Origins and Migrations in the Extended Eastern Himalayas Brill’S Tibetan Studies Library

Origins and Migrations in the Extended Eastern Himalayas Brill’s Tibetan Studies Library

Edited by Henk Blezer Alex McKay Charles Ramble

VOLUME 16/4

The titles published in this series are listed at brill.nl/btsl Origins and Migrations in the Extended Eastern Himalayas

Edited by Toni Huber and Stuart Blackburn

LEIDEN • BOSTON 2012 Cover illustration: Tadé Mihu, an Idu Mishmi shaman (igu), leading the soul of a deceased person to the place of origins during a brohfee ritual. Anini, Arunachal Pradesh (photograph by Toni Huber, 2007)

Library of Congress Cataloging-in-Publication Data

Origins and migrations in the extended eastern Himalayas / edited by Toni Huber and Stuart Blackburn. p. cm. -- (Brill’s Tibetan studies library, ISSN 1568-6183 ; v. 16/4) Includes bibliographical references and index. ISBN 978-90-04-22691-3 (hbk. : acid-free paper) -- ISBN 978-90-04-22836-8 (e-book) 1. Tibeto-Burman peoples--Himalaya Mountains--Origin. 2. Tibeto-Burman peoples-- Migrations--History. 3. Tibeto-Burman peoples--Himalaya Mountains--History. 4. Tibeto-Burman peoples--Himalaya Mountains--Social life and customs. 5. Mountain people--Himalaya Mountains. 6. Himalaya Mountains--Emigration and immigration--History. 7. Himalaya Mountains--Social life and customs. 8. Himalaya Mountains--Ethnic relations. 9. Ethnology--Himalaya Mountains. I. Huber, Toni, 1956- II. Blackburn, Stuart H.

DS25.5.O75 2012 954.96--dc23

2012000054

ISSN 1568-6183 ISBN 978 90 04 22691 3 (hardback) ISBN 978 90 04 22836 8 (e-book)

Copyright 2012 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing, IDC Publishers, Martinus Nijhofff Publishers and VSP.

All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher.

Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change.

This book is printed on acid-free paper. CONTENTS

List of Illustrations ...... vii

Introduction Toni Huber and Stuart Blackburn ...... 1

Trans-Himalayan Migrations as Processes, Not Events: Towards a Theoretical Framework Geoff Childs ...... 11

Where the Waters Dry Up – The Place of Origin in Rai Myth and Ritual Martin Gaenszle ...... 33

Where did the Question ‘Where did My Tribe Come From?’ Come From? Robbins Burling ...... 49

Coevolving with the Landscape? Migration Narratives and the Environmental History of the Nyishi Tribe in Upland Arunachal Pradesh Alexander Aisher ...... 63

Micro-Migrations of Hill Peoples in Northern Arunachal Pradesh: Rethinking Methodologies and Claims of Origins in Tibet Toni Huber ...... 83

Apatani Ideas and Idioms of Origins Stuart Blackburn ...... 107

0LJUDWLRQ1DUUDWLYHV2I¿FLDO&ODVVL¿FDWLRQVDQG/RFDO Identities: The Memba of the Hidden Land of Pachakshiri Kerstin Grothmann ...... 125 vi CONTENTS

The Language, Culture, Environment and Origins of Proto-Tani Speakers: What is Knowable, and What is Not (Yet) Mark W. Post ...... 153

Glimpses of the Ethnolinguistic Prehistory of Northeastern India George van Driem ...... 187

Origin and Migration Myths in the Rhetoric of Naga Independence and Collective Identity Marion Wettstein ...... 213

Oral Histories and the ‘Origins’ of Current Peoples: Dynamic Ethnogenesis, with Remarks upon the Limitations of Language-Family Subgrouping F. K. L. Chit Hlaing ...... 239

Cords and Connections: Ritual and Spatial Integration in the Jinghpaw Cultural Zone Mandy Sadan ...... 253

Origin and Return: Genesis and the Souls of the Dead in Naxi Myth and Ritual Charles F. McKhann ...... 275

Migrating Brothers and Party-State Discourses on Ethnic Origin in Southwest China Koen Wellens ...... 299

Contributors ...... 321

Index...... 325 LIST OF ILLUSTRATIONS

Figure 1.1 Map of contributor case study locations ...... 3 Figure 2.1 View south on the Koshi River from Chintang village in Dhankuta district ...... 35 Figure 2.2 Map of the Ma:mangme journey ...... 38 Figure 2.3 The Kiranti drainage model of genealogy ...... 40 Figure 5.1 Map of the northern Subansiri region ...... 86 Figure 5.2 Partly Tibetanized Na living near Taksing, northern Subansiri...... 100 Figure 5.3 Bokar man wearing Tibetan wool tunic and nalong earrings, Monigong, Siyom ...... 101 Figure 7.1 Map of sites related to the Memba in the eastern Himalaya ...... 126 Figure 7.2 Mechukha valley view north ...... 127 Figure 7.3 Map of Pachakshiri and adjacent Tibetan sites ...... 131 Figure 7.4 View of Lhalung ...... 134 Figure 7.5 Picnic atmosphere during pilgrimage ...... 141 Figure 7.6 Map of the Mechukha valley ...... 142 Figure 7.7 Pemajeling with meditation caves to the right ...... 145 Figure 8.1 Map of the cultural-geographical context of the Tani languages today (approximate) ...... 154 Figure 8.2 Riyu (Minyong) village, East Siang District ...... 155 Figure 8.3 Galo ѪLEy (ritual specialist) and bòo (assistant) atop VDFUL¿FLDODOWDU'DO଎ғܺYLOODJH ...... 155 Figure 8.4 *DORHOGHU1\DDGRR5ܺEDD'DDU଎ғܺYLOODJH ...... 156 Figure 8.5 *DOREULGH'DDU଎ғܺYLOODJH ...... 156 Figure 8.6 0LODQJHOGHU$DPܺQ0RRGԥ0LODQJYLOODJH ...... 156 Figure 8.7 Provisional Tani family tree ...... 160 Figure 8.8 0RGHUQUHÀH[HVRI3URWR7DQL ƾRғµ¿VK¶ ...... 161 Figure 8.9 Cross-branch innovations, contact and the genetic position of Galo ...... 162 Figure 8.10 Taxonomic structure in the Tani lexicon (Pugo dialect of Galo) ...... 167 Figure 8.11 Representative set of terms employing topographical deixis in Lare Galo ...... 167 Figure 8.12 Environmental source of topographical deixis ...... 168 Figure 8.13 Well-assimilated Indic loanwords in Lare Galo (Asm = Assamese) ...... 173 Figure 8.14 Assamese loanwords in Karbi-Anglong Mising ...... 174 viii LIST OF ILLUSTRATIONS

Figure 8.15 Development of Galo modal of necessity via language-internal means following borrowing of an Indic lexical verb meaning ‘want/need’ ...... 177 Figure 8.16 Rhythm and the opposite typological drifts of Munda and Mon-Khmer ...... 179 Figure 8.17 Rhythm and typological drift in Tibeto-Burman ...... 179 Figure 8.18 Selection of Tani forms with scarce or unattested PTB cognates ...... 180 Figure 8.19 Milang forms with scarce or non-occurring TB/Tani cognates ...... 180 Figure 9.1 The fallen leaves diagram for Tibeto-Burman ...... 188 Figure 9.2 Geographical distribution of the major branches of Tibeto-Burman ...... 189 Figure 9.3 $XVWURDVLDWLFZLWK*pUDUG'LIÀRWK¶VWHQWDWLYH calibration of time depths for the various branches of the language family ...... 190 Figure 9.4 The portion of the Y chromosome phylogenetic tree .....200 Figure 10.1 Longterok/Ongterok stones near the village of Chungliyimti ...... 225 Figure 10.2 One of the trees near Makhel said to be the origin of the southern Naga groups ...... 228 Figure 10.3 Detail of a tree near Makhel said to be the origin of the southern Naga groups ...... 229 Figure 10.4 J.H. Hutton’s Migrations of Naga Tribes map ...... 230 Figure 13.1 Naxi Gods’ Road Map, showing demon pole (foreground) and god pole (back) ...... 282 Figure 13.2 Tibetan Wheel of Existence ...... 283 Figure 13.3 Demon pole of the Gods’ Road Map ...... 285 Figure 13.4 Human realm on the Gods’ Road Map ...... 286 Figure 13.5 Xiuqiu (the garuda) in the Wish-granting Tree (Haiqyi-bbaddaq-zzer) devouring Shv (nagas) ...... 287 Figure 13.6 God Pole of the Gods’ Road Map ...... 288 Figure 13.7 Map of Naxi and Moso ancestral roads ...... 289 Figure 13.8 Bundled corpse on heath platform, with earthly possessions, offerings of food and wine, and Gods’ Road (hempen cloth strip) leading out through the ceiling to rooftop altar ...... 293 Figure 13.9 Gods’ Road depicted in rooftop altar ...... 294 Figure 13.10 &RUSVHLQFRI¿QRQWHUUDFHZLWK*RGV¶5RDG ...... 295 Figure 13.11 Wish-granting Tree at village cremation ground ...... 296 Figure 14.1 &KDUWRIWKHGLIIHUHQW;ƯIƗQDXWRDQGH[RQ\PV ...... 305 Figure 14.2 Map showing Qiangic Languages ...... 308 INTRODUCTION

