scale less than a century long that meets these criteria. A method which studies multiple time scales, including long-term ones, will not only explain better the causes and ideological characteristics of particular rebellions and insurrections. It will also enable the student to distinguish more clearly between genuinely new patterns of collective violence and grievance, and repetitions of historic cycles of resistance and accommodation that occasionally included some forms of collective violence.” Stern, “New Approaches,” 13.

Chapter 3 The Evolution of Environment-Conflict Research: Toward a Livelihood Framework Introduction

Building on the literature review in chapter 2, this chapter argues that progress has stalled among qualitative environment-conflict research because the level of analysis adopted by most of this work—the state level—is inadequate to capture the empirical complexity of environment- conflict links on the ground. Qualitative work on environmental change-conflict links conducted by the Toronto Group, the Bern-Zurich Group, and researchers who have followed in these traditions has unconvincingly used a state-level of analysis. Although quantitative researchers have begun to disaggregate their studies to the sub-state level, qualitative researchers have yet to do so.240 The chapter proposes a detailed household-livelihood framework of analysis for future qualitative environment-conflict research that will foster more fine-grained understanding of the complex relationships between human-induced environmental change and violent conflict. This framework is central to this dissertation’s research on the D. Chuschi because it allows us to better understand the particular impacts of local-level environmental and demographic pressures, and disentangle various structural and political influences on the livelihoods of district residents.

The chapter begins by providing a brief critique of the state-centric focus of past environment- conflict research projects. This paves the way for the development of a household-livelihood framework for future research. The final section of the chapter outlines the essentials of this framework, and offers observations about its use for environment-conflict research.

240 In the past five years, disaggregated quantitative studies of environment-conflict links have emerged. See Henrik Urdal, "Population, resources, and political violence: A Subnational Study of India, 1956–2002," Journal of Conflict Resolution 52, no. 4 (2008), 590-617; Clionadh Raleigh and Henrik Urdal, "Climate Change, Environmental Degradation and Armed Conflict," Political Geography 26, no. 6 (2007), 674-94; Halvard Buhaug, "The Future is More Than Scale: A Reply to Diehl and O'Lear," Geopolitics 12, no. 1 (2007); Lars-Erik Cederman and Kristian Skrede Gleditsch, "Introduction to Special issue on “Disaggregating Civil War”," Journal of Conflict Resolution 53, no. 4 (2009), 487-95, and the project, Disaggregating Civil Wars, http://www.icr.ethz.ch/research/ecrp, accessed 3 Oct., 2011.

Environmental change and the role of the state – mediating scarcity, undermining capacity, or actively exploiting

Both the Bern-Zurich Group and the Toronto Group recognize the intermediary role played by the state and institutions to forestall or mitigate negative social consequences before they contribute to conflict-generating processes such as grievance formation or collective mobilization.241 The lack of such “social ingenuity” interventions, they argue, can significantly increase the probability that scarcities will lead to violent conflict.242 As well, regions marginalized by the central state may lack the administrative and law-enforcing apparatus of the state and “institutions founded on the rule of law, legitimized and accepted by local actors,” Baechler notes, thus making them more vulnerable to the impacts of resource scarcity and less capable of resolving environmental conflicts.243 In other cases, well-functioning, traditional social institutions that regulate access and use of resources may be disrupted by central state actions.244 Research indicates that when the role of traditional institutions are disrupted without adequately replacing these institutions through an effective state presence, worsening scarcity and an increased potential for conflict develops—a pattern that is particularly unsettling in areas marginal to the central state.245 The consequences of a weak, ineffective, or non-existent state presence is magnified, according to Baechler, in states with a weak civil society and a lack of political pluralism.246

The Toronto Group also argues that environmental scarcities can weaken the state by directly impacting state elites, by altering their relationship with the state, or by strengthening groups of elites that can challenge the state’s power. Scarcities provide economic and political opportunities for predatory elites to capture resource rents, ignore state dictates or laws like taxation, or “to penetrate the state to make it do their bidding.”247 The consequences of predatory

241 The core of the chapter was published in: Tom Deligiannis, "The Evolution of Environment-Conflict Research: Toward a Livelihood Framework," Global Environmental Politics 12, no. 1 (2012), 78-100. 242 Homer-Dixon, Environment, 107-09; and Baechler, Violence, 101-104. 243 Baechler, Violence, 103. 244 Baechler, Violence, 101. 245 For an example of this pattern in Peru, see Paul Trawick, "Comedy and tragedy in the Andean commons," Journal of Political Ecology 9, no. 1 (2002), 35-68. See also, Elinor Ostrom, "Coping with Tragedies of the Commons," Annual Review of Political Science 2, no. 1 (1999), 493-535. 246 Baechler, Violence, 103-4. 247 Homer-Dixon, Environment, 101-102.

81 behavior by elites can provoke defensive reactions from those who rely upon resources for their survival, eroding trust in the state, worsening social segmentation like class and ethnic divisions, and making it easier for challenger groups to develop.248 Thus, for this group, scarcities may not only lead to greater immiseration, but also stoke patterns of blame among groups and break bonds of trust within states and society.

Since the publication of the findings of the Toronto Group and the Bern Zurich Group, little research has been done to refine their findings, in spite of calls from scholars to address controversial or under-explored dimensions of environment-conflict research.249 Colin Kahl’s work stands as a notable exception, making two advances. First, Kahl argues that demographic and environmental stress (DES) can aggravate and deepen the inherent insecurity of weak states, possibly triggering or aggravating the “security dilemma” between groups—where actions taken by groups to ensure their security, perhaps in the face of the impacts of demographic or environmental stress, “can set off an action-reaction spiral that leaves all parties worse off and less secure.”250 Second, Kahl argues that DES provides state elites with opportunities to “engineer and direct violence downward toward social groups.”251 These “top-down” dynamics are in contrast to the “bottom-up” dynamics of much of the Toronto Group or Bern Zurich Group’s hypotheses, where environmental scarcities, among other factors, weaken states and societies, opening “political space for social groups to direct violence upward toward the state or sideways toward one another.”252 Kahl’s “state exploitation” hypothesis recognizes that the social segmentation and increasing grievances arising from DES provides both “incentives and opportunities to instigate violence.”253 Rising grievances or growing numbers of aggrieved citizens threaten to undermine stability and control by state elites, providing an incentive to rulers to find a way to stabilize their base of support, “mobilize new supporters, and co-opt or crush political opponents” to remain in power.254

248 Homer-Dixon, Environment, 102. 249 Matthew and Dabelko, “Environment”; Gleditsch and Diehl, Environmental; Schwartz, Deligiannis, and Homer-Dixon, Environment and Violent Conflict; and Richard A Matthew, Michael Brklacich, and Bryan McDonald, "Analyzing environment, conflict, and cooperation," Understanding Environment, Conflict, and Cooperation (Nairobi: UNEP, 2004), 5-15. 250 Kahl, States, 47; and Homer-Dixon, Environment, 96. 251 Kahl, States, 12; and Peluso and Watts, Violent, 22-3. 252 Kahl, States, 12. 253 Kahl, States, 50. 254 Kahl, States, 50.

Shifting the focus from the state to household livelihoods: A household-livelihood framework for environment-conflict research

The centrality of the state to this strand of environment-conflict research is both inherent in the causal models and forms the referent object for many research efforts. The state and its stability are the essential focus. The descent of the state into anarchy, the rise of violent groups that threaten to overturn state order, or the rise of groups which threaten the integrity of the state— these form the dependent variables of the environment-conflict research done by the Toronto Group, the Bern Zurich Group, and those who have followed their approach.255 Similarly, the social effects of scarcity, and the variables identified by scholars as interacting with and mediating these impacts, are frequently explained in terms of their broad societal impacts— general agricultural decline, the increasing division of groups or elites in states, etc. Much of this research focuses almost exclusively at the state and societal level.

However, the impacts of scarcities are not inherently or exclusively felt at the state level. Scarcities initially affect individuals, families, and communities personally and directly, before being translated into broader state or societal effects. Localized immiseration or social impacts may result in little or no national impacts. Conflicts may happen at levels far below the level needed to pull a state into anarchy, or to threaten the integrity of the state or its rulers. Researchers should not assume that impacts automatically scale upwards to impact the state, though in some cases they may.

The state-level bias in case-study research leads to local processes of environmental scarcities, and their local social effects, being understudied and inadequately understood by most qualitative environment-conflict research to date.256 Scholars have not begun with an explicit focus on the livelihoods of those studied or those who are impacted by changing demographic and environmental conditions. Instead, the research offers underspecified generalizations or incomplete accounts of the impact of environmental change. There is, in fact, an unspecified black box in the causal chain of much environmental conflict research between the generation of

255 Homer-Dixon, Environment, chapter 7; Baechler, Violence, chapter 4; and Kahl, States, 30. 256 Quantitative studies of environment-conflict links faced similar problems in the past; however, in the past five years such research is increasingly disaggregated to the sub-state level. See criticisms by Shannon O'Lear and Paul F Diehl, "Not Drawn to Scale: Research on Resource and Environmental Conflict," Geopolitics 12, no. 1 (2007), 166-82; and responses by Buhaug, “The Future.”

83 environmental scarcities and the social effects that they cause. Within this black box are specific impacts on people’s livelihoods of environmental scarcities and people’s adaptations to them. These dynamics have yet to be adequately theorized.257 As a result, the models have sought to assess the aggregate impact of individual or small-scale environmental scarcity-social effects without a clear understanding of the local dynamics that generate them. They have sought to understand the sum of societal impacts without understanding the individual processes that generate them. A closer look at those livelihood processes shows a more complex picture.

The critique presented here has three important implications for the validity of existing qualitative environment-conflict research. First, the state level of analysis has led researchers to ignore the impacts of environmental scarcities on a variety of local conflicts and their implications for societal and state stability. State-focused research uses a limited set of dependent variables, emphasizing conflicts that appear to pose the greatest threat to the state. Yet this overlooks how environmental change influences the extent and consequences of small-scale local conflicts.258 These conflicts may only kill or injure handfuls of people and are rarely reported accurately. They also may be difficult to document because they are often widely dispersed and occur over many years. Over time, however, their local impacts may undermine the fabric of society, alter migration patterns, and affect social and ethnic group solidarity and cohesion in certain areas of states. The grievances spawned can stimulate social unrest, and violence from local conflicts can exacerbate or condition patterns of violence during insurgencies and civil war in ways not readily apparent to those examining the conflict from the state level.259

The patterns of interaction in many of these small-scale local conflicts can be characterized as simple scarcity conflicts involving distributional conflicts between local groups over crucially important renewable resources, worsened by supply and demand changes in resource availability.260 These simple scarcity conflicts are exemplified by conflicts among cultivators,

257 Peluso and Watts, Violent, 20 offer a similar criticism of Baechler and Homer-Dixon’s work. 258 Quantitative environment-conflict researchers have used conflict databases with a violence threshold of 25 or more battle deaths since the late 1990s. However, even this threshold probably fails to capture a significant number of local conflicts that kill or injure a handful of people at a time. 259 Discussed in more detail below in chapter 6, 7, and 8; Catherine André and Jean-Philippe Platteau, "Land relations under unbearable stress: Rwanda Caught in the Malthusian trap," Journal of Economic Behavior & Organization 34, no. 1 (1998), 1-47; and Kalyvas, Logic, 390. 260 I adapt a term here from Homer-Dixon, Environment, 106-108. Homer-Dixon’s usage applied to inter-state conflicts.

84 among fishers, or between herders and cultivators. In many cases, resource scarcities may have social effects like those identified by Homer-Dixon at the state level, but at the individual, household, or group level. This can include household economic decline or immiseration, various forms of migration, or local social segmentation. Baechler also noted the possibility for local transformational conflicts—what many also label as “modernization,” “developmental,” or “market-penetration” conflicts—where the transformation of resources use or resource exploitation shifts from one type of human-nature relationships to another type. It is possible to trace how various simple scarcity conflicts lead to violent conflict, such as changes in land use between herders and cultivators, upheaval resulting from of the introduction of export agricultural crops, conservation enclosure conflicts, or the impacts of subtle changes brought about by long-term patterns of market penetration into areas with little or no previous market- based relationships.

Second, the analytical concentration on the state level among many researchers, without an adequate understanding or examination of local processes, suggests that many of the hypotheses in qualitative environment-conflict research are built upon shaky empirical ground. Local studies were not aggregated to the state or societal level to determine social effects in state-level studies. Detailed examination of data for state or societal level impacts reveal significant uncertainties about local processes, raising questions about the validity of environment-conflict hypotheses. The Toronto Group’s study of rebellion in Chiapas, for example, marshaled a variety of data that demonstrated linkages between environmental scarcities and conflict.261 Data limitations ultimately forced the authors to rely mostly on aggregate state level data and episodic local data to prove their thesis,262 a point some critics have seized upon to question their conclusions.263 Although space limitations preclude a detailed examination of the limitations of the Chiapas case, the lack of local data in the study raises questions about the validity of the analysis. The Chiapas study merely suggests that environmental scarcities played an important role. It fails,

261 Philip Howard and Thomas F Homer-Dixon, Environmental scarcity and violent conflict: the case of Chiapas, Mexico, American Association for the Advancement of Science (Toronto, 1995). 262 Howard and Homer-Dixon, Environmental Scarcity and Violent Conflict: The Case of Chiapas, Mexico, 8 and footnote 17. 263 See the critique by Hartmann, "Will the Circle be Unbroken?"; and Aaron Bobrow-Strain, "Between a ranch and a hard place: Violence, scarcity, and meaning in Chiapas, Mexico," in Violent Environments, ed. Nancy Lee Peluso and Michael Watts (Ithaca, NY: Cornell U. P., 2001), 155-185. Both seize on data gaps in the Toronto Group’s Chiapas case. But a careful reading of Bobrow-Strain fails to refute Howard and Homer-Dixon’s 1995 analysis.

85 however, to make a convincing case supporting the hypothesized causal mechanisms. As with other cases by the Toronto Group, the Bern-Zurich Group, and scholars such as Colin Kahl, the extensive use of broad-scale or episodic data provides strong indications that scarcities cause negative social effects and contribute to social violence.264 However, data gaps and assumptions of trends weaken the analysis; the hypotheses would be more convincingly evaluated with sustained and detailed examination at lower levels of analysis. By contrast, many qualitative studies on environmental conflict by political ecologists do a better job of outlining local social effects of human pressure on the environment, due to extensive use of local field studies, ethnographic methodologies, and concerns about social justice. Although much of the political ecology tradition significantly under-theorizes processes of violent conflict,265 future qualitative work on environmental change-conflict linkages could benefit from their methodological approaches.

A third implication of the livelihood critique made here is that it may have led some environment-conflict researchers to over-predict the likelihood of environmental scarcities causing conflict. A state level of analysis underestimates the myriad ways in which local stakeholders respond to and adapt to the impacts of environmental scarcities on their livelihoods, often in ways that ameliorate negative impacts. State-level models have underestimated local level agency. Broad Neo-Malthusian accounts are particularly vulnerable to this criticism, because they are excessively linear in their presentation of how scarcities can lead to conflict and they accept as generalizable causal models that more likely describe special conditions.266 Robert Ford notes that numerous observers of agro-ecological and demographic trends in Rwanda predicted imminent collapse for over forty years, “but it never happened when and like they predicted,” although “others were able to show that considerable coping was possible and actually achieved.”267 Assumptions of linear relationships between scarcity and negative social

264 Examinations of the Bern Zurich Group’s case studies and Kahl’s cases are beyond the scope of this chapter; however, a state-frame of analysis generally prevails in their work as well. See Tor A Benjaminsen, "Does Supply- Induced Scarcity Drive Violent Conflicts in the African Sahel? The case of the Tuareg rebellion in northern Mali," Journal of Peace Research 45, no. 6 (2008), 819-36 for a critique of Baechler and Kahl. 265 Kahl, States, 25. 266 See, for example, Robert D Kaplan, "The coming anarchy," Atlantic Monthly 273, no. 2 (1994), 44-76. 267 Robert E Ford and Kim T Adamson, "The Population-Environment Nexus and Vulnerability Assessment in Africa," GeoJournal 35, no. 2 (1995), 212. Michael Thompson noted similar assumptions of impending disaster that never happened in analyses of Nepal’s environmental situation. See Michael Thompson, “Not Seeing the

86 effects made it seem that Rwanda was always on the edge of abyss, obscuring the fact that predicting the outbreak of violence in Rwanda required focusing on the interaction between the local social effects of environmental scarcities and particular economic, political, and cultural variables.268 Elinor Ostrom, in critiquing how similar ‘tragedy of the commons’ arguments over- predict resource destruction, also notes that such broad-level models essentially accept “extreme assumptions” as general theories.269 Broad-level Neo-Malthusian models may accurately describe reality in those few cases when the conditions are right. However, a better understanding of local level empirical reality demonstrates that the predictive utility of state- level models of environment-conflict is probably less frequent than many believe.

The final section of this chapter seeks to correct the shortcomings in past environmental change- conflict research by arguing for the use of a household-livelihood framework as an initial level of analysis needed to understand the impact of environmental change on rural populations and, thereby, to conflict. The state of livelihoods over time and the factors that influence livelihood adaptation and change then become the analytical starting points for research on environmental change and conflict. Local, regional, national, and international levels of analysis are then integrated into this initial analysis of household-livelihood change to understand how and when conflict will emerge. This allows for a finer appreciation of how influences on rural livelihoods change over time. This approach also helps disentangle the underappreciated possibilities of human agency among those confronting the impacts of environmental change.

Over the past ten years, household-livelihood analysis has entered the mainstream with development practitioners, population-environment researchers, and climate change adaptation researchers.270 However, only limited steps have been taken to integrate a household-livelihood framework into environment-conflict research. Early work in this area can be traced to Indra de

People for the Population: a Cautionary Tale from the Himilaya,” in Environment and Security: Discourses and Practices, eds. Miriam R. Lowi and Brian R. Shaw (New York: St. Martin’s Press, 2000), 192-206. 268 Baechler, Violence. 269 Elinor Ostrom, Governing the Commons: The Evolution of Institutions for Collective Action (Cambridge: Cambridge University Press, 1990), 183-4. I would like to thank Ronald Mitchell for pointing out how my argument here parallels aspects of Ostrom’s critique about the tragedy of the commons. 270 See “Livelihoods Connect,” Institute of Development Studies, Sussex, UK: http://www.eldis.org/go/topics/dossiers/livelihoods; de Sherbinin et al 2008; and International Institute for Sustainable Development, International Union for Conservation of Nature and Natural Resources, and Stockholm Environment Institute 2003.

Soysa and Nils Petter Gleditsch, who argued that poverty stimulates rural conflict, hurting rural livelihoods and entitlements, and further aggravating the poverty and immiseration cycle.271 Facing livelihood loss, subsistence crises, or “the hopelessness of surviving at the margins,” many turn to criminality, banditry, or forms of collective violence like rural rebellions.272 Lief Ohlsson argued that, although poverty can generate conflict, it is the processes that lead to the rapid loss of rural dwellers’ livelihoods or their inability to attain or maintain adequate livelihoods that prove key. These forces directly cause poverty for many rural dwellers.273 To Ohlsson, the loss of livelihood is the “missing link” in describing “causal mechanisms linking both poverty and environmental factors to conflict.”274 Researchers must detail those processes that lead to increasing inequalities and rapidly cause people to lose their livelihoods, leaving deprivation and marginalization in their wake. Since many rural livelihoods depend on agriculture, Ohlsson argued, the failure of agriculture to sustain rural livelihoods is crucial. Environmental and demographic factors such as the degradation of arable land and population growth—particularly youth bulges—are important causes of livelihood loss, poverty, and, hence, conflict.275 Ohlsson returned the focus to the causes of environmental scarcities as the sources of livelihood loss, compared to de Soysa and Gleditsch’s political-economic arguments. Importantly, though, both perspectives focus on the rapid loss of livelihoods as a key causal process.

Following Ohlsson, a joint International Institute for Sustainable Development and International Union for the Conservation of Nature (IISD/IUCN) Task Force expanded on the ways in which livelihood changes affect the security of local communities. They define livelihoods as the “activities undertaken to translate resources—whether natural or human—into a means for living at the group or individual level, including the protection of goods and services.”276 Access to natural resources is crucial, they write, because it underpins all livelihoods. Environmental scarcity trends or sudden environmental shocks imperil livelihoods, according to the Task Force.

271 Indra de Soysa et al., To Cultivate Peace: Agriculture in a World of Conflict, International Peace Research Institute (Oslo, 1999), 16 and 32. 272 de Soysa et al., To Cultivate Peace, 35-36. 273 Ohlsson, Livelihood, 6-7. 274 Ohlsson, Livelihood, 3. 275 Ohlsson, Livelihood, 6-7. 276 Richard Anthony Matthew, Mark Halle, and Jason Switzer, Conserving the Peace: Resources, livelihoods and security (International Institute for Sustainable Development Winnipeg, Canada, 2002), 15-16

The impact of these trends or shocks on livelihoods depends upon the degree of vulnerability of those affected, which is partly a function of their “exposure to harm, and capacity to endure and recover.”277 Groups and individuals respond to trends and shocks to their livelihoods by adopting various coping strategies, including “development of new livelihoods, increased demand for productivity from the remaining livelihoods, conflict or migration in search of additional resources, or cooperation and trading with other groups.”278

The Task Force report’s distinction between shocks and changing trends extended Ohlsson’s focus on sudden shocks to livelihoods. Shocks resulting from natural disasters are certainly important causes of livelihood loss, but so are more subtle, long-term changes in the natural resource base upon which many rural livelihoods depend, such as degradation and depletion of natural resources, gradual reduction in resource availability due to increasing consumption or population growth, and changes in resource availability for specific groups due to distributional changes. People use various strategies to adapt to livelihood changes and the ability to adapt is partly a function of the underlying vulnerability of the livelihoods in question. Finally, the Task Force recognized that a linear relationship does not exist between environmental scarcities, their impact on livelihoods, and negative livelihood outcomes.

A deeper examination of the livelihood literature and recent research on resilience, vulnerability, and adaptation allows development of a more comprehensive framework involving rural household livelihood as an important component of environment-conflict research.279 At the core of this framework are the livelihood resources that make up household assets, capabilities, and entitlements—the dimensions of vulnerability which influence how households use those assets to respond to external pressures, and the strategies employed by households to reallocate land, labor, and capital resources in response to change, opportunities, and limitations (see Figure 6).280 A household is essentially a “social group which resides in the same place, shares the same

277 Matthew, Halle, and Switzer, Conserving, 16-17. 278 Matthew, Halle, and Switzer, Conserving, 17. 279 Coleen Vogel, "Foreword: Resilience, vulnerability and adaptation: A cross-cutting theme of the International Human Dimensions Programme on Global Environmental Change," Global Environmental Change 3, no. 16 (2006), 235-36, and other authors in this special issue. 280 Annelies Zoomers, ed., Land and Sustainable Livelihood in Latin America (Amsterdam: Royal Tropical Institute, KIT Publishers, 2001), 15.

89 meals, and makes joint or coordinated decisions over resource allocation and income pooling.”281 The people and activities in a household encompass the capabilities of the household, drawing upon the portfolio of assets available to the household and its community.282 Scholars recognize five primary asset categories for households: human assets, social assets, physical assets, financial assets, and natural assets.283 Traditional environment-conflict research has mostly focused on describing natural and physical assets, and looked to the state or institutions for an accounting of financial or human assets. Ethnographic research and micro-studies in development research have been effective in pointing to the importance of social assets for household livelihood efforts.284 In traditional highland districts like Chuschi, for example, reciprocal networks of labour exchange, called minka and ayni are crucial for household survival and variable labour requirements. Anthropologist Billie Jean Isbell described these processes at work in the D. Chuschi in the early 1970s: “minka is when an individual calls for aid, usually in the form of labor of some kind, and those who respond to his call are “lending ayni,” for which they expect repayment in comparable labour or service … A strict accounting is kept of debts and credits. The comuneros who is ostracized from participating in the mutual aid network cannot survive without recourse to cash for hired labor, and most comuneros do not participate in the cash economy of the nation.”285 Households draw upon a range of assets and capabilities as they strive to meet their various consumption and economic necessities.286

281 Frank Ellis, "Household Strategies and Rural Livelihood Diversification," The Journal of Development Studies 35, no. 1 (1998), 6. 282 Robert Chambers and Gordon Conway, "Sustainable Rural Livelihoods: Practical Concepts for the 21st Century. 296," Institute of Development Studies: IDS discussion paper, no. 296 (1991), 7. 283 Alex De Sherbinin et al., "Rural household demographics, livelihoods and the environment," Global Environmental Change 18, no. 1 (2008), 40. 284 See, for example, Mayer, Articulated, chapter 4. 285 Isbell, To Defend Ourselves, 167-68. 286 Zoomers, Land, 14.

Figure 6: Rural Household Livelihood Framework for Environment-Conflict Research

Diagram created from: DFID 1999; Scoones 1998; Collinson 2003; Chambers and Conway 1991; Adger 2006.

In developing a livelihood strategy, household decision makers are influenced by contextual and structural factors. Contextual conditions and trends include history, climate, agro-ecology, seasonality, demographic change, etc. These factors often operate over long temporal periods. They are often exogenous to the household or national context and can have transformative impacts on household-livelihood resources. For example, the steep slopes and easily eroded soils of mountain ecosystems condition the livelihood opportunities of those living in mountainous regions.287 Structural and process factors also influence household-livelihood resources. This refers to some of the familiar structural and process dimensions discussed by Homer-Dixon and Baechler where state or institutional capacities alter or transform the impacts of environmental scarcities. In this case, structural and process factors can include a wide range of man-made

287 Homer-Dixon, Environment, 105.

91 influences from governments, markets, laws, policies, investment and trade relationships, and civil society or international organizations. These forces can affect and condition the ways in which households use and develop their livelihood assets and can mediate the types of livelihood strategies employed. They also operate at various scales, from the local to the international, complicating attempts to disentangle their roles and impacts.

The assets, capabilities, and entitlements of rural households, as influenced by contextual and structural factors, determine the sensitivity of households to hazardous conditions and the household’s capacity to respond to risk, shocks, and stress—in essence, the household’s vulnerability and resilience to change.288 The impacts of resource scarcities remain important in determining vulnerabilities. However, such scarcities affect only part of a household’s endowments. The impact of scarcities on livelihoods must be examined in the context of other factors. Livelihood vulnerability thus has two crucial aspects that must be considered in any analysis: the external dimensions involving contextual or structural factors and the stresses and shocks to which livelihoods are subject and the internal dimensions involving their ability to cope or adapt.289 The likelihood of a household experiencing a particular stress or shock, combined with the particular assets and attributes of the household determines the degree of exposure and sensitivity of that household to the risk of a given environmental change. The specific strategies and abilities of a household to adapt or cope is determined by various drivers that include both the household’ attributes and broader conditions, processes, and institutions.290 In particular, because the interaction of political and economic processes in societies help determine the distribution of power and wealth between groups and individuals, and “the processes that create, sustain and transform these relationships over time,” a close relationship exists between vulnerability and power.291 Political-economic processes that disempower

288 Ellis, “Household,” 14; Barry Smit and Johanna Wandel, "Adaptation, Adaptive Capacity and Vulnerability," Global Environmental Change 16, no. 3 (2006), 286; and W Neil Adger, "Vulnerability," Global Environmental Change 16, no. 3 (2006), 268-81. 289 Chambers and Conway, Sustainable, 10; and Adger, “Vulnerability,” 270. 290 Smit and Wandel, “Adaptation,” 286-287; and John Pender, "Rural population growth, agricultural change, and natural resource management in developing countries: A review of hypotheses and some evidence from Honduras," in Population Matters: Demographic Change, Economic Growth, and Poverty in the Developing World, ed. Nancy Birdsall, Allen C Kelley, and Steven Sinding (Oxford: Oxford University Press, 2001), 355. 291 Sarah Collinson, ed., Power, Livelihoods and Conflict: Case Studies in Political Economy Analysis for Humanitarian Action, Humanitarian Policy Group Report 13 (London: Humanitarian Policy Group, Overseas Development Institute London, 2003), 3; and Adger, “Vulnerability,” 270.

92 households can greatly affect their vulnerability and condition the “space of vulnerability” for households.292 Finally, determining livelihood vulnerability requires that we examine how household assets and capabilities change over time, both as a result of the changing circumstances of household activities, and as a result of the ever changing influences of contexts and transforming structures.

Ellis’ work on how households diversify in the face of change provides important insights into how the nature of external environmental change influences the type of livelihood strategy households adopt in response. He distinguishes “ex-ante risk management from ex-post coping with crisis.”293 In the face of natural disasters or sudden, unexpected environmental events, households turn to coping strategies to deal with the shocks to their livelihoods.294 Coping happens during or after the event, and is oriented towards preserving existing livelihoods in the face of unexpected disruptions.295 A variety of specific coping strategies are employed by households to deal with shocks, including migrating, depleting assets, making claims on other assets, protecting existing assets, reducing current consumption, or shifting to lower quality consumption. Some, like Ohlsson, note that livelihood shocks can drive members of households to cope by joining criminal networks or insurgency groups. Household responses to slow stresses or risks, by contrast, differ from after-the-fact shock coping strategies. “Stresses are pressures which are typically continuous and cumulative, predictable and distressing,” writes Ellis, “such as seasonal shortages, rising populations, or declining resources.”296 In this case, households take the deliberate decision to adapt in the face of the on-going risks or stresses to make permanent changes to their livelihoods.297 Households or individuals may use one or more adaptation strategies, including temporary or seasonal migration, intensification or extensification of agricultural production, making claims on other assets, depleting assets, hoarding or protecting existing assets, stinting, or the use of diversification or complicating strategies.298 In both cases,

292 Collinson, Power, 3; and Michael J Watts and Hans G Bohle, "The Space of Vulnerability: The Causal Structure of Hunger and Famine," Progress in Human Geography 17, no. 1 (1993), 52-53. 293 Ellis, “Household,” 13-14; and Smit and Wandel, “Adaptation,” 287. 294 Ellis, “Household,” 10. 295 Zoomers 2001, 15. 296 Ellis 1998, 10-11. 297 Zoomers, Land, 15. 298 Ellis, “Household,” 11.

93 the capacity to adapt or cope is not static, but is “flexible and respond[s] to changes in economic, social, political and institutional conditions over time.”299

Finally—and perhaps most importantly for environment-conflict research—adaptation and coping strategies can have a range of positive and negative outcomes for household livelihoods: “positive if it is by choice, reversible, and increases security; negative if it is of necessity, irreversible, and fails to reduce vulnerability.”300 Households seek a low-risk framework in which to operate. But risk adaptation strategies can reduce some risks while increasing others. An adaptation strategy might reduce the risk of starvation or declining agricultural yields by increasing agricultural intensification on household plots or by extending household production to marginal lands in and around the household’s community. But this may increase the risk of shocks like landslides, as steep marginal lands are put into production or land cover that preserves vulnerable land is cleared. Or, such an approach may increase gradual land erosion and nutrient loss in existing holdings. A risk adaptation strategy might satisfy short-term risks or stresses, while also reducing the natural resilience of the natural assets necessary for long-term, sustainable household livelihoods. Thus, there are feedback loops from coping strategies and stress/risk management strategies to underlying contexts or conditions. Over time, risk adaptation strategies can have significant detrimental consequences on the underlying natural environment which households rely upon, as has happened in many parts of Haiti because of extensive deforestation on steep hillsides. Similarly, following Homer-Dixon, livelihood coping or adaptation strategies can also have an impact on transforming structures and processes, increasing the costs of government or policy activities, or providing opportunities for powerful sectors of a society to manipulate the consequences of these livelihood strategies for their own purposes, and thereby undermine markets and other social institutions in their wake. Scholars need to begin to disentangle the range of outcomes of adaptation and coping strategies for environment-conflict research, including both positive and negative consequences.

Although some environment-conflict scholars like Ohlsson have noted the implications of coping strategies in the face of livelihood shocks, the consequences of strategies of household adaptation to environmental and demographic change for environment-conflict analysis have not been

299 Smit and Wandel, “Adaptation,” 287. 300 Ellis, “Household,” 14-15.

94 adequately analyzed. In fact, it is likely that rural household livelihoods more often adapt to stresses and risks—particularly those emerging from developing environmental scarcities—than cope with shocks. Yet, the implications of household adaptation strategies and the social effects of positive or negative livelihood outcomes that emerge from these adaptations deserve more careful examination by environment-conflict scholars.

Understanding household diversification strategies, for example, helps illustrate how a more fine-grained understanding of livelihood adaptations enriches environmental change-conflict research. Diversification adaptations are widespread among rural people as a way of dealing with environmental and demographic stresses and risks.301 When employing a diversification strategy, households combine various production activities to reduce risk, such as simultaneously diversifying their cropping, engaging in non-farm income-generating activities, and receiving remittances from migration.302 However, such multi-pronged approaches complicate the ability of researchers to determine the social effects of environmental scarcities in at least three ways.

First, when households facing environmental scarcities adopt diversification strategies, outcomes may differ from those expected by environment-conflict scholars. When changes in a household’s assets are traced over time, we often see that gradually-building environmental scarcities stimulate diversification responses by rural households, leading to economic activities that forestall immiseration—undercutting the linear relationship between scarcities and household immiseration that much of the literature proposes.303 Or, diversification may only temporarily forestall immiseration until the impacts of environmental scarcities are impossible to alleviate by other adaptations. Scholars must be aware of these multiple possible outcomes and avoid assuming any direct linear relationship between scarcities and negative social effects.

Second, diversification by households experiencing environmental scarcities may generate unanticipated negative social effects, often far removed geographically or temporally from where the scarcity-household interactions took place. Different types of migration (temporary, permanent, return, repeat, circular), for example, have long been a key part of diversification

301 Chambers and Conway, Sustainable, 16. 302 Ellis, “Household,” 14-15. 303 Brush, “Diversity” 271.

95 strategies employed by rural households facing environmental and demographic change.304 Homer-Dixon and Baechler have noted how environmental scarcity-induced migration can lead to social stresses and contribute to group-identity conflicts.305 These accounts emphasize, however, how environmental scarcity-induced permanent migration is leading to conflicts in receiving areas. This is only one of many possible types of scarcity-migration relationships. From the household-livelihood perspective, it is evident that households use a variety of migration diversification strategies to deal with scarcities. The negative social effects of these strategies can be found in both receiving areas and areas of origin. Temporary or semi-permanent migration to urban or frontier areas, for example, may expose migrants to radical ideologies and mobilizing influences. Upon return, migrants may bring these influences back with them to their areas of origin, reducing subsequent social stability in these areas.306 Similarly, families often send members to urban areas for education opportunities or to find wage employment to return remittances. Such rural-urban migrants are possible targets for radical groups or radical ideologies spread through urban educational institutions, unions, or social groups. In Peru in the 1970s and 1980s, many youths targeted for recruitment by the Sendero Luminoso guerrilla group were young rural migrants to urban areas seeking educational or employment opportunities. Once radicalized, these recruits returned to their rural homes to prepare the groundwork for insurgency.307 Negative impacts on the natural resource base in the areas of origin of many migrants are also evident as a result of migration diversification strategies. Studies have shown that migration can increase environmental scarcities in some areas, because migration-induced labor shortages make it impossible for households to maintain natural assets like farm terraces, speeding soil erosion while increasing livelihood vulnerability for households.308 The empirical

304 Ellis, “Household,” 70-73; de Sherbinin et al., “Rural Household,” 45-46; Richard E Bilsborrow, "Migration, Population Change, and the Rural Environment," Environmental Change and Security Project Report 8, no. 1 (2002), 77-78; and Robert McLeman and Barry Smit, "Migration as an Adaptation to Climate Change," Climatic change 76, no. 1-2 (2006), 31-53. 305 Homer-Dixon, Environment, 93-96; and Baechler, Violence, 92-96. 306 See Benjaminsen, “Does Supply-Induced Scarcity,” 829-830. 307 Degregori, “Harvesting,’ and Palmer, Shining Path. 308 Bilsborrow, “Migration,” 37; Moshe Inbar and Carlos A. Llerena, “Erosion Processes in High Mountain Agricultural Terraces in Peru,” Mountain Research and Development, 29(1), Feb. 2000: 72-79. See also: Paolo Tarolli, Federico Preti, Nunzio Romano, “Terraced landscapes: From an old best practice to a potential hazard for soil degradation due to land abandonment,” Anthropocene 6, June 2014: 10–25. de Sherbinin et al. note that “remittances may have negative impacts on the environment by increasing investment in environmentally detrimental practices such as extensive pasturage or the transformation of agricultural lands into peri-urban real estate.” de Sherbinin et al., “Rural Household,” 46.

96 consequences of diversification strategies like migration are considerable. Scholars must examine each case in detail from the household-livelihood perspective to untangle the consequences of livelihood adaptation to environmental scarcities.

Finally, a detailed examination of patterns of livelihood diversification in the face of environmental scarcities can shed lights on the attributes of households that use these strategies, which households are vulnerable to the impact of environmental scarcities, and how differences in patterns of environmental scarcity help condition household diversification strategies. The impacts of scarcities are not felt equally by all rural households, given the large variations in assets, capabilities, and entitlements. However, existing environment-conflict research has done little to differentiate its analysis of the impacts of environmental scarcities on rural households. Livelihood scholars recognize that patterns of differentiation vary globally and by asset holdings.309 Some have argued that diversification in Asia and Latin America displays a U- shaped association between level of income diversification and level of income:310 “households with little land have become integrated into labor markets as wage workers, and agriculture is now a small component of income, being mainly for self-consumption; medium peasants are less reliant on off-farm income; and better-off households are diversified, but in a variety of activities ranging from wage-work to self-employment and investment in small business.”311 Patterns in Africa, by contrast, are somewhat different, with overall levels of diversification lower than in Asia, though non-farm income sources like remittances are higher in parts of Sub-Saharan Africa bordering South Africa. These patterns suggest that the impact of environmental scarcities on diversification is also strongly affected by variation in access to markets and the degree of household access to infrastructure, services, and institutional assistance, like roads, market services, supplies, agricultural extension services, power, etc.312 Where land is abundant or adequate for household livelihoods, the limits to diversification may be markets and services for rural households.

309 Ellis, “Household,” 10. 310 Ellis, “Household,” 10; and Kirsten Appendini and Annelies Zoomers, "Land and Livelihood: What Do We Know, and What Are the Issues," in Land and Sustainable Livelihood in Latin America, ed. Annelies Zoomers (Amsterdam: Royal Tropical Institute, KIT Publishers, 2001), 34, footnote 11. 311 Appendini, "Land,” 27. 312 Ellis, “Household,” 10.

Ultimately, environment-conflict research should do more than diagnose how and why environmental and demographic change contributes to conflict. It should offer scholars and policymakers insights into what interventions would be most effective at mitigating the negative social effects of scarcities. Integrating and exploring the implications of a livelihood framework to environment-conflict analysis offers a crucial first step toward this goal.

Effectively integrating a livelihood framework into environment-conflict research, however, requires researchers to surmount several obstacles. First, data problems abound at the local level, in the collection of local agro-ecological and livelihood data and in the distribution and use of this data by researchers and policymakers. Second, local studies need to work across level of analyses and be scaled-up to assess whether there are wider impacts on stability and scarcity. Assessing impacts across scales poses significant analytic challenges. Finally, the high degree of interactivity and multi-causality in analyzing the impact of environmental scarcities on household livelihoods makes it difficult to assess the relative influence of different factors and processes.313 Considerable ink has been spilled over the past fifteen years on methodological debates in environment-conflict research.314 Yet, beyond agreement that diversity in research methods can help map causal mechanisms and test hypothesized relationships, no solution has been found for the difficult problem of differentiating causal primacy among diverse processes, factors, and drivers. A new methodology or ontology for researchers may be needed, perhaps based upon complexity theory and complex modeling of the type used by social-ecological systems research.315

Conclusion

This chapter argues that the state level of analysis that has dominated qualitative environment- conflict research appears to have generated considerable uncertainty about the validity of hypothesized connections and considerable under-specification about the myriad pathways that exist in human-environmental change interactions. This chapter has proposed a household- livelihood framework for future research to correct some of these problems. Such an approach

313 See Schwartz, Deligiannis, and Homer-Dixon, Environment and Violent Conflict. 314 Levy, "Is the Environment a National Security Issue?; Homer-Dixon, Environment, Gleditsch and Diehl, Environmental; and Schwartz, Deligiannis, and Homer-Dixon, Environment and Violent Conflict. 315 See Carl Folke, "Resilience: The Emergence of a Perspective for Social–Ecological Systems Analyses," Global environmental change 16, no. 3 (2006), 253-67.

98 will also lead to a better appreciation of the many previously-ignored local violent conflicts that have their roots in human-environmental change interactions.

The household-livelihood framework outlined above provides a heuristic guide for field research in the D. Chuschi discussed in the following chapters, as we attempt to reconstruct historical changes in household livelihoods over many decades. The data requirements for modeling household livelihood decision making are substantial; attempting to do so for entire communities and then reconstruct these patterns historically is impossible, particularly in Peru’s Sierra, where little historical household data is available. While a detailed application of the household- livelihood framework may be useful for future studies, it was not possible to rigorously apply the framework during fieldwork in the D. Chuschi. The historical trend data is largely non-existent; instead, we have created snapshots of household livelihood conditions using data from different periods, both from archival research and ethnographic field interviews. The framework thus provides a crucial ordering function to the project’s field data, ensuring that questions about changing household subsistence survival are central to the discussion in the following chapters. The framework also helps illustrate the key structural constraints and opportunities that conditioned household decision-making in the D. Chuschi in the decades before the Shining Path insurgency.

Chapter 4 Historical Structural Changes and Transformations in Rio Pampas Affecting Natural Assets and Livelihoods in Chuschi and Quispillaccta Introduction

In order to understand if households in the D. Chuschi were facing increasing livelihood pressure in the decades before Sendero’s uprising, this thesis will first explore long-term historical structural changes and transformations affecting peasant livelihoods and natural assets. In the D. Chuschi, changes in the availability of household assets operated in the context of changing structural conditions to lead small-holder farmers in Quispillaccta and Chuschi to adjust household livelihood strategies in ways that contributed to inter-community conflict. One of the most important household assets is in the natural asset category, and comprises land and water necessary for growing crops and raising livestock. For farmers whose livelihoods largely depend upon what they can produce, the availability or scarcity of land has been crucially important in moderating household livelihood strategies. Until well into the twentieth century, there were few livelihood alternatives for residents in Chuschi and Quispillaccta other than small-holder farming and herding. Patterns of change over time in land assets and the creation of new and competing entitlements for assets thus provide insight into crucial mechanisms of livelihood change over time in the two communities, and social conflicts these changes helped to spawn.

Changes in assets and entitlements were particularly important later in the 19th and 20th century when key transformations in economic, demographic, and political factors began to alter the structural context for small-holder farmers in Chuschi and Quispillaccta. However, understanding patterns of change in land and water assets and entitlements over time requires tracing developments to the pre-Colonial period, when the Incas transformed ethnic settlement in the Rio Pampas in ways that heightened conflict among communities in the area. Broad transforming structural processes like Spanish Colonialism and economic shifts that led to resource capture by mestizo elites also dramatically altered settlement patterns and reduced arable land for communities, pressures only partially offset by the dramatic crash in indigenous populations in Peru that accompanied Spanish Conquest. Indigenous populations in the area would take centuries to rebuild to pre-Conquest levels.

99 100

Conquest led to the creation of new and competing entitlements to natural assets in Quispillaccta and Chuschi. Spanish elites and the Catholic Church appropriated some of the best agricultural land, codifying their control of these resources in the new colonial state and through the establishment of highland estates. This process continued at various times in the following centuries, particularly in the 19th and early 20th century as various export-led commodity booms created incentives for other elites to purchase or capture community lands. These highland estates on lands claimed by communities since pre-Colombian times further complicated ownership entitlements in some areas by housing generations of serfs on the estates who themselves developed cultural identities as workers of the land and expectations of entitlements over estate lands – lands that were still claimed by the surrounding communities.

By the twentieth century, regional economic developments and growing populations in both communities reduced land per capita and helped to alter settlement patterns in the two communities in a way that heightened inter-community conflict. Reduced cultivated land availability in the district, in lower elevation areas near the Rio Pampas, combined with growing prices for livestock and other livelihood opportunities to stimulate a spread of settlement patterns in the two communities to higher elevation areas, where new permanent settlements heightened conflicts over uncertain community boundaries and rights to contested land holdings. In some cases, communities began to come into conflict with former plantation serfs. As plantations were sold or abandoned, competing claims to the assets arose between communities claiming original ownership to the lands and former plantation serfs who had worked the land for several generations. In other cases, the communities began to agitate to recover entitlements to lands whose output had been devoted to the Catholic Church. These conflicts helped shape fights over land and water assets in the twentieth century. This story is outlined in the following field work chapters 4-8,

Chapter 4 outlines how elite land capture and historical structural changes impacted livelihoods of people living in the area from pre-Conquest to the end of Spanish colonialism. The chapter also argues that existing accounts of rural change in the D. Chuschi underemphasize the impact of elite capture of cultivated and grazing lands, compared to other areas in Ayacucho, and thereby present an incomplete understanding about how elite capture of district lands negatively impacted livelihoods and conflicts in the district. By the time researchers first arrived to study the district in the mid-1960s, few elite landowners were left. Landholding seemed particularly

101 egalitarian and free from elite domination, compared to other areas in Ayacucho. As new land reforms swept away the last vestiges of elite land control, including lands controlled by the Catholic Church, the District seemed largely free from the influence of abusive elites. This chapter and chapter 5 argue that this was partly an illusion – that the pernicious impacts of past elite resource capture were still stimulating conflict in the district, particularly conflict between district communities. Chapter 4 focuses in particular on the livelihood impact of elite capture of cultivated lands in the D. Chuschi, while chapter 5 examines the capture of district grazing lands and conflictual efforts by communities to regain control of these areas from the 1800s to the 1980s.

Natural asset capture and entitlement capture in Western Cangallo – Pre-Conquest to 20th Century 4.2.1. Pre-Colonial and Colonial developments

The patterns of change in natural asset holding and competition over inherently limited cultivated land in the steep valleys of Pampas River basin began in the pre-colonial era and were significantly altered by Spanish colonial exploitation and reforms. More than a half century before the Spanish arrived in Peru, competition between the Inca empire and the Chanka and associated tribes in Southern Peru led to the rearrangement of tribal groups inhabiting the Rio Pampas region. The Andahuaylas region, east of central Ayacucho, was the stronghold of the Chanka, and their influence spread west to the Pampas River. Various independent groups inhabited the Pampas River valley after the collapse of the Wari state, and some may have sometimes associated with the Chanka when it suited their purpose, including during the Chanka’s struggle with the Incas in the 15th century.316 The Inca’s defeat of the Chanka in the

316 Brian S Bauer and Lucas C Kellett, "Cultural Transformations of the Chanka Homeland (Andahuaylas, Peru) During the Late Intermediate Period (Ad 1000–1400)," Latin American Antiquity 21, no. 1 (2010), 109; Frank Meddens and Cirilo Vivanco Pomacanchari, "The Late Intermediate Period Ceramic Traditions of Ayacucho, Apurimac, and Huancavelica: Current Thoughts on the Chanca and Other Regional Polities," Ñawpa Pacha: Journal of Andean Archaeology 38, no. 1 (2018), 47. A widely accepted account by Luis Lubreras argued that the entire Pampas Valley, from its headwaters in the western cordillera of the Andes to the Apurímac River in the east, had long been a stronghold of the Chanka ‘Confederation.’ However, recent archeological work in the area, cited above, has led to a re-assessment of Lumbreras’s arguments of widespread Chanka control of the Rio Pampas, including areas around the D. Chuschi. Rather than a Chanka-led confederation or proto-state in the area, there is persuasive evidence that the Pampas valley was inhabited by numerous competing groups who sometimes cooperated with the Chanka when it served their purposes. This conclusion supports early archeological work around the D. Chuschi which failed to find evidence of a Chanka presence in the area. See Isbell, To Defend, 62. Lubreras’ argument can

102 late 15th century led to expanded Inca control over the Andes and dislocation for the Chanka and associated peoples.317 The Incas integrated the defeated Chankas and associated groups into their expanding empire, resettling many Pampas River valley tribes to other parts of the empire, while bringing in other loyal tribes (mitimae) to settle vacated areas of the Pampas River basin.318 According to Klaren, settler groups “replaced rebellious groups and others whose loyalty was suspect and had been deported to other settled regions for re-education and reintegration into the empire.”319

This practice of shuffling ethnic groups around the empire was a central strategy for maintaining Inca control over their empire; however, it also served to exacerbate inter-community tensions in several ways.320 Few areas of the expanding Inca Empire were subjected to such a disruptive population shuffling as the Pampas River basin, with at least 20 different ethnic groups being settled into the area by the Incas in order to attempt to forestall renewed rebellion.321 For the Incas, resettlement was a strategy to “educate and convert” local populations into loyal subjects. The resettlement practice, however, also split local societies into opposing communities and undermined the maintenance of large-scale polities that could someday pose a renewed threat to the Incas.322 The Inca resettlement strategy also served to exacerbate intra-communal tensions that were inherently competitive and sometimes conflictual by virtue of the Andean production system. The exploitation of Andean landscapes into different vertical production activities – farming plots and raising livestock at different altitude levels spread over a wide area – forced households to rely on reciprocal exchange networks like kinship groups (ayllus) and ethnic groups in order ensure livelihood survival. Kinship groups expanded claims over territory in order to access and control as many production zones as possible. Competition between kinship

be found in: Luis G. Lumbreras, The Peoples and Cultures of Ancient Peru (Washington, DC: Smithsonian Institution Press, 1974). 198. 317 Klaren, Peru, 16-7. 318 Steve J. Stern, Peru's Indian Peoples and the Challenge of Spanish Conquest: Huamanga to 1640 (Madison: Univ of Wisconsin Press, 1993), 20; Jaime Urrutia, Huamanga: región, proceso e historia 1536-1770 (UNSCH, Ayacucho, 1984), 35. 319 Klaren, Peru, 19. 320 Klaren, Peru, 25-6. 321 Urrutia, Huamanga, 26, 35. 322 Stern, Peru's Indian Peoples, 12.

103 groups, ethnic groups, and between communities was a common outcome of adjacent, overlapping, or inter-mingled group claims; though, coexistence and mutual agreements to allow passage over each other’s lands was also equally common.323 The Inca resettlement strategy in the Pampas River basin thus heightened existing group divisions by introducing additional ethnic groups to the area, and to the competition for cultivated lands and pasture lands.

The land conflict between Chuschi and Quispillaccta most likely originated with this process of Inca resettlement of the Pampas River basin. The standard account of the origins of Chuschi and Quispillaccta outlines that each community was moved by the Inca from other parts of the empire – Chuschinos descended from Aymaraes Indians moved from Apurimac, while Quispillacctinos descended from Canas Indians moved from Cuzco. Colonial era documents held by the communities support this account, and provide the first documentation of the conflict between the two communities. A document in Quispillaccta’s archives from 1567 outlines the legal fight over land between the “Canas Indians” and the “Aymaraes Indians”.324 An inspection report of the area a few years later by a Spanish colonial official in 1574 notes that the Aymaraes were resettled in the Pampas River basin after the defeat of the Chanka.325 Additional details of the Aymaraes’ origins emerge from a decree dated 1593, prepared by the Corregidor Blasco Núñez de Vela – the equivalent of a chief district magistrate with supervisory responsibility over the Indian towns in the area. Núñez’s decree similarly details the land conflict between the

323 Stern, Peru's Indian Peoples, 24-5; Enrique Mayer, Articulated, 50-1. Most Andean scholars use the term ayllu, instead of “kinship groups”, a term that captures only one facet of the meaning of ayllu, which evolved over time to refer to many groups beyond immediate relatives, and the privileges and obligations that go along with being part of an allyu. Isbell notes that ayllu can refer to “a barrio, the entire village, one’s family, or even the district, the department, or the nation.” Isbell, To Defend, 105. While the ayllus of the Incan or pre-Incan period were largely composed of extended family groups, notes Jacobson, with Spanish colonial rule, “ayllus underwent a gradual transformation into settlements defined by geographic location and their claims to land … Although kinship ties, real or symbolic, continued to be important, they became less rigid, and the ayllus were increasingly inhabited by Indian peasants from different regions.” Nils Jacobsen, Mirages of Transition: The Peruvian Altiplano, 1780-1930 (Berkeley: University of California Press, 1993), 87. While Isbell’s study is most concerned with the kin dimension of ayllu, the geographical dimension of ayllu as it relates to barrios in Quispillaccta and Chuschi has not been well explained in her study of Chuschi. As outlined below, a geographic understanding of group movement in these communities helps to untangle the evolution of the conflict between Quispillaccta and Chuschi over the centuries. Mayer carefully explores the relationship between household economies and kinship in chapter 1. 324 John Earls and Irene Silverblatts, “Ayllus Y Etnias de la Region Pampas-Qaracha: El Impacto del Imperio Incaico,” III Congreso Peruano el Hombre y la Cultura Andina, 31 de enero-5 de febrero 1977: actas y trabajos, 3rd Congreso Peruano del Hombre y la Cultura Andina, Lima, 1978: 163. The Canas are variously referred to in these early colonial documents as the “Cochas” kinship group, the “Canas” Indians, or as the “Quispellacta” kinship group. 325 Isbell, To Defend, 63.

104

Aymaraes inhabitants of Chuschi and the Canas inhabitants of Quispillaccta.326 The Aymaraes from Chuschi charged that Canas Indians from west of the river were usurping land given to them by Topa Inca Yupanqui, the Inca ruler at the time of the defeat of the Chanka.327 The Canas defended their land claim by producing a document prepared by the previous Corregidor Damián de la Bandera, stating that they had been relocated to the area from south of Cuzco by the Inca ruler Huayna Capac, the successor to Tupac Inca Yupanqui and the last Inca ruler to complete his reign before the Spanish conquest.328 In his decision, Corregidor Blasco Núñez affirmed that the Canas had a right to 10 topos of land to the west of the river bordering Chuschi – Tacsay Mayo or Qunchalla Kuchu.329 This roughly corresponds to all the land west of the river in the Qunchalla/Chuschi valley.330 A 1602 document outlines a renewal of the land dispute between the Aymaraes Indians of Chuschi and the Canas Indians. These documents suggest that both the Aymaraes and Canas were resettled into the area by the Inca after the defeat of the Chanka Confederation in the latter part of the 15th century, the Aymaraes arriving first, followed by the Canas descendants of Quispillaccta.

326 Isbell, To Defend, 65. 327 Isbell, To Defend, 65. There are various spellings of Topa Inca Yupanqui. I adopt Stern’s form. 328 Isbell, To Defend, 65; Marcela Machaca Mendieta, “Vigencia y Continuidad de Cultura y Agricultura Andina en Quispillaqta” (Ingeniera Agronomo Tesis, Universidad Nacional de San Cristobal de Huamanga, 1991), 233. Machaca cites the Communal Archive of Quispillaccta. Huayna Capac died in 1525 or 1527 after contracting smallpox introduced by the Spanish, which had spread to the Andes by 1525. 329 Isbell, To Defend, 65. Place names for natural features in Chuschi and Quispillaccta vary from source to source and over time, significantly complicating efforts to combine an analysis of historical records, oral history, scholar studies, and topographical maps. The name one chooses for an area can confer legitimacy of control for one group or another. So, deciding on one name or another is a politically fraught exercise when trying to disentangle ownership disputes. I am sensitive to these challenges in the D. Chuschi, where there has been a high level of contestation over land and water resources. The small stream separating Chuschi and Quispillaccta, for example, goes by the names Qunchalla Kuchu (Machaca, “Vigencia y Continuidad,” – a Quispillactino), Taksa Mayo (Cesar Ramon et al. La Comunidad de Chuschi (Lima: Instituto Indigenista Peruano, 1967); Isbell, To Defend), Rio Chuschi (Instituto Geográfico Nacional 1993, Departamento de Ayacucho 2006, and La Serna, The Corner of the Living), or Chocloqocha Mayo (Isbell, To Defend,). I refer to it as the Rio Chuschi, and the small valley that opens from the Rio Pampas below as the Qunchalla/Chuschi valley. 330 A topo of land is an Andean unit of measure used to quantify fields during colonial inspections. While the exact surface area of a topo varied according to soil quality, elevation, topography, or other factors that affected agricultural productivity, Wernke argues that it is roughly equivalent to 3,496m2. A topo is still used as a unit of measure in Cuzco, according to Watters. His study notes that one topo is equivalent to a 40mx80m field, or 3200m2 – roughly a third of a hectare. Watters, Poverty, 347. Steven A. Wernke, Negotiated Settlements: Andean Communities and Landscapes under Inka and Spanish Colonialism (Gainesville: University Press of Florida, 2013), 253-4. Using Wernke’s accounting, ten topos is thus 34,960m2 - or about 3½ hectares. Calculations with a topographical map of the Qunchalla valley show that ten topos are roughly equivalent to the area of the maize zone lands west of the Rio Chuschi/Qunchalla, from the edge of the Rio Pampas to the heights above Chuschi and Quispillaccta, but not including the higher altitude grazing lands.

105

Other anthropological work in the area, however, suggests an alternative interpretation of the origins of the Aymaraes in Chuschi, and counters their claim to have been resettled in the area by the Incas. John Earles and Irene Silverblatts believe that the Aymaraes were in the Pampas River basin since the 13th century, well before the Inca defeated the Chanka.331 While they do not offer archeological or documentary evidence to place the Aymaraes of Chuschi in their present location since before the Inca conquest, they argue that linguistic and kinship evidence from Chuschi and surrounding communities supports the conclusion that their presence in the Pampas river basin predates Inca mitimaes – ethnic groups like the Canas who were resettled in the area by the Inca after the defeat of the Chanka. They do not dispute that Quispillacctinos are descended from Canas Indians placed there by Huayna Capa.332 However, they question whether the Aymaraes were really mitimaes settled in the Pampas River basin by the Inca Topa Inca Yupanqui, as stated in colonial documents quoted by Isbell and others. Instead, they argue that the claim in Chuschino colonial documents to their being Aymaraes mitimaes from the Inca period was most likely a way of legitimizing their claims to land in the area, because rights to land derived from tenancy granted by the Incas was considered legitimate under Spanish colonial law. While the Canas of Quispillaccta had documents to prove their land rights, Chuschi’s Aymaraes had no way to prove that their residency predated the Inca period. So, Earls and Silverblatts believe that the Aymaraes constructed a claim to have been placed in the area by Inca Topa Yupanqui, in order to ensure that they had legal land rights under Spanish colonial law.333 Until further archeological work is done in the area, the origins of Aymaraes in the Rio Pampas remains contested. Evidence clearly indicates, however, that the competition for land between the inhabitants of Chuschi and Quispillaccta goes back at least to the early colonial era, and likely pre-dates Spanish conquest of Peru.

Spanish conquest and the disruption and reorganization of indigenous societies throughout the highlands that took place in the decades following conquest added a new layer of complication to the land competition between Chuschi and Quispillaccta, decimating populations, altering physical settlement patterns, heightening conflicts between ethnic groups, and increasing

331 Earls and Silverblatts, “Ayllus Y Etnias,” 160-1. 332 Earls and Silverblatts, “Ayllus Y Etnias,” 164. 333 Earls and Silverblatts, “Ayllus Y Etnias,” 164-5.

106 livelihood stress on households through disease and forced labour requirements. The first decades of Spanish control were marked by social disruption and colonial plunder of the indigenous population on a vast scale.334 However, by mid-16th century, the colonial exploitation system in Peru was in decline and in need of reform if it was going to continue to supply Spain with significant economic benefits in the future. The days of easy economic plunder in Peru were over and the Spanish crown grew increasingly concerned about declining revenues from tribute, reduced silver production from the mines in Potosí, increasing Indian unrest in parts of the colony – particularly in Ayacucho and in the former Inca strongholds in Cuzco – and disappearing opportunities for the Spanish immigrants who continued to pour into the colony.335 The demographic collapse of indigenous societies in Peru following conquest was also reducing the ability of the Spanish to extract surplus wealth and find the labour necessary to keep the “economy of plunder” operating.336

While estimates vary about the size of the pre-conquest indigenous population in Peru, there is no dispute that the consequences of the Spaniards’ introduction of new diseases to native populations with no resistance, combined with brutal exploitation and control in the decades after the defeat of the Inca – literally working Indians to death in the silver mines and rural estates – resulted in a spectacular crash in Indian populations in the Andes.337 Noble David Cook estimates that a pre-conquest Indian population of about 9 million in 1520 collapsed to slightly over 1 million by 1570, declining further to around 600,000 by 1620.338 See Peru population chart, Figure 7. Rates of decline were not even throughout Peru, however. While coastal and lowland areas suffered population declines between 75-85% in the period from 1520-1570, highland regions lost about 50% of their populations.339 The “relatively stable” populations of the central highlands, compared to coastal areas, began to significantly decline in the latter

334 Klaren, Peru, 39-53. 335 Klaren, Peru, 56. 336 Term is from Klaren, Peru, 56. 337 Klaren, Peru, 48-9. 338 Noble David Cook, Demographic Collapse, Indian Peru, 1520-1620 (Cambridge University Press, 1981), 114. 339 Cook, Demographic, 94. Cook argues (208) that coast and lowland areas were more vulnerable to infectious diseases than cold, isolated highland communities.

107 decades of the 16th century and early decades of the 17th century.340 Data for the Corregimiento of Vilcashuaman, the trusteeship area that included the Rio Pampas encomiendas encompassing present-day Chuschi and Quispillaccta show that by the second half of the 17th century Indian populations collapsed spectacularly to less than 25% of their levels in the 1570s.341 The impacts of disease and repression depleted Indian populations and caused severe disruptions to clan- based livelihood exchange systems (ayllu system) in communities in the Rio Pampas and throughout Ayacucho, even as population decline eased pressures on productive land.342

340 Cook, Demographic, 200. Cook describes the populations of the central highlands as relatively stable compared to the northern or coastal highlands. His characterization is puzzling, however, since he also notes that Indian population levels declined by an estimated 50% in the central highlands between 1520 and 1570. Stern, drawing in part on Cook’s data, also concludes that the Indians in Huamanga “faired relatively well” in adapting to Spanish conquest by the mid-16th century – that their “post-conquest decline was not as irrevocably devastating as in other Andean areas.” Stern, Peru's Indian Peoples, 44-5. However, it is hard to accept that the halving of indigenous populations in the area was not traumatic for ethnic groups in the central highlands, something Stern admits in a different publication. See Steve Stern, “The Social Significance of Judicial Institutions in an Exploitative Society: Huamanga, Peru, 1570-1640,” in George A. Collier et al., eds., The Inca and Aztec States, 1400-1800: Anthropology and History (New York: Academic Press, 1982), 305. 341 The two Repartimientos – or land divisions - of Quichuas & Aymaraes and Totos, located in the Corregimiento of Vilcashuaman and comprising the area of present-day Chuschi and Quispillaccta, illustrate these declines. Between 1573 and 1630, counts of tribute Indians declined 76% and 83% respectively. Data in Appendix C, Stern, Peru’s Indian Peoples, 204. 342 Mayer, Articulated, 101. Mayer provides a detailed illustration of the “deleterious effect on the level of household welfare” obligations in Huanuco in the 1560s.

108

Figure 7: Peru Population:Peru Population: Pre-Conquest Pre-Conquest (1520) to 2007 (1520) to 2007

31500000

29500000

27500000

25500000

23500000

21500000

19500000

17500000

15500000

13500000

11500000

9500000

7500000

5500000

3500000

1500000

-500000

1520 1549 1580 1610 1641 1671 1702 1733 1763 1794 1824 1855 1886 1914 1945 1975 2006

Peru

Sources: See Appendix 2.

Like past conquerors, consolidation of Spanish rule disrupted and reorganized indigenous communities in ways that would have lasting impacts and aggravate inter-group conflict in many highland areas. One of the most important impacts resulted from the efforts in the highlands to concentrate diffuse Indian settlements into Spanish-style nuclear settlements. By the 1560s, forty years after conquest, Spanish occupation of Peru was in crisis. Silver production was declining precipitously at the mines in Potosi, and Indian “rebelliousness and resistance to labour demands” increased throughout the highlands.343 In the former Incan strongholds of Cuzco, in the south-eastern highlands, a brewing neo-Incan rebellion posed the “most-immediate” threat to Spanish control of the conquered territories.344 Local rebellions were also fermenting in other corners of Peru. In northern Ayacucho, Huancas Indians, long bitter enemies of the Incas, also

343 Stern, Peru’s Indian Peoples, 71. 344 Stern, Peru’s Indian Peoples, 71.

109 plotted with neo-Incan conspirators to revolt against the Spanish.345 In south-central Ayacucho, meanwhile, Indian rebelliousness coalesced around a millenarian Indian rebellion that promoted the return to a pristine native spirituality cleansed of Incaic or Spanish influences. The Taki Onqoy rebellion, as it has come to be known, fed on impoverishment, social dislocation, and most of all the profound disillusionment among the diverse Indian groups of the area who had adapted to colonial conditions but who despaired “that the contradictions of colonialism would lead to unbearable results.”346

In order to stabilize and reform the Peruvian colony and ensure the establishment of a secure source of future wealth for the Spanish crown, Francisco de Toledo was appointed in 1569 by the Spanish Crown as the Viceroy of Peru to undertake far-reaching reforms of how indigenous populations were governed and exploited. Toledo’s reforms significantly affected land habitation patterns of Indian communities like Quispillaccta and Chuschi with lasting impacts well into the 20th century. Among Toledo’s reforms was the establishment of the reducción strategy for Indians throughout Peru. Reducción consolidated “dispersed Indian settlements into church-based towns of four hundred inhabitants or more.”347 Consolidated towns facilitated religious instruction of Indians. As Gose notes, the establishment of new towns based on the Spanish model went hand in hand with the destruction of old settlement patterns: New towns were to be as far away as possible from pagan settlements and their shrines. Once construction of new settlements began, Indians were to destroy their ‘old towns,’ not only to salvage building materials but also to make reducción definitive and irreversible. A grid pattern was to define the new towns. Onto the central plaza were to face the church, community buildings, the town council, court, and jail. … Indians were to construct their houses within the grid plan’s blocks with their doors facing the street.348

345 Stern, Peru’s Indian Peoples, 69-70. 346 Stern, Peru’s Indian Peoples, 68. See also Jeremy Mumford, “The Taki Onqoy and the Andean Nation: Sources and Interpretations,” Latin American Research Review 33(1): 1998: 150-165. 347 Peter Gose, Invaders as Ancestors: on the Intercultural Making and Unmaking of Spanish Colonialism in the Andes (Toronto: University of Toronto Press, 2008): 119. See also Jeremy Ravi Mumford, Vertical Empire: The General Resettlement of Indians in the Andes (Durham, N.C.: Duke U.P., 2012). 348 Gose, Invaders, 121.

