UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Social inequality and resource management : gender, caste and class in the rural Himalayas Tsering, Tashi 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2015_february_tsering_tashi.pdf [ 3.28MB ]
Metadata
JSON: 24-1.0167050.json
JSON-LD: 24-1.0167050-ld.json
RDF/XML (Pretty): 24-1.0167050-rdf.xml
RDF/JSON: 24-1.0167050-rdf.json
Turtle: 24-1.0167050-turtle.txt
N-Triples: 24-1.0167050-rdf-ntriples.txt
Original Record: 24-1.0167050-source.json
Full Text
24-1.0167050-fulltext.txt
Citation
24-1.0167050.ris

Full Text

	   	   SOCIAL INEQUALITY AND RESOURCE MANAGEMENT: GENDER, CASTE AND CLASS IN THE RURAL HIMALAYAS  by  Tashi Tsering   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Resource Management and Environmental Studies)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  November 2014 © Tashi Tsering, 2014 	   ii	  Abstract  The management of irrigation water and other resources, as practiced by traditional farming communities in developing countries, is often presented as a model of an equitable system – especially when compared to systems managed by states. This study demonstrates that the resource management practices in two Himalayan farming communities are, in fact, inequitable in terms of local gender, caste and class roles. This thesis examines inequalities in the social organization of irrigation systems in two villages in Spiti Valley in India’s Himachal Pradesh state. Its key finding is that the social organization of irrigation management, particularly in terms of farmers’ gender, class and caste backgrounds, is best understood as part of a broader division of labor for farming and related resources (such as for the management of fodder, dung and firewood), which are all embedded in the local socio-economic structure. This finding, which is based on participatory observation and interviews with farmers, as well as an analysis of historical and legal documents, underlines the importance of studying management of different resource sectors relationally rather than compartmentally. In particular, this study identifies key functional linkages between the social organization of farming and different resource sectors and develops theoretical approaches to the study of resource management in rural communities.  	   iii	  Preface  This dissertation is an original and independent work by the author, T. Tsering. All of the content of the thesis is unpublished except for a section in Chapter 3 that discuss the historical origins on Khangchen households in Spiti Valley, India. That discussion is included in a published article by the same author as follows:   Tsering, T. (2013). On the Unknown History of a Himalayan Buddhist Enclave: Spiti Valley Before the 10th Century. In T. Takeuchi, K. Iwao, A. Nishida, S. Kumagai, & M. Yamamoto (Eds.), Current Issues and Progress in Tibetan Studies: Proceedings of the 3rd International Seminar of Young Tibetologists (pp. 523-551). Kobe: Kobe City University of Foreign Studies.  The fieldwork for this study was covered under UBC Ethics Certificate number H10-01404.    	   iv	  Table of Contents Abstract ....................................................................................................................................... ii	  Preface ........................................................................................................................................ iii	  Table of Contents ....................................................................................................................... iv	  List of Tables .............................................................................................................................. x	  List of Figures ............................................................................................................................ xi	  Acknowledgements ................................................................................................................... xii	  1. INTRODUCTION .................................................................................................................... 1	  A gap in understanding ............................................................................................................... 1	  Structure of thesis ....................................................................................................................... 5	  2. RESEARCH QUESTIONS, OBJECTIVES AND FINDINGS .......................................... 10	  Theoretical considerations ........................................................................................................ 11	  Water management in Tibetan societies .............................................................................. 11	  Intersectional analysis of gender and other dimensions of social difference ...................... 12	  Approaches to the study of resource management .............................................................. 15	  Methodology ............................................................................................................................. 18	  My past work and later training at UBC ............................................................................. 19	  Outsider yet insider: reflections on my position as a researcher in Spiti ............................ 25	  Fieldwork in Spiti Valley .................................................................................................... 31	  Trips to all the villages of Spiti Valley ................................................................................ 35	  Interview and participant observation data: from frustration to insight .............................. 37	  3. VILLAGE SOCIAL STRUCTURE: A HISTORICAL BACKGROUND ........................ 43	  Introduction ............................................................................................................................... 43	  	   v	  Spiti Valley: the setting ............................................................................................................. 45	  Social stratification and socio-economic structure of Spiti Valley ........................................... 51	  Khangchen households ........................................................................................................ 54	  Dhutul households ............................................................................................................... 56	  “Caste” or Marab households .............................................................................................. 58	  Zo blacksmiths ........................................................................................................................ 60	  Beda musicians ....................................................................................................................... 62	  Historical background to local social structure and relations ................................................... 64	  Origins of household system of taxation during the Tibetan Empire .................................. 68	  Functional similarities: Khangchen households as military regiments .................................. 71	  Structural similarities: administrative organization of the Tibetan Empire and Spiti ............ 73	  Council meetings (‘Dun ma) of Spiti and the Tibetan Empire ............................................... 75	  Fire-raising stations: a military alert system in Spiti and of the Tibetan Empire ................... 76	  Other similarities .................................................................................................................... 78	  Cross checking hypothesis in the history of Spiti ............................................................... 79	  Administrative power structure in the periphery: Ladakhi and British rule ........................ 81	  Impacts of colonial British rule .............................................................................................. 85	  Socio-economic development of a frontier region under independent India ...................... 87	  Conclusion ................................................................................................................................ 95	  4. INEQUALITIES WITHIN WOMEN-MANAGED IRRIGATION SYSTEMS ............... 99	  Introduction ............................................................................................................................... 99	  Irrigation system of Spiti Valley and its social organization ............................................ 104	  Section 1: History of water management customs and their implications on society ............ 106	  Section 2: The codified customary laws of irrigation in Zibug and Khyung villages ............ 113	  	   vi	  Document 1: Riwaj-i-Abpashi of Zibug village: ............................................................... 115	  Document 2: Riwaj-i-Abpashi of Khyung—Hara fields: .................................................. 115	  Document 3: Riwaj-i-Abpashi Khyung—Dangpo (Source 1) and Nyipa (Source 2) fields: ........................................................................................................................................... 116	  Analysis of Document 1: codified irrigation customs of Zibug village ............................ 118	  Analysis of Document 2: Riwaj-i-Abpashi of Khyung Hara fields .................................. 121	  Analysis of Document 3: codified irrigation customs of Khyung’s Dangpo and Nyipa fields ........................................................................................................................................... 124	  Section 3: Irrigation-related activities and roles in a calendar year ........................................ 129	  Serpent spirit cake ritual (klu gtor) .................................................................................... 131	  Drawing of irrigation channels in the field (rnang slang) ................................................. 133	  Yurshag Shichu (yur zhag bzhi bcu) or 40 days to irrigation ............................................ 135	  Irrigation channel repair ....................................................................................................... 136	  Collection of festuca (rtsa skya) ........................................................................................... 138	  Watering of fields .............................................................................................................. 139	  The first watering of the year: Yurma .................................................................................. 140	  Kirzin (skar ‘dzin), the irrigation inauguration ritual and the exclusion of caste members . 142	  The order of performing Yurma in the village ..................................................................... 145	  The second watering: Rhagti ................................................................................................ 147	  The water source ritual: Chutsa (chu rtseg) .......................................................................... 148	  The third and fourth watering: Sumti and Zhiti .................................................................... 149	  Determination of watering turns (chu res) and regular watering .......................................... 150	  The germinating water: Minchu (smin chu) ......................................................................... 151	  Conclusion .............................................................................................................................. 152	  	   vii	  5. GENDER, CASTE AND CLASS ROLES IN FARMING ACTIVITIES ....................... 155	  Introduction ............................................................................................................................. 155	  Theoretical considerations: embeddedness ............................................................................. 156	  Gender, agrarian socio-economic class and farming labor ..................................................... 159	  Section 1: Class, caste and gender in farming ........................................................................ 165	  1.1 Farming and labor for Khangchen householders: the rich peasants ............................ 165	  1.2 Farming and labor for Dhutul householders ................................................................ 167	  1.3 Farming and labor for Zo householders ...................................................................... 169	  1.4 Farming labor for Beda householders ......................................................................... 171	  Section 2: Farming related activities of a year-cycle .............................................................. 173	  2.1 Warm-sun month (‘de brgya zla ba) ............................................................................ 177	  2.1.1 Dachang (mda’ chang), the “arrows and beer” festival: .............................................. 177	  2.1.2 Beer for guests (mgron chang) .................................................................................... 179	  2.2 Cultivate-the-fields month (zhing ‘debs zla ba): April-May ....................................... 180	  2.2.1 Scattering ash ............................................................................................................... 180	  2.2.2 Picking of rocks, ensuring moisture in soil and mixing manure ................................. 181	  2.2.3 Springtime circumabulation and purification (bum ‘khor) .......................................... 182	  2.2.4 Cutting of yak hair and seeding ................................................................................... 183	  2.2.5 The first plow: Rolha (rol kha) .................................................................................... 184	  2.2.6 Plow training (gnya’ byang) ........................................................................................ 186	  2.2.7 Work on: (sder kya?) ................................................................................................... 187	  2.2.8 Leveling the fields (sha la) .......................................................................................... 187	  2.2.9 Drawing of irrigation lines (sna lang) .......................................................................... 187	  2.3 Blue skies green earth month (gnam sngon sa sngon zla ba): May-June .................... 187	  	   viii	  2.4 Flower month (me tog zla ba): June-July .................................................................... 188	  2.4.1 “Purlog” weeding ........................................................................................................ 189	  2.4.2 “Bumkhor” circumambulation .................................................................................... 189	  2.5 Seed-ripening month (‘bras bu zla ba): July-August ................................................... 190	  2.6 Harvest month (sel zla ba): August-September ........................................................... 191	  2.6.1 Deity propitiation ritual (lha gsol) ............................................................................... 191	  2.6.2 Namgan harvest festival .............................................................................................. 191	  2.7 Threshing month (khu yu zla ba): September-October ............................................... 193	  2.8 Shorter days month (nyi rug zla ba): October-November ........................................... 195	  Concluding discussion: ........................................................................................................... 197	  6. CUSTOMARY REGULATIONS FOR THE COLLECTION OF FODDER, DUNG AND FIREWOOD .............................................................................................................................. 201	  Introduction ............................................................................................................................. 201	  Fodder, dung and firewood in Spiti ........................................................................................ 203	  1: State laws related to Spiti’s commons resources from the 19th to the 20th century ............ 211	  1.1 Under Ladakh rule (Until 1842) .................................................................................. 212	  1.2 Laws under British rule ............................................................................................... 218	  1.3 Laws under the Indian Government ............................................................................ 221	  2: Village customs of managing FDF collection .................................................................... 227	  Collection of FDF in a calendar year ................................................................................ 228	  Conclusion .............................................................................................................................. 238	  7. CONCLUSION ..................................................................................................................... 242	  Studying different resource sectors relationally ..................................................................... 242	  	   ix	  Theoretical implications .......................................................................................................... 251	  Other research recommendations ............................................................................................ 261	  REFERENCES .......................................................................................................................... 264	  APPENDIX 1: LOCAL MONTHS AND CALENDAR SYSTEM OF SPITI VALLEY ... 296	     	   x	  List of Tables Table 1: Social stratification in Tibet and Spiti ............................................................................ 53	  Table 2: Social stratification in Spiti Valley ................................................................................. 54	  Table 3: Historical timeline of Spiti Valley .................................................................................. 66	  Table 4: Composition of Hara irrigation users ........................................................................... 122	  Table 5: Main agricultural activities (March-July) ..................................................................... 175	  Table 6: Main agricultural activities (July-November) .............................................................. 176	  Table 7: High status farming roles .............................................................................................. 196	  Table 8: Low status farming roles .............................................................................................. 197	  Table 9: Some types of dung used in Spiti ................................................................................. 206	     	   xi	  List of Figures Figure 1: Geographical location of Spiti Valley (Map source: www.demis.nl) ........................... 47	     	   xii	  Acknowledgements  During the course of my study and research, I have greatly benifited from many individuals and scholars who have generously provided support and shared their knowledge and expertise of the field. This study also benefited through the critical engagement of fellow graduate students as well as through the guidance of many experts.   I must begin by thanking all the people of Spiti who have contributed to this study and tolerated my intrusion into their daily lives. When I first visited Spiti, Lobsang Gelek took me to different corners of the valley and made sure that I had every possible reason to choose Spiti as my fieldsite. I thank Lobsang for his keen support of my research and for introducing me to the people and places of Spiti. Once my research started, the family members of Amchi Ngawang Tsering, Ngawang Samdup and Dr. Tsering Norbu were incredibly generous and supportive of the project. I cannot thank them all enough. During the course of my research in India, many people received me kindly and provided assistance: Kalzang Chokit, Rigzin Palmo, Kunsang Lhamo, Kalsang, Lobsang, Tanzin Dawa, Meme Teku, Lama Kalu, Chering Rapten, Evi Tashi Yangzom, Lossar Rinchen Dhondup, Phuntsok Rai, Drimed Lodoe, Thinley, Dorje Bodh, Amchi Tsering Tashi, Lari Kusho Lama, Munshi Padma Dorje, Meme Tashi Dhondup and Lama Dorje Busso. I am also grateful to Lochen Rinpoche, Nono Sonam Angdui, Tsering Dorje, OC Handa and Vidya Sagar Negi for their advice and assistance regarding my research.  Here in Canada, my deep respect and heartfelt thanks goes to my wonderful thesis committee members: Tsering Shakya, Gunilla Oberg and Leila Harris. They come from very different 	   xiii	  disciplinary backgrounds, which worked wonderfully in the case of this interdisciplinary study. Their critical input and helpful suggestions on the many draft chapters have significantly improved the quality of this thesis. When I first joined UBC, I was initially planning to do a  study of international rivers. I am grateful to my former committee members, Richard Paisley, Hans Schreir, Pitman Potter and Milind Kandlikar for their support and understanding. I must also thank Felice Wyndham for serving on my committee for a year. I am grateful to Dan Martin, Brandon Dotson, Melvyn Goldstein and John Bray for their critical feedback and suggestions on the section concerning the historical origins of Spiti Valley’s Khangchen system in Chapter 3.  Funding for this study mainly came from UBC through the University Graduate Fellowship (2007, 2008 and 2009) and through the Dalai Lama Trust Graduate Scholarships (2010, 2011). In addition, I am grateful for funding support from Richard Paisley, Pitman Potter, John Lefebvre, D’Arcy Richardson, Isabelle Henrion-Dourcy and Tsering Shakya. Travel grants from UBC COSMOS International Research Travel Award, UBC Center for India and South Asia Research and UBC RMES helped pay for research and conference expenses.   Lastly, I want to thank my wife and family, whose love and support made it possible for me to complete this project. 	   1	  1. Introduction A gap in understanding On a crisp early morning in a Himalayan Buddhist village in Spiti Valley, a group of women—young and old—gathered near the village irrigation reservoir to water the fields. After some time, one of them released water from the reservoir by removing a rock that was used to block the outlet hole. Water gushed out of the reservoir and poured onto an unpaved rocky surface that branched into several channels. Some women rushed to their fields while others stayed on to ensure that an adequate share of water entered the channels that supplied the fields. Two of these women had a brief argument as one of them used her adze hoe to push a loose rock near one channel mouth and the rock affected the flow of water into another channel. Another woman threw a large rock, the size of a soccer ball, into a newly constructed concrete channel to divert water into a side channel. The chaos that ensued after the opening of the reservoir subsided as the reservoir became empty and the women dispersed to water the fields. (Field observations, Tsering, Spiti 2007)  The group of women I witnessed that morning in 2007 belong to a historically landless class of households known as Dhutul (smoke-maker: dud = “smoke”, ‘thul = “to make”). Traditionally, the land belonged to the Khangchen (Large Household, khang = “house”, chen = “big”), who still own the majority of the land in the Spiti Valley. While the Dhutul have access to irrigation water for only one day in a six-day cycle, the Khangchen 	   2	  households have access five out of the six days.1 The socio-economic effects of this uneven distribution of land and irrigation water are amplified by the fact that the 50 Dhutul households significantly outnumber the 13 Khangchen households.2 In this community, there are also outcaste households, some of which do not have any right to communal irrigation water.3 The effects of these class-based inequalities are entangled with gender-based differences: while women do most of the irrigation and farming work, the land and water rights are legally held by men.  This picture of inequality, characteristic of the irrigation system of a Spiti village, is starkly different to past field research in the region – which describes local irrigation practices and systems as equitable (Gutschow, 1997, 1998; Labbal, 2000; Mankelow, 2003; Tiwari & Gupta, 2008). These studies have described the local irrigation system as equitable, mainly because of features such as the proportional allocation of rights and responsibilities between farmers according to their landholdings or the size of their field (i.e., the larger the field, the greater the rights as well as the responsibilities), transparency in allocation (farmers can see how much water is being used and by whom), 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  1 When I observed the incident described above in 2007 there were 50 Dhutul households and 13 Khangchen households. In 2010, the number of Dhutul households sharing irrigation water had increased. The number of Khangchen households, however, remained the same. This case study of inequality in Zibug village in Spiti is far more extreme than other villages in the region, compare for example to that of Rinam village in Zangskar (Gutschow & Gutschow, 2003, p. 130).	  2 There is one main reservoir in the village, which does not contain sufficient water for all the fields. The reservoir is emptied each day to irrigate the fields on a turn-by-turn basis. Water distribution is organized into a six-day cycle; that is, every farmer gets water on every sixth day. For 5 days, water is shared among the Khangchen households (3 Khangchen households receive water each day). The water on the sixth day is divided among 50 Dhutul households. Therefore, in terms of quantity, Khangchen households get approximately 1/3rd (33%) of the daily stored water whereas a Dhutul family’s share is 1/50th (2%).	  3 As explained elsewhere in this study, the notion of “caste” in Spiti is different from that of the Hindu caste system. In Spiti, “caste” refers to two groups of households (i.e., musicians and blacksmiths), who enjoy lower social status as compared to other household types (i.e., Khangchen and Dhutul). Traditionally, Beda (i.e., musician) households did not own any farming land. It was only after the introduction of the 1969 Nautor land distribution program that they were granted the right and opportunity to own their own land. 	  	   3	  and autonomous decision-making (farmers make decisions about the irrigation system themselves). These features are not unique to the Himalayan region nor to Tibetan Buddhist societies in general.  Other “traditional” or “indigenous” irrigation systems, in contrast to state managed systems, have also often been described as models of an equitable system (e.g., Coward, 1979; Mabry & Cleveland, 1996; Trawick, 2001, 2003).4 While farmer-managed irrigation systems are generally more equitable in contrast to those managed by the state, I argue that allocation, decision making, as well as other aspects of irrigation practices at the village level are also inequitable in light of gender, socio-economic class and caste differences. For example, the principle of proportionality observed in many farmer-managed irrigation systems refers to the proportionality of farmers’ rights and responsibilities within the irrigation system according to their landholding only, and thus overlooks the un-proportional rights and responsibilities based on their gender, caste and other social differences. Similarly, aspects of local irrigation practice, such as autonomous decision-making and transparency in water distribution, have also been seen as markers of equity that have nothing to do with internal social differences between farmers. While feminist scholars have debunked the assumptions of equitable irrigation and resource management systems (discussed in the next section), the assumption still persists in Tibetan Studies.5   	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  4 There are traditional irrigation systems where access to irrigation water is not equitable. For example, Haagsma (1995) notes that land distribution and access to irrigated land is “very inequitable” (p. 46) in Ribeira de Duque on the island of San Antao, in Cape Verde, where water rights are not based on plot size. As Haagsma speculates: “Probably, historical factors play a part. One supposition is that land owners have increased their area under terraces in the past hoping for a better water supply, but without formally acquiring water rights” (p. 46). 	  5 The assumption also extends to other resource management practices (see, for example, Mishra, Prins, & Wieren, 2003) as well as to agrarian life (Norberg-Hodge, 1991) in general. 	  	   4	  This lack of attention to the social aspects of equity is pervasive in many studies of farming and resource management practices on the Tibetan Plateau. This is a significant gap, particularly given that water and related resource issues have commanded significant scientific and international media attention. Social science studies of irrigation and related farming and resource management in the Himalayas and the Tibetan Plateau are relevant to broader international interests in water and related climate change. This is particularly important in the context of the Tibetan Plateau, including the Himalayas, as these areas are the sources of many major rivers that flow into about a dozen countries, including India and China.6   This study provides a detailed historical, ethnographic and intersectional study of gender, caste and class roles of farmers living in Spiti Valley in the Indian state of Himachal Pradesh. This study, to my knowledge, presents the first detailed intersectional analysis of inequality in the social organization of irrigation and, by extension, in farming and related resource management practices on the Tibetan Plateau. It analyzes power relations, as they pertain to local livelihood activities, between farmers in terms of their gender and other social differences. As such, this analysis responds to calls for the examination of inter-user group dynamics (Beck, 1999, Agrawal, 2003; Agrawal & Sivaramakrishnan, 2000), particularly at the intersections of gender and other social differences (Harris, 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  6 The Tibetan Plateau is the world’s highest (with 4000 m being the average altitude) and largest (2.5 million square kilometers) plateau that serves as the source of the largest constellation of international rivers (Indus, Satluj, Brahmaputra, Mekong, Yellow, Yangtze, Irraweddy, Salween) in the world. Downstream countries include Pakistan, India, Nepal, Bangladesh, Thailand, Burma, Cambodia, Laos, Vietnam and China. The plateau is also now referred to as the world’s “Third Pole” – because it has the largest ice fields outside of the Arctic and Antarctic. Downstream supply concerns are mainly focused on the impact of climate change on the glaciers that feed the rivers and upstream water control, as well as how these are related to certain development projects. This is an area that I have personally researched and contributed to since 1998. A brief discussion of this work is provided in this chapter.	  	   5	  2008; Nightingale, 2011; Hawkins, et al., 2011), within historical and other regional contexts (Agrawal & Sivaramakrishnan, 2000).  Here it is important to briefly clarify how this study uses the term “equity” as different from that of “equality.” Within the context of resource management, “equality” refers to a conditon when all users have the same rights and responsibilities in accessing and managing local resources. As the case of Spiti village demonstrates, the power relations between different user groups are not equal due to social differences in terms of their gender, caste and class. This study uses the terms “(in)equity” and “(in)equitable” to highlight and examine structural inequalities of power and privilege between different user groups as they access and manage local resources.  Structure of thesis This thesis has five main chapters that show how the social organization of irrigation, farming and resource management practices are linked to each other and are embedded in power relations between the different social groups residing in Spiti. Chapter 2 describes the research questions, presents a literature review, and discusses the methodology of the study. The remaining main chapters (Chapters 3 to 6) have three common descriptive and analytical considerations. First, they probe into the relevant historical and legal contexts to shed light on the evolution and changing nature of the relationships between different user groups (chapter 3), specifically regarding irrigation management (chapter 4), farming (chapter 3 and 5), and fodder, dung and firewood (FDF) management (chapter 6). Second, all the chapters provide an intersectional analysis of gender, caste and class 	   6	  relations in terms of broader social relations (chapter 3), irrigation management (chapter 4), farming work (chapter 5), and FDF management (chapter 6). Third, all the descriptions of historical contexts and resource management practices follow a chronological order. These descriptions, although provided in separate chapters, can be overlaid to provide broader narratives of (a) the history behind local resource management customs and (b) farmers’ main livelihood activities from early April to late October. Below follows a brief description of each of these chapters.  Chapter 3 describes the setting and context of the study within the broader regional history and cultural geography of Spiti. This description informs how Spiti’s social structure and local power relations have been shaped by the policies of different regimes that ruled the region, which in turn were affected by Spiti’s peripheral location. Specifically, I trace the origin of Spiti’s Khangchen households as “tax payers” (Treba, khral pa) to the period of the Tibetan empire of the 7th-9th centuries. In order to support this hypothesis, I draw on the relevant history of the region as it is portrayed in English, Tibetan and Hindi language sources, which include relevant ancient Tibetan texts and studies on primary source materials. I also draw from knowledge acquired from my fieldwork, including interviews with farmers concerning the history of the region. This hypothesis provides a specific historical and legal context for understanding the origins of the privileged powers of Spiti’s Khangchen households as the sole group of farmers who own village agricultural fields and irrigation sources.7 The discussion then moves on to later regimes. First it provides a description of Spiti’s administrative structure under 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  7 By linking the origin of the institution of Khangchen system to the period of the Tibetan empire, this hypothesis also extends the current scholarly understanding of Spiti’s history, which is said to begin from the 10th century (Lahuli, 2002; Jahoda, 2009; Tsering, 2013).	  	   7	  Ladakhi rule. Then it discusses changes in socio-economic conditions and power relations among farmers that occurred during the British rule and later under Indian rule. This discussion shows how the seemingly rigid local socio-economic and power structures are always changing in response to larger political, economic and social forces.  Chapter 4 provides an intersectional analysis of the gender and equity dimensions of irrigation management practices and laws in the two case study villages. The Spiti case study is unique because irrigation is the general domain of women, whereas in many other parts of the world, irrigation is mainly the domain of men or of both men and women. The case study thus demonstrates the significance of intersectional analysis of equity issues among different groups of women, especially as contrasted with the common ‘men versus women’ approach. The study provides a detailed analysis of the related history, legal texts and farmers’ practices, to shed light on issues of equity between different user groups. This study, however, does not shed sufficient light on the relational logic by which different groups perform their roles. By this I mean that a descriptive understanding of irrigation roles and activities, including broader historical and legal contexts, does not explain the ongoing participation of marginal user groups in the irrigation system. These explanations are explored in the following chapters.   Chapter 5 consists of two sections. The first section describes how men and women of the different classes of households are responsible for different farming-related activities. The second section then goes on to demonstrate how these different roles come into play through a chronological description of all annual farming-related activities. These 	   8	  descriptions are provided with an underlying objective to demonstrate how the irrigation-related roles are part of a broader division of labor in farming-related activities and are embedded in local social structures and power relations. Building on the last chapter’s intersectional analysis of gender in irrigation management, this chapter further explicates how gender is imbricated in all levels of the social structure and in farming-related roles, in particular. This chapter forwards three key arguments: (i) the division of agricultural labor is based on class, caste and gender relations; (ii) gender not only dwells within and outside the household (Agarwal, 2007) but is also imbricated in all levels of the social structure in terms of men and women’s agricultural labor roles; and (iii) the different types of farming-related roles ascribed for men and women appear to be equitable in terms of number but are in fact highly unequal in terms of the desirability and prestige associated with the roles.   Chapter 6 builds on the previous chapters to show that the social organization of irrigation and farming-related activities are not only linked to each other but also to the management of the other resources of the village (in this case FDF). The chapter provides a chronological description of customary practices and regulations related to the collection of FDF. It shows how FDF regulations are integrated with broader farming practices in terms of the timing of the regulations and the management roles and how this has different implications for different user groups. It also shows how, for example, women’s role in irrigation is related to men’s role in collecting traditional fuel sources. The first section of the chapter continues the thrust of historical analysis used in the study (especially in chapters 3 and 4) by examining how the laws and policies of different 	   9	  regimes (mainly Ladakh, British and Indian) affected different user groups in terms of their access to these resources. The next section provides a chronological description of FDF collection practices and regulations from the beginning of the growing season to the end of the harvest season. This case study shows how traditional resource management customs that are benign and equitable on the surface can in fact be unfair and constitute “very dangerous techniques” of controlling labor and access to resources. Confirming those arguments forwarded by intersectional analysis, this chapter unravels how one dominant group of households has always benefited from these customs and how others, mainly poor households, continue to be disproportionately affected by the same customs.   In the conclusion, I address how the central argument of this study, along with the consequent approach undertaken in the main chapters, provide answers to my three research questions (detailed in Chapter 2). In addition, I will explicate the key functional linkages between the different resource sectors that are observed in this study. Finally, I present the main theoretical and policy implications of the study.   	   