TONI HUBER AND STUART BLACKBURN

7KLV ERRN LV WKH ¿UVW WR EULQJ WRJHWKHU FRQWHPSRUDU\ UHVHDUFK RQ Tibeto-Burman-speaking hill peoples in a region stretching from eastern Nepal through Arunachal Pradesh, Nagaland and the hill tracts surrounding Assam, to upland Southeast Asia and southwest China. This region of investigation we provisionally refer to as the extended eastern Himalayas. The volume also represents the only multi-disciplinary study of the closely related topics of origins and migration in this greater region of Asia to date, bringing together current research by anthropologists, folklorists, linguists and historians. The book is one result of our joint research project “Historicising the Production and Transmission of Origin and Migration Narratives in Arunachal Pradesh, Northeast India”, undertaken by the editors at the Humboldt University of Berlin during 2007. This culminated in the international conference “Origins and Migrations Among Tibeto-Burman-Speakers of the Extended Eastern Himalaya” (Humboldt University, 23-25 May, 2008), at which versions of the chapters contained here where presented. Like the conference out of which it emerged, the present collection of essays might best be considered exploratory in terms of both its regional IRFXVDQGLWVDSSURDFKWRVSHFL¿FWRSLFV2XUSULPDU\SXUSRVHLVWRLQLWL- ate a dialogue between scholars concerned with the region but whose work is conducted through different disciplinary approaches. Thus, to paraphrase our contributor Mark Post, a basic goal of the volume is to begin to understand what kinds of data, hypotheses and theories we are currently able to offer one another, and/or may eventually be capable of offering one another in the future when considering questions about origins and migration. The result is a diverse set of investigations into theoretical and methodological concerns, myth and ritual, narrative in social context, language and linguistic relationships, identity formation, and local communities in relation to modern states. We chose to focus on origins and migration because they struck us as unusually prominent throughout the region and because they encompass a wide range of ideas and practices. Considering origins and migration raises questions that have both bedevilled scholars for centuries and continue to spark controversy among local populations and states 2 TONI HUBER AND STUART BLACKBURN today. Such questions generated the numerous and often competing perspectives on origins and migration throughout the extended eastern Himalayas now in circulation. These include a multitude of narratives and claims from within local communities, studies by several generations of scholars native to or settled in the region itself, discourses generated by nation-states and their agents, as well as explanations by both colonial and post-colonial scholars from outside the region. As a whole, the essays in this volume engage with and critically investigate all of these differing perspectives. Focusing on the extended eastern Himalayas and their hill populations as a region, this book explores new and meaningful units of comparison. On the one hand, the idea of the extended eastern Himalayas is a response to the recent lifting of restrictions on conducting research in the adjacent hill regions of Arunachal Pradesh and Nagaland in northeast India. New UHVHDUFK¿QGLQJVIURPWKHVHDUHDV VHHFKDSWHUVE\$LVKHU%ODFNEXUQ Grothmann, Huber, Post and Wettstein in this volume) have again raised the question of how best to locate and understand the societies in these hill areas within larger frames of reference for comparison and interpretation. While we make no claim that the region of the extended eastern Himalayas is a uniform culture area, its diverse peoples do exhibit a sub- stantial degree of similarity in material culture, ritual practices and oral traditions, including ritual speech. On the other hand, we experiment with the extended eastern Himalayas as a regional focus while conscious of current debates about the virtues of new concepts of cross-border and inter-area regions (e.g., the Zomia concept), and differing views of hill peoples as either marginalized within states, or as strategic political and cultural agents who choose remote hill areas to avoid state enclosure and loss of autonomy (see Scott 2009). These discussions have emerged largely from the scholarship on the hill regions and frontier zones of eastern South Asia, upland Southeast Asia, and southwestern China (e.g., see van Schendel 2002, Scott 2009, Duncan 2004, Giersch 2006). We should also point out that all the essays in this book draw on research among Tibeto-Burman-speakers, which RQO\UHÀHFWVWKHLUSUHGRPLQDQFHLQWKHFKRVHQUHJLRQDQGQRWDSUHUHT- uisite imposed by the editors. Populations who speak Indo-Aryan, Daic/ 7DLDQG$XVWURDVLDWLFODQJXDJHVDUHVLJQL¿FDQWPLQRULWLHVLQWKHH[WHQGHG eastern Himalaya (see the chapter by van Driem), and multilingualism is widespread, too. Comparative research on origins and migration among those populations might uncover different patterns to those suggested by our authors. INTRODUCTION 3

ARRANGEMENT AND THEMES

Given the great diversity of contributions in this volume, we have chosen an overall grouping by region. The chapters are arranged geographically from west to east along the line of the Himalayas and their extensions eastwards, depending upon the case study location or predominent focus DUHD RI HDFK FRQWULEXWRU VHH ¿JXUH %H\RQG WKLV KRZHYHUWKH chapters fall into four thematic groups: theoretical and methodological DSSURDFKHVODQJXDJHLGHQWL¿FDWLRQDQGQDUUDWLYH

Figure 1.1. Map of contributor case study locations.

Theoretical and Methodological Approaches

One set of contributors raises issues of theory and methodology for the study of origins and migration in the extended eastern Himalayas. Robbins Burling considers a fundamental point about why local populations in the region have even bothered to ask—let alone answer—the ‘Where did my tribe come from?’ type of question. Drawing upon various examples from the Garo Hills in Meghalaya and adjacent areas, he concludes that such questions and possible answers to them can just as easily be generated via the stimulation of outside interventions (e.g., Western missionaries and their school systems), as they might arise 4 TONI HUBER AND STUART BLACKBURN from local social and political imperatives (e.g., such as concerns about territory and boundaries). His central caution for scholars dealing with the topic is that “[m]igration stories may reveal much about the aspirations and ethnic values of the people. They can hardly be taken as reliable historical memories.” Burling also rightly reminds us that not all populations throughout our study region are equally concerned with questions of origins and migrations. As his point of departure, Geoff Childs notes that very little of the existing scholarship on Himalayan migrations has so far been articulated with formal theories that attempt to explain migration. He advocates the application of migration typologies, network theory and the notion of ‘social capital’, as well as analysis of ‘push and pull’ factors to better understand why people move, who is most likely to migrate and how migrations are perpetuated across space and time. Using case study data on village settlement in the Nepal highlands, Childs demonstrates that while current social theories of migration require ethnographic and sociological data sets to empirically test them, their application can nevertheless be a powerful tool for considering “migrations as diachronic processes [which] help elucidate how social structures and ethnic identities form over time.” Toni Huber emphasizes that sweeping claims made about the origins and migration of hill peoples dwelling in the extended eastern Himalaya have often been based upon methodological poverty: uncritical use of reported oral narratives as a single type of evidence. Demonstrating an alternative approach, one employing a very diverse range of source materials for a case study in northern Arunachal Pradesh, Huber reveals the relevance of many small-scale movements or ‘micro-migrations’ over time for providing a more convincing understanding of ethnolinguistic distribution within the region. His approach and data also offer a challenge to the frequently published claim that many highland peoples dwelling in Arunachal Pradesh have their origins on, and migrated down from, the Tibetan Plateau. In common, Burling, Childs and Huber argue against the ‘single event’, en masse (and often long distance) migration that tends to dominate local narratives and scholarly writings alike. All three authors demonstrate instead the importance of an analytical focus on migration as ‘process not event’. There is general consensus here, along with F.K.L. Chit Hlaing’s chapter, that the preservation or continuity of distinct ethnic communities/identities across time and space that are implied by such ‘single event’ migration narratives are generally untenable INTRODUCTION 5 and misleading. As Chit Hlaing puts it in relation to his study of Chin, Kachin and Kayah in Burma, “one cannot in general ask usefully where such and such a people in their current identity and under their present ethnonym were very far back in time.” As a corollory, he reminds us that there is no simple or direct equation possible between an ethnic category and a language.

Language

Authors emphasizing the study of languages and linguistic data for understanding origins and migration constitute another thematic group of chapters. George van Driem and Mark Post both probe far back into the ethnolinguistic past of northeast India, a zone at the centre of the extended eastern Himalayan region. As van Driem reminds us, this zone is a complex hill and mountain region where the world’s two most populous families of languages—Tibeto-Burman (if we include Cantonese and Mandarin) and Indo-European—meet. Van Driem’s chapter uniquely surveys linguistic data together with RQJRLQJ¿QGLQJVE\DUFKDHRORJLVWVDQGPRUHUHFHQWUHVHDUFKLQJHQHWLF studies of regional populations. Reading across all this data, he points out that both the linguistic and genetic divides between Tibeto-Burman and Indo-European remain sharp in the Himalayas, that these divides run throughout the southern sub-Himalaya and Terai rather than the high mountain zone and that a marked discontinuity between Neolithic and Bronze Age traditions in the hills and on the plains correlates with this. In contrast, in the zone of northeast India where the mountains and hills of the Indo-Burmese borderlands meet the Brahmaputra basin, van Driem points to a far more complex and older genetic divide between Tibeto-Burman and Austroasiatic populations. This may indicate, according to various scenarios, an earlier ethnolinguistic assimilation of Austroasiatic by Tibeto-Burman peoples in this region. More research results from linguists, geneticists and archaeologists will help to elucidate further aspects of these early ethnolinguistic dynamics within the extended eastern Himalayan zone. In contrast to the breadth of van Driem’s survey, Mark Post focusses on the possible origins of a little-known unit within the extended eastern Himalaya, the Tani cultural-linguistic area of central Arunachal Pradesh (case studies in this volume by Aisher, Blackburn, Huber and—in part— Grothmann all fall within the Tani-speaking area). Using a wide variety of linguistic tools, Post works backwards from today’s Tani languages 6 TONI HUBER AND STUART BLACKBURN and their cultural and environmental context to consider what we can and cannot (yet) know of the Proto-Tani past. His preliminary conclusions suggest this past may not have been too different from what we currently know to be typical for most Tani speakers within the recent historical period: hill-dwelling, Tibeto-Burman-speakers, relatively isolated from outside groups while maintaining contacts within their region. Post highlights the present dearth of reliable research on Tani-speaking societies and their area contributed by other disciplines (anthropology, folklore studies, religious and ritual studies, archaeology, human genetics, geography), research which will aid linguists to gain both further clues and more certainty about the origins and development of the Tani cultural-linguistic area.