110

The implementation of the reducción strategy also established tribute obligations for Indian communities, with the building or expansion of community churches and municipal offices usually the first tributary project. At a more fundamental level, however, the strategy was rooted in notions of “ordering, modern rational persuasion, religious conversion, and political subjugation.”349 It was the Spanish colonial version of modernization, with the transformation of livelihoods that this implied. Reducciones, Gose notes, “were to embody an entirely different and more elevated way of life, including Catholicism, urbanity, public order, rational governance, improved personal hygiene, and morals.”350 The lawlessness and rebellion brewing in Peru provided the political will to enact and enforce such wide-ranging reforms in the highlands.351

The physical and spatial implications of Toledo’s reducción reform were significant for highland communities, with impacts evident to the present day.352 In the case of Quispillaccta and Chuschi, these reforms aggravated competition between the Canas and Aymaraes for hundreds of years. Previously dispersed settlements of kinship groups of Canas and Aymaraes, depleted by disease and forced labour requirements, were consolidated by Toledo’s reforms into concentrated towns based on the Spanish model. The groups appear to have been relocated during the reducción reforms to their current town-centre sites (referred to as the Matriz), next to each other, on opposite sides of the Rio Qunchalla/Chuschi – the small brook which divides the deep horseshoe shaped valley that runs into the Pampas River. (See Figure 3)

Their close proximity raises the question of why Spanish reformers would place competing ethnic groups in such close proximity, virtually guaranteeing future enmity and conflicts between the communities? Interviews and documentary evidence suggests that their current location was probably the outcome of competition between rival Spanish encomenderos– Spanish colonial elites who were given trusteeship by the Spanish crown over particular groups of Indians as a

349 Gose, Invaders, 119-122. 350 Gose, Invaders, 122. 351 Gose, Invaders, 120. 352 Mayer, Articulated, 35-6.

111 reward for service to the Spanish crown – who used the Indians under their control to expand their power and resource holdings over scarce cultivated land in the area.

In the case of Chuschi and Quispillaccta, it appears that manipulation and competition by local Spanish encomenderos led each tribal group to be consolidated into nuclear settlements in the same small valley, on either side of a tributary that runs into the Rio Pampas, the Rio Qunchalla/Chuschi. In the decades following the conquest of Inca Peru, Spanish conquerors, their descendants and various elites from Spain consolidated control over Indian populations throughout Ayacucho, to enrich themselves and help facilitate the Spanish crown’s exploitation of Indian Peru. However, colonizing Spanish elites also competed among themselves for Indian riches, and these conflicts probably also heightened inter-group conflict in the Pampas River area. Documents reviewed by Isbell appear to indicate that in the early colonial period the Aymaraes and Canas were under the control of different Spanish colonizers – or encomenderos.353 The encomenderos “could collect tribute in the form of goods or labour” from the Indians “in the name of the crown”, and “in return they assumed responsibility … for protecting and Christianizing” the Indians.354 Historian Steve Stern notes that encomenderos in Ayacucho frequently allied with the Indians under their control against the holdings of neighbouring encomenderos, encouraging their Indians to usurp neighbouring lands and thus expand the encomenderos’ holdings – the repartimiento.355 For the Indians, these alliances with encomenderos provided some protection against excessive colonial exploitation, while the encomenderos enhanced their ability to demand favours from his clients.356

Competing Spanish encomenderos likely used their control of the Indians under their charge in ways that heightened the competition between the Canas and Aymaraes for control of the lands in the area, ensuring that relocated ethnic groups under the Toledo reforms were pushed to the

353 Isbell, To Defend, 63-5. 354 Klaren, Peru, 41. 355 Stern, Peru’s Indian Peoples, 119-20; 263. Stern defines a repartimiento as the “encomienda district or jurisdiction.” 356 Steve Stern, “The Social Significance,” 297. Wernke suggests that colonization worked both ways because Spanish elites required the help of Indians to achieve their goals, leading to various compromises with native groups and allowing Indians to exercise a degree of control through Andean practices and institutions over their Spanish overlords. Wernke, Negotiated Settlements, 13-15.

112 edges of trusteeship holdings.357 Mumford notes that combining different repartimientos into one reducción was one approach pursued by Toledo’s inspectors.358 In the case of Chuschi and Quispillaccta, however, the Indians of two different and partially overlapping repartimientos were essentially brought together into two side-by-side reducciones, a solution that facilitated both Toledo’s reform goals and the local interests of encomenderos and their Indian chiefs.359 Encomenderos appeared to use the relocation of the Indians as a way of expanding and consolidating their holdings by locating communities in scarce fertile tributary valleys that flow into the Rio Pampas. The dispute highlighted in the decree by Corregidor Blasco Núñez de Vela in 1593 notes Aymaraes complaints that the Canas were being aided in the usurpation of their land by “Negro slaves” belonging to the encomendero Pedro de Rivera. Rivera’s family had trusteeship over the Canas Indians, suggesting that he approved of using his own black slaves to help expand the Canas territory at the expense of the Aymaraes, who were part of a different trusteeship or repartimiento, held by the encomendero Juan de Mañueco.360

With good cultivated land scarce in this part of the Rio Pampas, and with each ethnic group under the control of different encomenderos, it appears that the reducción strategy offered encomenderos the opportunity to relocate the Canas and Aymaraes as close as possible to the Qunchalla/Chuschi valley to maximize their holdings of temperate maize land and the grazing land in the high-altitude plains above the valley. For Toledo reform inspectors, relocating both communities in such close proximity also facilitated religious conversion and control by the Catholic church and simplified tribute collection in the area.

357 Machaca’s discussion of the colonial history of Quispillaccta similarly notes that the actions of encomenderos provoked the rivalries between groups in the area during the colonial era. Machaca, “Vigencia y Continuidad,” 235. 358 Mumford, Vertical Empire, 121. 359 I describe this arrangement as two side-by-side reducciónes rather than one reducción because two different communities were created on either side of the Rio Chuschi by the reducción officials. Each town was built around a central square, with church and official buildings located facing the square. While Mumford notes that Toledo’s reforms intended to “marginalize” encomendero interests (Mumford, Vertical Empire, 127), Stern’s work highlights the self-interested competition among encomenderos in colonial Huamanga. Given that the rational solution to consolidating repartimientos in the area would have been to create one reducción, the unusual arrangement in Chuschi strongly suggests that local interests – including local encomendero interests - had a role in influencing the reducción process. 360 Isbell, To Defend, 65. A map of encomiendas in Ayacucho supports this analysis, showing encomiendas of Totos (which controlled Canas Indians) and the encomiendas of Quichuas-Aymaraes (with the Aymaraes Indians) next to each other in the western end of Vilcas. Miriam Salas Olivari, “La Ciudad-Región de Huamanga: de los tiempos prehispánicos a le era colonial y republicana inicial,” in Entre la Región y la Nación: Nuevas aproximaciones a la historia ayacuchana y peruana (Lima: Instituto de Estudios Peruanos, 2013), 58.

113

Local evidence confirms that both communities were moved to their current location from other areas nearby. Marcela Machaca notes that during the colonial reducción Quispillaccta was relocated to the lower zones of the community holdings – the area known as Qichwapampa or Qichapampa.361 Interviews with Quispillacctinos note that the location of the Quispillaccta Matriz moved east several times to its current location from the mountain side several kilometers to the west.362 Similarly, Isbell notes that the modern history of Chuschi maintains that the town centre was relocated to its current location from the plain of Calcabamba, which is a few kilometers to the east.363 Both communities eventually came to be located in the same small valley and claimed land on their respective side of the Rio Qunchalla/Chuschi that flows into the Rio Pampas. As a result of the relocation of the communities, conflicts continued for centuries, primarily in the grazing lands in the high puna above, where boundaries remained unclear and contested and where control of land and water rights was crucial for livelihoods, as we will see below in chapters 5 and 6.

Interestingly, members of both communities and peasants from surrounding communities maintain links to their pre-reducción kinship identities, both culturally and spatially. In interviews, members of both communities divide the towns and annexes in the immediate area according to their kinship ties and allegiances with either the Aymaraes or Canas (Chuschi or Quispillaccta). These relationships partly determined the axis of conflicts between groups throughout the history of the area, with the dominant axis of conflict between Chuschi and Quispillaccta reproduced in conflicts between annexes of Chuschi and Quispillaccta. To this

361 Machaca, “Vigencia y Continuidad,” 235. In terms of Andean production zones, the qichwa zone is the temperate area best suited for maize production, generally lower than 3600m, according to most authors. See diagram of Andean production zones. 362 Interview, Sept. 30, 2004, Quispillaccta, Peru. 363 Isbell, To Defend, 65. She notes that there is a chapel today in Calcabamba, but “no evidence of house structures or refuse indicating a nucleated village.” Oral history accounts cited by Isbell of the move to Chuschi’s present location explain that the move was instigated by the disappearance of a religious statue from a chapel in Calcabamba to the modern site of Chuschi. (Isbell, To Defend, 65). Sánchez Villagómez argues that the creation of this “legend” would have facilitated the consolidation of dispersed Indians by religious authorities, presumably during Toledo’s reducción reforms. Marté Sánchez Villagómez, "(Re)Pensando Lo Olvidado, Sendero Luminoso Y La Violencia Política En El Perú: El Caso de Las Comunidades de Chuschi y Quispillaccta Durante la Década Del 80," Maestría De Antropología tesis, Universidad Nacional Mayor De San Marcos, Setiembre, 2004, 30-1. Interestingly, an informant in Quispillaccta told me that religious authorities pushed the Chuschinos to move the town further west from Calcabamba to its present location, even though the Chuschinos knew the land was claimed by the Canas and that trouble would follow the move. Interview, 17 Oct., 2004, Quispillaccta, Peru.

114 day, Quispillacctinos identify their community by describing the main town (Matriz or Llacta, located across the river from Chuschi), along with 11 barrios – hamlets located in other parts of the community. Each hamlet is home to distinct kinship groups (ayllus), but all self-identify as part of the community of Quispillaccta: Soccobamba, Pirhuamarca, Llacctahurán , Yuraqcruz, Huertahuasi, Tuco, Cuchoquesera, Saint Jerónimo de Pampamarca, Catalinayuq, Puncupata, and Unión Potrero.364 In conversations, it is common for Quispillacctinos to identify themselves by their barrio and as Quispillacctinos – members of the community of Quispillaccta, with ties to a specific geographic area of the community. “We are Canas Indians; they are Aymaraes,” elders in Quispillaccta maintained.365 Chuschinos also readily admit their Aymaraes ancestry and their links to surrounding communities, but they no longer maintain a single corporate identity like Quispillaccta. The main town of Chuschi is often discussed in relation to its annexes of Uchuiri, Chaqolla, and Cancha Cancha, though the latter obtained status as an independent peasant community in 1964.366 Historical and ethnographic evidence indicates that Chuschi shares kinship relationships with several communities both south and north of the Rio Pampas, such Cancha Cancha, Uchuyri, Chacolla, Tomanga, and Sarhua.367 In fact, during the colonial period, Chuschi was the administrative centre for the parish of Chuschi, which included 7 villages: Chuschi, Quispillaccta, Cancha-Cancha, Huarcaya, Tomanga, Auquilla, and Sarhua.368 Informants say that Chuschi and its annexes have supported each other in conflicts with Quispillaccta because they share kinship ties.369 “With Chuschi, we are family; we are

364 Machaca, “Vigencia y Continuidad,” 19. A 1967 study of Chuschi lists 12 barrios for Quispillaccta, differing from the list above only by the addition of Qachir and using Portrero, rather than Unión Potrero. Ramón C. César et al., La Comunidad de Chuschi (Ayacucho, Instituto Indigenista Peruano, Sub-Proyecto de Investigación Zona Cangallo, Octubre, 1967), 23. Quachir and Portrero may have been combined to form Unión Potrero at some point in the past. Some studies refer to Qachi Portrero, while the Peruvian topographical map of the area refers to the hamlet of Portrero next to the hamlet of Jachiripampa. While many annexes likely refer to pre-Colonial settlements, some of the barrios are clearly more recent creations as Quispillacctinos settled and secured control of areas of the district in the twentieth century. This is discussed in more detail below in section 6.3. 365 Quoted in Doris Castillo Gamboa, "Conflictos Comunales Case: Quispillaccta y Cancha Cancha, 1940-1980," (Practica Pre-Profesional, Especialidad Historia tesis, Universidad Nacional de San Cristóbal De Huamanga, Facultad de Ciencias Sociales, 1997), 81. 366 Fanny Bolivar de Colchado, Estudio Distrital de Chuschi, in Victor Hugo Sarmiento Medina et al., Los Distritos de Vischongo, Chuschi, Concepcion, y Ocros (Ayacucho: Instituto Indigenista Peruano, 1967): 13. 367 Earls and Silverblatts, “Ayllus Y Etnias,” 161-66; Castillo Gamboa, “Conflictos,” 46. 368 Isbell, To Defend, 62. The last four, located south of the Rio Pampas (the village of Sarhua and its annexes Auquilla, Huarcalla, and Tomanga) were separated from District of Chuschi in 1910 with the creation of the province of Victor Fajardo. 369 Interview, Quispillaccta, Peru, 18 Oct., 2004.

115

Aymaraes,” notes a resident of Cancha Cancha.370 Over time, the Aymaraes that settled the area have developed separate administrative communities, particularly following colonial Spanish consolidations. Isbell notes a document from 1586 which describes the parish of Chuschi as composed of the villages of Chuschi, Cancha Cancha, Sarhua, and Moros.371 However, we should be careful to not over-emphasize the degree of ethnic distinctiveness between communities today in the area. It is simplistic to conclude that ethnic divisions between Aymaraes and Canas were the root of centuries of conflict between Chuschi and Quispillaccta. As the ethnographic work of Earls and Silverblatts indicate, there has been inter-marriage and mixing between kinship groups over the centuries, and this complicates any attempt to draw clear ethnic distinctions between Quispillaccta and Chuschi.372

In sum, it appears that Spanish conquest and the reduction strategy pushed competing ethnic groups to settle in the Qunchalla/Chuschi valley that serves as the current location of the town centres of Quispillaccta and Chuschi. Reducción forced disparate Indian kinship groups to leave their dispersed settlements, which ranged widely over the various eco-zones, from the high puna to the banks of the Rio Pampas.373 The actions of Spanish elites no doubt aggravated the conflict dynamics between different ethnic groups in this area. While intermingling of ethnic groups did happen at times, the close proximity and competition for scarce temperate maize lands ensured heightened conflict between the two communities for hundreds of years. With the population crash of the 16th and 17th century, pressure on temperate zones and high-altitude puna grazing lands in the area probably abated to some extent for many decades, until Mestizo elites began establishing small haciendas on community borders and community populations recovered in the 19th and 20th century.

While the Rio Qunchalla/Chuschi provided a clear boundary between Quispillaccta and Chuschi in the Qunchalla/Chuschi valley, boundaries in the higher puna zones remained unclear and in dispute. Unfortunately, there is no definitive evidence of the extent of Canas and Aymaraes settlement before Spanish conquest, or what the boundaries were between groups before the

370Quoted in Castillo Gamboa, “Conflictos,” 82. 371 Isbell, To Defend, 62. 372 Earls and Silverblatts, “Ayllus Y Etnias,” 165-6; and personal communication with John Earls, 29 Sept., 2011. 373 Machaca, “Vigencia y Continuidad,” 184-5.

116 upheavals of Toledo’s reducción policy and the influence of local encomenderos. The dispersed barrio settlements of present-day Quispillaccta give some indication of how ayllus may have settled the area in the pre-Conquest period.374 While Quispillacctinos maintain in oral testimony vigorous claims over these lands as original community lands and have struggled over the centuries to have these areas recognized as part of their community’s patrimony, many of the same areas remain contested by neighbouring communities like Chuschi and Cancha Cancha, as we shall see below in chapters 5 and 6. It is quite possible that claims by Quispillaccta and Chuschi overlapped in many areas, particularly in the grazing zones, possibly originating in overlapping repartimiento land grants.375 Mumford notes that Toledo’s reducción inspectors dealt with overlapping repartimiento grants in other parts of the highlands.376 Unclear boundaries or overlapping land titles were also common problems in the late 19th and early 20th century, according to Jacobsen.377 It is not unreasonable to surmise that a similar scenario in the decades after the chaos of Conquest served to aggravate relations between Chuschi and Quispillaccta.

The livelihood impacts of reducción in Chuschi and Quispillaccta are difficult to discern after all this time, and on top of the existing disruption of Spanish colonization, disease impacts, and Spanish forced labour drafts. Consolidation of the ethnic groups to the current town centres no

374 Marcela Machaca presents an interesting diagram of Quispillaccta’s barrios by altitude and agricultural zone. Most of the annexes are located in a band between 3500m and 4000m, which corresponds to the extreme upper end of the maize zone, and sits comfortably in the tuber and grain growing zone – the Suni or upper Qichwa zone. This is several hundred meters above the altitude of the town center that was formed following Toledo’s reducción. The settlement band of Quispillaccta’s barrios could be indicative of their pre-Hispanic settlement patterns, or it could be indicative of more recent population shifts in the 20th century driven by push and pull factors and climatic change in the latter half of the 20th century. My conclusion is that both explanations are probably correct – that many barrios harken back to the pre-Hispanic settlement of the area, but that resettlement of the area was also driven by more recent events for some barrios, as discussed below in chapter 7. Archeological and ethnographic evidence from numerous communities in Peru indicate that the Spanish frequently moved communities to lower elevations when enacting resettlement policies. See Stephen B. Brush, Mountain, Field, and Family: The Economy and Human Ecology of an Andean Valley (n.p.: University of Pennsylvania Press, 1977), 49-50; and Mitchell, Peasants, 7. This continued practices employed by the Incas, who also appear to have moved groups to lower elevations, partly as a consequence of their conquest of the area and partly as a result of their greater interest in growing maize. See Geoffrey O. Seltzer and Christine A. Hastorf, “Climatic Change and its Effect on Prehispanic Agriculture in the Central Peruvian Andes,” Journal of Field Archeology 17(4), 1990: 409-10. Gradual climatic warming in the Andes in the late 15th and early 16th century facilitated the movements, but were not the cause. 375 Overlapping repartimientos was not unusual, according to Mumford, because “a repartimiento was more of a social and political unit than a territorial one.” Mumford, Vertical Empire, 29. 376 Mumford, Vertical Empire, 121. 377 Jacobsen, Mirages, 237.

117 doubt increased travel times for peasants who farmed different parcels of land over a wide area or cared for animals in the puna grazing lands. Hiking from the current town centre (Matriz) to the high-altitude puna grazing lands takes hours. It is quite likely that some community members continued to live part of the year in the high zones in rudimentary huts to care for pasture animals, much as they continued to do until the mid-20th century, according to informants. While demographic collapse alleviated some pressure on cultivated lands in the area, reducción probably increased agricultural intensity on the lands located closest to the new town sites. This area is the prime maize growing zone, and informants in Quispillaccta indicate that the area known as Mollebamba, steep lands fronting the Rio Pampas and slightly west of the town centre, contain some of the oldest farm plots in the community.378

The ecological and livelihood impacts of increased agricultural dependence on cultivated lands around the new nuclear Rio Pampas settlements would not be felt until community populations began to recover to pre-Conquest levels in the early twentieth century. However, contemporary critics of Toledo’s reducción policy recognized that concentrating dispersed households could exhaust the ability of the land to sustain those resettled:

It is a question in all these kingdoms, discussed by many, whether it is right that the Indians’ pueblos be brought together and … large civilized pueblos created. Those who say no, and that it would be a great irritation for the Indians … give as the principal reason the large quantity of Indians and that the land is thin in valleys and places to plant and harvest food, and that being as there are now [only] ten or twenty houses together … they are no more than the land can sustain.379

Toledo brushed aside these concerns, noting that the land sustained many more Indians under Incan rule.380 However, Incan rule allowed for more dispersed settlement patterns and a wider mix of livelihood strategies that did not heavily rely on cultivated lands. The concerns of opponents to Toledo’s reforms would be partly realized centuries later, as discussed below in chapter 7.

378 Interview, Sept. 30, 2004; and Machaca, “Vigencia y Continuidad,” 21. 379 Mumford, Vertical Empire, 91. 380 Mumford, Vertical Empire, 91.

118

4.2.2. Mestizo elite and Catholic Church: impacts in the 19th and 20th Centuries

The second major influence on natural asset holdings in the history of Chuschi and Quispillaccta comes from the resource capture of cultivated and pastures lands by elites and Catholic Church authorities. Coming on top of the upheaval that accompanied Toledo’s reducción and the imposition of tribute burdens in the decades after conquest, resource capture directly impacted the livelihoods of peasants in Ayacucho for hundreds of years by completely removing some of the most productive maize and pasture lands from community hands. Landholding in Chuschi and Quispillaccta was much more equitable than other parts of the department, with few large haciendas and a majority of the land held by small holders and communities in common. These trends do not alter the fact, however, that elites in the D. Chuschi and in neighbouring districts appropriated some of the best cultivated and grazing lands, and these seizures had livelihood impacts on both communities. The capture of valuable natural assets heightened competition for remaining resources among communities in the area, particularly in the 20th century as populations recovered to their pre-Conquest levels and the communities began to integrate into the cash economy. The processes of resource capture and the structural factors that facilitated elite capture of lands in and around Chuschi and Quispillaccta thus aggravated conflict between the communities over time, particularly in the 20th century. Violence between Quispillaccta and Chuschi thus was not a simple case of competing communities fighting for a limited resource pie. Understanding the causes of conflict between Chuschi and Quispillaccta and reasons for the violence in the latter half of the 20th century requires an understanding of how resource capture impacted livelihoods in these communities.

4.2.2.1. Regional Structural Change, Patterns of Elite Capture, and Land Holding Patterns in Cangallo

Conflicts between communities and highland estates are a common theme in Peruvian history. They make up the essential narrative in the indiginista movement that began in the late 19th century to expose the exploitation of many peasants in Peru at the hands of the landed elite.381 In Ayacucho, wider national economic development trends and geographic reality conditioned the

381 Klarén, Peru, 245-6.

119 rise of landed estates in the department, and their development path during the colonial period and later following Peru’s independence in the early 19th century. Since colonialism, dynamic economic activity generally took place outside Ayacucho, in the mining centres of Potasi and Huancavelica, or later around urban centres of Lima or departments to the south. With the decline of mining in the neighbouring department of Huancavelica at the end of the 17th century and the dislocation and economic disruption that followed the wars for independence in the early 19th century, economic activity remained rather insular and fragmented in Ayacucho.382 The northern half of the department was the centre of large haciendas, with a servile population of hacienda workers (peons) controlled by a small elite class of Mestizo notables. Peons came from the ranks of those dispossessed of their land by the hacienda’s creation, often from neighbouring communities. Estate owners included the Catholic Church, former colonial military officers, and private citizens. Independent peasant communities in these areas had to navigate around the considerable influence of the landed elite, who also controlled the political levers of power in the department.383 Smaller estates – or fundos – occupied the small valleys surrounding the department capital Huamanga, and south to the high-altitude lands bordering the Rio Pampas and beyond – the location of Chuschi and Quispillaccta.384 The high-altitude grasslands of the southern half of the Department of Ayacucho, broken in places by deep river valleys, was home to few large haciendas. Instead, a small number of fundos existed in lowland areas, surrounded by traditional peasant communities. These estates employed few peons.385

The economic and political disruption caused by the wars of independence in the early 1800s severely stressed the landed estates in Peru’s Sierra. Economic activity was concentrated in the landed estate sector. Those who backed the Spanish lost their estates, while commercial dislocation after the war forced many others to sell their holdings to other estate owners.386 There was little additional usurpation of peasant community land during those first decades

382 Klarén, Peru, 139-40; Degregori, Ayacucho 1969-1979, 28. 383 Degregori, Ayacucho; Virgilio Galdo Gutiérrez, "Ayacucho: Economía y Sociedad (Siglo XIX)," (Ayacucho, Perú, Consejo Nacional de Ciencia y Tecnología, Universidad Nacional de San Cristóbal de Huamanga, 1988), 27. 384 Degregori, Ayacucho 1969-1979, 28. 385 Degregori, Ayacucho 1969-1979, 28; and Gutiérrez "Ayacucho: Economía y Sociedad,” 51. 386 Joanna E. Drzewieniecki, "Indigenous Politics, Local Power, and the State in Peru, 1821-1968" (PhD Thesis, State University of New York at Buffalo, 1996), 171-2.

120 following independence, as hacienda owners consolidated control over existing hacienda holdings.387 The post-colonial history in the area of the current D. of Chuschi remains obscure. However, the most significant policy development affecting rural communities in the early post- colonial period emerged from the limited agrarian reform of 1828. A new law gave Indians the right to obtain property titles for the lands they occupied and opened up the possibility of buying lands in the private market.388 This reform reaffirmed peasants’ rights to community lands and strengthened their claims on lands that had been taken over by the church or landed elite. This would also form the foundation for future claims of entitlement to contested lands by many communities. In Ayacucho, however, local elites often ignored these reform laws, and usurpation of peasant lands continued unabated, particularly in the north of the Department with their large lowland haciendas.389

In the late19th century, severe economic crises, compounded by the devastating impacts of the War of the Pacific, 1879-1883, gave way to a wool export boom. This led to a significant rise in wool production in the highlands, and an equally significant rush by foreign and domestic capital to seize the best lands from Indian communities for export production.390 In Ayacucho, however, the economic growth taking place in other parts of Peru was not repeated. While some rural estates were established in the late 19th century, Ayacucho began to see a long process of economic decline in the landed estate sector in the northern half of the department and the economic dislocation of the department’s trade patterns, as more dynamic economic poles to the north and south of the department accelerated the process of drawing economic vitality outwards from Ayacucho. Economic growth in Peru in the latter half of the 19th century, particularly from 1880 onwards, led to the slow stagnation of the northern half of Ayacucho and its agricultural and manufacturing centres.391 Without an economic sector of significant dynamism to engage foreign and national capital, economic activity in northern Ayacucho was displaced by development in other regions. Manufacturing declined and even wheat exports to Lima, the

387 Jacobsen, Mirages of Transition, 108. 388 Jacobsen, Mirages of Transition, 124-5. 389 Eric Mayer, "State Policy and Community Conflict in Bolivia and Peru, 1900-1980," (PhD thesis, University of California, San Diego, 1995), 260-1. 390 See Jacobsen, Mirages of Transition, chapters 5 and 6. 391 Degregori, Ayacucho 1969-1979, 28-9.

121 primary agricultural export of Ayacucho’s northern haciendas, began to lose ground to wheat producers in the fertile central valley.392 Southern Ayacucho did see the expansion of wool and cattle production in the highland grazing lands; however, trade patterns flowed south to Nazca, Ica, and Chala, and not to northern Ayacucho and the department capital.393 In fact, road penetration in the late 19th and early 20th century into northern and southern Ayacucho facilitated the export of economic activity out of Ayacucho because no dynamic economic centre existed in the department.394 This led to an increasingly fragmented Ayacucho, with the northern and southern sectors economically subordinated to other neighbouring departments, and economic flows pulling outward.395 The consequences of these developments for the central Ayacucho provinces of Cangallo (which includes the D. Chuschi) and Victor Fajardo in the 19th and early part of the 20th century were significant. Located in the isolated middle of the department, with lands crisscrossed by steep river valleys and towering high-altitude grasslands, and lacking adequate communication and road networks, these provinces found themselves in an economic “no man’s land” with some of the worst poverty in the country.396

Patterns of economic development in Ayacucho in the 19th and early 20th century described above thus conditioned the particular land holding structures in various parts of the Department of Ayacucho– large haciendas with many peons in the northern half of the department produced cereals or new crops like coffee, cocoa, or herbs in the moist eastern flanking jungles for Lima and the central valley. In the south of the department, livestock production dominated the highlands, as communities and small estates exported animals to the coast via southern routes. In the isolated middle of Ayacucho, peasant smallholder communities competed for limited arable land with a few estates, the peasants producing mostly for their own consumption, with few linkages to the wider national economy aside from periodic livestock sales. These patterns continued well into the 20th century and are reflected in land holding patterns during the crucial

392 Degregori, Ayacucho 1969-1979, 28-9. 393 Gutiérrez, "Ayacucho: Economía y Sociedad,” 54; and Degregori, Ayacucho 1969-1979, 28-9. 394 Degregori, Ayacucho 1969-1979, 29-30. 395 Degregori, Ayacucho 1969-1979, 30-1. 396 Degregori, Ayacucho 1969-1979, 30.

122 transformations of the 1960s and 1970s when various agrarian reforms were implemented in Peru.

The latter half of the 20th century witnessed the slow decline of the landed estate sector in northern and southern Ayacucho, a process finalized by the 1969 Agrarian Reform. Much of the land reverted to control of peasant communities. While it is difficult to pin-point the exact moment when large estates began their slide into economic decline, patterns of unequal landholding in Ayacucho that characterized the hacienda economy changed for good with the 1969 Agrarian Reform. According to the 1961 Peruvian census, those holding 500 hectares or more land made up only 0.2% of total landowners, but they controlled almost two-thirds of the productive land in Ayacucho – 57.6% of the cultivated area. By contrast, almost 32% of landholders were farming less than 1 hectare of land, working only 1.7% of the total land area in Ayacucho.397 Before land reform, a minority of large landholders in Ayacucho held most of the productive land, while many peasants had tiny holdings.

By 1972, several years after agrarian reform, little seemed to have changed in Ayacucho. Those holding less than 1 hectare of land accounted for 33.3% of the total landholders in Ayacucho, but they collectively held only 1% of the cultivated land in the department. By contrast, those holding more than 500 hectares of land accounted for merely 0.18% of total landholders in Ayacucho, but held 72.6% of the cultivated land in the department.398 However, simple comparisons of land holdings by size category do not directly speak to the degree of inequality in landholdings in Ayacucho at this time because, by 1972, more than 60% of the properties over 100 hectares in Ayacucho were actually held communally by campesino communities – principally as pasture land.399 Large landholdings that used to be held by a minority of landlords now belonged to peasant communities. Although differences in the census figures make a direct comparison with 1972 challenging, the 1961 census notes that less than 30% of the landholdings over 5 hectares were held communally or by campesino communities – 28.4% of the total land

397 Dirección Nacional de Estadística y Censos, Volume V, 216. 398 Richard Webb and Graciela Fernández Baca, Perú En Números 1999 (Lima: Cuanto, 1999), 25. 399 Degregori, Ayacucho, Raices, 179.

123 area in Ayacucho – compared to the more than 60% held by these groups by 1972.400 By the late 1970s, the process of land reform during the late 1960s and 1970s had gradually eliminated the old class of elite landlords with large estates that had dominated northern and southern Ayacucho since conquest. Land reform had increasingly transferred the land in these areas into the collective control of campesino communities.

In the Province of Cangallo, the process of land redistribution as a result of agrarian reform probably had less of an effect than elsewhere in Ayacucho because Cangallo had one of the lowest proportions of large estates among Ayacucho’s provinces, and a larger proportion of total land held by “middle peasants” – those working from 0 to 5 hectares of land.401 Compared to Ayacucho as a whole, Cangallo’s land distribution was much more evenly distributed both before and after Agrarian Reform. In 1972, after the 1969 Agrarian Reform law, 85% of the land owners in Ayacucho worked 5 hectares or less and farmed about 9.7% of the total land, while 2.6% of the

400 Strict comparisons are difficult because the two censuses counted landholding differently. In 1972, 58.5% of communal or community holdings were greater than 2500 hectares, while the 28.4% figure of campesino holdings in 1961 was all land over 5 hectares. So, the 1972 percentages would be even larger if we counted all community or communal holdings over 5 hectares. Further evidence for the general conclusion of a large increase in community or communal holdings comes from examining the percentage differences in total land held communally or by campesino communities between 1961 and 1972. By this measure, we see that large landholdings increased by more than 80% for campesinos. (This figure is calculated by finding the percentage increase between landholdings over 5 hectares held communally or by campesino communities in 1961 and comparing this with the area of land over 2500 hectares held by campesino communities in 1972. So, again, it is a conservative estimate.) See Dirección Nacional de Estadística y Censos, Volume 5, 227-228; Hector Maletta and Katia Makhlouf, "Perú, Las Provincias En Cifras, 1876-1981," (Lima: AMIDEP, 1987), 41. 401 Degregori labels landholders holding 1-5 hectares of land as “middle peasants” (campesinos medios). Degregori, Ayacucho, Raices, 84.

124

Figure 8

Upward slope of the line from left to right is a general measure of division in land holdings between large and small units. The greater the slope from left to right, the greater the inequality in favour of large land holders. Lines that are largely horizontal indicate greater evenness in landholding between large and small landholders. Data from: Hector Maletta and Katia Makhlouf, Perú: las provincias en cifras, 1876-1981, Volumen III, Estructura Agraria, Series Estadísticas No. 2 (Lima: Universidad Del Pacifico/AMIDEP, 1987.

land owners with holdings over 20 hectares farmed around 88% of the land.402 By contrast, the farming population of Cangallo was almost entirely made up of small or middle land-holders, and Cangallo’s landholding distribution figures were little changed from 1961 to 1972. In 1961 90.5% of Cangallo’s landholders held less than 5 hectares, working just over 40% of the total farmed area of the province. Those working landholdings larger than 500 hectares accounted for

402 Degregori, Ayacucho, Raices, Table 39, 179.

125

0.04% of landowners, farming 36.1% of Cangallo’s land.403 By 1972, about 92% of all the farm units in Cangallo were less than 5 hectares in size, farming about 44% of the total land, while less than 0.5% of the landholdings greater than 20 hectares farmed about 41% of the land.404 (See Figure 8: Large and Small Landholding as a Percentage of Total Agricultural Land in 1961 &1972.)

Of all the provinces in Ayacucho, Cangallo had the largest proportion of total land owners who were small holders or middle peasants.405 (See Figure 9 below.)

Figure 9: Department and Provincial Comparison of Percentage of Land Held by Those With Less Than 5H, 1972 100 90 80 70 60 50 40 30

Percentage Percentage of Total 20 10 0

Provinces Compared to Ayacucho Avaerage

Data: Degregori Ayacucho, Total Land Area Land Holders raices de una crisis, 179.

In addition, both before and after Agrarian Reform, Cangallo had the second lowest number of very large land holders in Ayacucho – those holding more than 500 hectares. In 1972, land holdings making up more than 500 hectares account for 37.38% of agricultural units in Cangallo,

403 Dirección Nacional de Estadística y Censos, Volume 5, 218. 404 Degregori Ayacucho, Raíces, Table 39, 179. 405 The rankings of those holding less than 5 hectares and more than 1 hectare are the same as the more general category of those holding less than 5 hectares.

126 the second lowest percentage in Ayacucho after the province of La Mar, which has a little more than 12% of the province in holdings over 500 hectares.406 And, by 1972, all the landholdings larger than 2500 hectares in Cangallo were held by campesino communities, meaning that they were part of campesinos productive resources in some way.

Figure 10

What this meant on the ground was that land was more evenly distributed in Cangallo than in other provinces in Ayacucho, with the possible exception of the province of La Mar. While this conclusion is largely based on land-holding patterns in the 1960s and 1970s, the patterns were probably consistent with historical patterns.

406 Degregori Ayacucho, Raices, Table 39, 179.

127

Historically, there were few large haciendas in the province, and even these had mostly disappeared by the time of the 1969 agrarian reform. Studies of the Province of Cangallo note that by the mid-twentieth century, haciendas could only be found in eastern Cangallo, on the west bank of the Rio Pampas, primarily in the districts of Concepción and Ocros, and to a lesser degree in the district of Vilcas Huamán.407 (See Figure 10) These haciendas inhabited the deep, warm valleys of the Rio Pampas, as the river spilled eastward into the warm, biodiversity-rich high-altitude jungle that rings the Amazon – the “ceja de selva” or eye-brow of the jungle. These haciendas produced sugar cane to make aguardiente and molasses, grew sweet potatoes, cotton, and cultivated fruit trees like apples, oranges, custard apples, and mangos.408 Some of these haciendas were in existence for centuries by the time scholars arrived in the area in the 1960s to study Cangallo. The Hacienda Pomacocha, for example, was established in the mid-17th century and run by the Monastery of Santa Clara de Huamanga until 1960 when the peasants of Pomacocha and its annexes invaded and took over the hacienda lands.409 The takeover of Pomacocha in the early 1960s signaled that the days of the landed estates were numbered in Cangallo, though haciendas continued to exist in large numbers in the province of Huamanga to the north until land reform in 1969.

To scholars and government officials, the dichotomy between northern and southern Ayacucho and Cangallo seemed clear in the 1960s and early 1970s: in northern Ayacucho large landed estates owned the bulk of land and largely conditioned unequal landholding relationships, while in Cangallo the few existing haciendas in the eastern part of the province were on the way out. Western Cangallo, including the D. Chuschi, appeared to be a stronghold of traditional, small-

407 Cesar Ramon Cordova et al., La Zona de Cangallo (Lima, Ministerio de Trabajo Y Comunidades, Instituto Indigenista Peruano, 1967), 41. 408 Rufino Gonzales Calderón, "Estudio Geo-Económico Social de la Provincia de Cangallo Del Departamento de Ayacucho," (Br., Fac. de Ciencias Económicas y Comerciales tesis, Universidad Nacional Mayor de San Marcos, 1962), 34. The Cordova et al. study lists the following haciendas in Cangallo: Pajonal, Ninabamba, Locería, Oquechipa, La Colpa, Ayrabamba, Santa Rita, Pakomarca, La Mejorada, Pirwabamba, Astania, Oqenay, Ibias, Matará, Pomacocha. Calderón lists some of these and also includes: Soccopa and Ccaccamarca. All these haciendas were located in the eastern districts of the province of Cangallo. 409 César Ramón Córdova, “Estudio de los Problemas en Pomacocha,” Ministerio de Trabajo y Comunidades, Instituto Indigenista Peruano, Ayacucho, Perú, 1968. It took until 1975 for community to finally take control of the former hacienda lands. See Ulpiano Quispe Mejía, Relaciones de poder y violencia política en las comunidades campesinas de Ayacucho, PhD Tesis, (Facultad De Ciencias Sociales, Universidad Nacional Mayor De San Marcos, 2011): 43.

128 holder farmers, who lived outside of the pernicious influence of landed elites, especially when compared to eastern Cangallo and the declining or disbanded estates in that region. (The D. Chuschi’s land holding pattern is discussed in detail below in section 7.3.2.) By 1972, the few large estates that had existed in the province and controlled almost a third of its productive land, were in the hands of campesino communities, or divided among the former estate workers.410 This narrative of Cangallo and the D. Chuschi as a stronghold of campesino smallholders who escaped the destructive clutches of landed estates has conditioned scholarly understandings of the area for decades and distorted our understanding of livelihood change in the area in the past 100 years, as discussed below in section 4.2.2.2.

4.2.2.2. Communities in Conflict? Revealing the Impacts of Landed Estates and Mestizo Elites in Western Cangallo

Data from agricultural censuses and the historical record appear to indicate that farming in Cangallo bore the legacy of the lack of large estates in the province over the preceding centuries. More so than any other province in Ayacucho, farming was done by middle peasants, working the land with fewer large estates around than in other parts of the department. This is partly a function of the terrain in Cangallo, with plenty of steep mountain peaks and fewer flat plains or valley zones that would have been attractive for large scale market agricultural production, and thus attractive targets for usurpation by elites. As well, the area is distant from markets, being located in the centre of the department, and even today 4-6 hours by car north or south of major urban areas either in Ayacucho or in neighbouring departments.411 Chuschi and Quispillaccta appear to be an ideal illustration of the landholding pattern in Cangallo, a fact picked up by researchers who entered the area in the 1960s. Ethnographic work in the 1960s highlighted the paucity of haciendas in the area, and made a point of noting the absence of pernicious influence

410 The number of campesino communities in Cangallo correspondingly increased from 1961 to 1981 as control of former estates was shifted to newly formed campesino communities, or communities received their official recognition. In 1961, there were 34 recognized campesino communities, which increased to 37 by 1972, and to 79 by 1981. Degregori, Ayacucho, Raices, 182. Interestingly, political conflict appeared much more severe in eastern Cangallo’s districts than in western Cangallo districts like Chuschi, given recent accounts by Heilman in Before the Shining Path. Political violence also appeared more severe in eastern Cangallo during the Sendero insurgency and the government’s counter-insurgency operations in the area, according to recent accounts by Theidon. Whether the presence or absence of landed estates in these two areas is key to these different trajectories remains an open question. However, detailed histories of the Sendero insurgency in western Cangallo remain to be written. 411 Elderly informants recall the trip by foot to the capital Huamanga to sell livestock or buy commodities, a grueling all-day hike

129 of elite landowner conflicts.412 In the early 1970s, for example, Isbell noted that “Chuschinos have not felt direct hacienda domination” because the nearest haciendas are fifteen kilometers away, in the neighbouring province of Huamanga.413 Her research noted only one conflict with a landowner in grazing lands on the north-west border of the D. Chuschi and the neighbouring District of Vinchos.414 (Discussed further below in section 5.3.2).

In light of the apparent equality of land-holdings, competition and conflict in western Cangallo communities like Chuschi and Quispillaccta has been characterized by scholars as occurring in the absence of landed estates – as a conflict over borders between communities of small holders – and fairly typical of patterns of conflict in Cangallo where landed estates were scarce. The conflict between peons and the hacienda Pomacocha was seen to be an exception in this province, and more characteristic of the types of conflicts common in northern Ayacucho. As Isbell was to later conclude, the D. Chuschi was “a region that had escaped many of the semi- feudal relationships of the hacienda system.”415 This was a significant conclusion for debates about whether semi-feudal land-holding relationships in Ayacucho contributed to the rise of Sendero Luminoso because it corroborated the conclusions of influential scholars like Degregori, who argued that Agrarian Reform had eliminated semi-feudal landholding relationships long before Sendero’s insurgency, thus playing little role in Sendero’s rise.416 For Isbell, Cangallo’s more egalitarian land-holding structure and paucity of landed estates probably explained why Sendero chose the D. Chuschi to begin its armed struggle against the Peruvian state.

I believe that Sendero chose Chuschi for its initial military operations precisely because of this absence of haciendas, which allowed SL to experiment with peasant communities that had strong communal structures, autonomy over their resources, and whose experiences with capitalistic market penetrations were minimal. By initiating their revolution in what they believed to be

412 Ramón C. César et al., La Comunidad de Chuschi, 28. 413 Isbell, To Defend, 51 & 181. Isbell did her first fieldwork in the D. Chuschi in 1967, before the Agrarian Reform law was put in place, returning to district in 1969 and 1974. Gorriti, in an influential history of the Shining Path’s early years, repeated Isbell’s conclusions about the lack of hacienda impact in the D. Chuschi. Gorriti, Shining, 18. 414 Isbell, To Defend, 240-41. 415 Isbell, “Shining,” 78. 416 Carlos Iván Degregori, How Difficult it is to be God: Shining Path’s Politics of War in Peru, 1980-1999, Steve J. Stern ed. (Madison WI: The University of Wisconsin Press, 2012): 114-15.

130

a region that had escaped many of the semi-feudal relationships of the hacienda system, Sendero perhaps hoped to avoid the mistakes made by the guerrillas inspired by the Cuban revolution who failed to gain the support of hacienda peons for their short-lived insurgency in 1965.417

A closer examination of the historical record of land conflicts around Chuschi and Quispillaccta, however, reveals centuries of conflict with landed estates in the area, disputes that continued throughout the twentieth century and into the 1970s and 1980s, after Agrarian Reform. Observers like Isbell noted some of these conflicts at the time, though the extent of their impact was not understood at the time. At a minimum, these new findings alert us to the role that land disputes played in conditioning local violence in the district, and possibly contributed to wider violence in Ayacucho. Reconstructing the historical record of land conflicts in the district reveals that several disputes over grazing and cultivated land did originate with hacienda encroachment or outright seizure of lands claimed by the communities as part of their historical patrimony. In several cases, mestizo elites seized lands from Chuschi and Quispillaccta for a string of small estates that used to sit on the northern border of the communities, throughout the Rio Cachi watershed. The southern holdings of these small haciendas or fundos encompass the bulk of the northern Sunni zone in the current district – those areas in this portion of the Rio Cachi watershed between 3500m-4100m (where cultivated pastures, tubers and grains grow), along with adjacent higher altitude grazing lands. (See yellow Sunni zone in the northern sector of the D. Chuschi, in Figure 13: Rio Cachi Watershed in Vinchos and Chuschi Districts.) These haciendas are marked on topographical maps created in the late 1950s, though many had disappeared by the late 1960s when researchers entered the area, having been taken over or bought out by nearby communities (including Quispillaccta), or simply abandoned by their owners. (The haciendas are marked on Figure 12: Major Land Disputes in the D. Chuschi in the 20th Century.) While it remains unclear when most of these estates were formed, their influence on the conflicts within and between Chuschi and Quispillaccta has not been adequately studied up to now.418 However, oral history, community records, and legal archives provide details of

417 Isbell, “Shining,” 78. 418 This may be a result of the level of analysis used by many scholars in the past. Most of the estates bordering Quispillaccta and Chuschi to the north sit in neighbouring provinces like Huamanga. Scholars have tended to study communities in the area by focusing on a province or district, so this approach may have obscured conflict patterns across district or provincial boundaries. The first studies of the area were done in the late 1960s by the Instituto

131 the disputes between Quispillaccta and Chuschi and neighbouring estates. These conflicts later figure into conflicts between Quispillaccta and Chuschi themselves, as will be explored below in chapters 5 and 6.

The reassessment of the influence of landed elites and their allies in and around the D. Chuschi described here dovetails with recent reassessments of the influence and impact of gamonales and mistis on local relations in the district in the decades preceding the Sendero uprising.419 Gamonales are the abusive provincial landowners who dominated local power relations for much of Ayacucho’s history. In Chuschi, wealthy local elites dominated the community’s authority structures and dominated the district’s positions of power until well into the 1960s and early 1970s.

As discussed in the literature review, scholars have debated the extent to which gamonalismo “or the old culture and methods of gamonalism” – abusive relationships by mestizo elites over peasants – was still influencing Ayacucho in the years before the Sendero uprisings, and the degree to which they contributed to the rise of Sendero.420 Degregori argues that in post- Agrarian Reform Ayacucho the old gamonalismo system had largely crumbled by the time Sendero engaged in violent insurrection, though “rural bossism (gamonalismo) … still existed [in] small local fiefdoms.”421 The implication of Degregori’s findings are that gamonlismo and the grievances it engendered had a limited influence as a cause of Sendero’s uprising, compared

Indigenista Peruano and by U.S. anthropologist Billie Jean Isbell. Both limited their focus to either the community, the district, or the provincial level. Given that many districts and provinces were created in the area in the mid or late 1800s, it appears that many estates and conflicts with neighbouring communities pre-dated the creation of district and provincial borders in the area. Hacienda owners may have even manipulated the location of district borders to further dilute claims or complicate legal challenges from communities like Quispillaccta and Chuschi who found themselves in a different province or district. 419 Marté Sánchez Villagómez, (Re)Pensando Lo Olvidado,; La Serna, The Corner of the Living. 420 Steve Stern, in How Difficult it is to be God, 16. Stern notes that, “Gamonalismo refers to rule by provincial landowners and their allied merchants, authorities, and intermediaries over Indian peasants and servants. The term evokes “feudal-like” relations of human ownership and physical abuse buttressed by ethnic hierarchies; the non- Indians, or mistis, who became petty or grand versions of gamonal masters included mestizos and “whites” of tainted social origin.” Stern, Shining and Other Paths, 16-7. In the D. Chuschi, according to Isbell, peasants did not use words like gamonales or mistis. Instead, those residing in the district were referred to as either comuneros (community members) or vecinos (neighbours). Quechua speakers referred to vecinos as qalas - literally, the “naked ones”. However, other recent research notes district residents using the term mistis, including research by La Serna, Sánchez Villagómez, and my 2004 fieldwork. The usage of misti discussed below in footnote 408. 421 Degregori, “Harvesting Storms,” 130.

132 to the impact of recent educational movements in Ayacucho and Sendero’s rigid ideological presence in the face of a weak and abusive Peruvian state presence in the Department.422

This conclusion may be changing, however, based on the findings of new research of local power relations in the years before Sendero’s uprising. As new micro-studies describe the local realities in communities in the decades before Sendero’s uprising, scholars are now appreciating the degree to which gamonalismo continued to operate in many communities up to the outbreak of the Sendero uprising, including in parts of Ayacucho like Cangallo where scholars have previously argued that there was a minimal presence of landed elites. Ulpiano Quispe Mejía’s recent study, for example, points out that all of Ayacucho’s provinces had similar local power structures – a majority of the best land was under the control of landlords, rural bosses, or medium landowners.423 In districts like Chuschi, Quispe Mejía argues, the elite were composed of landowners, medium landowners, and rich peasants who filled the positions of authority and acted as liaisons between the community/district and provincial, department, and central government elites. In addition to agricultural activities, these ‘mistis’ or ‘mestizos’, as they were known to peasants, engaged in the cattle trade, operated as local merchants, transport sector, and gradually came to enter the department’s public administration, particularly in the judiciary and education sectors.424 Isbell’s study of Chuschi from the early 1970s notes a similar power structure in the district. These mistis, or ‘Qalas’ in Quechua as community members in Chuschi referred to them, “did not participate in communal rituals and reciprocal community exchanges and did not define themselves first and foremost as Chuschinos, Isbell noted.425 In Chuschi, these elites dominated the community for decades. At their worst, they grabbed some of the best land, abused their authority positions, and tormented community members.426 While there were

422 Degregori, “Harvesting Storms,” 132; Degregori, “How Difficult it is to be God,” 114-5. 423 Quispe Mejía, Relaciones de poder, 66. 424 Quispe Mejía, Relaciones de poder, 66-7. 425 Isbell, To Defend, 71. Isbell notes that the derogatory usage of qala to refer to elites in Chuschi literally means ‘naked’ or ‘pealed’, in Quechua. “A Quechua-speaking informant told [Isbell] that qala used to refer to villagers who had gone away and come back wearing shoes instead of sandals. Then, when so many mestizos took up residence in Chuschi” and they didn’t participate in communal reciprocal traditions, the term qala was applied. The qala, Isbell notes, “have peeled off their indigenous identity.” In Chuschi, qala and misti are used interchangeably today, though Quispe Mejía ascribes the terms to different levels of elites. 426 La Serna discusses the recent history of some of Chuschi’s abusive mistis. See La Serna, The Corner of the Living, 108-119.

133 no abusive mistis or qalas in Quispillaccta427, Chuschi’s abusive elite similarly negatively impacted Quispillaccta by their control of the district municipal positions and the powers associated with these positions for many decades.428 Finally, local mistis within the district and in districts bordering the D. Chuschi had a significant negative impact on local livelihoods and on the conflict between Quispillaccta and Chuschi, aggravating conflicts over land holdings in the district for decades. The current study provides the most comprehensive discussion to date of this last aspect and compliments recent work by La Serna to explore the pernicious impact of mistis in the district.

The examination of the causes and consequences of resource and entitlement capture in the district begins with an analysis of elite appropriation of cultivated lands fronting the Rio Pampas, before exploring the long history of capture and conflict over high-altitude Sunni and Puna grazing lands in the northern half of the district.

4.2.2.3. Elite Capture of Scarce Cultivated Lands in the District of Chuschi

Good cultivated land in the district is not abundant, as will be explored in more detail in chapter 7. This fact meant that elite seizure of cultivated land probably had an outsized impact on

427 Degragori notes that “[m]istis is a Quechua term that refers to mestizos, especially those linked to traditional local powers.” Degragori, How Difficult it is to be God, 198, ft. 70. Interviews with Quispillacctinos note that there were no abusive ‘mistis’ or ‘qalas’ in Quispillaccta who took advantage of community members like in Chuschi. Some Quispillacctinos will tell you they never had mistis in their communities. Others will say that they had the assistance of some Mistis, for example, but that the Mistis in Quispillaccta were not abusive like those in Chuschi. In noting this fact, some Quispillacctinos tacitly distinguish between mistis or qalas – who are seen as abusive elites – and mestizos, those Peruvians from inside and outside the community who do not engage in all communal reciprocal traditions, but who also do not take advantage of Quispillacctinos with their activities. Arguedas, in his study of Puquio, Ayacucho in the 1950s, noted a similar distinction. See José María Arguedas, Puquio: A Culture in the Process of Change,” in José María Arguedas, Yawar Fiesta (Austin: University of Texas Press, 1985), 159. The 1967 IEP study (Ramon et al., La Comunidad de Chuschi, 75) states that, “Quispillacctinos note with pride that there are no “Qalas” in their community; they are all equal, indigenous comuneros, and that no one orders or commands us.” My translation. Ninety-year old Victor Galindo Tucno, whose mother was from Chuschi and whose father was from Quispillaccta, said that Quispillaccta’s Mistis included some from the family Romani and Mamerto Pariona; however, they were not abusive. Chuschino shopkeeper Victor Calderon Jimenez (himself from a well-to-do Chuschi family) agreed that Quispillaccta’s Mistis were not abusive, mentioning Lucho Romaní, Mariano Alcoscer, and Mamerto Pariona as Quispillaccta’s notable Mistis. Pariona was one of Quispillaccta’s Personeros – or elected community representatives – during the 1960 conflict with Chuschi. La Serna, The Corner of the Living, 68. Pariona was still living in Quispillaccta when I did my fieldwork in 2004. Though elderly, he participated in communal work parties, such as planting fields for the ancianos organization. His appearance was strikingly different from other Quispillacctinos – tall, fair, with a European facial structure. Unfortunately, he refused repeated interview requests. 428 La Serna, The Corner of the Living, Ch 3.

134 communities in the area, though it is difficult to determine the extent of this burden. The appropriation of some of the best cultivated and grazing lands in Ayacucho by mestizo elites and the Catholic Church constituted some of the most frequent and long-lasting disputes with the communities who laid claim to these seized lands. This legacy can also be found in Chuschi and Quispillaccta. At some point after conquest, the Church in Quispillaccta and Chuschi appropriated the yearly bounty of a number of cultivated land parcels in the district. Community members cultivated these parcels for the Church, and the crops harvested from these fields, along with livestock taken from similarly appropriated livestock herds, called cofradia herds, were used to support various Catholic festivals in the district. This continued for many decades.429 Overall, the impact of resource capture by the Catholic Church was greater in absolute terms in Chuschi with respect to cultivated lands than in Quispillaccta, particularly in the limited Quechua natural region below 3,500m.

429 I have not been able to establish when the Church grabbed entitlements to the production of specific cultivated lands.

135

Figure 11: Agro-ecological Zones in the District of Chuschi

Quechua lands are limited in the D. Chuschi. According to a recent survey of land use in the district, lower Quechua lands best suited for maize make up only about 9% of the usable agricultural land area of the district.430 (See Figure 11 & Table 1)

430 Asociación Bartolomé Aripaylla (ABA), Agricultura sostenible para la adaptación al cambio climático en regions vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, Ayacucho, Septiembre 2011: 19.

136

Table 1: D. Chuschi Land-Use According to Agro-Ecological Zone

Agro-ecological Altitude Agricultural Use Percentage of Area in Hectares Zone (m)** Agricultural Land (aka. Production Zone) Area (%)* Semi-humid Puna 3900-4500m Pasture, camelid 16 7872.8 husbandry, water catchment Semi-dry Puna 3900-4500m Pasture and high- 55 27062.8 altitude agriculture Suni 3600-3900m Rain-fed tuber and 12 5904.6 pasture, partial irrigated tuber and grains Upper Quechua 3400-3600m Agriculture in 6 2952.3 small areas Lower, arid 2400-3400m Irrigated Maize 9 4428.5 Quechua cultivation

* Source: ABA, Agricultura sostenible para la adaptación al cambio climático en regiones vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, Ayacucho, Septiembre 2011: 19. The table in the source does not specify whether these figures represent the percentage of arable land in the district, the percentage of land in use, or the percent of total district land area. I have used the total district area, 49,205 hectares, to convert these percentages to area in hectares in the column to the left. **Altitude designations were not originally included in the source from which this data is taken. There is significant variation in the Peruvian Sierra in assigning altitudes to different agro-ecological zones, depending upon the geographic zone one examines. However, the ABA study used agro-ecological classifications based on the work of Mario Tapia, so it is reasonable to assume that the authors used Tapia’s altitude classifications when calculating land area in different zones. As a result, Tapia’s corresponding altitudes for the zonal classifications for the south-central Sierra are used in this table. These zonal definitions are slightly different than those in the map of agro-ecological zones above. See Mario E. Tapia, Ecodesarrollo en los Andes Altos (Lima: Fundación Friedrich Ebert, 1996): 67.

This recent figure is probably higher than the historical average because climate change and peasant cultivation practices have altered growing patterns in the district, allowing maize, grains, and bean crops to be grown at increasingly higher altitudes. (Discussed in more detail in chapter 7)

Research in the late 1960s indicated that the Church was entitled to the harvest of 7-10H of irrigated maize lands in Chuschi, and about 1-2 H of land in Quispillaccta.431 This amounts to

431 Ramón C. César et al., La Comunidad de Chuschi, 27-8; Isbell 71-2. David Scott Palmer reported in 1972 that the church held 40H of irrigated land in Chuschi. David Scott Palmer, "Revolution from Above": Military Government and Popular Participation in Peru, 1968-1972," Unpublished Dissertation (Ithaca, NY: Latin American Studies Program, Cornell University, 1973): 258. Even accepting Palmer’s higher figure, however, the Church’s holdings of irrigated maize land is still only slightly more than 1% of total maize lands.

137 less than 1% of land in the lower Quechua zone of the district. The Church’s land-holdings in Chuschi were distributed in about 10-12 fields in the prime maize growing area below the village centre, while the Church’s Quispillaccta holdings were similarly south of the town centre and spread over about 5 fields.432 Corn grown in these fields was used to celebrate various saint day festivals in Chuschi and Quispillaccta.

Until the 20th century, when population levels returned to pre-Conquest levels, the impact on livelihoods in Chuschi and Quispillaccta from the Church’s cultivated land entitlements was probably modest. In the 20th century, demand-induced changes as a result of population growth in the district increasingly pressured the supply of good cultivated land in the Quechua zone (discussed in detail below in section 7.2). By the late 1960s, anthropologist Billie Jean Isbell noted that the Catholic Church was the wealthiest landholder in the district.433 However, by this time, ending the Church’s control over the irrigated maize land would have done little to alter the paucity of cultivated maize lands that had developed in the district. Distributing the Church’s holdings would have helped possibly a few dozen families produce more maize; a welcome impact no doubt for those who benefited, but modest overall given that there were over 1400 households in the district by 1940.434 While food produced from church holdings in the maize zone was distributed to community members during religious festivals, participation in the festivals entailed significant costs to some community members, a burden that may have hastened the decline of community participation in the Catholic Church festivals in the latter half of the 20th centuries as livelihoods faced increasing stress in the district.435 The symbolic

432 Ramón C. César et al., La Comunidad de Chuschi, 27-8. 433 Isbell, To Defend, 71. 434 Ramón C. César et al., La Comunidad de Chuschi, 19. This study notes that there were 1,428 households in the district according to the 1940 census. Isbell notes that each of the thirteen irrigated maize fields held by the Church in Chuschi was roughly equivalent to the total amount of cultivated land held by each family in Chuschi, about 1- 2H. Isbell, To Defend, 72 & 192. The church’s holdings in Quispillaccta amounted to less than 1% of total maize lands, if we use Coronel’s total of 240H of maize land in Quispillaccta. Given the expansion of maize lands up the altitude slope in the last 50 years, it’s possible that the church holdings may have been proportionately larger in the 1950s and 1960s. The expansion of maize lands is discussed in detail below in chapter 7. 435 One informant noted that sponsors of Saint-day festivals faced significant personal costs for the prestige of sponsoring, including killing animals and using precious harvests to feed fiesta participants. A 75kg sack of corn was processed for chicha (corn beer), and several animals were slaughtered for meat. At times, sponsor-ship burdens required community members to even sell land to provide food and alcohol for the fiestas. Over time, community members realized the high costs of participating in Catholic fiestas and participation declined, a process possibly accelerated and reinforced by increasing conversion of community members in the 1960s to Evangelical Protestant sects, which prohibited alcohol consumption and did not require expensive outlays for religious fiestas.

138 importance of prime irrigated land being controlled by outsiders was a long-standing source of grievance for many in the district.436 As the winds of change began to blow in the favour of small-holder farmers in the 1960s, community members in both Quispillaccta and Chuschi began to demand that the Church’s entitlement to these lands end. By the early 1970s, with Agrarian Reform sweeping Peru, both Chuschi and Quispillaccta succeeded in taking back control of church holdings of cultivated land.437 While some of this land continues to be used by the community to produce maize for Saint-day festivals (see photo below), other uses have been found by Chuschi and Quispillaccta. Some Chuschi lands formerly controlled by the Church were used for the construction of the road into Chuschi in the early 1960s, and the construction of additional schools in the Matriz in the 1960s and 1970s. Other parcels are now used directly for community benefit, such as providing maize for disadvantaged members of the community. (See photo below)438

Interview with Jacindo Tucno, Quispillaccta, 16 Oct., 2004. Isbell, writing in the early 1970s, noted that poor and wealthy comuneros seemed particularly attracted to Protestantism: “With the Protestant ideology of personal advancement, comuneros can escape the obligation of participation in the complex of reciprocity [between community members] and displays of generosity [including when sponsoring fiestas] that consume so much of their economic surplus.” Isbell, To Defend, 240. The relationship between the burdens of participating in the Catholic fiesta system and the conversion to Protestantism seems clear. Conversion offered relief from the increasing burdens of the fiesta system, particularly during the post-war period when pressures on land production were increasing and households were increasingly interested in other priorities like education and purchasing consumer goods. Mitchell came to a similar conclusion in his study of Quinoa, noting that people increasingly turned to Protestanism in Quinoa in the 1970s and 1980s as a response to ecological and economic forces. Mitchell, Peasants, 171-77. 436 Isbell, To Defend, 239. 437 Isbell, To Defend, 238-9; Palmer, "Revolution from Above," 258-9. 438 At a communal assembly in Quispillaccta in 1979, parcels in the maize zone formerly held by the church were formally assigned to various barrios in the community by lottery. Quispillaccta, Libro de Actas de las Assembleas, 1979-81, 20 Sept., 1979.

139

Above: Chuschi community members sowing the Señor de los Temblores (Lord of the Earthquakes) field with maize, in Chuschi, October 30, 2004. Until the early 1970s, the Catholic Church controlled numerous fields like this one in the irrigated maize zone of Chuschi and Quispillaccta. This field is about 0.5H in size, according to field measurement. The community continues to use the harvest from this field for community fiestas. In the foreground stands former community President, Juan Carhuapoma Llalli. Photo by author. Below: Communal planting in Quispillaccta of an irrigated maize field that used to be part of the Church’s entitlement in Quispillaccta. When the community regained control of the land, this parcel was passed to the barrio of Huertahuasi, which assumed responsibility for planting the field. Some of this field’s yield is subsequently provided to the elderly citizens association in Quispillaccta, whose members are seen here planting maize. October 20, 2004. Photo by author.

140

While the impact of resource capture of maize lands by the Catholic Church may have been modest on its own, this impact was amplified by resource capture from mestizo elites of high- quality irrigated land in Chuschi. The presence of mestizo elites in Chuschi probably dates back to the colonial period, given Chuschi’s prominence among neighbouring villages as the parochial centre of the Curato de Chuschi.439 Village priests were outsiders, assigned to the district by the Catholic Church. For centuries, they wielded significant power in the village through their control of spiritual matters, the religious-prestige hierarchy, and their management of religious festivals.440 Chuschi’s designation as the headquarters of the newly created D. Chuschi in 1857 no doubt helped cement the presence of mestizos in Chuschi, because they occupied the key administrative posts in district as Governor and Lt. Governor until the 1960s.441 The position of Municipal Mayor in Chuschi was bureaucratically even more important in the district, given the mayor’s role in allocating and managing the budget, village development, and maintaining order.442 As with the district administrative positions, mestizos dominated the position of Mayor until the early 1970s because it required that the office-holder be able to speak and write Spanish, skills rare for peasants in Chuschi and Quispillaccta until the late 1960s or early 1970s.443 No records exist of mestizo elites holding maize lands in Quispillaccta. As noted above, informants indicate that Quispillaccta never had mistis living in the community like Chuschi. In Chuschi, however, mestizo resource capture of good cultivated lands is first documented in the 1830s, when the hamlet of Cancha Cancha launched a lawsuit against neighbouring Chuschi, claiming that mestizo elites with the support of community authorities appropriated about 16H of land in the important maize zone around Callcabamba, an area that lies between Chuschi and Cancha Cancha.444 Given the close kinship ties between indigenous community members in Chuschi and Cancha Cancha, this lawsuit no doubt shows that indigenous community members in Cancha Cancha were litigating against the mestizo elites then

439 An 1802 survey of Ayacucho listed 3,379 Spanish and Mestizos in the Province of Cangallo, the second lowest level in the Ayacucho zone. Cited by Galdo Guttiérrez, "Ayacucho: Economía y Sociedad,” 6. 440 Isbell, To Defend, 84-89. 441 La Serna, The Corner of the Living¸ 108. District authorities reported to bureaucratic authorities in the provincial capital, Cangallo, and to the Prefect in the department capital, Huamanga. 442 Isbell, To Defend, 88-89. 443 Isbell, To Defend, 71; La Serna, The Corner of the Living, 108. 444 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 42.