10	  2. Research Questions, Objectives and Findings There are three main objectives of this study. First, the study seeks to understand who (e.g. according to one’s caste, class and gender background) does what (which activity) in the irrigation system of Spiti Valley. Second, it examines factors that keep such an inequitable irrigation system functioning. Third, it examines the social organization of the irrigation system within its historical context. These objectives inform the three main research questions of the study:   1) How do farmers in the Spiti Valley allocate and manage irrigation water based on caste, class, gender, and linked differences?  2) What factors and socio-economic processes help explain the ongoing participation of disadvantaged groups?  3) How did the farmers of Spiti Valley come to practice such an inequitable irrigation system? This question can be broken down into two parts: a) When and how did Khangchen households became the sole group of users with legal rights to village irrigation sources and farming lands? b) How have contemporary inter-user group relationships evolved historically?  The study is informed by the theoretical framework of “agrarian environments” (Agrawal & Sivaramakrishnan, 2000) and directly relates to recent theoretical debates and approaches suggested by feminist political ecologists (e.g., Gururani, 2000; Harris, 2008; Nightingale, 2011; Hawkins et al., 2011). The framework of agrarian environments is chosen because it is regionally focused on South Asia and explicitly 	   11	  emphasizes the historical and political dynamics of resource use, control and conflict within communities in terms of local gender, class and caste relations, and how these relations deal with the larger forces of state, economy and environment.8 While the body of work represented in agrarian environments focuses on India, similar theoretical approaches have been used more broadly by feminist political ecologists (e.g., Harris, 2008; Nightingale, 2011; Hawkins et al., 2011).  The main theoretical finding of this research is that the social organization of irrigation management between farmers – which is based notably on gender, caste and class roles - is best understood as part of a broader social division of labor developed for farming and related resource management practices. It is within this understanding, along with an understanding of how these relationships have evolved due to changing historical and political conditions, that this thesis sheds light on all of its research questions.   Theoretical considerations This study directly contributes to research and interdisciplinary debates on water management in Tibetan society, on approaches to the study of resource management in rural contexts, and to the project of intersectional analysis of gender and other social differences in resource management.  Water management in Tibetan societies Research and studies on irrigation management in Tibetan societies can be divided into 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  8 This study uses the word “State” loosely to encompass pre-modern states or kingdoms (Ladakh), colonial powers (British), as well as modern states (India). 	  	   12	  three types based on periods of study and sources of data (Shakya, 1994).9 The earliest studies of water management in Tibetan areas are based mainly on travel books and gazetteers (Bell, 1928; Carassco, 1959) but suffer from a lack of field research access in Tibet. The second phase of studies also lacked fieldwork access in Tibet but were nevertheless unique as these were based on interviews conducted with Tibetan refugees in Nepal and India (Aziz, 1978; Ekvall & Downs, 1963), who arrived after China’s takeover of central Tibet in 1959. The third and most recent phase of studies on irrigation, to which this study directly contributes, is based on fieldwork in Ladakh after the opening of the region for tourism in 1974 (Norberg-Hodge, 1991; Gutschow, 1997, 1998; Labbal, 2000; Vohra, 2000; Gutschow & Mankelow, 2001; Gutschow & Gutschow, 2003; Mankelow, 2003; Tiwari & Gupta, 2008). These studies, which are highly relevant to Spiti because of its geographical and cultural proximity to Ladakh, include comparative case studies (Gutschow, 1998), analyses of government watershed programs (Mankelow, 2003; Gutschow & Mankelow, 2001), descriptions of the effects of globalization and tourism on irrigation (Tiwari & Gupta, 2008), and case study descriptions of irrigation practices, customs and taboos (Vohra, 2000).   Intersectional analysis of gender and other dimensions of social difference  Analysis of inequality in irrigation systems has by no means been ignored in literature outside the field of Tibetan Studies, but the number of such studies remains very small beyond those that focus on gender (e.g., Moser, 1993; Levy, 1996; Jordan & Zwarteveen, 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  9 Many people have written about irrigation in Tibetan societies. See, for example, Bell, 1928; Carrasco, 1959; Ekvall & Downs, 1963; Aziz, 1978; Norberg-Hodge, 1991; Gutschow, 1997, 1998; Labbal, 2000; Vohra, 2000; Gutschow & Mankelow, 2001; Mankelow, 2003; Sorenson, 2004; Angchok & Singh, 2006; Gupta & Tiwari, 2008; etc.	  	   13	  1997; Meinzen-Dick & Zwarteveen, 2001; Zwarteveen, 2006; Udas & Zwarteveen, 2010). Although it is commonly acknowledged that gender operates in interaction with other social categories such as class, caste and ethnicity (Meinzen-Dick & Zwarteveen, 2001; Wallace and Coles, 2005),10 there is still an ongoing need to further intersectional analyses of these issues in ways that emphasize how differences interact to reinforce one another, whether in the realm of development (Nightingale, 2011), or in terms of irrigation in particular (Harris, 2008).   Intersectional analysis takes analysis of inequality in irrigation and resource management beyond binaristic understandings of male-versus-female. Instead, such an analysis posits that although male-female based inequities represent an important issue, an examination of this issue is not sufficient to understand and address issues of inequity because sex-gender is only one of the many important intersecting layers of social relations. Intersectional theory argues that – and aims to show how – gender and other categories of social difference, such as race, class and ethnicity, do not act independently but interact with each other on multiple and often simultaneous levels, thereby ultimately contributing to systematic social inequalities. This analysis focuses on how people’s different identities or roles interact with each other and thus result in multiple advantages for some individuals and multiple disadvantages for others, with many others experiencing a mix of advantages and disadvantages. For example, it posits that experiences of an economically poor black women cannot be understood by only considering her experiences as economically poor, as a female, or as a black women, as if these categories were independent of each other (Crenshaw, 1991). Instead, intersectional 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  10 For an excellent discussion of debates in intersectionality in feminist literature, see Fincher (2004). 	  	   14	  analysis attempts to understand her experiences by considering how her different roles or identities interact and reinforce each other, thus putting her in a position where she faces multiple dimensions of disadvantage (i.e., of class, gender and race). In this way, intersectional analysis also avoids problems of assuming that all women face similar situations, as it is often the case that certain high-income or elite women may be in a very privileged position, particularly in comparison to lower caste women. The intersectional analysis of this study emphasizes both intra-category (different groups of women) and inter-category (gender, caste and class) differences (McCall, 2005).   This study will therefore conduct an intersectional analysis of equity in irrigation and farming-related resource management practices and analyze how certain user groups, such as Khangchen men, are multiply advantaged in these activities while others, such as women of outcaste households, face multiple disadvantages. Specifically, it has been pointed out that the intersection between gender and socio-economic class (Sultana, Mohanty, & Miraglia, 2013) or poverty (Harris, 2008), and observations of “ways in which boundaries between bodies, spaces, ecologies and symbolic meanings of difference are produced and maintained relationally through practices of work and ritual” (Nightingale, 2011, p. 153) are important areas for analysis and theorization. Building on these debates, this study unravels, for example, the covert but deeply significant ways in which members of underprivileged groups, especially those locally deemed “impure” (i.e., women and caste members), are discriminated against in ritual aspects of farming and related resource management practices.  For example, in chapter 4, I have identified the irrigation inaugural ritual as an activity that is always undertaken by non-caste members - 	   15	  although there are no explicit customary rules that bar caste (women) members from performing the ritual.    There are three key benefits of the intersectional analysis used in this study. First, as mentioned above, it provides a methodology of analyzing how different types of social inequalities interact with each other. Second, the study directly contributes to the lack of focus on inter-user group dynamics in the literature on irrigation (Beck, 1999; Harris, 2008) and resource management studies (Agrawal, 2003; Nightingale, 2011). Third, the understanding of how gender and other social differences interact to reinforce inequitable social relations in resource management and farming-related activities has important policy implications. This analysis contributes to policy interventions that argue that, unless development initiatives are tilted in favor of the poor, women and marginalized groups, such interventions are likely to compound local inequities (Beck & Fajber, 2006; Ribot, 2008). For example, as shown in the following chapters, government socio-economic development interventions in Spiti Valley have benefited different user groups differently; traditionally privileged groups, such as men of Khangchen households, are at an advantaged position to avail themselves of economic development opportunities because of their having better education and social connections, especially compared to the traditionally disadvantaged groups such as women and outcaste members.   Approaches to the study of resource management In recent years, political ecologists have made a set of arguments against the two main schools of thought that explain cooperative communal action in irrigation management. 	   16	  The first school, following Scott’s (1976) “moral economy of the peasant” argument, assumes that farmers cooperate out of a moral conscience arising from the small community’s need to cope with risks as well as their collective dependence on local resources, both of which are institutionalized in customs and tradition (Cleaver, 2000). The second school follows the work of Elinor Ostrom and uses an institutional-economic approach to study the cooperative management of irrigation by farmers. Both schools of thought tend to neglect issues of local politics within rural communities and also to overlook how these communities are changing in relation to broader regional contexts. In response, political ecologists have argued that rural communities are better understood not only within their historical contexts, but also within the local specificities of culture, politics and economy as these interact with state and market (Agrawal & Sivaramakrishnan, 2000; Agrawal, 2003; Mosse, 1997). This study contributes to this debate by applying the theoretical framework of “agrarian environments” (Agrawal & Sivaramakrishnan, 2000) to the study of irrigation in the Spiti Valley. This framework emphasizes the historical and political dynamics of resource use, control and conflict within communities in terms of gender, class and caste relations, and how the larger forces of state, economy and environment affect these relations. This framework is, as I demonstrate in the following chapters, closely linked to the work of certain ‘feminist political ecologists’ such as Andrea Nightingale, Bina Agarwal, Leila Harris and Shubhra Gururani.  While agrarian environments provides a helpful theoretical framework for uncovering micro-level issues of politics that exist between user groups in relation to macro-level 	   17	  contexts of regional history, politics and economy, it misses a critical intermediary level of analysis: the local economy. As a result, the agrarian environments framework is not sufficient if one seeks to understand why people, especially marginalized user groups, participate in an inequitable resource management system. For example, the answer to why a member of a landless farmer group would participate in an irrigation system that discriminates against her cannot be meaningfully found in macro-level contexts of history, state or the environment. The answer must be found in the material and procedural contexts of the local economy, such as payment for labor services, reciprocity, etc. In order to complement this methodological gap in the agrarian environments framework, and in order to describe the functioning of these relationships across farming and related resource management practices, this dissertation draws on the Polanyian idea of economy as an “instituted process” (Polanyi, 1957).   Polanyi viewed “economic life as a totality of relations and institutions that goes beyond the transactions of goods and services” (Gemici, 2008, p. 19, emphasis in original) and differentiated between formal and substantive meanings of economy (Polanyi, 1957).  This formal approach to economy is based on neo-classical economic theory, which assumes a condition of scarcity of means to satisfy unlimited wants in which rational individuals compete to maximize their interests. The substantive approach, on the other hand, views the economy “as an instituted process of interaction between man and his natural and social environment” (Polanyi, 1957, p. 248). By defining economy as an “instituted process”, Polanyi stressed the economic and non-economic institutions that facilitate the recurrent movement or circulation of goods and services. “The instituting of 	   18	  the economic process vests that process with unity and stability” (Polanyi, 1957, pp. 249-250). This is achieved through three forms of integration: reciprocity, redistribution and exchange: “Reciprocity denotes movements between correlative points of symmetrical groupings; redistribution designates appropriational movements toward a center and out of it again; [and] exchange refers here to vice-versa movements taking place as between ‘hands’ under a market system” (Polanyi, 1957, p. 250). Unlike the formalist approach, Polanyi’s substantivist or anthropological approach is not interested in individual behavior but is concerned, rather, with organized structures, institutions and processes - an approach which is appropriate for this study as it focuses on relations between user groups.   While it is beyond the scope of this study to conduct a thorough examination of Spiti’s village economy as an instituted process, it borrows these theoretical principles to explain (in chapter 5) how different institutions (Chumpa or irrigator, Thongpa or plowman, Mugthul or group labor) bind different user groups across different livelihood (farming and resource management) activities through temporal (repetitive yearly activities), social (reciprocity), and economic (remuneration) processes. Polanyi’s substantivist theoretical approach therefore provides a critical middle level explanation that cannot otherwise be ascertained between the micropolitics and the macro-historical contexts of the agrarian environments framework that is required to answer the research questions.   Methodology All observers of human behavior are conditioned by their own life experiences. In this 	   19	  case, my life experiences as an exile-born Tibetan with significant past involvement in research and international advocacy work on Tibet’s water resource issues has impacted this research, particularly in terms of the choice of research topic and field site. Similarly, my past academic training has had a significant influence on my theoretical position and research approach. Needless to mention, the realities of my fieldwork experiences also sculpted this research by influencing the analytical focus of this research and data collection.   My past work and later training at UBC Before joining UBC in 2006, I was engaged full-time in research and international advocacy work on Tibetan environmental issues that I started in 1998. During that time, although I kept a sustained research focus on Tibet’s water resource issues, my perspective was shaped by advocacy work to promote justice and environmental protection. In particular, there are two aspects of my past research and advocacy work that led me to this study. First, all of my past studies were based on secondary data, such as those published by the Tibetan Government-in-Exile, the Chinese government, the World Bank, and studies presented in peer-reviewed literature. As an exile Tibetan, I had never conducted fieldwork or lived in a traditional Tibetan society on the Tibetan Plateau. I wanted to gain field experience in order to gain a first-hand understanding of Tibetan water resource issues and a deeper understanding of the nature and functioning of Tibetan society. Also, as an advocate of the human development approach to the water resource management issues of the Tibetan Plateau (Tsering, 2003, 2011, 2012b, 2012c), I had long felt that there was a lack of studies that looked into the social aspects of water 	   20	  resource issues on the Tibetan Plateau. As someone who has regularly advocated for the primacy of interests of the local people and their environments in debates over water resource issues, I felt that the lack of detailed information about how they managed water locally was a big knowledge gap. The second aspect of my past work that shaped this study is that I believed in a discourse that assumed Tibetans were ecologically aware and lived in harmony with nature. When I started my research and advocacy work on Tibetan environmental issues, this discourse was widespread among Tibetans (Atisha, 1991; Central Tibetan Administration, 1992 & 2000) and Western supporters (Yeshi, 1991; Rowell, 1990; Aptez & Edwards, 1998) and at the time also constituted a widespread perception of other indigenous peoples such as the Australian Aborigines and the native North-Americans (Ellen, 1986; Brosius, 1999). Consequently, my knowledge and assumptions were shaped by this dominant discourse, as I had not been exposed to critical literature on the topic.   Since coming to UBC, I have read a great deal of literature that directly confronts the assumption that Tibetans (or other pre-modern peoples) are ecologically more noble, such as Brosius (1999), Agrawal and Sivaramakrishnan (2000), Ellen (1986), Huber (1997), Huber and Pederson (1997) and van Beek (1996, 2001). For example, Huber (1997) traces the genealogy of the “Green Tibetans” representation to the mid-1980’s, as a creation of exile Tibetan elites in Dharamsala, the seat of the Tibetan Government in Exile. Huber situates this discourse as a part of the global environmentalist Zeitgeist and the greening of “identities” during the 1980s and 1990s. Perhaps, being “advised by various well-meaning foreign supporters”, exile Tibetan elites not only generated the 	   21	  Green Tibetans representation but also “continue[d] to manipulate and disseminate [it]” as a way to raise awareness about Tibet’s environmental issues and to gain “international sympathy for the cause” (pp. 106-108). In other words, Huber argues that the representation of Tibetans as an environmentally aware people cannot be regarded as a form of benign romantic essentialism; quite the contrary, it was, to use Brosius’ term, a “strategic essentialism” (Brosius, 1999, p. 280) generated to garner international support for Tibet (Avedon, 1984; Gyatso & Shakya, 1998). In brief, my studies here at UBC have taught me the importance of critically understanding the genealogy of historical contexts and causes and have provided theoretical and methodological tools to overcome such assumptions.   Critical study of one’s own culture: dilemmas of the “native anthropologist” The application of critical approaches to the study of Tibetan society was initially a moral dilemma for me. My family and exile Tibetan upbringing have always emphasized the importance of preserving Tibetan identity and culture. Faced with the same dilemma of critical study, a Tibetan friend asked in frustration: “How can we afford to criticize our own culture, instead of glorifying and protecting it, at a time when our nation is undergoing one of the darkest periods in its history?” While I greatly respect his sentiment, I personally feel that a more important question to put forward concerns whether and how this study will benefit the people it describes.   Nationalistic glorification of one’s culture will invariably lead to the attribution of certain reified characteristics - such as the portrayal of all Tibetans as “peaceful Tibetan 	   22	  Buddhists” – in a manner that suggests that such characteristics are natural rather than socially constructed.11 In addition, essentialism often leads to the appropriation of that image as either the subject or object of political action (Brosuis, 1999; Phillips, 2010). For example, the use of contrasting essentialized images of peaceful Tibetans being repressed by violent Chinese soldiers has been a common strategy used by Tibetan rights activists to gain international sympathy. Furthermore, the politics behind these essentialized discourses undermine the role of those people who do not fit neatly into the category, such as the silenced stories or Arrested Histories (McGranahan, 2010) of members of the Tibetan armed resistance against China (Shakya, 1999).   While essentialist descriptions can be attractive for romantic, strategic or nationalist reasons (Brosius, 1999), I believe that, ultimately, a more nuanced understanding will be more useful for the people, even if the description may appear to be critical. Moreover, the approach of this study is not an exercise in the criticism of Tibetan society but rather the implementation of a set of critical theoretical and methodological tools that help us to gain a richer and more in-depth understanding of this society. As well, my own concern for Tibet is born from a broader concern with issues of inequality (and this is precisely the same interest that animates this study) as well as a desire to uncover certain inequalities as they operate across various socio-political dimensions. Political ecology, agrarian environments, and intersectional approaches all aim to help us better understand societal processes so that they might ultimately be changed for the better.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  11 Images that portrayed Tibetans as peaceful and living in harmony with nature (nomads milking their animals, or people praying to mountains, etc.) were common in the brochures and other information materials used by Tibetan rights groups. This is not unique to Tibet but common in broader environmental and indigenous rights movement literature (Huber, 1997; Lopez, 1998; Brosius, 1999).	  	   23	   Another example of how essentialism can be problematic, especially for those who do not fit neatly into the essentialized image, concerns this study in relation to my background as a Tibetan. Anthropological fieldwork has traditionally been a domain of the “Euro-American, white, middle-class male” (Gupta & Ferguson, 1997), individuals who traveled to isolated regions of the world to conduct, as many believed previously, objective scientific study of the native-Others. This essentialized image of fieldwork became “real anthropology”, providing the basis for an idea that all other kinds of ethnographic fieldwork that do not fit this image, such as that done by “native” anthropologists, are less than “real.” This idea has been challenged by a number of scholars (Narayan, 1993; Gupta & Ferguson, 1997; Weston, 1997; Bunzl, 2004). Gupta and Ferguson (1997, p. 16) point out that the problem originates from “ideas about Otherness [that] remain remarkably central to fieldwork ritual.” Echoing this point, Weston (1997) identifies the split between “Other” (the natives) and “self” (ethnographer), or “field” and “home”, as the site where the “virtual anthropologist” is produced. Pointing to the colonial contexts of the image, as well as the distinctions between “real” and “native” anthropologist, Narayan (1993) asks how “native” is “native anthropologist,” and how “real” is “real anthropologist” in today’s much different and fast-changing world?12 Instead of sticking to such inappropriate distinctions, Narayan (1993, p. 682) argues that in today’s changed setting, “it is more profitable to focus on 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  12 Some of the various issues that are completely different in today’s modern global context include the following: it is hard, if not impossible, to find places and cultures that are unknown or isolated; economic globalization and information technology, especially the internet, have greatly transformed societies around the world; the practice and idea of the objective scientific study of cultures has been debunked; the ethnographic gaze is not only used in reference to remote rural societies but also to modern societies and to the discipline of Anthropology itself; there are an increasing number of non-white male anthropologists dominating the field; etc. 	  	   24	  shifting identities in relationship with [the] people and issues an anthropologist seeks to represent.”  Bunzl (2004) highlights an important point about these debates and offers a solution to overcome the key problem of the Self/Other or “real”/“native” divide in anthropological fieldwork. Bunzl notes that “even the most radical attempts to rethink the concept of ‘native ethnography’ [Such as Narayan and Weston] have fallen short” of providing methodological solutions to “the foundational Self/Other divide that organizes classical fieldwork and produces the native anthropologist as a virtual member of the discipline” (p. 436). A solution to this problem, Bunzl suggests, is in the use of the Boasian tradition of fieldwork - which seeks to understand other cultures as products of particular historical developments. This approach dislodges cultural difference from its position as the enabling principle of ethnography and turns it into the very phenomenon in need of historical explanation. It retains anthropology’s empirical focus on the present and helps scholars produce neither native nor non-native history but simply history—in this case the history of a particular present.    Bunzl’s “Neo-Boasian anthropology” resonates closely with the research approach of this thesis and its aim to understand and explicate how Spiti’s contemporary resource management customs and embedded social relations are rooted in its genealogical past and have evolved historically. Bunzl proposes this approach as a way to overcome the difference between the “insider-outsider” dichotomy by producing (historical) explanations that are hidden from both the outsider and the insider. Although this study, 	   25	  like Neo-Boasian anthropology, also presents historical explanations regarding contemporary social relations and practices, when I conducted fieldwork I was not aware of Bunzl’s argument that the epistemological stance of such an approach overcomes the dilemma of the insider/outsider subject positions. Consequently, my fieldwork reflections often consisted of moments when I felt the insider-outsider differences, although not in the context of the “native anthropology” debate but more in the context of local knowledge and connections (i.e., the more the local knowledge and connections one has the more of an insider one is). It is therefore appropriate to share some reflections on the “shifting identities” (Narayan, 1993) that colored my experiences during my fieldwork. In doing this, I acknowledge that I was in some ways an insider, and in some ways an outsider. My goal has been to uncover explanations that have too often remained hidden to both.   Outsider yet insider: reflections on my position as a researcher in Spiti As a researcher from a “foreign country” (farmers generally label all or most foreign countries as “phyi rgyal” and show no interest in naming specific countries) visiting traditional Tibetan Buddhist villages for the first time, I was clearly an outsider when I arrived in Spiti. I had no local knowledge and do not remember having met anyone from Spiti Valley, with the exception of my brother-in-law and his brother, before the commencement of the field research. As a Tibetan born and raised in India, I have long wanted to learn or experience how people actually live on the Tibetan Plateau and what Tibetan Buddhist villages are like. The fact that I spoke fluent Tibetan and Hindi helped me greatly in absorbing new knowledge. At first, I could not follow Spiti’s local dialect. I 	   26	  was keen to learn it and, as it turned out, the dialect was not very hard to pick up as it was similar to the western Tibetan dialect that I speak.13 Right from my first field trip, I enjoyed using the local colloquial expressions that I was learning and kept a list of local words and expressions. My attempts to speak the local language usually had a disarming and humorous impact on the locals, which I found helpful in striking up a conversation or for breaking the ice in the interview process. Similarities in Tibetan cultural contexts also helped me participate in local religious and social gatherings in a more intimate manner than would have otherwise been the case for outsiders. The similarities also extended to more than language, religion and social variables. My name (there is a farmer in Khyung with the same first and last name and several others sharing my first or last name) as well as my physical features (such as eye, hair and skin color) were also similar to those of the locals. Upon at least two occasions, farmers told me that I looked like a local Spiti person (spi ti pa or “Pitiya” in local dialect). In fact, the finest compliment that I received in Spiti was that I sounded like a local person, with a “soft” accent (“lab ‘bol mo drag” in the local dialect).   In Spiti, being Tibetan comes with its own power implications. For hundreds of years, the people of Spiti had looked to Tibet as their cultural and religious center. For this reason, Tibetans are generally regarded as co-religionists (chos grogs) and are treated with respect. On certain occasions this relationship could get awkward, at least for me, when 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  13 My parents and relatives speak a western Tibetan dialect. Although my own Tibetan is a mixed dialect, acquired completely in an exile context, especially in schools which children from all over Tibet attended, my accent still sounds closer to the western Tibetan dialect than to any of the other Tibetan dialects. This became clear to me after a radio interview in 2008. The interviewer, a reporter of Voice of Tibet radio, was a speaker of the central Tibetan dialect (“U kay”) and she remarked that I spoke a western Tibetan dialect (“Toepa Kay”). The main difference between Spiti’s dialect and the exile version of the Toepa dialect that I speak is that the former is more archaic. 	  	   27	  people treated me with a respect that I felt I did not deserve. For example, one time, I was introduced to a group of elders as “His Holiness the Dalai Lama’s man” (rygal ba rin po chen gyi mi). This made many of the elders greet and touch me as if I were a holy person. On another occasion, the principal of a local school, a wonderful man from South India, invited me to speak about the Dalai Lama to his students on the occasion of the Dalai Lama’s birthday. I gladly obliged thinking that I would be speaking to a small group of students in a classroom. The actual event was held out in the open in front of the main school building. A special stage was set up, in the middle of which was a table on which a portrait of the Dalai Lama was kept with a flowerpot on each side. An unlit butter lamp offering was placed in front of the portrait. I was made the chief guest of the school event and was given the privilege of lighting the butter lamp on behalf of the school. First the principal gave a speech for a few minutes from the podium in front of the stage. Then it was my turn. After our talks, students and teachers performed dances and sang songs to us. These incidents made me uncomfortably reflect (internally) on my Tibetan identity and the school principal’s Indian identity as symbolizing the colonial forces to which the native children and staff members performed dances and paid respect.   It is also important to note that I was not a complete outsider. I was partly an insider due to the fact that my sister is married to a man from Spiti Valley. The family members of my brother-in-law were very gracious. For example, they called me a relative (spun kyag), and, on several occasions, I found myself being included in their Nyiring (nye ring) and Phiming (phi ming) kinship institutions.14 Nyiring refers to one’s relatives, by blood and 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  14 Within Nyiring relatives, blood relatives, especially those through the male line (rus or bone), are considered closest. 	  	   28	  by marriage. Phiming includes Nyiring and one’s close network of friends (and their household members). The role of Phiming is to provide social support during sickness, deaths, marriage, etc. and to function as an institution for internal resolution of family disputes, such as the one I will describe below.15   Having these relationships was immensely helpful for this research. My brother-in-law’s family belongs to the (traditionally landless) Dhutul class of farmers. Dhutul households represent an important part of Spiti’s demography, as they had increased exponentially in number and had accumulated substantial socio-economic status in the last half century (Jahoda, 2008). For example, this family includes a national award winning teacher, a local leader representing the Congress Party of India, an Amchi or doctor of Tibetan medicine, the first and only Ph.D. from Spiti, the founder of Spiti’s only elders’ home, and the vice principal of a popular private school, amongst others. My association with this family provided privileged access to information that would not otherwise be available to outsiders.   Since this family belonged to the traditionally landless Dhutul group of households, positioned at the middle of Spiti’s socio-economic structure (described in the next chapter), I felt that I was privy to a unique perspective, which in turn resulted in plenty of insights into local social relations. This middle level vantage point was illuminating for this research in several interesting ways. For example, my Nyiring informants did not 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  15 Kinship terms and institutions of Spiti, such as Nyiring, Phiming and Phaphed are similar to but slightly different from those of the other Himalayan Tibetan areas that I have researched such as pha spun of Ladakh (Crook, 1994, pp. 501-507) and dga’ nye of Dingri (Aziz, 1978, pp. 189-197). For example, pha phed of Spiti is like pha spun of Ladakh, with the exception that a farmer’s pha phed could also include Zo (“caste”) members. Similarly, phi ming is similar to dga’ nye of Dingri. 	  	   29	  hesitate to describe the various injustices and unequal treatments experienced by Dhutul farmers at the hands of the Khangchen farmers. Here, it is important to emphasize that many of them spoke of local farming and resource management customs as not only unequal but also inequitable and unfair. Their subjective perspective of “equity” was expressed clearly with observations such as: “it is unjust” (“drang gsum mi ‘dug” or “drang po mi ‘dug” ) and “[the Village Council] discriminates against the poor” (“med po la dbang btang ‘dug”). Similarly, many of them harbored biases, consciously or unconsciously, against those who were less privileged than them - for example towards the caste members. Some of them told me about local practices of discrimination against caste members as if these were appropriate - as if lower caste community members deserved such treatment.   At the same time, my closer association with Nyiring locals might have been a hindrance in terms of accessing the other groups. For example, one of the Nyiring men, who I am personally fond of, is a monk who had once challenged the decisions and authority of the village council and the village deity. This incident had led to the involvement of the Indian police as well as a declaration made on the part of the village council ordering that the most severe punishment be enacted against the monk: complete social ostracism from the village (me lam chu lam). I think it is possible that my close association with this maverick monk could have affected the way some of the village leaders, mostly Khangchen men, interacted with me. Another example was the fact that most interviewees hailing from caste households could not share their accounts of inequality as openly as Dhutul or Nyiring members did with me. Two of the elderly low caste women 	   30	  whom I interviewed explicitly expressed fear of the implications that the recounting of their personal experiences of discrimination might involve and chose instead to speak about other topics. I think that had I been a Nyiring of a caste householder, or even a female researcher, I would probably have heard more frank accounts from them. Similarly, had I been a Nyiring of a Khangchen householder, I would probably have heard more frank accounts about their privileges and responsibilities towards the village—even if it was not conveyed in those terms.   An incident that occurred in the autumn of 2010 was particularly indicative of my background both as an avid participant observer and as a partial insider of the local community. One of the male relatives of my brother-in-law, here called Sonam, was at the brink of divorcing his wife. One day, male Phiming members of his wife came from their village to Sonam’s village. That evening, Sonam’s male Phiming members gathered at his house. Two middlemen, described to me as those “who should not belong to either side or whose character is trusted by both parties to be unbiased in their roles”, were selected to go back and forth between the two groups, conveying the messages and trying to guide the parties towards a compromise. While this whole process was interesting to me,16 as an ethnographer I was also constantly mindful of issues of observer bias (i.e., 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  16 This process was interesting for many reasons. Briefly, the members of Sonam’s phi ming who gathered together that evening were all men, about thirty of them, representing more than half of the village. They all huddled together in the main room (ma khang or “mother house”). Many of the men were drinking chang beer and having a merry time; perhaps the occasion constituted a social get together for some of them. The number of men gathered also seems to symbolize the stature of the household. The success of the process, measured in terms of how amicable, swift and long-lasting the compromise was, depends greatly on the skills of the middlemen, who at least in this case, were not only conveying messages between the parties but also counseling - in a patriarchal and communitarian language - as well as making personal requests to both sides. The middlemen and the phi ming members also serve as witnesses to the process and as a kind of  social sanction to the outcomes, including the consequences of immediate or future failure of the compromise.	  	   31	  tendency of researchers to see more of what they want to see or are interactively influenced in time).  