,GHQWL¿FDWLRQ

The chapters by Kerstin Grothmann, Marion Wettstein, Mandy Sadan and Koen Wellens emphasize different ways in which claims about RULJLQV DQG PLJUDWLRQV FDQ EH UHOHYDQW IRU QHJRWLDWLQJ LGHQWL¿FDWLRQ Their case studies not only reveal this to be particularly so within the operational frameworks of modern nation-states and their essential FDWHJRU\JHQHUDWLQJ DQG ERXQGDU\¿[LQJ SUDFWLFHV EXW DOVR H[SODLQ that origins and migrations are important contemporary issues for many communities in the extended eastern Himalayas. Grothmann presents oral and written sources on the origins and migration of the Memba, a small, strongly Tibetanized population who were formerly active on both sides of the McMahon Line prior to modern Indian and Chinese state formation along this highly contested Himalayan frontier. She shows how the Memba, drawing upon the widespread Tibetan Buddhist refuge theme of the ‘hidden land’ (see also Childs in this volume), continue to employ both narratives and rituals related to claimed origins and migrations in negotiating their current identity vis-à-vis Tibetans to the north, who have a history of viewing them pejoratively, and their immediate Tani-speaking neighbours, whom they perceive as potential competitors within an Indian state system that KDVDORWWHGWKHPDOORI¿FLDOHWKQLFFDWHJRU\ODEHOVFRQQHFWHGZLWKULJKWV and access to resources. Wettstein’s case study of Naga groups highlights a strategic resort to origin and migration claims as a cultural resource for attempting to establish collective identity in a nationalist political struggle within the modern Indian state. After documenting how the blanket colonial ‘tribal’ INTRODUCTION 7 category Naga was eventually adopted for asserting unity as a single people who thus—according to Naga rhetoric—should be independant from the ‘Indians’, Wettstein reveals the pitfalls and challenges of deploying claims of common Naga origins and migration in support of this national political project. There are very different types of claims in circulation which cannot really be reconciled, and the more precise and localized the claims themselves are the more they risk factionalism in an already fraught ethno-political landscape. Indeed, even making public one’s alternative claims about origins and migrations can lead to censure DQG¿QHVLQWKH1DJDFRQWH[W Using the case of the Jinghpaw or Kachin of northern Burma, Mandy Sadan argues that origin and migration narratives can lose their inherent complexity and negotiated nature when dissociated from their original ritual context, and ‘stripped down’ in contemporary forms that are oriented towards the modern state and its discourses. They then become misunderstood, thus leaving the ethnic identities they are meant to support open to criticism. Sadan employs unique materials from local research groups who are documenting fast-disappearing Jinghpaw ritual practices and dialects as a means to understand the original—and often now obscured—questions that local origin and migration narratives were actually intended to address. Koen Wellens investigates ethnic minority groups of southwest China, such as the Premi and Namuyi, to reveal the dominant role that the modern Chinese party-state’s monopoly on versions of history can play in shaping local perceptions of these same groups’ myths of origin DQGVHQVHRILGHQWLW\:HOOHQVLGHQWL¿HVYDULRXVPHFKDQLVPVLQFOXGLQJ a ‘mythologizing template’ about migrating brothers—one occurring widely, in fact, throughout the extended eastern Himalaya—that facilitate local internalization of state discourse. His study clearly reveals WKH WRWDOL]LQJ QDWXUH RI RI¿FLDO FDWHJRUL]DWLRQ DQG KLVWRULRJUDSK\ LQ contemporary China.

Narratives

Finally, four authors explore narratives of origins and migration among Tibeto-Burman-speaking groups in Arunachal Pradesh, Yunnan and southeastern Nepal. Although the material they analyse is similar— stories of ancestral migration and the origin of the world among these SHRSOHDUHEURDGO\DOLNH²HDFKHVVD\¿QGVGLIIHUHQWHPSKDVHVZLWKLQ it. Stuart Blackburn considers concepts of origin among the Apatani, in 8 TONI HUBER AND STUART BLACKBURN

FHQWUDO$UXQDFKDO3UDGHVK,QERWKFRVPRJRQ\DQGJHQHDORJ\KH¿QGVD central narrative of differentiation from a prior unity. Apatani stories, he concludes, express the idea that the various objects and different people in the world are related through a common source or a shared ancestry. Alex Aisher examines a similar set of oral stories among the Nyishi, who are closely related to Apatani. Retelling their myths of the beginning of the world and their migration, including a genealogy stretching for 32 generations, he emphasizes the theme of exchange. The Nyishi, Aisher suggests, use their oral history to guide their relations with the landscape, with the spirit world, with the dead and with other tribes. Awareness of this ‘coevolution’ is the touchstone of Nyishi thinking. The remaining two essays look at the journey to the land of the dead, which is a belief and narrative theme common to many groups in the extended eastern Himalayas. Both Charles McKhann, on the Naxi in Yunnan, and Martin Gaenszle, on the Rai in southeastern Nepal, map P\WKVRIPLJUDWLRQRQWRWKHSK\VLFDOODQGVFDSH$QGERWK¿QGWKDWWKH journey of the dead is a return to the point from which the ancestors migrated. McKhann, like Blackburn and Aisher, highlights the dynamic between differentiation and unity. Gaenszle makes the interesting point that a return to the land of the dead, to the point of origin, has an ‘inherent tension.’ Reuniting with the ancestors may put one in touch with power and fertility, but it is also fraught with danger. The past is powerful, perhaps too powerful at times. All four essays also point out that these stories not only express ideas EXWDOVRLQÀXHQFHEHKDYLRU+HUHZHFDQVHHKRZWKHVHHVVD\VRYHUODS with the others in this book. Narratives, like other cultural practices, help to reinforce social hierarchy, to negotiate alliances with sometimes- KRVWLOHQHLJKERXUVWRGH¿QHLGHQWLW\DQGWRSRVLWLRQHWKQLFJURXSVLQ rapidly changing political contexts. At the same time, these stories assist scholars to grasp the historical process of social and identity formation. Finally, although we must be cautious about any straightforward relations between language and identity, and about grand claims of migrations, there is little doubt that language, identity and narrative are key to understanding the Tibeto-Burman-speaking peoples in the extended eastern Himalayas. In addition to these four broad themes that organise the book, readers ZLOO ¿QG LQ WKH LQGLYLGXDO FKDSWHUV D ZHDOWK RI QHZ UHVHDUFK GDWD RQ various hill peoples of the extended eastern Himalayas. Together these contributions represent a valuable resource for better understanding the region and its Tibeto-Burman-speaking populations, past and present. INTRODUCTION 9

ACKNOWLEDGEMENTS

The chapters herein were originally presented at the conference “Origins and Migrations Among Tibeto-Burman-Speakers of the Extended Eastern Himalaya” (Humboldt University of Berlin, 23-25 May, 2008). We would like to thank our colleagues Prof. Dr. Michael Oppitz (Zürich/Berlin), Dr. Sarit Chaudhuri (Itanagar) and Ms. Atsuko Ibata (New Delhi) for their presentations during the 2008 conference, which could not be included in this volume. Our thanks go to the Deutsche Forschungsgemeinschaft (Bonn) for enabling our joint research via its 0HUFDWRU3URJUDPDQGIRULWVJHQHURXV¿QDQFLDOVXSSRUWRIWKHVXEVH- quent conference. The Humboldt University of Berlin, and especially the Institute for Asian and African Studies, contributed important additional funding and logistical support for staging the conference, for ZKLFKZHDUHJUDWHIXO7KDQNVDUHGXHWRWKHVFLHQWL¿FUHYLHZHUVHOHFWHG by Brill for their insightful feedback on our manuscript. Finally, many ZHOOGHVHUYHGWKDQNVJRWR.DWMD6ZDUDW5HEHFFD*QXHFKWHODQG$¿D Adu-Sanyah of the Central Asian Seminar at the Humboldt University, who undertook technical editing work on the manuscript, and to designer Tara Daellenbach for completing the camera ready copy. 10 TONI HUBER AND STUART BLACKBURN

REFERENCES

Duncan, C.R. (ed.). 2004. Civilizing the Margins: Southeast Asian Government Policies for the Development of Minorities. Ithaca: Cornell University Press. Giersch, P. 2006. Asian Borderlands: The Transformation of Qing China’s Yunnan Frontier. Cambridge: Harvard University Press. van Schendel, W. 2002. Geographies of knowing, geographies of ignorance: Southeast Asia from the fringes. Environment and Planning D: Society and Space 20(6): 647-68. Scott, J.C. 2009. The Art of Not Being Governed. An Anarchist History of Upland Southeast Asia. New Haven: Yale University Press. TRANS-HIMALAYAN MIGRATIONS AS PROCESSES, NOT EVENTS: TOWARDS A THEORETICAL FRAMEWORK