141 controlling Chuschi and complicit in the seizure of cultivated lands around Callcabamba. Such litigation was in fact fairly common in Ayacucho at the time.445

By the 1830s, we thus have documentary evidence of an active mestizo presence in Chuschi, creating hardship and land problems for indigenous community members. Mestizos likely further moved into the area after 1870 with the rise of new classes of commercial landowners and businessmen in the southern half of Ayacucho, fueled by the wool boom and economic linkages to the coast.446 In Chuschi, holdings of prime maize land by mestizo families compounded the appropriation of maize land by the Catholic Church. By the late 1960s, 4 mestizo families owned about 12-13 hectares of land in Chuschi.447 In an interview, Ernesto Jaime, a prominent mestizo authority in the 1960s and 1970s, reported that his father held about 5 hectares of land.448 While these mestizo holdings were probably distributed throughout different ecological zones in the district, Isbell notes that one mestizo family that controlled 3-4 hectares of land held most of that land in the maize zone.449 Some mestizo holdings were even organized into fundos.450 A topographical map of the D. Chuschi produced in the 1950s shows one of the likely mestizo holdings, the Fundo Choccechanca, which sits in prime irrigated maize land south of the town centre and fronting the Rio Pampas. (See Map: Topographical Close-up of Chuschi District Maize Zone North of Rio Pampas)451 Most of the production from mestizo holdings was destined for sale in the local or regional market452, thus removing the production

445 Gutiérrez, "Ayacucho: Economía y Sociedad,” 53. 446 Degregori, Ayacucho1969-1979, 30. Jose Jaime, the wealthy uncle of prominent mestizo authority Ernesto Jaime, came to Chuschi from southern Ayacucho, according to informants. Interview with Victor Calderon Jimenez, Chuschi, 2 Nov. 2004, Field Notes, Book 2. 447 Ramón C. César et al., La Comunidad de Chuschi, 31. The holdings were in between 50-60 yugadas, which their study defines as about ¼ hectare. 448 Interview, Ernesto Jaime, Chuschi, 18 Oct., 2004. Jaime indicated that his father held about 20 yugadas of land in Chuschi, about 4 times the holding of the average community member. See details of average land holdings in the next chapter. 449 Isbell, To Defend, 72. 450 Interview, Ernesto Jaime, Chuschi, 18 Oct., 2004. 451 I have not been able to find any information about who owned the Fundo Choccechanca. No evidence of this fundo existed when I conducted my field work in the area in 2004. It could be one of the fundos referred to by Ernesto Jaime in our interview. 452 Isbell, To Defend, 72.

142 from the village. These holdings were modest compared to total maize lands in the district, accounting for less than 1% of the total lower maize zone.

The livelihood impact of maize land appropriations by mestizos, like the Church appropriations, was probably modest until the middle of the twentieth century. Even with the Church holdings added, the cumulative livelihood impact of distributing this land to community members would have only benefited a small percentage of households in the district. The impact was not equally felt throughout the district, however. The resource capture of maize lands primarily impacted peasants in Chuschi and Cancha Cancha, tying up dozens of parcels of prime land close to the main village settlement. The material impact was thus greater on Chuschinos than on Quispillacctinos, and heightened the symbolic grievance of outsiders controlling prime land as livelihoods became increasingly squeezed in the latter 20th century. The manner of appropriation no doubt further heightened grievances against mestizo elites in both communities. While a detailed account remains elusive about how mestizo elites came to hold prime maize lands in Chuschi, it most likely arose through a combination of extra-judicial actions, seizures, or forced sales from the late 1800s.453 This was the typical pattern in the southern highlands. That the holdings were so small by the mid-twentieth century demonstrates the isolation of the region, and the paucity of good maize lands for elites to capture.

4.2.3. Conclusion

Landholding in the D. Chuschi appeared relatively egalitarian and free of elite domination, according to scholars who first studied the district in the late 1960s and early 1970s. Although only a few abusive misti elites remained in the district, this chapter outlined the disruptive long- term legacy of elite capture of cultivated and grazing lands going back to the pre-Conquest period that altered peasant livelihoods in important ways, focusing settlements around relatively scarce cultivated lands in the Rio Pampas zone. Some of the best Rio Pampas cultivated lands were also captured by local elites and the Catholic Church following Spanish Conquest, further marginalizing local livelihoods. Chapter 5 picks up this story to illustrate the impact of elite

453 Community members in both Chuschi and Quispillaccta noted that the sale of land to people outside of the community was forbidden, given that the communities held communal title over land and distributed usufruct rights to control land only to community members. While it’s not clear when this rule was formalized, its existence highlights the illegal nature of mestizo and Church land holdings in the minds of community members.

143 capture of highland grazing lands in the district and the decades long conflicts by communities in the district to regain control of these lands, conflicts that increasingly pitted communities in the district against each other.

144

Chapter 5 Elite Capture of High-Altitude Pasture Lands in the District of Chuschi Introduction

The borders of the District of Chuschi were largely fixed when foreign and Peruvian anthropologists entered the area in the late 1960s and early 1970s to study the district’s communities. Those early studies noted continuing disputes between Chuschi and Quispillaccta over borders between the two communities in three areas: grazing lands just north of the town centre, in the north-east corner of the district, and in the north-west corner of the district.454 (See Figure 12) These disputes were seen as part of the long-running rivalry between the two communities, stretching back hundreds of years, and reaching a particular intensity in 1960 where a clash left dead and injured on both sides.455 However, reconstruction of the historical record of grazing land control in the district shows that the late 1960s and early 1970s represented the culmination of decades and even centuries of effort by indigenous community members in both Quispillaccta and Chuschi to reclaim lands – especially grazing lands in the northern parts of the district – that had been usurped since the colonial era by mestizo elites and the Catholic church. Reconstructing both the efforts of communities to regain lands they felt belonged to them and reconstructing the disputes that these efforts provoked provides important insights into the sources of grievances and perceived injustices at the heart of conflicts between district communities in the decades preceding Peru’s dirty war. This history not only helps to explain the evolution of current population settlement patterns in the district today, but provides partial explanations for why later government repression in the area during the Sendero insurgency fell especially hard on Quispillaccta. Chapter 5 sets the scene by describing the district’s high-altitude livestock zone and outlines the importance of herding to household livelihoods. After piecing together the early history of elite capture of grazing lands in the district, this chapter then focuses on disputes in the north-west and north-east portions of the

454 Ramon et al., La Comunidad de Chuschi, 26. 455 Isbell, To Defend, 49, 65-6.

145 district, leaving the account of disputes for grazing lands in the centre of the district for chapter 6.

Source: Instituto Geográfico Nacional, Perú.

Figure 12: Major Land Disputes in the District of Chuschi in the 20th Century

Livelihoods and the high-altitude grazing zones of the district

The D. Chuschi is largely a zone of high-altitude grazing lands. While most of its cultivated lands sit in precariously steep river valleys fronting the Rio Pampas in the extreme south of the district, the district’s essential character is defined by vast expanses of semi-arid highland grazing lands in sharply peaked mountainous terrain between 3600m and 4800m. While there is considerable variation throughout the intermountain valleys of Ayacucho in terms of precipitation, temperature, evaporation, soil quality, and vegetation type, the lower part of this grazing zone marks the upper limits of the heavily exploited woodland zone where cultivated crops predominate. The zone then transitions to the tundra or alpine meadow zone above the tree

146 line known as Puna. In the Puna zone frosts can strike 9-12 months of the year, temperatures generally average in the single digits Celsius, and rainfall is variable, depending upon the local relief. During the early months of year (the Andean summer) rain or fog may predominate, but hail is not unknown.456 Vegetation has adapted to these harsh conditions to hug the ground “in rosette or ‘cushion’ form for protection from the cold” on well drained flat areas. Sloping surfaces find a variety of bunch grasses growing that are used by animals and humans alike, while in poorly drained areas “puna peat moors” or seasonal ponds – bofedales – predominate.457

Today, these high-altitude lands in the district are growing areas of population settlement and principally hold communal and family herds of livestock for both communities and their annexes. According to Table 1 above (D. Chuschi Land-Use According to Agro-Ecological Zone), grazing lands in the D. Chuschi over 3600m make up over 80% of the land of the district (Puna and Sunni area in blue and yellow respectively in Rio Cachi Watershed map, Figure 13). Limited cultivated agriculture is also (increasingly) practiced in this zone, in the slightly lower elevation Sunni lands, which cover the zone from approximately 3600m to 4100m (yellow zone on map, figure 13).458 The yellow Sunni zone in the north-east of the district in the map below, Figure 13, falls within the Rio Cachi watershed, while the yellow Sunni zone directly north of the town center of Chuschi is referred to as the Kimsa Cruz area and drains south into the Rio Pampas. In the northern half of the district, the Rio Cachi watershed flows north into the District of Vinchos and becomes the Rio Chicllarazo. Entering into Vinchos, the watercourse flows into a highly desirable, lower elevation Quechua production zone.

456 Kent V. Flannery, Joyce Marcus, and Robert G. Reynolds, The Flocks of the Wamani: A Study of the Llama Herders on the Punas of Ayacucyho, Peru (New York: Academic Press, Inc., 1989): 16-17. 457 Flannery et al., Flocks, 18-19. 458 The Map, Rio Cachi Watershed in Vinchos and Chuschi Districts, was created by combining portions of two maps of the natural regions in Vinchos and Chuschi Districts. See: Perú, Departamento de Ayacucho, Plan Para el Uso Sostenible de los Recursos Naturales Altoandinos Y Prevencion de Los Efectos Climatologicos Adversos del Distrito de Chuschi del Departamento de Ayacucho, N.P., 2006: 105; and Perú, Departamento de Ayacucho, Plan Para el Uso Sostenible de los Recursos Naturales Altoandinos Y Prevencion de Los Efectos Climatologicos Adversos del Distrito de Vinchos del Departamento de Ayacucho, N.P., 2006: 85.

147

Figure 13: Rio Cachi Watershed in Vinchos and Chuschi Districts

Source: Perú, Departamento de Ayacucho, Plan Para el Uso Sostenible de los Recursos Naturales Altoandinos Y Prevencion de Los Efectos Climatologicos Adversos del Distrito de Chuschi del Departamento de Ayacucho; and Perú, Departamento de Ayacucho, Plan Para el Uso Sostenible de los Recursos Naturales Altoandinos Y Prevencion de Los Efectos Climatologicos Adversos del Distrito de Vinchos del Departamento de Ayacucho.

Throughout the Sunni lands of the district, cultivation of highland tubers and variable cultivation of grains is common but risky. In parts of the Rio Cachi watershed the land is relatively flat and suitable for planting potatoes and other Andean tubers, broad beans, and cereals. Maize cultivation is negligible because of frequent frosts. In fact, in the higher areas of Sunni zone of the Rio Cachi watershed all cultivated agriculture is risky because of frequent frosts. This area is

148 instead primarily used for livestock production.459 In the lower Puna zone from 4100m-4800m the only possible cultivation is the production of bitter potatoes, which are soaked and freeze- dried to produce Chuño. Instead, the Puna zone is primarily used for livestock grazing lands.

Today, much of the Rio Cachi watershed in the district is controlled by Quispillaccta as grazing lands for sheep and cattle. Quispillaccta also controls the high-altitude grazing lands around Tuco to the south-west, in the Puna between 4000m and 4800m. These grazing areas are too cold for raising cattle or sheep, but provide ample pasture for community and individual herds of llamas and alpacas.460 Chuschi and its associated barrios also control large areas of high Puna grazing lands in the north-west of the district, from Chicllarazo to the north-western borders of the district, and in the eastern Puna zone of the district, between Kimsa Cruz and the neighbouring district of Maria Pardo de Bello to the east. Depending upon the altitude and landscape, cattle, sheep, llamas or alpacas graze in these zones.

In the past twenty years, climate change has been altering production possibilities in the highland grazing areas of the district. Machaca reports that some farmers have successfully grown maize in test fields in the Rio Cachi area for more than ten years, a feat previously unheard of, given the extreme risk of frost and hail in the area.461 As well, during field work in 2004, I observed broad bean fields in Tuco at an altitude around 4200m, production that was historically never possible at such a high-altitude. Numerous comuneros indicated in interviews that the weather has changed considerably in recent decades. The Puna zone, for example, has warmed considerably in recent decades to allow vegetation and crops to grow in areas that were previously too cold for such plants.462

459 Machaca, “Vigencia y Continuidad,” 30, 60, 177-8. 460 Machaca, “Vigencia y Continuidad,” 30. 461 Machaca, “Vigencia y Continuidad,” 125-6. 462 Interview with Roberto Human Machaca, Nov. 1, 2004; interview with Dionisio Conde, Tuco, Nov. 4, 2004. Conde, who was 70 when interviewed, noted that there were few bushes and no trees in Tuco when he was young. Now, however, they’ve expanded tremendously throughout the higher zone. Interview with Policarpo Casavilca Quispe, Tuco, Nov. 5, 2004. Quispe indicated that in Tuco’s increasingly warm climate snow falls are now a rarity when they were once very common. The warmer climate has allowed Quispe to plant broad beans for the past five years, where they were once unknown in Tuco.

149

Alpacas graze in Tuco, Quispillaccta, at an altitude of approximately 4200m. Cultivated fields visible on slopes in the background. Photo: Author, Nov. 5, 2004, Tuco, Quispillaccta.

150

On the side of the mountain are fields of recently planted broad beans in the barrio of Tuco at an altitude of approximately 4200m, in the Puna zone. Tuco is the highest barrio in Quispillaccta. Changes in Tuco in recent decades illustrate how warming has expanded cultivated agriculture possibilities in high, cold zones in the District to areas that were previously too risky for cultivated crops. In the foreground are walled-off fields for grazing livestock. The division of the puna into walled livestock fields has expanded considerably in recent years. The area has also seen the increasing growth of large shrubs and trees, which did not exist in past decades according to elderly informants. Photo: Author, Tuco, Quispillaccta, Nov. 5, 2004.

For communities in the D. Chuschi, livestock are raised as a household risk management and investment strategy, rather than a food source for consumption. While wool is used from sheep and llamas, animals are rarely killed for food. Some animal products are regularly used by households in the district. A district study in the late 1960s found that half of the milk and cheese production and all of the wool production from livestock were consumed by families, and the remainder sold.463 More importantly, Machaca notes that livestock represent wealth, an emergency fund, and a source of savings for community members and for the community as a

463 Genaro Colchado A., La Ganaderia en Los Morochucos, Maria Parado de Bellido y Chuschi (Ayacucho: Instituto Indigenisto Peruano, 1968), 70. Unsurprisingly, this study, which was focused on assessing the uses of animal products in these districts in Cangallo and the limitations to increased commercialization of animal production, found numerous limitations to the commercialization of animal production in the area. The study failed to note the use of livestock as a risk management strategy for households and instead viewed them as under-utilized.

151 whole.464 In fact, herding is now the principal economic activity for households in the district; the size of one’s herds, rather than the amount of cultivated land held, traditionally represented the key indicator of wealth among households.465 Herding is a common part of household livelihood strategies in highland Andean communities. Stephen Brush notes that in the community of Uchucmarca, in the north central Andes, “beef, cheese, and milk are only rarely consumed in the village, but over half the village households keep a few head of cattle as a living bank account on the hoof, that can be converted to cash by selling to … itinerant merchants…”466 In the D. Chuschi itinerant cattle merchants have long visited the communities to buy cattle. Community members also sold animals themselves in Ayacucho if they needed cash or goods.467 Household studies in the Andes indicate that herding requires more energy and is less productive overall than cultivated agriculture; however, herding is less risky given crop hazards and variability and more dependable for low rates of return for households.468 Benefits to households are compounded by using children as herders, further saving adult time and energy for other livelihood activities.469

Equally important is the community role served by livestock herds. Each community in the district held community herds.470 For decades, most of these herds were controlled by the Catholic Church (discussed further below in section 5.3.1). However, herds that escaped church control or which were later recovered from the church were often used to help finance communal projects. As well, individual households might be asked to sell an animal from their household herds to help finance a community project. Machaca notes that at various times in the past,

464 Machaca, “Vigencia y Continuidad,” 177-8. 465 Isbell, To Defend, 200. Isbell’s research was primarily focused on Chuschi, but this conclusion is probably true for Quispillaccta and the other communities in the district as well. Land-holdings do matter for determining wealth and status, however, as Isbell herself notes in her research See page 74. In my own interviews with comuneros in both Chuschi and Quispillaccta the amount of land held was clearly a part of their determination of whether or not a community member was wealthy. In recent decades, economic activities in the district have changed somewhat, making assessments of wealth somewhat more complex. 466 Brush, Mountain, 79. 467 Castillo Gamboa, “Conflictos,” 30. She notes that these cattle purchases began in the community in the 1940s. 468 Mayer, Articulated, 17 469 Mayer, Articulated, 18. I also observed the use of children as herders among households in the Quispillaccta barrio of Llacctahurán during my fieldwork. 470 In Quispillaccta some of the community herds were divided among various barrios in 1972, when control of the communal herds was recovered from the Catholic church. Machaca, “Vigencia y Continuidad,” 179, 183.

152

Quispillacctino households were asked to sell one animal each to help finance the purchase of former hacienda lands or to pay legal bills in on-going boundary disputes.471 Livestock herds have also been actively used by communities to guarantee the sanctity and security of community boundaries. Machaca notes that this is one of the oldest and most important roles of the community’s livestock.472 Communities recognized that possession is nine-tenths of the law, both in terms of ensuring no one usurped their grazing lands, and in helping to legitimize their own disputed holdings.473

The history of elite capture of high-altitude grazing lands in the District of Chuschi

It is unsurprising that communities in the district developed strategies to forestall encroachments on their important grazing lands because there is a long history of usurpation of community livestock and grazing lands in the area. In some cases, grazing lands and entitlements to those lands were captured outright by elites, while at other times community members were forced to care for livestock herds while the benefits of those herds were captured by elites. This long history of resource capture pre-dates Spanish conquest of Peru. Machaca notes that Quispillaccta’s community herds descended from herds managed by the community but designated for use by the Incas and the Sun deity.474 Spanish conquest led to the capture of much of the grazing lands and herds in what is now the northern half of the district. This capture went hand-in-hand with the reducción policy of Spanish Viceroy Francisco Toledo, which concentrated the district’s population in the town centres near the Rio Pampas. Reducción probably reduced the importance of pastoralism as a household livelihood strategy and shifted much of household efforts to cultivated agriculture throughout the areas where it was implemented. This not only facilitated control of the population and freed more peasant labour

471 Machaca, “Vigencia y Continuidad,” 185 & 236. 472 Machaca, “Vigencia y Continuidad,” 179-80. 473 This impulse has continued until recent decades. As recently as 1981, Quispillacctinos are recorded in communal meetings encouraging their fellow community members to inhabit the land near their boundaries in order to safeguard it for the community. Quispillaccta, Libros de Actas de las Assembleas, 1979-81, March 16, 1981. 474 Machaca, “Vigencia y Continuidad,” 185-6.

153 for producing tribute, but it also facilitated the capture of livestock opportunities for the Spanish.475

In the decades following Conquest, Spanish and mestizo elites captured much of the traditional grazing lands of the communities in the area, establishing a string of small haciendas in the northern grazing zone of the district. Almost all of the haciendas were located north of the current district, in what is now the district of Vinchos and the neighbouring District of Chiara; however, they also controlled extensive grazing and high-altitude cultivated lands in what is now the northern half of the D. Chuschi.476 The haciendas were largely situated in valuable Quechua and Sunni lands close to the Rio Apacheta, Rio Chicllorazo, and Rio Allpachaca. These three rivers continue north and merge into the Rio Vinchos.477 (See map) While the detailed history of how Spanish elites seized much of the grazing lands in the area in the post-Conquest period remains unclear, by the early 1800s most of the high Sunni and lower Puna zone of the current district had been incorporated into various haciendas. Communities now concentrated in the Rio Pampas zone still had access to some puna grazing areas, but now grazing opportunities were severely constrained and faced interference from elites. For much of the next century, communities in the district faced constant battles with the haciendas over control of these lands, and among themselves over efforts to regain control of these grazing and high-altitude cultivated lands. As the hacienda economy declined in the area between the late 19th century and the mid- 20th century, the conflicts shifted to battles between communities over efforts to gain control of hacienda lands as owners sold or abandoned their holdings.

The Rio Chachi watershed and the grazing and cultivated lands leading from this area to the north-west are prime lands in the area, apart from the area immediately fronting the Rio Pampas. Today, the area is also the site of a large regional water project, Proyecto Rio Cachi. Built in the late 1990s, the Rio Cachi hydrological project provides a significant portion of water and

475 Mayer, The Articulated Peasant, 268; Watters, Poverty and Peasantry, 57. Mayer writes: “The colonial rationale behind reducciones was to reduce family access to production zones, limiting them to those required for minimum subsistence levels in order to expand available labor time for tribute.” 476 In one case, the Hacienda Santa Catalina, the hacienda maintained a dwelling in the current district boundaries, in what is now the barrio of Catalinayuq in Quispillaccta. Machaca, “Vigencia y Continuidad,” 236. 477 Detailed information about the Rio Cachi basin can be found in Gobierno Regional De Ayacucho, Proyecto Especial “Río Cachi”: Memoria (Ayacucho: Perú, 2006). The key haciendas are noted on Figure 12: Major Land Disputes in the District of Chuschi in the 20th Century.

154 electricity to the capital city of Huamanga.478 These lands in the north-east of the current district were the source of much of the conflict in the area from the mid-1800s to the late twentieth century. By the early 1800s, the total area of the Rio Cachi watershed was controlled in several small estates by the mestizo hacienda-owning families Villavicencio and Ruiz.479 The land around the current Quispillaccta barrio of Catalinayuq, called Ñawpallacta by Quispillacctinos, was controlled by the Villavicensio family in the Hacienda Santa Carolina.480 The Hacienda Kikamachay controlled lands around Unión Portrero and Pillcoccasa.481 Lands around the current barrios of Pampamarca and Cuchoquesera were controlled by the Ruiz family who controlled the Hacienda Putaje.482 At the western edge of the Rio Cachi basin, sat the Hacienda Yaruca, controlling lands at the north-west edge of the district and abutting Chuschi’s high- altitude grazing lands. By the 1920s, this hacienda came under the control of Emilio Del Solar and Manuela Flores.483 Other nearby haciendas like Milpo, Ingahuasi, and Allpachaca similarly controlled lands in the basin that communities long considered their patrimony. These haciendas and others like it in the area cultivated tubers and raised livestock like cattle, sheep, horses, and mules.484

478 Quispillaccta ceded 392.88H of land in Cuchoquesera and Pampamarca to build the Rio Cachi reservoir, significantly impacting communal grazing operations in these barrios. Some residents complain, however, that the Ayacucho government failed to keep promises made in exchange for allowing the construction of the reservoir, for irrigation and public works projects in the area. In fact, some argue that the project increased poverty in the area, while providing benefits for communities downstream. Asociación Bartolomé Aripaylla, Yakumama – Madre Agua Lagunas de Lluvia y Comunidades Criadoras del Agua (Ayacucho: ABA- Ayacucho, 2014): 94-5; 102-3. 479 Machaca, “Vigencia y Continuidad,” 185. 480 Machaca, “Vigencia y Continuidad,” 236. 481 Machaca notes that old dwellings and corrals in Unión Portrero indicate that the land was controlled by mestizo elites for about 7 or 8 generations, dating mestizo control in that area to the early 1800s. It is not clear who owned Kikamachay (also spelled Quicamachay). According to Castillo Gamboa, “Conflictos,” 51, Fundo Pillcoccasa was controlled by the Vivanco family, who sold the land to Luis Humberto Vasallo. He later sold the land to Quispillaccta, as discussed below in section 5.3.3. 482 Machaca, “Vigencia y Continuidad,” 236; Castillo Gamboa, “Conflictos,” 52. This area was known in the colonial era as San Jerónimo de Pampamarca and San Juan de Cucho Quesera. At some point in the 20th century, the Hacienda Putaje was controlled jointly by the Ruiz and Garcia families. Spelling for local place names varies according to the source and can thus be confusing. For example, “Putaje” is also spelled as Putaqa or Putacca; “Kuchukersera” is known as Cuchoquisera or Cuchoquesera. Where possible, I try to use the names and spellings most commonly used by the communities in the district. 483 Miguel La Serna, "Los huérfanos de la justicia: Estado y gamonal en Chuschi antes de la lucha armada," in Entre la región y la nación: nuevas aproximaciones a la historia ayauchana y peruana, Roberto Ayala Huaytalla ed., (Lima: Instituto de Estudios Peruanos, 2013), 248-9. 484 Gutiérrez, "Ayacucho: Economía y Sociedad,” 49.

155

The limited agrarian reform in 1828 and the decline of the hacienda economy in Ayacucho (discussed above) appeared to have led to efforts by communities in the district to regain control of lands lost to haciendas in the area. As noted above, the reform allowed Indian communities to buy and hold title to land for the first time. For hacienda owners in the area, both local and regional economic trends were working against their continued existence. The small haciendas in the centre of the Department, such as those in the Rio Cachi basin, became increasingly isolated in the region as economic activity increasingly oriented outward from Ayacucho to the north and south. At the same time, local hacienda operations appeared to be suffering from mismanagement, making them increasingly uneconomical. Machaca notes that unsustainable increases in herd size and little concern for overgrazing by the mestizo-controlled haciendas in the Rio Cachi basin degraded the pasture lands of haciendas, leading estates to be increasingly marginal economic operations.485 According to Machaca, this spurred the land owners Villavicencio and Ruiz to propose selling parts of their holdings.486 The local ecological and economic decline of the haciendas in the area helps to explain why some mestizo elites in this area were increasingly getting out of the livestock business in the late 1800s and early 1900s, while in other parts of highland Peru, elites were pouring capital into the livestock industry as the wool boom took off and actively dispossessed many communities of their best grazing lands. As increasingly marginal haciendas disappeared in Ayacucho throughout the first half of 19th century, conflict trends began to shift away from community-hacienda conflicts to conflicts between communities over borders and land holdings. This trend that scholars have identified throughout Ayacucho played out in the D. Chuschi well into the 20th century.487

485 Machaca , “Vigencia y Continuidad,” 185. Jacobsen, Mirages, 308-310, notes that it was common for livestock estates to pasture more animals than ecologically or economically viable on estates in the southern highlands at the time. “Under conditions of high risk, insecurity, and disputed rights to resources, the efficient estate maximized livestock capital, even if this approach lowered productivity,” Jacobsen writes. He notes that this often led to a downward economic spiral on many estates at the time, particularly impacting small to medium estates. The small estates in the Rio Cachi basin may have similarly suffered. 486 Machaca, “Vigencia y Continuidad,” 185. It is not clear when Ruiz holdings were sold. 487 Gutiérrez, "Ayacucho: Economía y Sociedad,” 56.

156

5.3.1. Community struggles to regain control of captured high-altitude grazing lands

While the record remains incomplete on how the communities in the area regained control of the Sunni and Puna lands, by the 1840s Quispillacctinos appear to have first seized the opportunity to buy the land of the declining haciendas – land they always considered to be their lands.488 In 1848 Quispillaccta began the process to purchase the lands of the Hacienda Santa Catalina from the Villavicensio family, a sale that was completed in 1870.489 Each community member in Quispillaccta contributed money or livestock to gather funds for the sale.490 In the decades to come, Quispillaccta purchased other hacienda-owned lands in the upper Rio Cachi watershed, including the lands around the hamlet of Unión Portrero – controlled by the hacienda Quicamachay, and land around Puncupata, Ccochapampa, Chontalla, and the hamlets San Jerónimo de Pampamarca and Cuchoquesera – all controlled by the hacienda Putaje.491 By the early 1920s, Quispillaccta had purchased a string of former hacienda lands in the upper Sunni regions of the Rio Cachi watershed, lands which now make up most of the north-east sector of the D. Chuschi. Not only were these some of the best pasture and high-altitude cultivated lands

488 Machaca, “Vigencia y Continuidad,” 236. Machaca notes that a variety of lands in the Rio Cachi watershed occupied by the haciendas Santa Catalina, Putaqa, Quicamaychay, Inga Wasi, Allpachaca, and others in the area were sought to be recovered by Quispillaccta, meaning that the community believed that they had some historical entitlement to these lands. 489 “Promesa de Venta,” 1848 y 1849, in Reproduccion textual y mecanográfica de los títulos de propiedad de los terrenos del pueblo de San Juan de Quispillaccta, Toma 3, 1927: 34-5; “Escritura de Ratificación en una escritura de venta que una compra-venta hecha por Don Mariano Villavicensio, a favor de la comunidad del pueblo de Quispillaccta, de la Hacienda Catalina, situada en los confiles del pueblo de Chuschi,” 1886 y 1887, in Reproducción textual y mecanográfica de los títulos de propiedad de los terrenos del pueblo de San Juan de Quispillaccta, Toma 3, 1927: 40-1. Both documents in Quispillaccta community archives. 490 Machaca, “Vigencia y Continuidad,” 185 & 236. Machaca notes that each community member contributed 300 pesos and 300 soles to help purchase the Hacienda Catalina. 491 Machaca, “Vigencia y Continuidad,” 236. I have not been able to find specific details about when these purchases were made. Castillo Gamboa, “Conflictos,” 51-2, notes that Puncupata, Ccochapampa, Cuchoquesera and Chontalla were lands originally controlled by the Hacienda Quicamachay, but later usurped by the Hacienda Putaje before being sold to Quispillaccta. She also says that San Jeronimo de Pampamarca was purchased from the Garcia family. Later documents and court proceedings note their close association with the Ruiz family of the Hacienda Putaje. Salomón Galindo said that Pampamarca was once the hacienda San Geronimo de Pampamarca; but known to Quispillacctinos as Pataquisira. In the 19th century the area was controlled by the hacendado Antonio Sotomayor until he was driven out of the area by the Chileans during the War of the Pacific with Chile in the early 1880s. He never returned. Interview, Salomón Galindo Achallma, 77, Cuchoquesera, 28 October, 2004. A similar account, also relying in part on Salomón Galindo, can be found in Oseas Núñez Espinoza in an online history of Cuchoquesera, , accessed 9-1-16.

157 in the area, their location in the head waters of the Rio Cachi meant access to important sources of water for irrigation in the semi-arid highlands of Ayacucho.492

Unfortunately, these purchases failed to solidify their entitlements over what the community believed were their traditional lands. Nor did it stem the illegal seizure of community assets and lands by mestizo elites. Soon after beginning the purchase of the Hacienda Catalina in the late 1840s, Quispillaccta established a community herd of livestock on the newly acquired lands of Catalinayuq. By maintaining community herds on purchased lands, Quispillaccta sought to generate funds to benefit the entire community and also hoped to ensure the security of the newly acquired lands.493 Similar community herds were established in the barrios of Cuchoquesera and Tuco in the decades to come.494 In 1890, however, the Catholic Church used its ecclesiastic powers to capture much of the benefits of Quispillaccta’s community herds. Community herds in Catalinayuq and Tuco were unilaterally declared as cofradia or brotherhood herds for the Church – herds designated for a particular Catholic Saint that would be controlled for the benefit of the Archbishop of Ayacucho. The local parish of Chuschi was instructed to manage these herds, while the community members in Quispillaccta were forced to continue the day-to-day care of the herds for the Church.495 The Church had now stepped into the power vacuum created by the disappearing haciendas to seize assets and entitlements from communities.496

The Catholic Church’s presence in the D. Chuschi dates to the colonial era. The burden of its exploitation and abuse of the community members in the district was a long-held grievance for many peasants. Accounts of the settlement of the current Matriz of Chuschi and Quispillaccta

492 Gutíérrez, "Ayacucho: Economía y Sociedad,” 52, notes that many of the conflicts over pasture lands in the Rio Pampas area were also about controlling irrigation possibilities for these pasture lands. 493 Machaca, “Vigencia y Continuidad,” 183 & 185. 494 Machaca, “Vigencia y Continuidad,” 183. 495 Castillo Gamboa, “Conflictos,” 29-30 & 53; Machaca, “Vigencia y Continuidad,” 186. Gamboa notes that a mestizo shop keeper in Chuschi helped to manage the Church’s cofradia. Machaca notes that the newly acquired lands and community herds were “awarded” to the Saints, and in this way, the lands and animals came under Church control. Castillo Gamboa notes that the herds in Catalinayuq benefitted Santos Elias and Santa Teresa; herds in Tuco were designated for Santos Marcos and Virgen del Carmen, while a third in Qiwilla were designated for Santos Lucas and Virgen del Rosario. I have been unable to locate Qiwilla or any additional information on this herd. 496 This may have been a trend in several parts of Ayacucho. Records indicate that there were six cases of litigation between communities and priests between 1886 and 1895 in Ayacucho, and none in the preceding years. Guttíérrez, "Ayacucho: Economía y Sociedad,” 53.

158 indicate the crucial role of the Catholic Church. The construction of a chapel in Chuschi’s village centre in 1554 was key to the relocation of the communities as part of the reducción strategy employed by Spanish authorities in the mid-16th century. A similar chapel was built in 1555 a short distance away, in what was to become the town centre of Quispillaccta.497 The existing Catholic church in Chuschi was built in 1774, and Chuschi became the colonial ecclesiastic centre for several villages in the surrounding area.498

Capturing the entitlements of Quispillaccta’s newly purchased Sunni and Puna lands was no doubt an additional blow to the community following the colonial reducción policy. Two of the three community herds in Quispillaccta were no longer available for the community’s needs. Only the small communal herds in Cuchoquesera escaped appropriation by the Church.499

The Church similarly captured entitlements from the community herds in Chuschi and its annex Cancha Cancha for decades. Like Quispillaccta, community members in Chuschi and Cancha Cancha also struggled for decades to recover and maintain entitlements to herding lands threatened by elites from the Hacienda Yaruca and Milpo respectively. (Discussed below in section 5.3.2) The capture of some of these entitlements by the Catholic Church was thus another galling imposition of elite resource capture to comuneros.500 The Church had long controlled a cofradia from Chuschi’s community herds in the puna of Chicllarazo, appropriating

497 Isbell, To Defend, 65; Municipalidad Distrital de Chuschi, Reseña Histórica del Distrito, web page, accessed 4 Oct., 2013, http://www.munichuschi.gob.pe/portal/ciudad/reseniahistorica.html. 498 Isbell, To Defend, 65. 499 Machaca, “Vigencia y Continuidad,” 186. A 1923 legal complaint, discussed below in footnote 490, notes that there were only 48 head of cattle in Cuchoquesera when they were seized by the Hacienda Putaje. This is small compared to the hundreds of animals that were kept in barrios like Tuco. 500 It is not clear when the Catholic Church began to control Chuschi and Cancha Cancha’s community herds – the herds that became the cofradia herds of the Church. In 1972 Chuschi expelled the Catholic priest from the community and seized the cofradia for the community. In an interview with Isbell, the priest indicated that he was “mystified” about the reasons for his expulsion, believing that he had served the community well for fifteen years. He also characterized the seizure of the Church’s herd as stealing from the Church, claiming that documents proved Church control of the herd since colonial times. “However,” Isbell noted, “comuneros felt that the Church had illegally taken what had belonged to their ancestors. The community was therefore receiving what had been theirs all the time.” Isbell, To Defend, 239. Isbell was probably speaking to Father Carlos Eulogio Chávez Abad, who was the parish priest in Chuschi from 1963 to 1972. His abuses and conflict with Chuschinos are discussed in Miguel La Serna, “In Plain View of the Catholic Faithful: Church-Peasant Conflict in the Peruvian Andes, 1963-1980,” Hispanic American Historical Review 95(4), 2015: 631-657. As with other La Serna writings, his heavy reliance on legal petitions as documentary records of historical events in the district should be accepted with caution.

159 the benefits of the herd until the early 1970s.501 In the late 1960s, one study assessed that Chuschi’s cofradia herd consisted of approximately 1000 head of sheep and 250 to 300 cattle. Sales of livestock were used to finance repairs to the local chapels and to provide food and drink for saint day festivals throughout the year.502

The Catholic Church’s control of community herds throughout the district continued well into the 20th century, and no doubt significantly impacted the well-being of the communities in the district. Valuable assets were taken from community members at a time when rising population in the district (discussed in detail below in section 7.2) and the increasing penetration of market forces were presenting new challenges and opportunities for the communities as the 19th century waned and in the first decades of the 20th century. While individual families often still controlled their own small family livestock herds in many communities, the continuing drain of resources from the cofradia herds removed assets and labour that could have been used to deal with community challenges in the early to mid-20th century, like the lack of potable water, limited education resources, and the paucity of income for investment in the communities and their agricultural activities – needs subsequently identified when outside experts entered the communities in the late 1960s.503 The Church’s resource capture also heightened competition for the remaining Sunni and Puna lands in the area, particularly in areas where hacienda land had yet to be recovered by communities, and on the boundaries between communities where overlapping claims to the same land had yet to be resolved. This will be discussed below in the remainder of chapter 5 and in chapter 6.

The recovery of some former hacienda lands by the early 20th century did not end the exploitative practices and outright resource capture by those remaining haciendas in the area. At various times communities in the district continued to deal with robbery and illegal seizure of their lands and livestock by the remaining hacienda elites in the area. These cases of resource capture similarly threatened limited community resources and drained scarce monies from

501 Ramon et al., La Comunidad de Chuschi, 27-8; Castillo Gamboa, “Conflictos,” 34-5. It is not clear where Cancha Cancha’s cofradia was kept. Cancha Cancha had a long-term grazing presence around Choccoro, in an area north-west of Unión Portrero, straddling the northern boundaries of the district, and this may have been the location of their cofradia herd. 502 Ramon et al., La Comunidad de Chuschi, 27-8. 503 Ramon et al., La Comunidad de Chuschi, 152.

160 community coffers because of the associated legal expenses. These cases demonstrated that communities in the district continued to struggle directly with haciendas or the legacies of area haciendas well into the 20th century. For example, in January, 1923, authorities in Quispillaccta charged that Don Manuel Ruiz Condomino, the owner of the Hacienda Putaje ordered his men to invade the Quispillaccta barrio of Cuchoquesera and capture the community’s herds.504 Quispillaccta’s Cuchoquesera herd was the only community herd that escaped Church control. While Quispillaccta’s livestock were led to the neighbouring Hacienda Putaje, Ruiz ordered Putaje’s livestock to be pastured freely throughout Cuchoquesera. It took a judicial complaint and days of pleading by Quispillacctinos to reverse the invasion and illegal seizure of their communal herd, though Ruiz kept several heads of cattle seized from Quispillaccta.505 Incredibly, legal proceedings that sprang from this land invasion would wind their way through courts in the Department for 49 years before being settled in Quispillaccta’s favour in 1972. Chuschi similarly confronted the abuse of hacendados during this period. In the early 1920s, the Del Solar family controlled the Hacienda Yaruca and began to extend the boundaries of Yaruca at Chuschi’s expense by forging land titles and seizing lands in the extreme north-west of the district, around the areas of Rumichaca.506 The lands in dispute include high-altitude grazing lands but also some of the only Sunni and low Puna lands in the area. Chuschi’s struggles against the Del Solar hacendados would last even longer than Quispillaccta’s battles with the Ruiz family, with unresolved legal issues remaining until the late 1990s.

By the early decades of the twentieth century, both communities thus still struggled with landowners to reclaim assets they considered to be their traditional patrimony, while feeling the pinch of the Church’s capture of community herds. Chuschi’s dispute with the Hacienda Yaruca

504 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 46. 505 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 48-9; Castillo Gamboa, “Conflictos,” 46. 506 La Serna, Los Huerfanos, 249. La Serna notes that the Del Solar family took control of the hacienda in the early 1920s. However, information from Isbell’s account of the conflict with the hacendados suggests that the same family controlled the hacienda for many decades before the 1920s. She relates that in 1975 the son of Emilio Del Solar, who controlled Yaruca in the 1920s, claimed that the land had been purchased by his great grandfather. Isbell, To Defend, 241. If correct, this suggests that Yaruca was in the Del Solar family for two generations prior to the 1920s. The precise boundaries of the disputed territories cannot be located on maps. Topographical maps from the 1950s show the Hacienda Yarua, which is likely the hacienda controlled by the Del Solar family. Isbell notes that the disputed lands were near Niñobamba, which can be found at the extreme north-west tip of the district’s boundaries on the map.

161 would continue for decades and consume considerable community resources, as discussed below in section 5.3.2. It threatened their main area of high-altitude grazing lands in the northwest of the district. For Quispillacctinos, too, Putaje’s invasion of Cuchoquesera was a significant threat to their community’s only communal herd and grazing zone in the Rio Cachi watershed. Members of both communities must have worried that the actions of the hacendados were a prelude to the complete usurpation of their grazing lands, given how the pattern of invasion conformed to the typical way that hacienda usurpation of land happened throughout the southern Sierra. Jacobsen notes a description of how violent hacienda usurpation typically happened in Azángaro province, in the Department of Puno at the turn of the century: The grabbing of land from Indians begins with the act of daily placing cattle and mules belonging to the usurper in the pastures and cultivated fields of the Indian. In this the colonos and employees of the latifundista use force, and they proceed to kill the few head of livestock of the Indian for their own consumption. Alternatively, they drive the Indian’s livestock to the latifundia’s central building complex, [kill the animals there], and distribute most of the meat and hides among themselves, while reserving the carcasses of the best-fed animals for the patron (hacienda owner) or as gifts to the provincial authorities. … In the following days the looting of the Indian’s hut begins with the object of weakening his economic situation. This goes on until, under the pressure of this display of force, the owner decides to sign the bill of sale. As a sale price they receive a small sum in money or kind according to the whim of the land grabber.507

While Quispillaccta’s success in recovering Cuchoquesera and most of their cattle illustrates how the fortunes were declining for some haciendas in the area by the mid-1920s, both communities would continue to struggle for decades against hacienda abuses in the area.

Struggles with the Hacienda Yaruca and Hacienda Putaje in the 1920s took place in a period of remarkable unrest in the southern highlands, presaging the decline of the hacienda economy in the coming decades. The wool boom of the late 19th century and the price spikes sparked by increased wool demand during World War One set off a land grabbing frenzy at the turn of the century in the southern highlands.508 This accelerated peasant militancy in the southern

507 Jacobsen, Mirages, 236. 508 See Jacobsen, Mirages, chapter 6.

162 highlands to their highest levels in over a hundred years in the decade between 1915 and 1925.509 Instability in wool markets following WW1 and into the 1920s exacerbated the social unrest in the southern highlands. “A ‘seismic wave’ of rebellions and other forms of peasant resistance engulfed nearly every highland province of Puno and Cuzco departments between 1920 and 1923.”510 Ayacucho similarly witnessed “a broad range of rural mobilizations against abusive authorities in rural Ayacucho[,]”511 and the spread of a wide-ranging indigenous rights political movement in the Department – the Tawantinsuyo movement.512 In Cangallo, the Tawantinsuyo movement “coalesced into a broad program of action and protest throughout much of the province,” particularly in eastern district of Carhuanca in 1923, where people rebelled against abusive authorities, unfair taxes, and lack of indigenous rights.513 The rising unrest and growing awareness of the ‘Indian problem’ pushed the Peruvian central government to undertake some of the most extensive reforms in over 100 years, recognizing Indian communities in the constitution in 1920, establishing a Bureau of Indian Affairs in 1921, creating investigative commissions to examine the unrest in the southern highlands, and increasingly inserting themselves between clashing highland peasants and landlords – all of which dovetailed with its own commitment at the time to modernizing Peru’s capitalist state.514 As the unrest worsened in the early 1920s – partly stimulated by the promise of reform that remained largely unfulfilled – the government grew increasingly worried that peasant unrest was getting out of control. A violent backlash was unleashed by the Peruvian military and local elites on peasant protesters and communities in the

509 Jacobsen, Mirages, 337. 510 Jacobsen, Mirages, 344; Watters, Poverty, 245. 511 Jaymie Patricia Heilman, "Under Civilian Colonels: Indigenous Political Mobilization in 1920s Ayacucho, Peru," The Americas 66, 4 (April 2010), 502. 512 Jaymie Patricia Heilman, Before the Shining Path: Politics in Rural Ayacucho, 1895-1980 (Stanford: Standord University Press, 2010), chapter 2. 513 Heilman, Before the Shining Path, 47. Though Heilman argues that the Tawantinsuyo protest movement was present in much of Cangallo, her book actually only offers evidence of unrest in the eastern part of the province, far removed from the District of Chuschi. 514 Thomas M. Davies Jr., "Indian Integration in Peru, 1820-1948: An Overview," The Americas 30, 2 (1973), 195; Watters, Poverty, 245; Jacobsen, Mirages, 345. Some of the government’s initiatives were a terrible step backward, however, such as its road building program throughout the country that forced all men between 18-60 to build roads in the highlands – a program that essentially reinstated forced labour for road building and killed thousands of mostly indigenous Peruvians in the 1920s. See Davies Jr., “Indian Integration in Peru,” 198-199.

163 southern highlands, killing thousands.515 In eastern Cangallo, provincial authorities called in the Peruvian military, which put down the Carhuanca rebellion with brutal force, killing many, looting and burning homes, and stealing livestock.516

While the Peruvian state was still unwilling in the early decades of the twentieth century to overturn the power balance among landlords and peasants in the countryside, modest reforms in the rights of indigenous communities in 1920s were somewhat expanded by legislative changes in the 1930s. These changes codified the formal recognition of indigenous communities, guaranteed the protection of community lands, promised a legal code, and required representation in local government.517 By the early 1940s, Peru’s institutions were increasingly recognizing and codifying the existence of indigenous communities in the highlands. The process of officially recognizing indigenous communities included submission of information about community land holdings. In the context of overlapping and contested claims by neighbouring communities and large land owners, however, extending official recognition potentially risked inflaming long-running land disputes in the highlands.

In the D. Chuschi, these government reforms appear to have had two important impacts. First, the changes encouraged Quispillaccta and Chuschi to obtain formal community recognition in the 1940s, no doubt with the hope that recognition would help to consolidate control over existing land holdings and claimed land holdings. This helped provoke over forty years of direct conflict between Chuschi and Quispillaccta over the high Sunni and Puna lands directly north of the Matriz, including very violent inter-community clashes in the 1960s, 1970s, and early 1980s. Second, both communities sought to expand their grazing land holdings through additional purchases of lands in the Rio Cachi basin. Quispillaccta’s August, 1942 purchases of lands in the north-east corner of the current district set off decades of conflict with the hacendados of Putaje and the peons and peasants working these lands, many of whom had kinship links to Chuschi. Chuschi’s attempt one month later, in Sept., 1942, to purchase a large portion of the hacienda Yaruca also began decades of legal battles that remained unresolved to the present day,

515 Jacobsen, Mirages, 347-8. More than 2000 peasants were killed in the two Cuzco provinces of Huancané and Azángaro alone, according to Jacobsen. 516 Heilman, Before the Shining Path, 50. 517 Davies Jr., “Indian Integration in Peru,” 201-2.

164 both with the owners of Yaruca and with the hacienda peons who later claimed the land. The details of these conflicts are reconstructed below, along with an account of deeper background causes in later chapters that exacerbated the conflicts in the high-altitude zones of the district, such as the causes of population movements to the area in the mid-20th century.

5.3.2. Chuschi’s Yaruca dispute on the north-west borders

Chuschi’s aborted purchase of portions of the Hacienda Yaruca appears to have worsened the already troubled relations between Chuschi and hacienda owners, the Del Solar Flores family, setting the stage for several decades of conflict over grazing lands in the north-west of the district. Frustrated by decades of conflict with Yaruca, Chuschi decided in 1942 to buy back control of almost half of the land controlled by the hacienda from Emilio Del Solar, the family patriarch. In addition to useful high-altitude grazing land, the area under dispute included some of the only Sunni lands in this portion of the upper Rio Chachi watershed bordering the Rio Apacheta, providing important sources of cultivated land and water for comuneros in the area.

Source: Instituto Geográfico Nacional, Perú.

Figure 14: Disputed Territory Between Chuschi and Hacienda Yaruca in NW of District of Chuschi

165

However, the sale was annulled by an Ayacucho judge because of competing ownership claims by other mestizo elites.518 Chuschinos considered the land as their own, so the annulment of the sale must have been particularly galling to the community. In the wake of the aborted sale, Emilio Del Solar’s sons Cesar Flores and Javier Del Solar set out in the following decades to tighten their control of the disputed lands and attempted to steal even more land from Chuschi.

With land invasions and limited land reform happening in other parts of Peru in the late 1950s and early 1960s, the community must have felt that it was only a matter of time before the land would be returned to them. The 1968 military coup in Peru and promises by the leader of the controlling military junta, Gen. Juan Velasco, that agrarian reform was coming to Peru surely heightened expectations that the disputed lands would return to Chuschi. The greater the portents of reform, however, the more the hacendados increased their legal and violent harassment of Chuschinos on the borders of the disputed territories, to try to coerce control of more land for Yaruca, to cement control of the land they held, and to intimidate Chuschi into relenting to the hacienda’s control of the disputed land. Between 1968 and 1975, the hacendados filed several spurious legal complaints about supposed Chuschino attacks and damage of their property as a cover for their own campaign of intimidation, violence, and abuse of Chuschinos and destruction of Chuschi property in the areas bordering the disputed lands. When not leading the assaults themselves, the hacendados often compelled their own peons undertake the violence and intimidation against Chuschi, possibly hoping to provoke a violent counter-response from Chuschi that would portray the Del Solar family as the victims. 519 They paired the strategy of violent intimidation with a vigorous legal defence that attempted to question the cultural, moral, and legal legitimacy of Chuschi’s land claims.520

Chuschinos maintained restraint and patience in the face of continued violence and provocation from Del Solar for several years after the 1968 coup. However, as years passed without resolution to the dispute, the community turned to direct action to regain control of the disputed

518 La Serna, Los Huerfanos, 249. A legal petition in 1978 by the Del Solar family noted that the land under dispute comprised 650 hectares. Chuschi’s legal filings refer to this land as Yaruca-Ingaflorida. Isbell notes that the disputed lands are adjacent to Chuschi’s communal grazing lands in Inga Wasi. See Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 43-4; Isbell, To Defend, 241. 519 La Serna, Los Huerfanos, 261-3. 520 La Serna, 256-8.

166 territory. In the first years following the new government’s announcement of agrarian reform in Peru, the community put its faith in the bureaucratic reform process to regain their land, cooperating with various regional government officials to detail their land holdings.521 Perhaps buoyed by the quick expropriation of the nearby Hacienda Milpo in Oct. 1969, 522 for Chuschinos, the Del Solar family’s days in control of Yaruca must have seemed numbered. In mid-1972 the community launched a formal lawsuit to regain the disputed lands. The trial continued for years without a resolution, hearing a variety of evidence from both sides about who rightfully controlled the contested lands.523 Frustrated by the continued control of Yaruca’s lands years after Velasco had promised agrarian reform, in 1975 the Chuschinos decided to engage in direct action to pressure the government’s agrarian reform office to speed up the decision on its lawsuit.524 Without warning, over 50 Chuschino men herded hundreds of animals from the community’s recently recovered cofradia lands onto the disputed territories, trampling Del Solar’s potato crops in the process in an otherwise peaceful invasion of the territory.525 While this invasion did speed up the review process of the litigation by convincing the government to dispatch a land judge to the area a few weeks later to gather evidence,526 Chuschinos would have to wait until mid-1976 before the court decided in their favour and awarded them the disputed territory.527

Unfortunately for Chuschi, legal victory did not resolve the dispute. Over the next two years the hacienda owners used a variety of legal subterfuge and delaying tactics to try prevent Chuschinos from actually taking physical control of the disputed lands. Given the shifting national political context at the time, with the deposing of Peru’s President Velasco in a coup in August, 1975, the reversal of progressive policies by the new leader, and the ending of agrarian reform in 1976,528 the hacendados perhaps thought that by refusing to relinquish control of the disputed territories,

521 La Serna, 254 & 263-7. 522 The expropriation of the Hacienda Milpo is described in David Scott Palmer, "Revolution from Above," 247-49. 523 La Serna, Los Huerfanos, 254-5. 524 Isbell, To Defend, 240. 525 Isbell, To Defend, 240. Isbell notes that it happened on April 16th, 1975, and was a largely peaceful invasion, aside from the shooting of a Chuschino’s dog by a hacienda employee. 526 Isbell, To Defend, 241. 527 La Serna, Los Huerfanos, 274. 528 Watters, Poverty, 359-60.

167 they could outlast or reverse the court’s decision. First Del Solar tried to claim that the disputed lands actually belonged to other relatives, thus invalidating the court decision. Then, the family tried to argue that they needed the land for their personal use, and that the Agrarian Reform laws prevented ‘peasants’ from seizing lands used by Del Solar for personal use. They also tried to change the name of the disputed territory and claim that Chuschi was trying to illegally seize the land. There were also claims that Del Solar was bribing the land judge to prevent the handover.529 Hand in hand with the legal stalling tactics, Del Solar continued to use violent intimidation on Chuschinos in the area to maintain control of the land.530

By late 1978, Chuschi’s patience had run out that a legal route would finally return the disputed territory to the community. The community again decided to take direct action to enforce the court’s decision. Hundreds of Chuschinos invaded the disputed lands, capturing Javier Del Solar in the process. Del Solar was allegedly captured with equipment for processing cocaine on the Yaruca estate. He was forcibly marched to jail in Huamanga by community members, and later faced charges of manufacturing cocaine.531 Unfortunately, for Chuschi, the community was ordered by police to leave the disputed territory soon after, again putting into doubt the resolve of the state to enforce the court’s decision on the land dispute. Time, however, was running out on the Del Solar family, because Sendero was becoming increasingly active in the area in the late 1970s. By 1981, when the area was firmly under Sendero control, the Del Solar case was abandoned, the hacendados presumably having fled the region. The dispute appears to have continued into the late 1980s during the Sendero insurgency, even though the hacienda-owning Del Solar Flores family had by then been replaced as the plaintiff in the conflict by the community of Yaruca, in the neighbouring District of Vinchos.532 It is possible that Yaruca’s peon’s were leading the legal case against Chuschi at the time, either in a bid to win control of the hacienda’s lands for themselves, or at the behest of the hacendados. The latter seems possible given that the Del Solar family reappeared in litigation against Chuschi in the late 1990s. According to La Serna’s 2007 interview with former community President Juan

529 La Serna, Los Huerfanos, 274-279; Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 44. 530 La Serna, Los Huerfanos, 277-8. 531 La Serna, Los Huerfanos, 278-80. 532 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 42-7.

168

Carhuapoma, the Del Solar family continues to frustrate Chuschi’s efforts to regain the contested lands.533

5.3.3. Quispillaccta’s dispute over Hacienda Quicamachay lands on the NE borders

The Chuschi-Del Solar dispute is particularly interesting because the conflict endured long after the mestizo hacienda owners disappeared from the area with the entry of Sendero. The shift in the conflict during the 1980s to a conflict between Chuschi and the former hacienda peons and community members from nearby hamlets happened in other areas of the district as well. The most serious example of this concerned the conflict between Quispillaccta and the Hacienda Putaje over lands purchased by Quispillaccta in the mid-20th century from the Hacienda Quicamachay. After decades of battling the hacendados of Putaje, Quispillaccta found itself struggling in the 1970s with former hacienda peons and community members from nearby communities for control of their purchased lands. In this case, when entitlement conflicts with mestizo elites began to wane in the twentieth century, the conflicts shifted to entitlement conflicts between communities, with lasting impacts of enmity between communities in the D. Chuschi.

In August, 1942 the community of Quispillaccta bought several thousand hectares of land from the Hacienda Quicamachay in an area on the north-east border of the current district. This purchase would trigger decades of conflict and legal battles with the owners of the nearby Hacienda Putaje and with peasants farming and living on the land at the time of the purchase. Quispillaccta’s eventual victory in this conflict would leave a lasting legacy of ill-will among those involved and worsen relations between Quispillaccta and Chuschi. Quispillaccta’s 1942 purchase of over 5000 hectares of former hacienda lands from Humberto Vasallo Barbaran and Sara Bedoya Palomino de Vasallo, the owners of the Hacienda Quichamachay was, in the community’s eyes, another step to recover lands long seized by mestizo elites. The sale encompassed areas around the Fundo Pillcoccasa and near Ccochapampa.534 About 300-400H of

533 La Serna, Los Huerfanos, 281. Sánchez Villagómez’s study appears to have mistakenly concluded that Chuschi won their disputed lands back. 534 I have been unable to locate the exact location of the purchased lands, but have estimated the location on the map, Figure 15, based on place names and similar map names.

169 the purchased land was being occupied by the Hacienda Putaje, according to a clause in the sale document.535

Source: Instituto Geográfico Nacional, Perú.

Figure 15: Disputed Territory in Land Conflict Between Quispillaccta and Hacienda Putaje in NE of District

However, the owners of the Hacienda Putaje, Vidal Ruiz Tello and César y Rodolfo García Espinoza, challenged the sale less than a year later, saying that the land really belonged to the Hacienda Putaje. Quispillaccta’s legal complaint against the Putaje Hacendados was eventually decided in the community’s favour in 1947 when the Dirección General de Asuntos Indígenas (Bureau of Indian Affairs), ruled that the purchased land belonged to Quispillaccta.536 The

535 Oseas Núñez Espinoza in an online history of Cuchoquesera, , assessed 9-1-16. Núñez Espinoza says that the Quicamachay sale brought the lands of Cuchoquesera to Quispillaccta. However, it is not clear if he means some or all of the lands now part of the barrio of Cuchoquesera. As noted above, Quispillaccta had a cofradia of animals in Cuchoquesera since the early 20th century and clashed with the Hacienda Putaje in the 1920s, according to Sánchez Villagómez. So, the timeline remains unclear on when ownership of Cuchoquesera was established by Quispillaccta. It is possible that part of the purchased Quicamachay lands added to Quispillaccta’s earlier holdings in the area. 536 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 49; Ministry of Agriculture archives, Ayacucho, Folio 97- 230.

170 government’s decision failed to settle the dispute, however. In 1951, frustrated by continued invasions of their lands at the hands of “hacendados and others” Quispillaccta decided to raise money to make a formal survey of their landholdings and send this to the Ministry in Lima.537

The conflict over the Quicamachay lands continued, however, in August 1965 when Putaje’s owner Vidal Ruiz Tello alleged that when Quispillaccta bought Quicamachay from Vasallo- Bedoya, 2900H of the 5300H sale was sold in error. This land actually belonging to the Hacienda Putaje, according to Ruiz’s complaint. This allegation was rejected by the court, but appealed by the Ruiz family and continued throughout the late 1960s.538

By the early 1970s, the hacendados of Putaje were no longer a party to the dispute. As occurred with land disputes with Yaruca in the north-west of the district, Quispillaccta’s battle to retain control over the purchased Quicamachay lands now continued with those peasants and peons who had taken over the Hacienda Putaje’s land holdings.539 In the intervening years, the national and local political context had radically changed in Peru with disruptive consequences in the area. The Peruvian military’s Agrarian Reform efforts sought to finally eliminate the old land holding system in the country and replace it with modern agrarian cooperatives. However, in places like the Rio Cachi watershed, where the hacienda system had long been in decline, it appears that the Agrarian reform accelerated the scramble among peasant communities in the area to consolidate control over contested lands, lands they considered their historical patrimony, or over lands that they had long worked as hacienda peons. Multiple competing agendas spurred conflicts in the area as each group sought to secure their interests and goals – from other communities in the area, from former hacienda workers, from government reform officials, from local elites, and from land owners struggling to maintain control over their holdings.

537 Quispillaccta, Libros de Actas de las Assembleas, 1947-62, Municipal Archives: meeting 15 April, 1951. 538 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 50. In an interview, Cuchoquesera resident Salomón Galindo Achallma said that the trial ended in 1970 in Quispillaccta’s favour. Galindo said that the hacendados of Putaje unsuccessfully tried to influence Agrarian Reform officials and use their influence to distort the Agrarian Reform process to their advantage to take control of the disputed land. Interview, Salomón Galindo Achallma, 77, Cuchoquesera, 28 October, 2004. 539 According to Cuchoquesera resident Salomón Galindo, Putaje’s hacendados had told their peons that part of Quispillaccta’s 1942 land purchase belonged to the Hacienda Putaje. So, when those Putaje peons took control of the hacienda in the Agrarian Reform in the early 1970s, they took over the claim to some of the disputed lands as well. Interview, Salomón Galindo Achallma, 77, Cuchoquesera, 28 October, 2004.

171

In some cases, the government’s agrarian reform only exacerbated conflicts in the area, as happened with the Hacienda Milpo, located in the Rio Cachi basin. The military government expropriated the Hacienda Milpo outright in October 1969 and sought to give all the land to the hacienda’s peons. The government also planned to create a massive collective farm with the asset of twelve haciendas in the Rio Cachi basin.540 However, this move only furthered unrest, uncertainty, and conflict in the area. The hacienda’s owners engaged in protracted legal battles with the Peruvian state that left the hacienda’s peons in a legal limbo for years, with no clear ownership or control of hacienda lands. The government sold the hacienda’s saleable assets while the court case continued. The impoverished peons struggled to maintain their livelihoods in the face of non-existent assistance from the government, succeeding only in further degrading the hacienda’s lands with poor practices.541 New conflicts also arose with neighbouring communities like Cancha Cancha, which long held grazing lands nearby in the area in Huaillaccasa, and with Quispillaccta over areas like Ñahuipuquio and Accoccasa.542 As discussed above in the literature review, Agrarian Reform measures aggravated long standing conflicts between communities in many areas of the highlands.543

In the case of the disputed Quicamachay lands, Quispillaccta suddenly confronted a much more complex series of conflicts when agrarian reform displaced the hacendados of Putaje. The Quicamachay conflict now morphed into three different disputes. Two of the disputes – over lands in Ccochapampa and around Pillcoccasa – revolved around struggles with peasants who had lived on or near the purchased lands when they were still controlled by Huberto Vasallo of Quicamachay. The third dispute involved former peons from the Hacienda Putaje over lands purchased by Quispillaccta as part of the 1942 Quicamachay purchase, but also claimed as part of Putaje.

540 David Scott Palmer, "Revolution from Above," 247-49. 541 David Scott Palmer, "Revolution from Above," 247-49. 542 The land conflict between Cancha Cancha and the Milpo hacienda continues to this day in the form of litigation between the former hacienda peons of Milpo who have formed the campesino community of Milpo and the community of Cancha Cancha. Roger Maquera, Las Comunidades Campesinas en la región Ayacucho, (Lima: Asociación SER/ Grupo Allpa, 2009): 18. . 543 See Linda J Seligmann, Between Reform & Revolution: Political Struggles in the Peruvian Andes, 1969-1991 (Stanford, CA: Stanford University Press, 1995), 60-1.

172

The Pillcoccasa dispute began when Quispillaccta purchased the approximately 200H of land around the Fundo Pillcoccasa in 1942 as part of the Quicamachay sale. At the time, the area was home to 43 families who served the Fundo as colonos. These employees worked and lived on the land for the hacienda, and were originally community members from Chuschi’s annex, Cancha Cancha.544 Cancha Cancha had relatively little grazing land available in the Colonial era, so for many decades some Cancha Cancha residents worked and settled in the hacienda lands of Putaje and Quicamachay, cultivating crops and raising livestock for the haciendas, while others paid the haciendas for the right to graze their own livestock on hacienda lands.545 Castillo Gamboa argues that Cancha Cancha residents considered the Pillcoccasa lands as part of their community’s historical patrimony, having controlled the area before haciendas were established in the area.546 Rocky, unirrigated, and with poor access, Pillcoccasa was not immediately settled by Quispillaccta after its purchase in 1942.547 Instead, the land was purchased along with all the colonos living on the lands. These colonos were allowed to continue living there in the decades to come to watch over it and help secure it for Quispillaccta.548 To the colonos, however, Quispillaccta’s purchase allowed them to reclaim control over lands they long considered their own.549 Castillo Gamboa claims that the conflict over Pillcoccasa accelerated in the 1960s, as

544 Castillo Gamboa, “Conflictos,” 50; interview with Pablo Conde, Quispillaccta, 25 October, 2004. 545 Castillo Gamboa, “Conflictos,” 36. A Chuschino informant I spoke to also noted that Cancha Cancha comuneros used to live in Pillcoccosa, and that some of those former colonos now live in the matriz of Cancha Cancha. Interview with Marcelino Roca, Chuschi, 24 Oct., 2004. 546 Castillo Gamboa, “Conflictos,” 50. 547 Castillo Gamboa, “Conflictos,” 51 and 54. 548 Interview with Pablo Conde, Quispillaccta, 25 October, 2004. Conde notes that the colonos of Pillcoccasa paid Quispillaccta authorities in sheep every September for the right to continue living in Pillcoccasa. 549 According to interviews with colono descendants done by Castillo Gamboa, “Conflictos,” 50, the hacendado Vasallo Barbaran first proposed selling Pillcoccosa to the workers living in Pillcoccosa; however, when the colonos couldn’t collectively agree to buy the land, Quispillaccta offered to buy the land along with the colonos. Descendants of these colonos claim that they contributed funds to help Quispillaccta purchase Pillcoccasa. However, I cannot find any record of this in the Quispillaccta records. “The hacienda owner Vassallo Barbaran had proposed the sale of Pillcoccasa to the settlers (colonos),” Castillo Gamboa quotes the son of one colono. “In that time one of the settlers was my father, Jesús Quispe [and] my uncle José Quispe, but the people were not in agreement over the idea of selling one cow per person to reach the price of Pillcoccasa. The oqes Quispillacctinos found out about this, and they proposed to buy the land together with the settlers. The settlers answered very well and accepted the offer and in this manner they bought the land together with the oqes Quispillacctas, giving the money to the Vassallos.” Gamboa, “Conflictos,” 50. “Oqes” is an expletive used by Chuschinos to refer to Quispillacctinos as “darkies.” It harkens back to the different ethnic identity between Quispillacctinos and Chuschinos. The frequent swearing in reference to the Quispillacctinos in this quote displays the disdain and hatred of the colono descendant for Quispillaccta. See also the discussion about Quispillactino Occes, as La Serna spells it. La Serna, The Corner of the Living, 62.