While my hybrid position as an “outsider-yet-insider” and other shifting identities had their advantages and disadvantages, I believe that I was able to make it more advantageous by constantly practicing reflexive thinking and by taking measures to counter the weaknesses of my own position. For example, I made a conscious effort to develop good relations with those farmers representing underprivileged groups by talking to them and showing interest in their activities. One of the best sites to develop closer relationship with the farmers, especially the underprivileged groups of the society, was the crop fields (zhing kha). Farmers who do the manual work in the fields neatly represented these groups, namely: women from all classes of society and men from Dhutul and caste households. I enjoyed working alongside them: digging dirt, carrying loads of fodder plants, taking out weeds, using adze hoes to clear irrigation canals, and pursuing my own research interests at the same time (i.e., asking questions, taking notes, etc.). Spending time with them in the field, including during the tea and lunch breaks, was helpful in developing friendships and cultivating trust on the part of the women and caste members, and this allowed these farmers to talk more openly with me than they might otherwise have done.  Fieldwork in Spiti Valley It is difficult to conduct fieldwork inside Tibet due to political conditions. This is especially difficult, if not impossible, for me as a Tibetan refugee from India. Therefore, 	   32	  in 2007, I conducted preliminary fieldwork in the Indian western Himalayan regions of Ladakh and Spiti because of its cultural and geographical affinity with Tibet. I visited Spiti Valley three more times (2008, 2010 and 2011) and spent over 6 months in the valley in total.   On my first trip in 2007, I did a larger tour of the Buddhist western Himalayas of India. I did this in order to gain a general understanding of the region’s culture and geography and to choose a specific field site for research. After the tour, I chose Spiti Valley as I observed that farmers’ lifestyle in Spiti is visibly more traditional and less affected by the forces of globalization and tourism as compared to that of Ladakh. For example, except in the villages of Kaza and Tabo, there were hardly any hotels or commercial billboard signs in Spiti’s villages. This isolation is partly enforced by geography and politics. The valley is surrounded by 6,000 m tall mountains on all sides and access to the region was shut off by the Indian government from 1962 to 1993 due to political sensitivities surrounding the Sino-Indian dispute. Most of the tourists who visit the region prefer to visit Ladakh and Zangskar due to better transportation and tourism infrastructure. I also noticed that the whole of the native population of Spiti were Buddhists, whereas in Ladakh, the natives were a mix of Buddhists and Muslims. Although the mixed population of Buddhists and Muslims is an interesting phenomenon, my research interest, for the time being, was focused on Tibetan Buddhist villages. I was also touched by the warm hospitality and expressions of support for my research that I received in Spiti. This support comes partly from and because of my family connections in Spiti (discussed above). Most importantly, the incident that inspired the research questions of this project, 	   33	  described at the beginning of this chapter, was observed in Spiti during this (2007) trip. In 2008, I briefly returned to Spiti Valley to develop personal connections in anticipation of the research. I chose the village of Zibug, where I witnessed the irrigation incident, and the neighboring village of Khyung, as the sites of my ethnographic study.  Most of the ethnographic data was collected during three months in 2010 (August-October) and two months (April and May) in 2011. The 2010 fieldwork was done during the last phases of local agricultural activities and consisted mainly of festivals, rituals and work related to the harvesting of crops. The 2011 fieldwork was done during the beginning of the farming year, during which farmers were mainly occupied with rituals and activities related to tilling and irrigation. This fitted well with the timing of my two previous fieldwork periods in Spiti, both of which were done in the month of July, during which I engaged in farmer’s work in their fields, including irrigation.   The climatic conditions of 2010 slightly affected my research plans and activities. According to local meteorological records, there was 300 mm of rainfall and  >60 mm of snowfall from May to September in Spiti that year. Local elders in Spiti told me that it was the wettest year that they could remember in their lifetime. I overheard villagers talk among themselves that some fields in the village of Gyu were swept by landslides and rising water levels. In the neighboring region of Ladakh, a disastrous flood caused by a cloudburst had killed many people and destroyed numerous houses. While the rains were not as destructive in Spiti, I also heard that many old houses, which are all made of mud, had collapsed.  	   34	   The excessive rains had also largely freed farmers of their irrigation duties. This deprived me of my plans to conduct participant observation of farmers’ irrigation-related activities. However, since the farmers were mostly free and staying indoors due to the wet weather, I had excellent opportunities to conduct interviews. I was also able to travel widely in Spiti to learn more about the region and collect data, which I will return to later.  The rains also affected the harvest season of 2010. The initial part of the harvest season, which is dedicated to peas, was very intense that year. The rains had adversely affected the peas in some fields while in others the pea harvests were bountiful. The perishable nature of the peas, the large sizes of the fields in relation to the available labor, poor coordination between farmers and buyers, as well as poor road conditions due to landslides and floods, rendered this season most stressful for farmers. During these days, as the mother of my host family lamented, “we don’t have time to be bothered even if our heads were on fire.” However, the latter part of the harvest season, which is dedicated to barley, was more relaxed that year. This is partly because barley is a more resilient crop and partly because the extra water had allowed the crop to grow for a longer duration. This proved fortuitous for this research as the harvesting of barley and its threshing-related rituals were done in a less rushed manner, which worked well for me as a researcher interested not only in observing but also in constantly asking questions about every detail, in addition to seeking opportunities for interviews.  When I arrived in Spiti in the beginning of April 2011 for my final round of fieldwork for 	   35	  this project, parts of the fields in Khyung and Zibug were still covered in snow. Work on the fields began within a week as I observed some farmers throw ash and dirt on the snow-covered fields so that the snow would melt faster and the fields would be ready for tilling. As the days went by, farmers became busy with farming-related activities and I had sufficient opportunities to observe who and how farmers tilled, as well as how they sowed seeds and watered their fields. This continued over the next several weeks.   Trips to all the villages of Spiti Valley In 2010 and again in 2011, I traveled throughout Spiti Valley to learn about the valley more broadly and to look for internal differences within the valley (e.g., upper Spiti as compared with lower Spiti). During these trips, I collected basic data from each village, such as the total number of households, the number of old Khangchen households, the number of Beda and Zo17 (the two “caste”) households, the number of village doctors and tantric practitioners, the names and number of village deities, and the approximate area of grazing land or village commons used for collecting firewood associated with each village. During these travels, I had some general questions about Spiti that I was trying to answer.   First, I was interested in learning if the number of old Khangchen households in each village had been a constant or if it had changed. In the course of my research on the history of Spiti Valley and my efforts to understand why Khangchen households enjoy privileged access to irrigation and farming resources, I heard repeatedly from farmers that their elders say “at first, there were 250 Khangchen households and 66 Za households.”  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  17 Zo is spelled bzo pa in Tibetan. Spelling of Beda is unclear. 	  	   36	  Farmers also say that these 250 Khangchen households still exist and have not changed in number. So I collected data on the number of old Khangchen households from every village to see how close or far apart the total number was in relation to the local belief that the total number should be 250. Another question I was interested in concerned whether there are any other households besides the three main household types  (Khangchen, Dhutul and caste households) that I observed in the two villages (here named Zibug and Khyung)18 where I conducted most of my ethnographic research. It was important to know if the social structure and relations that I observed in Zibug and Khyung villages were representative of other villages in the valley. If, for example, other villages have additional groups of households, I would know that the social structure and internal power relations of my case studies could not be described as being representative of the valley. During these trips, I was able to confirm that the village social structure comprised of the three basic household types in the valley.   Second, I was interested in finding out if it was customary throughout the valley to ascribe irrigation-related responsibilities to women. It was important to know this because, in the two village field sites, only women did the work of watering the fields. If irrigation-related roles are allocated to women throughout the valley, then the answers to my questions must also be applicable to and ascertainable from the perspective of all the villages. This study highlights similarities in the significance and requirements of irrigation and firewood collection, and also suggests how these and other features of these activities, including related cultural values, contribute to a gendered division of labor in 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  18 I have used fictitious names, Zibug and Khyung, to the two field site villages in order to protect informants.	  	   37	  which women become responsible for irrigation and men for firewood collection (discussed in Chapter 4 and 6, respectively). These trips helped me to confirm that the tradition of only women doing irrigation work was indeed common throughout the Spiti Valley. My understanding of the gendered nature of irrigation and firewood collection as being related grew stronger through interviews and discussions with farmers, as well as through my examination of British colonial records from the 19th century (discussed in chapter 6).    Interview and participant observation data: from frustration to insight I began conducting interviews from the second week of April 2010. Kalzang, the daughter of my host family helped me update and organize a list of all the people in the two village field sites according to gender and household type: Khangchen, Dhutul, Zo and Beda. My plan was to interview randomly chosen people representing all the gender, caste and class groups (4 men and 4 women from Khangchen households; 4 men and 4 women from Dhutul households; and 3 men and 3 women from Caste households). In addition, I wanted to interview local elders, scholars and representatives of local institutions such as the village deity, village council, and members of the local monastery, as well as the Tibetan medicine practitioners and astrologers. As the research progressed, however, I decided that it was better to be more flexible with the selection of interview subjects. While I was determined to and did indeed interview men and women from all social groups in order to get representative perspectives, I realized that simply getting a set number of interviews from people representing all groups would not be sufficient. Moreover, I was not satisfied with the data I was collecting in many of the interviews. In 	   38	  addition to getting viewpoints from all user groups, I was also interested in digging deeper, not only into issues of micropolitics between user groups but also into history and culture. For deeper issues concerning micropolitics, I had to rely more on people (e.g., male community leaders from Dhutul and Zo caste households) who were not afraid, and who were willing and had the time to share their views. For history and culture, I had to rely not only on elders, scholars and leaders but also on specific villagers who were unique experts in their own ways. For example, I learned many things about local culture and folk views on history from two senior Beda women whose names were not in the randomly selected list. Interviews with the Beda women were far less informative: one of them was relatively new (having come as a bride) to Spiti. I also interviewed, discussed and learned many things from farmers of other Spiti villages. For example, I interviewed a retired school teacher from Kyuling Village, upon a suggestion made by the current Nono (traditional ruler) of Spiti, who also himself provided valuable historical information for this study.   While I collected data on all kinds of topics (i.e., marriage customs, kinship, history, politics, demography, weather, folk stories, etc.), I focused my interviews and analysis on irrigation practices. Towards the beginning of my fieldwork, I was a bit frustrated with the brevity of and similarities in farmer’s descriptions of irrigation management. Farmers have a tendency to treat irrigation management roles as too ordinary and simple to deserve detailed description. Even those who admitted that irrigation is an important and complex task gave descriptions that were straightforward and not new to me. These were concerned with the mechanics of letting water flow from reservoirs into canals and then 	   39	  into the fields, or about irrigation customs and institutions (mainly water turns and serpent spirits). I tried to probe further about issues of conflict and competition between irrigation water user-groups, but farmers always claimed that these issues were non-existent. Even the elder farmers said that they did not remember a single case of internal conflict regarding irrigation water or that of a violation of irrigation customs. This was surprising, as I had assumed that such cases are bound to occur, if not very often, then at least occasionally.   Another unexpected and frustrating experience with the interview data was that I was not getting a clear understanding of how and why, if at all, irrigation tasks are organized according to gender, caste and class roles. The problem began to emerge as it became clear that interview data and field observations were failing to provide an explanation of how irrigation labor relationships are organized on the basis of gender, caste and class. For example, data collected on water management did not add up to provide answers to questions such as: “why is the watering of crops only done by women and not by men?” or “why do some of the underprivileged farming households, including caste households, have access to irrigation water, while others are denied?” In other words, the data I collected on the social organization of irrigation management could not fully explain how and why the system was organized the way it was.  I asked some of my knowledgeable informants for help on this topic but in vain. I never got a clear answer. Their answers were usually pithy, such as “women do irrigation work because that has been the custom” or “caste and Dhutul women participate in irrigation 	   40	  management because they get land in return.” Moreover they would not show much interest in answering these questions. Their lack of interest seemed partly due to the fact that the conceptual compartmentalization of irrigation management roles according to different social categories was too alien to the farmers, and partly because irrigation activities were too banal for them. A key reason, however, as I figured out months later, was that the question – the research question itself – was not framed properly. My assumption that the social organization of irrigation management tasks can be coherently understood within a framework of caste, gender and caste relationships proved wrong in the field. I came to realize that my assumption was wrong not because these relationships do not exist in irrigation management, which of course they do, but because (as I came to later understand) these relationships are part of a broader social division of labor in the village that included farming as well as the management of other resources. I thus had to expand the scope of my analysis from trying to understand the functional logic of inequities in the social organization of labor in irrigation management only to seeing, and continuing the intersectional analysis of, these relationships as part of the larger social organization of labor for farming and related resource management practices.   What I therefore learned and show in this study is that irrigation management roles, through the lenses of caste, gender and class relationships, are a part of and embedded in relations of power in the village socio-economic structure. This sociological understanding and description forms the core argument of this thesis for three key reasons. First, a broader sociological approach, using a Polanyian approach to economy as an instituted process (in Chapter 5), provides clear and rational answers as to why and 	   41	  how inequitable relations operate in irrigation management (my core research questions, especially #2). This approach also provides answers to various case study specific questions such as why only women practice irrigation in Spiti or why certain landless and outcaste members of the village have a small piece of farmland and a share in irrigation water. Second, by expanding explication beyond irrigation to the broader agrarian livelihood practices of the villages, this study provides a more nuanced and complete understanding of the social organization of labor within the village context. Third, and more importantly, it is consistent with the application of formal theoretical principles of understanding micropolitics between user groups (Agrarian Environments, Polanyi, Intersectional theory) and bears important policy implications, especially for international development work. I speak to the implications of this work more specifically in the conclusion (Chapter 7).  Finally, some notes on the non-academic textual data collected and used in this study. From the beginning of the project, I have been obsessively collecting all kinds of textual data on Spiti. Several of my local contacts have helped me in this effort. I had asked them to collect “anything and everything related to Spiti”, including books, magazines, brochures, newspaper clips, etc. During fieldwork trips to India, I went to twelve local libraries (in Shimla, Dharamsala, Manali and Spiti), where I spent several days reading and copying relevant information. I also traveled to the different villages of Spiti Valley collecting textual (published and unpublished) and other materials related to Spiti, particularly concerning agriculture, society, politics and history. As a result, I collected a substantial body of non-academic textual sources on all aspects of Spiti: history, food, 	   42	  music, culture, stories, songs, local laws, etc. Among the English language sources, the most relevant and the most extensive source of textual material related to Spiti and this study are the colonial British records and other publications (accounts of travel to Spiti, officer’s personal memoirs, etc.). These records have the most detailed accounts of all aspects of local history, economy, society, customs, administrative records, etc. These records include historical and legal documents such as the 1873 Spiti Regulation Act and the 1864 British House of Commons records outlining why Spiti was annexed to British India. Relevant Tibetan language sources include works on local history (a series of articles by Jhampa, 2008-2010; Tsetan, 1987; Gergan, 1976; Gyalpo, 2006; etc.) and histories of local temples (Tsering, 2000; Rigzin & Lodie, unpublished; Bhoti, 2010) as well as the biographies of Rinchen Zangpo19 and Rangrig Repa.20 The main Hindi language sources include a plethora of local publications, including different issues of Kunzom and La-Dartse magazines, as well as a “Rare Book Series” on Lahul and Spiti published by the Himachal Academy of Art, Culture and Language. These publications include many articles on local history, culture and farming practices. An important Hindi language source on local history is Rahul Sankrityayan (1994, 2002, 1948/2006). I have also collected and learned significantly from many unpublished primary sources, such as a detailed account of an injustice experienced by a local Dhutul monk imposed by the village council; file records of local Zo blacksmith leaders on their experiences of injustices from villagers; local revenue officer records of village irrigation and other resource management customs, etc. Except for sensitive materials, I will eventually make these resources publicly available either through a website or through a local Spiti library. 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  19 Rinchen Zangpo is credited with having built many of the local temples, monasteries and other structures in Spiti Valley in the late 10th century and possibly also in the early 11th century.	  20 A famous 17th century saint from Spiti Valley. 	  	   43	  3. Village Social Structure: A Historical Background Introduction Central to issues of socio-economic inequality in the villages of Spiti Valley is the fact that most of the cultivated fields and associated irrigation sources belong to a minority group of households, the Khangchen. It was only in recent decades that the Indian government granted farming land to the traditionally landless Dhutul and caste households. Until then, Dhutul and caste household members depended on Khangchen households for sources of work.21 Even today, as I demonstrate in the following chapters, the Khangchen households not only own most of the village fields and associated irrigation sources but also dominate local farming and resource management customs and practices. This socio-economic inequality raises important questions regarding the legal and historical basis on which Khangchen households came to enjoy such privileged access to local resources. In interviews and conversations with farmers representing different local groups, the only relevant answer I received was that Khangchen households are the oldest households of the village. However, no one knows exactly when Khangchen households acquired their privileged status.22 In this chapter, I will propose a hypothesis that seeks to answer this question.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  21 The two villages where I conducted fieldwork are considered historical or old villages. There are many newer villages, some of them very small, where any type of households can exercise ownership over the fields. It is only in the major villages of Spiti that the village fields are owned exclusively by one group of households. While some of the fields in these villages may belong to other households, these are recent creations that are negligible in size and number. 	  22 In their detailed description of irrigation systems in the neighboring region of Zangskar, Gutschow and Gutschow (2003, p. 117) also note that the “Zangskari house[hold] appears to be eternal, for its members can no more recall its origins than imagine their demise.” 	  	   44	  This chapter has two main parts. The first section provides a brief introduction to Spiti Valley (mainly its geography, culture and history) and to the village socio-economic structure. The second section provides some historical background to the study. Specifically, it aims to historicize these social relations and structures, to be able to speak to how they change, and to highlight what aspects of these relations and structures seem to have been more enduring.  Since this study concerns the micropolitics of village resource management and farming-related customs based on an individual’s gender, class and caste relations, there is a risk that these social relations, along with the local socio-economic structure, will be perceived as unchanging. It is therefore an important goal of this chapter to argue that the seemingly rigid local socio-economic and power structures that characterize the villages of Spiti Valley are always changing in response to larger political, economic and social forces. I argue that the social features that have persisted for long historical periods, such as the privileged status of the Khangchen households, are precisely those that were adopted (usually with certain modifications) or allowed to continue by the different regimes that held power over the region. This approach directly responds to calls for analysis of how changes at the micro-level are a part or a manifestation of – as sculpted by and in response to – larger political and economic changes (Marcus & Fischer, 1986; Agrawal & Sivaramakrishnan, 2000). It is also consistent with shifting anthropological paradigms of describing Himalayan communities.  Several pioneering anthropological studies conducted in Tibetan Buddhist societies in the Himalayas have been criticized for presenting the cultural and historical ‘setting’ of their case studies as isolated societies untouched by forces of state and market (van Beek, 1996; Tsering, 2008).  While this flaw has been redressed in many recent studies, Charles 	   45	  Ramble (2008, p. 23) warns that there is an occasional tendency in recent studies to suppress the uniqueness of each individual case study. Martjin van Beek (1996), on the other hand, argues that the very practice of presenting objectified representations of people and places is part of the problem inherent in attempting to understand and address issues of identity representation, which is a major issue in the region (Shakya, 1993).  Instead, van Beek advocates a study of the historical construction of objectified representations of people and place rather than the representations themselves. These debates in Tibetan and Himalayan studies represent broader debates concerning essentialist portrayals of indigenous cultures (Ellen, 1986), ascriptions of cultural characteristics to geographical spaces (Gupta & Ferguson, 1992), and portrayals of traditional societies as being untouched by larger political economies (Marcus & Fischer, 1986, pp. 77-110). Attending to these debates, this chapter shows how the society and culture of this relatively isolated valley are “always in a flux, in a perpetual historically sensitive state of resistance and accommodation to broader processes of influence that are as much inside as outside the local context” (Marcus & Fischer, 1986, p. 78, emphasis original).23   Spiti Valley: the setting Spiti, the valley of gods! In summer, Spiti is filled with greenery of crops, like a valley of turquoise. In autumn, Spiti is filled with mustard flowers, like a valley of gold. 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  23 The only exception to the treatment is van Beek’s (1996) call to analyze how the objectified identity of the region and people have been historically (and bureaucratically) constructed. While I have learned much from this approach to dismantling objectified identities, it is not undertaken in this chapter because the method does not fit the goals of this research.	  	   46	  In winter, Spiti is covered in snow, like a valley of silver. ~ Local Spiti saying ~  Spiti Valley is located in the present day Indian state of Himachal Pradesh.  Currently, it has a population of around 12,000 people 24 living in a 7,100 sq. km.25 territory.  Geographically, it is situated in the rain shadow of the Himalayas, which deprives the valley of the Indian monsoon rains. Spiti’s extreme altitude (many villages in Spiti are situated higher than 4,000 meters above sea level, with the lowest villages in Spiti being over 3000m above sea level) and arid landscape distinguishes Spiti and its neighboring region of upper Kinnaur from other regions of Himachal Pradesh.26   Partly due to its high and arid geographical conditions, Spiti is relatively isolated.  Most tourists as well as anthropological researchers visiting the region choose to visit the neighboring region of Ladakh, which has similar cultural and geographical environments but is connected by an airport facility and better road infrastructure - and thus a stronger tourist economy.  The road connecting Spiti to the nearest Indian town of Manali, much of which is unpaved, is normally closed from November to April every year due to snow. The 1962 border war between India and China resulted in Spiti, along with its neighboring region of Ladakh, being closed to tourists and Indians from outside the region.  While Ladakh was opened to tourists in 1974, Spiti was kept off-limits to tourists, 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  24 A local official who coordinated the 2010 population census said that Spiti’s population is 11,852. However, a government of Himachal Pradesh website presents the local population as being 12,445 (Accessed on January 3, 2014: http://hplahaulspiti.nic.in/fact_file.htm).	  25 Census of India, 2001. Series 3: Himachal Pradesh, Primary Census Abstract (p. 25).	  26 The latter are mostly green with vegetation due to their location on the south face of the Himalayas, a geographical situation which allows them to benefit from more sunlight and to receive the Monsoon rains.	  	   47	  as well as to non-local Indians, until 1993.  The opening of Ladakh attracted substantial anthropological research in the region. In recent decades, research in Ladakh has continued, and includes a dedicated international association of scholars and even a peer-reviewed Ladakh Studies journal.  In contrast, very little anthropological and sociological research has been carried out in Spiti (exceptions include Jahoda, 2007, 2008; and Dollfus, 2004).  The most detailed academic research related to Spiti’s history and culture has been done in the field of Buddhist Studies, mainly through the study of Buddhist art, architecture, and manuscripts found in its ancient temples and monasteries.    Figure 1: Geographical location of Spiti Valley (Map source: www.demis.nl)  	  Spiti is most well known for its old Buddhist temples and monasteries.  These sites have attracted some of the pioneer scholars of Tibetan Buddhism (Tucci, 1935/1988; 	   48	  Snellgrove, 1957/1995) and continue to draw researchers to Spiti (Klimburg-Salter, 1997; Tropper, 2008).  Several of Spiti’s temples and monasteries are believed to be more than a thousand years old.  During the 10th century, the Buddhist rulers of the newly founded Guge kingdom supported the Great Translator Rinchen Zangpo (lo tsa ba chen po rin chen bzang po) in a project whose objective was to translate all the major Sanskrit Buddhist texts into the Tibetan language and to build temples, monasteries and stupas in the region (mainly Ladakh, Zanskar, Spiti, Upper Kinnaur, Guge and Purang).  These initiatives also promoted significant cultural and scholarly exchanges between Tibetan and Indian scholars and artists (Klimburg-Salter, 1997), as evidenced by standardized Tibetan translations of Sanskrit Buddhist texts from this time as well as by the Indian (Kashmiri) influenced art and iconography featured in the Buddhist temples of the Alchi Monastery in Ladakh and the Tabo Monastery in Spiti.  Spiti was a key site of these initiatives (Klimburg-Salter, 1997; Tucci, 1935/1988),27 which laid the foundation for what is known in Tibetan Buddhist historiography as the “Second Diffusion of Buddhism”.28  The legacy of this period is such that even today Spiti and its neighboring regions (including Ladakh in the Indian state of Jammu and Kashmir, Lahaul and upper Kinnaur in Himachal Pradesh, and the Ngari region of Tibet) share many aspects of 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  27 Spiti became a key site of the scholarly, artistic and architectural activities of the “Second Diffusion of Buddhism” partly because it was located at the crossroads of Kashmir, India and Guge and partly because it was a stronghold of Bon religion (Tucci, 1935/1988, p. 15; Laxman S. Thakur, 2008), which the proponents of Tibetan Buddhism, specifically the Tibetan rulers who had established themselves in the region during this period, sought to subdue (Tucci, 1935/1988; Klimburg-Salter, 1997). Although many aspects of Bon religion - such as the propitiation of mountain deities and serpent spirits - are still practiced in Spiti and other Himalayan Tibetan Buddhist regions today, these deities and cults are either subdued or incorporated into the Tibetan Buddhist pantheon. 	  28 According to standard Tibetan Buddhist history, Buddhism first came to Tibet during the reign of Yarlung dynasty kings but much of Buddhist heritage was destroyed during and after the reign of the last anti-Buddhist emperor, Lang Darma (‘u dum btsan po, who reigned from 841-842, according to Dotson, 2009, p. 143). This phase of Tibetan Buddhism is known as the “First Diffusion of Buddhism”. Tibetan Buddhism then reemerged mainly from Western Tibet, starting from the late 10th century, commencing the period known as the “Second Diffusion of Buddhism”. 	  	   49	  religious culture, social customs, institutions, belief systems, and language.  The main features found in all these regions are the dominance of Tibetan Buddhism in all aspects of life, belief in mountain deities and serpent spirits, and agriculture as the economic mainstay of the people.29 Local notions of gender and gender roles in Spiti are also similar to those in the Tibetan-speaking regions of the Himalayas (Gyatso, 1987; Huber, 1994; Tsomo, 2004). In the following chapters, I show that local notions of gender roles relate very well to the larger literature on feminist political ecology and critical agrarian studies pertaining to South Asia in general.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  29 Another important feature of the Western Himalayan region of India, and Tibet more broadly, is the institution of monasteries.  Monasteries play an important role in Tibetan Buddhist societies (Mills, 2003; Jahoda, 2007).  Spiti has five monasteries: Tabo, Dhangkar, Kee, Tengyud and Gungri.29  Tabo (built in 996 A. D.) is recognized as the world’s oldest continuously occupied Buddhist monastery. These monasteries played two key roles: they were centers of religious practice and education, and they were powerful socio-economic institutions which owned large tracts of agricultural land.  Each monastery also has its own base of affiliated villages. For example, several villages are affiliated with each monastery.  Whenever there is a need for religious ritual performance (e.g., during sickness or death), farmers avail these services through their local or affiliated monastery.  Farmers of affiliated villages in turn are responsible for supporting the monastery through different kinds of taxes, including the enrolment of their children as monks (Jahoda, 2007).  Although the monasteries in Spiti are relatively small, academically inclined monks are able to achieve the highest levels of educational training because these centers of religious education are part of the larger Tibetan monastic education system.  For example, it was customary for scholarly monks from Spiti to attend larger monasteries in Tibet, where all the major monasteries – Tashi Lhunpo, Sera, Gaden and Drepung – accommodated monks from Spiti and nearby regions into their residential colleges for natives of the Ngari region, of which Spiti historically formed a part (Tsetan, 1987). Since the closing of the India-Tibet border, however, monks from Spiti join monasteries established by Tibetan refugees in different parts of India.  The monasteries in Spiti traditionally held large tracts of agricultural land, which were cultivated by tenant farmers (Jahoda, 2007).  However, much of the monastic estate lands were given to tenant farmers after the passage of the 1950 Punjab Tenants Act, which imposed a ceiling on the size of land holdings that any one individual or institution could occupy (Jahoda, 2008, p. 13).  Although the powers of the monasteries have declined in recent decades due to the rule of modern secular states, they still remain powerful.  For instance, even today, all non-caste households are required to admit their second son into the monastery.  During my trip to Spiti in 2007, I learned from a Dhutul man of Tashigang village that he had to pay a fine of 40,000 rupees to Kee Monastery because he had chosen to send his younger son to school rather than to the monastery.  Another example of the monasteries’ power and influence is evidenced by village laws that prohibit farmers from farming or other livelihood activities (such as collecting dung) when the head of the local monastery visits the village.  Irrigation water that day flows freely without use.  Everyone in the village must be engaged in activities (reception, cooking, dance, music, decoration, serving, cleaning, etc.) which honor the visiting lama. 	  	   50	  While Spiti shares its high altitude mountainous desert landscape and Tibetan Buddhist culture with its Tibetan and Ladakhi neighboring regions, it also has distinct cultural and socio-economic characteristics.  Some of these distinct attributes are a product of more recent history, while others are more ancient. Among Spiti’s distinct cultural features with ancient roots are its local dialect and calendar system.  Although people in Spiti speak a Western Tibetan dialect, an Indian linguist who has compiled a substantial dictionary of Spiti words (Mathews, personal communication, May 6, 2011) mentioned that the Spiti dialect – with its own local variations – has its own distinctive linguistic characteristics that are not found in the Ladakhi and central Tibetan languages.30 Perhaps a more noteworthy and relevant cultural feature of Spiti is that Spiti has a unique system of timekeeping, with the names of the months of the calendar being based on ecological, seasonal and agricultural cycles (see Appendix 1 and chapter 5).31 Compared to both the Tibetan and Ladakhi systems, this calendar system is unique in terms of the timing of the year cycle (including the timing of the new year)32 and the names of the months (Gergan, 1978).   As indicated from Spiti’s naming of the months according to agricultural seasons, farming has been the main form of livelihood practice in Spiti from time immemorial. In addition to agriculture, people also engage in other livelihood activities such as weaving, 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  30 The local language is called Bhoti. Bhoti is a generic term in Hindi for Tibet, meaning “of Tibet”; Tibet is known to Hindus as “bhot”.  There are local variations in Spiti’s Bhoti: people in lower Spiti use more honorific language in their speech as compared with those in upper Spiti and those of Pin Valley. 	  31 Another unique aspect of Spiti, at least contemporary Spiti, is that it is rare to find polyandrous relations, which are common in its neighboring regions.  In the two villages where I conducted this study, there was only one household with a polyandrous arrangement. 	  32 For example, while the Ladakhi new year starts in the 11th month of the Tibetan calendar (Gergan, 1978), Spiti’s starts in the 10th month. 	  	   51	  trade and pastoralism.33 Today, the government also plays a central role by employing a significant number of the adult population as road construction workers, teachers, nurses, and bureaucrats, as well as in a host of other capacities. Therefore, although I use the word “farmer” to generically describe the people of Spiti in this study, it must be pointed out that today people are engaged in a host of livelihood (income) sources in addition to their common occupation of farming. For example, the father of my host family is a local doctor, a farmer and a proficient local architect, all of which are closely tied to activities in the village. At the same time, he also works as an employee of the state government as a forest guard.  