GEOFF CHILDS

INTRODUCTION

Many studies of ethnic group origins focus on two questions: where did the (name of group) come from, and how did the (name of group) get to their current homeland? Some scholars address these questions by examining migration routes recounted in myths of origin or shamanic journeys. Although ontological narratives provide important emic perspectives on how a group came into being, they often leave the impression that, at some distinct point in time, the group in question moved en masse to their cur- UHQWKRPHODQGVHWWOHGDQGWKHQRVVL¿HGLQWRDGLVWLQFWHWKQLFFRPPXQLW\ The punctuated equilibrium model of migration, characterised by long periods of stability interrupted by brief periods of mobility, glosses over a fundamental, well-documented feature of human movements: although migrations can start as singular events, they typically evolve into protracted processes. As archaeologists have noted, small-scale movements of people are historically more frequent yet less easy to pinpoint than large-scale migrations. Nevertheless, small-scale movements can act as important ini- tiators of social and cultural transformations (Hegmon, Nelson, and Ennes 2000). Focusing exclusively on a seminal migration event, rather than the diachronic process, misses a great part of the story of how communities form over time. The purpose of this paper is to develop a theoretically-informed approach to studying how migration, as a protracted process, has shaped settlement patterns and social structures in the Himalayan region. Although scholars have unearthed a wealth of data to document the settling of Himalayan valleys (e.g., Oppitz 1968; Jackson 1978; Ramble 2008), none have developed an explanatory framework that rests upon formal theories of migration.1 The situation is partially accountable to the fact that migration theories are designed to be empirically tested using 1 Wim van Spengen’s geohistory of Tibetan trade routes (2000) draws on Braudel’s theoretical melding of history and geography to document the development of trade routes between Tibet and South Asia. Albeit not a study of migration per se, van Spengen provides an excellent example of how a scholar can use theoretical insights to provide a more robust and historically contextualized explanation of Himalayan settlement patterns. 12 GEOFF CHILDS data from surveys and interviews—data that simply does not exist in the historical archives. Nevertheless, I contend that systematic research on contemporary migrations can yield novel insights about historical migrations. To accomplish this objective, I draw upon migration theories to address three fundamental questions: Why do people move? Who in a JLYHQFRPPXQLW\LVPRVWOLNHO\WRPLJUDWH"+RZGRQHWZRUNVLQÀXHQFH the scale and direction of migrations? The theoretical discussion forms a backdrop for the second part of the paper: a case study on the settlement history of Sama, a village in Nepal.

PART I: THEORETICAL PERSPECTIVES

'LVFLSOLQDU\RULHQWDWLRQVH[HUWDSRZHUIXOLQÀXHQFHRQWKHW\SHVRIWKHR- retical questions a researcher studying migration is likely to ask (see %UHWWHOO DQG +ROOL¿HOG )RU H[DPSOH D GHPRJUDSKHU PD\ ZDQW to investigate how migration affects birth and death rates of both sending and receiving populations, whereas an economist is more likely to focus on the macroeconomic forces that motivate people to move in the ¿UVWSODFH)RUWKHSXUSRVHRIWKLVSDSHU,DPFRQFHUQHGZLWKTXHVWLRQV more rooted in anthropology, namely, how migration is facilitated by VRFLDOQHWZRUNVDQGKRZPLJUDWLRQLQÀXHQFHVWKHVRFLDORUJDQLVDWLRQRI FRPPXQLWLHV,GUDZXSRQWZRERGLHVRIWKHRUHWLFDOOLWHUDWXUH7KH¿UVW seeks to explain motives behind migration, while the second examines how social networks facilitate the movement of people.

Why Do People Migrate?

$Q\VWXG\RIPLJUDWLRQPXVW¿UVWLGHQWLI\WKHQDWXUHRIWKDWPLJUDWLRQ Brettell (2000:99-102) provides a summary of migration typologies that VWDUWVZLWK¿YHLGHQWL¿HGE\*RQ]DOH] Seasonal migration is a movement by individuals or families that usually occurs once a year in response to seasonal labour opportunities. Temporary, nonseasonal migration is usually undertaken by young, unmarried individuals who leave their places of origin for varying lengths of time in order to gain skills, experience, education, or resources before returning to settle down. A temporary, nonseasonal migrant generally sets out with the intention to return, although may end up leaving permanently. Recurrent migration LVDQLQWHQVL¿FDWLRQRIWHPSRUDU\QRQVHDVRQDOPLJUDWLRQ7KLVW\SRORJ\ describes people who continually leave home for varying periods of TRANS-HIMALAYAN MIGRATIONS 13 time throughout their productive years. Margolis (1995) refers to recurrent migrants as sojourners, and coined the term ‘yo-yo migrations’ to emphasise that many migrants never intend to remain detached from their original communities. Permanent migration covers those who move permanently from one place to another. Permanent migrations often start as temporary or recurrent migrations by individuals, but then develop into the movement of entire families. Continuous migration, a relatively rare phenomenon, usually involves nuclear families that move from job to job and live in temporary residences. Such people have little if any contact ZLWKDKRPHFRPPXQLW\7RWKHVH¿YHFDWHJRULHV*RQ]DOH]DGGVFRQÀLFW migration (1989) which creates a distinct category of migrant, the refugee. Although refugees do not necessarily move willingly, the realities they face in terms of making a living and forming social networks are similar in many ways to other types of migrants (Malkki 1995). Migration typologies emerge within a context of migration auspices, or ‘the social, economic, political and historical contexts within which migration begins and proceeds’ (Grieco 1998:706). Migration auspices can be framed as push factors (those that impel people to leave a place) and pull factors (those that induce people to move to a place). The most widely analyzed push and pull factors in contemporary research are wage differentials that stimulate the movement of people from poorer to wealthier nations. There is certainly merit in using such an approach to study relatively recent migrations throughout the Himalayan region, for example, research on social changes induced by Sherpa migrations to Darjeeling (Ortner 1989, 1999), or the impacts of circular labour on subsistence strategies and old-age care in Helambu (Goldstein and Beall 1981; Bishop 1998:71-80). On the other hand, the wage differential model of migration has less explanatory value the further one moves back in time. In the context of historical trans-Himalayan migrations we therefore need to focus on other push factors, including safety threats (political turmoil, including war or the threat of persecution), KHDOWKWKUHDWV HSLGHPLFV VXEVLVWHQFHWKUHDWV GURXJKWÀRRGLQJDQG crop failure), and demographic factors (increasing population density LQGXFHGE\QDWXUDOJURZWKDQGRULQPLJUDWLRQ /HJHQGVDQG¿UVWKDQG observations provide evidence that people have moved throughout the region in response to some of the push factors listed above. To cite WKUHHH[DPSOHVWKH6KHUSDVFODLPWRKDYHÀHGWKHLUKRPHODQGLQHDVW- ern Tibet to avoid a war (Oppitz 1968; Wangmo 2005); a settlement in Shöyul, Nepal, was abandoned due to the failure of its irrigation system (Ramble 2008:45); and some people in southeastern Tibet abandoned 14 GEOFF CHILDS their villages in response to an epidemic (Bacot 1912). Pull factors include the availability of under-utilised resources in areas that lay on the margins of political control. This resource-based pull factor entails either the existence of uninhabited land, or the existence of land utilised in a limited manner so that potential migrants can envision negotiating access rights or driving previous occupants away. Microeconomic theories of migration treat people as rational actors who make individual choices to maximise net returns on their labour. One microeconomic theory, however, moves away from treating individuals as autonomous entities by focusing on the household as the decision-making unit of analysis. The ‘new economics of migration’ emphasises that decisions are typically made collectively in order to PD[LPLVHEHQH¿WVWRWKHKRXVHKROG 6WDUNDQG%ORRP %HFDXVH the household mediates between the individual and the outside world, a focus on households allows one to better understand how decisions are made in consideration of economic, political, and social conditions, as well as cultural norms (Brettell 2000:107). Using the household as the unit of analysis has certain advantages. Massey and colleagues cite numerous studies to document how migration decisions are not mutually exclusive (either move or stay put). Rather, they often represent household-level strategies to diversify income sources by retaining some members at home who engage in local production while sending others outside, either temporarily or permanently, to capitalise on external opportunities (Massey et al. 1993:439). The new economics of migration retains the assumption that people are rational actors, and that the intent of migration is to maximise returns on labour. In addition it recognises that people in households respond to economic and political circumstances collectively, rather than individu- ally, when deciding who should migrate, and to where. Both macro- and microeconomic theories acknowledge that the propensity to migrate is not uniform across any society, but varies in relation to a range of individual characteristics. Simply put, in any given environment some people are more likely to migrate than others. This point is abundantly clear when examining a relatively recent trans-Himalayan PLJUDWLRQWKHÀLJKWRI7LEHWDQUHIXJHHVWR6RXWK$VLDVWDUWLQJLQ Based on the composition of the exile population it is safe to conclude that the original migrant population was comprised of more males than females. Furthermore, a large percentage of the refugee population inhabited border regions (e.g., Kyirong and Tingri) from where it was comparatively easy to cross into exile. A disproportionate number of the TRANS-HIMALAYAN MIGRATIONS 15

DULVWRFUDF\DQGFOHULV\ÀHGNQRZLQJWKH\FRXOGEHWDUJHWHGIRUSHUVHFX- tion. In comparison very few members of the lower strata of society left 7LEHWSUHVXPDEO\LQÀXHQFHGE\WKHKRSHWKDWWKH\ZRXOGEHQH¿WIURP China’s redistribution policies. Gender, age, proximity to the border, and social status all had a bearing on the propensity to migrate. Gender is an important variable to consider with respect to the propensity to migrate. Cerrutti and Massey (2001:187-88) point out that neoclassical theories of migration tended to treat women as passive agents who had little if any role in the decision-making process; a woman’s propensity to migrate was pegged to that of her husband. Brettell further argues that neoclassical approaches implied that ‘women represented the traditional pole of the continuum and men the pole of modernity’ (2000:109). Many researchers clearly recognise that women do have LQÀXHQFH LQ PLJUDWLRQ GHFLVLRQPDNLQJ DOEHLW WKHLU LQÀXHQFH YDULHV from one society to another and may not be readily apparent (Riley and *DUGQHU ,QJHQHUDOVWXGLHV¿QGWKDWPHQ¶VLQWHQWLRQVDUHPRUH closely related to their income-earning roles, while women’s intentions are more closely related to their family-support roles (De Jong 2000; Cerrutti and Massey 2001).