173

Quispillaccta sought to consolidate greater control over their purchased lands, and as the area opened up to greater commercial exchange and improved transport with the completion of a road to the district in the mid-1960s.550 As noted above, however, continued disputes with the hacendados of Putaje in the mid-1960s also spurred Quispillaccta to consolidate control over their land holdings. It also appears, however, that the colonos and peasants from Cancha Cancha increased their efforts in the late 1960s and early 1970s to take control of Pillcoccasa, perhaps emboldened by the rhetoric and actions unleashed by agrarian reform in the area. Increasing numbers of peasants from Cancha Cancha began to settle in the area and refused to leave as Quispillaccta attempted to secure the land.551

Another dispute with former workers from the Hacienda Quicamachay centred around lands near the peasant community of Ccochapampa that Quispillaccta long considered their own.552 As in Pillcoccasa, the former hacienda workers (referred to as feudatarios in legal filings) were allowed to stay on in the land after it was purchased by Quispillaccta, on the condition that they would integrate themselves as members into the community of Quispillaccta. However, Quispillaccta claimed that the former feudatarios from Quicamachay did not follow through on their promises and instead looked to integrate themselves into the nearby community of

550 Castillo Gamboa, “Conflictos,” 51. Isbell, To Defend, 46, notes that the road from Huamanga to Chuschi was completed in 1966, entering the district from the south-east via Pampa Cangallo and along the Rio Pampas. A road into the district through the Rio Cachi watershed in the north east was completed some years later. 551 Castillo Gamboa, “Conflictos,” 52. In an interview, Quispillactino Pablo Conde remarked that as numbers of those living in Pillcoccasa increased, they became emboldened and refused to pay rent to Quispillaccta, seeking instead to obtain control of the land. Interview with Pablo Conde, Quispillaccta, 25 October, 2004. Oseas Núñez Espinoza says that the Cancha Cancha dispute was complicated by the actions of a former foreman of the Hacienda Quicamachay, Darío Ochoa, and his decedents, who worked land in the area called Incacapilla that was claimed by Quispillaccta. Darío Ochoa’s descendants married comuneros from Cancha Cancha, further enmeshing Ochoa’s dispute with Quispillaccta with conflicts between Quispillaccta and Cancha Cancha over Pillcoccasa. The Ochoa family’s involvement in the conflict continued well into 1981, when Sendero began to operate in the area. Oseas Núñez Espinoza, online history of Cuchoquesera, , assessed 9-1-16. There is no clear history of the conflicts, disputes, lawsuits, and land conflicts in the north-east corner of the district. 552 In an April 1975 community meeting in Quispillaccta, community members affirm that the disputed area was always Quispillaccta’s land, and was formerly referred to by the community as Rograhuaycco. A 1979 legal filing by Quispillaccta noted part of the dispute with Ccochapampa revolved around a 1x5km parcel of land beside the creek Ñahuipuquio, passing through Rupa-quesera until Almapampana-ccasa. Sánchez Villagómez, "(Re)Pensando Lo Olvidado, 53. The area appears to lie to the north of the Cerro Chontalla on topographical maps. Quispillaccta, Libros de Actas de las Assembleas, 1973-75, Municipal Archives: meeting April 1975.

174

Ccochapampa, seizing control of the disputed lands and destroying property belonging to Quispillacctinos in the area.553

A third dispute nearby with former feudatarios from the now defunct Hacienda Putaje centred on pasture land in an area near the territories under dispute with Ccochapampa.554 Reigniting the long-running disputes with the Ruiz family, this conflict led to a one-day physical conflict between residents of Quispillaccta and Putaje in 1974.555

In community meetings in 1974, Quispillacctinos expressed their frustration with ongoing land conflicts around the Quicamachay lands. The community decided to again undertake a survey of Quispillaccta’s territories and to press SINAMOS (the Sistema Nacional de Movilización Social or National System for Social Mobilization) – a newly formed bureaucratic organization formed by the Velasco military government to advance and control revolutionary change in rural areas – to help Quispillaccta find a definitive resolution to the disputes with its neighbours.556 Ultimately, entreaties to the government failed to find a resolution to the community’s disputes in the area. Communities continued to seek definitive resolutions through the court system or by affecting realities on the ground. After several years of legal battles and government intervention, the conflict with Putaje was settled in 1976, largely in Quispillaccta’s favour.557 However, disputes with Cancha Cancha over Pillcoccasa and Ccochapampa heated up and continued for years throughout the late 1970s and into the early 1980s. Both Cancha Cancha and Ccochapampa revived the arguments of the hacendados Ruiz that the Quicamachay lands

553 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 51-2. Sánchez Villagómez dates the dispute to 1976; however, there are references several years earlier to the dispute in the records of Quispillaccta’s community assembly. The conflict was clearly on-going for several years in the late 1960s and early 1970s before formal legal filings were made in 1976, as indicated in the above note. 554 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 50-1; Castillo Gamboa, “Conflictos,” 47 and 53; and Quispillaccta, Libros de Actas de las Assembleas, 1973-75, Municipal Archives. The area is referred to as Chontaya in Castillo Gamboa (Chontalla on topographical maps), and as Rupa quesera Huayco in Sànchez Villagómez. It may be referring to lower areas of the Rupaquesera brook, since Huayco is Quechua for valley and the lower elevation areas of the brook Rupaquesera flow toward Putaje’s lands in the Sunni zone. 555 Castillo Gamboa, “Conflictos,” 53. 556 Quispillaccta, Libros de Actas de las Assembleas, 1973-75, Municipal Archives: meetings 17 Feb. 1974, 27 April 1974, & 20 August 1974. 557 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 51; Salvatierra et.al., La Vida, 205.

175

Quispillaccta purchased from Vassallo were never Vassallo’s to sell, thus invalidating the legitimacy of the 1942 purchase by Quispillaccta.558

Several times in the late 1970s, rival community groups attacked crops and settlements in disputed areas, and then planted their own crops and built their own settlements to assure the legitimacy of their claims. This often provoked a legal and physical counter-attack in a tit-for-tat dynamic that went on for years. In 1976, large groups of Quispillacctinos entered disputed areas in Pillcoccasa on two occasions, plundering and destroying the huts and corrals of the Cancha Cancha residents, and then staking their own presence in the area, building huts, planting crops, and pasturing animals throughout the disputed areas.559 Two years later, in January 1978, the entire community of Cancha Cancha responded by gathering together to counter-invade the disputed area, resulting in a serious physical clash with Quispillacctinos in the area. Charges and counter charges led to trails over the land invasion and assaults that lasted for several months in 1979 before Cancha Cancha abruptly abandoned the legal cases in early-1979, leaving Quispillaccta victorious.560 Little more than a week after Pillcoccasa was occupied by Quispillaccta in Sept. 1976, former feudatarios from Ccochapampa entered into disputed territories in their conflict with Quispillaccta, occupying lands and destroying 14 huts belonging to Quispillaccta community members.561 With no formal legal title to prove their ownership, however, Ccochapampa lost the legal fight over the disputed lands in mid-1977 and was forced to pay a large fine to Quispillaccta in compensation.562 The dispute with Ccochapampa did not end, however. In 1978 the community convinced regional authorities to allow it to separate from Quispillaccta and the D. and become part of the District of Vinchos. In early 1979, Ccochapampa residents again invaded Quispillaccta territory in the disputed area. Legal appeals

558 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 52 and 56. 559 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 54; Castillo Gamboa, “Conflictos,” 55-57. 560 Quispillaccta, Libros de Actas de las Assembleas, 1976-79, Municipal Archives: meeting 17 May, 1979; Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 55. I have not been able find any details about why Cancha Cancha suddenly quit their legal efforts to regain control of Pillcoccasa or how the dispute was resolved. Salvatierra et.al., La Vida, 205, says that the dispute with Cancha Cancha was resolved in 1982 through extra-judicial conciliation. 561 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 51-2. 562 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 52; Quispillaccta, Libros de Actas de las Assembleas, 1976- 79, Municipal Archives: meeting 13 July, 1977. The community meeting notes that Ccochapampa has given Quispillaccta 200,000 Soles in compensation for legal expenses.

176 by Quispillaccta failed to settle the dispute, and trials over boundaries continued well into 1981 and 1982, the period when Sendero began operating in the area.563 Perhaps frustrated by the lack of resolution of the conflict, in early 1982 Quispillacctinos in the area responded by seizing control of cattle and land in the disputed area, leading Ccochapampa to file a legal complaint.564

Outside political events, however, finally appeared to be making an impact on the long-running dispute. By the beginning of 1982, the insurgency and counter-insurgency against the Shining Path was heating up in the Rio Pampas and Rio Cachi area. In October and November 1981, Sendero shuttered the Chuschi municipality, leading all the local officials to resign their posts.565 At the same time that Sendero celebrated taking control of the district, Peru’s paramilitary police forces began a counter-offensive in Ayacucho, declaring a state of emergency in the province of Cangallo in October, 1981, significantly increasing patrols in the area, and arresting suspected militants.566 All the communities in the area began to feel pressure from both the Shining Path and the Peruvian security forces to settle communal disputes once and for all, as will be discussed further below in Chapter 8. In March 1982, community records note that Quispillaccta and Ccochapampa met to define the final boundaries in the dispute. In the coming months, the long history of land invasions and legal appeals ends.567

High altitude grazing lands in the Rio Cachi zone of the D. Chuschi have been an essential part of household livelihoods in the area for centuries. However, the communities in the district have had to confront centuries of capture of these lands by mestizo elites and the Catholic Church. This chapter has outlined the long history of the capture of puna lands in the north of the district, and the conflicts of Chuschi and Quispillaccta to regain control of these lands, particularly in the twentieth century. Lingering disputes over seized hacienda lands have enmeshed both communities in decades of litigation and frequent clashes on the ground over disputed territories

563 Quispillaccta, Libros de Actas de las Assembleas, 1976-79, Municipal Archives: meeting 15 Feb., 1979; Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives: meeting 11 August 1979; 31 Jan., 1982. 564 Quispillaccta, Libros de Actas de las Assembleas, 1982-85, Municipal Archives: 4 March, 1982. 565 Isbell, To Defend, 84. 566 Gorriti, Shining, 143. 567 Quispillaccta, Libros de Actas de las Assembleas, 1982-85, Municipal Archives: 8 March 1982; Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 52; Salvatierra et.al., La Vida, 205, say that the deal with Ccochapampa was sealed on 23 April, 1982.

177 in the north-west of the district and north-east of the district. These clashes and contentious litigations continued right up to the outbreak of Sendero activity in the district. Chuschi’s conflict seemed to abate when Sendero became active in the community, but reignited with the former hacienda peons (perhaps fronted by the hacendados) when Sendero was driven out of the District. Quispillaccta was largely successful in conflicts and litigation in the late 1970s over disputed lands in the north-east of the district, outcomes that no doubt grieved the losing parties on the eve of Sendero’s activities in the district.

178

Chapter 6 Controlling the Centre: The Struggle to Control the High Sunni and Puna Zones of the Rio Cachi Basin in the District Introduction

By the middle of the twentieth century, long term structural changes in the D. Chuschi combined with pressures released by national legislative changes to set the stage for bloody land clashes between the communities of Quispillaccta and Chuschi in the 1960s and 1970s. The proximate cause of these clashes revolved around gaining control of high-altitude grazing and cultivated zones immediately north of the district Matriz and south of the Quispillaccta barrios of Puncupata and Unión Portrero. These lands were grazing and marginal cultivated zones that were seeing increasing permanent settlement from the district Matriz, and were important sources of water for downstream barrio lands around Puncupata, Catalinayuq, Cuchoquesera, and Unión Portrero to the north, and water for upper cultivated zones above the district Matriz. While the evidence appears to indicate that this area was long inhabited by a few comuneros from both Quispillaccta and Chuschi (and its associated barrios), growing settlement of these areas by mid-20th century set the stage for increasingly fierce legal and physical clashes for control in the latter half of the 20th century.

This chapter outlines the nature of these clashes and the origins of intra-district migration from both Quispillaccta and Chuschi during the twentieth century that aggravated conflicts for control of this zone. Various factors that increasingly drew comuneros to permanently settle in this zone are also discussed, such as the desire to increasingly focus on livestock husbandry, increasing educational opportunities in the area, and better linkages to Huamanga to the north. The chapter also examines clashes in the 1960s and 1970s between Chuschi and Quispillaccta over this zone, and the outcome of those clashes. The chapter concludes by examining the historic peace settlement between Chuschi and Quispillaccta in late 1981, which conclusively settled this long- running conflict. The implications of this settlement and other land conflicts in the Rio Cachi puna zone are explored in the chapter’s conclusion. In the end, the settlement was a major success for Quispillaccta, solidifying their control over much of the disputed territory in the centre of the district, both grazing lands and high-altitude sources of water for these areas.

179

Disputes in the 1940s and 1950s to control the centre

Conflicts over the centre of the district raged for decades over the high punas and mountain peaks that divide the Rio Pampas watershed in the south from the Rio Cachi watershed in the north. Formal legal clashes over these areas first appear in the documentary record after the community of Chuschi sought to obtain official community recognition in late 1940. This effort inflamed disputes between Chuschi and Quispillaccta over each community’s claimed landholdings. Quispillaccta denounced Chuschi for usurping lands “that had belonged to them ‘since time immemorial’” but which had been taken by the haciendas Putaje, Quicamachay, and

Source: Instituto Geográfico Nacional, Perú.

Figure 16: Disputed Zone between Chuschi and Quispillaccta, 1940 Shaded area of the 1940 disputed zone is approximated based on a copy of a topographical map of the disputed territories found in César et al. 1967. The original map was produced by the community of Quispillaccta, and listed as being in the possession of the legal representative of Quispillaccta. The date of the original is unclear. The map may have been the outcome of a community decision in April 1951 to survey their lands, as recorded in the records of Quispillaccta’s community decisions. The Quispillaccta barrios (in blue lettering) of Unión Portrero and Yuraqcruz are shown in their current locations. These settlements appear to have been expanded after the conflicts with Chuschi between the 1940s and 1970s, as discussed below in chapter 6. Quispillaccta, Libros de Actas de las Assembleas, 1947-62, Municipal Archives: meeting 15 April, 1951.

180

Allpachaca. 568 For this reason, Quispillaccta asked that it too be recognized and registered as Chuschi had.569 This event, in fact, is the first modern legal complaint between Chuschi and Quispillaccta, and the precursor to four decades of direct conflict between the communities. The 1940 dispute centred around a 40 km2 portion of the district that lay between Quispillaccta’s Rio Cachi land purchases in the north and the district Matriz in the south. An official sent by the Ministry of Indigenous Affairs who was sent to the area in October, 1941 to try to settle the dispute noted that the area in dispute lay generally to the east, northeast, and southeast of Kimsa Cruz.570

As noted above, these were primarily lower puna grazing lands, with limited and risky cultivated potential. These lands were significant to both communities for two reasons. First, these lands provided relatively close access to highland grazing zones from the Matriz for both Quispillaccta and Chuschi and from the increasingly settled Quispillaccta barrios to the north. Chuschinos main grazing area lay several hours walk to the north-west, while Cancha Cancha had a small communal grazing zone around Huaylla Ccasa and Choccoro, in an area on the north-central border of the district.571 The livelihood opportunities to access this pasture zone was especially important to those living in Chuschi’s associated barrios of Uchuyri, Cancha Cancha, and Chacolla, located in the extreme south-east of the district, near the Rio Pampas. Barrios like Cancha Cancha had less grazing land than Quispillaccta and Chuschi.572 Having access to grazing zones above the Matriz could also save hours in travel time to and from pastures and make the difference between having livestock or not for a family.573 So, for Chuschi and its associated hamlets, it was highly advantageous to have pastures located close to the main

568 La Serna, The Corner of the Living, 65; Archivo del Ministerio de Agricultura, Ayacucho, Tomo 1, Folio 42. These haciendas all controlled Rio Cachi lands that currently sit in the north-east of the D. Chuschi, and lands in neighbouring districts to the north-east. 569 La Serna, The Corner of the Living, 65. 570 Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 59. 571 Castillo Gamboa, “Conflictos,” 34-5. 572 Castillo Gamboa, “Conflictos,” 35. 573 Castillo Gamboa quotes one elderly community member from Cancha Cancha who noted the historically limited grazing opportunities for Cancha Cancha residents compared to Quispillacctinos: “Before there was good agricultural production, but livestock were few. We did not have much; there were 80 cows among 4 families. Now there is nothing; not even one. The families that have the most have up to 5 cows. There is not much in comparison with what Quispillacctinos have; they produce a great quantity of livestock and have a communal livestock business. They are doing well.” My translation. Castillo Gamboa, “Conflictos,” 36.

181 cultivated zone in the district. For Quispillacctinos, access to nearby grazing zones for those living in the Matriz or those settling in the barrios of Unión Portrero, Puncupata, Pampamarca, and Catalinayuq, was similarly beneficial.

The contested lands were also significant because they provided control and access to important sources of water for the Quispillaccta barrios to the north, and for agriculture and pasture zones north of the Matriz for both communities. Water resources are technically owned by the Peruvian state, according to the 1969 General Water Law.574 But in remote districts like Chuschi where the Peruvian state has historically had little presence, functional and administration control of water resources has been contested by communities. Some Quispillacctinos contend that the Spanish colonial reducción policy stripped Quispillaccta of its water resources, leaving Quispillaccta water poor, while neighbouring Chuschi was rich in water resources.575 Quispillaccta’s purchase and settlement of areas in the north-east of the district in the 19th and 20th century brought the issue of access and control of water resources to the forefront. These barrios lay several hundred meters lower than the surrounding peaks, which form the headwaters of the Rio Cachi basin in this area. Streams and high-altitude wetlands provide important water sources for pastures and lower cultivated zones for these barrios. Some areas, like Catalinayuq, had little water in the lower Sunni areas and relied on water flowing from higher altitude zones to the south to supplement resources in dry months.576 Streams and a canal brought water from Ingahuasi for consumption and irrigation in Catalinayuq to the north, for example.577 The barrios of Catalinayuq and Cuchoquesera were the site of two of Quispillaccta’s communal herds – the former under the control of the church, while the latter was the only livestock herd under the sole control of the community of Quispillaccta until the 1970s. Further to the south, streams around Cconchulla, located at the head of the Qunchalla/Chuschi

574 Paul Trawick, The Struggle for Water in Peru: Comedy and Tragedy in the Andean Commons (Stanford, CA: Stanford University Press, 2003), 37. 575 ABA, Yakumama – Madre Agua Lagunas de Lluvia y Comunidades Criadoras del Agua, 49. 576 ABA, Yakumama – Madre Agua Lagunas de Lluvia y Comunidades Criadoras del Agua, 101-2. 577 César Ramón Córdova, Estudio de la Vivienda en los Centros Poplados de Inkaraqay, Chuschi, y Catalinayoq, Ministerio de Trabajo y Comunidades (Instituto Indigenista Peruano, Ayacucho, 1968), 37. Even in years when the rains were good, the streams and canal from Ingahuasi would often dry up for several months from August to October, according to local informants quoted in the ABA study. This has changed now that ABA has helped develop rain-water harvesting projects in the area. ABA, Yakumama – Madre Agua Lagunas de Lluvia y Comunidades Criadoras del Agua, 101-2.

182 valley, similarly provide key springs for cultivated agriculture to the fields below, located above and around the Matriz.578 Informants indicate that Quispillaccta struggled to obtain sufficient water in the past for the fields above the Matriz, often clashing with Chuschi for control of shared waters in the area. Juan Galindo recalled that in Totora des Huascarumi, Chuschi built a canal to divert water from springs in the area for their use, forcing Quispillaccta to divert water from Cconchulla for their fields.579 Pascuali Huamani noted that there was a dispute over the Milluyacu spring, located in the area below Mutuma, when Chuschi diverted it for their fields. Quispillacctinos had to divert the spring for their own use secretly at night.580 Disputes over shared water sources in the area, particularly as demand-induced pressures increased with increasing settlement of the Sunni and Puna zones, were a key dimension of the conflict between communities and would arise again in the 1960s when the conflict heated up.

Government intervention to solve the conflict between Quispillaccta and Chuschi produced an agreement in November 1941, but failed to put an end to the competition to control the lands around Kimsa Cruz. After legal representatives from both communities were summoned to Lima in May, 1941, months passed before an official government representative arrived in the communities in October 1941 to hammer out an agreement. The government’s representative failed to determine conclusive ownership of the disputed lands, so he proposed to solve the conflict by simply dividing the disputed zone in half.581 Importantly, however, the agreement failed to deal with the fate of comuneros from both communities living within the disputed area. Comuneros from each community were allowed to remain in place, even if they now found themselves in lands controlled by the other community.582 While this solution avoided the difficult problem of transferring populations from one side to the other of the line of control, it simply set the stage for future conflicts as settlement of the area from both communities increased over the next twenty years.

578 Machaca, “Vigencia y Continuidad,” 22-3. 579 Interview with Juan Galindo, Oct. 11, 2004. 580 Pascualia Huamani, interview, 12 Oct., 2004. The areas of disputed water resources are noted on the topographical map, underlined in red. Huamani is the widow of Sebastian Mendíete Tucno, a Quispillactino killed along with his brother Martin in the 6 May, 1960 confrontation with Chuschi discussed below in section 6.4. 581 La Serna, The Corner of the Living, 65-7. 582 Three provisions of the agreement are outlined by Sánchez Villagómez, "(Re)Pensando Lo Olvidado,” 58-9.

183

While the Peruvian government lauded its own efforts to reach a “definitive end”583 to the conflict between Chuschi and Quispillaccta, the boundary remained “imaginary” because the government did not implement the agreement on the ground by marking boundaries in both communities.584 In fact, disputes on the ground about land ownership during the next twenty years further inflamed relations between communities in the area. In April 1951, Quispillaccta’s community assembly approved a plan to collect money from community members in order to have a blueprint of their landholdings drawn up. Frustrated that their lands continued to suffer invasions from “hacendados and others,” the assembly vowed to send the plan to the Ministry of Indigenous Affairs in Lima and to the Department prefecture.585 Two years later, Quispillaccta successfully petitioned Department authorities for full administrative and economic autonomy from Chuschi, complaining that Chuschi’s district authorities were failing to cooperate with Quispillaccta on district public works projects.586

Factors drawing district members to settle the high Sunni and Puna zones

Until the late 1950s, there were few legal complaints between Chuschi and Quispillaccta over lands around Kimsa Cruz. For a time, the conflict appeared to have subsided, even while both communities faced challenges on other fronts in the district. Developments both nationally and locally, however, were setting the stage for renewed conflicts in the 1960s. As peasants around Peru began to agitate for agrarian reform and popular peasant movements began to seize land in the central Sierra in the late 1950s, increasing numbers of comuneros from both Quispillaccta and Chuschi were settling permanently in the high Sunni and Puna zone north of the district Matriz, moving from nuclear settlements near the Rio Pampas to barrios spread out in the Rio Cachi watershed. Various push and pull factors encouraged this trend during the middle decades of the twentieth century. This settlement transformation in the district would have important consequences for intercommunity conflict in the second half of the 20th century.

583 Quoted in La Serna, The Corner of the Living, 67. 584 Ministry of Agriculture archives, Ayacucho, Chuschi Tomo 1, Folio 252-279. 585 Quispillaccta, Libros de Actas de las Assembleas, 1947-62, Municipal Archives: meeting 15 April, 1951. Unfortunately, this volume was damaged in storage, destroying important records of this decade. Community records in Chuschi were similarly destroyed when Sendero burned the district municipal building in the early 1980s. 586 La Serna, The Corner of the Living, .67.

184

While push factors will be examined in the following chapter exploring supply and demand pressures in the district, in this section I will focus on the pull factors that drew settlers to these previously lightly settled zones.

Arriving in the D. Chuschi today from the north, through the Rio Cachi watershed, a visitor encounters an area heavily settled compared to what it looked like one hundred years ago. The settlement change is especially pronounced in the Sunni zone of the Rio Cachi watershed controlled by Quispillaccta, where today a majority of Quispillacctinos have settled permanently in areas that were previously almost devoid of permanent residents. According to 1993 Census data, almost 60% of Quispillaccta’s residents live in the high-altitude barrios of Tuco, Unión Portrero, Yuraqcruz, Pampamarca, Puncupata, Catalanayoc, and Cuchoquesera.587 Many Quispillacctino families who live permanently in these barrios still have homes in the Matriz, but these are used only occasionally when tending fields in the lower maize zone or by the elderly and younger family members who need to be close to schools and social services offered in the Matriz. Settlement in Chuschi’s high-altitude Sunni and Puna zones has somewhat increased over the past century, but not to the same degree as in Quispillaccta. Over 80% of residents of Chuschi and its associated barrios of Cancha Cancha, Chacolla, and Uchuyri still live in nuclear settlements near the Rio Pampas that date from the colonial reducción resettlement.588

The district’s high-altitude Sunni and Puna zone was largely devoid of permanent settlements until the middle of the 20th century, according to informants. Belasario Galindo Conde, 96 when interviewed in 2004, recalls coming to Pampamarca as a teenager in the 1920s. At that time, there were no permanent residents or settlers in Pampamarca, he noted. People simply came to the area temporarily to care for animals.589 Nearby, few people were living Cuchoquesera when

587 INEI, Censos Nacionales 1993: IX de Población y IV de Vivienda, Centros Poblados 1993, Lima. Peruvians were increasingly returning to their former highland residences by the time of the 1993 census, with the end of violence in the region from the civil war. This continued throughout the 1990s. I visited a community in Huanta in 2000 that was only just returning to build houses after most of the community fled the region during the civil war. However, population levels in the District of Chuschi were fairly stable by 1993. The total district population was only a few hundred people higher in the 2007 INEI census – 8080 in 1993, compared to 8281 in 2007. 588 INEI, Censos Nacionales 1993: IX de Población y IV de Vivienda, Centros Poblados 1993, Lima. Based on 1993 INEI census data, about 13% of residents of Chuschi and its associated barrios live in settlements in the high- altitude puna zone of the Rio Cachi basin. 589 Interview, Belasario Galindo Conde, Pampamarca, 29 October, 2004. Chuschino Victor Galindo Tucno, 99 when I interviewed him in 2004, said few people lived in the upper zones when he was young. With a mother from

185

Putaje’s Hacendado Manuel Ruiz ordered the illegal seizure of Quispillaccta’s independent communal herd in 1923. Only a handful of families lived in the barrio when Pablo Conde was born in Cuchoquesera in the early 1930s.590 Salómon Galindo said that his father came to Cuchoquesera around 1918. The notable livestock farmer and one of the founders of Evangelical Protestantism in Quispillaccta was born in Cuchoquesera in 1927. Even into the 1940s, few lived permanently in the area, he said. Instead, families had corrals in the wide-open pastures and rotated through the area with their animals at different times of the year. Galindo and his family split his time between the Llacta (town centre) and Cuchoquesera until 1970, when he settled permanently in Cuchoquesera.591 The shepherd in charge of Catalinayuq’s cofradia was the first permanent settler in that barrio, according to an informant.592

By the early 1940s and into the 1950s, settlement in the high-altitude Sunni zone of the district increased substantially. Both Quispillacctinos and Chuschinos began to increasingly build permanent housing in areas that were previously only temporarily settled to care for livestock. In the early 1950s there were about thirty people living permanently in Catalinayuq when Emilio Rejas moved there to live with his in-laws.593 Belasario Galindo built his house in Pampamarca in the early 1950s, when only a handful of residents were living there at the time (six or seven). What is now the main barrio centre was then empty pampa without trees, he said. However, settlement increased in the area after this time.594 Settlement also began to increase in the barrio of Unión Portrero in the 1950s, according to José Espinoza Flores.595 By 1961, according to 1961 census figures, 720 Quispillacctinos were living in the Rio Cachi annexes, which was a

Chuschi and a father from Quispillaccta, Galindo Conde would often travel to the area when visiting his relatives in Quispillaccta. Interview, Victor Galindo Tucno, Chuschi, 24 October, 2004. 590 Interview, Pablo Conde, Cuchoquesera, 25 October, 2004. 591 Interview, Salómon Galindo, 77, 28 October 2004. In an interview, Galindo told me that he was 80. However, in a video interview, Galindo says that he was born in July, 1927. https://www.youtube.com/watch?v=iQWVvPP4- pM, accessed 6 September, 2018. 592 Interview Emilio Rejas Casavilca, Catalinayuq, 29 October, 2004. 593 Interview Emilio Rejas Casavilca, Catalinayuq, 29 October, 2004. Pascualia Huamani noted that about twenty families lived in Catalinayuq in the early 1950s. Interview, Pascualia Huamani, Quispillaccta, 12 October, 2004. 594 Interview, Belasario Galindo Conde, Pampamarca, 29 October, 2004. 595 Interview, José Espinoza Flores, Quispillaccta, 31 October, 2004.

186 little more than a quarter of Quispillaccta’s population.596 In other high-altitude Sunni and low Puna zones in the district, outside the Rio Cachi basin, settlement was also increasing. Settlement was increasing in the 1940s and early 1950s in Quispillaccta’s south-west barrios of Tuco, Huertahuasi, and Llacctahurán .597 About 20 people lived permanently in Tuco in the late 1940s, mostly to care for livestock, according to Roberto Huaman Machaca.598 Others only came to the area for short stays to care for livestock. But permanent settlement had increased to several dozen residents by the 1957, when Tuco resident Policarpo Casavilca Quispe left the area to work in Ica and other cities.599 By 1961, according to 1961 census figures, 628 Quispillacctinos were living in the high-altitude annexes of Tuco, Huertahuasi, Pirhuamarca, and Llacctahurán.600

596 Data in Bolivar de Colchado, Estudio Distrital de Chuschi, 12. The Rio Cachi barrios add up to about 28% of Quispillaccta’s total population. However, the data does not include population figures for the small, isolated Rio Pampas barrio of Soccobamba. So, the proportion in the Rio Cachi was probably slightly lower. 597 Only about fifteen people lived in Uchupata, above the barrio of Llacctahurán, in the 1930s and 1940s, when Fermín Qallocunto Tomaylla was a boy. In the late 1940s and 1950s, he noted increasing numbers of permanent houses were being built in Llacctahurán. Interview, Fermín Qallocunto Tomaylla, Quispillaccta, 31 October, 2004; Huertahuasi was largely empty in the early 20th century, according to Pedro Conde Nuñez, while Melchor Quispe Nuñez noted that about thirty people lived there in the 1940s. Settlement increased slowly in Huertahuasi in the latter half of the century. Interview Pedro Conde Nuñez, Quispillaccta, 11 October, 2004; interview Melchor Quispe Nuñez, Quispillaccta, 31 October, 2004. 598 Interview, Roberto Huaman Machaca, Tuco, 1 November, 2004. 599 Interview, Policarpo Casavilca Quispe, Tuco, 5 November, 2004. 600 Cited in Bolivar de Colchado, Estudio Distrital de Chuschi, 12. Residents from the small barrio of Socobamba, in the extreme south-west of the district fronting the Rio Pampas, were not counted by the 1961 census.

187

Table 2: Quispillaccta’s 1960 Population: Main Town and Annexes

Quispillaccta’s Matriz Population 1215 (1961 Census)

Quispillaccta Annex Population (1961 Census)

Cuchoquesera 133 Pampamarca 42

Yuraqcruz 83

Catalinayuq 187 Puncupata 130

Unión Portrero 145

Tuco 180 Huertahuasi 109

Pirhuamarca 121

Llacctahurán 218 Total 2563

Source: 1961 Census data cited in Bolivar de Colchado, Estudio Distrital de Chuschi, 12.

Changes in Chuschino settlement of the Rio Cachi high Sunni zone are difficult to reconstruct because today these areas are almost entirely controlled by Quispillaccta and the communities sometimes gave different names to the same area. However, informants from Quispillaccta noted that some Chuschinos had long lived in high zones above Unión Portrero and Puncupata, in Dos Corras and east to Ingahuasi, where Chuschinos had built a chapel and kept cofradia livestock. They also had some settlements and pastured livestock to the south around Pucacoral (near current-day Yuraqcruz) and Cconchulla.601 These areas linked to Chuschi’s livestock areas above the Matriz, around Condorbamba and Acco.

601 Interview, Quispillaccta, 16 October, 2004; Interview, Chuschi, 2 November, 2004. A Chuschino told me that Chuschi notable Ernesto Jaime used to pasture his animals in Cconchalla (Cconchulla in map below, Figure 17) like it was his own hacienda.

188

On topographical maps we can see that Chuschino settlement of parts of the high Sunni zone was linking its high-altitude grazing lands to the north-west, with its puna holdings above the Matriz. Chuschi’s territory was long separated with Quispillaccta in the middle, and Chuschi sought to control these lands to link these areas together, according to a Quispillacctino informant.602 Regardless of whether there was a deliberate plan to settle the high Sunni zone above the Matriz to form continuous Chuschino holdings, settlement patterns in the mid-20th century and the location of conflicts in the early 1960s, illustrate that the key conflict over the upper Rio Cachi basin was precisely those lands between the two dynamic centres of each community. See disputed zones in map below, Figure 17.

Source: Instituto Geográfico Nacional, Perú.

Figure 17: Disputed Zone between Chuschi and Quispillaccta, late 1950s to 1980s

602 Interview, Quispillaccta, 11 October, 2004.

189

Informants in both communities noted that increasing settlement from the Rio Pampas zone to the high Sunni and Puna zones in the 1950s and 1960s was a necessary strategy in order to ensure community control of the land. Again, as in other disputed zones, the idea that possession is nine-tenths of the law held sway, given the uncertain legal status of the disputed territories. Numerous informants noted that people moved to the high zones with the encouragement of communal authorities in both communities.603 With legal battles being waged by Chuschi in the north-west and by Quispillaccta in the north-east of the district, increasing settlement in Quispillaccta’s newly purchased lands was thought to help cement claims of ownership.604 Establishing or moving communal livestock herds to newly purchased or disputed areas was another strategy employed by the communities to ensure control of the land. Quispillaccta had done this in the late 1800s and early 20th century in Catalinayuq and Cuchoquesera. According to Quispillacctino informants, the same approach was used in Tuco, where the authorities moved the communal herd further up the valley in the 20th century to ensure control of the land, and with the establishment of a communal herd in Jelluy (Jelluy Huayjo), in an area that later became the barrio of Yuraqcruz.605 Quispillaccta’s strategy to encourage settlement in order to cement community control of contested lands continued off and on for decades, up until the early 1980s. In 1979, Quispillaccta’s communal authorities, encouraged peasants from the land-poor barrio of Pirhuamarca to move to the Yuraqcruz area in order to help consolidate control over lands claimed by Quispillaccta but still being occupied by Chuschinos.606 Chuschi pursued a similar strategy with the establishment of a communal herd in Llachocc, where they also built a small

603 Interview, Llacctahurán, 9 October, 2004; interview Quispillaccta, 9 October, 2004; Interview Ernesto Jaime, Chuschi, 19 October, 2004. 604 Interview, Llacctahurán, 9 October, 2004. 605 Interview, Jacindo Nuñez, Quispillaccta, 10 October, 2004; Interview, Emilio Conde Achalma, Quispillaccta, 11 Oct. 2004. 606 Oseas Núñez Espinoza in an online history of Cuchoquesera, , accessed 9-1-16. Peasants from Pirhuamarca had also established a neighbourhood in Cuchoquesera. The Sankaypata massacre survivor whose story is told in the preface was a shepherd of the community livestock herd in Cuchoquesera at the time of his capture. However, he self-identified as a resident of the barrio of Socobamba in my interview (which is very close to Pirhuamarca), and was identified as a member of the Pirhuamarca barrio in Cuchoquesera. Salvatierra et al., La Vida, 259. So, there was sufficient migration from Quispillacta’s stressed Rio Pampas barrios to the high puna barrios of the district in the late 20th century to establish identifiable migrant identities in the new settlements, even though they all considered themselves Quispillacctinos.

190 stone chapel, another strategy both communities used to try to cement ownership of a disputed area.607

Encouragement and entreaties to protect community land would not have been enough, in and of themselves, to convince comuneros in Quispillaccta and Chuschi to settle high Sunni and Puna zones of the Rio Cachi basin in the mid-20th century. Making risky livelihood decisions to permanently move to these areas also required that comuneros see tangible livelihood benefits from the move. For those who chose to make the move, the benefits were the prospect of obtaining abundant pasture land for ever-more valuable livestock herds, the realization through experimentation that cropping of tubers and grains was risky but provided the benefits of excellent harvests, and the promise that services like schools and roads would soon reach the area – particularly in the early 1960s.

The prospect of obtaining abundant pasture lands for increasingly valuable livestock herds was a strong draw to bring Quispillacctinos and Chuschinos to the high Sunni and Puna zones of the Rio Cachi basin in the district in the mid-20th century. Jose Nuñez explained in an interview that he moved to Catalinayuq in the 1940s from the Matriz near the Rio Pampas because there was land there for cropping and grazing opportunities. Quispillaccta’s authorities were allowing the land to be distributed and divided.608 By the early 1960s, however, there was little land left in Catalinayuq for new settlers, according to Jaoaquin Conde Pacotaype.609 Census figures from 1961 list 187 residents in Catalinayuq in 1961 – or about 30-35 families.610 Cropping was risky in the district’s Rio Cachi zones. However, increasing experimentation by new settlers showed promise and also helped to draw resident to the area. “There was an element of experimentation to cropping the high zones,” Emilio Conde Achalma explained in an interview. “People experimented with fields, saw that production was good, and increasingly moved to the high

607 Interview, Jose Nuñez, Quispillaccta, 16 October, 2004; La Serna, The Corner of the Living, 69. La Serna notes that Chuschinos also had a small chapel in the Suyoccacca area. Quispillaccta also built chapels in disputed areas, in Jelluy Huayjo, and attempted to build one in Llachoc early in the 1960 hostilities. The historical record is unclear about when these chapels were built in the central disputed zone, a point La Serna makes about Chuschi’s Llachoc chapel. 608 Interview, Jose Nuñez, Quispillaccta, 16 October, 2004 609 Interview, Joaquin Conde Pacotaype, Quispillaccta, 31 October, 2004. The informant was a community authority in Quispillaccta in the early 1960s. 610 Bolivar de Colchado, Estudio Distrital de Chuschi, 12.

191 zones.”611 The prospect of available grazing and crop lands in the district’s Rio Cachi zone was thus providing a viable livelihood option for residents based in the Rio Pampas community centres and helping to draw settlers north.

Further enhancing the pull to the Rio Cachi zone in the mid-20th century was the prospect that commodity prices were increasing in Peru in the second half of the 20th century, potentially increasing the livelihood benefit from making the risky decision to put more effort into livestock and crop activities in the higher zones of the district. From the early 1950s, the real value and sale price of cattle increased in Peru for about a decade, before stagnating in the 1960s and 1970s.612 Cattle herds increased sharply in the early 1950s in Ayacucho, before stagnating in the 1960s. Cattle numbers in Peru, however, continued to generally increase in Peru up to 1980.613 (See Figure 18 below.) Both Chuschi and Quispillaccta raised cattle for occasional meat sales or barter trade, according to a survey in 1966.614

611 Interview, Emilio Conde Achalma, Quispillaccta, 11 Oct. 2004. 612 Republica Peruana, Dirección de Investigación y Desarrollo, Estudio de la Política de Abastecimiento y Precios de la Carne de Vacuno (Lima: Ministerio de Economía y Finanzas, Marzo 1972): Graph 6, 71. 613 Ministerio de Agricultura, Compendio Estadístico Agrario, 1950-1991 (Lima: Ministerio de Agricultura, Oficina de Estadística Agraria, 1992): 522. With the exception of sheep, production of all major animal commodities increased in the 1950s – fowl, alpaca, goats, llama, porcine, and cattle. See tables on page 526. 614 Saramiento Medina et al., Los Distritos de Vischongo, Chuschi, Y Ocros, 34. The smaller annexes of Chuschi - Cancha Cancha, Uchuyri, and Chaqolla - did not sell their cattle for commercial sales, but consumed their meat, according to the IEP study. No doubt due to their limited access to grazing land, these annexes probably only occasionally sold livestock for cash.

192

Figure 18

Data Source: Ministerio de Agricultura, “Compendio Estadístico Agrario, 1950-1991,” (Lima: Ministerio de Agricultura, Oficina de Estadística Agraria, 1992): table 5.146, 522.

Crop prices also climbed between the 1940s and late 1970s throughout Peru. However, the increasing cost of living and regional differentiation in commercial commodity production during this period meant that the real value of crops produced provided little monetary benefit for most peasants.615 One important exception to the slow or stagnant growth in commodity values in the 1960s and 1970s was dairy products. The value of dairy products continued to increase substantially in the 1960s and 1970s.616 This led many peasants in Peru to shift

615 Watters, Poverty, 153; Klaren, Peru, 310-11. In one study of Cuzco noted by Watters, after a short increase in the early 1950s, peasants faced declining terms of trade up to 1980. See Watters, Poverty, figure 9.1, 156. 616 Watters, Poverty, 153. Watters cautions that the data for Peruvian livestock production and prices are problematic for this period.

193 production away from subsistence or foodstuff production to dairy cattle.617 District residents produced cheese for sale; however, milk was not sold commercially outside the district but consumed locally because of the lack of refrigerated storage and transport.618

For district residents, therefore, rising commodity prices in the early 1950s and the prospect of further increases were a definite draw to secure grazing lands in the Rio Cachi zones of the district. While crops and cattle meat production became risky by the 1960s because of stagnating prices, increasing imports, and price controls, the risk was partially offset by the increasing demand and increasing value of dairy products. Commodity price and demand signals thus provided strong lures to settle grazing areas of the district in the early decades after World War 2.

Further enhancing the draw to the district’s Rio Cachi zone was the prospect of infrastructure development in the area. Educational opportunities for district residents were expanding from the 1940s to the 1960s, reflecting a national trend of strong demand for expanded educational opportunities.619 In fact, the largest social movement in Ayacucho in the 1960s and 1970s was a movement composed of secondary school students seeking free education.620 While the politics of education and literacy in Ayacucho was complicated in the 1940s and 1950s, many peasants saw education as integral to notions of progress and development, and essential for casting off decades of oppression and manipulation by educated elites.621

School construction was as an important priority for the comuneros in the D. Chuschi; however, schools were mostly located in the Rio Pampas region until the late 1950s and 60s when efforts

617 Klaren, Peru, 310. 618 Saramiento Medina et al., Los Distritos de Vischongo, Chuschi, Y Ocros, 34. Cheese and milk production were commonly observed during field work in 2004; though, production and sale of fresh milk was limited by the lack of refrigeration in the district. 619 Klaren, Peru, 333; Degregori, How Difficult it is to be God, 161. Klaren notes that postwar economic growth, middle-class expansion, and increased pressure from peasant, migrant, and union groups drove the push for educational expansion during this period. However, Degregori notes that from 1960, education spending began to decline in real terms in Peru, leading to a growing gap between popular aspirations for education and the Peruvian state’s fulfillment of those education desires. 620 Degregori, How Difficult it is to be God, 115. 621 Heilman, Before the Shining Path, 96-7; Degragori, How Difficult it is to be God, 95-97; Isbell, To Defend, 69- 70.

194 began to locate schools in distant annexes. In early 1947, Quispillacctinos approved a motion to build a new school next to the small, aging structure in the Matriz that served as the community’s only school.622 But, with little support from the state and the expectation from authorities that community members would provide all the labour and some material for construction, progress was slow and demanding. More than ten years later, as tensions began to increase in the Rio Cachi zone between Quispillaccta and Chuschi, Quispillaccta’s new school was still without a roof. Community tensions may have spurred Quispillaccta to new sacrifices to complete construction. Between Dec. 1959 and August 1960 – a period that saw massive violent clashes between Quispillaccta and Chuschi - Quispillacctinos voted to suspend community celebrations three times and use the saved money to complete the school’s construction.623 While the vast majority of the more than 700 students in the district in the mid-1960s still attended schools located in nuclear settlements near the Rio Pampas, Quispillacctinos could reasonably expect that further settlement of the Rio Cachi zone would lead to schools being built in the area. Quispillaccta had already built a small school in Catalinayuq to serve community members in the area as settlement increased. Julian Vilca Mejía told me that when he returned to the district in 1961, after spending twenty years working in Lima, Catalinayuq’s school was helping to draw settlers to the Rio Cachi zone.624 By 1966, it had 34 students.625 Services were slowly following settlement, and community members in the early 1960s no doubt believed that the trend would continue in the future.626

622 Quispillaccta, Libros de Actas de las Assembleas, 1947-62, Municipal Archives: meeting Feb. 1947. There were several additional schools across the river in Chuschi which also served Quispillaccta’s families. Salmon Galindo told me that the first primary school in Quispillaccta was built in 1938. Interview, Salomón Galindo Achallma, 77, Cuchoquesera, 28 October, 2004. 623 Quispillaccta, Libros de Actas de las Assembleas, 1947-62, Municipal Archives: meetings Dec. 1959, January 1960, and 31 August, 1960. 624 Interview, Julian Vilca Mejía, Llacctahurán, 8 October, 2004. 625 Saramiento Medina et al., Los Distritos de Vischongo, Chuschi, Y Ocros, table 1, 60. I have been unable to determine when the school was built in Catalinayuq, nor when the schools were built in the Matriz. A survey in 1968 found about 50 families living in Catalinayuq. César Ramón Córdova, Estudio de la Vivienda en los Centros Poplados de Inkaraqay, Chuschi, y Catalinayoq, 26. 626 Studies of education in the district done in 1966 by IEP showed that the 1960 clashes led to hundreds of students leaving schools located in the Rio Pampas region, with the majority leaving from Chuschi and Quispillaccta’s schools in the Matriz. Ramon et al., La Comunidad de Chuschi, 130. Quispillaccta families had long sent their children to schools in Chuschi’s Matriz, a trend that continues to this day. Parents no doubt pulled their children from school out of fear for their safety following the clashes in 1960, a trend that probably increased demand for new schools in the settled Rio Cachi zones. As well, land for new schools in the Rio Pampas settlements was scarce given the limited supply of valuable arable land in the area, while there was abundant land for new schools in the

195

The prospect of road construction into the district, and particularly into the Rio Cachi zone, also helped to draw settlers to the area in the 1950s and 1960s. Isbell’s early research on Chuschi noted the impact of road construction in opening up the communities in the district, both commercially and culturally.627 The expectation of increasing road development in the district no doubt contributed to community residents making livelihood decisions that would position them to take advantage of new commercial opportunities once the road was completed, including settling in the Rio Cachi basin to pursue expanded herding opportunities. Based on the experience with road construction in the southern portion of the district and attitudes observers recorded about expanded road construction in the Rio Cachi zone of the district, road construction probably contributed to increasing settlement of the higher zones in the late 1950s and early 1960s.

Roads first entered the district from the south-east in 1961, though residents no doubt knew that they were coming a few years before. In 1961, the government of Ayacucho completed a road to Cancha Cancha, via the Rio Pampas to the east. The road was completed to Chuschi that same year by communal labour. An extension of the road to the Matriz centre of Chuschi and Quispillaccta’s came in 1966.628

Chuschi’s new road link increased the town’s commercial importance in the area, and widened livelihood opportunities for comuneros and elites in the district.629 The changes brought by road access also increased conflict in the communities of the district, particularly between comuneros and elites. Some informants noted that various mestizo elites in Chuschi opposed the building of

new permanent settlements in the Rio Cachi zone. The scarcity of arable land in the Quechua zone of the district was another factor that argued for new school construction to take place in the high-altitude grazing zones. Additional schools in the Rio Pampas zone were not built, in fact, until the 1970s, when both communities used land confiscated from the church holdings in the area to build new schools. To this day, new infrastructure projects in the Rio Pampas zone generate conflict when they force the expropriation of valuable and limited Quechua zone lands. During field work in October and November 2004, the district’s road system was being expanding to various barrios that had no road links. In a community meeting in Chuschi on October 24th, several comuneros complained about losing precious Quechua zone lands to community-approved development projects. Heilman’s study of Carhuanca, Ayacucho noted similar disputes around school construction. Heilman, Before the Shining Path, 102. 627 Isbell, To Defend, 22 and 33. 628 Ramon et al., La Comunidad de Chuschi, 11; Isbell, To Defend, 46. There is some dispute among authors about the date when the road reached Chuschi. The IEP study from 1967 says that the community completed the section of the road from Cancha Cancha to Chuschi in 1961. Isbell, by contrast, says that by 1961 the road had only reached Cancha Cancha and didn’t reach Chuschi until 1966. 629 Isbell, To Defend, 49; Palmer, Revolution from Above, 256.

196 the road because they feared that the changes it brought would undermine their power.630 Other research shows that these same elites were also best placed to take advantage of commercial opportunities brought by the road. According to Palmer, some elites in Chuschi took advantage of the road to monopolize commercial transport to and from Chuschi for several years in the early 1960s.631

In the early 1960s, district residents also looked forward to the new road connections in the northern, Rio Cachi zone. A few kilometers north of the district, the Via de los Libertadores (Highway of the Liberators) was being built, “connecting [Huamanga] directly with the coast at Pisco for the first time.”632 Communities south of the highway began building spurs to connect with the new highway, opening up a more direct route to the markets in Huamanga for commodities.633 The prospect of a northern route into the district excited residents and no doubt helped spur migration to the area in anticipation of new livelihood opportunities. In 1967, Quispillacctinos told researchers that they were very excited about the conclusion of the Huamanga to Ica highway. It would permit easier trips to Huamanga and the coast, and possibly allow them to establish a weekly market in the barrio of Tuco that would draw community members from the D. Chuschi and the neighbouring districts of Paras and Totos.634 Tuco lies on the western edges of the district. So, for Quispillacctinos, a road into the Rio Cachi zone would enable them to have their own transportation link from the far corners of their community to major urban areas outside the district. Controlling the key Rio Cachi zones of the district therefore facilitated these linkages to occur. Emerging road networks in the early 1960s thus provided important livelihood imperatives for settling and controlling the Rio Cachi zones of the

630 Interview, Ernesto Jaime, Chuschi, 18 Oct., 2004. Jaime noted that Chuschi’s priest also worked with Chuschi’s Misti elites to oppose road construction to the town, attempting to frighten Chuschinos by saying that road construction would bring more thieves to Chuschi. His account was confirmed by Chuschino Alejandro Allcca Vilca, who said that Jaime argued in assemblies that the proposed road was good for the people. Allcca said that Jaime was hated by Chuschi’s Mistis for his constant agitation and education of comuneros against Misti domination. Mistis derisively called Jaime, “Llama cejo” – Llama eyebrow. Interview, Alejandro Allcca Vilca, Chuschi, 23 October, 2004. 631 Palmer, Revolution from Above, 256. 632 Palmer, “Rebellion in Rural Peru: The Origins and Evolution of Sendero Luminoso,” 134. 633 Palmer, Revolution from Above, 247. Palmer notes that the owners of the hacienda Milpo built a road in 1967 to connect to the Via de los Libertadores. Marked on the topographical map, the road sits only a few kilometers north of the northern boundary of the District of Chuschi. 634 Saramiento Medina et al., Los Distritos de Vischongo, Chuschi, Y Ocros, 13.

197 district, particularly for Quispillacctinos, who did not have direct access into their community lands up until this time without crossing into the territories of Chuschi and its annexes. Given centuries of conflict between the communities, the importance of new road access to their community free of interference from neighbouring communities cannot be overstated.635

Inter-community clashes in 1960 over control of disputed high Sunni and Puna zones

The rising peasant unrest in the late 1950s throughout highland Peru, but particularly in the central Sierra, also contributed to altering the opportunity structure for peasants in the district by raising expectations of agrarian reform in the years to come. While peasant movements in Peru the 1950s and early 1960s have been discussed above, their impact can be seen in the increasing clashes that they helped spawn between communities and landholders in and around the district in this period. Every Peruvian political reform to affect peasants since the early 1830s had repercussions in the district, in the form of clashes, livelihood changes, petitions to the state, etc., as peasants maneuvered to create livelihood opportunities in the changing political context, particularly in relation to securing more permanent tenure and entitlement rights to the land they believed belonged to their community. In light of the attractions of settling the Rio Cachi regions of the district discussed above in section 6.3, and the pressures pushing residents to the area from the Rio Pampas settlements, discussed below in chapter 7, the stage was set for more aggressive attempts to assert control over the region by competing communities.

Over the course of about two months from late March to early May 1960, serious violent clashes erupted between Quispillacctinos and Chuschinos (including affiliated annexes) over control of three regions of the central sections of the Rio Cachi Sunni and Puna zone in the district: an area around Loreta, on the north-west border between Chuschi and Quispillaccta; an area on the west of the disputed zone stretching from Yuracc-Coral, Suyoccacca to Cconchulla, at the top of the valley with the Matriz; and an area in the east of the disputed zone stretching from Ingauasi, Pallcca, Llachocc, Tapacoccha, Kimsa Cruz, and Arapa. (See map.) Determining which community escalated the conflict in the late 1950s was challenging for regional authorities at the

635 Road building continues in the district today. During fieldwork in 2004, a road was under construction from Quispillaccta’s barrio of Llacctahurán to the Matriz. Since then, new roads have been built to connect even the most isolated barrios like Socobamba to the Matriz.

198 time, as La Serna points out.636 It remains difficult to point to one side or the other as the instigator of the conflicts, given the slow migration to the area over many years. Authorities from both Chuschi and Quispillaccta claimed in various official complaints that residents from both communities engaged in land invasions, crop plantings, and corral construction on land claimed by the other. Quispillacctinos complained of invasions between 1957 and 1958 around Loreta and in the western disputed portion of the Rio Cachi basin, stretching from Cconchulla to Jeulla. Their complaints included claims that Chuschino settlers were laying claim to Quispillaccta’s water resources in the area, water that fed land in Puncupata and Unión Portrero.637 In late 1959 they added complaints against Chuschino invasions to the east, around Accopampa.638 Chuschinos claimed similar violations in late 1959, including threats of physical violence, by Quispillacctinos in an area called Huacctacancha.639 As Ernesto Jaime – the district Mayor in 1960 and a key leader of Chuschi’s violent response to Quispillaccta’s mobilizations – explained to me in 2004, Chuschi had more land as the district capital and was thus the target of invasions by the annexes. “Why are you invading this land? This land belongs to Chuschi,” he would tell the invaders. “We need the land,” came the reply.640

636 La Serna, The Corner of the Living, 68-9. La Serna’s account of the conflict between Chuschi and Quispillaccta in 1959 and 1960 relies almost entirely on court records of complaints and counter-complaints by the communities. This is problematic history, because he writes a history of the clashes using these records as if the details described in the complaints were fact. Little research is done on the wider historical context of these events in La Serna’s research. Clearly, there is tremendous room for exaggeration, over-statement, and outright lies by both communities in official legal complaints. Both were trying to convince authorities of the veracity of their claims and portray themselves as victims of the other’s actions. We should be careful of accepting the details of legal complaints as fact, as La Serna does in several publications without multiple sources of confirmation. La Serna’s account is also published in Miguel La Serna, “To Cross the River of Blood: How an Inter-Community Conflict is Linked to the Peruvian Civil War, 1940-1983,” in Christine Hunefeldt and Misha Kokotovic eds., Power, Culture, and Violence in the Andes (Portland, OR: Sussex Academic Press, 2009): 110-144; and Miguel La Serna, The Corner of the Living: Local Power Relations and Indigenous Perceptions in Ayacucho, Peru, 1940-1983, Ph.D. Dissertation, University of San Diego, CA, 2008). 637 La Serna, The Corner of the Living, 68-9. Interview with Pedro Conde Nuñez, Quispillaccta, 11 October, 2004 638 La Serna, The Corner of the Living, 68. 639 La Serna, The Corner of the Living, 68. I am unable to locate this area on various maps of the disputed region. 640 Interview, Ernesto Jaime, Chuschi, 18 Oct., 2004. Numerous Quispillacctinos noted that Jaime was a key planner and leader of Chuschi’s violent response, a conclusion that agrees with La Serna’s analysis. Informants also named Roberto Miranda as a key leader of the violence. His role is not discussed by La Serna. Interview, Quispillaccta, 12 October, 2004.

199

At the end of 1959, provincial authorities told both sides to refrain from land invasions and property destruction in the disputed zone until an investigation could be completed.641 However, several months later, in April and May, 1960, the bloodiest clashes to date erupted. In early April 1960, Quispillacctinos decided to take matters into their own hands and forcefully push out Chuschino settlers from some disputed areas. Between March 30th and April 3rd, hundreds of Quispillacctinos pushed Chuschinos from the central and eastern portions of the disputed zone, around Accoccasa, Lachocc, Yuracc-coral, Pallcca, and Ingahuasi. Settlements were destroyed, corrals were demolished, and even Chuschi’s chapel at Lachocc was raised to the ground.642 A few days later, between April 5th-7th, hundreds of Quispillacctinos again attacked, clearing away Chuschino settlers in the western and south-central portions of the disputed lands, Suyuccacca, Qenhua, Tapaccocha, Quimsacruz, Pachanca, Cconchalla, and again at Lachocc.643 These widespread and spontaneous mobilizations of Quispillacctinos to seize disputed Rio Cachi lands provoked a violent response from Chuschi.644

In the weeks that followed Quispillaccta’s mobilization, Chuschi’s Misti authorities met to plot their revenge on Quispillaccta and to mobilize Chuschi’s comuneros masses.645 On 16 April, 1960, in an assembly in Chuschi, dozens of Chuschinos pledged their support to initiate litigation against Quispillaccta for their actions, and vowed to defend their lands “materially or personally until the end of the litigation.”646 Quispillacctinos even heard rumours during these weeks that Chuschinos were purchasing handgun ammunition, according to accounts dictated to provincial authorities several months later.647

Chuschi’s response to Quispillaccta’s mobilizations came a few weeks later on May 5th and 6th with attacks on Quispillacctino settlements in Ccullahuaycco in the south-west of the disputed

641 La Serna, The Corner of the Living, 68-9. 642 La Serna, The Corner of the Living, 69-70. 643 La Serna, The Corner of the Living, 70. 644 Don Pastor Galindo Ccallocunto, from Puncupata, noted in an interview that the Quispillaccta mobilization was not organized by Quispillaccta authorities, but was a spontaneous mobilization of community members. Galindo, in his early 20s at the time of the conflict, was an authority in Quispillaccta in the 1970s. Interview Pastor Galindo Ccallocunto, 11 October, 2004. 645 La Serna, To Cross the River of Blood, 116. 646 Quoted in La Serna, To Cross the River of Blood, 116-117. 647 La Serna, To Cross the River of Blood, 117.

200 zone and around Kimsa Cruz in the east. Almost 200 hundred Chuschinos on foot and on horseback first destroyed the small ranching settlement of Ccullahuaycco, in an area that is now part of the barrio Yuraqcruz.648 Some residents of the area fled to Tuco for assistance, according to Juan Galindo. “Cculla doesn’t exist anymore,” they said.649 The next day several hundred Chuschinos confronted hundreds of Quispillacctinos at Kimsa Cruz. When the bloody day’s confrontation was over, three Quispillacctinos lay dead from gunshots, while at least a dozen more were wounded by gunshots.650 Dozens of other community members on both sides were wounded in the melee by clubs and rocks.651 Quispillacctinos would get their own fatal revenge two years later, chasing down two Chuschinos in the Rio Cachi zone and beating one to death. Both may have played a role in the violence of May, 1960; the dead man, Miguel Pacotaipe, had a dwelling destroyed when Quispillacctinos cleared Suyoccacca of Chuschi settlers.652

The outcome of the violence between 1959 and 1962 clearly benefited Quispillaccta’s interests in controlling the Rio Cachi basin of the district. Both sides will tell you that Quispillaccta’s superior numbers during the clashes turned the tide in their favour, despite the use of fire arms by Chuschi’s mistis. Chuschi managed to destroy holdings in Ccullahuaycco and draw blood in Kimsa Cruz in 1960, but they were unable to reverse Quispillaccta’s clearing of Chuschino settlers in much of the high zones of the Rio Cachi basin, particularly in the western section. In the eastern section of the disputed zone Quispillaccta also made gains, particularly in the highlands above Pampamarca, while further south both communities faced each other across a tense line of division from Llachocc, Pallcca, to Kimsa Cruz. Ccullahuaycco remained in Quispillaccta’s hands, along with gains in Cconchulla to the east, above the main valley. The state proved unable to mediate a permanent solution to the 1960s clashes; in fact, they were even incapable of prosecuting anybody from either Chuschi or Quispillaccta for the death, bloodshed,

648 La Serna, To Cross the River of Blood, 119. 649 Interview with Juan Galindo, Oct. 11, 2004. 650 The Quispillactino dead included brothers Martín and Sebastien Mendieta, and Antonio Galindo Espinoza. I interviewed Sebastien Mendieta’s widow, Pascualia Huamaní, in 2004. 651 La Serna, To Cross the River of Blood, 119-120. 652 La Serna, To Cross the River of Blood, 121-124.

201 and property damage suffered.653 While litigation continued in the next decade by both sides to cement gains or reverse loses, control on the ground was all that mattered in the end.

Détente and maneuver – continuing struggles to control disputed zones between Chuschi and Quispillaccta, 1960s and 1970s

In the almost two decades between the bloody events of May, 1961 and the takeover of the district by the Shining Path in late 1981, the dispute over the Rio Cachi disputed zone flared repeatedly, resisted attempts during the Agrarian reform era to find a peaceful solution, and again threatened to lead to all-out conflict between the two communities. While many community members simply tried to get along in order to make a living, at various times in the 1960s and early 1970s, incidents of property destruction, violent threats, and land invasion were recorded in the judicial record – primarily from the Loreta zone in the north-west of the disputed area, around Ccullahuaycco and Conchulla in the south-west of the disputed area, and around the line of division in the east between the two communities, particularly around the grazing area of Llachocc, where high-altitude streams feed Pampamarca and Catalinayuq.654 The events of 1960 had by no means settled the conflict, though large-scale open conflict instead gave way to isolated incidents of hostility between the communities over disputed territories.

National agrarian reform in the early 1970s introduced new dynamic opportunities in the highlands for campesino communities, motivating many to seek control of disputed or usurped lands. As discussed above, the reality of land reform rarely matched the results; however, the promise of a more sympathetic government led many peasant communities to seek redress for past injustices and settle outstanding land issues. During the agrarian reform years in D. Chuschi, community members from both sides made several attempts to involve the reformist bureaucracy in an effort to resolve the dispute between Quispillaccta and Chuschi. Chuschino Marcelino

653 The lack of state capacity and presence in the area in the early 1960s was reflected by the failure to hold anyone responsible for the violence of 1960. While a few community members on both sides spent a small amount of time in jail, no one was ever prosecuted for the murders, shooting, or beatings that took place. Chuschino Victor Galindo Tucno told me that after the Kimsa Cruz violence, he remembers Chuschi’s lawyer laughing at Chuschi’s comuneros, and telling them that they could have killed 20 Quispillacctinos and nothing would have happened. Interview, Victor Galindo Tucno, Chuschi, 24 Oct., 2004. The lawyer’s comment also reflects the long-standing reality in the highlands that Mistis in the district could abuse and even kill comuneros with near impunity. 654 La Serna, To Cross the River of Blood, 126-7.

202

Rocha Cayllahua, who was among the group of authorities that settled the conflict with Quispillaccta in 1981, recalled a failed attempt in 1971 by the Liga Agraria (Agrarian League) to resolve the dispute.655 Quispillacctinos also looked to the new reformist government for solutions to the dispute, agreeing in 1972 to participate in an effort organized by SINAMOS (The National System for the Support of Social Mobilization) to find a resolution to the land disputes between the communities. As in the past, the community may have calculated that the new reform bureaucracy could help solidify land claims in contested areas. Quispillacctinos were angered when Chuschinos spurned the SINAMOS peacemaking effort.656 This may have led the community in early 1973 to ask SINAMOS to undertake a census of their land, animals, and community members.657 A census could provide some legitimacy for their holdings.658 These and later efforts at reconciliation throughout the 1970s failed, however. Chuschinos rejected attempts at conflict resolution and continued to appeal to provincial and court authorities to try to seek redress for their claims that Quispillaccta usurped their lands.659 As one Chuschi community leader later told me about Chuschino views at the time, “Chuschinos weren’t interested in finding a solution in the 1970s.”660

In mid-May, 1980, four students from the University of Huamanga entered Chuschi’s town hall and burned the ballot boxes for the upcoming national elections. This act of anti-regime violence is often held up as the first shot in the Shining Path’s insurgency in Peru.661 However, in the D. Chuschi, tensions had already begun to heat up weeks before Sendero militants attacked the voting office. Since March 1980, conflict was again erupting in the disputed Rio Cachi zone between Quispillaccta and Chuschi. Over the course of the next eighteen months, hundreds of Quispillacctinos and Chuschinos were involved in back and forth land invasions, violent threats,

655 Interview, Marcelino Rocha Cayllahua, Chuschi, 24 Oct., 2004. The Liga Agraria was a new national body of peasant federations set up during then President Velasco’s agrarian reform efforts. This failed conciliation the Liga Agraria was confirmed by Emilio Nuñez of Quispillaccta in an interview in 2004. 656 La Serna, To Cross the River of Blood, 127. 657 Quispillaccta, Libros de Actas de las Assembleas, 1973-75, Municipal Archives: meeting 21 January, 1973. 658 Peru’s first national agrarian census was released in 1972. It is also possible that the Jan. 1973 entreaty to SINAMOS was an attempt to correct what the community felt was their faulty representation in that census. Some Peruvian scholars note that there were various flaws in the 1972 agrarian census. 659 La Serna, To Cross the River of Blood, 127. 660 Interview, Chuschi, 2004. Identity protected. 661 Gorriti, The Shining Path, 17-20.

203 the destruction of property, and the creation of new settlements in disputed areas. Both communities filed numerous legal petitions to provincial authorities as a result of the renewed hostilities. This time the conflict was concentrated in the north-west of the disputed zone, around Loreta, and in the eastern section of the zone – areas around Ingahuasi, Pallcca, Llachocc, Tapacocha, Arapa, and Quimsa Cruz.662

National and regional political developments were again providing a structural opening for peasant actions in each community. With national elections in 1980, Peru was emerging from more than ten years of military control of national politics, a retreat from power by Peru’s Revolutionary Government of the Armed Forces that began in 1977.663 Most of Peru’s leftist parties joined the political process and were actively courting rural voters with their platforms of change. Ayacucho’s countryside saw vigorous political competition among competing leftist groups. In Cangallo, however, the Shining Path, which had rejected participation in the election and had been organizing armed insurrection for several years in rural communities, deployed its militants to harass and intimidate other leftists groups stumping for voters in communities up and down the Rio Pampas.664 In eastern Cangallo they took advantage of the absence of all local police forces, which had been withdrawn from the area in 1978 by Department police officials after days of violent clashes with students and teachers over high school reform. The school reform violence had an important impact on the political atmosphere in Cangallo, according to Heilman: “The … strikes triggered an immediate radicalization of politics throughout Cangallo province and particularly in the eastern districts of Vischongo, Vilcashuamán, Huambalpa, and Carhuanca … Senderistas throughout eastern Cangallo took the 1978 police retreat as their opportunity to finalize their preparations for war” over the next two years.665 Community members in the D. Chuschi were no doubt aware of the political winds swirling around them as each resorted to direct action to gain control of the disputed Rio Cachi zone in 1980 and 1981. The extent of the invasions, property destruction, and mobilizations by each side reached levels comparable to the conflicts in 1959-60. Tensions appeared to be nearing their peak, in June

662 La Serna, To Cross the River of Blood, 130-135 gives details of the legal claims and counterclaims of violence and destruction from each event. 663 Heilman, Before the Shining Path, 174. 664 Gorriti, The Shining Path, 10-13. 665 Heilman, Before the Shining Path, 176.