In addition, during the four years when I was engaged in my fieldwork, he started and completed building a large four-story house, which he plans to run as a guesthouse for tourists.  Social stratification and socio-economic structure of Spiti Valley Literature on social stratification in Tibetan (Carrasco, 1959; Goldstein, 1971; Aziz, 1978; Wiley, 1986), Ladakhi (Crook, 1994) and Spiti (Jahoda, 2008) societies shows similarities in socio-economic class structure, albeit with minor regional and local differences. Households with farming land and irrigation rights, with associated tax 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  33 Before roads connected Spiti, people made most of their own clothing, shoes and ropes - mainly from yak and sheep skin and wool (Harcourt, 1871, p. 77; Interviews, 2010 & 2011).  Today, because people wear modern clothing (pants, shirts, etc., with most women wearing Indian salwar kameez), weaving is limited to making carpets and ropes, an activity which is mainly done during the winter months.  Women do most of the weaving, although men also engage in spinning wool with a spindle.  In the past, trading was also an important activity in Spiti.  Spiti traditionally formed a trade link (mainly for wool) between the plains of India and the highlands of Tibet and Ladakh (Thakur, 2001, pp. 5-17), which was the main reason for the British takeover of Spiti in 1846 (British Parliament, 1864, p. 26).  Historically, people from Ladakh, Changthang, Tibet and India used to come to Spiti during an annual trading fair at La Dartse, near Zibug Village, during the late summer (Interviews, 2010 & 2011).  Today, some farmers engage in the selling of Chumurti horses and yaks to Ladakhi and Tibetan nomads in the Changthang.  During my fieldwork in 2010, a group of nomads from Changthang also came to Khyung, Spiti, with sheep and horses for trade. As for pastoralism in Spiti, Charudutt Mishra and his colleagues (e.g., Mishr et al., 2003; Mishra, Wieren, Ketner, Heitkonig, & Prins, 2004) have written extensively on the subject.	  	   52	  obligations, form the basic corporate entities of the Tibetan socio-economic system. A Tibetan village is typically comprised of several classes of households. These are, starting from the top: aristocrats, tantric practitioners (Joba or Ngagpa), Tibetan medicine practitioners (Amchi), taxpayer farmers, landless Dhutul (or Dhuchung) farmers, and outcastes. The minor differences at the regional and local level are mainly demonstrated in the compositions of different classes of people or households. Some regions and villages within a region have a greater percentage of one particular class of households, while others have less or none of these households. For example, the outcaste group of Beda, which is present in the western Himalayan regions of Ladakh and Spiti is absent in central Tibet. An example of differences in class structure within a region would be the fact that many villages in Spiti have only two kinds of households (i.e., Khangchen and Dhutul), while many others have three kinds of households (i.e., Khangchen, Dhutul and outcastes). These kinds of differences are further outlined in the chart below, which compares the social stratifications characteristic of Tibet and Spiti. The percentage of aristocrats in Spiti Valley (only one household)34 is significantly lower than that of Tibet. Another important difference is that Spiti has only two types of outcaste households (i.e., blacksmiths and musicians), whereas Tibet has several types of outcastes.      	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  34 There are a few minor aristocrat households in Spiti. These are the “small Nono” of Pin Valley, Nono of Mane Village, Nono of Demul Village, as well as some households that claim to be descendants of the Lord of the Fort (mkhar spon). The status of these households is lower than that of the main Nono (of Kyuling Village) of Spiti and generally equivalent to the Joba and Amchi households. 	  	   53	  Table 1: Social stratification in Tibet and Spiti  For the purposes of this study, the socio-economic structure of Spiti Village is best classified into three classes of households: Khangchen, Dhutul and outcastes (which are today known as “castes”). The following table shows three different types of social stratification characteristic of Spiti. At a broader level, the society is grouped into two endogamous classes: the outcastes and the Chechang. While this is an important classification for the locals, this study uses the next level of classification that further divides the majority of Chechang households, which form 95.4% of the population (according to a 2001 Census), into taxpayer Khangchen households and the traditionally landless Dhutul households. This is an important differentiation for the purposes of this study because socio-economic power dynamics between these two groups constitutes an important aspect of farming and resource management customs. Although the society can be further classified in more detail, as is shown in the third column of the table, such  Social classes % of population in traditional Tibetan society (Gombo, 1983, p. 68) % of population in Spiti Valley (according to 2001 Census) King or ruler  N/A Aristocrats 3% 0.0003% (one Nono household) Taxpayer Khangchen [including Amchi and Joba] 40%  Dhutul (or Dhuchung) 50%  Outcastes 7%  5.6%  	   54	  detail is not necessary for the purpose of analyzing the social organization of farming and resource management customs.   Broad stratification Standard stratification Detailed stratification   Nono Chechang/Yarab Khangchen Joba, Amchi and minor aristocrats   Khangchen  Dhutul Dhutul   Zo blacksmiths Caste/Marab Caste/ Marab Beda musicians Table 2: Social stratification in Spiti Valley 	  Khangchen households Locals believe that Khangchen households are the founders and the oldest households of the village.35  That is why these households are also referred to as the “Old Households” (khang chen rnying pa).  These households were also called “tax payers” (khral pa) because only this group of households was historically responsible for paying taxes to the ruler.36  The farming estates of these households formed the basic unit of agricultural production and taxation which constituted the traditional economy.  As tax-payers, these 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  35 Another definition of Khangchen is the “house of the eldest brother or son”, the person responsible for running the farming estate. When the eldest son marries and takes charge of running the farming estate, his parents (often with an aunt or younger sister) retire into a Khangchung (Khang = house, chung = small). Carassco (1959, p. 32), along with British colonial records, mention yet another household, Yangchungpa (“still-smaller-one”), where the members of Khangchung move if/when their grandson becomes head of the Khangchen and their son must move into the Khangchung.35 Khangchung and Yangchung households are thus extended residences of the Khangchen household. 	  36 Taxes include grains, corvee labor and military service. These taxes were abolished or exempted by the Indian government.	  	   55	  households (i.e., the men of these households) had legal rights to all of the arable land and irrigation sources of the village.  These men also represented the Village Council and hence were the decision makers; they thus constituted the most powerful group of households in the village.37  How did the Khangchen households come to own all the local resources and decision-making powers, and how does the Khangchen household system perpetuate this arrangement?  According to common belief, the Khangchen households have privileged ownership over all farming land and irrigation water because they are (or are believed to be) the male descendants of the oldest households of the village.  Locals believe that there were a total of 250 such households in the beginning, which were known as the “250 soldiers”38 (Interviews, 2010, 2011).  The old Khangchen households of today thus represent the descendants of the 250 soldiers.  Local elders mentioned that these 250 Khangchen households still exist and that the number has not changed much in history, if at all.39   	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  37 This was the case until after the introduction of the 1968 Nautor land grant program, described below, which made it possible for poor farmers to become “full” tax payers to the village and thus to assume membership in the Village Council.  	  38 “dang po dmag phed rang gsum bgrya byung pa red zer gyi ‘dug”. Here, “phed” is the local pronounciation for “phyed” or half. I am not sure about the correct spelling of “rang” but most probably, it is “dang”, which would render the statement as “dang po dmag phyed dang sum brgya byung pa red zer gyi ‘dug”.  The expression “half less than 300 hundred (phyed dang sum brgya) is consistent with classical Tibetan expressions such as “sde pa phyed dang brgyad” for “half less than eight sets” of four alphabets, making up the 30 alphabets in the Tibetan language.	  39 During my fieldwork in 2010 and 2011, I travelled to all the villages in Spiti and collected data on the total number of Historical Khangchen households in each village, which totaled 269. Here, it must be noted that the numbers were collected by asking random villagers and trusting that there was a degree of veracity in their responses. Generally, all mature members of the village would have this knowledge and information. Also, there were several known cases in the historical memory of villagers that explained how the numbers had increased. In some villages, the numbers had increased through the fissioning of rich Khangchen households. In others, the number had increased after a Dhutul farmer bought land from the village and had joined the ranks of Khangchen householders.	  	   56	  Maintaining the continuity in the number and size of land holdings was an important characteristic of the Khangchen household system (Wylie, 1986, p. 6).  This continuity is maintained by a system of primogeniture, whereby the property is passed on to the eldest son when he gets married. This practice continues from one generation to the next, with local laws existing to ensure that the property is not divided into smaller pieces. For example, while it is customary to give small pieces of land of Khangchen property to support certain family members, such as a monk brother (who is given a “monk field” or dra zhing) or elder parents, these pieces of land must be returned to the Khangchen household after the death of the family members to whom these were given.   Dhutul households Dhutul households are generally described as those “who have nothing but a house, being literally a smoke-maker” (Gazetteer, 1883, p. 439).  Members of Dhutul households were traditionally landless farmers who were dependent on Khangchen households for work (Lyall, 1874, p. 185).40 Although this class of household is socio-economically weaker than the Khangchen, members of both Dhutul and Khangchen households are considered equally pure, and known as Yarab (“good manners”), as opposed to the ritually impure members of caste households, who are known as Marab (“vulgar”/“immoral”).   The Dhutul class of farmers comprises the majority of the Spiti population.  According to local history, in the beginning, there were 600 “Za” or Dhutul households along with the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  40 Historically, there were two types of Dhutul households in Spiti: those that were tied to a Khangchen household (rang gi dud ‘thul) and those that were independent (Gazetteer, 1883, pp. 479-480).  These two types of Dhutul are also found in other Tibetan societies (Goldstein, 1971; Aziz, 1978).  However, no distinction is made between the two types of Dhutul households in Spiti today.	  	   57	  250 “soldier” or Khangchen households. Other historical data on Dhutuls is generally not available.  That this group of households did not pay taxes to the ruler is probably why the British did not seek to provide census figures specific to this category.  The Indian Government does not differentiate between Khangchen and Dhutul households because both categories (as well as everyone else in Spiti) have been exempted from taxation.  Jahoda (2008) has shown that the number of Dhutul farmers has increased significantly since Spiti became a part of India, which is consistent with my field observations.  The total population of Spiti has increased from 4,222 in 1951 (a figure based on independent India’s first national census) to 11,852 in 2010 ( a figure based on the last national census).  Given that the total number of Khangchen households has changed very little (as mentioned above) and since the caste population makes up only a small section of the local population (5.6%, according to the 2001 census),41 we can deduce that the number of Dhutul households has increased the most over the last few decades.42  This is consistent with studies done by Jahoda (2008) on demographic changes in Spiti’s Tabo village, which show that the majority of the increase in population belongs to the Dhutul category, formed mainly through the fissioning of Old Khangchen and Dhutul households.  The main factor that made the increase in population socio-economically feasible was the 1968 Nautor land grant program (discussed below). After this program was introduced, all of the landless households in Zibug and Khyung villages were given land, which made it possible for many of them to formally enter the ranks of the Khangchen class by paying “full” village taxes.  Thus, the socio-economic status or power of the Dhutul households 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  41 The total population of Spiti in 2001 was 10,679 (5,574 males and 4,805 females), out of which “schedule caste” members comprised 600 (317 males and 283 females).  (Data collected from Kaza office of the Additional District Commissioner, Spiti Valley). 	  42 Here, I am referring to the absolute number of individuals, not the percentage.  I do not have sufficient data to compare in terms of percentage. 	  	   58	  rose significantly after the passage of the Himachal Pradesh Nautor Land Rules act of 1968.   “Caste” or Marab households  At the bottom of the village social structure are the two outcaste Marab households: Zo blacksmiths and Beda musicians. It is unknown when and how Zo households came to Spiti but the Beda were introduced under Ladakhi rule (discussed below). Today, Zo and Beda households are more commonly known – in local speech and government records, as well as in academic publications – as “caste” or “Scheduled Caste” households. During my field research, I observed that farmers use the word “caste” more often than “Marab”, even while speaking in the local language.43   The use of the term “caste” to refer to these two groups of households in Spiti is also standard in academic publications (e.g., Mishra et al., 2003; Gutschow, 2006; Jahoda 2007 & 2008; Tsering & Ishimura, 2012), where it is normally clarified or implicitly assumed that the term specifically refers to local outcaste groups and does not concern the Indian caste organization.44 This study will also use the term “caste” to refer to Zo and/or Beda households/members; this is not to imply, however, that the system is the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  43 The local use of the English word “caste” could have become popular only after the Indian Constitution Order of 1950, which designated “Bhot” (or Tibetan) people of Spiti under “Scheduled Tribe” and Beda and Zopa under “Scheduled Caste” categories. In other words, according to Indian government records (including the Census), all the Chechang members belong to “Scheduled Tribe” and all the Marab members belong to “Scheduled Caste”. These categories have become popular among people because members of Scheduled Tribes and Scheduled Castes receive subsidized government aid and special “reservation” quotas in government offices and institutions. A telling example of the popularity of these terms is that local leaders often use the word “tribal” to describe Spiti as a “backward” region deserving of government aid, on the one hand, and attractive enough to generate tourism, on the other. 	  44 Caste is also used in studies related to Ladakhi (Gutschow, 2006) and Tibetan (Aziz, 1978; Ortner, 1973) societies more broadly. 	  	   59	  same as the Hindu caste system. Here I want to emphasize that the uncritical use of the term “caste” can reify caste biases. Although an option to avoid such reification is to not use the term, I chose to actively use it because this study directly concerns socio-economic inequities experienced by Zo and Beda members precisely because of their outcaste status, which is more strongly expressed by the term “caste” than “marab.” More importantly, I believe that the reification of caste inequities happens when the term is used in an uncritical manner. This study, on the contrary, reveals and critically analyzes routine discrimination experienced by outcaste members in the domains of farming and resource management.  The social beliefs and practices concerning caste members in Spiti are similar to those concerning outcastes in Tibet (Carrasco, 1959; Aziz, 1978; Gombo, 1983; Childs, 2003) and other Himalayan Buddhist regions such as Ladakh (Rather, 1997) and Khumbu in Nepal (Ortner, 1973).  In addition to lower social status, caste members are believed to be irreversibly defiled -  according to local notions of purity. Members of Chechang households follow two main rules of exclusion against caste members: prohibition on marital relations and on sharing the same drinking cup. In the rare cases of marital relations or “mixing of mouths” between Chechang and caste members, the result is that the Chechang person involved permanently loses his or her status.45 Finally, it must be 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  45 As mentioned earlier, the caste groups form a separate endogamous group in Spiti. Here it must be noted that while members of Chechang and caste households are socially allowed to marry within their own social group, those belonging to more respected or higher status households, such as local doctors (Amchi) and minor aristocrats, prefer to marry among themselves or within the local elites. Similarly, within the caste group, Zo blacksmiths prefer to marry within their own social group, as Beda musicians are the lowest in social status. While marital relations between different classes of houses within Chechang households and within caste households are socially allowed, marriage between Chechang and caste members is strictly prohibited. Such a marriage would entail permanent loss of status for the transgressing Chechang member. Similarly, although people of Spiti do not normally share drinking cups or eating bowls 	   60	  noted here that, in Spiti, caste is a stronger determinant of a person’s status than is gender. This became clear to me after I interviewed a Zo man, who sat on my bed during the interview process. This came to the attention of the mother of my host (Dhutul) family. Speaking about the Zo man I interviewed, she said that he is “bad manners” (ma rabs). Consequently, she said that she never let them (i.e., caste members) sit above or next to her, a woman.   With this brief introduction, I will now describe the two caste groups of Spiti.   Zo blacksmiths The blacksmiths are an important class of people in Spiti’s society. Locally known as “Zo” (Zo or Zo ba for “craftsman”), this group of individuals is traditionally responsible for making all kinds of metal implements used in Spiti society.  Colonial British records speak highly of the skills of the blacksmiths of Spiti, who are known for making good quality “pipes, tinder boxes, bits, locks and keys, knives, choppers, hoes, plough-shares, and chains” (Lyall, 1874, p. 198).  Since many Zo households traditionally had farming land (Zo zhing) in the village, which they received in return for their services, the Zo households paid a small “smoke tax” or Dhutral (dud khral), just as the Dhutul household members did.  Therefore, Zo households were also considered as belonging to the Dhutul 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  with one another, especially those of lower classes (e.g., a Khangchen man will not drink from the same cup as that of a Dhutul man), the “mixing of mouths” between Chechang members or between caste members is socially acceptable. However, the “mixing of mouths” between Chechang and caste members is strictly avoided because even that would entail the permanent loss of status for the Chechang person who “mixed mouths.” A Dhutul man told me a story he had heard: once a group of people working on road construction took a tea break.  During the tea break, a young Chechang man sipped out of the cup of a young Beda woman by mistake.  When news of this incident spread, the young man’s parents sent him to the Beda family to be wedded to their girl and to live as a Beda man.  Although I could not confirm if the story is indeed true, the narrative reveals social mores and biases against the Beda caste.	  	   61	  category in terms of village tax contribution.  However, for the purposes of this study, Zo households are viewed as belonging to the “caste” category because caste is a more important determinant of their social status than is Dhutul.   According to demographic data from 1868, there were 100 blacksmiths living in Spiti in 1868, comprising 3.3% of the total population of the valley, which was 3,024 (Lyall, 1874, p. 198).46 The combined population of Zo and Beda (see below) in Spiti has increased from 4.8% in 1868 (Lyall, 1874) to 5.6% in 2001 (government census data). To extrapolate an approximate current population of Zo people in Spiti from this limited information, assuming that they are more than the Beda people (Interviews, 2010, 2011), we can say that they constitute approximately 3.8% of the local population (+/- 1%).47  Zo households are scattered throughout Spiti Valley with varying numbers in different villages: with more in some (e.g., three households in Khyung in 2010), less in some (e.g., two households in Zibug in 2010) and none in others (e.g., Hanse in 2010).  Today none of the Zo households in Zibug and Khyung villages engage in producing metal implements; this is due to the fact that mass produced tools are imported from the plains of India.  Instead, their main source of livelihood is farming. However many members are also employed in the service sector (with government and NGOs) and some work as construction contractors.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  46 The number of people representative of blacksmith households in 1868 (when the total population of Spiti Valley was 3,024) was 100.	  47 According to 1868 data, Zo constituted 3.3% of the total population. In the same year, the total caste population constituted 4.8%.  Since there are only two caste groups in Spiti, we can also calculate the percentage of the Zo population within the total caste population: if 4.8% of the caste population represented 100%, then the Zo population of 3.3% would equal 68.75% of the total caste population. Applying this percentage (68.75%) to 2001 caste data (i.e., 5.6% of the population), we get 3.85% (i.e., if 100% of caste is 5.6%, then 68.75% would equal 3.85%).	  	   62	   Beda musicians  Beda refers to the smallest and weakest group of households in Spiti.  Owing to their small number, they are found only in some of the villages.  In both Zibug and Khyung villages, there were two Beda households each.  Their traditional profession is to play music during ceremonial occasions, such as in the presence of religious and political dignitaries or during festivals and rituals.  The Beda are often described as “wandering minstrels” in British colonial publications.  In return for their services, they were traditionally provided alms (bsod nyoms) in the form of food, grain and fodder (see chapter 6) by the Khangchen households mainly and, to a lesser extent, also by the Dhutul households.  Until they received Nautor land, they had historically neither paid any tax to the village nor engaged in agriculture.  Today, Beda musicians in many villages (including both Khyung and Zibug villages) are free from the customary obligation to play music for villagers in exchange for alms. Instead, Beda musicians provide their services on a “daily wage” (dihari in Hindi) basis. 48  The traditional Spiti musical ensemble, called “Hirib” or “Hiribpa”, involves the playing of three musical instruments: a Beda man playing a reed pipe (“sunna”), a Beda woman playing a large tambourine (about two feet in diameter, called “dawu”), and a Zo man playing a pair of kettle drums (“daman”).49  Often, especially at marriage parties where 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  48 Chapter 5 describes the historical circumstances that led Beda musicians of Zibug village to discontinue their customary obligations and to start the contemporary practice of playing for money.	  49 The names of musical instruments have Persian meanings: “Sunna” is probably derived from Urdu/Persian words “shenai”/“surnay”). The tambourine or “dawu” is probably derived from the Persian word “daf” as it is the same instrument, same size and both use goat-skin. The kettledrum, daman, is “damana” in Urdu/Persian. 	  	   63	  the musicians get paid well, there are more than three musicians and singers (Beda women).  It is customary for the Hirib musicians to play at the public festivals of the local monasteries (religious dance, ‘chams) and for the village deities at the local temples. They also play music and sing for the people, mainly for the Khangchen households, during birthdays, weddings, parties and village festivals.  The songs sung by Beda women at weddings and parties are regarded as auspicious (rten ‘brel glu) and ancient (dang po’i glu). While Spiti women are known for their large knowledge of songs, only the Beda women know the many ceremonial songs of Spiti’s social and religious culture.  Thus, when it comes to Spiti’s social and religious culture, the Beda musicians form an indispensible class of people.50 Yet they are treated with contempt and the Beda women (be mo), in particular, are at the worst receiving end of society’s prejudices and even hatred (Rather, 1997). An example was described to me by a local community leader (interview, 2010), who opined that it is not uncommon for Hirib musicians, especially the Beda woman, to be physically beaten by drunken men for not playing music or singing well during parties. He remembers witnessing a Beda woman playing the tambourine with blood all over her face during a community party.51  One of my field informants had learned from a local elder that Bedas were introduced into Spiti when it was under Ladakhi rule (i.e., between 1630 and 1842) to play music for the Nono during ceremonial occasions at the Cliff Castle (brag mkhar) in Dhangkar village.  Thus, according to my field informant, the first Beda of Spiti Valley lived in 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  50 Once a year, there is a party of the deities (lha yi chang ka) which is held on the 17th day of the first month of the Tibetan calendar in Khyung and Guling villages. On this day, the deities dance to Beda women’s singing and those songs can only be sung on that day.	  51 For a historical account of mistreatment of Beda women in the hands of Zo men, see Egerton (1864/2011, p. 50).	  	   64	  Dhangkar village, which is where the fort of the Governor was situated.  Gradually, the population of Beda increased and they moved to different villages.  This theory seems plausible because Beda are unique to Ladakh among Tibetan Buddhist regions and these musicians also served the kings and royalty of Ladakh (Rather, 1997, pp. 216-217; Jest & Sanday, p. 6).  The fact that the etymology of Beda is unclear in the Tibetan language as well as in the local Spiti dialect—unlike Zo, Duthul and Khangchen, which are all Tibetan—also suggests their non-Tibetan origin.52   Having described the three different classes of households, I will now provide a discussion of the historical context in which this structure was shaped and how power relations have changed through the centuries.   Historical background to local social structure and relations The following historical background to Spiti’s social structure and changing social relations is presented in three sections.  First, I present a hypothesis that the institutional basis of Spiti’s Khangchen household as tax-payers (khral pa) and soldiers (dmag) was founded during the period of the Tibetan Empire (7-9th century).  To make this link, I first cite 13th and 16th century Tibetan texts that establish that Spiti formed a part of a military administrative district of the Tibetan Empire.  Drawing on studies based on primary textual sources from the time of the Tibetan Empire, I make linkages between the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  52 According to Rather (1997, p. 215) and Bhasin (2004, p. 136), Beda is a Tibetan word meaning “to live separately.” They say that “Be refers to separately and da means reside”. This meaning, however, does not make sense for native speakers and locals. None of the Beda interviewees and local scholars I asked were able to provide a meaning or spelling of the word with any certainty. There is no standard spelling of ‘Beda’ in Tibetan, which is evident in the “various different spellings” used in Tibetan language publications (Jahoda, 2009, p. 53). The word is most likely non-Tibetan. There are also Beda castes in other parts of India. 	  	   65	  structure, function, size and service requirements of the military administrative system of the empire and those of Spiti under the Ladakhi rule (as described in British colonial records and in interviews with local people).  I also show how the similarities between the taxation systems also find significant support in local oral history, language and social customs.  The detailed hypothesis presented in the first section is important for this study, as well as for the early history of Spiti more broadly, because it suggests that the dominant social strata of the contemporary Spiti village, the Khangchen households, was instituted during the time when Spiti was under the Tibetan Empire. The second section describes the administrative power structure of Spiti under Ladakhi and British rule.  This discussion is relevant because it directly relates to Spiti’s social structure and historical administrative system, which are central to this chapter. The third and final section describes some of the major socio-economic changes that occurred in Spiti after it became a part of India.  These changes were mainly introduced or controlled by the Indian Government, which considers Spiti to be a sensitive border region.  A discussion of these changes is relevant here because they have dramatically changed the socio-economic, educational and infrastructure conditions in Spiti, leading to significant changes in gender, caste and class roles.        	   66	  Time line Powers  Main influence Social relations Pre-7th century Zhangzhung   7-9th century Tibetan Empire Soldier-Khangchen system of taxation Soldier-Khangchen system of taxation 10th century-1630 Guge Campaign to introduce Tibetan Buddhism Cultural Tibetanization of the region 1630-1683/84;  1705-1735; and from 1758-1842* Ladakh Peripheral region: largely ignored Nono and Beda 1846-1947 British Documentation of local customs Consolidation and legitimization of Khangchen privileges 1947-now India Distribution of land to landless; education; commercialization of agriculture; integration of local economy with larger political economy Empowerment of underprivileged groups; socio-economic improvement Table 3: Historical timeline of Spiti Valley53   	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  53 Major powers that ruled Spiti for more than 50 years. Dating for * based on Jahoda, 2007.	  	   67	  Spiti has a relatively coherent historical narrative from the 10th century onwards (Lahuli, 2002). Before the 10th century, the only thing generally known or accepted in Hindi (e.g., Sankrityayan, 1948/2006, 1994, 2002), Tibetan (e.g., mkhas pa’i dga’ ston; Gyalpo, 2006; Shastri, 2007) and English (e.g., Thakur, 2000; Petech, 1997) language sources is that Spiti once formed a part of the Zhangzhung kingdom (periodization unknown). This seems to have been the case until the region became a part of an expanding Tibetan Empire sometime in the mid 7th century (Sankrityayan, 1994, 2002).  For the purposes of this study, I will begin the historical discussion from the period of the Tibetan Empire because there are striking similarities between the military administrative system of the Tibetan Empire (on which there is a substantial body of literature based on primary sources such as texts and records from that period) and Spiti’s historical taxation and administrative system, especially those aspects associated with the Khangchen households.  Based on these similarities or linkages, I hypothesize that the institutional basis of Spiti’s Khangchen system of taxation was introduced during the time of the Tibetan Empire (7th to 9th century AD).   Before I present the different arguments for the hypothesis, two important explanations should be made to prevent misinterpretation about this hypothesis.  First, this hypothesis strictly applies to the case of Spiti and not to other Tibetan Buddhist regions.  I point this out because similar taxation or administrative systems in other regions could have different origins. For example, Goldstein (1971, p. 14) has noted, in the case of Samada village in central Tibet, that a military service (dmag khral) system of taxation was 	   68	  introduced in the early eighteenth century during the time of Mi dbang Pho lha nas.  Second, this hypothesis only concerns the historical foundation of the institution of the “250 soldiers” of Spiti Valley, not the ethnic or geographic origins of the soldiers per se.  This is important to point out because there are many examples of prehistoric cave art and petroglyphs in Spiti (Thakur, 2000) that bear evidence that the valley was inhabited much earlier on.  Currently, there is no information regarding whether the original group of 250 soldiers was selected among people already living in the valley or whether they were sent from elsewhere to guard the border region of Spiti.54   Origins of household system of taxation during the Tibetan Empire  A main source for this hypothesis springs from old Tibetan language sources, mainly lDe’u chos ‘byung (12th/13th century)55 and mKhas pa’i dga’ ston (16th century).56  Spiti is mentioned in these texts as part of the military administrative system of the lower Zhangzhung region of the Tibetan Empire.  The reference in lde’u chos ‘byung, which is also quoted by Dotson (2006, p. 162), Iwao (2007, p. 212) and Shastri (2007, p. 59) says:  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  54 Although the local historian Tsetan (1987) states that the people of Spiti migrated from the neighboring region of Guge (which constituted a kingdom from the 10th to the 17th century), he does not mention when they came.  Moreover, while some scholars have argued that Tibetans migrated into the western Himalayan region (Denwood, 2008 & 2009), including Spiti Valley (Thakur, 2000, Sankrityayan, 1994; 2002), during the time of the Tibetan Empire, there is no information about the ethnic or geographic origins of the 250 soldiers.	  55 Some scholars (e.g., Shastri, 2007) consider the text to hail from the 12th century, while others (Dan Martin, personal communication, 2012) believe that the text was composed in the 13th century. 	  56 Mkhas pa’i dga’ ston is available online via the Tibetan Buddhist Resource Center, see Threngwa (1986).	  	   69	  In the border region between Tibet and Nyipa, there are five thousand-districts of the lower Zhang-zhung. These are the two, Gu-ge and Gu-cog, the two, Spyir-tsang and Yar-rtsang and the little thousand-district of Spyi-ti.57  A similar reference to Spiti as part of the administrative district of the Tibetan Empire is provided in mkhas pa’i dga’ ston (p. 189):  In the border of Tibet and Nyipa, the two, Gu-ge and Cog-la, the two, Spyi-tsang and Yar-tsang, and the little thousand-district of Ci-di (Spiti) are the five thousand-districts of lower Zhangzhung.58   The Old Tibetan Empire was divided into various regions such as “the four horns” (ru bzhi) of central Tibet, Nyipa, and Zhangzhung, which were composed of units of “thousand-districts”.59 During the Old Tibetan Empire, the thousand-districts (stong sde) were the “fundamental units” (Uray, 1982, p. 545) of military and economic administration (Takeuchi, 2004; Iwao, 2007; Dotson, 2006 & 2007). A thousand-district was comprised of a thousand households, responsible for supplying a thousand soldiers, i.e., one soldier from each household (Dotson, 2009, p. 39). It is thus possible to deduce from these Tibetan language sources and studies based on primary source materials (Uray, 1982; Takeuchi, 2004; Iwao, 2007, Dotson 2006, 2007 & 2009) that Spiti was not only a 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  57 The Tibetan language source reads: “Bod dang sum pa’i so mtshams na/zhang zhung smad kyi stong sde lnga yod de/gug ge gu cog gnyis/spyir rtsang yar rtsang gnyis/spyi ti stong bu chung dang lnga’o.” Iwao (2007, p. 212) presents a similar translation. I am grateful to Melvyn Goldstein for editing and improving my translation to keep the style and language closer to the original. 	  58 The Tibetan language source reads: “Bod dang sum pa’i mtshams na gug ge cog la gnyis spyi gtsang yar gtsang gnyis ci di stong bu chung ste zhang zhung smad kyi stong sde lnga.”	  59 The thousand-districts sometimes added up to ten thousand-districts (khri sde, Iwao, 2007).	  	   70	  part of the Old Tibetan Empire but that it also provided taxes, including one soldier from each of the tax-paying households.    Before I discuss some similarities between the administrative systems of Spiti’s Khangchen household and that of the Tibetan Empire, I will first clarify that the two differently spelled names for Spiti in lde’u chos ‘byung and mkhas pa’i dga’ ston refer to the same region.  Shastri (2007) discusses the different spellings used for Spiti in Tibetan language sources. These include spyi ti, spi ti, pi ti and ci di.  Many writers believe that the authentic spelling is spyi ti, which is also consistent with the spelling in the oldest Tibetan language source, lde’u chos ‘byung (see, Shastri, 2007; Gergan, 1976; Rossi, 2002: Jhampa, 2008-2010; Tsetan, 1987).  An important reason for the lack of a standardized spelling is that most Tibetan language readers will pronounce spyi ti as “Chi ti” or ci ti, according to the central Tibetan system of pronunciation, whereas locally it is known as “pi ti”.  The people of Spiti do not pronounce the (superscribed) “s” and they likewise often do not pronounce the (subscribed) “y” (signs) that are featured in Tibetan texts (Jhampa, 2008-2010).  Thus, if we take out “s” and “y” from spyi ti, we are left with “pi ti”, the local pronunciation.  The lack of a standardized Tibetan spelling for Spiti can also be seen between the local monasteries: Kee Monastery uses spyi ti in its publications (Tsering, 2000), whereas Tabo Monastery uses spi ti on its website.  Thus it is safe to conclude, as others cited above have done, that the references to “spyi ti” in lde’u chos ‘byung and “ci di” in Mkhas pa’i dga’ ston are to Spiti, especially because there is no other place in the region with the same name.  	   71	  Functional similarities: Khangchen households as military regiments According to an elder from Spiti’s Hansa village (interview, November 5, 2011), the historical Khangchen households are also known as “dmag” (meaning “army” or “soldiers”) because each of these households was required to provide a soldier in times of war. This claim is supported by British records as well as by the fact that I heard several stories from my interviewees about Spiti men having to go to Ladakh to serve as soldiers during times of war.60  For example, Trebeck (1841, p. 64) writes that “[e]very family [or household] subject to taxes possesses a matchlock, a sword, and a bow and arrows.” This observation suggests that members of these households did indeed serve as soldiers.  