How Do Migrations Perpetuate?

Most migrations commence with the movement of a few enterprising individuals or pioneers, people who have a higher propensity to take risks or explore new options. If successful, the information they gather DQGWKHQHWZRUNVWKH\IRUPLQÀXHQFHWKHSURSHQVLW\WRPLJUDWHIRURWKHUV OHDGLQJWRWKHSHUSHWXDWLRQRIDPLJUDWLRQVWUHDPDQGWKHLQWHQVL¿FDWLRQ of social networks between places of origin and destination. Network theory is devoted to studying the ways that social networks form and facilitate the continuing movement of people (Massey et al. 1993). Networks are multilocal. In other words, ‘they encompass a variety of geographical destinations’ (Wilson 1998:395-96). At a surface level it is easy to envision the migrant network as a dyad composed of a sending and a receiving community. However, networks are typically far more complex and involve links with numerous localities. The multilocal nature of networks allows them to operate as a form of social capital, GH¿QHG DV µWKH FDSDFLW\ RI LQGLYLGXDOV WR FRPPDQG VFDUFH UHVRXUFHV by virtue of their membership in networks or broader social structure’ (Portes 1995:12). As social capital, people use networks to learn about ZKHUHWRJRKRZWR¿QGHPSOR\PHQWDQGKRZWRPDQDJHWKHLUGDLO\ 16 GEOFF CHILDS lives in unfamiliar surroundings. Networks thereby reduce the costs and risks associated with migrating, and in the process help develop and maintain migration streams. Networks can be based on weak ties, strong ties, or a combination thereof. The strength of ties varies according to ‘the amount of time, the HPRWLRQDOLQWHQVLW\WKHLQWLPDF\ PXWXDOFRQ¿GLQJ DQGWKHUHFLSURFDO services which characterizes the tie’ (Granovetter 1973:1361, cited in Grieco 1998:705). Weak ties represent relationships between acquain- tances and people of common origin, in contrast to strong ties which are relationships among close friends and kin. Strong tie networks are more densely clustered and are comprised of individuals who have stronger emotional bonds, interact on a regular basis, and share the same information. Weak ties are more diffuse and are comprised of people who have sporadic contact with each other and few emotional commitments. Two important points can be made with respect to strong and weak WLHV7KH¿UVWLVWKDWPLJUDWLRQDXVSLFHV GLVFXVVHGDERYH LQÀXHQFHWKH nature of social networks that develop at migration destinations (Grieco 1998). When individuals rather than families or communities move they tend to form weak networks among themselves, while seeking to enhance their ties with members of the host community in order to adjust to the new environment. One result is rapid assimilation. In contrast, when entire social units move (e.g., families or communities), they tend to form stronger social networks among themselves and resist assimilation by establishing distinct ethnic communities (Grieco 1998:706). The second point is that networks expand through ‘the strength of weak ties’ (Wilson 1998:397-98). According to Granovetter (1982) weak ties act as information conduits between more densely clustered networks of kin and friends. Diffuse social networks in the form of weak ties constitute a form of social capital. A rise in the number of migrants increases the quantity of weak ties thereby increasing the volume of social capital available to the migrant group as a whole. Furthermore, the strength of weak ties is greater among higher socioeconomic groups because they have access to more information and resources (Wilson 1998:398). Members of higher status social or economic groups are in better positions to disseminate information through weak networks because of the volume of information they possess, as well as the resources that facilitate their movements and communications with other members of a weak tie network. Massey integrates network theory with macro- and microeconomic perspectives through the concept of ‘cumulative causation’, or the TRANS-HIMALAYAN MIGRATIONS 17 dynamic interplay between migrant communities and their places of origin (Massey 1990; Massey et al. 1993). Like network theory, cumulative causation examines the complex feedback mechanisms that contribute to the self-perpetuating movement of migrants. Ideological forces in a sending community, referred to as the culture of migration, change over time in response to political and economic forces, and the ensuing PLJUDWLRQSURFHVV7KHFXOWXUHRIPLJUDWLRQFRQWLQXDOO\LQÀXHQFHVWKH propensity for individuals to migrate by shaping peoples’ receptiveness to the idea that moving elsewhere can lead to a better life. If receptiveness to moving becomes deeply embedded, that is, if a strong culture of migration develops, then the propensity to migrate increases for more people. The culture of migration is therefore central to the argument embedded in the cumulative causation model that ‘each act of migration alters the social context within which subsequent migration decisions are made, typically in ways that make additional movement more likely’ (Massey et al. 1993:451). To summarise, various push and pull factors shape the environment within which migration decisions are made. Except in the case of forced PLJUDWLRQVSHRSOHJHQHUDOO\ZHLJKWKHFRVWVDQGEHQH¿WVRIPLJUDWLRQ options, which can include moving en masse (a household or entire community), or only sending selected members outside. Migration decision- making is best described as a relationship between the propensity to PLJUDWH DVLQÀXHQFHGE\PLJUDWLRQDXVSLFHVWKHFXOWXUHRIPLJUDWLRQ individual human capital attributes, and household characteristics) and WKHPRWLYDWLRQWRPLJUDWH DVLQÀXHQFHGE\VRFLDOQHWZRUNVDQGYDULRXV push and pull factors). The most appropriate unit of analysis for migration decision-making is the household. The next objective of this paper is to use the theoretical perspectives presented above as a means for shedding new light on the settlement history of Sama, a village in the highlands of Nepal.

PART II: THE PROCESS OF POPULATING SAMA

Sama is a village of roughly 500 inhabitants situated at an elevation of 3500m in the Nubri Valley, Gorkha District, Nepal. The village is populated by an admixture of ethnic Tibetans who migrated from the north, and ethnic Ghales who entered the valley from the south. The following sections detail Sama’s social and religious organisation, sketch the village’s settlement history, and deploy theoretical perspectives to better understand the long-term processes that formed the village’s unique 18 GEOFF CHILDS

FKDUDFWHU ,Q WKH ¿QDO VHFWLRQ , DUJXH WKDW WKH QHWZRUNV HVWDEOLVKHG WKURXJK SUHYLRXV LQPLJUDWLRQV LQÀXHQFH WKH FRQWHPSRUDU\ SDWWHUQ RI out-migration and thereby continue to play a role in shaping the social and religious organisations of Sama.2

Social and Religious Organisation

Sama’s society is comprised of four patrilineal descent lineages (rgyud pa). Listed in order of descending prestige these are Ngadag (‘Possessing Power’), Pönzang (‘Good Rulers’), Yorkung (‘Irrigators’), and Chumin (‘Low and Inferior’). The Ngadag are a lineage of householder lamas (sngags pa) who descend from Tibet’s medieval emperors. The Pönzang, as their name implies, were once associated with political leadership. They claim to descend from Kyika Ratö, a legendary illegitimate son of D7LEHWDQTXHHQZKRÀHGWRWKH+LPDOD\DV0RUHFRQFUHWHO\3|Q]DQJ members claim to have migrated from Barpak, an ethnically Ghale village to the south of Nubri. Yorkung and Chumin members comprise the majority of Sama’s population. According to oral accounts they are descendents of Sama’s original settlers. Sama is a Nyingmapa Buddhist community. The Ngadag lamas are especially devoted to the teachings of the Changter School (byang gter, ‘Northern Treasure’) that is associated with Rigzen Gödemchen, a fourteenth century cleric who revealed many hidden teachings (gter ma) that, according to legend, had been concealed in the eighth century by Padmasambhava. Rigzen Gödemchen’s revealed corpus includes keys to opening, and guidebooks for entering, beyül (sbas yul), or ‘hidden lands’. BeyülDUHUHPRWHGLI¿FXOWWRDFFHVVYDOOH\VVLWXDWHGLQWKH+LPDOD\DQERU- derlands. Tibetans believe that beyül were concealed by Padmasambhava, who designated them as settlement destinations to be opened during times RISROLWLFDOFRQÀLFWPRUDOGHJHQHUDWLRQDQGUHOLJLRXVGHFOLQH2QHVXFK hidden land, Kyimolung, is situated within the Nubri Valley.