204

1981. According to a Chuschino complaint to provincial authorities at that time, Quispillaccta had even approved in a general assembly a plan to invade the disputed lands around Loreta/Huaracco, Ingahuasi, Pallcca, and Pucahuasi666 – disputed areas in the north-west and eastern portions of the contentious Rio Cachi zone. The stage appeared set to repeat the deadly clashes of 1960.

Peace in the midst of civil war – final settlement of land disputes in the District of Chuschi and the impacts of this settlement in the context of Sendero’s insurgency in the area 6.6.1. The 1981 peace agreement between Chuschi and Quispillaccta

As the planting season began in the district, violence between Quispillaccta and Chuschi appeared to escalate in September and early October, 1981.667 But suddenly, on Oct. 16th 1981, Quispillaccta’s communal assembly was informed that a formal agreement had been reached with Quispillaccta to mark the boundary in the disputed zones of the high Sunni and Puna.668 Months of behind-the-scenes efforts by community authorities dramatically shifted the district toward a historic peace settlement. We now know that the peace agreement came together over several months, from May to October 1981, at the same time that violent confrontations in the Rio Cachi zone were reaching their peak. Existing scholarly accounts of the rising conflict between Chuschi and Quispillaccta in 1980-81, during those initial months of Sendero activity in the district, have failed to describe the peace settlement and explain its wider significance for conflict and violence in the area during the first years of the Sendero insurgency.669

The ground work for the agreement was laid by community leaders in both Quispillaccta and Chuschi several years earlier, and by changing attitudes toward maintaining the high costs of

666 La Serna, To Cross the River of Blood, 132. I did not find any record of such a decision in Quispillaccta’s community records. 667 La Serna, To Cross the River of Blood, 130-35.

668 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 16 October, 1981. La Serna’s account of the clashes in 1980-81 outlines records of complaints in the back and forth clashes between the communities up to 12 October, 1981. His account ends there, however, curiously omitting any discussion of the peace agreement that was signed only days later. 669 None of the scholarly accounts of events in the D. Chuschi during 1980-83 – Isbell, Sànchez Villagómez, or La Serna - give an account of the peace agreement and its significance.

205

continued litigation. Quispillaccta’s President at the time of the 1981 deal, Emilio Núñez, said that his predecessor as President, Armando Tomaylla, started to talk to Chuschinos about a solution in 1978-9.670 When Núñez took over as Quispillaccta’s President in 1980, he found a receptive partner on the Chuschi

Peacemakers: Quispillaccta President Emilio Núñez Conde (left) and Chuschi side for negotiating a deal in President Marcelino Rocha Cayllahua (right), who were instrumental in negotiating the 1981 peace agreement between Chuschi and Quispillaccta. Photo Chuschino President Marcelino sources: Salvatierra et. al., 205; https://solucionespracticas.org.pe/aprendiendo-a- construir-su-propia-casa. Rocha. Both were young community leaders, in their early 30s, and tired of the long conflict that continued to drain community coffers. Foreign development professionals working in the area at the time also noted that campesinos were increasingly aware of the lost development opportunities from spending scarce community funds on continued boundary litigation.671 In an interview, former Chuschi President Juan Carhuapoma noted that Chuschinos were keenly aware of the drain on community coffers from the on-going litigation with the Hacienda Yaruca and its former peons. This helped to motivate

670 Interview, Emilio Núñez Conde, Quispillaccta, 13 October, 2004. 671 Vera Gianotten and Ton de Wit, "Organizacion Popular: El Objetivo De La Investigacion Participativa,” Boletín de Estudios Latinoamericanos y del Caribe 35 (December 1983), 115. Gianotten and de Wit worked on rural development programs for the Netherlands’ Royal Tropical Institute in the D. Chuschi between 1978 and 1983. Their article discusses their experiences in trying to design and deliver rural development programs in the district, as part of Huamanga University’s Centro de Capacitación Campesina (Peasant Training Center, CCC) a program for rural extension assistance and training that was part of the University’s Allpachaka research station. The CCC worked with various peasant communities in the Rio Pampas and Rio Cachi river basins, in cooperation with foreign development agencies like the Swiss Agency for Cooperation and Development COSUDE (Agencia Suiza para el Desarrollo y la Cooperación). The Allpachaka research station was a former hacienda north-east of the D. Chuschi, noted on topographical maps of the area. The lands making up Allpachaka were claimed by many peasant communities in the area, including by Quispillaccta, according to informants. Allpachaka’s destruction by Sendero in 1982 marked an important and controversial moment in Sendero’s attempts to take over Cangallo. A detailed discussion of Allpachaka and its destruction can be found in, Michael L. Smith, Rural Development in the Crossfire: The Role of Grassroots Support Organizations in Situations of Political Violence (Ottawa, IDRC, 1991), 33-43. See also Kurt Burri, Los Inicios de la Cooperación Técnica del Gobierno Suizo, 1964-1974 (Lima: Agencia Suiza para el Desarrollo y la Cooperación (COSUDE), 2000).

206

Chuschi to make peace.672 These changes no doubt helped set the stage for the efforts of Chuschi’s and Quispillaccta’s authorities to make peace.

While various legal and political officials urged the leaders to strike a deal, tensions on both sides made it difficult to find a solution collectively.673 As conflict continued in the contested areas, representatives had to overcome community skepticism and their own suspicions of the other side. According to Quispillaccta authority Emilio Núñez, Ernesto Jaime, the former District Mayor and Chuschi leader in the 1960 conflict, laughed at them when they sought to begin negotiations, saying that the conflict was unsolvable.674 One meeting in early 1981 in Huamanga between Chuschino and Quispillacctino authorities devolved into fisticuffs in the Plaza de Armas.675 Rocha and Nuñez, however, committed to working together, with minimal outside interference, to try to find a solution to the conflict.676 In spite of the skepticism of some, the vast majority of Chuschinos were in favour of starting the peace process.677 Quispillaccta’s support for the process is recorded in the records of the community assembly in May 1980, where the community approved a statement for the record committing each side to putting forth two representatives to analyze and negotiate a peaceful resolution of their problems.678 The representatives were even going to explore the possibility of ‘unification’ of the communities.679

672 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. 673 Juan Carhuapoma Llalli, Chuschi’s President after Rocha, said that lawyers from each community first suggested that the Presidents of Chuschi and Quispillaccta meet in Huamanga to try to solve the dispute. Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. Both sides also mentioned the mediation attempts of a student, Gerardo Allcca Nuñez, who tried to participate in the dialogue to find a solution. 674 Interview, Emilio Núñez Conde, Quispillaccta, 13 October, 2004. 675 Interview, Marcelino Rocha Cayllahua, Chuschi, 24 October, 2004. Rocha says that Huamanga police and Chuschino Victor Galindo, who was a Liga Agraria member, calmed the authorities and chastised them for the fight. “How long is this going to continue?” Galindo asked them? Rocha said that Galindo’s comments made him think more seriously about finding a peaceful solution. 676 Interview, Marcelino Rocha Cayllahua, Chuschi, 24 October, 2004. 677 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. Carhuapoma said that 90% of Chuschinos were in favour of beginning the process to find a solution. 678 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 10 May, 1980. 679 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 10 May, 1980. The voices for peace were not unanimous. One Quispillactino, who at the time was Quispillaccta’s Justice of the Peace, is recorded in the minutes of the meeting as rejecting unification and calling on Quispillaccta to return Tapacocha to drive out Chuschinos.

207

In early 1981, Quispillaccta’s official community records show that the community remained committed to finding a negotiated solution to the problems with Chuschi, even as conflicts began to escalate in the disputed zone.680 Tensions had flared in a new area in the south-west of the district, where Quispillaccta’s Tuco barrio holdings abutted Chuschi’s high puna grazing zone. Herminio Nuñez told Quispillaccta’s assembly on March 16th, 1981 about the destruction of his house and corrals in Huiscahuaycco.681 It is possible that the back and forth invasions in the months leading to the final agreement and the opening of a new front in the conflict in the south- west were unsanctioned maneuvers by community members on each side to claim as much territory as possible before a final agreement was signed. In a community meeting soon after an early March 1981 incursion by Chuschi, Quispillacctino Antonio Huamani reiterated the importance of inhabiting Quispillaccta’s land near their boundaries, no doubt to protect against invasions by their neighbours.682

As the July/August dry season passed to the beginning of the planting season in September and October 1981, tensions in the district began to increase significantly. Sendero was increasingly active in the community, with its moralizing and conscious-raising efforts.683 In August, Sendero militants had executed two Quispillacctino cattle thieves in Chuschi’s main plaza.684 Invasions and property destruction in the disputed zones increased significantly in September and early October. On October 16th, however, Quispillaccta’s community assembly was told that a formal agreement had been reached with Chuschi.685 They agreed to have both sides leave the disputed areas so that both communities could quickly define the final boundaries. Provisions

680 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 16 March, 1981. Chuschi’s records of their community assemblies were destroyed when the Shining Path burned the district office on April 9, 1983. This act is noted by Marté Sánchez Villagómez, "El Horror Olvidado. Memoria E Historia De La Violencia Política En Ayacucho (1980-2000)" (Tesis Doctoral defendida en la Universidad Autónoma de Barcelona, 2015), 284. 681 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 16 March, 1981. On the topographical map of the area included above, Huiscahuaycco is spelled Huiscahuayocc, and is highlighted in the far west of the disputed zone. 682 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 16 March, 1981. The Chuschino incursions of 3 March 1981 are described in La Serna, To Cross the River of Blood, 135. 683 Isbell, “Shining Path and Peasant Responses,” 83. 684 Isbell, “Shining Path and Peasant Responses,” 83. 685 An open question is whether the violence and invasions of September and October pushed both sides to recognize that they had reached a hurting stalemate, and that continuing violence would be unbearable?

208 were also agreed upon for penalties in the event of violations of these borders. As well, all outstanding official complaints or claims were to be dismissed by both sides once the agreement was reached.686 In another assembly in Quispillaccta on October 18th, President Nuñez informed the community that both parties had reached an agreement on most issues, but they needed to finalize the agreement to make it a reality by defining the borders. This process would start with Pucahuasi first, before dealing with Huancarumi, Ingahuasi, Pallcca y Conyacalle, Huaracco, and Kimsa Cruz – the latter encompassing the lands in the north-west and eastern area of the Rio Cachi disputed zones.687

Some Quispillacctinos were unhappy about the peace process, their complaints recorded by the assembly minutes. Several barrio presidents complained that Quispillaccta was losing out in the deal because they did not have enough land compared to their neighbours in other barrios. Nuñez dismissed these complaints, saying that continuing a trial was too expensive and that both communities suffer by continuing legal actions. In the end, the assembly voted to continue with the peace process.688 Some Chuschinos were similarly unhappy with the deal. In an interview, Rocha said that some Chuschinos accused him of selling out to Quispillaccta for some bulls – that he was bribed into making the deal. Others, thanked him for making the deal, however. 689

Negotiators moved quickly in late October and early November 1981 to settle the disputed boundaries. On 22 October they defined the boundaries for Pucahuasi and Huancarumi, setting out the agreement first in text, with markers to be placed later.690 On October 28th they met in

686 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 16 October, 1981. The penalty set out in the records in the event of a violation is that each side would forfeit their entire claims, to be ruled upon by a land judge. This is a significant penalty and ensures that each community would police its own citizens to protect its claims. 687 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 18 March, 1981. I have been unable to locate Pucahuasi, Huancarumi, or Conyacalle on community maps. Pucahuasi and Huancarumi are probably in the south-west of the community, where Quispillaccta’s Tuco barrio lands abut Chuschi’s high puna grazing zone and where clashes broke out in early 1981. In this area, the topographical map notes Pucruhuasi and Cerro Pucamachay, so it is reasonable to conclude that these disputed areas are nearby. The map shows numerous high-altitude wetlands, which were typically the areas fought over because they provided water and pasture areas when some wetlands dried up in the dry season. In retrospect, it was smart to begin the boundary marking in this area, because the depth of tensions on both sides was likely least evident in the area where conflict had only recently arisen. 688 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, meeting 18 March, 1981. 689 Interview, Marcelino Rocha Cayllahua, Chuschi, 24 October, 2004. 690 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 22 October, 1981.

209

Kimsa Cruz, the site of the May 1960 violence that claimed the lives of three Quispillacctinos, and set the boundaries from Kimsa Cruz to Ingahuasi.691 A fight almost broke out among negotiators as they set out the Kimsa Cruz boundary, according to Chuschi authority Juan Carhuapoma. However, tensions cooled on both sides, partly with the assistance of a Swiss NGO representative who was present at the meeting. Both sides agreed to suspend the process for a bit to let tempers cool. 692 The next week, on November 4th, boundaries were set from Ingahuasi to Pallcca.693 Two weeks later November 18th, the final boundary in the disputed zone in the north-west was set, between Chuschi’s Huaracco barrio and Quispillaccta’s Unión Portrero.694 In less than one month, authorities from Chuschi and Quispillaccta defined boundaries that had been a source of conflict for decades. The complex process of defining the boundaries was complicated by the legacy of resource and entitlement capture in the area from the colonial era. Carhuapoma noted that at times during the boundary marking there were obvious overlapping claims from community land titles to the same area, so negotiators had to compromise and split the difference, using natural boundaries to complete the demarcation in many places.695

With all the boundaries now settled, each community presented the final peace agreement to their community assemblies at the end of November 1981. The communities agreed to present the ratified peace deal to Peruvian government authorities in December. Opposition to the deal continued from some in both communities, however. Remarkably, former Quispillaccta President Armando Tomaylla Nuñez from barrio Catalinayuq, who had initially broached the idea of a deal with Chuschi during his time as community president, now condemned the deal, calling on President Emilio Nuñez to quit so that a new community president could press for more land.696 Some Quispillacctinos, particularly those from Catalinayuq barrio were upset that

691 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 28 October, 1981. 692 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. 693 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 4 November, 1981. 694 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 18 November, 1981. 695 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. 696 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 30 November, 1981. Oseas Núñez Espinoza’s online account of events at that time notes that Tomaylla did not want to give up an inch of Quispillaccta’s land to Chuschi during his time as community president in the late 1970s, a statement that somewhat

210

Quispillaccta had made concessions in the boundary demarcation that gave Chuschi part of Pallcca and Arapa.697 These highlands contain the headwaters that feed the pastures of Catalinayuq, an area where water is limited. President Nuñez responded that Catalinayuq was welcome to continue the litigation if they had a basis to continue. In the end, when the proposed agreement was put to a vote, the peace agreement was ratified by Quispillaccta’s assembly.698 In Chuschi, some community members also continued to voice opposition to the deal at the ratification meeting. In the end, however, the momentum for sealing a deal prevailed and a majority of Chuschinos voted to ratify the final boundary agreement, according to authority Juan Carhuapoma.699 When the ratification was concluded, both communities held a communal feast in Quispillaccta to celebrate.700 The agreement would eventually be presented to Peruvian government authorities and a formal plan drawn up with government officials in Lima, according to Quispillaccta’s assembly records.701 The last entry about the settlement comes from late January, 1984, a little more than a year after the Peruvian military had begun its brutal counter- insurgency campaign in the district, when Quispillacctinos were informed that the formal plan of the agreed boundaries had been completed with government officials.702 In the final accounting, the deal that finally solved centuries of conflict between Chuschi and Quispillaccta largely favoured Quispillaccta in its outcome. Even though Quispillaccta made concessions in the eastern zone around Pallca, Tapacocha, and Arapa, they were largely successful in pursuing claims to Loreta – near Unión Portrero, in areas now part of Yuraqcruz, and north of the main valley around Conchulla and Kimsa Cruz. Most Quispillacctinos feel that they came out of the conflict ahead, though they always maintain that the disputed territories

contradicts President Núñez’s account above. , assessed 9-1-16. However, as noted above, 697 This point is made by Oseas Núñez Espinoza in an online history of Cuchoquesera, , assessed 9-1-16; Jose Nuñez, from the barrio of Catalinayuq, blamed Emilio Núñez for giving Chuschi Pallcca. Interview, Jose Nuñez, Quispillaccta, 16 October, 2004. 698 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 30 November, 1981. 699 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. 700 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. 701 Quispillaccta, Libros de Actas de las Assembleas, 1979-81, Municipal Archives, 30 November, 1981. 702 Quispillaccta, Libros de Actas de las Assembleas, 1982-85, Municipal Archives, 25 January, 1984.

211 were always theirs to begin with.703 Certainly, when one examines the official boundary between communities set out in the 1981 agreement, Quispillaccta controls the main central portion of the disputed Rio Cachi high Sunni and Puna zones, creating a contiguous community from the Rio Pampas in the south to the borders of district in the north-east Rio Cachi zone. (See map, Figure 19: Current boundaries of Quispillaccta (black) within the District of Chuschi overlaid with historic conflict zones.) Chuschi is essentially divided in two, with a small high Sunni-Puna zone in the south-east connecting to their Matriz holdings on the Rio Pampas, separated from significant holdings of high pastures in the north-west of the district which connect to the Matriz by a trail through Quispillaccta’s land.704 Some Chuschinos are still stung by their loss of holdings in this zone, apparent in interviews that I conducted in 2004. Numerous informants repeated that Quispillaccta stole their land, and became angry when asked about the conflicts in 1960 and 1980-1, something that I never experienced when talking to Quispillacctinos about the land dispute.

703 This point is made repeatedly to me in interviews and a similar sentiment is expressed by Oseas Núñez Espinoza in an online history of Cuchoquesera, , assessed 9-1- 16. Núñez Espinoza says that Catalinayuq lost control of parts of Pallcca and Arapa because they didn’t completely inhabit the land, allowing an opportunity for Chuschinos to encroach upon the territory and later claim it as their own. 704 Oseas Núñez Espinoza also makes this point. , assessed 9-1-16.

212

Source: Instituto Geográfico Nacional, Perú.

Figure 19: Current boundaries of Quispillaccta (black) within the District of Chuschi overlaid with historic conflict zones

In the end, the weight of realities on the ground, the high costs of continuing litigation over disputed territories, and the explosive political context in the region at the time served to support the visionary leaders on both sides who sought to end the long and costly conflict. Even though Chuschi had pushed back in 1980 and 1981 by invading some of the disputed areas earlier occupied by Quispillaccta, the community was not able to occupy and settle as much of the disputed territory as Quispillaccta. Quispillaccta had larger numbers in their favour and had been using these numbers to their advantage to settle the disputed zones since the 1940s.705

705 The question of whether population growth in Quispillaccta was faster than in Chuschi and whether this played a part in tipping the balance on the ground is explored in the next chapter below in section 7.2, examining supply and demand-induced resource pressures. Certainly, early studies of the district made in the mid-1960s noted that Quispillaccta was going through a “demographic explosion.” Ramon et al. La Comunidad de Chuschi, 151. Though

213

Chuschi’s President at the time of the agreement, Marcelino Rocha Cayllahua, admitted that Chuschinos felt like they were losing with the agreement because they were fewer in number.706 As well, in the major confrontation between communities in 1960 and 1981, the legal records cited by La Serna noted greater numbers of Quispillacctinos involved in the fights.707 Both sides were also tired of spending precious community treasure to continue seemingly endless legal battles over disputed territories, as discussed above. Chuschino Juan Carhuapoma noted that debilitating financial impact of the decades-long Chuschi litigation against the hacendados of Rumichaka-Yaruca was particularly influential in convincing Chuschinos that the time was ripe to settle with Quispillaccta.708 Quispillacctinos were also aware that the costs of continued litigation would drain funds that could be used to address the needs of both Quispillaccta and Chuschi. As then Quispillaccta President Emilio Nuñez said in support of the peace deal: “We’re all poor, we’re all campesinos. How long will we continue fighting! We should be preoccupied with bringing progress to our communities. We settle this peacefully and we make history.”709

Finally, the extraordinary structural conditions that community members found themselves in from 1978 onwards cannot be under-estimated for its influence on pushing the communities to an agreement; though, the precise impact remains contested. At the moment when Peru’s most violent insurgent group was planning to initiate armed conflict against the state by direct action in the D. Chuschi and the surrounding districts, community leaders decided the time was ripe for settling one of the longest documented inter-community conflicts in Peru. While there is likely a connection, the influence of Sendero on the impetus to reach peace remains unclear today. The evidence about Sendero’s role is contradictory. La Serna claims that Sendero chose Chuschi as

the historical demographic data is not very good for each community, my reconstruction of demographic trends in each community seems to confirm that Quispillaccta’s birth rates in the 1950s and 60s exceeded those of Chuschi. Migration trends (also discussed below in section 7.4.1) obviously also influenced demographic trends in the district, and it appears that out-migration was higher in Chuschi than in Quispillaccta for much of the post-war period. 706 Interview, Marcelino Rocha Cayllahua, Chuschi, 24 October, 2004. 707 This is also not conclusive; Quispillaccta could have simply been better at mobilizing their community members, speaking to higher community cohesion than Chuschi, which had been for decades hobbled by the impact of abusive Misti elites controlling the authority positions. 708 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004. 709 Quoted in Salvatierra et al., La Vida, 205. My translation.

214 the start of the violent insurgency “because they hoped to channel inter-village hostilities into the violence of armed struggle.”710 By this logic, Sendero would not have been in favour of a peace deal. La Serna presents no evidence to support this claim, however, and the evidence that exists seems to contradict his claim. If Sendero was looking to aggravate the inter-village hostilities for their own violent struggle, negotiating a peace deal would seem risky for community leaders.

Arguments are more persuasive that Sendero was in favour of a deal; though solid evidence of this remains elusive. Isbell believes Sendero chose Chuschi to start the insurgency because the district’s communities would probably be more receptive to their message, given their “strong communal structures, autonomy over their resources,” and minimal experiences with “capitalistic market penetrations.”711 If this was the case, Sendero would have likely supported a peace agreement if it would facilitate community members being more receptive to Sendero’s message, and help create a more consolidated cadre of supporters.712 Recall that Sendero’s first violent actions in the district aimed to solve community problems – eliminating thieves, bullies, and adulators – demonstrating the value of their organization and strength to community members. So, acting as a peacemaking in inter-community conflicts logically fit with their early actions. It is possible that the Sendero cadres working on consciousness raising in the district helped push the communities to make peace for their own interests to build support among community members in the service of their larger war aims against the Peruvian state. However, there is no documentary evidence for this “soft persuasion” influence on the deal-makers.

A related reason why the communities felt that the time was ripe for a deal was that they believed it dangerous to continue the conflict in the face of impending civil unrest. In this case, the communities may have calculated that the risks of hostilities and violence breaking out in a time of heightened political unrest made continuing with the conflict dangerous. Leaders may have realized that they were nearing a precipice – that continued conflict could lead to much

710 La Serna, The Corner of the Living, 158. More recently, based on new interviews with Elena Iparraguirre, the imprisoned number two leader of Sendero, it appears that Sendero chose this area to begin the armed struggle because of the region’s high poverty. See Starn and La Serna, The Shining Path, 87. 711 Isbell, “Shining Path and Peasant Responses,” 78. Isbell’s argument focuses on the same factors that Eric Wolf hypothesized for the strata of peasants most likely to rebel. See Wolf, Peasant Wars, 291. 712 I call this the “house divided” thesis – in Sendero’s logic, if a house is divided against itself, that house cannot stand in its fight against the Peruvian state.

215 worse violence or catastrophe. Negotiation experts note that such moments are opportune for initiating a negotiated settlement.713 Self-preservation, therefore, may have spurred the move to peace. Again, however, we have no evidence of this.

A similar reason why communities may have felt the dangers of continuing the conflict were unbearable is that they were directly threatened into making an agreement, either by Sendero or by the Peruvian State. Field work in the D. Chuschi provided intriguing but somewhat inconclusive evidence on this explanation. When interviewed about the peace deal, Chuschi’s then-President Marcelino Rocha Cayllahua said that the ranking Sendero commander in the Chuschi zone, Comrade Medina, came to district in July or August of 1981 and threatened both sides to settle the conflict or face consequences. Medina, Rocha said, told them that “now no one will fight for land because we are in a time of war.”714 Meeting with community leaders and teachers in the district, Medina said that they would “hear from him” if they did not reach an agreement. The obvious implication is that this threat thus motivated the parties to expedite the efforts to find a solution.715 Chuschi’s next President, Juan Carhuapoma, confirmed that Sendero’s threat motivated the parties to find a solution, though he was not present at the meeting with Medina, he told me.716 If true, Sendero’s threat would go a long way to understanding why an agreement was reached in the district at that particular time, and would directly contradict La Serna’s analysis.

However, there is no evidence to prove the veracity of the claim of Sendero threats towards Chuschi’s leaders; we can only take Chuschino informants at their word, even though there are good reasons to question whether claims of Sendero threats are in fact a way for Chuschino leaders to seek political cover for their deal. Chuschi’s leaders faced criticism for making the deal, as noted above, so claims Sendero threats ‘motivated’ a deal would partially exonerate Chuschi’s leaders from any criticism or censure from their own community members about making the deal. In such a circumstance, they simply had no choice but to make a deal given

713 William Zartman and Saadia Touval, “International Mediation,” in Leashing the Dogs of War, eds. C. Crocker et al. (Washington: US Institute of Peace Press, 2007), 445. 714 Interview, Marcelino Rocha Cayllahua, Chuschi, 24 October, 2004. 715 Interview, Marcelino Rocha Cayllahua, Chuschi, 24 October, 2004. 716 Interview, Juan Carhuapoma Llalli, Chuschi, 24 October, 2004.

216

Sendero’s threats. Unfortunately, evidence about Sendero threats to force a deal only come from interviews with Rocha and Carhuapoma; in numerous interviews with Quispillaccta’s past authorities and with community members in both villages, no one mentioned Sendero’s threat. Had such a threat taken place, one would expect that it would have been widely known and made to both communities, precisely because a public threat makes the threat more powerful to restrain community members from rash actions in the disputed zone. In addition, a threat in July or August of 1981 fails to explain why violence escalated as it did in September and October, 1981?

There is evidence, however, that the Peruvian military threatened community leaders in 1983 with violence if clashes between communities continued. Quispillaccta’s assembly archives record a meeting in June, 1983, between a Captain from Peru’s Army – likely the infamous Captain Chacal - and the authorities of Quispillaccta, Chuschi, Cancha Cancha, and Uchuyri. The military warned community leaders that there must not be fights of any kind between communities, or else the authorities would be punished. Furthermore, all district residents were to be guaranteed the right of free transit across any community lands, according to Quispillaccta records.717 Given that the military had, by this time, killed dozens of community members in its vicious counter-insurgency campaign, this threat to community authorities was obviously taken seriously. It was a tangible reason for maintaining and implementing the peace agreement ratified the year before. While threats from Peru’s military had nothing to do with the push to finalize an agreement in 1981, military threats were no doubt partly responsible for the implementation and adherence to the agreement in the years that followed.

Conclusion

Various structural changes both within the D. Chuschi and wider trends in Peru and abroad helped condition the momentum for reversing resource capture in the D. Chuschi in the latter half of the 20th century. As populations in the communities recovered to levels approaching pre- Conquest numbers and economic opportunities increased for livestock and agricultural product sales in the late 19th and early 20th century in the highlands, elite and church control of lands inside the communities meant that communities had to look elsewhere to expand production. As

717 Quispillaccta, Libros de Actas de las Assembleas, 1982-85, Municipal Archives, 12 June, 1983.

217 popular pressure for agricultural change built during this time, Quispillaccta, Chuschi, and many other communities in the area began a long process to recover lands they historically claimed as community patrimony, or sought to purchase lands from mestizo land owners looking to sell small estates of marginal or declining economic viability. These changes, in addition to pressures discussed below in chapter 7, triggered increasing migration from the colonial population centres on the Rio Pampas to new settlements in the Rio Cachi zone of the district. However, unclear or overlapping land claims, particularly in the Sunni and Puna grazing areas where boundaries between community holdings many never have been clearly demarcated, along with the legacy of interference and resource capture from small haciendas, set the stage for serious inter-community clashes at various points in the 20th century, particularly in 1961 and the late 1970s. Elite capture of precious cultivated land in the Rio Pampas zone probably exacerbated these tensions in the twentieth century, until the land reform pressures in the 1970s pushed the last abusive elites out of the district.

When conflicts were settled in the late 1970s and early 1980s over the Rio Cachi zone lands in the district and over the Quicamachay lands, more than one hundred years of efforts to regain control of seized lands in the Rio Cachi basin by communities in the D. Chuschi largely came to an end. While all the communities in the district had sought to gain control of Sunni and Puna lands in this zone, Quispillaccta was clearly the beneficiary of these struggles, significantly expanding its presence in the upper Rio Cachi basin from the late 1800s to the 1980s. Much of the land Quispillacctinos considered as their historical patrimony was regained, though holdings from the Hacienda Ingahuasi and the Hacienda Allpachaca remained beyond their control. Chuschi, by contrast, remained frustrated in their attempts to seize contested lands in the north- west contested with hacendados from the Yaruca estate. The desire for further expansion in the Sunni zones of the Rio Cachi watershed was, by the latter half of the twentieth century, increasingly clashing with the reality of other communities and former hacienda feudatarios living in these areas. There is little doubt that peasants living in the area hungered to establish control over more land for their communities, as the continued border disputes in the early 1980s indicated. Any opportunity for further expansion of community land-holdings was seized. The reportedly 2000 peasants who, at Sendero’s urging, invaded the University of Huamanga’s experimental agricultural station at Allpachaca in August 1982 to plant crops and seize

218 livestock718 were likely just as motivated by the opportunity to seize control of additional grazing and cultivated lands from the former hacienda, as they were by Sendero’s ideological exhortations to collectivist struggle.

However, the process of reclaiming lands lost to Spanish and Mestizo usurpers was not beneficial to every group involved. There were clearly winners and losers on both sides in the long historical struggle to regain high-altitude grazing lands. Communities members in Chuschi, Cancha Cancha, Putaje, and Ccochapampa lost out to Quispillaccta in the contests described above. When efforts to gain control of these lands were abandoned or settled in the late 1970s and early 1980s, their claims were left unfulfilled and their aspirations unrealized. Added to that were grievances born from decades of damage done to peasant livelihoods from the destruction of property during back and forth community clashes. The arrival of Sendero and the Peruvian military in the early 1980s temporarily froze the enmity among some over their decades-long efforts to gain control of the disputed lands, until it spilled out in a new direction during the counter-insurgency campaign.. Similarly, there were also losers on Quispillaccta’s side who may have been angered by settlements reached with Chuschi and its annexes. Some barrio residents in Catalinayuq and Puncupata smarted from the loss of parts of Pallcca, Arapa, and Tapacocha, as noted above. As well, other Quispillacctinos from barrios near disputed zones may have harboured resentment for the casualties, damage, and losses suffered in the land invasions and conflicts during the 1960s and 1970s. La Serna, for example, notes how the sons of Martin Mendieta, who was one of the three Quispillacctinos killed in the May 1960 battle at Kimsa Cruz, rustled cattle from Chuschinos in revenge for their father’s killing. This sparked deadly revenge against one of their associates by Chuschinos during Sendero’s initial killing of thieves and moral degenerates in the district.719

In the end, resource and entitlement capture in the district had consequences that echoed for centuries. These impacts were aggravated by supply and demand-induced resource pressures discussed in the next chapter. Resolving the impacts of resource and entitlement capture was a messy process in the D. Chuschi that generated decades of conflict between communities in the

718 Isbell, To Defend, 85. According to Isbell’s interviews, this event is remembered fondly by many peasants, who know doubt welcomed what the return of these former hacienda lands to peasant control. 719 La Serna, The Corner of the Living, 160-1.

219 area for several decades in the 20th century, up to the early years of the Sendero uprising. Changing local, regional, and national contexts significantly complicated efforts to recover lands that communities believed were rightly theirs. The causal influences extended beyond the capture of valuable lands and the riches that flowed from them, as is discussed in chapter 8. The land conflict helped to condition patterns of violence during the bloody months of 1983 when Peru’s Army finally confronted Sendero in the district.

220

Chapter 7 Supply and Demand Pressures in the District of Chuschi Introduction

The struggles in the latter half of the twentieth century to regain control of high-altitude grazing lands in the D. Chuschi described in the preceding chapters stemmed from a variety of causes at the local, regional, and national level, as political, economic, and social changes swept the highlands for several decades. The analysis provided thus far, however, provides only a partial answer about why these contests between communities in the district became so heated and violent in the second half of the twentieth century. Gradual changes to peasant livelihood opportunities from at least the start of the twentieth century also began to increasingly challenge households in the district by mid-century. Supply and demand changes began to impact the natural assets held by many households, and increasingly constrained peasant livelihood returns from cultivated lands. Households adapted to these increasing strains on cultivated assets in various ways, including diversifying household activities through commodity production, intensifying production on cultivated lands, migration, and shifting to focus efforts on livestock husbandry in the higher zones of the district.

These adaptations did not happen in isolation to the local demand and supply pressures but also responded to regional and national policy trends as the Peruvian government introduced various development policies. In the context of local supply and demand pressures on natural assets in the first half of the twentieth century, livelihood adaptations that included shifting more household efforts to livestock husbandry in the high puna zones of the district made good sense to many households. Livelihood adaptations to pressures on natural assets thus increased the political stakes for communities to establish control over contested Puna grazing areas described in preceding chapters.

This chapter begins by outlining demand-induced pressures on natural assets, pressures that stemmed in part from local population growth in the first half of the twentieth century, which reduced cultivated land availability and increased the intensity of household exploitation of local natural assets like land. Evidence suggests that these pressures increasingly stressed household cultivated crop returns in the twentieth century. Impacts did not happen in isolation, however. People adapted to these pressures, sometimes in ways that worsened impacts on local natural

221 assets, and in other ways that increasingly diversified household livelihood activities. This chapter also outlines the extent of these household adaptations, and their implications on local livelihoods. Although diversification activities by households enabled many to deal with the stresses from demand and supply pressures on natural assets, they were not without consequences as increasing internal migration to the high puna zones of the district heightened tensions between communities over access and control to increasingly important grazing areas of the district.

Demand-induced pressures on livelihoods in the District of Chuschi

In January, 1967, a group of anthropologists and government experts from the Instituto Indigenista Peruano (IIP), part of the Ministry of Labour and Communities, traveled to the D. Chuschi to conduct an intensive survey of the district and its needs. Like many remote Andean regions, the D. Chuschi’s limited interaction with the Peruvian state up to that point meant that much was not known about livelihoods in the district. Over the course of the next month, their study provided an important snapshot of the livelihoods and cultural traditions of residents in the district. Anthropologist Billie Jean Isbell arrived in the district soon after and built on this work as part of her multi-year study of the peasant community of Chuschi. Taken together, these efforts provide a rich scholarly foundation to study the state of livelihoods in the D. Chuschi, compared to other nearby districts.

As snapshots of livelihoods in the district in the mid-1960s, however, these studies do not provide a long-term perspective to understand how trends in key livelihood assets and capabilities changed over time. In particular, snapshot studies in the 1960s found the district well into several decades of strong population growth, with attendant impacts on the availability of some natural assets like cultivated land and associated impacts on livelihood adaptations like migration. By reconstructing demographic trends in the district over many decades we can begin to appreciate the impact and adaptation to demand-induced pressures on natural assets in the district.

The extent of the dramatic increases in the D. Chuschi’s population in the 20th century is only evident in examining long-term population trends. According to Isbell, the main town of Chuschi was experiencing several decades of population decline in the late 1960s, even while the

222 wider district population continued to increase over the same period. Chuschi’s population decline was attributed to out-migration to Lima and other areas where education and economic opportunities were increasingly available.720 Isbell’s focus on the declining population of Chuschi’s main town, however, ignored equally important long-term trends in the district showing a rapid increase in the district’s population in the early 20th Century and strong growth in the second half of the 20th century. While all of Ayacucho experienced rapid population increases in the late 1800s, according to demographic reconstructions, Cangallo’s demographic take off began somewhat later and was steeper in the early years of the 20th century. By the mid- 1960s Cangallo’s population growth was slowing; however, this was capping off a tripling of the province’s population in less than one hundred years. (See chart, Figure 21: Population of Department of Ayacucho and Province of Cangallo, 1791-2007. Unless noted on chart, data and

720 Isbell, To Defend, 68.

223 sources for charts in this chapter can be found in Appendix 2: Population of Peru, Ayacucho, Province of Cangallo, and D. Chuschi)

Figure 20:

224

Figure 21:

Population trends in the D. Chuschi similarly showed strong increases in the early decades of the 20th century, though the D. Chuschi’s steep take-off in growth was somewhat later than the wider province and department.721

721 District population figures are somewhat uncertain before 1940, and should be taken with some caution. The steep increase noted in the early 20th century could be a function of the available data.

225

Figure 22:

(See graph, Figure 22: Population of Province of Cangallo and D. Chuschi, 1876-2007.) The D. Chuschi’s population trajectory also differed from provincial trends in the latter half of the twentieth century, with Cangallo showing declining rates of growth in the 1960s and then absolute population declines in the 1970s. The rate of increase in the district’s population did slow somewhat in the 1960s when researchers first entered the area. However, population growth rates took off again in the 1970s and continued for the next three decades, with total population in the district growing by almost 25% by the mid-1990s. (See chart, Figure 23: Population of Province of Cangallo and D. Chuschi, 1940-2007.) From 1940 to 1980, on the eve of Sendero’s insurgency in the area, population growth in the D. Chuschi had increased over 20%, in spite of out-migration from the area over this same period (out-migration from the district is discussed in more detail below in section 7.4.1) and in spite of over a decade of declining population in the rest of the province. The data thus indicate that population trends in the district differed markedly from wider trends in Cangallo.

226

Figure 23:

As well, in contrast to Isbell’s observations about declining population trends in Chuschi’s main town in the late 1960s, population growth continued in areas outside of the main town in the district over the latter half of the twentieth century. The district’s population at mid-century was overwhelmingly located in areas outside of the town centre, according to researchers.722 Total population continued to grow until early 1990s, in spite of the upheaval in the decades before and during Sendero's uprising.

The data also appear to indicate that Quispillaccta’s population growth outpaced the growth of the community of Chuschi and its annexes in the latter half of the twentieth century. Population data below the district level is difficult to obtain because Sendero burned the municipal archives

722 Ramón C. et al., La Comunidad de Chuschi, 18. Some families also shifted their residences depending on the time of the agricultural season, particularly in Quispillaccta, where families appeared to have stronger ties to barrios. Thus, the effort by census takers to tie residents to one town or barrio was problematic because it did not consider the shifting household residency as agricultural demands caused families to move around the district. Isbell’s general observation about declining residency in the town of Chuschi due to migration was thus somewhat questionable without more detailed observations of people’s livelihoods.

227 in Chuschi in the early 1980s, destroying birth and death records for the community. Limited birth and death records are available for Quispillaccta from their community Public Registry archives. This data has been supplemented with baptismal records for both communities held by Chuschi’s Catholic Church in order to reconstruct population patterns from the mid-1800s to the start of Sendero’s insurgency. Baptisms were very important to the Catholic faithful, and families were keen to baptise children soon after birth for fear that ever-present illnesses would take the lives of their children before they could attain baptismal salvation. Baptismal records can thus serve as a proxy for birth records in the district for Quispillaccta and Chuschi, with some caveats.723 Their value declines after the late 1950s for several reasons. First, Evangelical Christian sects began to penetrate the district in the early 1960s and families turned away from the Catholic Church, particularly in Quispillaccta. The reluctance of Quispillacctinos to baptise their children in Chuschi’s church, a function of the enmity between communities, also somewhat undermines the utility of this data for Quispillaccta, particularly when tensions between the two communities heated up in 1960. These conflicts tended to discourage Quispillacctinos from dealing with Chuschi and its Catholic Church.724 Finally, a number of conflicts between comuneros and Catholic priests in Chuschi in the late 1950s and 1960s may have further discouraged community members from baptising children in the church.725 The baptismal record is thus most useful in noting the rise in births in the 1930s, 1940s, and 1950s. When baptismal records are combined with existing birth records to reconstruct sub-district population trends, it is clear that communities in the district were experiencing strong population growth from the early 20th century, and that Quispillaccta’s growth began out-pacing growth by

723 Jacobsen makes a similar point in his study of the province of Azángaro, in the Department of Puno. Jacobsen, Mirages, 26. In counting baptisms in Quispillaccta and Chuschi, I have grouped together the main town and affiliated annexes for each community from the district, using the cultural and kin-group affiliations described above for Quispillaccta and Chuschi. Because latter conflicts over land in the puna zone saw each side face off with affiliated annexes as allies and integral parts of their community self-identity, it makes sense to group together birth data in this way. Chuschi’s birth figures thus also include births that took place in Cancha Cancha, Uchuyri, and Chacolla. I excluded Sarhua and its annexes, though they were identified by informants as cultural allies of Chuschi, because Sarhua and its annexes separated from the District of Chuschi in 1910 with the creation of the province of Victor Fajardo. 724 Quispillaccta had its own Catholic Church, but the Parochial centre of the District was in Chuschi’s church. By the mid-twentieth century, Quispillaccta’s Catholic Church did not have a regular attending priest, unlike in Chuschi, so many families would have had to go to Chuschi’s church for baptisms. 725 These conflicts are discussed in Miguel La Serna, “In Plain View of the Catholic Faithful: Church-Peasant Conflict in the Peruvian Andes, 1963-1980,” Hispanic American Historical Review 95(4), 2015: 631-657.

228

Chuschi and its annexes after the 1950s. (See graph, Figure 24: Chuschi and Quispillaccta: Births and Baptisms, 1850-2000.) Observers at the time similarly noted these developments, with informants attributing Quispillaccta’s successes in the 1960s Kimsa Cruz conflict in part due to their greater numbers, as discussed above. As well, IIP researchers noted in the mid- 1960s that Quispillaccta was undergoing a major demographic explosion, with much higher growth compared to Chuschi.726

Demographic trends in the D. Chuschi thus demonstrate that population growth took off in the early decades of the 20th Century, with Quispillaccta’s growth apparently outpacing Chuschi and its annexes after mid-century. The district’s growth somewhat lagged growth in Cangallo, Ayacucho, and Peru generally, raising the prospect that the impacts of this growth in the district similarly lagged impacts in areas outside of the district.

Figure 24:

Birth data is taken from surviving Municipal registers, and baptism data is taken from records held in Chuschi’s Catholic Church.

726 Ramón C. et al., La Comunidad de Chuschi, 151.

229

The causes of population growth in the district, and in Peru more generally from the late-1800s, are a likely outcome of Peru’s progression through the epidemiological transition during this time. The epidemiological transition “describes changing patterns of population distributions in relation to changing patterns of mortality, fertility, life expectancy, and leading causes of death.”727 Epidemiological transition theory was first proposed by Omran in 1971, and argues that population dynamics during the transition moved from a period of high and fluctuating mortality rates and variable population levels, primarily as a result of the impacts of infectious diseases, to an “age of receding pandemics” where fewer people died because of diseases, seeing instead an increasing shift in mortality causes to “chronic” diseases like cancer.728 General mortality rates declined as pandemics declined, leading to increased average life expectancy and more sustained population growth with the transition.729 The specific reasons why infectious diseases declined during the transition are multifaceted, but scholars particularly emphasize our increasing understanding of germ theory in the late 1800s and the wave of public health and medical interventions that swept through developing countries during this period to provide cleaner water, basic sanitation services, and changes in medical practices.730

The epidemiological transition in Peru in the late 19th and early 20th century kick-started long- term population growth in the country, leading the country’s population to increase from 2 million in 1876 to more than 7 million by 1940. Up until this time, with few Peruvians living in urban centres, high rural fertility and low life expectancy kept growth slow.731 The country’s annual average growth rate between 1791 and 1850 was 0.8%; however, between 1876 and 1940 that growth rate had increased to 1.5%.732 While immigration played a small but important part in this growth, Contreras argues that the introduction of new Peruvian health policies were largely responsible for the country’s population growth in the early decades of the twentieth

727 Robert E. McKeown, “The Epidemiologic Transition: Changing Patterns of Mortality and Population Dynamics,” American Journal of Lifestyle Medicine; 3(1 Supplement), 1 July 2009: 19S. 728 McKeown, “The Epidemiologic Transition,” 20S. 729 McKeown, “The Epidemiologic Transition,” 20S. 730 McKeown, “The Epidemiologic Transition,” 20S. 731 Paul Gootenberg, "Population and Ethnicity in Early Republican Peru: Some Revisions," Latin American Research Review, 26(3), (1991): 129. 732 See Appendix 2: Population of Peru, Ayacucho, Province of Cangallo, and District of Chuschi.

230 century.733 Public health advances in the late 1800s were promoted by Peruvian elites as part of a broader civilizing and modernizing drive in the country.734 Mortality reduction campaigns began in 1870 with medical personnel from the Peruvian military being sent to each province. In 1887 the Peruvian government released its first health regulations, and in 1896 created a government health Directorate, changes spurred by various epidemics that swept the country in the 1870s and 1880s. The Directorate worked on areas of hygiene and demography and sought to reduce the mortality rate and increase population in Peru through public education and hygiene improvements.735 Vaccination efforts also began to take off in the country in the last decade of the 19th century and the first decades of the twentieth century, with vaccination officials travelling to provinces across the country, along with an expansion of health care workers nationwide.736 The Health Act of 1905 established mandatory vaccinations, inoculating 150,000 of Peru’s 3 million residents that year.737 Progress was slow in the early years of the twentieth century; higher in major urban centres, and slower in rural highland areas where the bulk of the highland population lived.738 However, by the 1920s, Peru’s annual rate of growth was around 2%, almost ensuring that the country’s population would double in 35 years. By the 1940 census, the national mortality rate had decreased to 27.1 per 1000 from 40-50 per 1000 in 1879.739

In the D. Chuschi, public health facilities were limited until the early 1970s. A rudimentary public health post was established in mid-1960s, staffed with a trained, bilingual health worker who dealt with minor illnesses, preformed home visits, and conducted vaccination campaigns.740

733 Carlos Contreras, Sobre los Orígenes de la Explosión Demográfica en el Perú: 1876-1940, Documento de Trabajo No. 6 (Lima: Instituto de Estudios Peruanos, 1994): 17. 734 Contreras, Sobre los Orígenes, 17-18. 735 Contreras, Sobre los Orígenes, 19. 736 Contreras, Sobre los Orígenes, 19-20. 737 Contreras, Sobre los Orígenes, 23. 738 According to Contreras, officials at the time reported that many highland families, fearing the vaccinations, appeared to have fled before government officials entered their communities, hampering public health efforts. Contreras, Sobre los Orígenes, 19-20. 739 Contreras, Sobre los Orígenes, 21. The official 1879 figure is 32.5 per 1000, but Contreras argues that this was probably underestimated by at least 10, and probably even higher in remote rural areas. 740 Ramón C. et al., La Comunidad de Chuschi, 119; Fanny Bolívar de Colchado, “Actitud de los Campesinos Frente al Personal Auxiliar de Salud,” in Instituto Indigenista Peruano, Dos Estudios en la Zona de Cangallo

231

In fact, the IIP researchers in 1967 urgently recommended the establishment of a health clinic and a public health education campaign in the district, noting that district inhabitants were still in the habit of depositing human excrement in the streets.741 While historical data on public health interventions in the D. Chuschi before the 1960s are sparse, it is likely that these limited public health interventions helped spur population growth in the district during the twentieth century. One informant told me that practically every family lost one or more children due to illnesses, especially in winter. But this slowed when the health post appeared.742 Mortality declines during the epidemiological transition spurred population growth widely in Peru in the early decades of the twentieth century743, but the impacts appeared to take a little more time to filter to remote regions like the D. Chuschi. Strong population growth by the 1940s was evident in the district and sustained for several decades, with attendant impacts on household livelihoods discussed below in section 7.3.

Demand-induced impacts on livelihood resources in the District of Chuschi

Population growth does not automatically lead to stress on peasant livelihoods. This conclusion from reviews of theoretical and empirical literature discussed above is affirmed when the implications of population growth in the early twentieth century in the D. Chuschi are examined. Mitchell is correct in noting the distinction between population growth and population pressure, where population pressure refers “to the entire relationship between population and resource production.”744 A demand-side focus on population growth in relation to some ecological threshold or attempts to draw conclusions about livelihood impacts from growth rates or per capita resource availability would be simplistic unless contextualized within the range of livelihood assets and activities of households over time. In order to understand the impact of population growth on community members in the D. Chuschi, we need to examine how this growth impacted key livelihood assets and opportunities for community members as the

(Ayacucho, 1968): 24. A doctor might only be present in the district during vaccination campaigns, according to the IIP study. Isbell notes that a health centre was being built in the district in the 1970s. Isbell, To Defend, 220. 741 Ramón C. et al., La Comunidad de Chuschi, 152. 742 Interview with José Espinoza Flores, Unión Portrero, 31 October, 2004. 743 Contreras, Sobre los Orígenes, 28. 744 Mitchell, Peasants, 19.

232 population pressure increased over the twentieth century, and the livelihood adaptions of households. As discussed above, various livelihood adaptations and institutional changes can ameliorate the impacts of population pressure. Andean smallholders are skilled at adapting to their circumstances, within the limits available to them in terms of assets, capabilities, entitlements, and structural constraints.

Untangling impacts and adaptations at the household level is extremely complex, however, particularly in the context of incomplete historical data, as is evident with this study’s focus on the D. Chuschi. Historical generalizations about impacts and adaptations for entire communities or groups in the district in this case cannot offer certain conclusions, only suggestions of patterns and consequences. In the D. Chuschi, cultivated land and animal husbandry was probably sufficient for most livelihood needs before rapid population growth in the twentieth century, in the context of a diminished post-Conquest population in the area. As growth accelerated in the twentieth century, the always limited resource base of cultivated land was spread more thinly among district residents and worked more intensively in ways that sometimes undermined the quality of the resource, leading to declines in some crop yields, and increasing degradation of some cultivated land. These trends show up in archival and fieldwork data in reduced per capita cultivated land over time in households, reduced yields per parcel, declines in the quality of yields, along with evidence of erosion problems in the heavily cultivated areas of the district near the Rio Pampas.

These impacts did not necessarily lead to household crisis for many in the district, however. Negative impacts were never preordained because some households could still apply their indigenous knowledge to work the land more intensively to maintain decent yields. For others, the impacts instead stimulated various livelihood adaptations in household economic and social activities that enabled many community members to continue to survive and thrive with a diminished stock of cultivated land assets – adaptations like permanent or temporary migration, shifting labour to other economic activities like livestock husbandry and/or commodity production, and increasing demands for the elimination of social-political drains on peasant livelihoods - like the impacts from abusive elites and religious authorities and the abandonment of cultural practices such as Catholic fiestas that drained precious resources. Parallel development trends outside of the district, such as increasing opportunities for wage labour in urban areas, in export plantations on the coast, or in the tropical lowlands to the east, enabled

233 particular kinds of adaptive responses. However, these livelihood opportunities also worsened stresses on natural assets for some community members because they pulled cultivators away from the district at times, contributing to labour reductions in production zone activities in the district that reduced investments in maintaining productive assets like cultivated lands. Some of these adaptions thus helped to worsen supply constraints for some community members. Many households also chose to adapt by increasingly moving to the high Sunni and Puna grazing zones to settle, to take advantage of new and expanding livelihood opportunities in that area. These demand and supply-side pressures thus contributed to inter-community conflicts over control of these areas discussed above.

7.3.1. Perspectives on the peasant household

A remarkable transformation has taken place over forty years on views about the livelihood sustainability of the Andean peasantry, from the time when outside experts first began to study highland communities in rural Peru in the late 1950s. At that time, the Peruvian government’s regional development plan for the southern Sierra described a “bleak, denuded region” crowded with unproductive subsistence farmers working tiny plots in tradition-bound ways that offered little hope for improvements in development or welfare levels.745 The report attributed the peasantry’s condition in the southern Sierra to the fact that many farmers were confronting a scarcity of good land – partly a result of severe inequality of land holdings - and using backward agricultural approaches oriented toward subsistence production, rather than producing for a market economy.746 This neo-liberal diagnosis of the peasantry’s woes, buttressed with a strong dose of Neo-Malthusian analysis, has been transformed by the work of dozens of anthropologists, economists, and agricultural specialists in the ensuing decades.

Instead, Andean peasants are now seen to have “skillfully fashioned” adaptations to their highly diverse Andean environment over thousands of years, allowing them to take advantage of different crop species and ecological zones in ways that minimize or balance risk and maximize the livelihood opportunities for highland communities.747 Farmers use crop and species diversity

745 Watters, Poverty, 40-1. 746 Watters, Poverty, 40-1. 747 Watters, Poverty, 34; Brush, Mountain, 18; Mayer, The Articulated Peasant, 25-6. See also, Pierre Morlon, “Peasant strategies to deal with risk,” LEISA Magazine 7(1&2), May 1991: 7-8; Earls, “The Character of Inca and

234 to take advantage of the diverse micro-climates and ecological zones of Andean regions, where every 100m increase in altitude is like a ten degree difference in latitude.748 Animal husbandry is also widely employed by peasants, a low energy activity with returns that can result in higher energy exchanges for households.749 Combined agricultural and livestock activities also “make the best possible use of available labor throughout the year” for households.750 Rather than living in subsistence isolation, peasant households are now seen to have survived and persisted because they are “articulated with other households, their communities, and the commodity markets” to varying degrees, thereby maximizing diverse livelihood opportunities.751 Before access to capitalist markets was common, households connected and traded through barter networks, economic relationships that continued as market transactions grew more common in the twentieth century.752 By the early 1970s, most peasant households were already deeply integrated into the market economy, analysts have now concluded, a process that accelerated from the early 1950s and continued into the 1980s.753 As Caballero summarized:

…the dominant opinion among economists and economic anthropologists who have concerned themselves with the Andean peasantry is that … their economy is efficient. Aspects that at first sight may appear to show inefficiency, such as widespread fragmentation of land, are found on closer inspection to be consistent with a strategy that makes best possible use of microecological variations, tends to diversify risks, and plans activities in such a way that labor requirements are not concentrated in certain weeks. The technology is better adapted to the local milieu than is usually assumed … Recent research has shown the importance of seasonality and risk … it is the articulating element of much of the peasant economic and social life, especially the sequence of activities and temporary migrations. Equally fundamental are the attempts to reduce risk in order to

Andean Agriculture,” Unpublished paper, Departamento de Ciencias Sociales, Pontificia Universidad Católica del Perú, 1998: http://macareo.pucp.edu.pe/~jearls/documentosPDF/theCharacter.PDF; and David L. Browman, “Agro- Pastoral Risk Management in the Central Andes,” Research in Economic Agriculture 8, 1987: 171-200. 748 Mayer, The Articulated Peasant, 15-16. 749 Mayer, The Articulated Peasant, 17-8. 750 Jose Maria Caballero, “Agriculture and the Peasantry Under Industrialization Pressures: Lessons from the Peruvian Experience,” Latin American Research Review 19(2), 1984: 26. 751 Mayer, The Articulated Peasant, xiii. 752 See Mayer’s chapter on barter, Mayer, The Articulated Peasant, 143-171. 753 Caballero, “Agriculture,” 20-1.

235

ensure survival. Fragmentation of plots, especially at different ecological niches, diversification of activities, and even certain types of resistance to the usage of modern inputs, can, and probably should be, interpreted as a way in which peasants confront risks.754

The conclusions above from Caballero, Mayer, and other Andean scholars thoroughly explain complex adaptations that make up the “mode of livelihood” of small-scale farmers in the Andes today.755 Their work also accurately describes how community members in the D. Chuschi made a living on the eve of Sendero’s uprising. However, for our purposes, as a guide to how the transformation of Andean livelihoods evolved to reach the state it was in before Sendero’s uprising, their analysis is incomplete. We must resist the temptation to see the state of the peasantry on the eve of Sendero’s uprising as static – as if that was the way in which peasant livelihoods were always organized. In fact, the degree of fragmentation of household land parcels, the degree of diversification of livelihood strategies, and the vulnerability of households to market or climatic vulnerability, have not always been as analysts above described them (and to be fair, analysts of Andean livelihoods would acknowledge this truth). Outlining the long- term trends that have led Andean households to the economic lifestyle that defines their current survival strategies means outlining the pressures that have molded various adaptive strategies among households – pressures such as those that led to the increasing fragmentation of cultivated land over time, pressures that have undermined the quality of some cultivated lands, pressures that have shifted cultivation or animal husbandry toward market commodity production, etc.

7.3.2. Trends in declining land availability

Land is a key resource for subsistence and semi-substance agriculturalists - one of the key elements determining the economic lifestyle of peasants according to Caballero, along with their social organization, need to satisfy household necessities, and relation to the market.756 “In the Andes” Caballero writes, “nature is decidedly heterogeneous with respect to climate and soils, with pronounced microclimatic variations, market seasonality, and diverse risks (drought, frost,

754 Caballero, “Agriculture,” 24-5. 755 This is Mayer’s expression of the question analysts have been trying to answer for the past 50 years. Mayer, The Articulated Peasant, xiv. 756 Caballero, “Agriculture,” 24.

236 landslides). The ecology is “vertical,” that is, mountainous, and land is scarce and of poor quality.”757 Most analysts of Andean peasant livelihoods take for granted that farmers have had to provide for their households in an environment where sufficient, good quality agricultural land for crop cultivation is scarce.758 But while this may describe the late 20th century reality for many Andean peasants, in some areas cultivated lands may have remained sufficient for many until the early decades of the twentieth century when Peru’s population recovered to pre- Conquest levels, particularly in remote regions where market penetration and elite resource capture of peasant lands for market production was less evident, like the western Rio Pampas zone. While high fertility and high mortality characterized Peru’s demographic profile before the impacts of the late 19th century epidemiological transition, Gootenberg notes that high fertility “was abetted by a relative abundance of economic resources, chiefly accessible land” after the colonial demographic crash, thus providing ample food supply for the population growth of the late 19th and early 20th century.759 Adequate endowments of land sufficient for subsistence needs were likely a reality for many peasants at the start of the 20th century, even in the context of widespread resource capture by Mestizo elites which severely impoverished other peasants.760

757 Caballero, “Agriculture,” 24. 758 Interview, Alberto Pasco Font, Grupo de Análisis para el Desarrollo, GRADE, Lima, Perú, 24 August, 1999. 759 Gootenberg, “Population and Ethnicity,” 130. Ample supplies of land as a result of the devastation of Conquest may have long been a reality in the highlands. Mayer quotes a villager from Tancor, Huánuco – in the Central Andes – speaking to Spanish tribute inspection officials in 1562, explaining why he farms in his wife’s village of Tancor and not in his home village of Chacapampa: “Not that there is no land in Chacapampa; with so many people dying off there is plenty of land all over… There is land everywhere because our people are diminishing very quickly, and because the Inca is gone and does not force us to produce food here in the villages for his tambos and storehouses.” The villager’s account makes clear, however, that colonial tribute obligations were putting crushing obligations on his livelihood at the time. Mayer, The Articulated Peasant, 81. 760 I do not want to downplay the impact elite resource capture of peasant lands in inflicting livelihood hardship on peasant communities. The tremendous diversity of experiences in Peru makes generalizations about the scarcity or availability of land at any given time difficult. The key point is that population crash after conquest did loosen constraints on land resources for many, though elite capture of many of these same lands for market production also impoverished many as well. Each local experience must be carefully evaluated before conclusions are drawn. At a minimum, there appears to be evidence to support the conclusion that in some places conquest loosened the availability of land for some, even in the face of resource capture by Mestizo elites, trends that would change again as communities regained usurped lands, populations grew, and new market realities developed in the 19th and 20th century. Jacobsen’s careful analysis of the province of Azángaro, in the Department of Puno, similarly describes a situation where peasants found themselves in an era of abundant land with low population densities. In Azángaro, Jacobsen notes, peasants probably controlled more than 50% of the agriculturally useful land in the late 18th century – the late colonial period. In spite of this situation, however, Jacobsen notes that government officials at the time spoke of the severe problems of precarious peasant livelihoods and land scarcity. Jacobsen, Mirages, 95. Officials attributed the hardship of the herding communities in the area to the “stagnation” of their herds, a result of the “infertility of the altiplano’s soil and [officials] assumed that the pastures simply could not support more animals.” Jacobsen, Mirages, 95-6. In reality, Jacobsen’s study concludes, peasant livelihoods were at risk in Azangaro

237

In the D. Chuschi, it appears that families began to feel the pinch of inadequate cultivated land endowments after mid- 20th century, compared to more abundant pasture lands in the higher zones of the district. This conclusion is supported both by interview evidence from elderly informants and through an examination of population and land-use data. About 27% of the district’s land is suitable for cultivated agriculture, according to a recent survey, with only about 9% of the total agricultural land making up the Quechua zone between 2400, and 3400m where maize production is possible. Agriculture in the 16% of agricultural land not in the maize zone is common but risky, mostly composed of rain-fed tuber and grain production.761 The Spanish reducción located the town centres of Chuschi and Quispillaccta around 3000m, in the middle of the Quechua and Sunni zones.762

Interviews with older residents in Quispillaccta and Chuschi recall a time in their youth when land was more abundant and parcel sizes were larger in the district. Inheritance practices of splitting a household’s land among the offspring, however, have steadily eroded land availability for households. Like many places in the Andes, parallel inheritance usually is practiced in the district, with mothers passing land to daughters and fathers to sons.763 With a limited supply of

because elites imposed severe labour and tribute obligations on peasants in addition to seizing the best lands for their own estates, obligations that made it extremely difficult for peasants to secure sustainable livelihoods. Land and legal reforms in the early 1800s described above opened the door for peasant communities to reclaim seized lands in the decades to come, no doubt easing livelihood stresses for some; however, the wool export boom of the late 1800s and early 20th century forestalled benefits for many others. Jacobsen, Mirages, 125. In areas of marginal or declining highland estates, like around the D. Chuschi, as communities recovered lands this eased pressures, particularly at a time of low population densities. But as communities grew in size in the first decades of the 20th century, livelihood pressures on increasingly less abundant cultivated lands no doubt increased. 761 See Chart, District of Chuschi Land-Use According to Agro-Ecological Zone, in chapter 3. As explained above, calculations for total cultivated land in the district are probably higher today than in the early 20th century because of climate change. 762 Isbell cites Chuschi’s altitude as 3154m; however, topographical maps show the main square of each community at just below 3000m. 763 Isbell, To Defend, 38. When inquiring about land holdings I did not specifically ask informants to detail land holdings by each parent. Specific questions about the amount of land held by family members is a fairly sensitive subject, given the fear that the information might be used against households by the government. Mitchell, Peasants, 55, found a similar situation when inquiring about land holdings for his study of Quinua, Ayacucho. So, informant statements about land holding are uncertain. In many cases, male heads of households explicitly spoke of land holdings and historical land holdings in terms of the male lineage. When informants used this language, I repeated the information in the text below. If informants spoke of family holdings, I repeat this information in the text.

238 land and multiple children in each household, inheritance practices can result in processes creating ever smaller parcelization of landholdings for households – or, minifundismo, as it is referred to in Peru.764 The 1967 IIP study of the district found that this process was driving the increasing division of household land holdings, and that requests for land from land-poor households had exhausted available communal holdings in the communities.765 IIP’s researchers assumed five-member families in the D. Chuschi, which would suggest between 3-5 children per family.766 However, census surveys in the 1970s found that the total fertility rate in Ayacucho was 7.4 in 1975, suggesting that families of five or more were common before the 1970s.767 Juan Galindo, 80 years old when interviewed, said that when he was young “the harvest was good enough that when they opened a new chacra (field), the grandfathers would say that this was enough because they couldn’t carry all the harvest.”768 Fermín Qallocunto Tomaylla of Quispillaccta, who was also 80 when interviewed, similarly noted that when he was young they would never sow potatoes in a 1 yugada field (around 1/7thH), only smaller chacras because these were sufficient.769 Others noted that some families did have large maize fields many years ago. “In the old days families would have chacras of one or two yugadas in size,” Julian Vilca Mejía told me, referring to when he was a young man in the 1940s and 1950s. “But population growth and inheritance have decreased the size of the fields. Today only a small number of people have chacras that size.”770 Chuschino elders also confirmed in an interview that declining parcel size was a problem that was leading people to struggle to make a living. Almost no one

764 Watters, Poverty, 88. “Minifundism is a system of agriculture that depends on a number of tiny, fragmented pieces of land that are worked primarily for subsistence. As it exists in Chilca [Cuzco], it is primarily a consequence of land hunger, which in turn is a result of the extensive ancient exploitation of land by surrounding haciendas and the rapid rate of population increase.” 765 Ramón C. et al., La Comunidad de Chuschi, 27. IIP’s researchers quoted the district’s governor as saying that, “the community’s communal lands had been distributed for years to community members who asked for land, and now there was no more land to distribute.” 766 Ramón C. et al., La Comunidad de Chuschi, 18. 767 Pablo Sanchez Zevallos, La Sierra Peruana: Realidad Poblacional (Lima: AMIDEP, 1988), 61. The national fertility rate in Peru was 5.6. By 1987, Ayacucho’s total fertility rate had fallen to 5.8, compared to a national rate of 4.5. 768 Interview, Juan Galindo, Quispillaccta, 11 October, 2004. 769 Interview, Fermín Qallocunto Tomaylla, Quispillaccta, 31 October, 2004. Yugada size is discussed below. 770 Interview, Julian Vilca Mejía, 8 October, 2004. A yugada is a traditional Andean measurement unit, usually referring to the amount of land that can be plowed in one day. While this obviously varies from parcel to parcel, Andean researchers in the 1960s estimated one yugada to be approximately 2500m2, or about ¼ hectare. There are strong reasons to suggest that this is an over-estimation, however, as I discuss below.