During my fieldwork, I found that only some of the Khangchen households still possessed such ancestral weapons as many had either sold or lost them.    A local custom linking Khangchen households with the “soldiers” is found in a tradition of the gathering of men of Khangchen households, which is referred to as Mag Zom (dmag ‘dzom or “gathering of soldiers”).  This meeting is generally convened when there is an infringement of traditional customs or an assault upon Khangchen mens’ power privileges.  During my fieldwork, I heard several references to the meetings of Spiti’s Khangchen household representatives as the “gathering of soldiers”.61 Interestingly, the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  60 One of these stories, for instance, presents a theory as to why Spiti celebrates their new year earlier than is the general Tibetan custom.  According to this theory, one year the men of Spiti were summoned by the king of Ladakh to fight a war.  The timing of the situation was such that the men were unlikely to return by the new year, if they were to return at all.  In order not to miss the celebrations, and for whatever other reasons, the people of Spiti decided to celebrate the new year months before the actual new year. Since that time, according to this story, Spiti’s new year came to be celebrated in the tenth month of the Tibetan calendar.  There are also two other theories as to why Spiti celebrates their new year in the tenth month of the Tibetan calendar. See Tsering & Jhampa, 2011.	  61 These meetings are almost always significant for the region. The nature of these meetings sheds light on the role of Spiti’s “army” as being one concerned with the contemporary internal protection of social-customs.	  	   72	  purpose of these meetings was orientated around safeguarding the traditional customs that privileged Khangchen households.  In one instance, a “gathering of soldiers living above Sheela creek” was held in 2007 to discuss how to deal with caste members who had admitted their children into Buddhist monasteries despite it being against the custom for lower caste members to enter the monastery (see above and Tsering & Ishimura, 2012).  A more recent “gathering of soldiers” was called as a show of force to dispute Kaza villagers, affiliated with another monastery, who had challenged the sale of a piece of land in the village in the name of the head lama of Kee Monastery. A third meeting was held a few decades ago to deal with certain Dhutul household men who were conspiring to challenge Spiti’s system of primogeniture in the Indian court of law. In all three instances, the will or the power of the “soldiers” prevailed. These instances are relevant here not only because the men of Khangchen households are called “soldiers” but also because these customary gatherings functioned as a show of force.  The tone in which the stories were narrated to me, especially when references were made to the “gathering of soldiers”, implied that there could have been violence if the other party had not capitulated.   Having established that Spiti’s Khangchen households were known as and functioned in practice as a regiment in terms of their obligation and solidarity, I will now move on to arguments that link Spiti’s administrative system, as well as some aspects of local culture, to those of the Tibetan Empire of the 7th-9th century.   	   73	  Structural similarities: administrative organization of the Tibetan Empire and Spiti Spiti’s traditional administrative divisions were comprised of five subdistricts or subunits.  Each of these districts was in turn comprised of a group of fifty Khangchen households.  In other words, the original two hundred and fifty soldier households were divided into “five groups of fifty households” (known as lnga bcu lnga), each comprising an administrative district or unit.  Spiti’s five groups are Upper (stod), Middle (bar), Lower (sham), Pin (spin or sprin, a tributary of Spiti river) and Religious Estates (chos gzhis).  Similarly, according to Tsuguhito Takeuchi (1994) and Brandon Dotson (2007), the Tibetan Empire of the 7th-9th centuries was also organized into subunits of fifty soldier households, known as tshan.    One issue correlating Spiti’s five subdistricts to the Tibetan Empire divisions is that the Religious Estates subunit in Spiti may not have existed during the time of the Tibetan Empire because the oldest monastery of Spiti, Tabo, was built in 996 A.D.  According to Tibetan tradition, the main development of monasteries took place during the period of the Second Diffusion of Buddhism.  Here, I would argue that this does not necessarily mean that the fifty tax-paying households were created anew with the institution of the Religious Estates.  It is more likely that the fifty households existed before that time as a subunit group of tax-payers, and regrouped later as the Religious Estates.  This is because the basis of household taxation and administration is the farming land.  All farming land and irrigation sources in Spiti belonged to taxpayers. In order to institute fifty new tax-paying households, these households had to be given new farming land – which would also most likely entail finding new irrigation sources and building canals. This would 	   74	  require a tremendous investment of resources from the state/kingdom, especially given the fact that there are many more and larger monasteries in the kingdom of Guge.62  Instead, it is more likely that the already existing tax-payers were assigned to different monasteries for their support. Additionally, the Tibetan Empire organized its administrative divisions according to geographical locations (Uray, 1982, p. 545), a practice which is also consistent with the administrative organization of Spiti.  Except for the Religious Estates subunit, all other administrative subunits of Spiti are neatly grouped according to geographical locations.  In the case of the Religious Estates subunits: all but two villages are neatly clustered together along the upper Spiti River.63  In addition to the structural and functional similarities between the two administrative systems mentioned above, there are also other interesting similarities. These include the use of certain terms or uses of language and mountain top fire-raising stations that are also recorded in textual sources from the time of the Tibetan Empire. There are also other factors such as the fort-like architectural design of Spiti’s villages, which again hints at their military past, and social customs of inheritance that are believed to have been written during the time of the Tibetan Empire.  After discussing these points, I will end 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  62 According to local belief, the builder of Tabo Monastery, Rinchen Zangpo, built a total of 108 temples in different parts of the kingdom. 	  63 The two villages that are not clustered with other villages of the subunit are Lithang and Tabo. Two subunits belonged to Lithang, situated in the middle of Spiti: Religious Estates and Lower Unit (sham).  Lithang farmers paid taxes to both the monasteries and to the state (see, e.g., Appendix I of Punjab Government [of British India], 1897, for a list of the distribution and division of villages into different administrative units). As for Tabo village, there is an interesting local story I heard at Kyomo Village, which is also part of the Religious Estates unit. Seven households from Kyomo village pay tax to Tabo Monastery - even though the village and the monastery are located at opposite ends of the valley. The reason for this, I was told by a villager, is that the monastery was originally planned for construction at Kyomo but was later built at Tabo. I was also told that the ruins of the initial monastery are still visible at Kyomo.	  	   75	  this section by considering how the hypothesis fits in with the broader regional historical context.  Council meetings (‘Dun ma) of Spiti and the Tibetan Empire In the course of my research and learning of the local language, I learned that the dialect of Spiti uses many words that are called brda rnying, or “archaic words”, by speakers of the central Tibetan dialect.  These archaic words are found in Tibetan texts or used only in honorific form for sacred objects or in religious contexts among speakers of the central Tibetan dialect.  For example, a Tibetan medicine practitioner from Spiti told me during an interview that the vocabulary of the Tibetan medical text, rgyud bzhi (the Four Tantras), will be more familiar to a speaker of the Spiti dialect than to a speaker of the central Tibetan dialect because the language used in the text is more similar to that of Spiti than to the central Tibetan.64   An important archaic word that links Spiti directly to the time of the Tibetan Empire is ‘dun ma.  In the local Spiti language, the Village Council meetings are called ‘dun ma.  This word is rarely used in the central Tibetan dialect to mean “meeting” but is rather more commonly used to mean “advice” or “wish”.  What is interesting in the case of Spiti’s usage of the word is that a very similar and common usage is found in the Old Tibetan Annals, Tibet’s oldest extant historical record dating to the time of the Tibetan Empire (Dotson, 2009).  The Old Tibetan Annals is a document written by the court 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  64 Although Tibetan medicine practitioners believe that the Four Tantras of Tibetan Medicine was composed in the eighth century by Yuthok Yonten Gompo, scholars have questioned this assertion.  Todd Fenner argues that the text was “written at a time when the [medical] traditions of India, China and Greece met and blended together within a Buddhist framework” (Fenner, p. 467).	  	   76	  historiographers of the Tibetan Empire, chronicling the main bureaucratic and administrative activities of the kings and ministers of the Tibetan Empire.  In this text, the word ‘dun ma is used at least 35 times to refer to the “council meetings” of the ministers and soldiers.65  Considering the military regiment feature of the Khangchen men and that the Village Council meetings (‘dun ma) were traditionally only represented by Khangchen men, the contextual usage and meaning of the word is strikingly similar to those found in the Old Tibetan Annals.  Thus, in the context of my proposed hypothesis, the use of the word ‘dun ma for Village Council in Spiti seems to be a remnant of an administrative terminology which characterized the military administrative system of the Tibetan Empire. Originally, this word would have referred to a “council meeting” of military officials.   Fire-raising stations: a military alert system in Spiti and of the Tibetan Empire  Several elders who spoke to me about Spiti’s early history mentioned that there are a few fire-raising stations situated on strategically chosen mountains tops and that these originally functioned as an alert system.  According to local history, whenever Spiti was under attack or under the threat of attack from foreign armies or militia, designated firemen (so pa) would light huge fires at these mountaintop structures as a means to send warning signals to other parts of Spiti. According to a teacher in Kaza, the main fire-raising stations were located at every site through which an invading army could enter 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  65 The word ‘dun ma is mentioned in records of the Old Tibetan Annals for the following A. D. years: 673-674, 674-675, twice in 681-682, twice in 682-683, 684-685, twice 694-695, 695-696, 698-699, 701-702, 702-703, twice in 704-705, twice in 706-707, 707-708, 708-709, 709-710, 710-711, 711-712, 714-715, 715-716, 717-718, 722-723, 723-724, 727-728, twice in 728-729, 730-731, 732-733, 733-734, 734-735, and 761-762 (Dotson, 2009).  This is excluding the use of a shortened version of the word, the syllable ‘dun, which is also used numerous times in the text. 	  	   77	  Spiti: near the head of the river valley (Losar village), near the end of Spiti’s traditional border along the river valley (Lari village), and near mountain passes such as Parang-La and Pin Parvati pass.  In addition, there were several other sites in between these sites. They were located such that they could be seen by as many distant villages as possible and so that further signals could be sent to warn as many villages as swiftly as possible.    The mountaintop fire-raising stations are relevant to this hypothesis because the Old Tibetan Annals have three separate references to them (see Dotson, 2009).  The first instance is most interesting.  According to the Annals, in the summer of the year of the dog (674-675 A.D.), the emperor resided in Zrid.  Then, Spiti’s neighboring region of Lcog-la - which is the same place as Gu-cog of lde’u chos ‘byung and Cog-la of mkhas pa’i dga’ ston (see above) - revolted against the empire and the emperor departed to Tshang-bang-na.  Thereafter the council meeting (‘dun ma) was convened, which took account of the fire-raising stations (see Dotson, 2009, p. 205, for an image of the original text of lines 60 and 61, or pp. 90-91 for an English translation and transliteration of the relevant lines).  The second instance is recorded in the year of the hare (691-692 A.D.), when the council convened at Sky Bra-ma-tang and made a selection of fire-raising stations (Dotson, 2009, p. 97).  Then, in the year of the bird (709-710 A.D.), the council (‘dun ma) convened at ‘On-chang-do, where they took account of the fire-raising stations of Rulag.  These references to fire-raising stations in the Old Tibetan Annals, especially the first reference of 674-675, give strong indications that Spiti’s fire-raising stations are a remnant of the Tibetan Empire’s alert system.  There is, however, one contradiction to this idea. 	   78	   Tsetan (1987) provides the only published account of Spiti’s fire-raising stations.  Tsetan (1987, p. 35) does not relate these to the time of the Tibetan Empire but refers to the fire-raising stations as evidence of a period when Spiti became a target of frequent attacks and robberies from its neighboring regions; this was due to the fact that the rulers of Guge and Ladakh did not pay attention to Spiti because of its remoteness.  According to Tsetan, these circumstances – Spiti being left without protection and under the mercy of frequent attacks – necessitated the building of these “small houses on top of mountains, called So-sa, the ruins of which still exist” (p. 35).  According to Tsetan, when these structures were lit up, the people of Spiti were required to go in the direction of the fire with bows, arrows, swords, spears and other weapons to fight against the intruders.  Thus, despite Tsetan’s different idea of the origin of Spiti’s fire-raising stations, his description of the function of these structures as a part of an early alert system is similar to those provided by my informants and as indicated in the Old Tibetan Annals.  As for Tsetan’s later periodization of the building of Spiti’s fire-raising system, it is not unlikely that the people referred to by Tsetan were rebuilding an older alert system.    Other similarities There are several other features associated with Spiti that could also be connected with the Tibetan Empire. These include a system of corvee labor that existed in Spiti during British Ladakhi and British times (Bray, 2008; British records) but also extends back to the time of the Tibetan Empire (Dotson, 2009). Since the time of the Tibetan Empire, corvee labor was one of the many forms of taxation, along with military service 	   79	  (discussed above) and the payment of grains, that taxpaying (khral pa) households were customarily obligated to contribute.  Another aspect of local custom that seems to have originated during the time of the Tibetan Empire is the system of inheritance based on patrilineal descent and primogeniture, which included legal mechanisms to keep the family lands intact.  According to Rebecca French (1990), these legal practices were “recorded as early as the period of the Tibetan Empire in the eighth century” (p. 470).  While I have not been able to locate any textual evidence associated with these specific legal records mentioned by French, I did observe in the Old Tibetan Annals that the Chief minister, Mgar Stong-rtsan, of the Tibetan Empire wrote the texts of the laws in the year of the hare, i.e., 655-656 A.D. (see Dotson, 2009, p. 85). Instead of simply discussing other examples 66 that link Spiti to the Tibetan empire, I will now cross-check the hypothesis from a broader historical perspective.   Cross checking hypothesis in the history of Spiti A key idea presented in the above discussion of the hypothesis is that the Khangchen system of administration in Spiti was instituted during the time of the Tibetan Empire. One method of checking this idea is to consider all the kingdoms, empires and states that ruled Spiti and to eliminate those that are unlikely to have introduced it. I have presented here a timeline of Spiti’s history according to the major powers that ruled over it.  Since Spiti’s Khangchen system is based on Tibetan administrative laws and practices, we need 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  66 Yet another example is the fort-like architectural design of Spiti’s villages. Houses in Spiti’s ancient villages, such as Khyung, were traditionally built very close or attached to each other and the whole structure was built such that there were a few (normally three) entrances with heavy gates that could be shut to keep intruders out. In other words, the village was a military fort (mkhar). In an interview with a Khangchen man of Khyung village, both the Tibetan word mkhar and the Hindi word kila were used to describe the historical housing structure of the village. 	  	   80	  to only consider the Tibetan kingdoms and empires that ruled Spiti, which leaves us with five options: the Tibetan Empire (7th - 9th century AD); Guge (late 10th - 1630); Ladakh (from 1634-1683/84, and again, from 1758-1842); Tibet (1683/84-1787); and Purig (1734-1758).67  From this list, we can eliminate Tibet (1683/84-1787) because the period of rule (3 or 4 years) is too brief to have introduced a lasting system of socio-economic administration.  We can also eliminate Purig (1734-1758) because the period of rule, 20 years, is still too brief.  Besides, as Jahoda (2009) points out, Purig’s rule over Spiti was most likely only nominal: “as in later times when Spiti was under the authority of the kings of Ladakh, the functionaries of [Purig] were rarely present in Spiti and that the immediate exercise of his power was confined to the collection of the collective total sum of revenues” (p. 46).    After eliminating Tibet and Purig from the Tibetan (Buddhist) kingdoms and empires that ruled Spiti, we are left with three powers that lasted for centuries: the Tibetan Empire, Guge kingdom and Ladakh kingdom.  Among these three powers, it is rather safe to assume that the two later kingdoms of Guge and Ladakh adopted or modeled their system of administration based on the earlier rule, i.e., that of the Tibetan Empire, especially because there is no published historical record that shows otherwise. The assumption that the Guge and Ladakh kingdoms adopted the administrative system of the Tibetan Empire seems plausible also based upon the fact that the Khangchen system of taxation is common to villages across the Tibetan Himalayas – an area stretching from Ladakh (Crook, 1994; Gutschow, 1998) in the western Himalayas, to Ding ri (Aziz, 1978) in 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  67 The dates, starting from the kingdom of Guge of the late 10th century to the Ladakh kingdom of the 18th century, are based on Jahoda (2009).	  	   81	  central Himalaya, all the way to eastern Himalayan regions such as Mdo mkhar of Mtsho sna rdzong, which is close to the border of present day Arunachal Pradesh and Bhutan (Personal communication, Tsering Dolma, May 1, 2013).68  Although a shared culture and practice do not necessarily equate with common rule, it is probable that this system of taxation was created during a time when the Tibetans ruled a vast region extending from the western to the eastern Himalayas.  Among the three powers that ruled Spiti for a long period, the Guge and Ladakh kingdoms were limited to Western Tibet and did not rule the central or eastern Himalayan regions.  Thus, again, it is probable that it was the Tibetan Empire that first established the system of administration.   Administrative power structure in the periphery: Ladakhi and British rule The oldest published record of Spiti’s internal taxation and administrative system comes from 19th century colonial British records (Moorcroft & Trebeck, 1841, pp. 69; Lyall, 1874; Egerton, 1864/2011).  When the British took control of Spiti in 1846, their policy was to continue to use the Khangchen household as the basic unit of taxation and administration as had been the custom under the Ladakhi rule.69 The British officials prepared detailed records of the local taxation and administrative system, which they described as having come from the Ladakhi system.  Using these records and other publications (Gergan, 1976; Jahoda, 2008), aided by interviews with local scholars and 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  68 The system or the household types can have different names in these different parts of Tibet. For example, in skyid grong, Khangchen households are known as khral ‘dzin, Khangchung households are called rgan tshang (“elder’s house”), and Dhutul households are called dud chung (see Childs, 2000).	  69 The kingdom of Ladakh had lost power over Spiti to the kingdom of Jammu in 1842 after a treaty was agreed upon between the Dogras and the Tibetans that confirmed the deprivation of Ladakhi kingdom’s powers. This episode of Jammu’s rule over Spiti is not discussed here because Jammu’s rule over Spiti was for a short duration and it did not affect social relationships in Spiti. Of interest though is the fact that, according to the 1883 Gazetteer (p. 498), Spiti started paying taxes to Jammu in 1839. Another noteworthy point about the Jammu rule is that a Sikh force plundered Spiti during that time (Gazetteer, 1883, p. 498).	  	   82	  elders, especially those who personally knew the last Nono (local ruler) in charge and had witnessed his court in session, I will now describe the administrative system of Spiti during the 19th and the first half of the 20th century.   Spiti’s local administrative structure is comprised of three levels: village, subunit groups of fifty households, and the Spiti Valley as a whole.70  This three-tiered administrative structure functioned from at least the beginning of the 19th century (Lyall, 1874, p. 181; Egerton, 1864/2011) to the middle of the 20th century (interview, 24 October 2010).   At the village level, all internal matters – legal, economic and political – were overseen and controlled by the men of the Khangchen households, whose meetings (‘dun ma) constituted the Village Council. The other two classes of households—Dhutul and caste—did not enjoy a decision-making vote in the Village Council since they neither owned land nor were required to pay taxes to the state.  A village chief, known as rgad po (“elder” or “old man”), headed the Village Council. The rgad po was assisted by two or three men (rgan las), including a bookkeeper (rtsis pa) and a storekeeper (gnyer pa). The rgad po and rgan las team represented the leaders of the village. They were mainly responsible for conducting Village Council meetings, maintaining village accounts and deciding upon the timing of communal activities (such as the commencement of the harvest season or the imposition of temporal bans on the collection of dung). Whenever the Village Council was unable to decide on a matter, it consulted the village deity (yul lha) through a spirit medium (lus g.yer). Therefore, although the Khangchen men held all 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  70 The local names for these three units are yul for village, Nga bcu nga for the five Kothees, and spyi ti lung pa for Spiti Valley.	  	   83	  the powers, the village deity and its spirit medium constituted a separate (local religious) institution that advised, and sometimes even changed, the Village Council’s decision.   Above the Village Councils were five administrative subunits or subdistricts of fifty households, which have received a new administrative name, “Kothee”, since the British rule.  Each group of fifty households was headed by a Great Elder (rgad po chen mo in the local language, or “lambardar” in the new administrative system),71 who acted as an intermediary authority between the ruler(s) of Spiti and the Khangchen households.  The Great Elders represented their group of fifty Khangchen households to the Nono and conveyed orders or messages from the Nono to members of the fifty households under their authority.  The Great Elders were also responsible for arranging the corvee labor requirements and for the collection of supplies on behalf of travellers (Lyall, 1874, p. 188).  At the level of legal administration, the Great Elders were members of the Nono’s (also known by a new official title, “Wazir”, under the British) privy council, which often functioned as the local court to adjudicate petty local legal cases.  Above the Kothee was the privy council of the Nono, represented by the Nono, the five Great Elders and the Nono’s secretary (known as “munshi”, who also served as the revenue officer of “Patwari”) (see Gergan, 1976, p. 323).  The powers of the Nono, or rather the limits on the powers of the Nono in the adjudication of legal cases, were set in the 1873 Spiti Regulation Act (Punjab Code, 1888, pp. 373-375). For example, serious crimes (such as murders or any offenses requiring fines in excess of two hundred rupees) 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  71 According to a local elder (Interview, 24 October 2010), these representatives were chosen on the basis of their ability to speak in public and their personality.	  	   84	  were to be reported to the Assistant Commissioner of Kulu.  Therefore the privy council of the Nono, which was directly under the authority of the Assistant Commissioner of Kulu, served as the local court responsible for dealing with petty internal cases.   Prior to the British rule, i.e., when Spiti was under the rule of the Ladakh kingdom, there was at least one other official besides the Nono who held a position above the Great Elders. This was the “Lord of the Fort” (mkhar dpon), who served as the head of the military and who was “responsible for order and the administration of police and justice” (Egerton, 1864/2011, p. 51-52).  The Nono, on the other hand, was the administrative head of civil and economic affairs.  The position of the Lord of the Fort (mkhar spon) was deemed unnecessary and eliminated by the British authorities who considered that only one ruler was desirable (Egerton, 1864/2011, p. 51).72  There was however another position, that of the Secretary of the Nono (Lyall, 1974, p. 188), which apparently was slightly lower than that of the Nono, which, as mentioned above, was continued under the British rule.   Thus the British allowed the traditional system of taxation and administration to continue, albeit by eliminating the position of the Lord of the Fort and by consolidating the powers of the Nono.73  This description of Spiti’s old administrative structure, derived from all known sources, complements the information provided in the two previous sections. In 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  72 Egerton (1864/2011) also mentions the “Little” Nono of Pin Valley, who wanted to establish a separate jurisdiction in Pin Valley, independent of the Kyuling Nono. This request was rejected. 	  73 The Spiti Regulation Act was passed in 1873  (Punjab Code, 1888, p. 373-375) and outlined the powers and responsibilities of the Nono, and his relation to higher regional authority. The final clause of the Act shows that Spiti’s administration was unique and different from that of the rest of British India: “No law hereafter passed by the Governor General in Council shall be deemed to extend to the Pargana of Spiti, unless the same be specially named therein” (Punjab Code, 1888, p. 375).	  	   85	  what follows, I will describe some of the major changes that occurred in local social relations, first under colonial British rule and then under India.  Impacts of colonial British rule One of the best ways to ascertain the social consequences of colonial British rule is by considering how its laws and legal practices would have impacted different sections of society.  The most important aspect of British laws, especially in rural places like Spiti Valley, was the revenue settlement, which determined the tax obligations of landholders.  The revenue settlement required not only the registration of all taxable land but also, in the case of what were the “tribal” areas of Himachal Pradesh, systematic documentation of all the local customs (including practices and laws). These later served as the legal administrative papers of the colonial government and aided in the resolution of conflicts.  Although Spiti was not a major source of revenue for the British, it was nevertheless affected by broader colonial policies and practices related to revenue settlement and collection.    In Spiti, the first revenue settlement was done in the same year Spiti was incorporated into British India in 1846.  The revenue settlement of 1846 lacked details and was conducted without sufficient understanding of local conditions – it had overlooked the existence of monastic tax system – which later became a cause of confusion for British officials and a matter of complaint on the part of the farmers (Lyall, 1874). In response, an official by the name of Lyall visited Spiti in 1867 to revise the revenue settlement and to ensure that the confusions and the complaints of farmers were addressed and resolved, 	   86	  which he claims to have done successfully. Lyall also reports documentation of the customary laws of every village in Spiti (Lyall, 1874, pp. 234-235). Records of these documents are still used by the Indian Government today, and are directly relevant to and therefore analyzed extensively in this study (in chapters 4 and 6).   An important effect of British laws was that they introduced the “private ownership” of land.  Prior to British rule, although tax-paying farmers (Khangchen households) enjoyed de facto ownership of land without interference from the state, in theory all land belonged to the king of Ladakh.74 The British laws changed the status of taxpayer farmers from being that of tenants under Ladakhi and Tibetan law (Jahoda, 2008; Pirie, 2005) to being that of formal landowners (Lyall, 1874; Jahoda, 2008). As a part of British policy regulating the expansion of agriculture to generate higher revenue, the British also introduced new laws in Spiti. These included the Nautor rule, whose objective was to break up wasteland for cultivation.  A part of the regulatory procedures for the Nautor land grant required approval from the local Village Council, which was exclusively composed of men of Khangchen households (see chapter 6).75  The British laws 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  74 Wylie (1986, p. 6) describes the land tenure system of Tibet, which is similar to that of Ladakh, in the following manner: “As a general rule, the tax estate belonged to the [tax-payer or Khangchen man] and his heirs forever. It could be transmitted to his heirs either during [his] lifetime or at death. The property was forfeitable [by the state authorities or officials on behalf of the king] only in the event of chronic failure to pay taxes, conviction of treason or failure to have an heir. The [tax-payer] was entitled to quiet enjoyment of his land and to manage the farm as he liked without interference from the authorities.” 	  75 If the proposal for the Nautor land grant was approved by the Village Council, the proposal would then require the final approval of the Nono, who had the authority to grant land on behalf of the government. Although no data on the amount of Nautor lands distributed during British rule are available, people mentioned cases of Nautor land grants occurring during the period of British rule. For example, the Nono had granted himself Nautor lands in Dhangkar as well as in Kyuling village (interview, 24 October 2010). I also heard a dramatic story of a man who was believed to be the first man in Pin Valley, probably in Spiti as well, to get Nautor land during the British rule. According to the story (narrated to me by a farmer from Guling village of Pin Valley), villagers refused to acknowledge his Nautor land rights and even decided to attack him. The man escaped in the middle of the night and travelled on foot to Kulu to seek justice from a higher authority, which most probably was the Assistant Commissioner of the district.	  	   87	  legitimized and consolidated the traditional powers of these men through a number of ways.  These included: codification of local customs that tended to privilege the men of Khangchen households, a change in their legal status from that of tenant farmers to official land owners, and the provision of new powers in the regulation of agricultural expansion. Consequently, the relative power positions of traditionally underprivileged groups, such as women and members of landless Dhutul and caste households, were undermined (Baker, 2003).  Socio-economic development of a frontier region under independent India When India became independent in 1947, local socio-economic conditions were similar to those that existed centuries ago. The administration of the valley was still organized according to five groups of fifty soldier households under the Nono. There were no paved roads, cars, dispensaries,76 telephones,77 or electricity.78  The literacy rate was extremely low, especially for women and caste members. People’s main livelihood was based on subsistence agriculture and a significant portion of the population was extremely poor.  Things began to change under Indian rule. Taxation was lifted so that no one had to pay taxes to the state anymore. In terms of local administration, India continued the British policy of governing the Himalayan “tribal” regions, including Spiti, through local customs, as recorded in administrative papers such as Riwaj-i-am and Riwaj-i-Abpashi. This policy allowed many aspects of Spiti’s traditional customs and practices, such as those of farming and local resource management (described in the following chapters), to 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  76 The first dispensary in Spiti was built in Kaza in 1958 (Gazetteer, 1975, p. 247).	  77 Telephone services were first implemented in the villages of Losar and Sagnam in 1999 by Bharat Sanchar Nigam Limited.	  78 The Rongtong power plant of Spiti was inaugurated in 1986 (interview, 21 August 2010).	  	   88	  continue as they had under Ladakhi and British rule. However, unlike the British and Ladakhi rulers, who took a distant approach to administration in Spiti and left local socio-economic conditions unaffected, India took a much more active role in the local administration of Spiti and invested heavily in the region to develop socio-economic and road infrastructure conditions, especially after the 1962 India-China war over border disputes. The Indian government’s active role in the region brought significant improvements to people’s socio-economic and living conditions, especially those of women, caste and Dhutul members.  Starting from the mid-1950s, a series of administrative and legal initiatives were implemented in Spiti. Jahoda (2008) has analyzed the socio-economic implications of some of these initiatives. He notes how the implementation of the 1950 Punjab Tenants Act set a limit on the amount of land held by the Tabo monastery (to 250 bighas or 50 acres), which resulted in the monastery’s registered tenant farmers gaining legal ownership of the land (Jahoda, 2008, p. 13). Another important change was the introduction of Indian Police in Spiti and the dissolution of the Nono’s official powers (interview, 21 August 2010).  During the same period, the Panchayat system was introduced to replace the traditional administrative subunits of fifty households (Kothi) as well as the village headman (rgad po).79 Unlike the traditional administrative institutions (Village Council, Nono’s council, etc.), which were only reserved for men of Khangchen households, the leadership positions in the Panchayat system were open to men, as well as to women, from all types of households. However, based upon conversations and 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  79 According to the new Panchayat system, Spiti was divided into several small groups of villages or Panchayats. In 2011, there were 13 Panchayats in Spiti.	  	   89	  interviews, it seems that men of Khangchen households continue to dominate this local level of politics.   The Panchayat system, however, has not fully replaced the traditional systems of administration. Today, there is a co-existence of the Panchayat system with the local Village Council headed by Gadpo (rgad po or elders) (Jahoda, 2008, p. 13).  The traditional administrative institutions did not disappear because they played an important role in aspects of people’s social, cultural and religious lives which are beyond the purview of the state.  In other words, this led to the separation of the public and private spheres of interest; here the governance of public affairs of the state were taken over by the Panchayat committee and the governance of the private or internal affairs of the village continued to be undertaken by the Village Council.  For example, farmers’ affiliation to local monasteries is organized on the basis of a fifty household subunit system, where different groups of fifty households are affiliated with different monasteries. 80  Similarly, the village headman or the Gadpo was responsible for matters concerning the village temple, the village deity medium, as well as the performance of various rituals related to farming (see chapter 5) and other resource management (see chapter 6) customs.  Therefore the new administrative system of Panchayat became formally responsible for administrative works related to the state - for instance, those related to development initiatives and public elections. On the other hand, the traditional Village Council became responsible for internal village matters and it continues to play a very important role in the customary lives of the farmers.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  80 Monasteries play an important role in farmers’ lives, including matters associated with the performance of rituals during important occasions, the compulsory admission of farmers’ sons to these monastic establishments and the collection/payment of monastic taxes.	  	   90	   Starting in the late 1950’s, the Punjab Armed Police (Spiti was then a part of the Indian state of Punjab) began road construction to Spiti (interview, 21 August 2010).  At that time, there was an exodus of Tibetan refugees arriving in India, many of whom were hired to do road construction work in the region. Then, in 1962, war broke out between India and China over border disputes. Spiti was closed off to all outside visitors and designated an “inner line” area due to its proximity to the border, even though Spiti was not a site of any militant encounter. On the one hand, this law resulted in Spiti becoming more isolated from the rest of India and the world at large, but on the other, it attracted the government’s attention regarding the need for road infrastructure development in the region.  Road construction work was speeded up, and thereafter came under the supervision of the Border Road Organization (BRO) of the Indian army.  Despite Spiti’s difficult terrain, altitude and long snowy winters, the BRO was able to complete the work relatively quickly and the first public bus service connecting Spiti to the nearest Indian city of Manali was introduced in 1972 (interview, 21 August 2010).  