Sama’s Settlement History

The earliest reference to Nubri dates from the late tenth or early eleventh century when the intrepid yogi Milarepa (1040-1123) reportedly

2 Elsewhere I have extensively described Sama in ethnographic (Childs 2004a), demographic (Childs 2008:105-133), and historical (Childs 2000, 2001) terms. I refer readers to those publications for further details, including discussions of primary source materials. TRANS-HIMALAYAN MIGRATIONS 19 visited the area to meditate in a cave near the present location of Sama. According to his account, Milarepa encountered a ‘land of a different language’ which he characterised as a ‘pitch-black realm of ignorance where the dharma had not yet spread.’ If Milarepa’s account is accurate,3 then Nubri in the eleventh century was inhabited either seasonally or permanently by a non-Tibetan speaking people that practiced a religion other than Buddhism. During Milarepa’s lifetime Gungtang, a kingdom founded by descendents of Tibet’s medieval emperors, extended its domain to encompass the Nubri Valley. Around 1280 Gungtang’s rulers marked their control over Nubri by erecting Black Cliff Fort; its remains are still visible across the river from Sama. The garrisoning of this fort probably represents the ¿UVWVHWWOLQJRI7LEHWDQVLQWKHYDOOH\DQGPD\FRLQFLGHZLWKWKHHVWDEOLVK- ment of Sama by the ancestors of today’s Yorkung and Chumin lineages. During the height of Mongol hegemony in Tibet (c. 1249-1349), the Gungtang rulers formed a marital alliance with Sakya, the most powerful Buddhist order due to their priest-patron relationship with the Mongol overlords. However, by the late fourteenth century the Gungtang rulers shifted their alliance to the Nyingmapa. They patronised the Changter School by inviting Rigzen Gödemchen to reside at Palbar Monastery in Kyirong. From there he reportedly set forth to ‘open’ Kyimolung, a hidden land situated within Nubri Valley. A guide to Kyimolung reputedly discovered by Rigzen Gödemchen states that ‘Tibetan is spoken in the upper part [of the Nubri Valley].’ If Rigzen Gödemchen was the author (rather than discoverer) of this guide,4 then Sama, which is situated in WKHXSSHUSRUWLRQRI1XEULZDVGH¿QLWHO\LQKDELWHGE\7LEHWDQVSHDNHUV DWWKHGDZQRIWKH¿IWHHQWKFHQWXU\ The downfall of Gungtang closely corresponds with another migration into Nubri, in this case the entrance of a single individual who has left a lasting legacy on Sama’s social structure. In 1620 the king of Tsang conquered Gungtang and exterminated members of its ruling family. According to beyül prophesies the demise of Gungtang signals the time

3 The historical validity of Milarepa’s descriptions are suspect because many of them circulated exclusively in oral form for centuries before being compiled and committed to paper in the late 1400s. His references to borderland peoples appear to be tropes as much as reliable descriptions. 4 Michael Aris 1988 raised the proposition that treasure revealers, such as Pemalingpa (1450-1521), composed many of the written teachings that they reputedly discovered. Similarly, because prophesies contained within guidebooks to various beyül contain very VSHFL¿FUHIHUHQFHVWRKLVWRULFDOHYHQWVWKDWRFFXUUHGFHQWXULHVDIWHU3DGPDVDPEKDYDYLVLWHG Tibet, I suspect that these texts were written contemporaneously with their discoverers; Childs 1999. 20 GEOFF CHILDS when descendents of Tibet’s medieval emperors should seek refuge in places like Kyimolung. The lamas of Tradumtse Monastery, located to the north of Nubri, belonged to precisely such a lineage. According to oral accounts the people of Sama invited Yönden Puntsog, the younger brother of Tradumtse’s head lama, to establish a temple and reside in their village. Through marriage and procreation Yönden Puntsog became the ancestor of all the Ngadag lineage members who live in Sama today. Based on the Ngadag lineage’s genealogy Yönden Puntsog was born close to 1600 and therefore moved to Sama around the time that the King of Tsang eradicated Tradumtse’s patrons, the Gungtang rulers. The presence of Yönden Puntsog or his descendent is historically attested in 1688 when a lama from Kutang, the lower part of the Nubri Valley, recorded meeting ‘the Ngadag lama’ in Sama. Katog Rigzen Tsewang Norbu, who visited Nubri in 1729, also mentions the Ngadag lamas of Sama in several of his writings. In 1642 the Mongol Gushri Khan bestowed the Fifth Dalai Lama with secular power over Tibet, and by the early 1700s Tibetans had organised their domain into districts (rdzong). Nubri became part of Dzongga District and paid taxes to the Tibetan government via the district com- missioner of Dzongga. Shortly thereafter the Pönzang ancestors entered Sama. Their descendents claim to be ethnic Ghales who migrated from the village of Barpak. Gurung society is comprised of four major clans, one of which, the Ghale, held the traditional position of leadership. Pignède conjectures that Ghale is etymologically related to rgyal, the Tibetan word for king, and supports this argument by noting that the traditional kings throughout Gurung territory were invariably Ghale (1993:168). According to one legend the Ghale ancestors came from the Tibetan Plateau and established several small kingdoms in central Nepal before being usurped by Rajputs, including those who established the Kingdom of Gorkha (Pignède 1993:162-165, 197-198). The Pönzang DQFHVWRUV ZHUH SUHVXPDEO\ *KDOH UXOHUV ZKR ÀHG %DUSDN VLWXDWHG between Gorkha and Nubri, to escape their enemies. Their lineage name, Pönzang (‘Good Rulers’), implies a traditional role of political leadership, making them prime targets for the usurping Rajputs. Also, during the 1700s Sama was part of Tibet thereby placing the Ghale refugees beyond the reach of their enemies. The Kingdom of Nepal, formed under the leadership of the Gorkha king Prithvi Narayan Shah during the late eighteenth century, incorporated Nubri into its domains in 1856 following a war with Tibet. A FHQWXU\ODWHU6DPDEHFDPHDVDIHKDYHQIRU7LEHWDQVÀHHLQJ&KLQHVH TRANS-HIMALAYAN MIGRATIONS 21 rule. From 1959 through the early 1960s hundreds of Tibetans passed through the village. Whereas most transited through the area on route to Kathmandu, several families settled in Sama. Since that time, with the exception of in-marrying men and women, nobody has settled in Sama as a migrant. To the contrary, the current trend is out-migration. These days many parents send their children to Kathmandu, Nepal, and various Tibetan encampments throughout India to attend secular schools or to reside in monasteries. The majority of these children never return except to visit. ,QVXPPDU\6DPDZDV¿UVWHVWDEOLVKHGE\PHPEHUVRIWKH

Theorising Sama’s Settlement Process

Destinations for migration are rarely the product of random choice. To the contrary, they are intentionally chosen on the basis of information passed through social networks. One facet of trans-Himalayan networks is represented by yogis who play a unique role as migration route pioneers and disseminators of information about potential destinations. The pioneering role of yogis is attested in a Tibetan cultural template, described by Huber (1997), for making landscapes amenable to human settlement. The process involves Padmasambhava who lays the groundwork by converting malevolent deities into ‘defenders of the faith’. Henceforth, a place is primed for being opened through ritual procedures by a yogi, such as Milarepa, who leaves imprimaturs throughout the natural surroundings—tangible evidence of his miraculous deeds that convert the landscape into a habitable realm. Once an area is opened, others can visit for meditation, pilgrimage, or even settlement. The residents of many areas throughout the Himalayas, including Nubri (Childs 2004), Solu-Khumbu (Wangmo 2005), Langtang (Ehrhard 1997; 22 GEOFF CHILDS