239 has 1H of land anymore, they told me. The barrios are fighting between each other today, they said, because there isn’t enough land.”771 “Splitting land among offspring has really hurt the amount of land in each family and now there’s not enough for families, especially if a family wants to pass on land to their kids,” 73 year-old Marcello Tomaya told me. Instead, he encouraged his kids to get an education to be able to get other jobs.772 Nemesio Nuñez’s father had lots of land in various parcels all over the community, he told me in an interview. But when it was divided among his 6 children, there wasn’t much left for each child, he said. Nuñez’s children don’t feel that they have enough land to live, so many have gone to the cities.773 Gregory Conde Nuñez, of the Quispillaccta barrio of Pirhuamarca, agrees that inheritance practices have reduced land availability, as shown by his family. His father had about 4H of land, but with six sons, he has been left with 3 yugadas of land. He’s now looking to buy land.774 Chuschino informants similarly described how inheritance reduced their family’s landholdings. Eighty-year old Victor Esteban Caillahua Mikuylla of Chuschi related that his grandfather’s 2-3 yugadas of land was sufficient for their family. Inheritance reduced his father’s holdings to 1.5 yugadas, but with two sons, Don Victor himself inherited only 1 yugada.775 By contrast, Chuschino Leon Huaman Conde noted that he has 4 yugadas of land, like his grandfather, because both his grandfather and father only had one son.776 Similarly, 75-year old Quispillacctino Roberto Huaman Machaca explained that he has 3 yugadas of land, like his father, who had only one son.777 Clearly, many community members believe that inheritance practices that divided land among offspring reduced land availability for households in the district during the twentieth century, in the context of a rising population in the district.

771 Interview with representatives of Chuschino Ancianos group, Chuschi, 15 October, 2004. 772 Interview, Marcello Tomaya, Quispillaccta, 9 October, 2004. 773 Interview, Nemesio Nuñez, Unión Portrero, Quispillaccta, 18 October, 2004. 774 Interview, Gregory Conde Nuñez, Quispillaccta, 20 October, 2004. Conde Nuñez was 45 when interviewed, so his father’s 4H of land would have certainly put the family among the larger land-holders in Quispillaccta in the second half of the twentieth century. However, as his family history shows, moving from land abundance to limited land resources could happen very quickly if a family had many offspring. 775 Interview, Victor Esteban Caillahua Mikuylla, Chuschi, 24 October, 2004. 776 Interview, Leon Huaman Conde, Chuschi, 31 October, 2004. Huaman Conde’s name betrays that he had relatives from Quispillaccta, like several of the Chuschinos interviewed for this study. In fact, my first research assistant, Quispillactino Hereberto Nuñez, was surprised during our interviews to come across so many Chuschinos with Quispillactino relatives. 777 Interview, Roberto Huaman Machaca, Quispillaccta, 1 November, 2004.

240

Detailed land-holding data does not exist for the D. Chuschi before the 1960s. However, piecing together the available data appears to show that district residents had access to ever-smaller amounts of cultivated land throughout the second half of the twentieth century, with per capita averages of around 1H per household likely common by the early 1970s. Calculations of population over time compared to

Figure 25: Cultivated Land (H) per Capita in the District of Chuschi 7 6 5 4 3 2 1 0

Maize land (H)/cap Tuber-Grain Land (H)/Cap Log. (Maize land (H)/cap) Log. (Tuber-Grain Land (H)/Cap)

available cultivated land show marked per capita declines in cultivated land in the early half of the twentieth century.778 (See Figure 25: Cultivated Land (H) Per Capita in the D. Chuschi.)

Survey data suggest that cultivated land holdings were even smaller than indicated above. In 1967, IIP researchers found that Chuschino families had between 6 to 8 yugadas of land per family, or about 1.58H of land using their calculations of 2250m2 per yugada. Holdings in Quispillaccta (Matriz/town centre) were smaller, with each family possessing 4 to 6 yugadas of land, or about 1.12H.779 Annexes to Chuschi and Quispillaccta, however, had smaller holdings. Chuschi’s annexes of Cancha Cancha, Chaqollo (Chacula on the topographical map), and

778 Data is drawn from ABA chart, District of Chuschi Land-Use According to Agro-Ecological Zone, in chapter 3, and compared with population figures noted above. As discussed above, population data for the district before 1940 is less reliable than later census data. 779 Ramón C. et al., La Comunidad de Chuschi, 31-2.

241

Uchuiri (Uchuyri on topographical map) had 3 to 4 yugadas of cultivated land, while Kallcabamba (Callcabamba on the topographical map) had 4 to 5 yugadas of land.780 Annexes in Quispillaccta had even less land. Three barrios of Quispillaccta surveyed by IIP researchers, Unión Potrero, Cuchoquesera, and Llacctahurán had average landholdings between 1 to 2 yugadas.781 In fact, another IIP study of the province of Cangallo found that every population centre in the D. Chuschi, aside from Chuschi’s main town, had less than 1H of land on average in 1967.782 The 1972 Peruvian agricultural census, conducted only a few years after the IIP studies, found that, of those holding less than 1H of land, the majority of Cangallo’s residents, had on average about 4.7 parcels.783 This means that many residents of the district had several parcels of land, generally less than ¼ hectare.

When we carefully consider how IIP researchers came up with land holdings estimates, however, there are strong reasons to suggest that their calculations also over-estimate household landholdings in the district. The IIP data on land holdings is based on the supposition that there are approximately four to five yugadas per hectare in the D. Chuschi, or that one yugada is about 2250m2.784 This measure is used in all their community studies in the province of Cangallo in the mid-1960s. It is also close to the 2500m2 measure used in other communities in Ayacucho and in Peru’s 1972 agricultural census.785 However, closer examination of this yugada measure and comparisons with field-size measures by other scholars indicate that the size of 1 yugada is less than 2250 m2, and that the size of average land-holdings is also smaller. A yugada is defined

780 Bolivar de Colchado, Estudio Distrital de Chuschi, 17. 781 The first two are located in the Rio Cachi watershed, with only high-altitude cultivated lands but ready access to abundant pasture lands. Llacctahurán is located in the Rio Pampas zone, on the edge of the maize zone, so its lack of cultivated land per capita is more serious from a household livelihoods perspective. 782 Ramon Cordova et al., La Zona de Cangallo, 40. This somewhat contradicts the data in the IIP study by Saramiento Medina et al., noted above, for holdings for Quispillaccta’s Matriz and for Kallcabamba. However, in both cases, the average was very close to 1H. Land holdings for the district, aside from Chuschi’s main town, were actually fairly representative of Cangallo as a whole, according to the study by Ramon Cordova et al. They found that more than 83% of Cangallo’s population centres had average land holdings of less than 1H per family. Less than 17% had land holdings between 1 and 5H. 783 Maletta and Makhlouf, Perú, las provincias en cifras, 1876-1981. 784 Ramón C. et al., La Comunidad de Chuschi, 32. 785 Mitchell, Peasants, 55 & 224, footnote 224. Peru’s 2012 Agrarian Census does not list yugada as an indigenous measure for Ayacucho, but instead lists a yunta. Like a yugada, a yunta is measured at 2500m2. See INEI, Informe Final de Actividades Consistencia, IV Censo Nacional Agropecuario 2012 (Lima: INEI, 2013): 7.

242 as “the quantity of land that can be plowed in one day.”786 This type of measure goes by other names in other parts of Peru. In Cuzco, Watters notes that community members still use the colonial term topo, while in the province of Azángaro, Puno, the term masa is used by peasants to refer to “the amount of land that a work party of three men can plow, sow, or harvest in one day,” according to Jacobsen.787 Variation in yugada size is to be expected, depending on factors such as the slope, the quality of the land, and the efficiency of the plowing team. In Cuzco, Watters found a topo equivalent to about 3200m2, while Jacobsen found one masa be about 760m2.788

In Ayacucho, comparative evidence suggests that IIP researchers over-estimated yugada size in the 1960s, and thus over-estimated land-holdings of households in the D. Chuschi. To arrive at their calculation that 1 yugada equals 2250m2, IIP researchers surveyed various community members and mestizo elites, and were told that there were about 4-5 yugadas in a hectare. They also measured a parcel that one community member indicated was one yugada to be 2240m2.789 Other researchers, however, have found that a typical yugada is often much smaller than a quarter hectare. Mitchell measured multiple fields in the Ayacucho community of Quinua, and found an average of 0.1294 hectares per yugada, or about 1298.7m2 per yugada.790 Renzo Salvador Aroni Sulca, studying the community of Huamanquiquia, in Víctor Fajardo province, just south of Chuschi, notes that 1 yugada is equivalent to 1800m2.791 Alejandro Tumbay Maldonado, studying the community of Pomabamba, in the District of Maria Parado de Bellido, 30 minutes west of Chuschi, similarly found 1 yugada equivalent to 1800m2.792 Another study in the District of San Marcos y Chavín de Huantar, in the Department of Ancash measured 1

786 Mitchell, Peasants, 55. 787 Watters, Poverty, 347; Jacobsen, Mirages, 410, footnote 15. 788 Watters, Poverty, 347; Jacobsen, Mirages, 410, footnote 15. 789 Ramón C. et al., La Comunidad de Chuschi, 32. 790 Mitchell, Peasants, 55. 791 Renzo Salvador Aroni Sulca, Campesinado y violencia política en Víctor Fajardo (Ayacucho), 1980-1993, Unpublished Thesis, Universidad Nacional Mayor De San Marcos, Lima, 2009: 57. There is no explanation in this study for how the 1800m2 /yugada figure was calculated. 792 Alejandro Maldonado Tumbay, Zonificación y Priorización del Potencial de Recursos Naturales de la Comunidad Campesina de Pomabamba, Distrito María Parado de Bellido – Ayacucho, Unpublished Thesis, Facultad De Geología, Minas, Metalurgia y Ciencias, Universidad Nacional Mayor De San Marcos, Lima, 2002: n.p. There is no explanation in this study for how the 1800m2 /yugada figure was calculated.

243 yugada at about 1700m2.793 Fieldwork in Quispillaccta and Chuschi similarly found that 1 yugada was far less than the 2250m2 standard used by IIP. A field near Quispillaccta’s Matriz belonging to Antonio Huamani Galindo, a Varayoc Mejor – or traditional village authority – was measured to be approximately 2850m2. Don Antonio believed that this chakra, which had a slope between 6 and 12 degrees, was about 2 yugadas, meaning that 1 yugada is about 1425m2.794 If we use the conversion of 1 yugada=1425m2 to assess landholdings during the IIP 1967 study of the D. Chuschi, we find that households in Chuschi’s annexes had an average of less than ½H of cultivated land, while residents of Chuschi’s main town had about 1H on average. Residents of Quispillaccta’s Matriz had about 2/3H of cultivated land; however, its annexes had less than 1/3H. Quispillaccta’s households, particularly in the annexes, appeared to have very small holdings of cultivated lands by the late 1960s.

Adaptations and their impacts

7.4.1. Migration

Confronting increasingly limited endowments of cultivated land in the D. Chuschi, residents adapted by intensification and extensification of production in cultivated land zones, adopting new technologies when available, and diversifying household income activities – including through temporary or permanent migration. Some of these adaptations worsened the supply of available cultivated land, as is evidenced by erosion of soils and land slides in the Rio Pampas zone, declines in the quality of crop yields for some over time, and falling yields of some cultivated crops. These developments are discussed in more detail below in section 7.6.

793 Orlando Cajamarca Chávez, Proyecto Desarrollo Agropecuario En La Cuenca Del Mosna: “Proyecto Mosna,” CARE Perú, 2001: 14. The author found that households had an average of 1.5 yugadas each spread among 5.8 chacaras, or about 3420m2 per household. 794 Field measurement and interview with Antonio Huamani Galindo, Quispillaccta, 1 November, 2004. An even smaller unit of measure commonly used in Quispillaccta and Chuschi is the melga. Few farmers have parcels of 1 yugada in size in the lower cultivated zones, particularly in the latter half of the twentieth century. So, a smaller field measure makes sense. A melga was measured by Quispillactino and agricultural engineer Gualberto Machaca to be about 250m2. In my interviews I mostly used yugadas, though informants sometimes responded to questions by using melgas. Gualberto Machaca Mendieta, Conservación y Recuperación de Suelos en Tres Microcuencas del Rio Pampas, con Énfasis en la Comunidad Campesina de Quispillaccta (2675-4780 m.s.n.m.) Ayacucho, (Ingeniera Agrónomo Tesis, Universidad Nacional de San Cristóbal de Huamanga, 1997), 68.

244

As population growth and demand pressures on cultivated land in the district increased in the twentieth century, many households sought to diversify livelihood activities. Temporary or semi-permanent migration was one strategy that many households pursued. While migration was not the only out-of-district livelihood diversification strategy to be pursued, it was one of the most important strategies because it provided a vital source of cash.795 However, migration also worsened labour availability for some in the district, with negative consequences for land management. Migration trends from the district thus deserve discussion in this chapter.796 It is important to note, however, that district residents did not migrate solely because of supply and demand pressures; reasons for migration were varied and complex. Migration offered an opportunity for households, relative to constraining opportunities for cultivated agriculture in lower parts of the district – particularly from the end of WW2 to the late-1960s.

Peru was transformed by a massive rural to urban movement of people beginning in the late 1940s and lasting into the early 1970s, stimulated by population growth in in the mid-twentieth century and post-WW2 economic liberalism. The proportion of migrants in Peru’s population increased from 9.5% in 1940, to 23.3% in 1961, and to 26.4% in 1972, Klaren notes.797 Public works projects and export-driven economic policies put in place by the Odria military dictatorship were a magnet for attracting poor peasants from the Sierra to coastal urban centres like Lima.798 Between “1940 to 1961,” Klaren notes, “the number of inhabitants of metropolitan Lima exploded from almost 600,000 to nearly 2 million, an average of just over 5 percent a year. A large portion of these inhabitants were migrants from the countryside, mainly from the departments of Ancash, Ayacucho, and Junin, who settled in barriadas (squatter settlements)

795 Mitchell makes the same point about the community of Quinoa. Mitchell, Peasants, 111. 796 Space constraints do now allow me to examine other strategies of livelihood diversification that district residents pursued, such as weaving or other commodity production activities. Nor do I examine economic activities in the livestock sector in the district in great detail. Textile production and the production of livestock-derived products like cheese are important activities for some households. Along with animal production, these activities have only increased in importance in the latter 20th century. However, this study is particularly concerned with the transition over the twentieth century to put more household emphasis into these activities, the reasons why the transition took place, and the consequences for conflicts in the district. As a result, I do not explore the livestock and herding economy aspects in as much detail in this dissertation as the cultivated agriculture sector. Ramón C. et al., La Comunidad de Chuschi, 52-55, discusses commodity production and livestock derived products in the district in the mid-1960s. 797 Klaren, Peru, 301. 798 Klaren, Peru, 302.

245 that increasingly ringed the city on three sides.”799 These patterns of rural to urban migration began to slow in the 1970s, partly due to the Agrarian Reform and partly due to the economic stagnation and crisis that the country faced, particularly after 1975. Movements of migrants to major urban centres like Lima dropped somewhat in the 1970s, while smaller movements to small urban centres like department capitals or to newly opened lands in the Amazonian lowlands and the eastern slopes of the Andes increased.800

These general migration patterns, however, mask significant variability in both the duration of migration and the destination of migrants from the D. Chuschi. Informant interviews highlight that district residents have a long history of temporary, seasonal, or semi-permanent migration to urban areas like Lima, Huamanga, and Ica – stays that sometimes lasted the winter months of July and August, and sometimes lasted for years. Isbell says that the first Chuschino to migrate to Lima left in the late 1930s.801 Interviews suggest that many district residents may have journeyed to Lima and other urban areas before or around this time. A detailed history of migration in the district remains to be completed.802 More than half of the district residents interviewed during field work in 2004 had migrated at some point in their lives, with the vast majority of these trips being to Lima. This is consistent with historical data. A study in 1966 found that almost 70% of district residents who emigrated between January and August traveled to Lima.803 A few travelled to other highland departments to work as agricultural labourers, while a small number travelled to the Amazonian lowlands to work in plantations. A small number also worked in highland mines, while three Chuschinos spent several years in the Army

799 Klaren, Peru, 301. 800 Ann Brigid Ackelmire, Agrarian reform and internal migration in Peru, Unpublished M.A. Thesis, The University of Arizona, 1987, 64-5. 801 Isbell, To Defend, 180. 802 My informant evidence contradicts some of the existing scholarship on migration in the district. For example, 96-year-old Belisario Galindo Conde of Quispillaccta told me that he first went to Lima when he was 15, and seasonally for a few weeks or months for many years afterward. Interview, Belisario Galindo Conde, Quispillaccta, 29 October, 2004. In another example, 90-year old Chuschino Basilio Chipana Nuñez said that he was about 15 when he first went to Lima. Interview, Basilio Chipana Nuñez, Chuschi, 2 November, 2004. Obviously, these interviews were capturing residents who chose to come back to the district after their time away; many others left and never returned. So, this is, at best, a partial glimpse into the reasons and dynamics of out-migration from the district. 803 Bolivar de Colchado, Estudio Distrital de Chuschi, 16. Her data also confirms the informant data on other migrant destinations: about 15% of migrants travelled to Huamanga; and about 8% travelled to the mining centre of La Oroya and to the Amazonian low-lands (Selva Satipo).

246 around the time of the 1941 war with Ecuador.804 Informant evidence and studies by IIP in the mid-1960s indicate that migrants were mostly young district men or teens.805 More than half of those interviewed said that their migration was preceded by other family members. Recent studies suggest that poorer migrants in Peru will avoid migrating to large distant cities, where travel and expenses are beyond their means, and instead migrate closer to home, to nearby urban or rural areas.806 These insights probably say more about current migration trends out of the district than about migration in the boom period after WW2, when work was abundant and living expenses cheap for migrants in Lima. One young informant did tell me in 2004 that would like to migrate to Lima, but that he could not afford it because he has little money and there’s no work in Lima. He said that he’d rather have the little that he has in Quispillaccta than take the risk of migrating.807 Richer district members had much more flexibility about destinations, length of stays, and reasons for migration. Some rich district elites had long sent family members to urban areas like Huamanga for education, a trend that later spread among community members in the 1960s and 1970s.808 Among non-elites, about one third of informants were away for more than one year, in some cases almost two decades, before returning to the district. The vast majority of people interviewed, however, engaged in various forms of temporary or seasonal migration lasting between a few weeks to a few months.809

This temporary seasonal migration aligns most closely with rural livelihoods in the district; their temporary migration is partly an adaptation to supplement rural household activities in the

804 Two Chuschinos spoke of being press-ganged into the Army after being caught by authorities in Lima without proper documents. 805 Ramón C. et al., La Comunidad de Chuschi, 24. Temporary migration was usually men who left their wives and children in the district, according to the IIP study. If that temporary migration status changed, then other family members would also travel to urban areas. Other family members in the district would take care of land and livestock left behind for a portion of the harvest. However, IIP researchers did report young women migrating to work as household help or as fruit vendors. 806 Ackelmire, Agrarain, 70-1. 807 Anonymous interview, Quispillaccta, 12 October, 2004. 808 Ramón C. et al., La Comunidad de Chuschi, 24. Notable Chuschino Misti Ernesto Jaime told me that he was sent to Huamanga for schooling until his late teens, and then went into the Army at 23. His military service was a formidable experience that gave him the confidence to return to Chuschi and successfully challenge the abusive elites that ran the district and oppressed community members. Interview, Ernesto Jaime, Chuschi, 19 October, 2004. His military experience also helped him in the deadly role he played for Chuschinos in the 1960 Kimsa Cruz violence, according to Quispillactino informants. 809 Isbell also reported this type of migration was common. Isbell, To Defend, 182.

247 district. Estimating how many migrants left during this period is almost impossible due to data limitations. The 1966 IIP study noted above counted 393 district residents migrating out between January to August, evenly split between Quispillaccta and Chuschi.810 When compared with population estimates for the 1960s, these figures give a very rough estimate of 5-10% of each community’s population migrating temporarily or permanently in the 1960s.811 While their specific reasons for leaving varied, a majority indicated that they wanted to make money to be able to buy things – cloths, tools, etc. – because wage labour was not possible in the district. Few district residents in the mid-twentieth century were integrated into the cash economy of Ayacucho. As noted above, most were subsistence or semi-subsistence farmers who sold little for cash, though they did often engage in the barter economy. Temporary migration for wage labour offered an opportunity to generate cash income that did not exist in the district. Interestingly, few explained their decision to leave because of push factors in the district, such as increasing land pressures. This is not unusual. While they sometimes characterized their reasons for leaving as a lack of opportunity in the district, subtle pressures like land stress might not be consciously recognized by community members as a reason behind their decision to migrate. What is clear, however, is that migration offered an opportunity for rural residents to link their rural livelihoods to urban livelihood opportunities without completely severing rural lives and identities.812

7.4.2. Extensification and intensification of land use

The expansion of cultivated lands to use all available fields in the Quechua and lower Sunni zone was probably one of the first household adaptations by many in the district. Extensification later

810 Bolivar de Colchado, Estudio Distrital de Chuschi, 16. I have counted Chuschi and its annexes together – Chuschi (60), Cancha Cancha (110), Chaqolla (12), Uchuiri (16), while the figures for Quispillaccta (195) are not broken down by population centre. 811 IIP estimated that the population of the district in 1966 was 7507, which gives an outmigration rate of about 5% (using 393 counted migrants) for the 8 months in 1966 when they tracked out-migration from the district. Their 1966 population figure, however, is an extrapolation based on 1961 census data. So, it is not a solid estimate. If we estimate Quispillaccta’s 1961 population at about 2600 and assume a similar migration number given for 1966 (195), we get an outmigration rate of about 7.5% (1961 census figures cited by Bolivar de Colchado’s IIP study for Quispillaccta - population centre (Llacta) and all its annexes except Socobamba - give a population of 2563 (Bolivar de Colchado, Estudio Distrital de Chuschi, 12); so, 2600 is a fair estimation.). It thus seems reasonable to assume an outmigration rate in this period between 5-10% for each community. 812 This linking of urban and rural livelihoods is brilliantly explored by Karston Paerregaard, Linking Separate Worlds: Urban Migrants and Rural Lives in Peru (Oxford: Berg, 1997).

248 spread to higher Sunni and Puna zones, as more high-altitude lands were brought under cultivation in the latter half of the twentieth century. These adaptations were more acutely necessary in Quispillaccta because of greater supply and demand pressures, though Chuschi and its annexes also faced similar pressures during this period.

Cultivated land in Quispillaccta was always much more limited than in Chuschi and its annexes. So, the rapid increase in the community’s population meant that people had to work what land they had as intensely as possible. In Quispillaccta, only about 2% of their land is cultivatable, with only about 240H of maize land available to households.813 For most of the twentieth century the majority of land cultivation in Quispillaccta took place in the Rio Pampas zone around the town centre and to the west. The low-altitude area fronting the Rio Pampas is known as Mollebamba, and the fields around Mollebamba were historically the main maize growing zone for Quispillaccta.814 The earliest aerial photos the district’s maize and cultivated areas from the mid-1950s show extensive cultivation and little unused land in this area, compared to the late 1990s.

813 Machaca, “Vigencia y Continuidad,” 73. Tirso A. Gonzalas, “The Cultures of the Seed in the Peruvian Andes,” in Stephen B. Brush ed., Genes in the Field: On Farm Conservation of Crop Diversity (Ottawa: Lewis Publishers, International Development Research Centre, and International Plant Genetic Resources Institute, 2000): 205; José Coronel et al., Estudio Comparativo de Comunidades Campesinas Altoandinas: Ayacucho (Ayacucho: Centro Internacional de Investigaciones para el Desarrollo/Asociación Peruana para el Fomento de las Ciencias Sociales, 1990): 29. Machaca notes that there are 470.48H of cultivated land in Quispillaccta – about 2.17% out of a total landholding of 20,989.48H. In last decades of the twentieth century, there has been an expansion of cultivated lands both to higher altitudes in the Rio Pampas zone and within the Rio Cachi zone, as discussed above. So, the 2.17% figure and Coronel’s figure of 240H of maize lands are probably higher than they were in the mid-twentieth century. According to 1960 census figures, there were about 500 families in Quispillaccta (if we use the conservative 5- person/family figure), which translates to less than 1/2H of maize land per Quispillactino household. 814 Machaca, “Vigencia y Continuidad,” 21.

249

1955 aerial photo of maize zone of Chuschi and Quispillaccta. Rio Qunchalla/Rio Chuschi divides land holdings in the valley. The land is heavily cleared in all but the steepest slopes, with very little vegetation evident in the cultivated areas. Photo is from August, 1955, at the height of the dry season, several months after harvest. Land use patterns are similar in the 1974 aerial photo below. However, the 1997 photo below shows extensive abandonment and regrowth of brush in lower zone fields. (Source: Servicio Aerofotografico Nacional, Peru. Annotations added by author.)

250

1974 Aerial photo of District’s maize zone. This photo was taken in May, at the height of the harvest. (Source: Servicio Aerofotografico Nacional, Peru. Annotations added by author.)

251

1997 Photo of Quispillaccta’s traditional maize zone, illustrating extensive abandonment of fields and regrowth of vegetation. Ground-truthing in this area during field work in 2004 confirmed the widespread abandonment of cultivation in the area. Photo taken in June, 1997, at the start of the dry season. (Source: Servicio Aerofotografico Nacional, Peru. Annotations added by author.)

With maize land limited, these areas were either never fallowed or fallowing quickly disappeared as land demand increased over the past century. Informants in both communities note that these lands between 2450 and 3600m fronting the Rio Pampas were usually never fallowed in years past, but cropped year after year. Generally, continual cropping is an indication of the intensification of agricultural activities. However, as in other parts of the Andes, the limited availability of maize land may have meant that these lands were rarely fallowed in the past or

252 else fallowing ended when land demand increased at some point in the distant past.815 Juan Galindo does not remember families fallowing land in the lower corn growing zone, though he allowed that perhaps some families did fallow. He noted that harvests used to be very good in the lower zone, but they fell terribly in the late 1960s and 1970s, a factor he attributed to greater heat in the area.816 Nemesio Nuñez said that his family never fallowed their maize fields or added livestock manure (guano in Spanish) in the maize fields of Mollebamba.817 Adding fertilizer like manure is an intensification strategy. Traditionally, some communities in the district like Quispillaccta seasonally rotate their livestock herds from upper zones to the lower zones after harvest to feed on crop stubble and to help fertilize the fields with their manure.818 However, individual households also learned to add manure to their individual plots at the time of planting to deal with declining soil fertility. But this practice was not widespread nor historically common according to informants until later in the twentieth century, suggesting that it was an intensification strategy pursued by some households as land pressure increased. Chuschino elders similarly noted that maize fields were not fallowed in the days of their youth and manure never added because the area is too far from their livestock zone in the far north- west of the district.819 Leon Huaman Conde says that his field in Chuschi’s maize zone has never been fallowed, and now it does not produce like it used to produce when he was young.

815 Machaca notes that low-zone maize lands in Mollebamba were fallowed for ten-year periods (Machaca, “Vigencia y Continuidad,” 20); however, none of my informants remember these lands being fallowed. Mayer notes that in the Mantaro valley, low-zone maize lands were cropped continuously when he studied the area. He speculated that in the past the maize zone in his study area may have had long cropping cycles and short fallow periods. However, few fallowed these lands at the time of his study in the 1970s, partly because of need but also because fallowed fields at these low elevations quickly become filled with difficult to remove, invasive African Kikuyo grass. I did not find evidence of Kikuyo grass in the District of Chuschi, suggesting that land need probably ended any fallowing of maize lands long ago. Enrique Mayer, Land-Use in the Andes: Ecology and Agriculture in the Mantaro Valley of Peru with Special Reference to Potatoes (Lima: CIP, 1979): 67 & 70. 816 Interview, Juan Galindo, 80, Quispillaccta, 11 Oct., 2004. 817 Interview, Nemesio Nuñez, 55, Quispillaccta, 18 Oct., 2004. Several other Quispillactino informants similarly noted that they never fallowed their fields in the maize zone. However, one informant indicated that he rests his maize field fronting the Rio Pampas whenever he judges that it needs to be fallowed, about every ten years or so, for about 2-3 years. Interview, Marcello Tomaya, Quispillaccta, 9 October, 2004. Another field I examined in Suytucoral, facing the Rio Pampas and owned by Don Theodosio Flores Galindo, had just been fallowed for one year and was now being planted in tubers. Galindo was planting corn in the upper portion of the field and tubers in the lower portion, indicating that in some areas, the division in cropping zones between tubers/grains and maize is not clear-cut. Interview, Don Theodosio Flores Galindo, Quispillaccta, 8 October, 2004. 818 Machaca, “Vigencia y Continuidad,” 57. 819 Interview with representatives of Chuschino Ancianos group, Chuschi, 15 October, 2004. The 1967 IIP study of the District of Chuschi, however, noted that community members used livestock manure on their crops, particularly on their maize crops. Ramón C. et al., La Comunidad de Chuschi, 36.

253

He also thought that the heat was greater today.820 While fallowing fields in the district’s lower maize zone appeared uncommon, farmers did undertake activities in this zone to try to maintain soil fertility, particularly given the steep gradient of the zone. Terracing is widespread in the area, and terraces require constant attention to maintain retaining walls and irrigation channels.821 When labour availability permitted, households would have regularly expended effort to maintain their terraced fields.822 As well, other forms of cultivation were practiced in the area along with maize to bolster soil fertility - inter-cropping with associative species like wheat, broad beans, amaranth, and squash, or even adding ash to the soil when chemical fertilizers were unavailable.823 These activities were no doubt practices with a long history in the Andes; however, in the context of increasing pressure on a limited land base, their adaptive importance became especially relevant.

Many informants noted that the need to add manure to their cultivated fields showed the increasing stress on their cultivated assets, particularly because adding manure to fields was not something traditionally done by households until the latter half of the twentieth century. Juan Galindo of Quispillaccta’s barrio of Huertahuasi said that he first saw someone adding guano to their fields in the mid-1950s. From then, little by little, they added more guano. Now, he notes that households are constantly moving guano around from grazing areas to their cultivated fields.824 Pablo Conde noted that guano was never added to maize crops when he was young; however, today, even with manure and artificial fertilizer cob size and quality is poor.825 Victor Galindo Tucno, one of the oldest community members interviewed, noted that they did

820 Interview, Leon Huaman Conde, 79, Chuschi, 31 October, 2004. 821 Trawick argues that the primary purpose of terraces according to peasants themselves is water conservation, rather than to maintain soil fertility. “Erosion control and soil conservation are benefits that people recognize [with terraces], but they do not usually mention these results on their own.” Trawick, The Struggle for Water in Peru, 92- 3. 822 Abandoned terraced fields, however, are a source of soil erosion, as discussed above. 823 Machaca, “Vigencia y Continuidad,” 125-130 & 130; Ramón C. et al., La Comunidad de Chuschi, 36. Machaca does not indicate whether ash was “added” to field by burning crop stubble, though this could be inferred. Later studies of soil erosion in the district by Gualberto Machaca Mendieta note that burning fields was considered a sign of poor land management; so, burning may have been discouraged. 824 Interview, Juan Galindo, 80, Quispillaccta, 11 October, 2004. Various informants from both villages noted that adding guano to fields became common practice in the 1950s. 825 Interview, Pablo Conde, 70, Quispillaccta, October 25, 2004. Changes in crop harvest quality is discussed further below in section 7.6.

254 sometimes add guano to the “head” of the field when he was young, but that this wasn’t necessary if one’s land was good and fertile.826 Clearly, the move to increasingly add manure to fields mid-century indicated increasing pressure on cultivated assets for many households. This increasing supply scarcity was confirmed by former Chuschi authority, Ernesto Jaime who said that land started to decline in quality in the late 1940s and early 1950s.827 Jaime said that these developments worried comuneros who turned to their authorities for help. But it wasn’t until the late 1950s when authorities in the district were able to offer assistance with the introduction of chemical fertilizers to the district. Artificial fertilizers helped to provide good harvests at first, according to Jaime. However, their use also introduced new agricultural problems after a few years, as discussed below in section 7.6.828 Fertilizer and pesticide use were still not widespread when IIP undertook their study in 1967. They reported that only a handful of mestizo elite agriculturists used artificial fertilizers on their fields, and only a few cultivators sprayed their crops with pesticides to prevent plagues.829

Cultivated lands higher up the altitude slope, where maize gives way to tuber and cereal production – the Sunni zone and low Puna zone - were similarly never fallowed, according to interviews with district residents, because cultivated lands were growing ever scarce. However, in these areas, farmers increasingly adapted by adding natural and artificial fertilizers and by maintaining crop rotation. As Chuschino Emilio Guillen Ballon commented, “When you have big chacras you can fallow parts of them, but with small chacras we need to work them all the time, like now; but we do rotate crops.”830 Chuschino Victor Esteban Caillahua Mikuylla concurred: “We didn’t fallow our maize chacras, but we did rotate other chacras, planting potatoes after wheat,” for example.831 Quispillacctino Joaquin Conde Pacotaype noted that he

826 Interview, Victor Galindo Tucno, 99, Chuschi, 24 October, 2004. 827 Interview Ernesto Jaime, Chuschi, 19 October, 2004. 828 Interview Ernesto Jaime, Chuschi, 19 October, 2004. Quispillactino Pablo Conde noted that he first began using pesticides on his crops in the early 1960s, when Swiss NGO technicians showed them how to use them. Interview, Pablo Conde, 70, Quispillaccta, October 25, 2004. 829 Ramón C. et al., La Comunidad de Chuschi, 36 & 41. IIP noted that there were two households that employed backpack pesticide spray units, and their success had inspired several other households to seek to acquire these tools. It is likely that other households in both communities employed small numbers of fertilizer and pesticides on a small scale from the early 1960s, but that this use was not examined by the IIP researchers. 830 Interview, Emilio Antonio Guillén, Chuschi, 19 October, 2004. 831 Interview, Victor Esteban Caillahua Mikuylla, Chuschi, 24 October, 2004.

255 used to never rotate crops when he was young. He first started rotating crops in the late 1980s and now they do it out of necessity.832 Machaca’s study notes that crop rotation is used in Quispillaccta predominately in the Sunni zone between 3600m and 4000m, and is only rarely used in the higher parts of the maize zone (Qichwa) and the lower parts of the puna zone (Sallqa).833 In her studies of household cultivation, crop rotation – following tuber cultivation with broad beans, barley, garlic, or wheat crops – was usually practiced for 4-5 years before fields were fallowed for 3-10 years.834 IIP’s 1967 study of the D. Chuschi similarly noted crop rotation in high-altitude tuber fields, but does not mention crop rotation or fallowing in maize fields. IIP noted that potato crops were followed in rotation by broad beans, other Andean tubers like oca and olluco, and then by wheat, before fields were fallowed for 1 to 3 years. Following the next crop rotation cycle, IIP reported, fields were fallowed for a longer period, between 4 to 8 years.835 Fallow periods varied according to when cultivators determined that fields had sufficiently recovered to be planted again.836 Some households clearly did not fallow their fields, suggesting that decisions to fallow were partly dependent upon a household having sufficient land or livestock endowments, as informants noted above. Crop rotation was likely always used in some zones, and after some time in newly opened chakras when new lands were available. But decisions to fallow were likely dependent upon the extent of asset endowments. As land endowments faced demand pressure, the frequency of fallowing likely declined for many households in the district.837

These patterns are consistent with evidence in other parts of the Andes. Mayer’s study of the sectoral fallow system in the Mantaro Valley – a community organized fallow rotation system,

832 Interview, Joaquin Conde Pacotaype, Quispillaccta, 31 October, 2004. 833 Machaca, “Vigencia y Continuidad,” 143. “Qichwa” zone is another spelling for the Quechua zone, the lowest elevation maize zone between 2400m and 3600m outlined above in chapter 4. 834 Machaca, “Vigencia y Continuidad,” 144. 835 Ramón C. et al., La Comunidad de Chuschi, 36. 836 Ramón C. et al., La Comunidad de Chuschi, 36. 837 This conclusion is also supported by a 2006 study of the challenges of climate change and sustainable development in the District of Chuschi. In areas of rain-fed cultivation in the district, the study noted that historic crop rotation patterns of these lands have changed with the increase in population pressure, leading to the elimination of fallowing on these lands. Gobierno de Ayacucho, Plan Para el Uso Sostenible de los Recursos Naturales Altoandinos y Prevención de los Efectos Climatológicos Adversos del Distrito de Chuschi, Provincia de Cangallo del Departamento de Ayacucho, (Huamanga, Ayacucho: 2006), 78.

256 where every household followed a set system of crop rotation and land fallow – found that pressures associated with “population growth and increasing land scarcity, independization of agro-life zones, and privatization of land” led to the progressive dismantling of the community controlled sectoral fallow system in the 20th century and an increase in land under cultivation.838 As land became less available because of population growth or as households decided that market opportunities made it advantageous to intensify production for monetary production, extensification and intensification of land cultivation was practiced, and fallowing was reduced or abandoned. In many areas, community organized rotation and fallow systems were completely abandoned in favour of systems where individual households made decisions about rotation cycles or fallowing.839 Similarly, a study of sectoral fallow systems in the Central Andes by Orlove and Godoy also suggested that potato or maize fields where fallowing was not being practiced was an “unfortunate consequence of land fragmentation,” and that these areas were “former sectoral fallow systems.”840 By the latter half of the twentieth century, this type of situation also prevailed in the D. Chuschi, with limited or reduced fallow periods among some households indicating agricultural intensification at the expense of soil quality.

Given the limited endowments of maize lands in the D. Chuschi by mid-century, particularly for Quispillacctinos, it is not surprising that informants said that maize lands were never fallowed by households. As maize land holdings declined for households, there were few options but to work the land continuously.841 This pattern likely held in the district until the 1970s when many households, particularly in Quispillaccta, began to conclude that it was increasingly not worth expending effort in the lower maize zone compared to opportunities in other production zones. Aerial photos of the area from 1974 show the lower maize zone being heavily used; however, photos from 1997 show widespread abandonment of many lower zone fields. As Julian Vilca

838 Mayer, Land-Use in the Andes, 67. 839 Mayer, Land-Use in the Andes, 68. 840 Ben Orlove and Ricardo Godoy, “Sectoral Fallowing Systems in the Central Andes,” Journal of Ethnobiology 6(1), 1986: 174. Orlove and Godoy do not specifically note the causes of land fragmentation; however, among Andean specialists, demand-pressures like population growth and increasing market production are widely attributed as the causes of land fragmentation in the twentieth century. 841 Maize was never sold by households but always consumed or used for the production of chicha – corn beer. Mayer also notes that maize land in the Mantaro Valley communities he studied was also rarely, if ever, fallowed.

257

Mejía explained, “the lower zone was heavily used, and they learned to crop pastures and crops in the upper zone.”842

Learning and gradually improving climatic conditions at higher altitudes in the district were allowing community residents who had little land or declining crop opportunities in the lower zones to shift cultivation to higher production zones.843 This shift should also be seen as part of the extensification strategies used by peasant households to adapt to pressures faced in lower production zones, processes also noted by Mayer in communities in the Mantaro Valley.844 However, this strategy has always been risky because higher altitudes require longer growing periods and frost can heavily damage crops at higher altitudes. Field interviews with Quispillacctinos and research by Machaca indicates that maize cultivation became increasingly possible at higher altitudes in the D. Chuschi in the latter half of the twentieth century. Maize is now regularly being grown in areas where it was previously rarely grown, like the barrios of Llacctahurán (3,750m) and Pirhuamarca (3,800m), above the widely accepted 3,500m altitude limit.845 In some parts, Machaca notes that it is possible to come across maize fields at altitudes of 4,000m.846 There are local variations of a few hundred meters for the maize zone in many places in the district, given the extremely high agro-ecological diversity in the Peruvian Andes and depending on the particular local micro-climate. The relationship between altitude and temperature can have a significant impact on crops like maize. The altitude temperature gradient

842 Interview, Julian Vilca Mejía, Quispillaccta, 8 October, 2004. 843 According to Mayer, a production zone is “a community managed set of specific productive resources in which crops are grown in distinctive ways. These zones include infrastructural features, a particular system of rationing - resources (such as irrigated water and natural grasses), and rule-making mechanisms that regulate how the productive resources are to be used. Complementing the management of these resources are individual production units (such as households) that hold access rights to specific portions of these resources.” Mayer, The Articulated Peasant, 245. 844 Mayer writes that, “[p]opulation expansion and greater need for food crops resulted in the growth of permanent villages at higher altitudes than the nucleus. These anexos (annexes), still subject to the “mother village,” became specialized in the production of crops adapted to the intermediate zone.” In some cases, Mayer noted, annex villages completely specialized in specific agro-ecological production zones, eliminating long-held patterns still evident in the District of Chuschi, where households engage in agricultural or livestock activities across multiple agro-ecological production zones at various altitudes. Mayer, Land-Use in the Andes, 58-60. Emphasis in the original source. 845 The 3500m maize limit is similarly noted by Mayer in the Mantaro Valley communities he studied. However, he similarly notes that some households grew maize in the “intermediate altitude and climate range, with low yields.” Mayer, Land-Use in the Andes, 52. 846 Machaca, “Vigencia y Continuidad,” 21.

258 in the district illustrates the risks of pushing cultivation to higher zones. Based on studies in Ayacucho, Earls calculated, for example, that an altitude increase of 10C results in a 9-day extension in the maturation cycle for maize.847 So, efforts to grow maize at higher altitudes in the district faced various micro-climate and altitude challenges. For this reason, establishing typical local patterns of agro-ecological limits through research and interviews is important to charting change over time in the district. Community members in Quispillaccta noted in interviews that maize was not grown in Llacctahurán before the 1950s.848 Machaca, who uses 3,600 as the normal limit for the maize zone in Quispillaccta, also notes that some farmers have successfully grown maize for more than a decade in “test” fields in the Rio Cachi watershed, in the upper puna zone of the district, close to the barrio of Unión Potrero at 3,700m. This is an area that historically never saw maize cultivation because of frequent freezes and unreliable precipitation.849 Clearly, the need for maize land and increasingly favourable climate conditions – likely a result of climate change since the early 1970s – have led community members to successfully introduce maize cultivation at higher altitudes in the community. As Nicolás Machaca noted, “where are we going to go if not up to the puna to plant corn, because our soil is quite reduced, especially in the Qichwa zone.”850

847 Earls, “The Character of Inca and Andean Agriculture,” 16. 848 Interview with members of the Quispillactino Senior’s Group (ancianos), 9 Oct., 2004. 849 Interview with members of the Quispillactino Senior’s Group (ancianos), 9 Oct., 2004, and Machaca, “Vigencia y Continuidad,” 125-6. 850 Quoted in Machaca, “Vigencia y Continuidad,” 22.

259

Figure 26: Altitudinal Shifts in Maize zone in District of Chuschi

1 2 1 1 1 1 1 1

Rio Pampa

s 4 1 1

1 1

Source: Instituto Geográfico Nacional, Perú.

Topographical Close-up of Chuschi District Maize Zone North of Rio Pampas

The highlighted yellow area indicates the zone below 3,600m traditionally suitable for maize cultivation, according to Machaca’s study of Quispillaccta. These fields fall steeply towards the Rio Pampas to the south, which runs west to east. At altitudes of 3,600m or below, Machaca’s crop surveys indicated maize as the dominant crop. Given that Machaca’s study was completed in 1991, her surveys have probably captured climate change-induced shifts of maize cultivation to higher altitudes. 1. The highlighted yellow areas of the map show that maize was possible up to the barrio of Llaqtahurán (Llactauran), because much of the barrio is in the yellow zone; however, informants indicate it rarely grew there until recent decades. 2. In the valley holding the Chuschi and Quispillaccta town centres (Matriz), informants also note that corn rarely grew above the point indicated by the arrow, where the road into Chuschi turned south. Above this altitude, broad beans, potatoes, and barley were traditionally grown. 3. Now, however, corn often grows to Matuma, indicated at point 3, an increase of several hundred meters in altitude. When the map was produced in the 1960s, cartographers coloured green those areas visibly under cultivation from aerial photographs. Green areas include areas above 3,600m where various non-maize grains and tubers are grown. 4. Fundo Choccechaca in prime irrigated maize zone of Chuschi. (Source: Machaca pp. 123-4;

Interview with Emilio Conde Achalma & Pedro Condé Nuñez, 11 October, 2004.)

Mayer concluded from studies in other parts of the north-central Andes that lower zones, where land can be more intensively worked, are likely to eschew organized fallow systems and instead be cropped continually with maize.851 In such zones, when extensification of land holdings is

851 Mayer, The Articulated Peasant, 257.

260 impossible because there is no more land to bring into production, the limitations on continuous, intensified cropping are instead likely to be the availability of irrigation, and the complex calculations of labour availability and household production zone priorities.852 Such factors were also operating in Quispillaccta’s maize zone, providing incentives for households to turn away from investing effort in the lower maize zone. Decreasing yields, limited or reduced access to water – both irrigated water and rainfall according to informants, and the comparatively better opportunities for cultivation and herding in higher zones, led many to increasingly abandon or limit cultivated activities in the lower zone after the 1970s. A field visit to Mollebamba in late 2004 – Quispillaccta’s low maize fields close to the Rio Pampas - during what should have been the prime maize planting period instead found field after field abandoned and overgrown. No new crops had been planted, and numerous fields of stunted corn and wheat from previous year’s planting remained unharvested. Farmers had not even bothered to feed the stunted crops from the previous year to their animals. Many fields were heavily overgrown, obviously unplanted for years. If the rains are good, some may still plant these chacras, according to my research assistant Hereberto Nuñez.853

852 Mayer, The Articulated Peasant, 258 & 261. 853 Field visit, Mollebamba zone, Quispillaccta, October 10, 2004.

261

New roads to Llacctahurán and Soccobamba. The Historic landslide upper road was being built during my fieldwork in area 2004.

Rio Pampas

Mollebamba Mollebamba

Aerial photo of Quispillaccta’s historic maize zone from Google Earth, 2016. The abandonment of cultivated fields has continued to grow in this area, though a small concentration of fields in Mollebamba continue to be cropped. The Santa Rosa landslide area is also clearly visible in this photo. (Source: Google Earth.)

Water is especially important for maize cultivation, particularly on the western slopes of the Andes where there is a frequent shortage of moisture throughout the growing season.854 Some maize areas in the district have seen less water availability and more precipitation variability in recent decades, impacting cultivation activities in the maize zone, particularly in the lower maize zone of Quispillaccta. The fields fronting the Rio Pampas in this area have the highest temperatures in the district, so water availability is key. “Maize plants have certain critical periods,” Earls notes. “For instance, the plant’s water requirements increase drastically at flowering time. If it does not receive adequate water at this time yield can be reduced by 50- 60%.”855 Respondents from both communities noted the impact of both decreasing rainfall and increased precipitation variability in the area. Pedro Conde Nuñez said his family does not farm in Mollebamba anymore because the rain is not like it was in years past.856 Emilio Conde

854 Mayer, The Articulated Peasant, 258. 855 Earls, “The Character of Inca and Andean Agriculture,” 16. 856 Interview, Pedro Condé Nuñez, Quispillaccta, 11 October, 2004.

262 thought that there is now less rain in this area now, which is why there is little corn grown in this area. This weather change has impacted this area’s crops.857 Juan Galindo of Quispillaccta said that the heat in Mollebamba is the problem. In the old days, the harvest there was good. But from the 1970s, the harvest there fell terribly and they stopped sowing there.”858 Gregory Conde Nuñez agreed that heat is the problem in Mollebamba, and even if there is rain there, production is half what it used to be.859 Joaquin Conde Pacotaype noted that the quality of rainfall had also changed along with the more intense heat in the lower zone: “In the old days, the rain was soft and now the rain and wind are hard and strong, and the sun is intense like in the city. It’s almost as if the sun is coming down closer to the ground,” he said. “Because of the weather, only chacras (fields) with irrigation do well; the others don’t do well and there is no security in cultivation, especially cultivating on the slopes, compared to the greater security with harvests in the old days.”860 Chuschi’s elders concurred that both rainfall patterns and the heat in the lower maize zone appears to have changed in recent years, impacting rainfed cultivation: “Forty years ago, the rains always seemed to start in September; but now there’s more variability. There’s not much rain now during sowing in September and October. The heat seems more intense in this zone. In those chacras where there’s no irrigation, sowing depends on the amount of rain, and if there’s no rain, they don’t sow them. So, there’s abandoned fields. Chuschinos are increasingly working the lower fields with fruit trees.”861 Such insights are consistent with maize field experiments done by Earls that show the importance of climatic conditions in the first month after sowing for the length of the vegetative cycle. Reduced water availability or increased uncertainty about water availability in the first month can dramatically alter the risks for maize crop yield and quality. 862 Cultivators in the district have had to adapt in the latter half of the twentieth century to uncertainties and risks in their cultivation in the lower maize zone, sometimes by shifting livelihood activities to other zones or crops.

857 Interview, Emilio Condé Achalma, Quispillaccta, 11 October, 2004. 858 Interview, Juan Galindo, 80, Quispillaccta, 11 October, 2004. 859 Interview, Gregory Conde Nuñez, Quispillaccta, 20 October, 2004. 860 Interview, Joaquin Conde Pacotaype, Villa Vista, 861 Interview with representatives of Chuschino Ancianos group, Chuschi, 15 October, 2004. 862 Earls, “The Character of Inca and Andean Agriculture,” 17.

263

A lack of adequate irrigation created additional uncertainty in Quispillaccta’s lower maize zone and may have helped to limit maize production in this zone. For centuries, this area was served by an irrigation channel drawn from the stream dividing Quispillaccta and Chuschi. The mountain directly west of Quispillaccta, which has historically been extensively cultivated, is criss-crossed with five main irrigation canals. Irrigated land is highly important for households. Economist José María Caballero believes that irrigated land is equivalent to twice as much rain- fed land.863 The lowest canal on the mountain, called Itanawayqu by Quispillacctinos, brings water from the Rio Qunchalla/Chuschi, beginning around 3200m to the south-west fields fronting the Rio Pampas.864 Unfortunately, at some point in the mid-twentieth century, a major land slide blocked this canal, cutting off irrigation to the lower maize zone of Quispillaccta.865 Hiking through the area today, the site of the land slide is evident, as a massive portion of the hill appeared to give way. The slide area is easily visible on the earliest aerial photographs from the

863 José María Caballero, Economía Agraria de la Sierra Peruana: Andes de la Reforma Agraria de 1969 (Lima: Instituto de Estudios Peruanos, 1981): 97. Caballero argues that 1H of irrigated cultivated land is equivalent to 2.1H of rain-fed cultivated land. 864 Machaca, “Vigencia y Continuidad,” 22-24; Coronel et al., Estudio Comparativo, 57. Machaca notes four main channels and one sub-channel in her diagram of the irrigation canals, Machaca, “Vigencia y Continuidad,” 22, while Coronel’s diagram shows four active canals, but does not show the damaged Itanawayqu canal. 865 Machaca, writing in 1991, noted that the landslide happened “more or less 20 years ago.” However, the slide’s visibility on aerial photos from 1955 demonstrate the slide more likely happened in the mid-twentieth century.

264 area, dating the landslide to before the first aerial photographs in 1955. With irrigation unavailable and ever-precarious rains in the area, many Quispillacctinos like Joaquin Conde Pacotaype, have given up farming in the lower maize zone fronting the Rio Pampas. He says

Close-up of landslide area in lower maize zone of Quispillaccta, taken from 1974 aerial photo. The slide site is also visible in the 1955 aerial photo; however, the resolution is much better in the 1974 image. Some farmers returned to planting their fields on the edge of the slide zone, as is evident in the photo. But a large portion below the new cliffs remain uncultivated. The Itanawayqu canal destroyed by the slide used to flow above and to the left of the river course, bringing water to fields in Ruqruqa and then onto Mollebamba in the lower left of the photo. These relatively flat fields front the Rio Pampas. (Source: Servicio Aerofotografico Nacional, Peru. Annotations added by author.) that he abandoned his chacra in the low zone (Chiwaypata) about forty years ago – in the late 1950s or early 1960s – because there is no irrigation there and not enough rain. “In the mid- 1950s many people abandoned their chacras in the low zone,” he said, “because the Santa Rosa cliffs collapsed and wrecked the canal that brought water to this area. The authorities never rebuilt the canal because they were afraid of another land slide in this steep area. So, they looked for pipes to bring water to the area, but never found any.”866

866 Interview, Joaquin Conde Pacotaype, Villa Vista, Quispillaccta, 31 October, 2004. Another Quispillactino, Melchor Quispe Nuñez of Huertahuasi, said that he has not sown a crop in the low zone (Chuihapata) for two years

265

More intense use of cultivated land in Quispillaccta that accompanied population growth in the mid-twentieth century also cleared trees and brush from many hillside areas in the cultivated zone. These trends probably increased soil erosion and may have contributed to destabilizing the soil to trigger the Santa Rosa cliffs landslide in the lower maize zone. Many informants noted that the land in the cultivated zone was much more heavily cleared in the mid-twentieth century, both because of the need for firewood and building material to supply an expanding community and because of the need for firewood for preparations for community fiestas.867 Community members in both Chuschi and Quispillaccta noted that they often had to travel to the steep slopes on the other side of the Rio Pampas to gather firewood because there was none in the district.868 Tomás Conde Quispe, from the remote Quispillaccta barrio of Soccobamba, fronting the Rio Pampas in the south-east of the district, said that when he was a boy the sides of the hills used to be bare. “There were fewer bushes and trees around this area because many people came from Llacta (main village) to gather wood.”869 Extensification and demand pressures led to increasing land clearing, probably worsening the stability of cultivated land.

Other evidence also indicates that the intensification and extensification of cultivated agriculture that accompanied growing populations in the district in the 20th century worsened land degradation on cultivated lands. Multiple assessments of soil erosion - by Peruvian experts over the last thirty years, through field measurements during field work in 2004, and from conversations with community members – indicate that it aggravated land degradation for some

now because the yields are poor and the land is tired. Barley and wheat are also poor there, he said. “Now just grasses and bushes grow there. The land used to be irrigated by a little irrigation channel that came from a spring. But now the spring comes out somewhere else; too low. And the cliffs in the area are sliding too much.” Interview, Melchor Quispe Nuñez, Huertahuasi, Quispillaccta, 31 October, 2004. 867 Interview, Theodosio Flores Galindo, Llacctahurán, Quispillaccta, 29 September, 2004; Interview Pedro Conde Nuñez, 11 October, 2004. Conde Nuñez said that the mountain above Quispillaccta’s main town used to be completely covered in fields, up to the high areas like Qispipata; however, now they don’t farm this area anymore. Don Theodosio noted that since the fiestas have ended in Quispillaccta, the trees have recovered because they are not cut anymore, and certain bushes like cerqa have recovered because they don’t use them anymore for home construction because there are tin roofs. 868 Interview with representatives of Chuschino Ancianos group, Chuschi, 15 October, 2004; Interview, Victor Galindo Tucno, 99, Chuschi, 24 October, 2004; Interview, Emilio Condé Achalma, Quispillaccta, 11 October, 2004. 869 Interview, Tomás Conde Quispe, Socobamba, Quispillaccta, 3 November, 2004. In his 40s when interviewed, his recollections would date these observations to the late 1960s or early 1970s.

266 in the cultivated zone in the district during the 20th century, both from heavier cropping of cultivated lands, but also because households increasingly abandoned the lower maize zone and shifted efforts to higher sunni and puna zones. These trends were particularly important for Quispillacctinos, who had less land on average than Chuschinos, and higher demand pressures by the mid-twentieth century. Soil erosion is challenging to assess with confidence in highly heterogenous agricultural zones like those found in the D. Chuschi, particularly given wide variations in cropping capacities among cultivators. The following insights are thus not definitive proof, but provide another layer of evidence about land degradation trends in the district during the 20th Century.

Soil erosion is held up by some as the single best proxy for various aspects of land degradation, and a major cause of falling agricultural productivity.870 However, “since the effects of soil loss vary depending on the underlying soil type, soil loss, by itself, is not an appropriate proxy measure for productivity decline.”871 Deep soils that are rich in nutrients can withstand soil loss better than thin, poor soils. Soils in the D. Chuschi are highly variable; however, there are significant areas where soils are thin and face significant erosion risks. General erosion studies of the district found erosion to be moderate grade 2 when fields are being cultivated. Erosion rates rise to unsustainable levels between 3-5 when fields are not being cultivated.872 The soils in the lower maize zone of Quispillaccta in Sunnipampa, close to the historic slide area, were found to be low in fertility, according to an analysis in the late 1970s.873 Fields around the heavily cultivated hill to the west of Quispillaccta’s Matriz have thin soils that have been exposed to erosion processes on the steep slopes over the decades of rain-fed agriculture, according to Marcela Machaca.874 In a study of the Rio Pampas watershed in the district in the late 1990s, including the major tributaries feeding into the Rio Pampas, Gualberto Machaca determined that

870 Michael A. Stocking and Niamh Murnaghan, Handbook for the Field Assessment of Land Degradation (London: Earthscan, 2001), 8. 871 Stocking and Murnaghan, Handbook, 14. 872 Coronel et al., Estudio Comparativo, 12. 873 Machaca, “Vigencia y Continuidad,” 54-55. Machaca notes that the soils in the low, Quechua zone of the district are classed as clay-loam, sandy loam, and silt loam. They have a PH between 5.74 to 8.45, and are high in calcium and salts. 874 Machaca, “Vigencia y Continuidad,” 67-8. This includes the high areas around Quispipata and steep areas around the barrios of Llacctahurán, Pirhuamarca, and Socobamba.

267 erosion varied from light to severe, depending upon the area examined.875 Soil loss in his experimental measures varied between 0.5 TM/H/Y to almost 3 TM/H/Y, within the FAO’s permissible range.876 However, Machaca notes, these rates are still worrisome because of the thin soils in the area.877

Field assessments in 2004 in various chacras in Quispillaccta’s maize zone similarly found wide variation in soil erosion rates, depending on location, slope, cultivation practices, etc. A field assessed near Llacctahurán , at the upper end of the maize zone and with a slope of 10 degrees, provided an estimated soil loss of 9.6 TM/H/Y – at the upper end of acceptable erosion for annual cropping areas.878 Multiple measures in another parcel in Quispillaccta’s maize zone fronting the Rio Pampas found much higher rates of soil loss. Two portions of the parcel were assessed; both had slopes between 10-15 degrees, though slopes were greater in the portion with higher soil loss. One small area had a very high loss rate of 138 TM/H/Yr, while a different portion of the field had loss rates slightly higher than acceptable, at 21 TM/H/Yr. 879 Another interesting measure of soil loss came from a small maize field outside of Quispillaccta’s Matriz. This small field about 250m2 (about 1 Melga) has always been planted with maize, according to Quispillacctino Esteben Nuñez. A large rock in the field has increasingly been exposed over decades of cropping. “It’s as if the rocks have grown from below,” Sr. Nuñez said, commenting on the growing exposure of rocks in his field over the years.880 Calculations of soil loss

875 Gualberto Machaca, Conservación y Recuperación de Suelos, 125. 876 Gualberto Machaca, Conservación y Recuperación de Suelos, 128. 877 Gualberto Machaca, Conservación y Recuperación de Suelos, 128. 878 The figure is the average of two measurements of Field 1, belonging to Theodosio Flores Galindo, Oct. 1, 2004, using tree root exposure and rock exposure. Factors that could lead to both overestimation and underestimation of soil loss using this method are discussed by Stocking and Murnaghan, Handbook, 54-55 & 62. Field measurements noted here and below were conducted through 15 measures of soil degradation spread across 8 different fields in the Quechua and Sunni zone of Quispillaccta, including tree root exposure, waterfall effect, rock exposure, pedestals, build-up against barriers, and build-up against tree trunks. Assessments were made according to the methodology and guidelines set out by Stocking and Murnaghan. See Appendix 1 for a table outlining the various field measurements of soil erosion used. 879 Both measures were done in Field 4, belonging to Virginia Nuñez, Oct. 2, 2004, using build-up against a barrier methodology, outlined in Stocking and Murnaghan. Factors that could lead to both overestimation and underestimation of soil loss using this method are discussed by Stocking and Murnaghan, Handbook, 68. 880 Interview, Esteben Nuñez, Quispillaccta, 6 Oct., 2004. Jon Hellin found similar reactions among farmers in Honduras, who often failed to recognize that soil erosion was taking place in their fields. “In Honduras, he writes, “it is not uncommon to hear farmers talk about ‘rocks growing out of the hillside’. Soil loss rates as high as 20 to 40 tonnes per hectare per year result in an annual lowering of the soil surface by less than 0.3 mm. As farmers cannot

268 measured from rock exposure over the past 50 years show soil loss rates of about 69T/H/Y, a troubling loss rate that is not sustainable.881 Sr. Nuñez says that he now often adds manure to fertilize the soil of this field, either at the top, on the borders, around the plants early in the growing cycle, or where he thinks that the soil needs additional fertility. He plants rows of corn up and down the slope of the field if he’s planting by himself. Labour shortages are a challenge for him; he doesn’t plant parts of the field where it is harder to work the land or where production is marginal, such as the steepest slopes. His yield of corn in this field is several times smaller than a similarly sized field in Llacctahurán, cropped by notable farmer Theodosio Flores Galindo, though Don Theodosio has more family assistance in his efforts. These variations in cropping ability and yield potential illustrate the wide variability in both erosion rates and cropping outcomes depending on prevailing conditions, labour availability, opportunity, etc. Given the steepness of many parcels, particularly in the Rio Pampas zone of the district, high rates of soil erosion can be expected, even for skilled cultivators. Measurements in a 13 degree sloped portion of a field facing the Rio Pampas show high erosion rates of 81T/H/Y.882 The measurements were made against a barrier constructed by the farmer only one year previously; an obvious attempt to preserve the soil at the top of his field.

While we cannot say for certain whether soil erosion led to increasing land degradation for households farming this area in the mid-twentieth century as demand pressures intensified in the area, the challenges of cultivated agriculture in this erosion prone area were an additional risk that households had to manage. Soil erosion risks likely added to land degradation pressures noted above to worsen supply scarcities in the Rio Pampas cultivated zone for households. Certainly, assessments of soil erosion in the latter decades of the twentieth century noted widespread problems in the district, particularly in the steep zones of the Rio Pampas area, as

see this erosion occurring, the explanation of rocks growing is a logical explanation for rocks becoming exposed.” Jon Hellin, “From Soil Erosion to Soil Quality,” LEISA Magazine 19(4), December 2003: 10. 881 Assessment of erosion was made in Field 6, belonging to Esteben Nuñez, Quispillaccta, on 6 Oct., 2004, employing rock exposure methodology outlined by Stocking and Murnaghan. Factors that could lead to both overestimation and underestimation of soil loss using this method are discussed by Stocking and Murnaghan, Handbook, 61-2. 882 Assessment of erosion was made in field 8, belonging to Theodosio Flores Galindo, Llacctahurán, on Oct. 8, 2004, employing build-up against a barrier methodology outlined by Stocking and Murnaghan. Factors that could lead to both overestimation and underestimation of soil loss using this method are discussed by Stocking and Murnaghan, Handbook, 68.

269 indicated by the pink erosion areas on the land-use map of the district below, Figure 27. Historic zones of steep Quispillaccta cultivation in the Rio Pampas zone to the west of the Matriz show widespread erosion, along with widespread erosion in puna grazing zones. In fact, a 2006 district plan on climate change and sustainable development listed most of the historic cultivation lands in the low Sunni and Quechua zones fronting the Rio Pampas as unsuitable for cultivated agriculture because of their steepness, erosion risk, and low natural fertility. Instead, the plan recommended that this historic centre of cultivated agriculture in the district be transformed into forestry plantations.883

Figure 27: Land-Use in the District of Chuschi Pink area shows eroded land. Source: Asociación Bartolomé Aripaylla (ABA), Agricultura sostenible para la adaptación al cambio climático en regiones vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, Ayacucho, Septiembre 2011: 19.

883 Gobierno de Ayacucho, Plan Para el Uso Sostenible de los Recursos Naturales Altoandinos y Prevención de los Efectos Climatológicos Adversos del Distrito de Chuschi, Provincia de Cangallo del Departamento de Ayacucho, 75-6.

270

Erosion problems in the D. Chuschi’s maize zone led to extensive planting of eucalyptus, a fast- growing foreign species, on steep hillsides above the town centres of Chuschi, Cancha Cancha, and Chacola, and in a small area above Quispillaccta’s Matriz, as indicated in teal on the map above. Such plantings created additional problems, however, and may have worsened land degradation in the district. Eucalyptus were originally introduced into Peru’s south-central Sierra in the mid-to-late 1800s as a source of fuel and timber.884 Informants recall eucalyptus first appearing in the district in the late 1950s, often planted on the borders of family fields.885 Dickenson notes that such plantings are common because farmers “can take advantage of water destined for irrigating crops.”886 Authorities in the district began to promote the planting of eucalyptus trees in the late 1950s to stabilize soil and to Four large eucalyptus trees planted on the border of a field belonging to Don Theodosio Flores Galindo, outside Llacctahurán. Don Theodosio said that these trees have ruined deal with the scarcity of wood as a this corner of the field, soaking up most of the rain that falls here. Interview, Theodosio Flores Galindo, Llacctahurán, Oct. 1, 2004. Photo by author. result of the widespread clearing of brush and trees described above.887 Areal photos of the district Matriz from 1962 show the steep slopes above Chuschi’s town centre crowded with fields; however, by 1974 the same area is increasingly overgrown as a young eucalyptus plantation takes root, apparently the first large scale planting of the trees for soil stabilization in the district.888 Gade notes that Peru’s Agrarian

884 Joshua C. Dickinson III, “The Eucalypt in the Sierra of Southern Peru,” Annals of the Association of American Geographers 59(2): 1969, 296. 885 Interview, Emilio Condé Achalma, Quispillaccta, 11 October, 2004. 886 Dickinson, “The Eucalypt,” 298. 887 Interview, Nemesio Nuñez, Unión Portrero, Quispillaccta, 18 October, 2004. 888 Today this eucalyptus forest towers above the town centre of Chuschi. Former Chuschi authority Marcelino Rocha Cayllahua told me that the community planted the eucalyptus above Chuschi in 1971. Interview, Marcelino Rocha Cayllahua, Chuschi, 24 October, 2004.