In addition to the strategic development of road infrastructure around the border region of Spiti, the Indian government also invested heavily in socio-economic and educational development programs. The strategic thinking of the government was aptly described by a senior local leader, who was also a retired Indian army officer: “Ensuring continued human habitation in ‘border regions’ and improving local livelihood with integrated socio-economic development programs is beneficial for the security of the country” (interview, 21 August 2010). Starting during the time when Spiti was closed off from the 	   91	  rest of the world (1962-1993) and continuing to this day, a series of government policy measures and development programs have been transforming a largely subsistence and barter-based economy into a cash economy that is more integrated with the larger political economy of the country. The first major socio-economic development initiative was the distribution of land to landless farmers - including distribution to women and caste members - through the 1968 Himachal Pradesh Nautor Land Rules. Significantly for the arid mountainous region of Spiti, the Nautor fields were connected with new irrigation canals to distant water sources, a project whose construction began in the early 1970’s (interview, 21 August 2010). The next major initiative was support for commercial crops. Today, green peas, introduced in 1986 (Mishra, 2000; interview, 21 August 2010), is the main commercial crop, grown throughout Spiti Valley. Farmers in lower Spiti have also successfully grown apples for at least two decades. Today, the extent of land cultivated in Spiti has more than doubled due to the Nautor land distribution, thereby bringing substantial cash income to the region (interview, August 2008).   Other major sources of cash income, in addition to those facilitated by the Nautor fields and cash crops, include government jobs, government aid money, and tourism. In the beginning, the locals were reluctant or did not want to work for the Indian government. As the British records tell, people of Spiti were not interested in things that were alien to their (Tibetan Buddhist) culture. This was confirmed in interviews, where people shared stories of how their family members first refused to work for the Indian government or to send their children to Indian government schools. Today, however, government jobs are 	   92	  highly coveted and a common source of livelihood. For example, according to a statistic published by Jahoda (2008, p. 15), “[I]n 1993-94, out of 4,021 individuals aged fifteen and over [in Spiti], 1,541 were on the payroll of government agencies.” Similarly, education through government schools is a standard practice in Spiti. When India became independent in 1947, there was barely any school in Spiti.81 In 2010, by contrast, there were about a hundred schools in the valley (Tsering & Ishimura, 2012). As for an example of government development aid in the region, Rs. 841 million was reportedly invested in Spiti between 1978 and 1997 through programs such as Integrated Tribal Development Projects and Tribal Sub Plans (Jahoda, 2008, p. 14).82 While the government’s extensive investments in the socio-economic development of the region are appreciated by the local people, many local leaders expressed their disappointment with corruption in the system; as one of the locals claimed, by the time government aid reaches people, there is only 10% of the funds left. Similarly for education development, while one can boast about educational accomplishments in Spiti in terms of numbers, the quality of education is nevertheless poor (Tsering & Ishimura, 2012).   The impact of government socio-economic and education development investments in Spiti is at once significant, complex and still unfolding. While the general improvement in the socio-economic and living conditions of the people is undeniable, the benefits of government development programs, such as the Nautor land distribution, have affected the society unevenly. Men, especially those from Khangchen households are the main 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  81 Bajpai (1987, p. 133) says that the first school in Spiti was established in 1932 in Kaza village. Local elders opine that there was “barely any school” (“na ke barabar” in Hindi) when India became independent. 	  82 This amount (Rs. 841 million) does not include funds invested into the region through another major project called the Desert Development Program.	  	   93	  beneficiaries, while women and caste members seem to have benefited less. This is not surprising considering the fact that men, especially those from the Khangchen and Dhutul households, are more educated and also better connected politically, as compared with women and caste members. Yet, curiously, caste and women interviewees seem to exhibit more gratitude toward the government, apparent from their more frequent use of the phrase “the gracious and kind government” (“Sarkar drinchen”), to give credit to the Indian government for the positive changes it has brought to their lives. This is understandable if one considers how the opportunities provided by the government are not only open to all the citizens, irrespective of caste, class and gender differences, but also to the fact that government programs (or positions) are often targeted at benefiting underprivileged groups. Therefore, despite the ability of the local elites to take greater advantage of opportunities provided by market and state led development programs, underprivileged groups have been greatly empowered under Indian rule.   As members of traditionally underprivileged groups became educated and started owning their own lands, traditional labor relations between different classes or groups of people in farming and resource management roles were also affected. As labor became more scarce, women, who customarily do most of the time-consuming farm labor, had to work more. Owing to the high demand for farm labor, it became common for girls from poor families to discontinue school in order to dedicate themselves to farming tasks (interview, 2010).83 As for Zo blacksmiths, their skills became obsolete as ready-made tools from the Indian plains became available in the market. Today, Zo households are actively involved 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  83 In order to address the issue of poor girls being kept at home for farm labor due to a lack of proper schooling and hostel facilities, some locals have started a girls hostel project that I am supporting: www.spitivalley.org 	  	   94	  in farming – both on their own fields and on other farmers’ fields in the role of plowmen and irrigators (discussed in the following chapters). Owing to the large number of landowners, many Beda musicians have decided that they are unable to perform according to the customary, non-monetary contractual relations of performing music on demand. Instead, these Beda musicians perform their musical services for secular celebratory occasions on a “daily wage” basis. However, Beda musicians continue to play music for religious and ritual purposes, especially for the village temple, according to custom. All these changes mean that the traditionally underprivileged groups are not as dependent on the Old Khangchen households for their livelihood as was formerly the case. Thus, the Old Khangchen households have lost some of their powers in village politics as many Dhutul and caste households, who have become economically well off, now pay “full” taxes to the village and thus have gained membership rights in the Village Council.   While Spiti has experienced significant socio-economic and educational development over the last few decades due largely to the investments made by the Indian government, there has also been a reinvigoration of Buddhism in Spiti in recent decades due to the involvement of Tibetan refugee lamas. Several senior Tibetan Buddhist teachers have visited Spiti several times, including six visits by the Dalai Lama. Some of these teachers stayed for long durations and gave extensive teachings. According to two local leaders, who spoke to me singly on separate occasions, the Tibetan Buddhist teachers who visited Spiti, especially the Dalai Lama, had a huge impact on the revival of Buddhism in the region. Referring to the implications of the exiled Tibetan leader’s visits to the 	   95	  historically peripheral (i.e., to Tibet and India) regions of Ladakh and Spiti, Tenzin Tethong, the Director of the Tibetan language service of Radio Free Asia, made a poignant remark on the changing nature of “Greater Tibet”: he noted that now the center is coming to the periphery (Conference lecture, July 23, 2013). The same analogy of the center actively coming to the periphery could also be noted in regard to India’s attention to Spiti - not only in terms of development investment in the region but also because India’s Prime Minister, Indira Gandhi, as well as several Chief Ministers, have also visited the region. All this reflects the attention Spiti has garnered in the last few decades, a particularly significant development after the centuries of neglect experienced by the region on the part of the central authorities.   Conclusion The central argument of this chapter is that the power relations embedded in the local socio-economic structure of Spiti Valley have always been changing due to, and in response to, the administrative laws and policies of the different empires, kingdoms and states that ruled the region. After presenting a description of the village socio-economic structure, I have demonstrated how certain classes of households were introduced and local power relations affected by the laws and policies of different powers, which in turn were often affected by Spiti’s border location.  A major contribution of this chapter is the original hypothesis presented which answers the question raised at the beginning - that regarding the historical origin and legal basis of the Khangchen household’s privileged access to local farming lands and irrigation 	   96	  sources. Drawing on ancient Tibetan texts, studies based on original source materials from the time of the Tibetan Empire and local oral history, as well as my own field research, I hypothesized that Spiti’s Khangchen system of taxation, which is directly associated with the privileges enjoyed by Khangchen households in terms of access rights to farming lands and irrigation sources, is based on the military administrative system of the Tibetan Empire that was established between the 7th-9th centuries A.D. This hypothesis is highly relevant for this study - not only because it provides a definite historical period and context for understanding the privileged powers of the contemporary Khangchen households but also for the light that it sheds on the origin of Spiti’s most prominent class of household.  After the collapse of the Tibetan Empire, the next major power that introduced important changes in local culture was the kingdom of Guge. Starting in the 10th century, the rulers of the Guge kingdom supported an active missionary campaign to introduce and establish Buddhism in the region, mainly through the construction of temples and monasteries, as well as through the translation of Sanskrit Buddhist texts into the Tibetan language. Spiti was an important site of these activities partly because it was located between the Guge kingdom and Kashmir. Kashmir was a center of Buddhist activity at that time, and a key source of Buddhist knowledge for the Guge kingdom. In addition to introducing Buddhism, the Guge kingdom also established monasteries in Spiti, and by implication, introduced the Monastic Estates taxation system. Guge was followed by the kingdom of Ladakh, which introduced the institution of the Nono, the local administrative ruler, who in turn introduced the Beda musicians into Spiti.  	   97	   While all the classes of households in the hierarchical village social structure – Khangchen, Dhutul and caste – were established by the time the British took over Spiti in 1846, British laws resulted in rendering the dominant Khangchen households more powerful, thereby weakening the traditionally underprivileged groups (such as women and members of caste and Dhutul households). Under the previous Ladakhi and Tibetan legal systems, although Khangchen men normally enjoyed full rights over their estates, they were in fact legal tenants, as all land belonged to the state. British laws privatized land ownership, making the Khangchen men registered owners of their estates.   In addition, the British also documented and legalized local customs, including those related to the management of resources such as irrigation, fodder and firewood. These measures provided lasting legal protection for ancient customs that privileged certain members of society over others. These customs were and still are highly inequitable because the privileged group of Khangchen men are a small minority compared to the rest of the village population, which is composed of women, children and elders as well as members of Dhutul and caste households. As discussed in the following chapters, these codified customs, in the form of government administrative papers, are still valid and continue to affect local farming and resource management customs.   Spiti saw the greatest changes, both in terms of material economic conditions and in terms of local power relations, under the rule of the current Indian State. These changes, which are still unfolding, are a result of the Indian initiative aimed at developing and 	   98	  integrating the local economy and administration with the larger Indian economy and administration. Certain policies have been implemented through significant economic investments and a variety of administrative measures introduced to consolidate state control over Spiti, which, since the 1962 India-China war, has been recognized as a sensitive border region. In terms of the impact on local power relations, these socio-economic development initiatives are being received favorably by the people, particularly by the underprivileged sections of the society, who often refer to the Indian government as “kind and gracious”. For the first time in history, women and members of Dhutul and caste households are not only able to own land and engage in commercial agriculture, but they are also able to acquire free education and get jobs in the service economy. However, the impact of the development initiatives on different sections of society are more complex and uneven than meets the eye. For example, the men of the Khangchen and Dhutul households have been able to take greater advantage of the opportunities provided by the state than men of caste households and women in general. Therefore, despite dramatic changes over the course of recent decades, the hierarchical social relations of power still persist, especially in the customary practices involving farming and resource management.    	   99	  4. Inequalities Within Women-Managed Irrigation Systems Introduction Gender has become an important concept in the field of environment and development studies, particularly in those studies that examine the social and political dimensions of water issues. Drawing on broader feminist literature on environment and development (e.g., Agarwal, 1986, 1998), the literature on ‘gender and water’ emphasizes the unequal power relations between men and women, mainly in terms of access, management responsibilities and decision-making roles (Meinzen-dick & Zwarteveen, 2001; Upadhyay, 2003; Adams, Watson, & Mutiso, 1997). The general premise of gender approach to water management is that society assigns different expectations, roles and responsibilities to men and women in the arena of water use, access and management, and that these are often tilted against women in patriarchal social, economic and political contexts. As gender analysis has become more ‘mainstream’ in the last few decades, many frameworks, methods and tools have been developed for use by development practitioners to measure, understand and improve gender dynamics or to overcome inequalities. 84 For example, Moser (1993) has developed methodological procedures and tools for development planning focused on gender (i.e., “gender planning”) which has been put forth “with a goal to emancipate women from their subordination, and to embrace equality, equity and empowerment” (p. 89). Similarly, Levy (1996) has identified thirteen elements in the form of a “web”, where every element provides the conditions for the institutionalization of gender. The idea here is for development practitioners to use this as a methodology for gender policy and planning, as well as a 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  84 For an overview discussion of these frameworks and methods, see Wallace and Coles, 2005, pp. 6-7.	  	   100	  tool for gender diagnosis and procedural operations.85  More recently, a set of critical arguments is presented as a response to these developments. At the level of discourse and practice, it is argued that ‘gender mainstreaming’ or attention to gender by organizations and institutions is increasingly being regarded as a technical issue, disassociating the gender agenda from its feminist transformatory objectives (Wallace and Coles, 2005). Although it has commonly been acknowledged that gender operates in interaction with other social categories such as class, caste and ethnicity (Meinzen-Dick & Zwarteveen, 2001; Wallace and Coles, 2005), it has been pointed out that there are a lack of studies that investigate the intersection of gender and other social difference (such as class, caste, race and ethnicity) in development studies in general (Nightingale, 2011), and in irrigation studies (Harris, 2008) in particular,86 It is within this particular framework, which calls for an intersectional analysis of gender and other social differences in irrigation management, that I am situating this chapter. This chapter provides an intersectional study of gender and equity in irrigation management through a detailed study of the historical contexts, legal texts, and farmers’ practices concerning irrigation management in the Himalayas.  Intersectional theory argues that although gender has powerful outcomes on people’s lives, it does not operate alone. Gender always operates in relation to other social differences such as race, class, caste and ethnicity and this in turn results in a 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  85 The thirteen elements in Levy’s (1996) Web of Institutionalisation are resource, mainstreaming location of responsibility, political commitment, policy/planning, procedures, pressure of political constituencies, representative political structures, staff development, women and men’s experience and interpretation of their reality, delivery of programs and projects, methodology, research and theory building. 	  86 For an excellent discussion of debates in intersectionality in feminist literature, see Fincher, 2004. 	  	   101	  multilayered system of disadvantage for some people and privilege for others. For example, in Canadian society, it can be said that aboriginal and black women from low-income households face multiple dimensions of disadvantage (of race, ethnicity and class) as compared to, for example, a white middle-class women. Therefore, intersectional approaches analyze the interplay of gender and other social differences and seek to unravel and shed light upon how certain groups are multiply disadvantaged while others are privileged. In this sense, the application of intersectional analysis to water management issues posits that although sex-gender based inequities around water issues are important, this alone is not sufficient to understand and address issues of inequity because sex-gender is only one of the many important intersecting layers of social relations that mediate different gendered water relations. Specifically, it has been pointed out that the intersection between gender and socio-economic class (Sultana et al., 2013) or poverty (Harris, 2008) is an important area for analysis and theorization. Validating these arguments, this chapter unravels intersectional differences among different groups of women, as contrasted with a men-versus-women approach, in terms of their socio-economic class and caste backgrounds.   This case study, which examines the inequality among women groups is significant in some ways. First, “too often, inequalities among women are overlooked in initiatives promoting women’s rights and interests” (Sultana et al., 2013, p. 13). This is true in general and for irrigation management in particular. Second, as the description of local irrigation activities in this chapter demonstrates, women are responsible for almost all of these activities. This is in contrast to the situation portrayed in the larger literature - 	   102	  which generally assumes women’s secondary status based on exclusion from irrigation practices. In Spiti, irrigation is the domain of women. However it is nonetheless the case that important processes of social differentiation and inequalities operate through their engagement and are reinforced by it, as well as being additionally reinforced through state practices.   This chapter makes several additional general contributions, adding to a growing body of ethnographic literature on the traditional Tibetan irrigation management practices and customs (Carrasco, 1959; Ekvall & Downs, 1963; Aziz, 1978; Norberg-Hodge, 1992; Gutschow, 1997, 1998; Labbal, 2000; Vohra, 2000; Gutschow & Mankelow, 2001; Gutschow & Gutschow, 2003; Mankelow, 2003; Sorensen, 2003; Gupta & Tiwari, 2008) which were discussed in the previous chapter.   This chapter is composed of three main sections. The first section provides relevant historical and geographical contexts that form the basis of some of the main features of the social organization of Spiti’s irrigation system. The second section provides a translation and analysis of the text of codified irrigation customs of Zibug and Kyung villages in Spiti Valley. I pay attention to the historical context in which the irrigation customs were codified in order to shed light on the nature of the language of the legal text and its impact on local power relations. I also analyze how unequal power relations in access to and management of irrigation water in terms of gender and landholdings are cemented in law. The third section, presented in order to highlight the inequalities and internal politics embedded in these activities, provides a detailed ethnographic 	   103	  description of irrigation-related activities performed by farmers (these practices are similar in both the villages) according to their ascribed gender, class and caste roles. The description of irrigation activities shows that women perform almost all of these activities. From an intersectional perspective, gender is unique because it exists in not only every household (Agarwal, 2007) but also in every socio-economic class, with each class often having a different set of expected roles. That is why one should not lump all women into one group and simply say that women perform irrigation-related tasks. Instead, one should analyze which categories of women perform which kinds of irrigation-related tasks; this is due to the fact that these tasks can be seen as hierarchical based on the ascribed roles of women belonging to different household types. In addition, inspired by Ribot and Peluso’s (2003) suggestions on empirically observing the differences between ability and right to benefit from things in understanding access to property, I analyze gender equity not only based on empirical observations of “who performs which irrigation tasks” but also on an analysis of “who are excluded and why”. I show how an analysis on exclusion, particularly covert exclusionary practices, can yield insights into deeper layers of micro-politics between different groups.   This chapter shows that the social organization of Spiti Valley’s irrigation system is based on social relations of gender, socio-economic class and caste. In the next two chapters, I show that these relations, including gender roles, can be better understood by situating these within the context of a broader division of labor. For example, in chapter 5, I show how agricultural labor roles – with irrigation being an important component of agriculture – are divided among farmers on the basis of their gender, caste and class 	   104	  backgrounds, with important differences in the intersection of gender and other social notions of difference. Similarly, in chapter 6, I argue that women’s role in irrigation management is partly a result of a broader division of labor between men and women, where men are responsible for the equally necessary and time-consuming task of collecting firewood. Throughout this thesis, I emphasize the historical context and cultural beliefs that provide meanings regarding the nature of these roles, especially as they pertain to the local people.   Having introduced the contributions and contents of this chapter, and its relation to the thesis, I will now briefly mention some important features of the social organization of Spiti’s irrigation system, which will then be further explained, in terms of their historical contexts, in Section 1.   Irrigation system of Spiti Valley and its social organization Kelly (1983) has noted that the “widely used concept of an ‘irrigation system’ typically conflates three distinct dimensions of agricultural water use: natural water flow patterns, physical networks of facilities and environmental modifications, and organization of configurations of irrigation roles” (pp. 880-881). It is therefore important to emphasize that this study is only concerned with the social organization of irrigation management.   Another analytical aspect of the study that must be clarified at the outset is that it is concerned only with irrigation systems related to old fields that are located next to the village. Generally, there are three main types of fields in Spiti: village fields, mountain 	   105	  fields, and Nautor fields. Village fields are the traditional or historical fields that are situated in and around the immediate vicinity of a village. These are the fields to which all irrigation and farming-related customary laws apply. Mountain fields (ri zhing) are located a few kilometers away from the village and are cultivated by one or a few villagers. Although farmers of mountain fields irrigate and cultivate their fields according to customs, they are not obligated to follow every custom. Finally, Nautor fields are those that were allocated by the Indian government to landless and poor farmers after the passage of the Nautor Land Rules Act. Nautor fields are located several kilometers away from most of the villages, including Zibug and Khyung.   Though Zibug and Khyung farmers own all three kinds of fields, this study is concerned only with the historical village fields. I have focused on the historical village fields because farmers’ work on these fields is regulated by stringent traditional customs. In the mountain fields and Nautor fields, farmers are not required to follow all the traditional customs. Thus irrigation practices in the historical village fields, even though these are performed in a contemporary context, can nevertheless be considered “traditional” practices. In addition, the codified irrigation customs of Zibug and Khyung villages, which are translated and analyzed below, only concern the village fields. Another advantage of focusing on village fields is that these are representative of the village. Mountain fields are not the ideal sites for a study of village customs because these fields are like a family compound, belonging to only one or a small group of households. Nautor fields are not only recent creations and situated far away from the villages but are composed of fields belonging to farmers of many different villages.  	   106	   Technically, as we shall see below, there are only two groups of irrigation users in the village fields: Khangchen households (representing one group) and members from the rest of village (lumped together and composing the second group). In terms of numbers, Khangchen households are the minority but they have most of the rights over irrigation water. Although the second group, comprised of Dhutul and caste households, are numerically much larger (numbers discussed below), they have very limited access to irrigation water. Other key features of the social organization of Spiti’s irrigation system include: 1) men of Khangchen households are the legal rights holders of irrigation water and they thus make most of the key decisions, such as the commencement timing regarding different irrigation activities; 2) women do most of the irrigation-related work; and 3) many women from Dhutul and caste households work as Chu pa (irrigators) for Khangchen households on a contractual basis. These features are explained in more detail in their historical and geographical contexts in the next section.  Section 1: History of water management customs and their implications on society The earliest historical records pertaining to Spiti say that it was once a part of the Zhangzhung kingdom, before the region became part of the expanding Tibetan empire in the 7th century.87 The legacy of the Zhangzhung period, or the period before Spiti came under Tibetan rule, is relevant to Spiti’s irrigation management customs because farmers have certain pre-Buddhist water-related belief systems and practices. 88 These include the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  87 The oldest textual evidence of Spiti belonging to the Zhangzhung kingdom is the 13th century Tibetan historical text lde’u chos ‘byung (discussed in chapter 3).	  88 These belief systems are common throughout the Himalayan regions, including among Indians, Nepalese and Tibetan peoples.	  	   107	  belief in transworldly serpent spirits called Lu (klu), or “naga” in Sanskrit, which the farmers believe reside in irrigation water sources (i.e., springs) and have the powers to control weather conditions, specifically precipitation (Interview, October 2, 2010). Since springs constitute most of the irrigation water sources for Spiti – the two villages where this study was conducted are entirely dependent on springs for irrigation water – the beliefs, rituals and practices surrounding the transworldly serpent spirits form an important dimension of Spiti’s water management practices and customs. These beliefs, rituals and practices are described in more detail below.   The historical context and hypothesis presented in the previous chapter are relevant to irrigation because, as will be shown below, in the study of de jure, rights holders of the traditional irrigation system are the men of the Old Khangchen households. In other words, according to traditional customs, all of the farming land and irrigation supplies, as well as the decision-making powers concerning the management of local resources, were under the control of the men of Khangchen households. When I asked these men why they have such privileges, several of them told me that their ancestors were the first founders of the village (Interviews, 2010, 2011). My hypothesis links the historical and legal basis of their privileges to the time of the Old Tibetan Empire because these privileges are representative of the privileges of tax-payers which were instituted at that time. Even today, for example, Khangchen households are known both as “tax-payers” (khral pa) and as “soldier” (dmag) households.  Here it is necessary to briefly mention some historical contexts discussed in the previous 	   108	  chapter. In the 10th century, after the collapse of the Old Tibetan Empire, Spiti became a part of a new regional and powerful Tibetan kingdom known as Guge. The kings of Guge kingdom played a leading role in reviving Tibetan Buddhism.89 Under Guge rule, Spiti became a key site of Tibetan Buddhist activity; this involved the construction and establishment of numerous monasteries, including the famous Tabo Monastery of Spiti, built in 996 A.D. (Klimburg-Salter, 1997).90 One of the customs associated with these monasteries is that the second or the middle son of the tax-paying Khangchen households must join the monastery as a monk. Even today, every non-caste household must enroll at least their second son in the local monastery. If they do not then they must pay a fine. From my field observations, it is common to find families where all the sons, with the exception of the eldest, who runs the household property, are monks.91 A latent function of the custom of monastic enrolment of younger sons was that this resulted in there being many unmarried women in society. These women formed the main labor force for farmwork (subsistence agriculture was the main source of livelihood). This partly explains why it is customary for women to perform the farm and irrigation work in 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  89 The first diffusion is referred to posthumously as the spread of Buddhism during the period of the Old Tibetan Empire.	  90 Although Spiti was under Tibetan rule before the 10th century, there is evidence that suggests that earlier people spoke a different language, often referred to as that of Zhangzhung. Based on a study of various names and words written on the walls of the Tabo Monastery at different periods of time, which show a gradual transition from an older language into Tibetan, Klimburg-Salter hypothesizes that the period of construction of the Tabo Monastery (996 A.D.) represented the starting point of the Tibetanization of the region. 	  91 Two points are worth mentioning here. First, historically, only boys or men from Khangchen households were allowed to join the monasteries. Then sometime after the middle of the 20th century, boys or men from Dhutul households were allowed to join the local monasteries. Boys and men from caste families are still not permitted to join the local monasteries. Second, there were no nunneries in Spiti until the latter half of the 20th century. According to the census of 1891, only 1.4% of the female population of Spiti was literate. For a historical discussion of education and social change in Spiti, with a focus on intersectionality, see Tsering & Ishimura, 2012.	  	   109	  Spiti.92  Another important factor that further explains why only women perform irrigation work in Spiti is its geographical and environmental contexts. Owing to its geographical location in the high Himalayas (both Zibug and Khyung villages are situated slightly higher than 4000 meters above sea level), Spiti has a short growing season and long snowy winters. Owing to these conditions, every household must collect a large enough stock of firewood and dung to ensure that their heating and cooking needs will be met throughout the long winter months. The small villages of Spiti are scattered along the river valley and each village has a relatively large commons areas within which farmers collect firewood and dung.93 Collection of firewood is a relatively challenging task, which, for various reasons (discussed in chapter 6), local custom has reserved only for men. One of the reasons is the physical difficulty of the task. The main sources of firewood are two thorny shrubs, Caragana versicolor (gra ma) and Lonicera (brab), which both grow in the highest and driest areas. The thorny shrubs are mainly uprooted with the hands, using brute physical force, and with the help of adze hoes. In addition, farmers must travel long distances with yaks to carry the collected firewood.94 The collection of firewood (and dung) from distant mountainous areas as an exclusive livelihood role of men has had direct implications for irrigation being the exclusive livelihood role of women. As mentioned earlier, farm work (except for activities that 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  92 This is not to say that traditionally monks did not perform agricultural labor. For example, during my fieldwork, I observed a man who was also a monk helping his elderly parents with the harvesting of barley. I asked several farmers if it is common for monks to do or assist in agricultural work. The answer I received was that monks may assist in agricultural labor but only in exceptional or rare cases.  	  93 During the interviews, when asked if the villages have at least 15 square kilometers of commons territory, everyone who gave a definitive answer said ‘yes’.	  94 Livelihood practices requiring long-distrance travel with yaks (e.g., for collection of salt or for trade), were the domain of men throughout the Tibetan inhabited regions.	  	   110	  require the use of yaks for plowing and threshing crops) is mainly done by women. During the short summer months, men have to go to the mountains regularly to collect (and stock) firewood, which leaves only women to do the irrigation work. The division of labor therefore had to be based on gender, as opposed to age, because both irrigation and the collection of firewood has to be regularly done by healthy adults as these tasks require skilled experience, as well as physical strength and endurance.   Returning to the historical context, the next important historical development that affected irrigation management customs - especially power relations between different user groups - occurred after 1846 when Spiti came under British rule. Under the traditional Tibetan (French, 1990), and Ladakhi (Pirie, 2005) legal system, the tax-payer Khangchen men served as tenant farmers. As noted earlier, under the British rule/policies, these tenant farmers became private landowners (Lyall, 1874, p. 181; Jahoda, 2008, pp. 10-11). This is related to the use of the term “Zamindar” in legal texts concerning the local irrigation system (discussed below). At this time, in 1867, Spiti’s traditional customs, including those of irrigation, were documented as part of a revised British revenue settlement. These documents have since served as local administrative papers and thus attained legal status. I will provide a translation and analysis of the relevant text concerning irrigation, known as “Riwaj-i-Abpashi”, “Riwaj-i-Abpashi” (the “Record of Irrigation Customs”),95 official copies of which are maintained by local revenue 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  95 Spiti’s irrigation customs were first codified in 1867 as part of a broader documentation project to prepare local administrative papers, known as Riwaj-i-Aam or the “Book of Customary Laws”, all of which were prepared during the same time (Lyall, 1874, pp. 234-235). As the name suggests, the Book of Customary Laws consists of all the main customs related to local livelihoods, including those related to the management of local resources (e.g., irrigation, farming, grazing, collection of medicinal plants, etc.) as well as social institutions (e.g., marriage, property ownership, religious institutions, etc.). These 	   111	  officers.96 Here I will briefly discuss how the process of codification impacted different sections of Spiti’s population, mainly in terms of gender and socio-economic (or class) background.   Baker (2003, p. 27), quoting Neeladri Battacharya, succinctly notes how the process concerning the documentation of customs in the Indian Himalayas had serious power implications for different sections of society: “[t]he ‘native voice,’ as it was filtered and interpreted during the codification and settlement process, was a ‘male, patriarchal voice, the voice of the dominant proprietary body speaking against the rights of non-proprietors, females and lower castes’.” This is evident in the language of the codified text discussed in Section 2. The power implications of the male patriarchical voice representing local customs were intensified by two other factors: 1) the use of legal terminologies that were borrowed from Indian contexts, and 2) the British policy of expanding agriculture (for tax collection) in India. In chapter 6, I discuss how these two factors rendered the men of Khangchen households more powerful while, at the same time, making women, Dhutul and Caste household members more disadvantaged at multiple levels. Here, due to space limitations, I will provide one brief example. Another example is apparent from the text of the codified irrigation custom translated below, which state that only the men of Khangchen households (called “Zamindars” in Urdu) are the holders of water rights or, practically, as owners of the irrigation system.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  administrative papers have since become the main legal basis for local resource management and social customs (Baker, 2003).	  