Lim 2004), and Helambu (Clarke 1980) refer to the sequential subjuga- tions of Padmasambhava and Milarepa as an important precondition for human settlement. Moving beyond the legendary nature of these events, we know for sure that Milarepa travelled widely and attained considerable fame. During his lifetime Milarepa visited hundreds of valleys throughout the Himalayas. Milarepa’s biography (Lhalungpa 1996) and collected poems (Chang DUH¿OOHGZLWKGHVFULSWLRQVRIWKHODQGVFDSHVWKDWKHYLVLWHGDQG the people he encountered. If the corpus of writings that is attributed to him merely approximates the extent of his endeavours, then it is safe to conclude that he traversed a vast terrain. Milarepa was thereby involved in an extensive, multilocal, weak tie network that involved religious practitioners, patrons, disciples, and villagers. Through his travels and teachings Milarepa placed himself in a position to convey a considerable amount of information to others about Himalayan geography. The inhabitants of Nubri, Solu-Khumbu, Langtang, and Helambu also conceive of their abodes as hidden lands (beyül). Padmasambhava reputedly concealed these valleys, and left prophesies stipulating when each beyül could be opened, and by whom. Treasure revealers (gter ston), especially those involved in the opening of hidden lands, con- ceivably played pioneering roles similar to Milarepa. Regional rulers often patronised these individuals, and facilitated their movements as part of a strategy to extend political domains (Ehrhard 1999a, 1999b). Domar Migyur Dorje (born 1675) is credited with opening the hidden land Namgo Dagam which currently lies in Nepal’s Langtang Valley (Ehrhard 1997). The news of his discovery prompted a wave of migration (Lim 2004:45). Similarly, the Sherpas envision their valley to be the beyül Khenbalung. According to tradition Rigzen Gödemchen (1337- 1408) opened Khembalung during his lifetime. Then in the mid-1400s ZKHQWKH6KHUSDVZHUHÀHHLQJ0RQJROLQLWLDWHGVWULIHLQ.KDPWKHODPD Sangye Paljor recalled a prophecy: Khembalung could be opened when the world was overcome by wars. Sherpa legend states that he dispatched his son and disciple to seek out Khumbu. They established hermitages that gradually developed into settlements (Wangmo 2005:22-23). Because the initial populating of Sama by Tibetans occurred after Milarepa’s visit, one cannot rule out the possibility that he played a role in prompting people to settle in the upper part of the Nubri Valley. The settlement process probably commenced as seasonal migrations of people periodically crossed passes from the north in search of resources, such as pastures. At some point the migration turned from seasonal to TRANS-HIMALAYAN MIGRATIONS 23 permanent when entire families began to settle in Sama. Because the locals practice lineage exogamy in marriage, Yorkhung and Chumin families must have settled simultaneously. Otherwise, settlers would have had to either import wives or marry incestuously. Pull factors would include the availability of land in a relatively lush, forested environment where agricultural production did not require the building and maintaining of an irrigation system. If the settling of Sama did coincide with the establishing of Black Cliff Fort by the Gungtang rulers, then the migration transpired under the auspices of a local polity’s territorial expansion. The settling of the Ngadag ancestor in Sama, an example of permanent migration, is consistent with seventeenth and eighteenth century movements of Nyingmapa practitioners toward the Himalayan borderlands in response to attacks on their teachings and institutions by Mongol and Tibetan enemies (Ehrhard 1999b). The Ngadag migration occurred under VSHFL¿FDXVSLFHVQDPHO\SROLWLFDOLQVWDELOLW\DVVRFLDWHGZLWKWKHGHPLVH of the Gungtang rulers. Push factors included a general atmosphere of persecutions against the Nyingmapas and the loss of Tradumtse’s most important patron, the ruler of Gungtang. A pull factor included Sama’s close proximity to Kyimolung, a beyül prophesised to be a destination for descendents of Tibet’s medieval emperors (e.g., members of the Ngadag lineage) under the precise auspices detailed above. The fact that Rigzen Gödemchen is credited with opening this beyül, and that he was supported by the same Gungtang ruler who patronised the Tradumtse ODPDVSRLQWVWRWKHUROHRIWKH\RJLDVDPLJUDWLRQSLRQHHUDQGWKHLQÀX- ence of weak tie networks in identifying potential places for migration. The Ngadag migration also illustrates the household-level nature RI PLJUDWLRQ GHFLVLRQPDNLQJ DV LW DSSOLHV WR D VSHFL¿F W\SH RIIDP- ily, Nyingmapa householder lamas (sngags pa). In many Nyingmapa communities such as Nubri (Childs 2004), Solu-Khumbu (Ortner 1989), Langtang (Lim 2004), and Helambu (Clarke 1980), these lamas control village temples and their associated estates. Succession operates according to the rule of primogeniture: the eldest son inherits his father’s estate. A younger son is ineligible to inherit his father’s position but by virtue of his bloodline has the potential—if given the opportunity—to form an independent, collateral lineage. Clarke (1980) documents the expansion of collateral lineages in Helambu through the founding of new temples by younger sons of lamas, a process that transpired from the seventeenth through the twentieth century. In the case of the Ngadag migration, the villagers of Sama invited Yönden Puntsog, the younger brother of Tradumtse’s incumbent, to establish a temple within their midst. The 24 GEOFF CHILDS decision to accept the offer was no doubt made as a household in con- VLGHUDWLRQRIWKHFRVWV ORVVRIRQHPHPEHU DQGEHQH¿WV H[SDQVLRQRI the lineage) associated with migration and founding a new temple. This case, and the case of Helambu, illustrates how birth parity affects the propensity to migrate among members of householder lama families; younger sons had a heightened propensity to migrate by virtue of their ability to form collateral lineages. The movement of Pönzang lineage ancestors to Sama is an example RIFRQÀLFWPLJUDWLRQ,IP\FRQMHFWXUHDERXWWKHWLPLQJDQGUHDVRQVIRU their move are correct, then a major push factor was the belligerence of political foes. Similar to the Ngadag case, the Pönzang ancestors’ social status was closely associated with their propensity to migrate under spe- FL¿FSROLWLFDODXVSLFHV6DPDZDVQRWDUDQGRPGHVWLQDWLRQEHFDXVHWKLV village and Barpak have a long history of trade relations and therefore a preexisting network that facilitated Pönzang lineage members’ incorporation into the community. However, because of the weak tie nature of this network, these ethnic Ghales assimilated to the point where they are culturally and linguistically indistinguishable from the original Tibetan inhabitants of Sama. 7KH SURSHQVLW\ WR PLJUDWH DOVR ÀXFWXDWHV LQ DVVRFLDWLRQ ZLWK FRQ- straints and opportunities imposed by political administrations. From the early 1700s until 1856 the majority of central Tibet’s residents were subjects of estates held by one of three landowning institutions: the government, the nobility, and the monasteries. Each institution registered its population and controlled the movement of its subjects. A person had to obtain a ‘human lease’ (mi bogs) to move from one place to another. The contract entailed payment of an annual fee to one’s lord in exchange for mobility (Goldstein 1971). However, a heavy tax burden and chronic indebtedness often prompted Tibetans to escape penury by slipping across the border into Nepal. The prelude to a 1958 government household register from Kyirong District alludes to this practice by warning: Heavy punishment will be levied to those persons who are found to throw blame onto others or who suppress facts of the smallest nature even to the size of a sesame seed. If any such misdeeds are found later the persons involved will take full responsibility as everything will be exposed like a chronic disease diagnosed. Similarly, arrangements of marriages, entering into religious life, and exchange of subjects are not permitted without prior permission, [this applies] HVSHFLDOO\WRWKRVHLJQRUDQWRQHVZKRÀHH to other lands thinking that they will be more secure and have a better life6XFKSHUVRQVPDNLQJÀLPV\H[FXVHVWRÀHHIURPWKHFRXQWU\PXVWEH stopped with tight security and the leaders and people have taken oath that TRANS-HIMALAYAN MIGRATIONS 25

such incidents will not be allowed to occur, for which cause the following persons have undertaken the oath: [seals of witnesses] (Childs 2008:281). Placing barriers on movements no doubt impacted the culture of migra- WLRQE\PDNLQJPRELOLW\DGLI¿FXOWWRDFKLHYHSURVSHFW8QWLO1XEUL ZDV D SDUW RI7LEHW VR ÀHHLQJ WKHUH ZRXOG QRW KDYH SODFHG D SHUVRQ beyond the reach of his lord. After 1856 Nubri became a part of Nepal and thereby a potential des- WLQDWLRQIRUSHRSOHÀHHLQJ7LEHW3HUKDSVVRPHSHRSOHGLGHQWHU6DPD in this manner, but if so they left no evidence in the form of new patri- lineages resulting from marriage and procreation. The situation changed dramatically in 1959 when Nubri became an important escape route IRUSHRSOHÀHHLQJWKHWXUPRLOLQ7LEHWDIWHUDIDLOHGUHEHOOLRQDJDLQVW Chinese rule. Tibetan refugees (forced migrants) represent the last wave of people to enter Sama. In this case a major push factor was the fear of repression under Chinese rule. Although most Tibetans merely passed through Sama, several families did stay. Not coincidentally, those who remained are nomads who had long-lasting trade relationships with the people of Sama. Once again, weak networks shaped the migration and settlement process. Thus far I have focused on the populating of Sama by its four main lineages and Tibetan refugees. In addition, an equally important but less visible process has introduced a continuous trickle of new residents into the village. Patrilocality, the customary post-marital residence pattern in Nubri, has resulted in the circulation of women among neighbouring villages. Because most marriages are arranged between families, the locus of the marital migration decision is the household, not the individual. Furthermore, social networks shape the propensity for women to migrate. For example, Ngadag families aspire to contract marriages with females from the most prestigious local lineages, notably the Lamzha lineage in the neighbouring village of Lho. Virtually all the women from Lho who married into Sama are members of this lineage, while conversely the majority of Sama’s women who married into Lho are members of the Ngadag lineage. The pattern of exchanges between the Ngadag and Lamzha lineages illustrates how strong networks shape the ongoing stream of marital migrations between these two villages. ,Q VXPPDU\ PLJUDWLRQ DXVSLFHV KDYH LQÀXHQFHG WUDQV+LPDOD\DQ migrations and shaped the ways that communities, such as Sama, have formed. Historically, the multilocal weak networks through which religious adepts moved facilitated the dissemination of information about potential migration destinations. Depending on circumstances, in some 26 GEOFF CHILDS cases entire families migrated to Sama (Yorkung, Chumin, and Pönzang ancestors), whereas in other cases only individuals made the move (the Ngadag ancestor Yönden Puntsog, and women who married into the YLOODJH ,QHYHU\FDVHWKHSURSHQVLW\WRPLJUDWHKDVEHHQLQÀXHQFHGE\ political and economic auspices, and varied along a range of variables LQFOXGLQJVRFLDOVWDWXVUHOLJLRXVDI¿OLDWLRQDQGJHQGHU