271

Reform government promoted eucalyptus for community agro-forestry plantations in the late 1960s and early 1970s.889 Interviews with district residents show that eucalyptus plantations also became an additional source of dispute in the long-running land conflict between Quispillaccta and Chuschi in the 1970s. Several informants describe an event in the mid-1970s when Quispillacctinos, under the cover of darkness, ripped out hundreds of eucalyptus seedlings planted by Chuschinos in a disputed area around Matuma, in the Kimsa Cruz disputed zone.890 To community members, eucalyptus did more than just stabilize the soil; they signified and guaranteed land ownership and control. Such conflicts over tree plantations continued into the early 1980s.891

Land and erosion control aside, however, eucalyptus trees also damage nearby crop production because of their heavy water usage and allelopathic effects from leaf litter and roots.892 The tendency of farmers to plant eucalyptus on the borders of their fields – a practice that informants noted also helped to mark boundaries and protected against people cutting their trees – negatively impacted surrounding crops by harming yields and reducing water availability for crops.893 In fact, given their impacts, it is highly questionable whether eucalyptus should have

889 Daniel W. Gade, Spell of the Urubamba: Anthropogeographical Essays on an Andean Valley in Space and Time (New York: Springer, 2016), 180. 890 Interview with Quispillaccta Ancianos Association members, Quispillaccta, 9 Oct., 2004. 891 Conflicts over eucalyptus continued as part of the inter-community struggle between Chuschi and Quispillaccta into the early 1980s, with both sides accusing the other of ripping out seedlings on contested land. See La Serna, To Cross the River of Blood, 130 and 132. 892 Studies show that the allelopathic impacts of eucalyptus trees negatively impact grains like maize and wheat in planted fields, reducing seedling germination, and stunting plant and root growth. M.A. Khan, “Allelopathic Effect of Eucalyptus on Maize Crop,” Pakistan Journal of Biological Sciences 2(1), 1999: 390-93; D. Blaize et al., “Effects of Eucalyptus on wheat, maize, and cowpea,” Allelopathy Journal 4(2), 1997, 341-344; As well, the fast growth of eucalyptus comes at a cost of significant water usage. In Kenya, the heavy water usage character of eucalyptus is so well known that they are named “drinking water” trees in the Mukungugu and Embu languages. Brandy Garrett Kluthe and Diana Chen, “Eucalyptus sp. at the intersection of environment and culture in Kenya,” Ethnobiology Letters 8(1), 2017: 18; If eucalyptus roots reach the groundwater table, Martin von Roeder notes, their annual evapotranspiration rates may be higher than annual rainfall, gradually depleting groundwater. “In cases where eucalyptus plantations replace grassland or shrubby natural vegetation,” von Roeder continues, “the water use of the plantations is in most conditions higher than the use of the original vegetation.” Martin von Roeder, The Impact of Eucalyptus Plantations on the Ecology of Maputa Land with Special Reference to Wetlands, Unpublished M.Sc. Thesis, Faculty of Landscape Planning, Technische Universitat Munchen, Munich, Germany, November 2014: 10. 893 Numerous informants spoke of the negative impacts of eucalyptus trees on crops. In late 2004, during my field work in the district, the NGO ABA was encouraging Quispillacctinos to cut down Eucalyptus trees for home construction and replace them with native species that do not harm crop production. Dickinson also notes the planting of Eucalyptus to mark the borders of fields in the early 1960s in the Central Valley of the Sierra. Dickinson “The Eucalypt,” 302.

272 ever been planted for erosion control in semi-arid climates like Ayacucho. “Under dry conditions ground vegetation is suppressed by root competition,” from eucalyptus trees, the FAO noted, making eucalyptus ill-suited for erosion control.894 Thus, “dense plantations of eucalyptus … are usually not recommended for erosion control, particularly in semi-arid climates.”895 Attempts to manage land degradation by introducing eucalyptus probably had negative livelihood impacts for cultivators in the district in the latter decades of the twentieth century.

Precipitation and temperature changes

Decreases in water availability are particularly important in the semi-arid environment of the D. Chuschi. Water impacts from eucalyptus plantings have been compounded by changes in weather noted by informants in cultivated zones in the district. These weather changes have increased uncertainty among cultivators in the district, particularly in the lower Rio Pampas zone. Melchor Quispe Nuñez, of the high-altitude barrio of Huertahuasi, said that “the rains used to be on-time and the hail and freeze used to be sparse. In June and July, they used to sometimes get snow. But now the rains are sporadic. Sometimes they get rains; sometimes they don’t,” he said. “Snow doesn’t reach the lower zones anymore,” he said, “only the high parts of Huertahuasi.”896 Victor Galindo Tucno, 99, said that now the rain is crazy, the heat intense, and the hail more frequent than in the old days.897 Belisario Galindo Conde, 96, has spent most of his life in the Rio Cachi barrio of Pampamarca. He said that the weather is less predictable now. “People knew the days when the freeze would come, as if it had an appointment; they also knew when the hail would come. And the heat was never too intense; rain was gentle. But now the rain is stronger and with lots of wind, and the heat is stronger and harder to take,” he said. “People knew when the snow would fall,” he said. “Now, there’s no snow, but there’s more hail. The snow didn’t harm the crops like the hail does now.”898 Villager after villager from both

894 M.E.D. Poore and C. Fries, The Ecological Effects of Eucalyptus, FAO Forestry Paper 59 (Rome: Food and Agricultural Organization, 1985), 54. 895 Poore and Fries, The Ecological Effects of Eucalyptus, 21. The FAO also notes on page 54 that eucalyptus will reduce the water yields in catchments when planted in areas without exiting trees, particularly when the trees are young and growing rapidly. Such impacts are particularly notable in the dry, cleared slopes the district’s Rio Pampas zone. 896 Interview, Melchor Quispe Nuñez, Huertahuasi, Quispillaccta, 31 October, 2004. 897 Interview, Victor Galindo Tucno, 99, Chuschi, 24 October, 2004. 898 Interview, Belisario Galindo Conde, Pampamarca, 29 October, 2004.

273

Chuschi and Quispillaccta commented on the change in the weather in recent decades – the absence of gentle, predictable rainfall between September and November essential to successfully sowing crops; the more frequent intense rainfalls and winds that damage crops and wash away soil from the fields; more intense heat in lower zones and less cold and snow in the high zones; and more frequent and unpredictable freezes and hail storms.899 Other studies have recorded similar observations from district informants.900 Recent scholarly research on observed and predicted climate change in the Andes has found that impacts are somewhat consistent with observed and predicted impacts from climate change in mountainous high-altitude zones like the D. Chuschi. One study of climate change in the Andes found temperatures had increased between 0.1 and 0.20C since 1969.901 Similarly, a recent review of climate change impacts in the Andes found an increase in average temperatures over the past 60 years and that temperatures in high-altitude zones have increased at a faster rate than sea-surface temperatures.902 The review found no clear trends in annual precipitation, however, suggesting that increasing temperatures are driving water stress impacts in many areas, as is evident with melting glaciers, for example.903 Scientists acknowledge that the significant variability of the Andean region, data

899 Policarpo Casavilca Quispe of Tuco, 63, offered other additional interesting comments: “In the old days, the rain came in September and October for the sowing. But now the regular rain doesn’t come until November. Now it can rain any time with much variability. There used to be lots of snow in the old days from Tuco to Llacctahurán (Quispillaccta’s western barrios). They used to make big snow balls.” He showed that snow balls around half a metre in diameter was normal. “But there is hardly any snow now; only in the high peaks. The heat is definitely greater today than in the old days when they were used to the cold. In the old days there was lots of freeze and they couldn’t plant anything in Tuco, only potatoes in Huertahuasi. But now they plant broad beans in Tuco. There’s also more hail now; and it’s much more variable than in the old days.” Interview, Policarpo Casavilca Quispe, Tuco, 5 November, 2004. 900 Asociación Bartolomé Aripaylla (ABA), Agricultura sostenible para la adaptación al cambio climático en regions vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, Ayacucho, September 2011: 12-13. 901 Imtiaz Rangwala & James R. Miller, “Climate change in mountains: a review of elevation-dependent warming and its possible causes,” Climatic Change 59(Issue 1 &2): 531. 902 Austrian Development Cooperation and the Swiss Agency for Development and Cooperation, Mountains and Climate Change: A Global Concern, Sustainable Mountain Development Series, (Switzerland: Centre for Development and Environment (CDE), Swiss Agency for Development and Cooperation (SDC) and Geographica Bernensia, 2014): 16-17. There is evidence from North America that global warming may result in more frequent damaging hail storms, as many community members reported - a seemingly counter-intuitive impact of global warming. I have not found studies about hail frequency and intensity in the Andes, however. Julian C. Brimelow et al., “The changing hail threat over North America in response to anthropogenic climate change,” Nature Climate Change 7(July 2017): 516-523; and John T. Allen, “Hail potential heating up,” Nature Climate Change 7(July 2017): 474-475. 903 Austrian Development Cooperation and the Swiss Agency for Development and Cooperation, Mountains and Climate Change, 16-17. Another study found declines in annual precipitation and rainy season precipitation in

274 limitations, and complexity in modeling climate impacts and projections in the area mean that large uncertainties remain in our understanding of the impacts and future path of climate change in the Andes.904

The limited historical scope of data and interruptions in data collection caused by the 1980s civil war leave an incomplete record of temperature and rainfall in the district. Records of average annual rainfall from a station situated near Chuschi’s town square show high variability but no clear decline in the 1960s and 1970s. The average of four stations in the upper Rio Cachi zone

southern Peru between 1950 and 1994. However, the authors thought that the impact was localized and probably not significant. Mathias Vuille et al., “20th Century Climate Change in the Tropical Andes: Observations and Model Results,” Climatic Change 59(Issue 1 &2): 81-2. 904 Wouter Buytaert, “Assessment of the Impacts of Climate Change on Mountain Hydrology: Development of a Methodology Through a Case Study in the Andes of Peru,” Mountain Research and Development, 32(3), 2012: 385- 386.

275

Figure 28:Average Annual Rainfall Chuschi Station and Rio Cachi Basin Stations (mm) 1400

1200

1000

800

600

400

200

Total Precipitation-Chushi Total Precipitation -Cachi Chuschi Mean Precipitation Cachi Mean Precipitation record somewhat higher annual average precipitation in the 1990s and 2000s, but also a great deal of variability from year to year, as noted above in graph, Figure 28: Average Annual Rainfall, Chuschi Station and Rio Cachi Basin Stations.

Rainfall data from the station near Chuschi’s town square show slight decreases in precipitation in October and November between 1963 and 1981, as illustrated in the graph below, Figure 29: Chuschi Station Precipitation, 1963-1981.905 This limited local data provides only partial support for observations from community members about changing weather in the district in the latter half of the twentieth century. Historical temperature data is unavailable for much of the

905 Data from Servicio Nacional de Meteorología E Hidrología. ONI chart from http://ggweather.com/enso/oni.htm, based on NOAA data.

276 district, unfortunately. Data from the Rio Cachi zone for the 1990s and 2000s does not conclusively show a rise in temperatures in that zone.906

Figure 29:

Precipitation data from Chuschi station, located a few hundred metres from Chuschi’s town square, shows slight declines in October and November precipitation (dotted lines, lower figure). If you remove the tremendously high Dec., 1981 rainfall data point, December’s trendline is also slightly downward. Important Niño events are overlaid, but do not correlate with major peaks or declines in rainfall. (Source: Data from Servicio Nacional de Meteorología E Hidrología. ONI chart from http://ggweather.com/enso/oni.htm, based on NOAA data.)

Climate change impacts are likely to have only begun to be felt in the early 1970s, when many households had already switched much of their economic activity to the Rio Cachi zone. So, the actual impact of changing weather patterns on household decisions to shift their focus to the higher zone may have been limited. However, the observed and recorded changes added to

906 Asociación Bartolomé Aripaylla (ABA), Agricultura sostenible para la adaptación al cambio climático en regions vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, Ayacucho, September 2011: 12.

277 existing strains for cultivators in the lower zones of the district, and probably helped cement household choices to centre their livelihood activities in the higher zones.

Declining agricultural yields and crop quality

Decreases in the quantity and quality of agricultural yields in the latter half of the twentieth century provide another piece of evidence of growing land degradation in the district in the decades before Sendero’s uprising. Declines in crop yields, Stocking and Murnaghan note, “may be an indicator that soil quality has changed, which in turn may indicate that soil and land degradation are occurring.”907 Interviews with dozens of district cultivators demonstrate that yields of maize and tuber crops like potatoes declined over the past half century as agricultural intensification and extensification spread through the cultivated zones. These developments further disadvantaged livelihood strategies focused on the lower cultivated zone in the district.

Researchers in the late 1960s and early 1970s reported low yields of key crops like maize and potatoes from the district. In 1966, IIP researchers interviewing selected residents recorded an average of just over 2.5 sacks/yugada (fanegas/yugada) of maize and 7.5 fanegas/yugada of potatoes.908 Figures from Peru’s 1972 agricultural census were even more worrisome, with average yields in the province of Cangallo at a little over 1 fanega/yugada of maize and about 1.7 fanegas/yugada of potatoes.909

However, we should be wary of seeing one or two years of low yields as proof of increasing supply scarcity for district households for several reasons. First, numerous scholars have noted that actual yields obtained from field measurement or careful interviews with cultivators are always higher than provincial, department, or national averages.910 Some of these differences may stem from methodological problems with government data collection. Mitchell says that the peasants he studied in Quinoa, Ayacucho feared the 1972 agricultural census as a prelude to land

907 Stocking and Murnaghan, Handbook, 11. 908 Ramón C. et al., La Comunidad de Chuschi, 34. Three district residents were asked yields in sacks per yugada. However, no information is provided to explain how these residents were chosen. 909 See table 48, Degregori, Ayacucho, raices de una crisis, 193-94. 910 Mayer, The Articulated Peasant, 20; Mitchell, Peasants, 77.

278 seizures or taxation, and systematically mislead census takers on land holdings, yields, etc.911 Similar dynamics may have also been at work in the D. Chuschi. Second, crop yields are highly variable in the Andes, given the heterogeneous environmental and cropping conditions.912 Morlon notes that “[r]isks, especially climatic ones, inevitably cause marked variations between years and, in the same year, between fields exposed to these risks to a greater or lesser extent.”913 To deal with risks, according to Morlon, farmers in Peru use two complementary strategies: 1) “reducing the risk level, and 2) spreading the risks as much as possible.”914 They reduce the level of risk by using different types of agricultural knowledge and techniques, like terracing or inter-cropping, while spreading the risks in time and space as much as possible through different allocations of household labour in various livelihood activities and by working different ecological zones.915 Attempting to manage risk inevitably results in highly variable yields, Morlon notes, with high yields some years and low yields other years: The result of this risk-spreading strategy is the multiplication of the number of farming combinations: cultivated sites (soil, climate, topography) multiplied by ways of working the soil, multiplied by species and varieties, multiplied by sowing dates etc., so that at least some of these combinations will be productive. Thus, the divergent yields (their dispersal in statistical terms) of the various fields cultivated by the same family result from the peasant strategy of risk spreading and from complementary use of different ecological environments. As the climate is unpredictable, the peasant cannot foresee which of the combinations will prove best … and, in fact, in any one year only a small proportion of these combinations prove to be "optimal" and produce high yields. Return from the other cultivated plots can be very low, and therefore the average yield, too.916

911 Mitchell, Peasants, 77. 912 Morlon, “Peasant strategies,” 8. Morlon writes that, “In the Andes of Peru and Bolivia, observations and measurements in the field reveal that yields are sometimes far higher than given in official statistics or obtained by quick surveys. Above all, they show the extreme variability in the yields of all crops, firstly, between years; secondly, between different categories of producers and even between producers within each category; and thirdly, between the different plots cultivated by one and the same family in different "production zones.” 913 Morlon, “Peasant strategies,” 8. 914 Morlon, “Peasant strategies,” 8. 915 Morlon, “Peasant strategies,” 8. 916 Morlon, “Peasant strategies,” 8.

279

Furthermore, some farmers may decide not to work some lands as intensely or thoroughly in order to redirect household assets toward other livelihood opportunities. Their actions may lead to falling yields and worsening land degradation because they ignore labour intensive land management practices like repairing terraces or expending time and energy to bring animal manure to fertilize distant fields. Mayer’s 1974 study of peasants in Tangor, in the Department of Pasco, showed that households decided to export labour to the cash economy, rather than intensify agricultural efforts for commodity production because off-farm work paid better.917 “Low returns on work effort compared to off-farm returns,” Mayer wrote, “lead to the abandonment of less productive lands in a country where agricultural land is scarce … The demand for outside money obtained from wages probably has a long-term negative impact on the ecosystem stability of Andean farm landscapes because they increasingly neglect maintenance work.”918 Yields thus reflect the vast range of high and low agricultural outcomes and reflect a range of possible livelihood and agro-ecological-climatic variables. They are not solely a function of land productivity and should not be considered solely a supply indicator – an indicator of the health of the resource, because they are actually a social-ecological outcome.

Agricultural yield data can still provide insights about the extent of land degradation and household impacts. To overcome the limitations discussed above, however, long-term yield trends and the range of socio-ecological factors behind those trends need to be determined. Ideally, yields should be tracked at the household level over time; however, decades-long local surveys are not practical and were not possible in the D. Chuschi. This study has instead supplemented yield data from IIP and Peru’s census with cultivator interviews about yield changes over time. Two complementary strategies were employed. First, interviews were conducted with dozes of elderly district residents, to determine how yields of maize and potatoes – the two most important crops in district households - have changed from the time of their youth to the present day. Responses have been plotted below in box and whisker plots for maize and potatoes to highlight the range of possible yields noted by respondents. Two groups of historical yield figures have been assembled for each graph, depending upon the age of informants, with informants over 70 years of age providing data before 1950, while younger respondents provided

917 Mayer, The Articulated Peasant, 26. 918 Mayer, The Articulated Peasant, 27.

280 yield insights for the period from the 1950s to the 1970s. Respondents were asked to compare historical yields to contemporary yields; these are plotted in the third category. Additional historical and contemporary yield data for the district are plotted on the graphs where data it is available. (For additional details about the boxplots used below, see Appendix 3: Notes on the Use of Box and Whisker Plots.)

Interviews with dozens of the district’s oldest residents comparing historical and contemporary yields of maize and potatoes demonstrate that yields have fallen for many over the past half century. From the 1930s to the 1970s, median corn yields fell by almost one quarter, before declining to less than half of historical yields in the present day, according to interviews (See Figure 30: Comparison of Historical and Contemporary Corn Yields in Sacks/Yugada in Quispillaccta and Chuschi According to Interviews). Potato yield declines were similar, with yields falling about a quarter from the 1930s to the 1970s, before falling in the contemporary period to about a quarter of the yields considered normal in the 1930s to 1950.919 (See Figure 30: Historical Potato Yields in Sacks/Yugada in Quispillaccta and Chuschi According to Interviews.) Interestingly, the continued decline in the 1950s to 1970s occurred concurrently with the introduction of Green Revolution technology to the district. These declining yields no doubt added additional livelihood stress on district households and pressure to find adaptive responses to challenges in cultivated agriculture. Given yield variability, these figures should also be read as an indicator that some households were probably also shifting effort to other livelihood activities, as outlined above.

919 Interview data have been ordered and organized into three time periods, as explained in Appendix 3. When a range of yields was given by informants, an average yield was calculated. Respondents were asked to provide yield figures in fanegas/yugada (sacks/1425m2). Maize yields were estimated as whole cobs and kernels. These are gross yields and do not account for seed maize and seed potatoes used in planting.

281

Figure 30

282

Figure 31

Memories can be faulty, however. So, comparisons of the interview data have been made with data collected by other researchers in the district, and by other studies in Ayacucho. This evidence suggests that the interview data on yields from the district is credible. Huamanga University agronomy professor Antonio Diaz Martinez, visiting the district in the late 1960s, found maize yields averaging less than 3.5 F/Y, and potato yields of about 20 F/Y.920 Along with the 1966 IIP yields noted above, researchers in the late 1960s were finding maize yields much lower than reported in this study’s field interviews, and potato yields within the range reported by district residents for the period. Some of the most careful yield studies were done by William Mitchell in the community of Quinoa, Ayacucho. Mitchell notes one farmer’s maize

920 Antonio Díaz Martínez, Ayacucho: hambre y esperanza (Ayacucho: Ediciones "Wamun Puma," 1969), 147. Díaz Martínez was radicalized in the years following his work in Ayacucho’s communities in the late 1960s. He later joined Sendero Luminoso and died in a prison riot in 1986. See Colin Harding, Antonio Díaz Martínez and the Ideology of Sendero Luminoso, Bulletin of Latin American Research, 7(1), 1988: 65-73.

283 yield in 1974 between 500 and 800kg, or a yield average of 650kg average - a little over 8.5 fanegas, which is very close to the median yield reported as normal by district informants between the 1950s and 1970s. Other informants Mitchell talked to said that this farmers yield was normal.921 Mitchell’s 1974 informant also harvested between 1000 and 2000kg of potatoes, working out to an average of 1500kg – or 20 fanegas, which is in the higher quartile of the yields in the district reported by my informants for this period.922 Based on these results and other surveys, Mitchell concluded that, in Quinoa, 1 yugada produces about 450-750kg of maize in a good year (6-10 fanegas), and between 1000-2000kg of potatoes (about 13-26.5 fanegas).923 These figures correspond very well to the median yield values reported by Chuschino and Quispillacctino informants for this period.

Field assessments in 2004 also support the interview estimates of contemporary maize yields in the district. A field belonging to the Quispillaccta barrio of Huertahuasi, located below the Matriz and facing the Rio Pampas, was being planted with maize during our visit in October, 2004. This field’s yield that season was promised to Quispillaccta’s Ancianos organization, to help supplement to the food supply of the community’s elderly residents. Multiple informants noted during planting that day that this field today produces about 4-6 F/Y of maize with cobs, while in decades past it produced almost 14 F/Y.924 Assessments of two Chuschino fields that are now worked collectively in order to provide maize for Catholic feast celebrations also supports contemporary interview estimates. The Dulce Niño Jesus field, in Chuschi’s maize zone fronting the Rio Pampas, produces about 12 F/Y of maize, at the higher end of contemporary maize estimates.925 Chuschi’s Santos Temblor field in the Rio Pampas maize zone (pictured above in chapter 4 as Chuschinos collectively planted it with maize) used to give about 10 F/Y of maize, but now yields less than half, about 4.5 F/Y. Again, these estimates closely

921 Mitchell, Peasants, 79. Mitchell more carefully surveyed yield figures, including determining the weight of seed maize used in the plot. This figure above is the average of the yield range cited by this farmer, before deducting seed weight. The figure is the total dried and husked cob weight – maize with cob. 922 Mitchell, Peasants, 79. 923 Mitchell, Peasants, 79. Mitchell notes that these figures do not account for post-harvest food losses, which have been calculated at about 15% of production for developing countries. 924 Interviews and field measures of Ancianos chakra, Quispillaccta, 20 October, 2004. This field was estimated at about ½ yugada, so I have converted the yield estimates to 1 yugada equivalents. 925 Interviews and field measures of Dulce Niño Jesus chacra, Chuschi, 30 October, 2004.

284 align with contemporary interview data. Gualberto Machaca’s study of soil erosion in the Rio Pampas zone of Quispillaccta in the early 1990s, however, reported potato yields much higher than those reported by this study’s informants for the contemporary period. His reported yields are comparable to yields reported by informants in the mid-20th century.926 Reconciling Machaca’s data with interview data about contemporary potato yields is difficult. It is possible that with labour requirements higher for potatoes than maize, the lower averages reported by elderly informants may reflect the challenges older residents have in planting potato fields. We know little about the cropping history of the fields Machaca measured; they may have also benefited from being fallowed when many residents abandoned the district during the 1980s civil war.927 Between 1990 and 1998, the Ayacucho average of potato production was about 13.5F/Y, higher yields than reported by informants in the district, but on the lower end of Machaca’s yield data. 928 Again, these different data reports may be reporting a range of possible potato yield outcomes from the district. Machaca’s data aside, however, the evidence assembled from field assessments in the district and scholarly assessments in the 1960s and 1970s appear to confirm the trend of falling yields of maize and potatoes in the 20th century, as identified by elderly informants in the district.

926 Gualberto Machaca, Conservación y Recuperación de Suelos, 109. 927 During fieldwork, we came across farmers working fields in the Rio Pampas area that had been abandoned for decades. We watched as Abraham Vilca Hachalma sowed tubers in a field opposite Llacctahurán in Lucerohaha, fronting the Rio Pampas, which he cultivated in 2003 for the first time in about 50 years. This small, high field, little more than 1/3 of a yugada produced 3 fanegas of tubers last year – or about 9 fanegas/yugada. That yield is higher than most of the interview yields provided by elderly residents and more in line with lower yields reported by Gualberto Machaca. So, it is reasonable to assume that higher yields are possible today in well-managed or well- rested fields in the Rio Pampas zone. Interview and field visit, Abraham Vilca Hachalma, Llacctahurán, 6 October, 2004. 928 Rolando Egúsquiza B., La papa: producción, transformación y comercialización (Lima: Universidad Nacional Agraria la Molina, 2000: 9.

285

Falling yields over many decades went hand-in-hand with gradual declines in the quality of crop harvests. Declining crop quality – smaller and more diseased potatoes and maize – is another indication of increasing supply stress on land assets of households in the district. The consensus was nearly universal among respondents in both Chuschi and Quispillaccta that crop quality today is much worse than in decades past, with size, taste, and crop quality all having declined. Elderly Chuschinos in their seniors (Anciano) group noted that corn cobs are much smaller today than they used to be, forcing them to often buy corn from other communities to make chicha – corn beer.929 In an interview, Pablo Conde of Quispillaccta pulled out a cob from the previous year’s harvest to show a typical example of the best cob size today, about 11cm. This size is after adding artificial and natural fertilizers like guano, which wasn’t necessary in decades past. Now they harvest lots of rotten corn, Conde noted, so they 930 Diseased potatoes infected by the potato tuber moth give it to their chickens. Quispillacctino Roberto (Phthorimaea operculella); informants say this is typical of the quality problems farmers confront today in their Huaman Machaca thought that the size of corn cobs yields. Photo by author. today is about ½ to ¾ of historical corn cobs.931 Basilio Chipana Nuñez, 90 years old when interviewed, also agreed that cobs in decades past used to be much bigger with huge grains. He noted that some varieties have disappeared from the community because they don’t produce well anymore.932 Pablo Conde noted that potatoes plants used to be so big that they could wander around a field and almost get lost among the leaves. Large potatoes were much more frequent in decades past, with some the size of softballs. However, they never see potatoes that size anymore.933 Today their harvest is also more

929 Interview with representatives of Chuschino Ancianos group, Chuschi, 15 October, 2004. 930 Interview, Pablo Conde, 70, Quispillaccta, October 25, 2004. 931 Interview, Roberto Huaman Machaca, Quispillaccta, 1 November, 2004. 932 Interview, Basilio Chipana Nuñez, 90, Chuschi, 2 November, 2004. 933 Several farmers noted that softball-sized potatoes were common in their youth, compared to their absence today. Interview, Pablo Conde, 70, Quispillaccta, October 25, 2004; Interview, Roberto Huaman Machaca, Quispillaccta, 1 November, 2004; Interview, Victor Calderon Jimenez, Chuschi, 2 November, 2004.

286 frequently attacked by pests and diseases or the potatoes are hollow.934 Farmers have to take care to pick out the wormy potatoes or their entire harvest will be infected in storage.935 The taste is also inferior compared to decades past, according to numerous informants. Some said that they often taste like green potatoes or garlic now when harvested. However, Salomón Galindo, a well-known farmer from Cuchoquesera, did not think that the crop taste had changed. He did say that people are now eating different varieties of potatoes, and that the older varieties of potatoes used to be sweeter. 936 Galindo also observed that native varieties produce little today, compared to hybrid varieties. While the quality of native crop varieties is often better, he thought that most people today chose quantity over quality.937

The causes of crop quality declines are complex, but there is evidence that both agricultural intensification and the adaptations that intensification spawned may have worsened crop quality for some households. Reductions in crop rotation and fallowing described above encourages pest growth, particularly of nematodes and Potato Weevils. So, production pressures to keep land in use as land became less available over the twentieth century may have had damaging impacts on crop yields and soil health.938 The disappearance of sectoral fallow systems of cropping in the district may have further worsened pest infestations because the move to individual household decision-making about when or if to fallow a field has consequences for nearby fields. If a household chooses not to fallow, pests in their fields will likely to migrate to their neighbour’s field, even if that field has been fallowed. Sectoral fallow systems probably increased the efficiency of pest control by ensuring comprehensive pest management across agricultural sectors in the community. Thus, the disappearance of these systems likely worsened

934 Asociación Bartolomé Aripaylla (ABA), Agricultura sostenible para la adaptación al cambio climático en regions vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, Ayacucho, September 2011: 13. 935 Interview, Pablo Conde, 70, Quispillaccta, October 25, 2004. Andean Potato Weevils and Potato Tuber Moths are the most common pests attacking potato crops in the district, according to informants. 936 Interview, Pablo Conde, 70, Quispillaccta, October 25, 2004; interview, Victor Esteban Caillahua Mikullya, 80, Chuschi, 24 October, 2004; Interview, Salomón Galindo Achallma, 77, Cuchoquesera, 28 October, 2004. 937 Interview, Salomón Galindo Achallma, 77, Cuchoquesera, 28 October, 2004. Galindo worked with Swiss agricultural extension workers in the late 1960s and in the 1970s on various Ayacucho government programs to improve native pastures and livestock practices. He is photographed in a history of Swiss agricultural extension efforts in Quispillaccta in a recent history of Swiss development efforts in Peru. Burri, Los Inicios, 88. 938 Andean Potato Weevil, Integrated Crop Management Toolbox, CIP, < https://research.cip.cgiar.org/confluence/display/icmtoolbox/Management+APW>, accessed 1 Nov., 2018.

287 pest damage to crops in the district.939 Marcela and Gualberto Machaca and their colleagues at ABA have also made convincing arguments that the loss of traditional cultivator knowledge – both in cropping practices but also the loss of traditional crop biodiversity – have had negative impacts on the ability of households to manage pest damage to crops.940 Though communities in the district remain reservoirs of high crop biodiversity compared to other highland communities,941 the increasing use of hybrid and introduced monoculture crops after the 1960s in the district may have also increased pest damage to crops.942 Finally, while the introduction of pesticides and artificial fertilizers likely helped to increase yields for many, problems with increasing pesticide resistance have been noted by informants in the district and may have eroded yield gains for some.943 Pablo Conde of Cuchoquesera said that he was introduced to pesticide use by Swiss agricultural extension workers in the early 1960s. At first, Conde said, they only needed a little pesticide. But now they use it and “the bugs are like friends with the pesticide.”944 Declines in crop quality have thus exacerbated yield declines faced by many households in the district.

Conclusion

In the end, the assembled evidence suggests that many cultivators increasingly faced supply and demand constraints of productive assets in the latter decades of the 20th century, particularly in the Rio Pampas zone of the district where several different indicators of land degradation

939 Orlove and Godoy, “Sectoral Fallowing Systems,” 180. The authors note that controlling potato pests does not necessarily require the fields to be left completely unused, because planting non-tuber crops like grains in a field also effectively acts to reduce tuber pests. So, non-tuber plantings are like fallow years when it comes to tuber pest management practices. 940 Gualberto Machaca, Conservación y Recuperación de Suelos, 68-9; Machaca, “Vigencia y Continuidad,” 163-4; ABA), Agricultura sostenible para la adaptación al cambio climático en regions vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, 45. The causes of losses in cultivator knowledge are complex, but include the impacts of migration, cultural discrimination and racism against indigenous knowledge, the breakdown of communal agricultural management systems, etc. 941 Coronel et al., Estudio Comparativo, 14-15, note Quispillaccta’s relatively high crop biodiversity. 942 The relationship between increasing crop monoculture and biodiversity loss and rising pest damage is well established. See Miguel A. Altieri et al., Agroecology: The Science of Sustainable Agriculture (Boulder, CO: Westview Press, 1995), 270-1. 943 This is a global pattern noted by experts promoting “ecoagriculture” approaches to farming. Jeffrey A. McNeely and Sara J. Scherr, Ecoagriculture: Strategies to Feed the World and Save Biodiversity (Washington D.C.: Island Press, 2003), 70. 944 Interview, Pablo Conde, 70, Quispillaccta, October 25, 2004.

288 developed in the latter half of the twentieth century. Land degradation is “an ‘umbrella’ term” describing various ways “in which the quality and productivity of the land may diminish from the point of view of the land user (and of society at large).”945 Cultivated agriculture in the Rio Pampas area of the district is inherently variable, and over several decades households faced resource degradation as a result of the consequences of existing steep slopes, thin soils, water shortages, extensification and intensification impacts during the twentieth century, and differences in household capabilities. Diminishing asset holdings as populations increased and the risks that households faced in their production activities in this zone resulted in many instead refocusing assets and capabilities to the higher zones and the Rio Cachi watershed areas of the district. Off-farm opportunities and adaption options reinforced these trends. Supply and demand pressures thus helped to stimulate the increasing settlement of district residents in the disputed high-zones.

Importantly, however, we should note that supply and demand pressures were not solely responsible for driving district residents to shift effort to higher zones. Wider structural changes in Peru’s economy were also crucially important. A complex calculation of opportunities, risks, and costs convinced many households to put less effort into cultivated agriculture in the Rio Pampas zone. Many others, however, continued to work these lands and adequately provide for their families. As notable farmer Theodosio Flores Galindo told me on our final walk back to the Matriz from Llacctahurán, past many chacras (fields) that were over-grown and unused, the lower Rio Pampas zone chacras were not abandoned because the land was poor. With proper effort and knowledge, these lands can be very productive. Some have been abandoned for almost 50 years, while others only a few years, he said. Many have been abandoned for 20-30 years. The reasons ranged widely, according to Don Theodosio. Some parcels were too close to the footpath and suffered too much animal damage. Others were owned by older residents who lived in distant barrios; working them wasn’t worth their effort. While other parcels were owned by recently deceased comuneros whose kids lived in distant cities.946 Don Theodosio was well aware of my inquiries with his fellow community members about increasing land scarcity and land degradation pressures. While not blind to these impacts, he was also confident in his ability

945 Stocking and Murnaghan, Handbook, 6-7. 946 Interview, Theodosio Flores Galindo, Llacctahurán, 7 November, 2004.

289 to work the land successfully, given enough assistance and his deep knowledge of traditional agricultural practices.947

In many ways, Don Theodosio’s perspective epitomized the conflict between Neo-Malthusian pressures and Cornucopian responses in the district. Adaptation and coping with risks and pressures was certainly possible, in Don Theodosio’s view. Other households, however, faced the same pressures and instead chose different paths for adaptation that reduced their reliance on cultivated agriculture in the Rio Pampas zone. Depending on the assets and capabilities at hand for households, various livelihood pathways were possible. Don Theodosio’s wisdom and the record of supply and demand pressures outlined above on the district’s Rio Pampas cultivated fields, confirm that binary arguments between Neo-Malthusians and Cornucopians around the impacts of environmental scarcities over-simplify livelihood realities. Both outcomes are possible, and some outcomes are in the messy middle; household assets and capabilities can lead to a range of possible pathways with different impacts.948 Today, many households continue to devote a great deal of energy to cultivated agriculture in the Rio Pampas zone. However, the shift by many others to settle the Rio Cachi zone of the district in the latter half of the twentieth century was certainly influenced by supply and demand impacts on household assets over many decades. This settlement aggravated conflicts between Quispillaccta and Chuschi in the latter decades of the twentieth century, with particularly severe consequences in the early 1980s, as discussed in Chapter 8.

947 Don Theodosio was also a relatively well-off farmer, with more cultivated lands than the average household in Quispillaccta. So, his cultivator optimism was also based, in part, on his abundance of household assets. 948 This theoretical debate in the environment-conflict literature and my conclusions about this debate are covered in more detail in Deligiannis, “The Evolution of Qualitative Environment-Conflict Research,” 44-51.

290

Chapter 8 Living Between the Sword and the Stone Introduction

In early 1983, the ink was barely dry on the historic agreement to settle centuries of land conflict between Chuschi and Quispillaccta when the outbreak of violence in the district soon led to Quispillaccta being labelled as a hotbed of support for Sendero Luminoso. Other long-running conflicts with Chuschi’s annex Cancha Cancha and with neighbouring communities in the northeast of the district had also only recently ended. Reconstructing the history of the early months of the violence in the district strongly suggest that the terrible violence in early 1983 cannot be decoupled from the outcome of historical land conflicts in the district. Decades of efforts to purchase and settle disputed lands in the Rio Cachi zone, described in Chapters 4, 5, and 6, led to increasing conflicts between Quispillaccta and neighbouring communities, as each sought to consolidate control over disputed lands. Supply and demand pressures on household livelihoods in the Rio Pampas zone of the district, described in Chapter 7, led to livelihood adaptations that increased settlement of disputed lands in the district, further heightening conflicts between Quispillaccta and other communities. Despite the efforts of community leaders in Quispillaccta and Chuschi to settle their long-running land conflict differences in the early 1980s, these conflicts re-emerged in the orgy of violence undertaken by the Shining Path and Peru’s security forces in 1983 and 1984.

This chapter begins by outlining the dynamics of violence in the district in late 1982 and early 1983, as Sendero began to kill those in the district who did not conform to their demands for obedience and support for the new Maoist project. As in other parts of Cangallo, authorities reluctant to commit to Sendero were the new targets of deadly retribution. These actions, however, were soon followed by a vicious Peruvian government counter-insurgency operation in the area, which sought to root out and destroy Sendero militants and their supporters. The following account illustrates that this violence was not indiscriminate, but targeted with particular harshness against Quispillacctinos, particularly in the recently consolidated Rio Cachi zone. Evidence is presented that suggests this targeted violence was in retribution for Quispillaccta’s land struggle victories. The chapter then assesses scholarly explanations of violence in the district to argue that prevailing interpretations have failed to adequately

291 appreciate the extent to which Quispillaccta was unfairly victimised for their land struggle successes with neighbouring communities. Certainly, every community suffered terribly attempting to find safe space between Sendero’s stone and the Army’s sword. But the extent of Quispillaccta’s suffering suggests that particular revenge dynamics were conditioning violence patterns during the civil war.

Violent revenge and the Peruvian Government’s counter- insurgency campaign in 1983

After years of clandestine efforts in Ayacucho to indoctrinate residents and recruit community members into Sendero’s cause, the insurgent group turned more forceful in establishing control of the region in late 1981 and 1982.949 With few police or government officials in the D. Chuschi, it was relatively easy for Sendero to force the closure of the district’s municipal offices in October 1981. Surrounding districts were similarly heavily infiltrated by Sendero and there was little government presence in much of the province of Cangallo at the end of 1981.950 After executing local cattle thieves and whipping delinquents in the D. Chuschi in 1981 and 1982, as described above, in late 1982, Sendero more forcefully moved to seize control of the district and eliminate local authorities reluctant to commit to the insurgency cause.951 On July 1st, 1982, armed Sendero militants moved to take control of the town centres in the district. They blew up the post office in Cancha Cancha, sacked the Artisanal House in Chuschi, beat anyone who resisted, and threatened to kill Chuschi’s Governor, only to be convinced to spare him in the face of community opposition.952 One month later, on August 2nd, a Sendero column and hundreds of

949 CRV, Inform, Volume 7, Section 2.2, 39-40. 950 Isbell, To Defend, 84-5. 951 CRV, Inform, Volume 7, Section 2.4, 35. Sendero’s first incursions into Cancha Cancha came in 1982, according to Peru’s TRC, when Sendero warned authorities to yield to Sendero and they whipped three residents for committing offenses against the community. La Serna’s account of the fatal punishment of thieves and delinquents in the district – something generally accepted by all the communities in the district - is the most comprehensive and is generally corroborated by other scholars like Sánchez Villagómez, Isbell, and my own interviews. It does not suffer from the speculative failings of his claims about greater Sendero support in Quispillaccta. See La Serna, 2012, Chapter 4. 952 Isbell, To Defend, 85. Sánchez Villagómez, "El Horror Olvidado,” 277- 282. A Quispillacctino communal authority at that time said that the Sendero militants were outsiders, unknown to the authorities, according to an interview with Sánchez Villagómez. Sánchez Villagómez notes that three district officials were captured by Sendero that day: Deputy Mayor Francisco Vilca; district Governor Bernardo Chipana; and Ramón Infanzón. All three were Chuschinos. Interestingly, according to informants, Infanzón comes from a family identified by informants in the communities as part of the district’s Mistis. Chuschi’s district authorities were targeted because

292 peasants from nearby communities looted and destroyed much of Huamanga University’s Allpachaka rural development research station, located a short distance from the contested lands in the north-east of the district.953 In early December, 1982, according to Isbell, Sendero organized a celebration in Cancha Cancha to mark the birth of the Popular Army. Participants arrived from all the surrounding communities, according to information obtained by Isbell, and five prisoners were put on public trial by Sendero, two from Cancha Cancha, and three from Chuschi. All were petty bureaucrats accused of abusing their positions of authority. Among those tried and later driven from the district was Chuschi’s Governor.954 By early 1983, Sendero planned to implement their third military plan, “Conquering Bases of Support,” between May 1983 and February 1984, with operations that would deepen the policy of destruction of local structures of state or traditional power and replace them with commissars appointed by the party.955 Now Sendero was prepared to use deadly violence against any remaining authorities or community members who opposed Sendero’s control of the district.

Conflict with district authorities and around Cancha Cancha would prove to be the key that unleashed terrible violence in the district from both Sendero and Peru’s military forces, with two Sendero incursions into the area in early 1983 leading to brutal Peruvian military counter- insurgency operations in May. Cancha Cancha’s reluctance to yield to Sendero, bolstered by the presence of a Peruvian military veteran who was an authority in the community at that time, made it a lynch-pin for killings in the district once the military counter-insurgency operations began. In late February 1983, Sendero militants returned to Cancha Cancha during the Catholic carnival celebrations; however, community members expelled the Sendero militants. Recognizing that their actions against Sendero increased risks for community members,

district Governor Chipana removed Sendero flags that had been put up in the town square a few days before. The removal of the flags was witnessed by a Chuschi school teacher, who was also a secret Sendero supporter, thus leading to Chipana’s targeting by Sendero, according to Sánchez Villagómez. Pio Taquire, a caretaker of the Artisanal House and Chuschi’s school was also beaten and tied up by Sendero for resisting their attempts to trash the Artisanal House and distribute the House’s tools to the community. Taquire had been a previous target of Sendero in Chuschi, condemned to wear a sign around his neck for philandering and wife beating, sometime before 1982. Sánchez Villagómez, "El Horror Olvidado,” 279-281. 953 Isbell, “Shining Path and Peasant Responses,” 85; Smith, "Rural Development in the Crossfire,” 38-9. Allpachaka is a four hour walk from Cuchoquesera, according to Núñez Espinoza, Online history of Cuchoquesera, , accessed Sept. 16, 2019. 954 Isbell, “Shining Path and Peasant Responses,” 86; La Serna, The Corner of the Living, 155. 955 CRV, Inform, Volume 7, Section 2.4, 35.

293 especially given the lack of police and military presence in Cancha Cancha, the community organized a peasant surveillance and patrol system to warn against further Sendero incursions.956 This defiance by authorities toward Sendero was increasingly being met with savage violence in early April in Ayacucho. On April 3rd, a large Sendero column massacred 67 comuneros in the community of Santiago de Lucanamarca, in Huancasancos province, several hours south of Chuschi. The killings in Lucanamarca were disciplinary revenge by Sendero against the community, which had formed a militia to oppose Sendero and killed the local Sendero commander in late March.957 In the D. Chuschi, Sendero again targeted those who dared to defy its military plan. On April 8th, a large Sendero column again entered Chuschi’s town centre, killing the traditional mayor (alcalde vara) Juan Cayllahua Tucno and several other community members, and burning several government buildings.958

Peru’s military soon responded to violence and threats against district authorities with brutal repression. Peru’s security forces entered the D. Chuschi in late 1982, initiating their counter- insurgency efforts in the area. In early April, 1983, Peru’s military established an Army base in the nearby District of Totos.959 Soldiers based at the Totos base would be responsible for much of the counter-insurgency bloodshed in the D. Chuschi.960

May would turn out to be the bloodiest month in the district during the Sendero insurgency as Peru’s military entered the conflict unleashing several waves of massacres – primarily of Quispillacctinos – during repeated sweeps of the district to search for Sendero militants.961

956 CRV, Inform, Volume 7, Section 2.4, 35. 957 Starn and La Serna, The Shining Path, 149-153 958 Sánchez Villagómez, "El Horror Olvidado,” 284-85. Peru’s TRC says that Sendero’s attack took place on April 8, leading to the death of Chuschi’s Governor and four peasants. CRV, Inform, Volume 7.2, Section 2.5, 39. Sánchez Villagómez says that the armed Sendero militants were accompanied by hundreds of peasants. However, we don’t know anything about who these peasants were or where they came from. This is an important historical gap in the story of the back and forth violence in the district in early 1983. 959 CRV, Inform, Volume 7, Section 2.8, 66. 960 CRV, Inform, Volume 7.2, Section 2.5, 39. 961 Testimony before Peru’s TRC and from survivors indicates that these Peruvian military operations in May were led by Captain Chacal (Jackal) – later identified by Peru’s TRC as Captain Santiago Alberto Picón Pesantes. Peruvian Truth and Reconciliation Commission Report, “Conference by Doctor Salomón Lerner Febres, President of the Truth and Reconciliation Commission, at the submission of a complaint on the death of four peasants in the locality of Totos, Province of Cangallo, Department of Ayacucho in April, 1983,” http://www.cverdad.org.pe/ingles/informacion/discursos/en_conferencias02.php, accessed 16 Sept., 2019. Decades

294

Quispillaccta’s high-altitude barrios were targeted in these sweeps - operations led by the Peruvian military, accompanied by dozens of residents of Chuschi and its associated annexes like Cancha Cancha. On May 14, 1983, an army patrol entered the Quispillaccta hamlet of Yuraqcruz, located in the centre of the district, in the hills above the main town. The group consisted of ten or fifteen soldiers, escorted by a group of Chuschi community members, pursuing a Sendero column that had carried out actions against Cancha Cancha. In Yuraqcruz, the soldiers forced the residents they encountered to lie on the ground while they checked their identification. Chuschi community members took advantage of the opportunity to loot the Yuraqcruz homes. Nine residents of Yuraqcruz were bound and led away. The next day, eight were later slaughtered by the military near Uchuryi, while one prisoner escaped by jumping into a ravine.962 Located in the puna above the main town, Yuraqcruz was one of the areas successfully seized by Quispillaccta in the land disputes with Chuschi in recent decades.

Sendero’s response a few days later again led to tragedy for Quispillaccta. On May 19th militants entered Quispillaccta’s Rio Cachi barrio of Cuchoquesera and press-ganged several community members into their plans to again attack Cancha Cancha. Sendero militants operating in the area forced community members from Cuchoquesera to participate in their actions, according to survivors.963 The next day, eight of the Cuchoquesera residents managed to flee from Sendero

later, investigators in Peru obtained an arrest warrant for the former Army Captain, to answer for his crimes in the Districts of Chuschi and Totos. Despite attempts by Peruvian legal NGOs to locate him, Picón Pesantes has fled justice. This is not unusual, but part of a pattern of former Peruvian military human rights abusers disappearing in the face of attempts to bring them to justice or escaping conviction. See Glatzer Tuesta, “Veinte años de impunidada y barbarie: Tras los pasos del Chacal,” Ideele 153(March 2003), http://www.idl.org.pe/idlrev/revistas/153/153chacal.pdf, accessed 16 Sept., 2019. See also: https://elperuano.pe/gespoboletinfiles/2018/12/19/1721524_1.pdf, accessed 16 Sept., 2019. 962 CRV, Inform, Volume 7.2, Section 2.5, 39-40. The captured comuneros were Narciso Achallma Capcha, Antonio Carhuapoma Conde, Valentín Núñez Flores, Julián Núñez Mendoza, Pedro Núñez Pacotaype, Reynaldo Núñez Pacotaype, Hilario Núñez Quispe, Máximo Vilca Ccallocunto, and another person, whose name is kept in reserve. According to Peru’s TRC, the prisoners were being walked to the provincial capital of Pampa Cangallo when the massacre took place. On route, the military let three other non-Quispillacctino prisoners continue under guard to the provincial capital but stopped on the road to torture the Quispillactino residents for information about Sendero. During the torture, one prisoner leaped into a ravine to escape, leading the soldiers to shoot the remaining 8 Quispillacctino prisoners. The military’s actions clearly indicated that they were targeting residents of Quispillaccta as suspected Sendero militants and sympathizers. The TRC’s account gives no details about the Sendero actions in Cancha Cancha that the military was responding to with their May 14th operations. 963 Salvatierra et al., La Vida, 275-6. That same day, Sendero killed four other residents of Quispillaccta’s Rio Cachi barrios for various reasons, including being opposed to Sendero’s agenda. See Salvatierra et al., La Vida, 246 and 270.The militants forced authorities in Cuchoquesera to cook for them and to kill a special Swiss cow that was being raised by the community livestock development organization, which had been established with Swiss aid, according to Salvatierra, Méndez, and Núñez. It is not clear how many Cuchoqueseran residents were forced into

295 when the militants came under pressure from a combined force of Peruvian military soldiers and Cancha Cancha community members. However, the escapees were then captured by residents of Cancha Cancha and handed over to the military, only to be later killed and buried in a secret grave.964 On May 21st, approximately 100 Sendero militants from platoon number 9 of Sendero’s Cangallo-Victor Fajardo Zonal Committee stormed back into Cancha Cancha, shooting, detonating explosives, and lighting houses on fire near the village square. Four community members were brutally killed by Sendero, including one community member who later died of his wounds.965 Two days later, on their way to the nearby provincial capital of Pampa Cangallo to denounce the attack, several community members from Cancha Cancha were surprised by a group of Sendero militants in the neighbouring town of Pomabamba. Cancha Cancha’s President, Indalecio Conde Quispe, was captured and killed.966 Outraged by the slaughter of their community president, Cancha Cancha residents took off in pursuit of the Pomabamba Sendero militants, catching up with them in Quispillaccta’s barrio of Cuchoquesera. In the ensuing melee, one Canch Cancha community member was killed - Albino Tacuri Condori – and several Senderistas were captured and handed over to the Peruvian police.967 Their fate remains unknown.

The Peruvian military’s response between May 28th and May 31st unleashed waves of killings and property destruction in Quispillaccta’s high puna barrios, including the Sakaypata massacre of 15 peasants in early June whose survivor testimony is detailed in the preface. Dozens of Quispillacctinos were captured, beaten, and some killed during the operations or soon after, as soldiers accompanied by community members from Chuschi and its annexes swept

the Cancha Cancha attack. Salvatierra, Méndez, and Núñez say that about 15 people were forced to participate; however, only 8 were later captured and killed by Peru’s military and community members in Cancha Cancha. Sendero frequently forced peasants into working and fighting for them in the early years of the insurgency in Cangallo. Heilman notes similar activities in the District of Carhuanca, which was then in eastern Cangallo. Heilman, Before the Shining Path, 189. 964 Paz y Esperanza, Boletín, Noviembre 2012, < http://www.pazyesperanza.org/pe/>, accessed 22 May, 2019. 965 The community member who later died of his wounds from the Sendero attack on May 21st was a former military soldier, Valentín Quispe Achas. Quispe was one of the organizers of Cancha Cancha’s resistance to Sendero’s incursions over the past year, according to Peru’s TRC. He seems to have been targeted by Sendero in this incursion, because masked insurgents stormed his house as soon as they entered Cancha Cancha. The other Cancha Cancha residents killed that day by Sendero were: Mariano Conde Cancho, Jesus Labio Conde, and Cirilo Achas Quispe. CRV, Inform, Volume 7, section 2.4, 36-37. 966 CRV, Inform, Volume 7, section 2.4, 36-7. Sánchez Villagómez, "El Horror Olvidado,” 292. 967 CRV, Inform, Volume 7, section 2.4, 37.

296

Quispillaccta’s settlements of Cuchoquesera, Pampamarca, Catalinayuq, Unión Portrero, and Llacctahurán .968 Community members from Chuschi and its annexes that accompanied the government soldiers looted and destroyed Quispillacctino houses as they went. Survivor testimony strongly suggests that the peasant groups accompanying the military on these sweeps took an active role in the looting and killing, seeking revenge on Quispillaccta. In some cases, community authorities were targeted in the attacks, authorities who had played a role in the recent land conflicts in the district. In other cases, those captured seemed to be random Quispillacctinos, in the wrong place at the wrong time.

Recent survivor testimony of these sweeps highlight the targeting of Quispillaccta. When soldiers arrived in late May to Cuchoquesera, they burned the communal store and the homes of several families.969 Salomón Galindo, who’s home was burned, was a well-known community development leader in the area. For several decades he had been working with Swiss NGO representatives and later with experts from the Allpachaka research station to improve pastures, herds, community development opportunities in and around Cuchoquesera.970 He had been singled out by Cancha Cancha in a dispute from 1980 for helping to organize Quispillacctinos from Cuchoquesera and Pampamarca to seize contested puna lands.971 Galindo would successfully evade both the military and their allies and the Senderistas during the months of killing in the district.972 However, fellow Cuchoquesera resident and authority, Francisco Núñez

968 Sources differ on the exact number of Quispillactinos caught up in the late May sweep. In recounting the details of the Sakaypata massacre, Peru’s TRC says that about 100 soldiers and community members from Chuschi and Cancha Cancha captured at least 16 Quispillactinos, with 15 later killed at Sakaypata. CRV, Inform, Volume 7, section 2.4, 30. Salvatierra, Méndez, and Núñez, says that 30 peasants were detained in the multi-day operation, eight from Pampamarca, eight from Unión Potrero, twelve from Catalinayuq, and two from the neighbouring district community of Putaqa. Salvatierra et al., La Vida, 214. In my interview with the survivor of the Sakaypata massacre, he noted that many of those captured were not necessarily captured in their home barrios, but instead were captured in various places in Quispillaccta – either on the way to or from agricultural work (like him) or on their way home after attending to community business in the town centre. 969 Salvatierra et al., La Vida, 265. 970 Burri, Los Inicios, 88; Oseas Núñez Espinoza, Online history of Cuchoquesera, , accessed Sept. 16, 2019; Interview, Salomón Galindo Achallma, 77, Cuchoquesera, 28 October, 2004. 971 Oseas Núñez Espinoza, Online history of Cuchoquesera, , accessed Sept. 16, 2019. The lands Quispillaccta invaded that day in 1980 was around Pukaramunwayqu, according to Núñez Espinoza. I have not been able to locate this on a topographical map. 972 Galindo was also targeted by Sendero because he was the earliest and the most successful proselytizer of the Evangelicalism in the district, and because he had worked extensively in the mid-1970s with Swiss development experts and with experts from the Allpachaka research station. Galindo told me that he started getting harassed by

297

Vilca, was not as fortunate. Núñez Vilca’s widow Alejandra says that her husband was targeted by community members from Cancha Cancha during the political violence in 1983 because he was a community lieutenant during the trial with Cancha Cancha in the early 1980s.973 Forced to again assume an authority position by the military who entered Cuchoquesera in late May, Núñez Vilca was detained by soldiers along with fellow Cuchoquesera authority Marcelino Espinoza Núñez, after being fingered by three peasants dressed as soldiers. Both were killed in the Sankaypata massacre.974 Demetrio Galindo Rocha from Pampamarca was another Quispillaccta peasant who had long worked with NGOs in the area to improve pastures and animal husbandry practices. He was also killed at Sakaypata, after being captured on his way home from the Matriz on official duties, during the military’s late May operations in the high puna.975

Sweeps by soldiers accompanied by peasants from Chuschi and its annexes also swept up seemingly innocent Quispillacctinos in the drive for retribution. Marcelina Quispe de Espinoza, whose husband would be captured and die in the Sankaypata massacre watched a peasant mob kill an elderly Quispillacctino in Puncapata in late May: “That morning they killed Jerónimo Vilca, an elderly gentleman. I went out with my four children and we hid …I watched as people passed with knife-pointed sticks, others with machetes and axes A whole crowd passed in front of us. Don Jerónimo had hidden in the middle of the ichu [grass] below the road. When he saw that people approached the place where he was, he jumped and started running desperately. Someone reached him, took out his ax and gave him an ax in the head and threw him [down]. So Jerónimo Vilca died.”976 Cirilo Galindo Huamaní of Pampamarca was returning to his home on May 28th with his young, twelve-year-old brother-in-law, after attending a military-required assembly in the Chuschi town centre. At Kimsacruz, the pair were captured by a combined

more radical community members for working with the outside development experts in 1975. Interview, Salómon Galindo, 77, 28 October 2004. As Smith notes, radical leftists increasingly criticized the Allpachaka station and its staff for collaborating with foreign imperialists, and for creating conditions that were creating a new class of rich peasants through Allpachaka’s activities. Smith, "Rural Development in the Crossfire,” 36-38. 973 Salvatierra et al., La Vida, 264. 974 Salvatierra et al., La Vida, 264. Francisco Núñez Vilca helped to dig the same hole as the only survivor of Sakaypata, whose testimony opens this dissertation’s preface. 975 Salvatierra et al., La Vida, 255-56. 976 Salvatierra et al., La Vida, 242.

298 group of soldiers accompanied by dozens of peasants who had already taken prisoner four other Quispillacctinos from Pampamarca. They were all beaten and bloodied by the mob, who wanted to kill them all right there by throwing them into the Wayunka ravine. "Let's start with this chibolo (kid)," they said. Instead, the military shipped them all to the base in Totos. Cirilo Galindo Huamaní’s young brother-in-law was released by the military; however, Cirilo and the other captives brought with him to Totos were all killed in the Sankaypata massacre.977

The military’s counter-insurgency campaign in the district in May, 1983 followed a plan enacted in other highland districts. Violent repression was combined with demands to resurrect state functions and offices. Community authorities banned by Sendero were reassigned by the military. A daily accounting of residents in each barrio was instituted, and self-defence militias - or rondas campesinas - were created in communities.978 District residents submitted their communities to military control in late May, even displaying white paper flowers at military- required assemblies to signal surrender.979 In the interests of survival, residents quickly conformed to the military’s demands, even if it meant having to live double lives, presenting a firm pro-government face when the soldiers appeared, or as Senderista supporters when the guerrillas returned – a strategy no doubt also employed by Chuschinos at times. Guillermo Vilca Galindo was a lieutenant governor of Quispillaccta in 1982. “When the terrorists arrived, we behaved like Senderistas; when the military came, we fought alongside the military. That’s how we behaved. Our lives depended on a thread, and there was no other way out.”980 Vilca Galindo was a former member of the armed forces but was still detained by the military for one week in 1983, accused of helping to organize community members to be Senderistas. He vigorously

977 Salvatierra et al., La Vida, 258-59. 978 Luis Medoza Achallma, appointed to be an authority in Catalinayuq by the military in late May, remembers the conjunction of deadly repression and demands for cooperation by the Army that month. Quispillacctinos would be pulled from community lineups to disappear in helicopters to the Totos army base, never to be seen again. But those left behind were exhorted to loyalty and activity against Sendero: The soldiers filled the helicopter with those fingered as Sendero sympathizers by a peasant dressed in military cloths, he explained. “To those of us left, the Captain asked: “Who are you with?” We responded in chorus: “With you my general!” Again, he asked us: “Are you going to organize against the path?” And, again, in chorus we all answered: “Yes, my general!”” Salvatierra et al., La Vida, 251. 979 Salvatierra et al., La Vida, 266. 980 Salvatierra et al., La Vida, 268.

299 argued with his military captors, telling them: "I am a political authority and I do what I have to do as an authority so that the whole population will respect me."981

In the months that followed, Quispillaccta quickly implemented the military’s demands for cooperation; however, Quispillacctinos continued to suffer disproportionately in the counter- insurgency violence in the district, leading some to question whether neighbouring communities were still telling the military that Quispillaccta was a Sendero stronghold. In June 1983, the community members of Quispillaccta met in Unión Potrero and decided to form the first self- defence committees. Each barrio would have a self-defence committee, and a mutual defense pact was made against subversive attacks.982 Quispillaccta also participated in the wider organization of anti-Sendero communities in the area. In September 1984, in the Rio Cachi community of Putaje, a large meeting was held with the participation of community members of Quispillaccta, Putaje, Wariperqa, Quchapampa, Condorpaqcha and other surrounding communities. In total, 24 villages came together. A plan of resistance was developed and they adopted solidarity commitments – mutual security arrangements. One of the agreements committed Quispillaccta to fight collectively against the Shining Path and come to the aid of the neighboring towns that were attacked.983 Their commitments were tested a few weeks later, on 24 September 1984, when dozens of Sendero militants entered Putaje, sacking and burning 41 houses. The self-defence committee of Putaje and a squad of soldiers from the military base of Casacancha, supported by community members of Quispillaccta, pursued the Senderistas across the D. Chuschi, finally catching up with them in the puna of Aqumate, near the border of Tuco and the District of Paras. Three Quispillacctinos were killed along with several soldiers and

981 Salvatierra et al., La Vida, 268. His two-faced strategy was common among peasants trying to survive the insurgency. Theidon writes about similar dynamics in the Huanta communities of northern Ayacucho that she studied: “Villagers learned that survival might well depend on showing one face to soldiers and another to the guerrillas. People lived their public and secret lives, masking their torn allegiances. Many people insisted that everyone became “two-faced” (iskay uyukuna), and one could never know which way anyone might turn.” Theidon, Intimate Enemies, 15. Vilca Galindo’s strategy of survival, though, failed his father in May 1983. His father, Martin Vilca Tomaylla was sleeping in Llaqtuahuran when military personnel passed his house and gave deceptive Sendero cheers. His father, acting like many trying to bend with which ever group was in town, replied with a similar cheer, not knowing that Sendero was not outside his door. He was arrested and subsequently killed in the Sankaypata massacre. The account of his father’s capture was in Peru’s CRV, Inform, Volume 2, 2.3, 29. 982 Salvatierra et al., La Vida, 215. 983 Salvatierra et al., La Vida, 215.

300

Senderistas.984 The community had now been bloodied together with the military in counter- insurgency operations against Sendero.

In spite of their alliance with the military, Quispillaccta struggled against continued rumours that the community supported Sendero. In 1984, when working on the orders of the Army to build a town square in Quispillaccta’s Matriz, Catalinayuq authority Luis Mendoza Achallma heard a rumour that the residents of Cancha Cancha and Chuschi were “coming to Quispillaccta to turn us into “dust” (ashes).” They were again encouraging the military to attack Quispillaccta by claiming that Quispillacctinos were all terrorists. A short time later, in Catalinayuq, Captain Jackal sent them a military guide to bring to the town centre, and warned them to take care of him or else be destroyed. As they returned to Quispillaccta’s town centre, dawn broke and they recognized their mysterious guide. He was the person who Quispillacctinos believed had been encouraging the military to again attack Quispillaccta. Heated discussion followed: “We asked him, ‘Why do you accuse us of being terrorists and massacre us? We are people of God.’ We told him, ‘Because of you, for calling us terrorists, they beat and torture us.’ When the group approached Pallqa, a neighborhood due north of Chuschi’s main town, the man escaped from the group in the direction of his village.985 The grievances against Quispillaccta ran deep and remained dangerous, even after the Peruvian army had regained control of the district.

Explaining violence in the district in the early 1980s in the context of historical land conflicts

The early 1980s are a critical moment in the centuries-long conflict between Quispillaccta and Chuschi because wider conflict currents in Peru began to intrude upon the district, eventually enveloping its residents in a brutal civil war for more than a decade. Understandably, scholars have wondered whether land conflicts in the district either contributed to the rise of Sendero in the area, or helped to condition patterns of violence in the district during the brutal early 1980s - during Sendero’s early violent actions in the district and the government counter-insurgency in the area. From the beginning, astute observers of developments in the district like Isbell suspected that the long history of local conflicts between Chuschi and Quispillaccta played some

984 Salvatierra et al., La Vida, 215. 985 Salvatierra et al., La Vida, 252-3.

301 role in the violence in the area during the Sendero insurgency – though the precise role was never explained in detail. Recent research by historians like Miguel La Serna has come to firmer conclusions based on an examination of past land conflicts in the district, linking Sendero’s violence against Chuschinos during the 1980s dirty war to Quispillaccta’s long-standing land grievances against Chuschi. The specific claims of scholars like La Serna that historical land grievances contributed to the rise of Sendero and help explain Sendero’s violence in the district in the early 1980s – particularly his claim that Sendero found greater support for its violence from aggrieved Quispillacctinos – needs careful consideration, however. Claims that Quispillacctinos used support for Sendero to seek redress for historical land grievances fails to adequately consider the implications of Quispillaccta’s success in the land struggles of the 1960s and 1970s, and the outcome of the 1981 peace settlement between Quispillaccta and Chuschi. In lights of the careful assessment of patterns of violence in the district in the early 1980s outlined above, some recent interpretations about the sources of Sendero’s support in the district and reasons for Sendero violence during the early 1980s appear questionable.

Research by Isbell, Sànchez Villagomez, Castillo Gamboa, La Serna, and my own conversations with district residents, confirms that many comuneros in the district, regardless of which community is examined, supported – or, at least, tolerated - Sendero’s efforts in 1981 and 1982 to rid the district of bad apples - to punish delinquents, thieves, and those who abused their power.986 No doubt, many found Sendero’s actions extremely distasteful. There was a strong tradition of avoiding the use of violence or displays of public conflict among comuneros in the district, as Isbell noted in the 1970s.987 Historically, peasants channeled much of their inter- community conflict in the back and forth litigation over land disputes. Using violence as a means of conflict management appears to have been rare, a pattern Theidon also noted in Huanta.988 Sendero’s willingness to use humiliation and public violence in 1981 and 1982 –

986 La Serna’s account of how different communities in the district supported Sendero’s specific actions against these bad apples is the most detailed to date, and rings true with my research. Building on past work by Isbell and Sànchez Villagomez, his account stresses that district peasants supported Sendero’s efforts to discipline those who violated peasant conceptions of moral order and those who abused their positions of authority for their own gain. See La Serna, The Corner of the Living, 149-157. Scholarly analysis of peasant responses to Sendero are discussed in Degragori, How Difficult it is to be God, 62-65. 987 Isbell, To Defend, 230-233. 988 Theidon, Intimate Enemies, 188-191. She writes: “Conflicts were constant battles within and between villages, and ritualized battles figured among the conflict management mechanisms practiced throughout Ayacucho.