96 There are several local officials, or Patwaris, in Spiti who are responsible for keeping land revenue records on behalf of different groups of villages. Zibug and Khyung villages belong to a single group and thus fall under one Patwari. This individual kindly shared and explained the technical terms used in the document. 	  	   112	   The most recent historical turn that had a significant impact on irrigation was the government of India’s land redistribution program, which followed the Nautor Land Rules Act of 1968. Since the passage of this law, many landless farmers, including many women, have received land. There are important ways that the Nautor land redistribution program impacted Spiti’s society that are worth noting here. First, by making landless farmers landowners, the traditional landowners lost their supply of cheap local labor. As a result of this development, today it is much harder for Khangchen householders to find contracted irrigators (known as chu pa) and plowman (thong pa), which are traditional instituted systems of farm labor.   Although the Nautor land redistribution program was aimed at benefiting the weakest sections of society, the people who most benefited from the program were the relatives or family members of the local officials responsible for granting Nautor lands, all of whom were educated men from Khangchen and Dhutul households (Fieldnotes, September 2010). Several Dhutul and caste farmers said that the first and the better and larger Nautor lands were distributed among the relatives and close friends of these officials. The poorest or the weakest sections of society, such as low caste (especially musician householders) were the last to receive land and their lands were also smaller and situated far from the village. In other words, although the granting of land benefited the weakest sections of the society, it benefited the more powerful ones more by granting them more and better lands.   	   113	  In sum, this historical discussion shows that the gendered nature of irrigation and farming labor relations is deeply rooted in history, and that it has been shaped under the influence of different kingdoms, states, empires and colonial interests. What is common among all these different kingdoms, states, empires and colonial powers is that they were all operating from a patriarchal mentality. Even the most liberating government’s (i.e., India’s) most generous and considerate programs (the Nautor land redistribution system), which were aimed at benefiting women, landless farmers and low caste members, resulted in compounding the traditional inequalities. I have also shown that the gendered nature of irrigation labor is also conditioned by environmental and geographic conditions. Furthermore, Khangchen men’s privileged rights regarding access to irrigation water have a historical and legal basis, for these men are (believed to be) the male descendants of the tax-paying households from the time of the Old Tibetan Empire.   Section 2: The codified customary laws of irrigation in Zibug and Khyung villages In this section, I will provide translations of Riwaj-i-Abpashi,97 along with an analysis of the text of the codified irrigation customs of Khyung and Zibug villages. A study of the texts is important for several reasons. First, these legal texts are the only documented record of local irrigation customs. The language of these texts provides insight not only into the historical contexts but also into details regarding how the farmers organize themselves in terms of sharing water, and in repairing and maintaining the irrigation systems. From an ethnographic perspective, it also provides insights into how the farmers gauge their time before the introduction of watches. Second, these texts are important 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  97 The translations were done with the assistance of the local Patwari, the land revenue officer, and Dr. Tsering Norbu, a local school teacher.	  	   114	  because these are de jure in nature, which has historical implications for power relations between farmers in terms of their gender and socio-economic status, as discussed earlier.  Third, as observed by Wacker (2008) in the case of Ladakhi irrigation records, “the practiced water rights of the villagers do not [always] correspond to the ones maintained in the official record” (p. 213). A comparison of the differences between codified customs and actual practice provides insight into historical changes that may have occurred in irrigation practices since the codification. It also sheds light on local politics as reflected in the internal agreements between farmers on how to deal with the changes.     The official document of codified irrigation customs of the two villages are handwritten in Hindi and are remarkably concise. The language contains many Urdu legal terms, the meanings of which the local revenue officer who maintains these records helped explain. In the translation below, I have attempted to stay as close to the meaning and style of the text as possible. The texts concerning irrigation customs for both the villages are provided under three sections: source channel, which identifies the main irrigation channel by its source name; watering turns, which describes how irrigation watering turns are divided among farmers; and channel repair, which describes how channel repair responsibilities are divided among the users. Below, I first provide translations of the three Riwaj-i-Abpashi documents (one for Zibug and two for Khyung 98) and then provide an analysis of the text which aims at shedding light on actual practice and issues of equity. In the following translation, only the names of villages, fields and people have been altered.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  98 Khyung has three groups of fields: Hara, Dangpo and Nyipa, located on different sides of the village. Each of these groups of fields has its own irrigation (rotation) systems, which are recorded separately but in two documents (one for Hara and one for Dangpo and Nyipa) in Riwaj-i-Abpashi. 	  	   115	   Document 1: Riwaj-i-Abpashi of Zibug village: Source channel: This channel was built by us Zamindars from the spring named upper Shajog.99 Watering turns: The watering turns is decided amongst the farmers. In one rotation, Zamindars get water for five days, who use it for irrigation on a turn-by-turn basis.  On the sixth day, the small households get water for one day, which is used by everyone together. After this, the turn starts again in the order mentioned above.  Repair of channels: When repair of a channel is needed, we Zamindars work together to do the repair. If anyone is absent, a fine is charged according to labor time and the amount is used for channel repair expenses.  Document 2: Riwaj-i-Abpashi of Khyung—Hara fields: Source channel: This irrigation channel was jointly constructed by the Zamindars from a stream named Rugolog.100  Watering turns: The water from this kuhl is shared between 17 large householders or Zamindars within themselves on a turn-by-turn basis from the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  99 The main source spring for Zibug’s irrigation waters (“Shajog” in the text) is also known as the home of the serpent spirits (klu khang). Water from this and other springs are collected in a communal reservoir, which is distributed among the farmers according to the described watering turn.	  100 The name of the stream, Rugolog (yur mgo log), is an interesting one. It literally translates as “Channel Head Collapse” and is also the name of the main village deity, which the farmers literally worshipped. According to local legend, which many believe to be true, a long time ago, Khyung farmers were unable to channel irrigation waters despite many attempts. The channels would either run dry or break apart. Frustrated with the situation, the farmers performed rituals for divine intervention. As a consequence, a powerful mountain demon (btsan) came from Ladakh in the form of a raven. The raven landed on a sacred site high above Khyung and poked its beak on the ground, from where a spring miraculously emerged. Today, the source spring of the Rugolog channel is considered sacred, and women of Khangchen households perform annual rituals here, just as Khangchen women of Zibug village perform rituals for the serpent spirits living at the source spring.	  	   116	  morning till 5 pm. On the first day, nine Zamindars get water. On the second day, another nine households get water. And this turn goes on in rotation from the third day. Since there are 17 households, one household gets water twice during the first round. On the second round, another household will get water for both days. In this way, all the 17 households get their turn for two-day waterings on a turn-by-turn basis. Small households get water after 5pm until it becomes dark. Water is shared among themselves on a turn-by-turn basis. After it gets dark, water is allowed to fill in the reservoir and the next day watering is done on a turn-by-turn basis. During the season, every small household gives water once to the big households. Repair of channels: When repair of a channel is needed, all the Zamindars work together to do the repair. If anyone is absent, a fine is charged equivalent to labor time and the amount is used for channel repair expenses.  Document 3: Riwaj-i-Abpashi Khyung—Dangpo (Source 1) and Nyipa (Source 2) fields:  Source channel 1: This irrigation channel was jointly constructed by us Zamindars from a spring [near] Ar. Watering turn 1: This group has 20 households, out of which nine are Zamindar households and eleven are small households. The nine Zamindar households get water on a turn-by-turn basis, in which three households divide water equally for each day. Thus a watering cycle is completed in three days. On the fourth day, another turn begins. The timing of water for the Zamindar household is [from 	   117	  morning] until shadow falls on Phalang [a pathway between the two mountains]. After that, until water is stopped for the next day’s collection in the reservoir, the small households use [leftover] water. Water is given to small households according to availability. This rotation system keeps going again and again.  However, from dawn until the [village] sheep and goats101 are sent to the “jungle”, water is given to Mrs. Tenzin Dekyi, wife of Dorje.  Twice a year, small householders get to use water [during the day] after the Zamindar rotation turn. And once or twice during the season, the small households give water to the Khangchen households even after the shadow [falls on Phalang].102  Channel repair 1: When the irrigation channel needs repair, one or two members from the Zamindar, as well as small households, go to perform the repair. If anyone is absent, a fine is charged according to a day’s labor wage. The fined amount is used for channel repair expenses and water is provided to the absentee. Source channel 2: We Zamindars jointly constructed this irrigation channel from a stream [near] Takmothang. Irrigation turn 2: same as channel 1. Channel pair 2: same as channel 1.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  101 Every morning, villagers sent their animals (cattle, sheep, goats and donkeys) out for grazing on the mountains under the care of two or three men (usually boys), who return with the animals in the evening. Thus the text above refers to the timing of these animals being taken/sent to the mountains, which is described more explicitly in the analysis below.	  102 The two days when small households get water during the daytime, mentioned in Riwaj-i-Abpashi of Dangpo and Nyipa fields, are the first two waterings (Yurma and Rhakti, discussed below). The first two waterings are critical for the crop’s healthy growth and these have to be done during the daytime.  The other clause, which says that once or twice a year the small households give water to the Khangchen households, is irrelevant because the practice does not exist.	  	   118	  Analysis of Document 1: codified irrigation customs of Zibug village The text uses the term “Zamindar” for the main rights holders of the village irrigation water. Zamindar is an Urdu word meaning “landowner”. However, it is important to point out that Zamindar is a masculine word, referring to a male landowner. The feminine or female equivalent of Zamindar is “Zamindarni”, which usually refers to the wife of the Zamindar. Therefore, the word Zamindar, as used in the text above, refers to the men of the Old Khangchen households. In the context of the Indian plains, from where the term is borrowed, Zamindars are aristocrats who held very large tracts of land (usually measured in square miles) that were leased out to peasants.103 In the case of Spiti, the Old Khangchen households were historically tenant farmers who had, on average, 3.5104 acres of land that were cultivated jointly by family members as well as by hired laborers. Thus the use of the term Zamindar to refer to Khangchen household men had three important consequences: it (a) changed their status from tenant105 farmers to landowners, (b) elevated their status, at least in title, to that of Indian aristocrats, and (c) established them as the legal rights holders to the village’s irrigation sources.   From my field observations and after consulting with farmers, I found that the “small households” mentioned in the text consist of a mixed bag of mostly Dhutul and caste households. These households have relatively small plots (<0.5 acre) of land near or 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  103 For a nice historical overview of the Zamindars, see “The Position of the Zamindars in the Mughal Empire” by Nurual Hasan, published in the Indian Economic and Social History Review, 1(4): 107-119.	  104 According to a retired teacher whose knowledge and accuracy of information is generally reliable (at least more so than the average Spiti informant), Spiti Zamindars have 15 to 20 bighas of land (5 bighas = 1 acre). 	  105 As mentioned above, the Khangchen households were tax-paying tenant farmers. The farming lands belonged to the state under the Ladakhi system. This changed under the British rule, which introduced a policy or law whereby farming lands were to be held independently by farmers (Lyall, 1874, p. 181; Jahoda, 2008, pp. 10-11).	  	   119	  around the main Khangchen fields.106 From a gender perspective, the patriarchal context of the text implies – according to local patrilineal customs of inheritance – that the men, not the women, of these “small households” hold the irrigation rights mentioned in the text. These points concerning local power dynamics, as they are reflected in the codified irrigation customs, also apply to Khyung’s irrigation system, as is discussed below.  The fact that the text identifies Khangchen men as the rights holders and lumps all other individuals and groups into the nebulous designation of  “small household” is reflective of the historical power structure of Spiti villages. Under the traditional power structure, which continued through the period of British rule, Khangchen men enjoyed all the rights and made all the decisions on behalf of the village. All other individuals and groups, including all the women and men of the Dhutul households, had no formal rights or decision-making authority over local resources.107 This was confirmed by several local farmers, including a Dhutul woman who claimed that traditionally all other individuals and groups “had no worth whatsoever” (brtsis rug lding se med kag).   In the description of watering turns, the text curiously does not mention the number of different households that share the water. It only mentions two groups of households: “Zamindars” (Khangchen households) and “small households”, thereby overshadowing 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  106 There are many ways that the small householders have come to own these small plots of land. According to a community leader from the village (Ngawang Samdup, personal communication, November 8, 2013), the majority were lands given to family members (younger brothers, sisters, aunts, etc.) of Khangchen households who had formed a separate home. Some households had established these rights through agreements that their ancestors had made with the village. In addition, a few were allowed to cultivate new fields under the Nautor Land Rules Act. Since these small household lands are not strictly regulated by the custom of primogeniture, the number of these small farming plots keeps increasing as lands get further divided between family members. 	  107 The formal decision-making mechanism for the village is the Village Council, which is comprised only of Khangchen men. After the 1968 Land Rules Act, however, many other households, including caste households, have begun to contribute full taxes to the village and thus have achieved representation in the Village Council.	  	   120	  the internal inequalities of irrigation access rights. In 2011, there were 13 Khangchen households and 60 small households. The 60 small households were comprised of 53 Dhutul, 4 Zo caste and 3 Beda caste households. While the 13 Khangchen households received water for five days out of the six-day cycle, the 60 small households received water for only one day out of the six-day cycle.108 Thus this system is highly inequitable in terms of the differences in the quantity and frequency of water available between Khangchen and small households.109   Although the text says that the repair work of the irrigation channels is jointly performed by the Zamindars (i.e., Khangchen men, but perhaps referring to the household here), this may not have been the case then and it is certainly not the case today. In actual practice, as I witnessed on one occasion, channel repair work (in Khyung) is done jointly by men, women, young children and Indian laborers representing every household (including “small households”) with water rights. Since my participant observation of channel repair work was of Khyung farmers, I will elaborate on this (immediately) below in the discussion concerning Khyung’s irrigation customs.110  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  108 The six-day watering cycle was divided as such: four Khangchen households get water on the first day; three Khangchen households on the second day; seven small households (belonging to Khangchen households) get water on the third day; three more Khangchen households share water on the fourth day; another three Khangchen households get water on the fifth day; and, on the sixth day, the sixty small households get water. This cycle is then repeated. Since the amount of water received by the four households sharing water for a day is relatively less compared to that of the other three Khangchen household groups, this group is compensated by three additional days of watering each year. These three days are: Mig chu, one day around the second watering, and on the fifteenth day of the sixth Tibetan month, which is also known as the fifteenth watering day (bcu lnga chu res).	  109 Because the system of primogeniture does not allow for the division of land, the number of Old Khangchen households remains relatively constant. However, the number of small householders seems to increase with time as non-Khangchen householders often tend to divide their land between their children. This renders the situation of inequity worse as years pass by. For example, the number of small households had increased from fifty during my first visit in 2007 to sixty in 2011.	  110 I was told by a Zibug farmer that the rules governing the social organization of channel repair work for Zibug are similar to those of Khyung.  	  	   121	   Analysis of Document 2: Riwaj-i-Abpashi of Khyung Hara fields  The text on the watering turns of the Hara fields says that 17 Zamindars have the main rights to the water. Again, the only other group mentioned is a nebulous “small household”. In actual practice, in 2011, there were 13 small households which received water after 5pm (see table below). These 13 small households are locally referred to as Chumed (chu med, literally “no water”). There is also another user group that is not mentioned in the text. This other group is called Tiping Langza (meaning, “those who raise irrigation tool”). In 2011, there were 30 of these households. The Tiping Langza farmers get water from the share of the 17 Khangchen households in return for certain contractual services that they provide, such as that of tilling and irrigating. Both Chumed and Tiping Langza groups are comprised of a mixed bag of Khangchung (i.e., along with the separate homes of parents of Khangchen),111 Dhutul and caste (both Zo blacksmiths and Beda musicians) households. Between Chumed and Tiping Langza, the former has better (permanent) irrigation rights and the latter has only temporary (contractual) rights.   In terms of equity, the distribution of water and power neatly follows an intersectional power hierarchy, with Khangchen men being at the top and exercising both de facto and de jure rights, and Beda caste women at the bottom, and having the least watering turns and no legal rights. In the table presented below, it is clear that access to irrigation water for different households is consistent with the local socio-economic or power structure, starting from the top to bottom: Khangchen/(Khangchung), Dhutul, Zo blacksmiths, and 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  111 When the eldest son gets married and assumes responsibility for the Khangchen household, the parents move into a smaller household known as a “Khangchung”. Many of these Khangchung households have their own small plots of land. Khangchung are technically a part of the Khangchen household. 	  	   122	  Beda musicians. Khangchen households have the most privileged access rights, followed by those of the Dhutul, followed by Zo caste members, and finally by the Beda caste members.  Number and type of irrigation users for Hara fields Number/type of households Irrigation rights 17 Khangchen 17 Khangchen  From morning to 5pm 13 Chumed small household 1 field shared by two Khangchen + 2 temple fields + 4 Dhutul + 5 Zo blacksmiths + 1 Beda musicians After 5pm (Leftover water or “chu ljug”) until the reservoirs are empty. Permanent right. 30 Tiping Langza small household  4 Khangchung + 11 Dhutul + 10 Zo blacksmiths + 5 Beda musicians Temporary right: water share from Khangchen households on contractual terms. Table 4: Composition of Hara irrigation users  Although women on the whole are a minority group (in terms of power), because gender operates in interaction with socio-economic background, women of higher socio-economic status enjoy higher prestige. This is partly because of the privileged irrigation water rights of their households, as compared to the women of the traditionally landless Dhutul and caste households. While one may argue that women of Khangchen 	   123	  households have more agricultural work responsibilities due to their larger fields (i.e., in comparison to poorer women farmers), one must bear in mind that the more mundane, difficult and time-consuming farming work is done by hired laborers, who are most often Dhutul or caste women. At the same time, from a socio-economic perspective, a larger field with more privileged irrigation access is more desirable than a smaller field with limited irrigation supply.  In terms of gender and caste intersection, as the chart above shows, there are six Beda irrigation rights, out of which one is permanent Chumed and five are temporary Tiping Langza rights. Beda women of Tiping Langza not only have to do irrigation work on their own temporary contractual fields but often also on the fields of Khangchen households, which are much larger in area. The point here is to highlight their underprivileged status, rights and responsibilities due to the social structural relations of power. As we shall see in the next section, caste women are also discriminated against or excluded in the performance of irrigation-related rituals because they are regarded as being impure. The text on the repair of channels, here again, speaks as if Khangchen men do all the work. As mentioned earlier, I had the chance to conduct participant observation of channel repair work with the farmers of Mara field in 2010. The de facto rule concerning who does irrigation channel repair is different from the official record. In reality, all Khangchen and small households that have a right to irrigation must send one representative to do the work. Although I found several members representing the Tiping Langza group, it seems to me that they came on behalf of their contractor Khangchen households. The people who came to work on irrigation repair were of all ages (from 	   124	  about 10 to about 75), gender and socio-economic background, and even included Indian workers.112   Analysis of Document 3: codified irrigation customs of Khyung’s Dangpo and Nyipa fields There are three notable things worth pointing out in the language of the codified irrigation customs of the Dangpo and Nyipa fields. One concerns a sentence (“Water is given to small households according to availability”) in the text that sheds light on the nature of the rights held by small householders in comparison to those held by Khangchen households. Second is a unique case of irrigation rights associated with an individual woman named Tenzin Dekyi. Third concerns the difference in the actual number of users mentioned in the text as compared to those in practice. I will discuss these three points one by one.   The irrigation customs for Dangpo and Nyipa follow the same unequal distribution pattern that we observed in Khyung’s Hara fields as well as in those of Zibug. The text says that Khangchen households get water from morning until the sun sets behind the Rhin-mo-kha-dang (srin mo kha gdang) mountain and its shadow falls on the Phalang area near the fields. After that, the text says that eleven small households get the leftover water. This pattern of small households getting “leftover” (chu ljug in local language) water after the Khangchen households are done watering is similar to that of the Khyung Hara fields. The text pertaining to the Dangpo fields has an additional sentence that raises 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  112 Among adult men, there were three Khangchen men, five Dhutul men, and one Zo blacksmith man. There were also nine adult women representing Khanchen, Dhutul and caste households. And there were thirteen young boys and girls, including six Indian boys. 	  	   125	  questions about the nature of the irrigation rights of small households. It says: “Water is given to small households according to availability.”  As it turns out, upon asking farmers what happens to the irrigation rights of small householders in times when there is a lack of water, it was discovered that whether the small householders get water during times of scarcity is dependant upon the good will and wishes of the Khangchen households. Khangchen householders are not required to leave a share of the (reservoir) water for the small householders. This is evident in the language of the codified customs. It says: “Water is given to small households according to availability” [emphasis mine]. By using words such as “given [by Khangchen men] … according to availability”, the text implies that the Khangchen households are not obliged to leave water for small householders in times of scarcity. This tells us two things. First, that Khangchen household men are the supreme decision makers regarding the use of irrigation water in times of scarcity. Second, and perhaps more importantly, it informs us that irrigation rotation based on the morning-evening “leftover” use system (used in Khyung’s fields) can be highly disadvantageous for the evening users or small households, who may not get any water at all, especially in extreme conditions of scarcity. A better alternative system of irrigation rotation for secondary rights holders in such a case is that of Zibug village, where small householders (secondary rights holders) have their own separate day of watering turns. Although Zibug’s example is not very appealing because of the high number of small householders, the system nevertheless protects these users from complete deprivation.  	   126	  The second thing worth noting in the text of Riwaj-i-Abpashi of Dangpo fields is the case of a unique water right associated with an individual woman named Tenzin Dekyi. The story behind Tenzin Dekyi’s unique water right is well known to the people of the village. It goes like this:113  Once two sisters went to release irrigation water to their fields. The sisters got into an argument over their share of water at a place where the main water channel divides into two channels: one flowing to the Hara fields and the other flowing into the Dangpo fields. The sisters argued that one was taking more water than her share. The argument escalated into a fight. In this fight, the sister from Dangpo fields, Tenzin Dekyi, was hit with a hoe by the sister from Hara fields and consequently died. Since Tenzin Dekyi had fought and lost her life over the water rights of Dangpo farmers, the nine Khangchen farmers of Dangpo fields thereafter decided to honor her sacrifice by giving the household to which she belonged a special water right. According to this decision, Tenzin Dekyi’s household was given the right to use the “overflow water” (zag ti) that flowed from the reservoir after it becomes full in the morning. The household can use this “overflow water” from dawn until the time when the sheep and goats of the village are sent to the mountains for grazing (in the morning). This water right was permanently given to Tenzin Dekyi, which meant that successive members of her household continue to have this right.  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  113 I heard this story in detail from two farmers of Dangpo fields: a Khangchen woman (Evi Tashi Yangzom, 2011) and a Dhutul man (Amchi Ngawang Tsering, field notes, 2010). The details described by them were essentially the same. 	  	   127	  Thus, ironically, the water right of Tenzin Dekyi was given after her death. The right actually belongs to the household where she lived at the time.114   The third notable thing in the codified irrigation customs of Khyung’s Dangpo and Nyipa fields is the issue of differences between codified customs and actual practice. Specifically, it provides a case where the actual number of Khangchen rights holders has decreased from what was recorded in the document. What is interesting is that, despite the decrease in the number of users, the number of watering turns remains the same as that mentioned in the text. In 2011, there were six Khangchen households in Nyipa (Source 2) fields but the watering system was done according to the nine Khangchen system recorded in the text. The number of households had decreased because three  households had ceased to exist. The fields and watering turns of two of the three deceased households now belong to their families – relatives who are the existing irrigation rights holders. As a result, two of the six Khangchen households had double turns. The share of the third deceased household does not belong to any single household but to the group as a whole, shared equally among the remaining six households. Thus the watering turn continues with the allotment of three shares per day over a three-day cycle, as written in the Riwaj-i-Abpashi. In the case of the Dangpo fields (Source 1), there are the same six Khangchen households as in the Nyipa fields, plus two households from the Hara fields who jointly share one field (half each, in Dangpo). Again, the watering turn remains the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  114 In addition, it should be pointed out from a gender perspective that since Spiti society is based on patrilineal inheritance, the irrigation rights of the household belong to the ‘head’ man of the household.	  	   128	  same as mentioned in the Riwaj-i-Abpashi: nine units divided into three watering turns per day in a three-day cycle.115   This example of the number of users changing while the watering system continues according to codified customs is similar to the reports of a case study of irrigation rights in the neighboring region of Ladakh done by Corrine Wacker (2008). In the Ladakhi case study (Wacker, 2008), the village of Tagmachig had an increase in the number of users but the allocation of water rights recorded in the legal documents were not altered and farmers had divided the water through an internal subdivision of the recorded rights, just as in the case of the Khyung’s Nyipa and Hara fields. This is done to ensure that overall water use remains the same and does not alter the water rights of existing users. The case of Nyipa and Dangpo provides some additional insights. When a household with water rights dies or ceases to exist, there are at least two possible ways this household’s water rights may be used: either the immediate relatives may take over the rights or the water rights may be shared by the remaining rights holders. This also suggests the unlikelihood of accepting new users (for example, a third party who is not related to the deceased rights holder).   Based on the textual analysis of codified customs, there are several take away points here. From an intersectional perspective, the powers and benefits (or the lack thereof) of irrigation rights of different categories of farmers, based on their socio-economic class, gender and caste backgrounds, neatly reflect the hierarchal social structure of the local 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  115 In this cycle, four households own one turn each (4 households x 1 field each = 4 fields/waterings), two households own two fields each (2 households x 2 fields = 4 fields/waterings), and another two households share one field (2 households x ½ field each = 1 field/waterings).	  	   129	  culture. The patriarchal social and legal contexts privilege men as rights holders and allocate women to a role as invisible workers. This is evident in the fact that nowhere in the text are women, or even anything to do with women, mentioned, or indeed even implied - even though women perform almost all of the irrigation-related work. The only exception is in reference to a special water right granted to a woman named Tenzin Dekyi, which in fact was given to her household (represented by men) as a result of her tragic death.   The above analysis of legal irrigation rights and actual practice also show, from an intersectional perspective, that men of the Khangchen households are privileged on multiple levels. While women in general are disadvantaged, we will see that in the realm of actual water management work (discussed in the next section), women of Dhutul and lower caste groups, are disadvantaged at multiple levels according to the broader socio-economic and gender power structures.   Section 3: Irrigation-related activities and roles in a calendar year What follows is a description of all the main irrigation-related tasks and rituals that are undertaken in the two case study villages. The discussion is presented chronologically from the beginning of the year to the end, as will also be the case in the next two chapters. This allows for the ethnographic descriptions to be chronologically overlaid, thus providing a more complete description of the farmer’s customary livelihood activities. My sources for the following section include participant observation and interviews. In particular, I had the privilege of conducting a detailed (more than 75 minutes) interview 	   130	  (on October 2, 2010) with the three irrigation group leaders (known as yur mgo pa, all three of whom were women older than fifty years) of Zibug village for the year 2010. The rituals and activities related to irrigation management in the two case study villages are similar. The types of activities, their purpose and meaning, as well as the basis of the division of the roles are the same. The only difference is timing: many of the activities are performed a few days apart due to various reasons. The following discussion of these two villages’ irrigation activities is therefore combined as one. After describing each irrigation-related activity, I will highlight issues of equity and micropolitics as they are related to the activity.  The focus of my analysis is on two things. As mentioned earlier, first, from an intersectional perspective, I analyze which categories of women perform which kinds of irrigation-related tasks. This is because these tasks, based on their desirability or prestige in relation to the ascribed roles of women belonging to different household types, can be seen as hierarchical. Second, I retain an empirical focus on issues such as who does (and who does not do) what (Ribot & Peluso, 2003, p. 154), as well as “when, where, and according to whom” (Emerson, Fretz and Shaw, 1995, p. 28). The objective here is to unravel the intersectional dynamics of gender, class and caste as they relate to irrigation management. In addition, I argue that attention to issues of who does not do what and why is critical to understanding how inequalities are cemented. 	   131	   Serpent spirit cake ritual (klu gtor) On the 25th day of the first Tibetan month,116 farmers offer ritual cakes to propitiate the local serpent spirits. Farmers believe that these serpent spirits (klu or “Naga” in Hindi) reside in the springs and have the ability to influence water flow and weather conditions (Interviews, October 2, 2010 & May 5, 2011). The belief in serpent spirits predates Buddhism and is found in other religious traditions of South and Southeast Asia, especially among the Hindus. These spirits are believed to possess features of both animals (e.g., vulnerable to injury and disease, as well as having a propensity toward aggression, etc.) and gods (e.g., possessing supernatural powers to control weather or to harm or benefit people). Thus farmers believe that these serpent spirits are sensitive beings, who can cause the springs to dry up or bring skin disease and other sicknesses to people who pollute the springs (Interview, October 2, 2010).117 Ritual texts are read and special ritual cakes are offered to appease the serpent spirits so that the spirits will keep the springs perennial and not harm the locals.  To perform the ritual, monks are invited to a home to read ritual scriptures and to make 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  116 In 2011, this day coincided with March 29 of the Gregorian calendar.	  117 Tibetan belief systems concerning serpent spirits are widely known (e.g., see Trungpa, 1978). To provide some insight into the nature of this belief system, here is a snippet of a description that was provided by a Spiti woman (Interview, October 2, 2010): “One should not wash or drink near these springs [inhabited by serpent spirits] because one might end up breathing or contacting the serpent spirit (klu yi kha rlang phog yong nga). Human contact also pollutes the serpent spirit, which gets diseased. Once the serpent spirits get diseased, the polluter will also get diseased. It is said that people with high spiritual merit (pa ra / bsod nams) cannot be harmed by the serpent spirit. However, serpent spirits wait up to a period of nine years and nine days (lo dgu chos dgu) until a time when the polluter’s spiritual merit goes down and then makes that person diseased. Once the polluter gets sick or diseased, he or she will remain uncured until the serpent spirits get healed from the disease. The only cure is through a special ritual – klu yi gdon drol – that must be undertaken to appease and heal the serpent spirit. This is an elaborate ritual that must be done carefully. If the person affected is very serious, then an even more elaborate ritual called sa thag gdon drol must be undertaken.”	  	   132	  ritual cakes.  During my field research, I observed that many of the households have their own little shrine dedicated to the serpent spirits. It was explained to me that certain households were particularly concerned with appeasing the serpent spirits, usually because the family members believed that the serpent spirits were responsible for giving them strange diseases. Any household can request that the monks perform this ritual upon their behalf. The ritual lasts one full day. After the ritual activities are completed at the home, the ritual cakes are taken to the spring and placed there as a symbolic gesture of offering to the serpent spirits that reside there. Since springs are the only sources of irrigation water in Zibug and Khyung,118 proper performance of this ritual is important to the farmers.   