302 including lethal violence – must have been a terrible shock to many district residents. However, scholars are probably correct that community members throughout the district accepted or at least tolerated this violence because of frustration that these abusers and local criminals were not being stopped by state authorities. Comuneros only began to actively turn on Sendero in late 1982 and 1983 when they began targeting district residents who many believed had not violated community norms and practices, and thus were not legitimate targets. (The coercive hammer of the military’s counter-insurgency operations in the district at this time no doubt hastened the shift against Sendero, as discussed above.) As Isbell wrote, comuneros turned on Sendero when “Sendero went too far. It went beyond killing known enemies and killed village authorities. It promised a better world where “truth without deceit is supreme.” But Sendero did not bring the promised truth; rather, it replaced the old power with its own authoritarianism.”989 Peasants rejected Sendero’s authoritarianism in the same way that they rejected traditional gamonales. 990

So, Sendero’s lethal justice may have offered district comuneros a form of peasant justice for violation of community norms in the face of an incompetent or absent state. La Serna convincingly argues that Sendero’s actions disciplined local leaders or power holders who violated the cultural component of peasant moral expectations – that a cultural moral economy existed in Ayacucho communities composed of “morally established assumptions that indigenous peasants brought to their relationships with all local power holders.”991 Specifically, La Serna argues that indigenous peasants had a “power pact” with local power holders, “in which peasants were willing to submit to and legitimize the dominion of local power holders provided that the latter lived up to these cultural expectations.”992 Community leaders or Mistis who violated peasant moral expectations through corruption, abuse, theft, possession of community land, or other moral failings were thus seen as legitimate targets for punishment or censure.

However, although the battles were widespread, the exercise of lethal violence was rare.” Theidon, Intimate Enemies, 190. Theidon quotes Carlos Iván Degragori who frequently heard the expression in rural Peru, “castigar pero no matar.” Punish but do not kill. Theidon, Intimate Enemies, 190-1. 989 Isbell, “Shining Path and Peasant Responses,” 94-5. 990 Heilman importantly notes that Degregori was mistaken in concluding that peasant acceptance of Sendero’s authoritarianism in the early 1980s was conditioned by peasant tolerance of gamonales in rural Ayacucho. Heilman, Before the Shining Path, 197. 991 Miguel de la Serna, “Murió comiendo rata: Power Relations in Pre-Sendero Ayacucho, Peru, 1940-1983,” A Contra Corriente: A Journal on Social History and Literature in Latin America 9, no.2 (Winter 2012): 4. 992 La Serna, “Murió comiendo rata,” 4.

303

Sendero’s targeting of such moral misfits thus faced little opposition among district residents, according to La Serna.

This argument tilts less toward James Scott’s moral economy argument and more toward notions of good governance and accountability. Though La Serna suggests that his argument explains why peasants were willing to subjugate themselves to some legitimate local power holders but not others, in practice this “subjugation” often amounted to what many of the governed in democratic societies tolerate from those who govern them. Mestizo elites holding land long- claimed by communities were not seen as legitimate, and were not offered moral legitimacy like landholding elites in Scott’s moral economy account. However, some local landholding elites could retain legitimacy by upholding the moral power pact in the eyes of local peasants. Local Chuschi mestizo Ernesto Jaime, for example, was accepted as a community development leader by Chuschinos, lauded for his role in directing the fight against neighbouring land grabbers like Quispillaccta, and respected for putting the welfare of the community of Chuschi above his own interests, even though he was widely known to be one of the largest landowners in the district. When he died in 2005, the community gave him a hero’s funeral, with a parade through town. Jamie was clearly seen as someone who lived, both personally and during his role as a community authority, with cultural legitimacy among Chuschinos, compared to other abusive mestizo elites in Chuschi like Humberto Ascarza.993 Jamie was a proud man in person, who wanted his voice to be heard in communal meetings, even when I met him in 2004 in his late 80s. Some Chuschinos privately whispered to me that he was a Misti, had stolen land from widows, and was arrogant, thinking himself better than other comuneros. But La Serna is correct when he says that Jamie was “flawed, to be sure, but generally respectful of the power pact.”994

Evidence collected by Isbell, La Serna, and Sànchez Villagomez strongly supports the conclusion that Sendero’s early violent acts in the district targeting violators of moral justice and moral power pacts were largely accepted. In eastern Cangallo, historian Jaymie Heilman notes that Sendero similarly killed abusive local elites in the early 1980s, violence that horrified

993 La Serna’s article, “Murió comiendo rata,” discusses the contrast between these two Chuschino elites. 994 La Serna, “Murió comiendo rata,” 27.

304 peasants, even if they “understood” why these elites were targeted.995 However, we cannot easily conclude that acceptance or tolerance of Sendero actions amounted to active or passive support for Sendero’s wider revolutionary goals. Clearly, actions in 1983 showed that when push came to shove and comuneros had to decide if revolutionary goals trumped indigenous peasant cultural norms, communities turned on Sendero.

The idea that support for Sendero stemmed from those who were aggrieved by violations of local “power pacts” also dovetails with research discussed above about how the impacts of peasant economic and social differentiation stemming from capitalist development in the Andes during the post-WW2 period may have influenced support for groups like Sendero. Here the material basis for grievances discussed in the revolution literature, combine with insights from Peruvian social change literature and empirical findings in highland districts like Chuschi. Economic and cultural differentiation was more palatable in a changing Peru if the newly upwardly well-off and ambitious peasants adhered to cultural moral power pacts in rural communities. However, when their ambitions and actions led them to violate cultural norms among peasant communities, because they took advantage of new-found wealth or positions to unfairly enrich themselves at the expense of the community or individual peasants, grievances built-up that were sometimes exploited by groups like Sendero. By the mid-late 1970s, the path to becoming an abusive Misti or gamonal in rural Ayacucho - behaving like a rural “boss” who violates the local moral power pact - might be a result of a historical landowning legacy or it could be the result of new-found development success. The reality was locally contingent. However, the reactions of peasants to violations of the moral power pact were similar in different communities. Such individuals were not seen as legitimate and thus worthy of protection from censure or targeting by groups like Sendero. Within communities, the continued existence of such abusive individuals into the late- 1970s, years after Agrarian Reform was supposed to rid the highlands of Mistis, might also explain why some would turn to support Sendero’s actions against other community members.

Less convincing, however, are La Serna’s claims that specific support for the Sendero uprising in the district was an extension of intercommunity conflict between Chuschi and Quispillaccta. Bloodshed from the 1960s was certainly on the minds of peasants in the district when Sendero

995 Heilman, Before the Shining Path, 192 & 196.

305 entered in the early 1980s, La Serna writes.996 He goes on to say that, “[t]he intense inter- community struggle provided an important incentive for peasants to support the Shining Path in the early years of the rebellion,”997 and “that indigenous Chuschinos and Quispillacctinos manipulated the political violence to revisit the historic intervillage feud.”998 In fact, La Serna’s argument goes a step further in linking historical land conflicts to differences in peasant support for Sendero between communities. La Serna repeatedly argues that Quispillaccta’s support for Sendero was broader and deeper than support in Chuschi, and that this support was directly linked to decades of inter-community conflict.999 It is no surprise that two of the four university students who burned the ballot boxes were from Quispillaccta, La Serna argues, because their actions that day “may have been fueled as much by young Quispillacctinos’ desire to burn down their rivals’ administrative center as it was by their desire to ignite the flames of communist revolution.”1000 “There were more terrorists in the zone of Quispillaccta, they were united with them,” La Serna quotes another Chuschino, Gregorio Cayllahua, as saying in TRC testimony.1001 For La Serna, revenge against Chuschi explains why so many Quispillacctinos seemed to have turned to supporting the Shining Path when the insurgency began. Here he echoes Gustavo Gorriti’s history of the conflict, where Gorriti claimed that Sendero “gained much more influence in Quispillaccta than Chuschi.”1002

By contrast, in discussing why Chuschinos supported Sendero to settle scores with Quispillaccta, La Serna comes to confusing conclusions. After many pages detailing the historical conflict

996 La Serna, To Cross the River of Blood, 135. 997 La Serna, To Cross the River of Blood, 136. 998 La Serna, The Corner of the Living,158. 999 La Serna, The Corner of the Living, 160-63. La Serna’s certainty of Quispillaccta as the source of Sendero support in the district is even sharper in an earlier publication, when he writes that “…even the very first episode of Shining Path insurgency in Peru, which we have taken to signify the Initiation of the Armed Struggle (ILA) against the Peruvian state, may have been fueled as much by young Quispillacctinos’ desire to ignite the flames of communist revolution in a metaphorical sense as it was to ignite the shingles of their rivals’ administrative center. The significance of this act would not have been lost on Chuschinos, serving as a political parallel to the burning of Chuschino religious centers in years past. All this leads us to the conclusion that in Chuschi the “river of blood” did not only lead to a classless society. It also led, quite literally, to Quispillaccta.” La Serna, To Cross the River of Blood, 136. 1000 La Serna, The Corner of the Living, 162. 1001 La Serna, The Corner of the Living, 162-63. 1002 Gorriti, The Shining Path, 19. Gorriti offers no evidence or citations to support this claim.

306 between Chuschi and Quispillaccta, La Serna writes that “it should come as no surprise that Chuschinos used the Shining Path insurgency as a pretext for revisiting this conflict.”1003 While he acknowledges that most of the victims in the early years in the district were from Quispillaccta, his detailed discussion of specific Sendero killings accepted by Chuschinos leaves the impression that Chuschi’s support for Sendero was limited in scope and time, primarily to punishing transgressions of the moral power pact.1004 His examples of Quispillacctino victims targeted by Sendero with Chuschi’s approval include two well-known community delinquents who were condemned even by their own community members in Quispillaccta.1005 With their deaths, he acknowledges that Chuschi’s support for killing morally degenerate Quispillacctinos was its way of settling scores with Quispillaccta over land clashes.1006

However, nowhere does La Serna investigate whether Chuschinos – and community members from Chuschi’s affiliated hamlets – supported killing Quispillacctinos for revenge who were neither widely accepted moral delinquents nor abusive authority figures – the main targets of Sendero’s early violence in the district. Nor does La Serna investigate Isbell’s suggestion that the violence in the district perpetrated by the Peruvian military was directed by ordinary Chuschinos at ordinary Quispillacctinos because of their historical land conflict.1007 In the end, the uncomfortable question that La Serna fails to examine is the extent to which the excessive targeted killing of Quispillacctinos, during Sendero’s rule of the district and particularly during the military counter-insurgency campaign in the area – killings which far outnumbered the killings of Chuschinos,1008 was largely spurred by score-settling of the historical land conflict between the two communities against Quispillaccta, the major victor in those conflicts during the 1960s and 1970s?

The deeper understanding of the specific dynamics of the historical land confrontations in the district over many decades outlined above provide the answer to this question. By early 1980s

1003 La Serna, The Corner of the Living, 160. 1004 La Serna, The Corner of the Living, 158-61. 1005 La Serna, The Corner of the Living, 160-161. 1006 La Serna, The Corner of the Living, 161. 1007 Isbell, “Shining Path and Peasant Responses,” 89. 1008 A fact noted by Isbell, “Shining Path and Peasant Responses,” 87.

307 when Sendero began to openly operate in the D. Chuschi, Quispillaccta had been far more successful in land conflicts with Chuschi and its annexes. This reality formed the context of social and political conflict in the district as Sendero’s operatives began to plan their insurgency against the Peruvian state in the district during this time. The peace settlement between Chuschi and Quispillaccta that was negotiated in late 1981 and finalized in February, 1982, formally recognized Quispillaccta’s control over most of the contested terrain in the Rio Cachi zone and left those who lost out in this struggle bitterly disappointed. Remarkably, La Serna’s account of the intercommunity conflicts between Chuschi and Quispillaccta never mentions this peace settlement, even though the deal that was reached finally established a secure and lasting line of division between the communities in the contentious Rio Cachi zone and in the hills above the Rio Pampas settlements. It was a deal that has withstood the test of the dirty war in the district after 1982, and it has survived the post-insurgency era to the present day. 1009

While the peace deal was able to finally settle the most contentious part of the long-standing conflict between the communities, the outcome of the settlement had important impacts on both communities because it froze in place a specific reality on the ground in the Rio Cachi zone – a reality that, in the end, favoured Quispillaccta in the long-running land conflict with Chuschi. Quispillaccta now essentially had legal control over much of the contentious Rio Cachi zone that they fought over with Chuschi. Added to this, Quispillaccta was similarly successful in the late 1970s in its conflict over Pillcoccasa, an outcome that disadvantaged comuneros from some of Chuschi’s sister annexes like Cancha Cancha. These outcomes provide the answer to the unanswered question above: Quispillaccta’s victories in contentious land disputes spawned grievances among the losers, grievances that some sought to settle during the Sendero insurgency, and more importantly, during the Peruvian military’s counter-insurgency campaign in the district. Violence was perpetrated against Quispillacctinos to a greater degree than against Chuschinos during early 1983 precisely because of Quispillaccta’s decades-long successes in land disputes with Chuschi and its annexes.

1009 In fact, in a recent publication listing unresolved border disputes between communities in Ayacucho, the centuries-long dispute between Chuschi and Quispillaccta is not listed. See Cuadro Nº 12, Conflictos Entre Comunidades Por Linderos No Delimitados, Región De Ayacucho, in Roger Maquera and Serafín Osorio, Las comunidades campesinas, 18.

308

Chapter 9 Conclusion

Qualitative fieldwork pays off – delayed temporal impacts of resource capture and revised understanding of conflict causes

The dissertation’s local, field-based approach to examining environmental change-conflict linkages has satisfied many of the suggestions of scholars like Tobias Ide, who have promoted field-based qualitative research as a way of advancing our understanding of debates about environmental change-violent conflict linkages, including debates about climate change and violent conflict.1010 Process-tracing and ethnographic research methods can highlight temporally deferred impacts of human-induced environmental change, Ide notes, and better contextualize change at the local level. This can allow scholars to “pinpoint such wider social, economic, and political dynamics which are highly relevant for understanding … [environmental change]- conflict links, but which are easily ignored by approaches focusing on the narrow relationship between two variables.”1011 In the D. Chuschi, the conflict impacts of local resource capture in the decades following Conquest, when highland estates were established by wealthy Mestizo elites in the puna of the Rio Cachi basin at the expense of local communities, would help cause conflict between communities centuries after peasants were originally dispossessed of grazing lands in this area.

This study captured the delayed temporal dynamics of conflict in the district. Conflict dynamics between actors involved in the original capture of grazing lands were largely irrelevant centuries later; conflict in the late twentieth century was not between the land-grabbing Mestizo elites and the communities who lost them, but between rival communities fighting for control of these former hacienda lands. By the 1960s and 1970s, the old Mestizo elites and the Catholic Church

1010 Ide, “Research methods,” 1-14. See also the sources in note 12. 1011 Ide, “Research methods,” 5.

309 who were the source of so much local enmity ceased to have much relevance in land affairs in the district, apart from the conflict between the community of Chuschi and the owners of the former Yaruca hacienda lands. Abusive local mestizo elites living in Chuschi’s main town were instead a source of continuing conflict for district residents, especially in the intra-community struggles in Chuschi described by La Serna.1012 These local Mistis and lawyers hired by the communities to advance their litigation also aggravated conflict between Quispillaccta and Chuschi in the 1960s and 1970s, as described above, and complicated efforts to find a peaceful solution to the land conflict between Chuschi and Quispillaccta.1013 It is likely that some elites probably aggravated conflicts between Chuschi and Quispillaccta for their own benefit, in ways that harken to insights from Colin Kahl1014; though, these linkages were not described in detail in this dissertation. Unsurprisingly, when peace was achieved between Chuschi and Quispillaccta,

1012 La Serna, The Corner of the Living, chapter 3. 1013 This was confirmed by numerous community members that I spoke with in both Quispillaccta and Chuschi. Lawyer Braro Albornoz from Huamanga was one Misti community members from both communities agreed helped worsen the conflict between Quispillaccta and Chuschi in the years leading to the 1960 Kimsa Cruz conflict. According to an account from a Quispillacctino anciano, Albornoz advised Quispillaccta to kill some Chuschinos because it would help speed up their land settlement litigation. When Quispillaccta rejected his advice, Albornoz angrily kicked out the Quispillacctino authorities from his office. He was soon working for Chuschi and encouraging them to kill some Quispillactinos to speed up their land settlement. Interview, Esteban Nuñez, Quispillaccta, October 15, 2004. A Chuschino noted that lawyers like Albornoz were bad because they did not want a solution. They were Quichka mikuy – someone who eats thorns. Interview, Marcelino Rocha Cayllahua, Chuschi, October 24, 2004. Albornoz seemed to have played the role of a typical tinterillo – literally, “one who deals with ink,” a phrase given to lawyers or amateur lawyers. Stephen Brush quotes Metraux’s description of the role of tinterillos: “…in regions where the Indians’ lot has not improved as a result of an agrarian reform land hunger sometimes assumes the form of an obsession. It gives rise to interminable lawsuits between Indians’ to the advantage of the notorious tinterillos – shady lawyers, who since the colonial era have earned a living by exploiting the Indians.” Quoted in Brush, Mountain, 61. Heilman also notes the damaging impacts of tinterillos in Carhuanca, in eastern Cangallo, in the mid-twentieth century. See Heilman, Before, chapter 4. 1014 Kahl, States, 227. As scarcities of cultivated and pasture lands increasingly challenged household livelihoods in the district after 1940, there was an increase in clashes within the community of Chuschi between Mestizo elites and peasants. Of course, there were other important changes also taking place among the peasantry that helped mobilize them to throw off the yoke of elite domination in Chuschi in the 1960s and 1970s, like increasing education, agrarian reform, greater national and regional political activity among peasants, and increasing economic development opportunities. Local environmental change was only one factor among many in the district that accelerated peasant activism in Chuschi and increasingly led them to challenge traditional Mestizo elites. This is a broader range of factors than proposed by Kahl in his elite exploitation thesis. Whether elites in Chuschi sought to aggravate conflict between Chuschi and Quispillaccta as a way of maintaining their control in the district remains unproven, however, even if there are hints of an aggravating role for elites based on field interviews and in the work of La Serna.

310 a new generation of peasant leaders brokered the final deal, free from the decades of corrupt and abusive influences from Chuschi’s mistis.1015

By pushing the temporal scope of this study to the pre-Colonial and Colonial period, this dissertation also reveals how many scholars misdiagnosed the roots of local conflict in the district, because they failed to understand the degree to which elite capture of lands in the district had conditioned livelihood realities between communities in the twentieth century. As a result, scholars failed to recognize how conflict over efforts to reclaim elite-captured lands could help drive violent conflict in the years before and during Sendero’s uprising. For decades scholars saw the conflict between Chuschi and Quispillaccta as an inter-community struggle over land, in a district largely free from the exploitative impacts of large haciendas. Billie Jean Isbell, the celebrated ethnographer of Chuschi stated that Chuschi “has escaped many of the semifeudal relationships of the hacienda system” because “[h]aciendas did not develop in Chuschi’s marginal land with its immense expanse of high puna communal pastures.”1016 Isbell was echoing the views of IEP ethnographic studies of the district in the 1960s. District boundaries helped to obscure the pernicious impacts of hacienda domination in the district because most of the historical haciendas impacting the communities in the D. Chuschi were located in neighbouring districts to the north.

These views had implications for scholarly analyses of Sendero’s rise in Cangallo in the 1980s and 1990s. Degregori argued that southern provinces like Cangallo had few large haciendas, compared to northern Ayacucho, so class conflict spurred by elite land-owner domination likely had little to do with the growth of Sendero in this region, compared to the impact of locally abusive elites or gamonales. Isbell even speculated that Sendero chose the D. Chuschi for its first armed actions precisely because of the absence of a history of hacienda domination.1017 As

1015 When examining conflict dynamics between Chuschi and Quispillaccta from the 1940s to mid-1970s, there were always some abusive Mistis in positions of authority in the district or as authorities in the community of Chuschi. Their presence correlates to actions which worsened relations between the communities during this period, and thus cannot be entirely discounted. Quispillaccta, as noted above, did not have any abusive Mistis, according to informants; though, there were a small number of community members or outsiders that some Quispillacctinos characterized as Mistis who helped navigate the legal side of land conflicts between communities, thus possibly contributing to the conflict’s durability and scope. 1016 Isbell, “Shining Path and Peasant Responses in Rural Ayacucho,” 78. 1017 Isbell wrote: “I believe that Sendero chose Chuschi for its initial military operations precisely because of this absence of haciendas, which allowed [Sendero Luminoso] to experiment with peasant communities that had strong

311 a result, scholars have over-emphasized the degree to which violent conflict in the district was spurred by historical and constructed ethnic differences – the Canas of Quispillaccta versus the Aymaraes of Chuschi – and the impact of abusive gamonales. Given the degree of historical ethnic mixing noted by scholars like John Earls,1018 can we characterize the conflict in the district as a conflict between neighbouring ethnic groups? Or, is it more accurately portrayed, as this thesis argues, as an inter-community land conflict aggravated by the impact of centuries of elite capture of patrimonial lands, supply and demand scarcities, and perceptions of ethnic differences between communities in the district? Key to explaining the violence in the district in the 1960s and 1970s, and the particular form of violence in 1983 during the military’s counter-insurgency campaign, was the way that conflict between Quispillaccta and Chuschi played out over efforts to reclaim former hacienda lands in the district. Since the pre-Conquest period, elites had also manipulated ethnic difference among communities in the area for their own purposes, heightening conflicts between Quispillacctinos and Chuschinos. Twentieth century land conflicts, aggravated by supply and demand scarcities in the district, similarly worsened divisions between groups in the district. Thus, historical patterns of resource capture and their social effects described in this dissertation were crucial for aggravating and setting the causal pathways in the latter decades of the twentieth century toward inter-community violent conflict in the district.

This conclusion does not speak directly to the question of whether scarcities in livelihood resources directly led to or contributed to the Sendero’s revolutionary uprising, which was the original question guiding this thesis. However, the evidence presented here demonstrates that environmental scarcities, combined with other important political-economic and social factors, aggravated group conflict between communities in the district in the decades leading up to Sendero’s insurgency. These linkages between environmental scarcities and inter-community violence are highlighted as the red pathway in the systems diagram below, Figure 32.

communal structures, autonomy over their resources, and whose experiences with capitalistic market penetrations were minimal. By initiating their revolution in what they believed to be a region that had escaped many of the semi- feudal relationships of the hacienda system, Sendero perhaps hoped to avoid the mistakes made by the guerrillas inspired by the Cuban revolution who failed to gain the support of hacienda peons for their short-lived insurgency in 1965.” Isbell, “Shining Path and Peasant Responses,” 78. District residents were no doubt aware of the long shadow of hacienda control of their patrimonial lands in the first years of Sendero’s uprising, let alone the impact of abusive Mestizo elites in the district, even if there were few functioning haciendas left in the area at that time. 1018 Noted above in footnote 333.

312

Revising theoretical links between environmental change and violent conflict

This dissertation’s findings about how environmental scarcities, combined with other political, economic, and social factors, aggravated group conflict between communities in the D. Chuschi informs long-standing debates around qualitative environment-conflict research, suggests revisions in hypothesized causal mechanisms, and supports some key findings. Processes described here support some of Homer-Dixon’s environment-conflict hypotheses, but suggest clarification and refinement are needed in others. Resource capture of grazing lands and Colonial relocation of communities to the lower Rio Pampas zone of the district stunted livelihood options for communities in the district. But the hypothesized process of resource capture described by Homer-Dixon1019 is not as helpful in explaining developments in the district compared to political ecologist research approaches that emphasize how “social relations of production and the social fields of power” resulted in particular “systems of access to and control over resources,” by elites in the district, and the consequences for social stability when these systems

1019 Homer-Dixon, Environment, Scarcity, and Violence, 74-75.

313 of control and power broke down.1020 Consumption-driven demand signals were acted upon by Mestizo elites who saw highland grazing lands as valuable commodities to seize, irrespective of patterns of population growth among peasants, and this drove resource capture during the Colonial and early Republican period in the district.1021 This case suggests that Homer-Dixon’s conception of resource capture should be revised to consider situations when population growth does little if anything to spur elites to grab resources. Instead, we should acknowledge that in the D. Chuschi case and in many others, the incentives of consumption demands are sufficient for driving resource capture.1022

The impacts on households of resource capture in the D. Chuschi during the Colonial period were moderated for several hundred years by the relative abundance of land for households following the severe population declines after Conquest. Supply and demand pressures on households as a result of heavy reliance on cultivated assets in the southern zone of the district began to be felt in the twentieth century, this dissertation concludes, as populations recovered in the district. Growing supply and demand pressures in the Rio Pampas zone of the district presaged increasing ecological marginalization1023 of households in this area, as yield declines and extensification of agricultural production signalled greater pressure on cultivated assets in this zone. Homer-Dixon and colleagues from the Toronto Group are thus correct in arguing that population-driven supply and demand impacts cannot be endogenized to political-economic factors and processes.1024 Detailed evidence presented in Chapter 7 highlight the increasingly negative impacts of such processes on households, particularly in Quispillaccta.

However, the livelihood immiseration from ecological marginalization predicted by Homer- Dixon’s theory failed to materialize for many district households because local and regional

1020 Quoted in Deligiannis, "The Evolution of Qualitative Environment-Conflict Research: Moving Towards Consensus." 42. 1021 These land seizures were happening, in fact, in the context of low or falling peasant population levels in the district. 1022 I have developed this point more extensively in Deligiannis, "The Evolution of Qualitative Environment- Conflict Research: Moving Towards Consensus." 44-47. This point does not invalidate Homer-Dixon’s conception of resource capture, which posits an interaction between population growth-driven demand changes and elite moves to capture scarcer, more valuable resource assets. I simply did not find this process at work in this study. 1023 Homer-Dixon, Environment, Scarcity, and Violence, 77-79. 1024 Deligiannis, "The Evolution of Qualitative Environment-Conflict Research: Moving Towards Consensus," 42.

314 adaptation options were available. To deal with worsening impacts of supply and demand scarcity, many peasants shifted livelihood activities to increasingly emphasize animal husbandry and moved to settle contested grazing lands in the northern puna zone of the district. As well, strategies of temporary or semi-permanent migration and petty commodity production were increasingly employed to supplement household livelihoods. While resource capture and ecological marginalization were linked in the D. Chuschi case, as Homer-Dixon suggests1025, they were temporally distant and not connected in ways predicted by his hypotheses. More importantly, household adaptations to the impacts of demand and supply changes in the twentieth century suggest that Homer-Dixon’s conception of ecological marginalization as a process for household immiseration overly simplifies dynamics around human-environmental pressure on households. Household adaptative capacity can moderate the impacts of ecological marginalization. This conclusion is developed more fully below in section 9.5 where I discuss this dissertation’s development of a livelihood framework for environment-conflict research.

Adaptations to demand and supply pressures by households in the Rio Pampas zone of the district did have conflict impacts, this dissertation concludes. More generally, the historical account of land conflict in the D. Chuschi affirms Homer-Dixon’s conclusion that “environmental scarcities can aggravate divisions or segmentation among ethnic, religious, and linguistic groups.”1026 Increasing group divisions may have worsened negative othering and contributed to the outbreak of violence.1027 Clearly, as communities adapted to increasing scarcities of grazing and cultivated lands, competition for these resources with neighbouring communities became more pronounced, and enmity between communities increased. These conflict processes were also stimulated by local and regional economic development trends which increased the value of control of grazing lands for community members in the twentieth century, particularly as peasants increasingly articulated their livelihood activities into the expanding cash economy in the highlands. Development work in the Rio Cachi basin, stimulated

1025 Homer-Dixon, Environment, Scarcity, and Violence, 78. 1026 Homer-Dixon, Environment, Scarcity, and Violence, 96. 1027 This conclusion appears to support Ide’s suggestion that group divisions and negative othering are key factors to pay attention to when attempting to discern environment-conflict causes. Ide, “Why do conflicts,” 62. However, Ide classifies negative othering as a structural variable in his 2015 QCA analysis, which ignores the empirical reality that group divisions and negative othering vary over time.

315 by efforts at the Allpachaka rural research station, probably contributed to increasing tensions over disputed lands as those shut out of access to these lands looked on from the sidelines while some Quispillacctinos profited from the increasing commercial opportunities that came to them as a result of their control of the disputed lands.1028 This suggests that the impacts of scarcities may have combined with increasing social and economic differentiation among peasants in the area to aggravate group conflict. Peasant differentiation as a possible causal mechanism of rural rebellion and revolution was hypothesized in chapter 2 as a possible source of Sendero support; however, the evidence presented here suggests scarcities and economic differentiation among peasants from different communities combined to instead aggravate community enmity, with violent consequences during the insurgency.

Homer-Dixon has argued that scarcities interact with other socio-economic and political variables to help produce negative social impacts. The violent clashes and the casualties from these local land conflicts only increased the fear, mistrust, and bitterness toward neighbouring communities in the twentieth century. These impacts did help cause households in the district to “turn inward and to focus on narrow survival strategies” for their communities, as Homer-Dixon predicted.1029 This conclusion is also supported by the empirical findings of two development specialists who worked in the district between 1978 and 1983. Vera Gianotten and Ton de Wit wrote in 1983 that boundary conflicts between Chuschi and Quispillaccta reinforced communal organization within communities in the district.1030 After clashes in 1960, to take another example, many families in Quispillaccta stopped sending their children to schools in Chuschi, and Quispillaccta increasingly pressed to build their own schools. Hardening attitudes toward

1028 I acknowledge that the evidence for this point is not developed sufficiently in this dissertation. Further research in the district is needed to comprehensively document how economic development patterns in the Rio Cachi zone of the district played out between Quispillaccta and competing communities in the 1970s. However, the evidence presented above points to this conclusion. In outlining why Sendero arose in eastern Cangallo and the province of Victor Fajardo, Theidon points to the impacts of frustrated economic and development expectations in the context of the economic crisis that hit Peru in 1976, particularly among students and teachers. Theidon, Intimate Enemies, 327. Her nod toward relative deprivation theory as a cause of radicalization of wage-earning teachers and aspiring student professionals can be seen as a parallel argument to my argument that rural developments in the 1970s in some places in the D. Chuschi was also leading some to be frustrated with their lack of progress compared to others in their community or in neighbouring communities. 1029 Homer-Dixon, Environment, Scarcity, and Violence, 96. 1030 Vera Gianotten and Ton de Wit, "Organizacion Popular: El Objetivo De La Investigacion Participativa,” Boletín de Estudios Latinoamericanos y del Caribe 35 (December 1983), 105.

316 each other resulted in fewer interactions between members of different communities in the 1960s and 1970s, divisions that would turn deadly in the early 1980s. The struggle for land between Quispillaccta and her neighbours thus appeared to be stimulating both group-identity conflict between different communities and Quispillaccta, but also exhibiting relative deprivation motivations among those angry at Quispillaccta for their success. As Homer-Dixon writes, “If the historical identity of a clearly defined social group is strongly linked to a particular set of natural resources or a particular pattern of resources use, degradation or depletion of that resource can accentuate a feeling of relative deprivation. Members of the group can come to feel that they are being denied their rightful access to resources that are key to their self-definition as a group. This relative deprivation boosts grievances that may eventually be expressed through aggressive assertion of group identity.”1031 Grievances among competing communities built up during Quispillaccta’s successful pursuit of land did not lead to insurgency, as Homer-Dixon’s theory suggested. Instead, the aggrieved used events during the Peruvian government’s counter- insurgency campaign to get back at Quispillaccta.

Scarcities can condition violence during violent conflict

These conclusions are important because they suggest that the key finding in this dissertation is not how environmental scarcities contributed to the onset of the Sendero rebellion, but instead how resource disputes and their outcomes conditioned patterns of violence in the D. Chuschi during the first years of Sendero’s uprising, and particularly during the Peruvian military’s deadly counter-insurgency campaign. Quispillaccta, largely the victor of the inter-community conflict over land and water resources, suffered disproportionately in the military counter- insurgency campaign after being characterized as a hotbed of Sendero support. However, given their success in competing with their neighbours over land disputes and the community’s economic development progress in the Rio Cachi zone, the logic of Quispillaccta being the key centre for Sendero support in the district – a common conclusion among scholars and Peruvian experts - is questionable. As a community, they had the most to lose by throwing their support behind Sendero, following their success in consolidating land in the district. While some Quispillacctinos certainly supported Sendero, the same was true for every community in the

1031 Homer-Dixon, Environment, Scarcity, and Violence, 147-8.

317 area.1032 There is no evidence that there were proportionally more Sendero supporters in Quispillaccta than any other community in the district. The most logical explanation for Quispillaccta being labelled as a centre for Sendero support and suffering disproportionately in the government’s counter-insurgency war is that grievances stemming from their success in the intercommunity resource competitions led some in neighbouring communities to claim that Quispillaccta was a Sendero stronghold in order to seek revenge for their land conflict successes – score-settling by counter-insurgency. Survivor testimony examined above seems to confirm this conclusion.

Thus, this dissertation argues that we cannot understand how patterns of violence developed during the insurgency unless local changes in livelihood assets are assessed over the long term. The violence perpetrated during the insurgency and particularly after 1982, when the Peruvian government began an aggressive counter-insurgency campaign in the area, was many times more deadly to district residents than group conflicts over land resources in the decades before Sendero’s insurgency. This violence disproportionately impacted Quispillaccta. Conflicts over environmental resources thus impacted the process and direction of violence in the civil war in the district once it began, with tragic consequences, particularly for Quispillacctinos.

Recent field research in other parts of Ayacucho provides evidence to support this dissertation’s conclusion about the impacts of local land conflicts on violence during the civil war. At the end of September, 1981, special Peruvian anti-terrorism police (Sinchis) entered the community of Sarhua, across the Rio Pampas to the south of the D. Chuschi in the Province of Victor Fajardo, beating comuneros, looting houses, and smashing communal property. They violently captured fifteen community members and threw them in jail in Huamanga. Accompanying the police was

1032 This finding also helps to explain why Sendero targeted several Quispillacctinos from the Rio Cachi barrios in May 1983 for execution because of their commercial and development activities. On May 20, 1983 Sendero killed Tomás Moreno Casavilca from Catalinayuq for being opposed to Sendero. Salvatierra et al., La Vida, 246. The same day Sendero also executed Cirilo Mendoza Mendieta, his wife Melchora Ccallocunto Conde and their son Albino Mendoza Ccallocunto in Catalinayuq. According to Albino’s sister, Sendero said that he didn’t help the party and preferred his business; that’s why he was killed. His parents were killed, because they were accused by Sendero of informing on Senderistas in the community. A note left after their killing said, “If you want to accuse, you will also be killed.” Salvatierra et al., La Vida, 270. Theidon notes that Sendero sometimes accused those who were ideologically opposed to Sendero of being snitches to justify their execution. Peasant differentiation within Quispillaccta as a result of land conflict success and economic development in the area may have also led to peasant differentiation within Quispillaccta and may have helped turn some who were not prospering towards Sendero. However, I do not have sufficient evidence to support this conclusion.

318 a local community member who accused the peasants of terrorism under the Peruvian government’s recently passed terrorism legislation.1033 A few months earlier, in early May, the community voted to expel “this comunero who wanted to be a gamonal”1034 from Sarhau for various abuses, after confiscating all his property and livestock and taking back communal land he had illegally seized in recent years.1035 The expelled ‘gamonal’ took his revenge on those who ejected him by accusing five Sarhuino authorities of terrorism.1036 When the accused authorities could not be located by the Sinchis, they roughly captured whoever was at hand in the town. After a few days, police released the imprisoned landowners for lack of evidence.1037 Some time after, however, Sendero militants in Sarhua killed the abusive landowner and secretly disposed of his body.1038 As Theidon notes, there were distinct regional patterns in Ayacucho to killings during the Sendero years: “Although Ayacucho was the “birthplace” of Sendero Luminoso, there were local patterns to the violence and its consequences – within and between communities, and between these communities and the Peruvian state.”1039 Local grievances

1033 Olga M. González, Unveiling Secrets of War in the Peruvian Andes (Chicago: The University of Chicago Press, 2011), 137-142. 1034 González, Unveiling Secrets, 127. 1035 González, Unveiling Secrets, 127-133. 1036 González, Unveiling Secrets, 133-137. 1037 González, Unveiling Secrets, 142. Several prominent Sarhuinos in Huamanga, including a border policy officer, were instrumental in convincing authorities that the terrorism accusations were false, and securing the release of the captured Sarhuinos. It is notable that pressure and persuasion could still overturn false terrorism accusations in 1981while Peruvian police headed the counter-terrorism operations in Ayacucho. The military’s arrival in Ayacucho a little over a year later, dramatically increased the brutality of Peruvian government repression. 1038 González, Unveiling Secrets, 146. 1039 Theidon, Intimate Enemies, 188. Sendero’s willingness to use violence was also used to settle scores in several other communities in Ayacucho. Theidon describes how an angry villager in Carhuahurán, in Huanta Province, denounced local authorities to Sendero as government snitches because he was punished for theft. Sendero rounded up eight authorities and a random group of villagers based on a list written on a matchbook and executed all but one comuneros, who managed to escape. Theidon, Intimate Enemies, 196. In San Miguel, Ayacucho, Nory Cóndor Alarcón and Nelson E. Pereyra Chávez examine how a decades-long intra-community land conflict between villagers led to a targeted killing by Sendero in 1984. Nory Cóndor Alarcón and Nelson E. Pereyra Chávez, “Desaparecidos en la penumbra del atardecer: disputas privadas, memoria y conflicto armado interno en San Miguel (Ayacucho),” Anthropologica 23, no. 34: 2015, 63-88. In another example, Sendero was used to settle old land disputes between the communities of Ocros and Pampas, communities in the neighbouring provinces of Huamanga and La Mar, located on the eastern edge of Ayacucho, on the western bank of the Pampas River. Violence from their historical land conflict helps to explain why Sendero murdered members of the community of Ocros in mid- January, 1984. Ocros largely won the land dispute in the 1970s against Pampa, according to research by Valérie Robin Azevedo, and this conditioned who was later targeted by Sendero and where Sendero chose to attack those who had recently allied with the military in their counter-insurgency efforts. Sendero’s attack set off a series of reprisals and revenge killing between the communities that killed 100 people in the first four months of 1984. See Valérie Robin Azevedo, Sur les sentiers de la violence: Politiques de la mémoire et conflit armé au Pérou (Paris:

319 around resources and power stimulated violence in Ayacucho along with wider national ideological struggles led by Sendero.

Environmental scarcities were thus crucially important for shaping patterns of violence during the civil war, this dissertation concludes, in addition to influencing processes that led to violent conflict onset. This finding is new in environment-conflict research and supports calls in recent reviews of climate change-conflict research to explore how the “dynamics” of conflict are impacted by climate change.1040 The dissertation’s findings also confirm work by civil war scholar Stathis Kalyvas on the sources and nature of violence in civil war, where “individuals and local communities involved in the war tend to take advantage of the prevailing situation” and denounce rivals “to settle private and local conflicts whose relation to the grand causes of the war or the goals of the belligerents is often tenuous.”1041 Violence was used instrumentally by Peru’s military “to generate compliance” with the government, and oppose Sendero, as Kalyvas suggests.1042 Serious human rights abuses were committed, but the violence was not “indiscriminate repression,” as Theidon describes the violence in Huanta, in northern Ayacucho.1043 Instead, Peru’s military set out to re-establish control over the district by indiscriminately targeting Quispillacctinos, knowing that some were probably Senderistas or sympathizers, but not knowing who specifically to target. For district residents, this had the effect of appearing selective and discriminate toward Quispillaccta, though the military did abuse

OpenEdition Books, 2019), Ch 3, paragraphs 84-104. https://books.openedition.org/iheal/8256, accessed Feb. 12, 2020. The historical land dispute between the two communities is outlined in: Heraclio Bonilla, La Defensa del Espacio Comunal Como Fuente de Conflicto: San Juan de Ocros vs. Pampas, Ayacucho, 1940-1970, Documento de Trabajo No. 34, Lima: Instituto de Estudios Peruanos, 1989. 1040 Joshua Busby, “Taking Stock: The Field of Climate and Security,” Current Climate Change Reports 4, no. 4: 2018: 341; Cullen Hendrix, Scott Gates, and Halvard Buhaug, “Environment and Conflict,” in T. David Mason and Sara McLaughlin Mitchell eds., What Do We Know about Civil Wars? (New York: Rowman & Littlefield, 2016), 243-244. Hendrix et. al. are correct in noting that “environmental factors may exacerbate contestation over property,” and that “migration induced by environmental factors is one way that could possibly spark a fight over property.” Their suggestion that state capacity, inequality, and unclear property rights can mediate such clashes is somewhat supported by my research. However, this dissertation’s findings suggest that greater attention should be paid to how potentially conflictual livelihood adaptations like migration are also driven by supply and demand scarcities, and that these pressures can heighten the stakes for conflict over disputed lands, processes ignored in Hendrix et. al.’s discussion. 1041 Kalyvas, Logic, 364. 1042 Kalyvas, Logic, 28. 1043 Theidon, Intimate Enemies, 5.

320 and kill peasants from Chuschi and its annexes as well at times.1044 Denunciations of Quispillaccta as a Sendero stronghold by some in Chuschi and its annexes in early 1983, stemming from decades of heated local land conflict, unleashed indiscriminate counter- insurgency violence against Quispillacctinos. From the military’s perspective, this violence was necessary to establish order, to turn them against Sendero and toward collaboration with Peru’s security services, precisely as Kalyvas’ theory of indiscriminate and discriminate violence suggests.1045 The violence was indiscriminate from the perspective of the Army and Chuschinos, but not from the perspective of Quispillacctinos, who saw that they were suffering more than other communities in the district.

The response from Quispillaccta seems to confirm that they felt targeted because it led the community to conclude that if they got rid of the Senderistas in their midst, then they would not be killed. However, if they did not eliminate the Senderistas, the entire community might be wiped out. So, Quispillaccta quickly conformed to the military’s demands for cooperation, and the community allied with neighbouring communities in the area against Sendero, to show their allegiance to the state. Some Senderistas were told to leave the community, while others were disappeared by their own community. Several Senderistas from Quispillaccta were quietly killed by the community when they refused to leave or reform their ways. Some of the bodies were thrown into the abandoned salt mine in the district.1046 Faced with the military’s ruthless pressure, Quispillacctinos “learned to kill our brother” to survive. 1047 Many communities confronted this terrible calculus in the months after the military’s counter-insurgency campaign intensified in rural Ayacucho.1048 Disentangling motives, causes, and impacts of violence at the

1044 Kalyvas, Logic, 145. Kalyvas writes: “In practice, the distinction between selective and indiscriminate violence hinges on public perceptions since it is possible to pretend to be selective by indiscriminately targeting isolated individuals. As long as people perceive such violence to be selective, it will have the same effects as selective violence. If people do not perceive it as selective, the results will be opposite, much like when they perceive selective violence to be indiscriminate.” 1045 Kalyvas, Logic, 150. 1046 Interview, Quispillaccta, 2004. Name and location withheld. 1047 The phrase comes from Theidon, who notes similar dynamics at work in Huanta, in northern Ayacucho, where community members had to kill the Senderistas in their midst to survive. Theidon, Intimate Enemies, 5. See her discussion in chapter 7. 1048 González discusses how community members in Sarhua killed a prominent local Senderista in March, 1983. Cleansing the community of this Sendero leader, rather than handing him to the military was seen as the safest

321 local level in the D. Chuschi thus adds to the growing scholarly effort to make sense of the complex reality on the ground in Ayacucho during the Sendero insurgency, and highlights different outcomes from the impacts of scarcities than this project first hypothesized.

Scarcities shape structures and processes, rather than trigger conflict

Another key finding of this dissertation speaks to debates among environment-conflict scholars about the causal significance of environmental scarcities and whether they can act as violent conflict triggers.1049 Environmental scarcities in the D. Chuschi did not act to trigger violent conflict, as Ide suggests; instead, various political and policy changes in Peru in the 19th and 20th century triggered local violent conflicts in the district, particularly in the three decades leading up to Sendero’s insurgency. This conclusion supports another of Ide’s hypotheses that policy changes toward natural resources can act to trigger conflict by altering opportunity structures that encourage groups to act.1050 Policy changes on land reform correlated to peasant activism and conflict over land resources going back to the early 19th century, decades before scarcities began to acutely impact household livelihoods in the district. Government reforms in the early 1800s allowing peasants to buy land led Quispillacctinos to seek to buy declining haciendas in the Rio Cachi zone. Reforms after World War 1 led to greater peasant activism and defence against encroaching haciendas in the area. Policy changes in the early 1940s to finally recognize peasant communities led both Chuschi and Quispillaccta to apply for formal peasant community recognition and boundary marking, setting off the first large-scale inter-community conflicts in the district in the twentieth century. Land reform movements in the late 1950s in other parts of Peru helped stimulate the violent clashes of 1960, while nation-wide land reform in the 1970s by President Velasco’s military-led government accelerated the struggle for contested lands in the district. These fights reached a peak in the late 1970s and early 1980s as Peru’s government faced economic crisis and a political opening to return to democratic elections in 1980. In each case, local communities reacted to policy changes or the events triggered by these changes in

course of action by Sarhuinos, because the militant could have implicated many in Sarhua with supporting Sendero. González, Unveiling Secrets, 206-07. 1049 Ide, “Why do conflicts over scarce renewable resources turn violent?” 62-3. 1050 Ide, “Why do conflicts over scarce renewable resources turn violent?” 63.

322 other parts of Peru to seek to reclaim lost lands or solidify control over lands that they believed were their own in the district. In some cases, economic developments combined to stimulate this increased political activism, as with the impact of increased wool demand during and after WW1. But, in each case, the proximate political impact of these economic and policy changes on conflict within the district was clear.

With regards to the Peruvian military’s counter-insurgency violence in the district in 1983 and the terrible toll suffered disproportionately by Quispillaccta, state decisions to deploy counter- insurgency forces in the district appeared to create opportunities for aggrieved community members in the district to get back at Quispillaccta.1051 In this case, a strong state capacity for counter-insurgency violence possibly altered the opportunity structure for score-settling.1052

Rather than violent conflict triggers, environmental scarcities are instead causally significant as a deep structural factor, altering causal pathways and impacting livelihood constraints and opportunities available to households. While some might characterize environmental scarcities as a root cause, this would be an overly narrow description of their causal impact because scarcities also act as dynamic pressures – “processes and activities that ‘translate’ the effects of [scarcities] both temporally and spatially”1053 into negative impacts on households, depending on the degree of scarcities and the adaptation and coping strategies employed by households. The structural impacts of environmental scarcities are not “static and invariate over time,”1054 as some scholars suggest, but instead change and transform in relation to the degree of human-pressure on the resources. Their causal structural role in the D. Chuschi case is analogous to tectonic plates, as Homer-Dixon has suggested.1055 Their precise causal role is likely contextually

1051 We do not yet know the motivations behind those who sought to settle scores by blaming Quispillaccta as Senderistas. Perhaps some thought it would lead to their reclaiming land now controlled by Quispillaccta in the peace deal, though this may have also been a misunderstanding on their part of the impact of the counter-insurgency policy. However, beyond instrumental impulses, their actions may have also been performative or driven simply by revenge. 1052 I would like to thank Steven Bernstein and Joshua Busby for suggesting that a changing opportunity structure and the role of the state were also crucially important for causing the violence in early 1983. 1053 See Wisner et.al., At Risk, 48. 1054 Ide, “Why do conflicts over scarce renewable resources turn violent?” 62. 1055 Homer-Dixon, Environment, Scarcity, and Violence, 18. Homer-Dixon’s use of this geo-physical analogy to describe the causal role of environmental scarcities harkens to Galtung’s similar usage to describe how cultural violence expresses its impacts. See Johan Galtung, “Cultural Violence,” Journal of Peace Research 27:3 (1990):

323 specific and difficult to generalize, however. It is conceivable that environmental scarcities could act to trigger conflicts in other contexts.

Livelihoods should be at the centre of environment-conflict research

Research on household adaptation to environmental scarcities and other livelihood pressures in the D. Chuschi also leads us to revise thinking about the most appropriate the level of analysis of environment-conflict research and about prevailing hypotheses about linear processes from environmental scarcities to violent conflict outcomes. Much of the research to date on environment-conflict linkages has chosen to use a state-level of analysis when attempting to determine if environmental scarcities can help cause violent conflict. This dissertation demonstrates that such an approach is misguided, because scarcities impact households and groups at the local level, and their local responses to these pressures are often lost if the focus is on the state level. This dissertation has instead developed a household model for environmental scarcity-conflict linkages. Situating the household livelihood model at the centre of the causal analysis of the impacts of environmental scarcities allows for a more complex understanding of the impact of scarcities in light of other factors noted by scholars like political-economic processes. Multiple levels of analysis are necessary to gain a full picture about why violence broke out; local and state level processes were key in early 1983. Multi-level analysis that includes the local level should become the norm in future environment-conflict research.

Equally important, the livelihood framework developed in this study allows for an appreciation of how household assets and capacities mediate social impacts through livelihood diversification. Environment-conflict research that posits linear processes from environmental scarcities to social effects and then to violent conflict is simplistic because households adapt and cope to stresses like environmental stresses or economic transformation by diversifying activities. Certainly, Homer-Dixon’s arguments that environmental scarcities can constrain agricultural and economic

294. Galtung described cultural violence as a “permanence,” in ways that are similar to how Homer-Dixon sees environmental scarcities. But scarcities are not permanent or unchanging. They evolve over time depending upon many influences. So, they can act as root causes or as dynamic pressures, to borrow the causal characterization of Wisner et al.

324 productivity are correct.1056 But there are likely few instances of a direct impact from environmental scarcities to social effects as he suggests. Livelihood diversification in the face of scarcities can also result in various unanticipated social effects that are far removed temporally or geographically from the impacts of scarcities, complicating efforts to deal with negative social impacts that may result. The livelihood model developed in this dissertation also cautions scholars about how livelihood diversification can lead to increased differentiation among those impacted by scarcities, because households are not equal and they cope and adapt differently to stresses. Differentiation can cause winners and losers among those impacted, exacerbating group divisions or facilitating less than desirable choices from those whose coping and adaptation strategies fail. Policy interventions for those impacted by environment change should seek to broadly enhance livelihood opportunities, reduce human security, and build adaptive and coping capacity, while avoiding maladaptation and outcomes that increase inequities.1057

In the D. Chuschi, communities forced to give up control of their puna rangelands during the colonial period under the pressure of newly established Mestizo haciendas adapted and refocused their livelihood strategies on cultivated agriculture on the hills fronting the Rio Pampas. Some herding continued, but probably with less livestock than before colonization. Diminished by the diseases of the Spaniards and their exploitation of labour, community cultivation efforts were probably sufficient for livelihood needs until the early twentieth century, when growing populations increasingly led households to look for new adaptive opportunities to supplement livelihoods. By this time, opportunities for settling the Rio Cachi zone were increasingly attractive, along with increasing wage labour from temporary migration and simple commodity production. Households in Chuschi and Quispillaccta adapted to scarcities in many ways,

1056 Homer-Dixon, Environment, Scarcity, and Violence, 80-85. 1057 W. Neil Adger et al., "Human Security," in Climate Change 2014: Impacts, Adaptation and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Wokring Group Ii to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, ed. Christopher B. Field, et al. (Cambridge: Cambridge University Press, 2014), 762; Jürgen Scheffran, Tobias Ide, and Janpeter Schilling, "Violent Climate or Climate of Violence? Concepts and Relations with Focus on Kenya and Sudan," The International Journal of Human Rights 18, no. 3 (2014), 376; A.K. Magnan et al., "Addressing the Risk of Maladaptation to Climate Change," Wiley Interdisciplinary Reviews: Climate Change 7, no. 5 (2016), 661; Colette Mortreux and Jon Barnett, "Adaptive Capacity: Exploring the Research Frontier," Wiley Interdisciplinary Reviews: Climate Change 8, no. 4 (2017) 7 of 12. Montreux and Barnett's review notes that the limited research on adaptation outcomes does not “support the assumption that systems with high adaptive capacity will adapt well.”

325 depending upon the prevailing economic, political, and social constraints and opportunities. At times, these adaptations did have important conflict implications. The impacts of scarcities in the district defied simple, linear predictions of social effects. Their adaptations increasingly shifted the impact of scarcities of cultivated lands to conflicts over high-altitude grazing lands in the district’s Rio Cachi zone as more comuneros settled in this zone. Some households no doubt prospered by these moves; however, others may have been unsuccessful with these adaptations, particularly those whose property was destroyed or livelihoods destroyed during inter- community conflicts.

Complex adaptive and coping responses of households modelled in this dissertation point to many possible outcomes, and instead suggest that we need to pay attention to both changes in household vulnerability and risk and to constraints on adaptation and coping, if we want to model likely social effects of pressures like environmental scarcities and their possible violent conflict implications. Increasing economic decline in the late 1970s and early 1980s in highland Peru, as reflected in declining government investment, dropping agricultural incomes, and reduced health outcomes described by McClintock1058 was ratcheting up the stress on peasant households, and may have increased the significance of struggles to control contested lands in the district. In the context of household livelihoods increasingly challenged by scarcities, these additional stresses probably reduced the adaptive options for households in the district, and may have further aggravated community conflict in the district.

1058 McClintock, Revolutionary Movements, 170-184.

326

Postscript I am optimistic about peace in the D. Chuschi – that a lasting peace will prevail between community members who for centuries came to see peasants from the next community as different; as rivals; as the other. For far too long, some people have tried to manipulate Chuschinos, Quispillacctinos, Canchacanchinos, and others in the district for their own selfish needs. The story outlined above shows the tragic consequences of some of these actions. However, I see hope when I think of Mario Silverio Huamani Conde, one of my research assistants during my time in the district, and someone who is emblematic of a new future for residents of all the communities in the D. Chuschi. Though now based in Huamanga, Mario has close family in both Quispillaccta and Chuschi. As we travelled around the district during my fieldwork in 2004, Mario was completely at ease visiting family and comuneros in both communities. His dual identity is much more common now in the district, decades after the violence of the early 1980s left deep scars among many. Peace now holds between former rivals, and community members are slowly building on shared identities in the district.

To be certain, memories and grievances are still raw. We saw the anger in some people when we asked about past conflicts between district communities. These raw memories and hardened attitudes toward their neighbours will not easily disappear. However, my interactions with various community members and past and present authorities in the district encourage me that many people in the district – hopefully most - have pushed past using violence to settle their disputes. They know better than anyone the consequences of using violence to settle disputes; it is not a cliché to say that violence begat violence in the district over the decades. Today, district authorities no longer only come from the Community of Chuschi and its annexes. Quispillacctinos now routinely hold the highest administrative offices in the district government. Abusive authorities can easily be removed in the next election. Local governance is much more representative today of those being governed.

Competition between Quispillaccta and its neighbours has not disappeared, however. In the past decade there have been repeated efforts by some Quispillacctinos to separate the community from the D. Chuschi, in order to create their own district in Cangallo. While unsuccessful thus far, the sentiment behind such efforts suggest that separation is seen by some as a better solution than building a shared future. There is scholarly evidence that competition between the

327 communities has moved into the political contests that are regularly held for district political offices.1059 But that should be seen as an encouraging development, because healthy political competition means that conflicts between communities can be settled within the governance structures, where they should be settled.

Cooperation among all the community members in the district is crucially important if the district is going to deal with the multiple challenges currently facing highland residents. While the livelihoods of district residents have improved markedly since Sendero began its insurgency in the area in 1980, the region remains one of the poorest in Peru.1060 Some households now produce products that they sell for cash, whether it is milk, or vegetables like garlic and potatoes.1061 However, most district agriculturalists only marginally produce for the local and regional market economy, and significant barriers remain for producers in the district to increase household production for cash markets.1062 Roads being built in the district during fieldwork in 2004, are now directly connecting even the most isolated barrios in Quispillaccta’s Rio Pampas zone, like Pirhuamarca, and Soccobamba, to Huamanga and Lima. These road links may enable greater market production in the future. As well, climate change is dramatically impacting highland districts like Chuschi, as noted above, altering productive relationships and long-held certainties about best practices for agricultural livelihoods. In semi-arid Ayacucho, possible

1059 Gustavo Flórez Salcedo, “Rivalidades comunales y contiendas electorales: micropolítica en las elecciones distritales de Chuschi. El caso de las comunidades campesinas de Chuschi y Quispillaccta,” in Alejandro Diez Hurtado ed., Tensiones y transformaciones en comunidades campesinas, (Lima: CISEPA-PUCP, 2012): 225-261. 1060 Oscar Espinosa de Rivero notes that Peru’s 1993 census listed Cangallo as having the fourth lowest per capital income level among all Peruvian provinces and the fourth highest level of the percentage of households in poverty among Peruvian provinces. Improvement has been slow. A 2002 UNDP report showed that Cangallo had the second lowest rate of human development in Ayacucho and the lowest rage of monthly per capita income. Óscar Espinosa, Desafíos para la inserción en el mercado por parte de familias campesinas y de pequeños productores de cereales en la provincia de Cangallo, Ayacucho. Programa Comercio Y Pobreza En Latinoamérica, Lima: Consorcio de Investigación Económica y Social (CIES), Febrero 2009), 7-8. 1061 Fieldwork notes. 1062 Espinosa de Rivero’s analysis notes some of the significant barriers farmers in neighbouring districts face in increasing grain production for local and regional markets. Similar barriers are faced by household producers in the District of Chuschi. During fieldwork in Quispillaccta’s Rio Cachi barrios I asked a dairy farmer if she produced cheese for Huamanga’s market. I had seen several farmers selling cheese on the streets of Huamanga in an earlier visit. She said she would like to produce cheese to sell in the market, but had no access to refrigeration or transport to enable her to produce and sell cheese in the capital. Such barriers are common to producers who are still relatively isolated, have limited resources and technical capacity, and located many hours from urban centres.

328 declines in the frequency and level of precipitation as a result of climate change could have severe impacts on household livelihoods.1063

Travelling through Quispillaccta’s Rio Cachi barrios, I also I noted with interest that larger portions of Quispillaccta’s productive punas were being partitioned and claimed by households – walled off from their neighbours puna holdings to ensure control and access to grazing lands, much like land in the lower cultivated zones have been for decades. These are new developments and the implications are not well studied. Parcelization of puna holdings may lead to better management and prevent over-grazing – particularly as climate change challenges farmers to manage their grazing lands. However, parcelization may also lead to greater wealth differentials among community members. Increasing household differentiation within communities like Quispillaccta or Chuschi could undermine traditional communal governance and reciprocity relationships. While Quispillaccta has managed to sustain its traditional governance structures, Chuschi has struggled to keep them in place in the post-Sendero era. In both communities, the desire of young people to leave the community for educational and professional opportunities challenges the maintenance of traditional governance structures.1064 Accelerated market access by remote or isolated producers may improve livelihood options for households; however, there’s no guarantee that it will improve income levels and reduce poverty, and it may increase the challenges communities face to retain their traditional cultural practices and communal control of local agricultural production.1065

1063 See Asociación Bartolomé Aripaylla (ABA), Agricultura sostenible para la adaptación al cambio climático en regions vulnerables de Bolivia y Perú: Informe de línea de base del programa en Ayacucho, Ayacucho, September 2011. 1064 Mariano Aronés Palomino, “Chuschi: el territorio del Chimaycha,” in Proyecto Qhapaq Ñan: Informes de investigación etnográfica, Pueblos y culturas en las rutas del Qhapaq Ñan, Ayacucho y Huancavelica Campaña 2003, Volume 1 (Lima: Instituto Nacional de Cultura, 2005), 132-36. 1065 Oscar Espinosa de Rivero notes that only a small number of potato producers in Cangallo have been able improve poverty indicators through greater production for local and regional markets. Espinosa de Rivero, Desafíos, 38-9. Mayer’s research highlighted that increasing commercialization of production impacts communal organization: “pressure toward commercial cropping often translates into a secessionist move [by households] to free a production zone from overall communal organization.” Mayer, The Articulated Peasant, 303.

329

We should applaud the efforts of community members and NGOs like the Asociación Bartolomé Aripaylla (ABA) that have diligently worked to help district community members face current and future challenges from the climate or markets, while also striving to strengthen community solidarity and preserve traditional cultural processes. ABA began in Quispillaccta more than twenty years ago. The NGO has recently been lauded for extensive efforts to build rainwater harvesting capacities in the district, by reconstructing and building small water reservoirs throughout the punas that take advantage of local bofedales –

ABA has helped spread knowledge and practice about temporary peat wetlands, which ABA helps managed seasonal reservoirs like this one in Quispillaccta’s Rio Cachi puna. Source: ABA communities turn into rainwater storage reservoirs, recharging local aquifers and storing water for dry periods.1066 They have also sponsored indigenous seed fairs to encourage the preservation and strengthening of local crop biodiversity and indigenous approaches to nurturing local agriculture, crucially important efforts to manage both pest and climate risk.1067 As well, ABA has started programs to rebuild local agricultural terraces in the district, to preserve soil fertility and overcome labour scarcities that handicap terrace maintenance. Such efforts to build local resilience and capacity among district agriculturalists are tremendously

1066 Asociación Bartolomé Aripaylla, Yakumama – Madre Agua - Lagunas de Lluvia y Comunidades Criadoras del Agua, Sistematización de la experiencia de la Asociación Bartolomé Aripaylla y Comunidad de Indígenas de Quispillaccta sobre la Crianza del Agua - Siembra y Cosecha del Agua de Lluvia (Huamanga, Ayacucho: 2014). M.S. Maldonado Fonkén, “An introduction to the bofedales of the Peruvian High Andes,” Mires and Peat, 15(5),2014/15, 1–13. 1067Marcela Machaca Mendieta, Crianza Andina de la Chacra en Quispillaccta (Ayacucho: Asociación Bartolomé Aripaylla, 1994); Tirso Gonzales, Nestor Chambi and Marcela Machaca, “Agriculture and cosmovision in the contemporary Andes: the nurturing of the seeds,” in United Nations Environment Programme, Cultural and spiritual values of biodiversity, ed. Darrell Addison Posey (London: Intermediate Technology 1999)., 211-217.

330 important and models for other communities. Such work should be encouraged and supported where possible. Recognizing shared challenges among district residents like these, and working to build livelihood resilience through such projects, helps to create stronger and more united communities in the district.

For a number of years, since it first dawned on me that grievances around Quispillaccta’s land conflict successes found deadly expression during the Peruvian military’s brutal counter- insurgency operations, I have puzzled about why sentiments persist about Quispillaccta being a Sendero stronghold. Clearly, there were Quispillacctinos who became active, militant supporters of Sendero. Some comuneros, as La Serna suggests in his discussion of early support for Sendero in the district, may even have supported Sendero as a way to strike back at Chuschi, their historical foe.1068 But there is no convincing evidence to prove that Quispillaccta’s support for Sendero was any greater than any other community in the district. In fact, La Serna also notes cases of Chuschinos using Sendero as a tool to take revenge on their rival Quispillaccta.

Scholars have a responsibility to take care in leveling judgments based on an incomplete understanding of history. There seems to be a strong scholarly consensus on the sources and scope of early tolerance or support for Sendero’s actions in the district. However, I find some scholars have been less than responsible in repeating and amplifying claims that Quispillaccta was the Sendero stronghold in the district based on an incomplete accounting of developments in the district. A thorough and judicious history of the district during the Sendero insurgency has yet to be written. I have agonised over whether the evidence I present in this dissertation is sufficient to prove that some of those who lost out to Quispillaccta in the decades-long fight for district lands quietly pointed fingers at Quispillaccta, when Peru’s military demanded to know where to find the Senderistas. I believe that the evidence I present builds on early writing by Isbell, and more recent work by others in the district to prove that such accusations were, at their core, tools for revenge, rather than a reflection of reality.

1068 La Serna, The Corner of the Living, 160-63.

331

So, I wonder whether the desire to exculpate Chuschi and tar Quispillaccta with the Sendero label by some scholars is instead an attempt to make up for an earlier historical wrong – the persistent myth among many Peruvians that Chuschi was a stronghold of Senderistas. Only a few months after the ballot boxes were burned, according to Deborah Poole and Gerardo Rénique, Chuschi “had become a household name throughout Peru. Evoking long repressed colonial fears of the highland peasantry’s legitimate claims for justice, Chuschi came to stand for coastal Peruvians as a symbol of Sendero Luminoso’s roots in Ayacucho. For Peru’s highland citizens, it resonated as a metaphor for all the frightening implications of the ‘ILA’, as Sendero’s acronym for ‘Initiation of Armed Struggle’. As the years went by and Sendero’s armed struggle intensified, the nature and scope of its early exploit in Chuschi were magnified accordingly. By 1986, the influential Peruvian magazine Sí was simply repeating what had become accepted historical fact when it described the students’ actions as a full-scale military ‘occupation’ of Chuschi by ‘armed graduates of Sendero’s first military school.’”1069 For Chuschinos and those who know the community and their suffering during the dirty war, the injustice of Chuschi being the symbolic and actual representative of Sendero in Ayacucho is both evident and galling. Is this sentiment behind the efforts of some scholars to repeat questionable claims?

We would do well to remember, as scholars working on events of such emotional and physical tragedy, the uses and abuses of history. As Margaret MacMillan has recently written:

History can be helpful; it can also be very dangerous. It is wiser to think of history, not as a pile of dead leaves or a collection of dusty artifacts, but as a pool, sometimes benign, often sulphurous, which lies under the present, silently shaping our institutions, our ways of thought, our likes and dislikes. We call on it … for validation and for lessons and advice. Validation, whether of group identities or for demands, or justification, almost always comes from using the past. You feel your life has a meaning if you are part of a much larger group, which predated your existence and which will survive you (carrying, however, some of your essence into the future). Sometimes we abuse history, creating bad or false histories to justify treating others badly, seizing their land, for example, or killing them. There are also many lessons and much advice offered by history, and it is easy to pick and choose what you want. The

1069 Poole and Rénique, Peru: Time of Fear, 57-8.

332

past can be used for almost anything you want to do in the present. We abuse it when we create lies about the past or write histories that show only one perspective. We can draw our lessons carefully or badly. That does not mean we should not look to history for understanding, support, and help; it does mean that we should do so with care.1070

I have sought to take care in telling this story, and I hope that a fuller historical accounting of events in the district will one day be written. If I have erred in my conclusions here, I take responsibility for those errors and I hope that care is taken to correct them. Care and empathy are needed to heal the wounds of the terrible tragedies in the district during Sendero’s rebellion, particularly if we hope that Quispillacctinos, Chuschinos – all residents of the district - can work and live together free from violence and hate in the future.

1070 Margaret MacMillan, The Uses and Abuses of History (Toronto: Penguin Group, 2009): ix-x.

333

Bibliography