Belief in serpent spirits is an ancient, prehistoric phenomenon shared by people throughout the Himalayas, Tibetan Plateau and in many parts of South and Southeast Asia. Although it is beyond the scope of this chapter to thoroughly analyze the politics of this belief system, it is interesting to note that the belief in serpent spirits is advantageous to the Khangchen households since it leads to the protection of their water sources. One of the key features of this belief system is that believers are deeply fearful of disturbing the natural integrity of the springs as well as any other place where the serpent spirits may reside. This leads people to abide by norms that guide against polluting the springs, which, in practical terms, benefits the irrigation rights holders the most. One farmer observed that there is also a tendency for the poorer sections of the community be more fearful of these spirits as they are more prone to diseases due to poor diet and hygiene. If 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  118 According to local informants, all but two villages get their irrigation water from springs. The two villages that do not depend on spring water are Shichiling and Shego, which are low-lying villages that get their irrigation water directly from the Spiti river.  	  	   133	  this is true, the implication of the belief system is uneven from an intersectional perspective. Additionally, the fact that only men of the Khangchen or Dhutul households read ritual texts and make ritual cakes reflect biases embedded in this irrigation-related ritual in terms of: gender (women’s role in this sacred ritual is minimal to non-existent); socio-economic class (e.g., historically all monks hailed from Khangchen households, although today there are also many Dhutul monks - but still none from caste households); and caste (neither gender of caste background engages in the reading of texts or the making of the ritual cakes).  Drawing of irrigation channels in the field (rnang slang) The next irrigation-related activity I observed and that farmers confirmed (Interview, October 2, 2010) is the drawing of irrigation channels in the fields. This takes place after the fields are plowed and leveled, normally during the second week of April.119 The “drawing” (ri mo bris) of irrigation channels in the field is a technical task that is led by the most experienced woman in the household. She is normally the Khangchen woman or her mother-in-law. During the performance of this task, she, as the leader, is known as Bal go tsug khan (bal mgo btsug mkhan)120 and her main task is to figure out where the channel lines should be drawn. She takes a number of factors into consideration for drawing the lines. These include the slope and the size of the field, the location of sluices, and the direction of water flow. Once the Bal go tsug khan figures out where the Nang (rnang) channels should flow, she traces a line by dragging her foot in the field. As she 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  119 This is normally during the end of zhing ‘debs (“field cultivation”) month or the second Tibetan month.	  120 Bal mgo btsug mkhan means “the one who initiates ‘bal’”. The meaning of bal is unclear.	  	   134	  traces the line, she sings: “rnang la, rimo la”, which is an instruction to draw channels (rnang) where she is drawing the lines (ri mo).   The leader’s customary instruction is followed by men or women known as bal lan rgyab mkhan,121 who dig channels along the line drawn by the leader. In the case of the large fields owned by Khangchen households, it is normally the men from Dhutul or Zo blacksmith households who assist in digging the channels with a yak plow. For smaller fields, a group of three or four women, normally from Dhutul and caste households, dig the channels with hoes. The hired plowman is known as Thongpa (thong pa), and he gets a plot of land for farming in return for his services (discussed further in chapter 5). The women helpers form a work team called Bhe (byed) and they also get reciprocal benefits for their assistance, such as the use of a plow (which belongs to the Khangchen household) on their fields or the sharing of irrigation water.   The dirt dug up by the plow or by adze hoes to make rnang channels is used to make earthen dikes on both sides of the channels. The rnang channels are designed in curved parallel lines, like lines of latitude, and situated about six feet apart. The curved design is made to affect the speed of flow and allows for easier manual watering control. After making the rnang channels, a series of smaller ‘vertical’ earthen dikes are constructed to form a series of rectangular subsections called sli ‘ungs.122 A sli ‘ungs subsection is the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  121 A translation of bal lan rgyab mkhan is “the one who answers ‘bal’”.	  122 The purpose of watering in subsections is to achieve higher efficiency (less wastage) and effectiveness (better control) of water. This is achieved by allowing scarce water to flow slowly into a manageable patch of field. This also allows the farmer to provide better care for the crops, especially when they are young and delicate (discussed below). Work is done at a pace coordinated with the water flow in such a way that every area and corner of sli ‘ungs subsections are sufficiently watered and tended to, without any wastage.	  	   135	  smallest unit area of irrigation, which, on average, is around 8 feet wide by 6 feet long in size. Women helpers do all the work associated with the channels, including making the rnang channels and sli ‘ung subsections, using adze hoes. The making of sli ‘ungs subsections completes the task of drawing channels in the field.  Activities related to the drawing of irrigation channels show that Khangchen women command great authority in the field, especially for their expertise in laying the ‘blueprint’ of the field irrigation channels. Women helpers, normally from Dhutul and caste households, do the manual work of digging, lifting and pushing dirt to make the channels and dikes according to the blueprint laid out by the expert Khangchen women.   Yurshag Shichu (yur zhag bzhi bcu) or 40 days to irrigation After making the irrigation channels in the field, it is customary to wait forty days before watering the fields. This custom is known as yur zhag bzhi bcu (“forty days to watering”). The forty-day break is a traditional practice that people still follow for barley and other subsistence crops. For pea, which is a commercial crop, people simply study the size of saplings and irrigate the field if and when they decide it is needed.123 During these forty days, farmers perform two key activities related to irrigation: repairing the irrigation channels and collecting plants (mainly Festuca or rtsa skya) which are used for the first watering as mulch. 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  123 The doing away of traditional customs for pea cultivation has led people to refer to green peas as “halla matar”. The term for green peas is “hara matar” in Hindi (“hara” is green, and “matar” is pea) but some locals call it “halla matar” as “halla” refers to the “noise” or “hype” that has developed around the crop.	  	   136	   Irrigation channel repair Repair of the canals mainly involves clearing rocks, building canal walls, collecting large pieces of sod and placing them on the canal walls. The codified irrigation customs mention that the Khangchen men do the channel repair work. However, the following detail of irrigation repair work will show that the work is done jointly by all the rights holders, including men and women of all ages and social backgrounds.   The decision regarding if and when to do irrigation repair work is undertaken by the village council leaders, who are all men. These men instruct the yur mgo pa (the head of irrigation works, who is almost always a woman) about a day when one person from every household will gather to repair the irrigation channels and reservoirs. The yur mgo pa is responsible for mobilizing labor for annual channel maintenance and repair. This includes conveying the message to the village by shouting loudly around the village on the eve of the gathering day. The yur mgo pa is also responsible for taking roll calls and collecting fines. The position of the yur mgo pa is served turn by turn among households. Women usually hold this position but it could be done by anyone, including Indian laborers supported by the household. In 2011, all of these positions in Zibug and Khyung were held by women.  A group of representatives (which could be a man, woman or child) from each household that gets water from the channel gathers the next day with pickaxes, hoes and spades.  The yur mgo pa (or someone standing in on her behalf) takes a roll call. Often the group 	   137	  will divide themselves into subgroups so that the work can be completed sooner and more efficiently. If there are three groups, one group will work on the head canals, another group will work on the mid canals and the last will work on the tail canals. Two individuals are selected, usually an elder and a helper, who will go straight to a designated spot near the head source to prepare tea. The three groups work – cleaning and repairing the canal – in the direction of the water source. In the end, they all meet at the designated spot for another round of roll call and a meal break with a communal tea. During one of these breaks (which I had a chance to observe firsthand), people take a rest on the ground and freely form into groups to eat food brought from their homes. There is no rule on group formation. However, groups tend to naturally form among those who are closer and friendlier to each other. For example, I noticed that one of the groups was mostly young boys, one group was all Indian workers, one group had a mix of young men and women from Khangchen and Dhutul households, one group had mostly adult Khangchen and Dhutul men, one group was composed of mostly adult women of Khangchen and Dhutul households, and one last group, which was the most unique, was that of the caste members. Caste members sat together under one group, and consisted of members of all ages and gender. I think that this is because people tend to share food within the groups. As is customary throughout the Himalayan regions and in South Asia in general, members of the caste group in Spiti are not allowed to share food with non-caste members (Interview, 2010 & 2011; Rather, 1997).   After the break, the groups continue to complete their leftover work. The groups that have completed their share of work will join other groups. Work must be done jointly. 	   138	  After all the work is completed, a final roll call is taken and the group then disperses. Those who were absent, I was told, would be charged a fine of Rs. 50 and the money would be given to the village council leaders after subtracting the amount spent on the communal tea.   Collection of festuca (rtsa skya) Farmers in Khyung and Zibug collect local varieties of Festuca and Polygonum (snya lo) during the first watering of the season. The use of these plants plays a crucial role in the most important (i.e., the first) watering of the season (described below). Polygonum is collected in late autumn after threshing the crops when all the farmers go to the mountains to collect plants (discussed in chapter 6). In the spring, farmers allot two or three days specifically for the collection of Festuca.   Festuca is a coveted and scarce 124 plant, the collection of which is banned in Zibug and Khyung - except in an area that is commonly owned by the two villages and on the specific days (two or three days a year) that have been specifically reserved for their collection. The dates are decided at a village council meeting and announced at the same time so that all the villagers have equal access to the plant. On that day, men, women and children from all household types go to the mountains to collect the plant. After that, the mountains are sealed again and further Festuca collection is banned. If sekya growth is extremely sparse, villagers might impose a complete ban on the collection of the plant 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  124 According to local farmers, the plant became almost extinct in recent decades because farmers uprooted the plant indiscriminately. While the upper parts of Festuca that grows above ground are used during the first watering, its root was used to insulate bedding. Festuca roots were spread on the ground, on which farmers made their bedding for warmth. 	  	   139	  until it revives.  In years that the collection of Festuca is banned, farmers will use leftover barley and mustard hay.   The plants are chopped into small sizes (3 to 6 inches) and spread evenly over the fields a day or two before the first irrigation. This work is done by or under the close supervision of the main women of the household. The purpose of laying the chopped grasses in the field is to allow for an even flooding of the fields; this in turn allows for better watering control and for protection of the small shoots of crop that may otherwise be destroyed by sudden, forceful watering that comes from the abundant spring melt water. The grass also acts as mulch, conserving moisture in the soil and improving soil fertility at the same time. The need for close attention, expertise and skilled hands make this a task that is solely performed by the head woman of the (Khangchen) household (Interview, October 2, 2010). As Gutschow (1998, p. 464) noted, the main irrigation role of the Khangchen women continues into the first two or three waterings because these “should be done so precisely that even the wealthiest farmers do not entrust it to hired labor”.   Watering of fields There are five main watering days in a year, which have their own names: Yurma (yur ma), Rhagti (sreg ti), Sumti (gsum ti), Zhiti (bzhi ti) and Minchu (smin chu). Other waterings do not have names. These five days, including their specific method of watering, are explained below. The system is similar to the watering system of Zangskar (Gutschow, 1998). The technique for the first four waterings is called Shag nang (gshag rnang), which is done by methodically opening and closing the sluices of each sli ‘ungs 	   140	  subsection in the first two waterings and by sets of nang channels in the third and fourth waterings. This technique requires that the farmer step inside the field to carefully adjust the water flow and ensure that it is evenly distributed into every corner of the field. This is a sensitive process because the shoots are young and delicate, and the soil is not accustomed to watering.  Another technique of watering that is used for Minchu (seed ripening water), as well as the other six to eight nameless waterings of the year, simply allows the water to flow by itself throughout the field. This is accomplished by simply opening the main sluice of the field. For watering by this technique, known as circular watering (kyir chu yi), farmers simply prepare the sluices in such a way that water will flow to all parts of the field by itself once the water flow is allowed through the main sluice of the field. There is no need for the farmer to step into the field to control the force or the direction of the flow. By the time the circular watering technique is applied, the shoots are bigger and more able to withstand the stronger and more irregular flow of water. The hired irrigators can therefore apply this technique.   The first watering of the year: Yurma  Here, the name Yurma (yur ma) means “Mother Watering” (yur for watering, ma for mother).125 Farmers compare Yurma to the foundation of a house (yur ma khang pas rmang yin). Just as the quality of the foundation of a house determines the stability and quality of the structure that is built on top, how well Yurma irrigation is done determines how well the crop will grow on the field. This is expressed in a local saying: Yurma is 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  125 The word yur ma can also mean “weeding”.	  	   141	  like laying down the foundation of a house (yur ma khang pa’i rmang gting gting yin). According to Gutschow (1998), “If the first watering is done properly, then the subsequent watering for the rest of the season is easy. Conversely, if the first watering is poorly done, the field suffers the entire season” (p. 464).   The first watering is also more challenging because the soil is soft and dry, which makes water flow harder to guide or control. Additionally, owing to the abundance of fresh melt water, water flow is more forceful at this time, and the saplings are therefore more susceptible to damage. So a highly skilled and careful irrigator is required for the first watering to not only manage the unruly flow and protect the young saplings but also to ensure that the watering is done in such a way that future waterings will be easier to control and more beneficial for the crop. Therefore, because Yurma watering requires precise control and technique, only the most experienced women of the household are allowed to oversee this process. Hired irrigators, known as chu pa, are only allowed to water the fields after the first few waterings.   Every village undertakes the actual process of Yurma in a coordinated manner according to stringent customary regulations. This involves starting on a designated day, guided by the Tibetan Buddhist almanac, and performing a ritual for irrigation inauguration. Thereafter the waterings are done on specific fields according to custom and decisions made by the village council, all of which are explained below. 	   142	   Kirzin (skar ‘dzin), the irrigation inauguration ritual and the exclusion of caste members The day and manner of inaugurating Yurma is determined according to the prescriptions of the Tibetan Buddhist almanac (lo tho), as is the case with the start of all the major agricultural activities (on this, see the following two chapters). A local elder, normally an Amchi (doctor, who is almost always a male),126 consults the calendar to determine an astrologically auspicious day. By consulting the almanac, he also finds out what kind of person should perform the ritual for inaugurating irrigation for that particular year. The ritual is known as Kirzin, which literally means “star catching” or “making sure of a propitious constellation” of stars127 by ensuring that the astrological sign (lo rtag) and element (khams) of the person who performs the ritual are the same as those prescribed by the almanac as the most auspicious for undertaking the ritual. Farmers also believe that a person who has both parents alive (pha ma mal ldan) should perform the ritual. This is because people having both their parents alive are regarded as fortunate and thus auspicious. If the gender of the person is not specified, preference is given for a female or a girl.128 If a suitable girl were not found, a boy meeting the criteria is chosen to perform the ritual. The ritual begins by reading prayers on behalf of the village. Then a person, usually a girl having astrological signs deemed particularly auspicious for the occasion, 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  126 Traditionally all of the village doctors are men from Khangchen households. Lately, some Dhutul men have also become doctors. And in Khyung, there is a unique case (the only one in Spiti that I know) of a female doctor. While Khangchen male doctors traditionally pass on their knowledge through a lineage system, from father to eldest son, the Dhutul and women doctors have acquired their training outside Spiti, mainly at an institute started by the late Amchi Sundar Singh in Manalai and at the Tibetan Medical and Astrological Institute in Dharamsala.	  127 According to Nitartha online Tibetan dictionary, skar ’dzin is “star catching, making sure of a propitious constellation”. Retrieved from http://www.nitartha.org/dictionary_search04.html.  	  128 Farmers I interviewed were unable to explain the rationale behind the preference for a female in the performance of this ritual. Perhaps it is for convenience since women do most of the work. It could also be due to certain belief systems related to the association of feminine qualities with water (Choezin, 2003). 	  	   143	  opens three sluices (awad) to allow water flow to the crops. When water flows through the sluices, the girl is taught to say certain prayers calling for a good irrigation season, a bountiful crop, and for good health. After that, the ritual is completed and the actual task of watering begins.  During one of the interviews, when an elderly woman was describing this ritual, I asked if the person or the girl considered auspicious for the performance of the irrigation ritual could be someone from a caste household. She did not answer with a ‘yes’ or ‘no’ but simply said that she had only seen non-caste members perform the ritual. I later posed the same question to several other farmers, including my research informants. Most of the farmers I asked gave answers similar to the elderly woman. Many simply said that they did not know. Only two of my informants, both educated men and informed about the intersectional focus of my research, concluded, after reflecting on the question and discussing possible scenarios, that non-caste (Chechang) farmers do not allow caste members to perform the ritual and that the main reason for this discrimination is that caste members are considered dirty or ritually impure. A scenario that helped them come to this conclusion is the following: what would farmers do if there is only one girl meeting the astrological prescriptions of the almanac and if she is from a lower caste? Would the farmers let the low caste girl perform the ritual or would they look for a Chechang boy who meets the criteria? When presented with such a scenario, two of my local informants, who can be considered as more modern and progressive individuals among the Spiti population, said that they thought that farmers would choose a Chechang boy who meets the criteria. Their answer is remarkable and strongly suggests that the 	   144	  non-participation of caste women in performance of the Kirzin ritual is a discrimination based on local notions of purity.129   Here, I draw on Nightingale’s (2011, p. 153) calls for examining “ways in which boundaries between bodies, spaces, ecologies and symbolic meanings of difference are produced and maintained relationally through practices of work and ritual.” An important aspect of this exclusionary discrimination in the performance of Kirzin irrigation inaugural ritual is that its practice, due to unquestioned customary practices, seems covert in nature. Since there is nothing to gain from being the chosen person for the ritual, no one really cares who performs the ritual as long as the astrological prescriptions of the almanac are fulfilled. People’s reluctance, or their apparent lack of a definitive answer to the question as to whether caste members can or cannot perform the ritual, shows that this discrimination is buried under the banality of customary practice. This study argues that even though the practice of this discrimination does not seem overt - i.e., it is not done (physically and verbally) in front of them - it is still arguably more powerful than the normal day-to-day aspects of discrimination, such as those that do not allow caste members to sit or eat with other social groups. It is in fact the very covert or overlooked nature of this discrimination, especially in the spatial context of sacred rituals related to scarce resources and important livelihood activities, that makes it powerful and deserving of attention. For it is these kinds of unquestioned practices of discrimination, particularly those in spatial contexts of sacred rituals, that rationalizes the overt everyday practices of discrimination. Here, irrigation water is a scarce and critical resource required for the 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  129 It must also be pointed out that another marker of social difference in Tibetan society, other than caste, where a group is generally considered to be ritually impure, is female-gender (e.g., Gyatso, 1987; Ortner, 1973; Huber, 1994).	  	   145	  survival of farmers living in the arid Himalayas. Water also has symbolic religious meanings associated with cleanliness and purification (e.g. it is used for purificatory rituals called khrud). Thus the ritual of irrigation inauguration is a highly sacred ritual, one that has deep spiritual meaning for the farmers. Martin Mills (2003), for example, calls this kind of ritual performed by farmers an expression of the “chthonic productive consciousness of the peasant”, which, he argues, “is truly hegemonic in Tibetan societies” (p. 345). Therefore, there are powerful and hidden symbolic and ritual meanings –  as well as socio-political implications – behind the exclusion of a certain group of people from certain spatial and ritual contexts. The spiritual significance of the skar ‘dzin ritual and the “chthonic productive consciousness” of the farmers are also evident in the fact that farmers emphatically mentioned that the ritual prayers are immensely important to them (Interview, October 2, 2010).   The order of performing Yurma in the village After the ritual opening of the three sluices, farmers begin the first watering of the year, which is done in a customary order and takes a few days to complete. In 2011, the fields of the village doctor (Amchi) and deity medium (lus g.yer) were irrigated on the first day in both the villages. On the second day, the fields of those households where someone is sick were irrigated. In addition, in Khyung Village, the field of another Amchi was irrigated on the second day. This Amchi hails from a Dhutul household and hence does not belong to a traditional Amchi (of descent or gdung rgyud) household. He is given the privilege of irrigation water on the second day in honor of his medical services to the village.  	   146	   The third day of Yurma was done differently in the two villages. In Khyung, the third day is reserved for the fields of small households in both Mara and Nyipa-Dangpo fields. 130 In Zibug, the third day of Yurma watering is done on the fields of the rest of the villagers, with the turns being based on a throw of dice. However, some of the fields may have a more urgent need for watering. Owners of these fields may request permission (dgongs pa) for urgent access to irrigation water from their irrigation group leader (yur mgo pa). The requests are usually accepted with a charge of Rs. 60 (2011 rate) per household. Those who have dgongs pa will be given water first in an order determined by a throw of dice. After that the remaining fields get water, with the turns likewise being decided by a throw of a dice.   The instituted processes of performing Yurma in the first three days show how the social (fields of sick farmers, local doctors, and those of the village deity) and the economic (farming activity) dimensions of the culture are embedded in each other. This is an argument that I will develop further in chapter 5. It is important to note that adult women do all the work on Yurma, as well as on the following waterings. This partly explains why the division of labor within the household for watering and for the collection of firewood is based on gender and not age. Both activities require adult workers.   	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  130 This custom of giving Yurma watering to small householders in Khyung is also codified in Riwaj-i-Abpashi, which mentions that “Twice a year, small householders get to use water [during the day] after Zamindar rotation turn.” One of these two days is during Yurma and the other is during Rhagti. 	  	   147	  The second watering: Rhagti In Spiti, the second watering is called Rhagti (sreg ti), referring to the condition of the saplings as “sun burnt” (nyima la sreg pa) and needing water. Farmers believe that the second watering must be done after an extended dry gap of 10-15 days, which creates an ideal condition for the crop to grow. A local proverb guides the timing of the second and third waterings: sreg ti la ki ya, gsum ti la stud (“Extend the duration for the second watering and shorten it for the third”).   Unlike Yurma, where the turn-by-turn order of watering customarily starts from the fields of the village doctor, deity medium, and other specific households (as described above), the water order for Rhagti starts from the highest altitude fields and on those fields where both laborers and crop conditions are ready. In other words, there is no specified order for Rhagti. The order is dependent on the conditions of the crops, farmers’ preparedness for irrigation, and altitude. Those that feel that their crops need watering earlier do so earlier, starting with those fields which are located on higher altitudes. If there is a serious shortage of water or if more farmers feel that they need water than what can be supplied at that time, a draw of lots is undertaken with the use of dice and an order is established. Rhagti is thus completed in three to four days. Thus in terms of equity, Rhagti does not exacerbate the general aspects of inequality characteristic of the irrigation system, such as women doing all the watering and Khangchen households having greater irrigation turns as compared with Dhutul and caste households.  	   148	  The water source ritual: Chutsa (chu rtseg) Once a year, women from Khangchen households perform an important ritual related to irrigation source control. There is no fixed day for this ritual as it may be done on any convenient day after the second watering. On this day, only Khangchen household women gather at the water source spring in order to propitiate the serpent spirits and to ensure that the spring rivulets flow into the main channel or the reservoir. The women may also give offerings of three white dishes and three sweet dishes (such as curd, butter, milk, sugar, honey, etc.) to the serpent spirits, which is a general practice in these rituals. An elderly Khangchen woman related that, in the past, the Khangchen women used to wear traditional attire, drink chang (local beer), sing songs and dance near the spring. It is not clear if the ritual singing and dancing was aimed at the appeasement of the serpent spirits or for their own entertainment. That they wore special attire and engaged in drinking and dancing tells that it had at least symbolic, if not ritual, significance.   The exclusion of Dhutul and caste women is not based on notions of purity because Dhutul and Khangchen women are ritually equal. Therefore, the performance of this unique ritual being the sole domain of Khangchen women marks their distinction as the highest socio-economic class of women. It symbolizes the power of the Khangchen women. For one, just as in the case of the first watering of Yurma, it shows that the Khangchen women do not trust any other women with the maintenance work of the irrigation sources. Second, it is also an occasion when Khangchen women discuss and decide on issues surrounding the management and use of the irrigation system. For example, during one of these meetings, held in recent years, the Khangchen women of 	   149	  Zibug village had decided to do away with the custom of letting irrigation flow freely for use on certain days (Ngawang Samdup, personal communication, November 8, 2013). According to the custom, on certain days (e.g., when there is rain or on occasions marking the arrival of high religious figures to Spiti or to the village), irrigation water is allowed to flow freely and anybody can use the water. The water is allowed to flow freely because the people are busy attending (“making tea” for) the religious figure, or because there is an over abundance of water due to the rain. Today, however, Zibug farmers do not practice this custom: water in the irrigation channels is never free for use by non-rights holders.   The third and fourth watering: Sumti and Zhiti Sumti (gsum ti) refers to the third watering of the year and Zhiti (bzhi ti) to the fourth watering. The order of watering for Sumti and Zhiti is the same as that of Rhagti. The method of Sumti is that this time, three sets of sli ‘ungs subsections are flooded together from the bottom of the field to the top. Similarly, for Zhiti, four sets of sli ‘ungs subsections are flooded, in the same order, from the bottom of the field to the top. The logic of gradually increasing the number of flooded subsections is that, with each subsequent watering, the saplings are now more able to withstand flooding.  In terms of equity implications, during this time, many Khangchen women pass over their irrigation responsibilities to contracted irrigators, who are usually Dhutul or caste women. Once the irrigator takes charge, they perform all the subsequent waterings - to the time of 	   150	  the final “germinating water”. In other words, these irrigators perform all the “chu res” or watering turns for the Khangchen households.  Determination of watering turns (chu res) and regular watering After the fourth watering, farmers determine irrigation turns for regular watering by throwing dice. This establishes the order of watering between households and will be done according to customs written in the Riwaj-i-Abpashi. The regular waterings (i.e., the watering after Zhiti and before the final watering, known as Minchu), do not have names. Hired irrigators (chu pa) belonging to Dhutul or caste households are mostly responsible for taking care of these waterings. In return for their contractual services, chu pa get a small plot of field and a share in the irrigation water. The chu pa’s contract is customarily valid for a year and is often renewed along with the hired tiller (thong pa, discussed in chapter 5).  In terms of equity, the caste and Dhutul women begin to take over their contractual irrigator (chu pa) responsibilities and perform the remaining waterings of the year, although some Khangchen women do all their own irrigation work themselves. Here the workload for Dhutul and caste women, who work as irrigators on behalf of the Khangchen households, must be noted, especially in terms of time. Watering responsibilities of irrigators include, in addition to watering the large fields of the Khangchen households, the opening and closing of the main reservoirs, some of which are situated quite far away and high up in the mountain. Irrigators must also open, close or adjust the different sluice gates of the secondary and other sub-channels to ensure that 	   151	  the right amount of water is flowing to the different channels and fields. The actual watering work on the fields is also time consuming, with the women having to manually channel the water with an adze hor or Tirping to different corners of the fields. All this takes several hours, often a whole day. Then in the evening, they must water their own fields in the same way. Dhutul and caste irrigators must regularly perform this routine, first watering all the Khangchen fields, along with performing the related watering responsibilities, and then watering their own field in the evenings. This routine is followed until the end of the last watering, known as the “germinating water”.  The germinating water: Minchu (smin chu) The final watering is called Minchu (smin = “to ripen”, chu = “water”) because farmers use it to help ripen the seeds. The final watering of the year is done several weeks before the harvesting.  The method of Minchu is different from that of Yurma, Rhagti and Sumti for this final watering does not require controlled precision and care. Water is allowed into the field in greater volume and velocity. Since the crops are fully-grown and sturdy, they are able to withstand such a flow. Locals also believe that the sudden surge of water assists in the quicker ripening of the seeds. Thus, there is less amount of work and time required for this watering. Farmers typically wait until at least a week after Minchu before harvesting. This provides sufficient time for the soil to become dry and ready for harvest.131 	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  131 This process is just the opposite of the system employed in Zanskar as described by Gutschow: “The final watering just before harvest is called ‘bru chu. This watering serves to loosen the soil and the roots, making it easier to pull out the entire plant as is customary in Zanskari harvests.” In Zibug and Khyung, it is customary to leave the roots. 	  	   152	   Conclusion Evidence from these case studies points to the need to go beyond general approaches to gender and equity in irrigation management. In the current literature on irrigation management, equity is generally theorized along an enquiry as to whether the fields get water with equal frequency and whether the allocation of water is proportional to the size of the fields (Trawick, 2001, 2003). This study shows that these two features are not sufficient to assess whether an irrigation system is truly equitable as these case studies satisfy these features but also have, for example, a small minority of farmers owning the majority of the fields and the water rights, while a vast majority of farmers have only negligible rights. In addition, these case studies highlight unequal decision-making powers and division of irrigation management roles in terms of gender and markers of social differences such as socio-economic class and gender. Moreover, in terms of theoretical approaches to gender and water, past approaches have mostly focused on “men versus women” issues as they are related to power relations in water use, access and management. While it is widely acknowledged that sex-gender is only one of the many intersecting determinants of social difference, any in-depth analysis of the multiple intersectionalities relating to water and gender issues is rare (Harris, 2008). These case studies from the Himalayan region of the Spiti Valley raise questions about general notions of irrigation management roles as they are dominated by men. The case studies show that in certain historical and geographical contexts, it is possible to find cases where women do most of the irrigation-related tasks. Although the discussion in this chapter does not shed light on why irrigation is the exclusive domain of women in Spiti, it does 	   153	  highlight a need for a “within-different-groups-of-women” approach to intersectional studies of gender and equity in irrigation management.  These intersectional case studies of equity and gender in irrigation management have provided a detailed analysis of the relevant history and legal documents, as well as the farmers’ actual practices. There are several important take away points in each of these sections. The analysis shows that Spiti’s unique case of water management - as the exclusive domain of women - can be understood as being the result of specific historical and geographical factors. It also thus shows that the gendered nature of irrigation management is more complex than a straightforward power interplay between men and women. Another important lesson gleaned from the case study is that while the codification of customary laws led to the protection of traditional ways of life, it also helped maintain, and indeed even intensified, internal inequalities associated with gender, caste and socio-economic class. Additionally, while government projects of poverty upliftment, particularly through the granting of Nautor lands to landless farmers, led to overall improvements in poor people’s livelihoods, the benefits of the Nautor land distributions were better received by the dominant social groups (i.e., men of Khangchen and Dhutul households) than by the underprivileged groups (especially by women and caste household members).  Finally, this study shows that, in addition to the four key irrigation activities (source control, delivery, use and drainage) identified by Kelly (1983), it is important to include an analysis of the religious ritual and social institutional dimensions of irrigation 	   154	  managemen