Cumulative Causation: How History is Shaping Contemporary Migrations

Currently, more people are moving out of Sama than moving in. A salient trend is for parents to send one or more of their children to distant schools or monasteries. Massey’s cumulative causation model helps explain how the current situation has developed in recent years. Until the 1980s very few parents sent children outside the village. However, the situation changed in response to demographic circumstances. As argued elsewhere (Childs 2004b), foreign patronage stimulated a building boom of monasteries founded by Tibetan refugees living in Kathmandu, Nepal. This in turn stimulated an increase in the number of recruits needed to ¿OOWKHUDQNVRIWKHQHZLQVWLWXWLRQV7KHH[SORVLRQRIPRQDVWLFLVPFRLQ- cided with a dramatic drop in the fertility rate among Tibetan exiles. As a result, many monasteries turned to the ethnically Tibetan highlands of 1HSDOWRUHFUXLWQRYLFHV2QHLPSRUWDQWODPDVXPPDULVHGWKHIXO¿OOPHQW of his recruitment needs in the following terms: During the time of the decrease in my merit, the Buddha Dharma in the land of snow (Tibet) was diminishing. This was especially the fate of [name of monastery] . . . At that time I became a refugee and did not know where to go. During this time of hardship, having the motivation to keep my monastery alive, I purchased a small piece of land in front of the Great Stupa of Boudhanath in Kathmandu, Nepal and built [in 1982] a small monastery, which I named [name of monastery]. $IWHUEXLOGLQJWKHPRQDVWHU\LQ%RXGKDLWZDVGLI¿FXOWWR¿QGFKLOGUHQ who wanted to become monks. I decided to ask some people I knew about ¿QGLQJFKLOGUHQZKRZLVKHGWREHPRQNV7KH\VDLG³1RZDGD\VDOOSHRSOH want to go to school and study or to do business and earn much money. It’s DOPRVWLPSRVVLEOHWR¿QGFKLOGUHQWREHFRPHPRQNV´ After hearing this, I was very depressed and disappointed but nevertheless I made a fruitful aspiration with good motivation and put many Chakras (mandalas) of the Sangha and Harmony of the Sangha into the main Buddha statue of the monastery. As a result of that, [person’s TRANS-HIMALAYAN MIGRATIONS 27

QDPH@EHFDPHWKH¿UVWPRQNIURP1XSUL9LOODJH>1XEUL9DOOH\@*RUNKD >'LVWULFW@1HSDO(YHQWXDOO\WKHQXPEHURIPRQNVJUHZDQGÀRXULVKHGDQG soon there were almost as many monks as in the monastery in Tibet.5 The monastery and associated secular school run by this lama in Kathmandu are now two of the primary destinations for Sama’s parents to send their children. The case illustrates a principle, embedded in Massey’s cumulative causation model, that each case of successful PLJUDWLRQDOWHUVWKHFXOWXUHRIPLJUDWLRQ:KHQ,¿UVWZRUNHGLQ6DPDLQ 1995 parents were willing to send sons outside the village to monasteries but were very reluctant to send either sons or daughters to secular schools. That situation has changed to the point where sending both boys DQGJLUOVLVFRQVLGHUHGDQRUPDODFFHSWDEOHDQGEHQH¿FLDOSUDFWLFH Ngedön Öseling Monastery in Kathmandu is another major destination for Sama’s parents to send their sons. In 1997, 17.3 percent of all males aged 5-24 (19 out of 100) from the village were monks residing in this particular monastery. How it came to be a major destination illustrates the importance of historical networks in establishing and maintaining streams of migrants. Urgyen Tulku, a high-ranking Nyingmapa lama, VSHQWWLPHLQ6DPDDIWHUÀHHLQJIURP7LEHW$VDZLGRZHUKHWRRNWKH daughter of one of Sama’s Ngadag lamas to be his second spouse. During the 1980s Urgyen Tulku established several monasteries in Kathmandu, including Ngedön Öseling whose abbot, Tsoknyi Rinpoche, is Urgyen Tulku’s son born of his bride from Sama. Because of these important connections many parents send their sons to be ordained as monks who study under Tsoknyi Rinpoche’s tutelage. The decision to become a monk is typically made by the parents in consideration of many factors. Pull factors can be both cultural (parents acquire merit for donating a child to the clerisy) and economic (the institution subsidises the rearing and education of a child). Family composition is important; monks are often redundant members of households with more than one son. Parents also use the monastic option as an economic GLYHUVL¿FDWLRQ VWUDWHJ\ 0RQNV RQFH HGXFDWHG DQG PDWXUH FDQ HDUQ income through the performance of rituals. Some of that income returns to village-dwelling families in the form of remittances. According to the typologies outlined above, the majority of move- PHQWV WRGD\ FDQ EH FODVVL¿HG DV QRQVHDVRQDO PLJUDWLRQ +RZHYHU distinguishing the intentions of movers is problematic because decisions are made by parents, not individual migrants. In reality many current moves are destined to become permanent due to a lack of pull factors

5 www.rinpoche.com/Nupri/englishletter.htm, accessed June 2005 28 GEOFF CHILDS

(i.e., vocations for monks and educated individuals) to attract return migrants. Furthermore, by becoming a monk a boy becomes ineligible WRLQKHULWKLVIDWKHU¶VSURSHUW\+HDOVRQXOOL¿HVKLVDELOLW\WRSHUIRUP physically demanding tasks that are necessary for survival in Sama but that contradict the cultural expectation that monks remain detached from mundane social and economic activities. Although monks can and often do visit their families, the absence of formal institution for celibate clerics has made it problematic for them to reside long-term in or near their natal villages.

CONCLUDING REMARKS

I have used this opportunity to demonstrate how certain theoretical frameworks can assist in the interpretation of historical migrations. In the process, I have argued that migrations are best analyzed as interre- lated processes rather than singular events, and that viewing migrations as diachronic processes helps elucidate how social structures and ethnic identities form over time. I have extended the analysis to the present to argue that a theoretically informed, historically grounded perspective on migration is an essential building block for understanding contemporary patterns of movement. Although Sama represents one unique case, I have referred to other Himalayan valleys in Nepal, notably Solu-Khumbu, Langtang, and Helambu, to highlight similarities in the ways that Tibetan communities lying south of the Himalayas came into existence. Parallels can also be found in the eastern Himalayas, suggesting that the phenomena described in this paper are more widespread. 6SHFL¿FDOO\6DUGDU$INKDPL UHFRUGVQXPHURXVLQFLGHQWVIURP the eighteenth century onward whereby Tibetans responded to political strife by following Nyingmapa lamas toward the hidden land Pemakö, which lies south of the Himalayan massif near the deep gorges of the lower Tsangpo (Brahmaputra) River. In the early twentieth century Bacot (1957) visited several abandoned villages in Kham, eastern Tibet, IURP ZKLFK SHRSOH UHSRUWHGO\ KDG ÀHG LQ VHDUFK RI 3HPDN| %DLOH\ (1957:34-38) later visited a village named Mipi where the remnants of this Pemakö-seeking expedition were living under adverse conditions, beleaguered through constant skirmishes with their non-Tibetan neighbours. Bailey also visited a village named Kapu where he found people from eastern Bhutan who had settled after searching unsuccessfully for Pemakö. Compelling similarities between all these cases include a Tibetan cultural template for migration that is linked to an emic concept TRANS-HIMALAYAN MIGRATIONS 29 of sacred space, the pioneering role played by itinerant yogis, the historical auspices that prompted many trans-Himalayan migrations, and the function weak networks play in disseminating geographical information. Together, these similarities help explain why many valleys situated south of the Himalayas were populated sequentially by people who came VHHNLQJUHIXJHIURPWXUPRLOZKRVHUHOLJLRXVDI¿OLDWLRQLV1\LQJPDSD and whose clerics are householder lama devotees of the Changter School whose corpus of teachings includes a particular conception of sacred space, the beyül. One topic left unaddressed is the role that women have played in trans-Himalayan migrations. Men in Tibetan societies are generally more mobile and therefore play a greater role in scouting migration destinations and initiating moves. However, transforming a potential destination into a full-time place of residence requires both men and women in order to constitute a productive and reproductive community. Unfortunately, the nature of historical data from the region tells us far more about male agents than female agency. The role of gender has received considerable attention in the Himalayan region with respect to marriage (Schuler 1987), social organisation (Watkins 1996), and religion (Schaeffer 2004), but still needs to be worked out in far greater detail with respect to migration in both historical and contemporary set- tings. At this juncture we can only assume that women had more than passive voices in the momentous, household-level decisions to pack up and move their families. To conclude, the deployment of theoretical frameworks for the study of contemporary migrations is considerably easier than for historical migrations due to the nature of the data on hand. Himalayan historical archives do not easily yield data that is amenable to the formal testing of theories. Nevertheless, my goal in this paper has been to demonstrate how a combination of archival and ethnographic data can be interpreted in light of migration theories. By doing so, one can attain a more nuanced understanding of migration as a process that has left tangible imprints on the social and religious fabric of contemporary communities. 30 GEOFF CHILDS

BIBLIOGRAPHY

Aris, M. 1988. Hidden Treasures and Secret Lives: A Study of Pemalingpa (1450- 1521) and the Sixth Dalai Lama (1683-1706). Shimla: Indian Institute of Advanced Studies. Bacot, J. 1912. Le Tibet Révolté: Vers Népémakö, la Terre Promise des Tibétains. Paris: Librairie Hachette. Bailey, F.M. 1957. No Passport to Tibet. London: Rupert Hart-Davis. Bishop, M.H. 1998. Himalayan Herders. Fort Worth: Harcourt Brace. Brettell, C.B. 2000. “Theorizing migration in anthropology: the social construction of networks, identities, communities, and globalscapes.” In C. Brettell and -+ROOL¿HOG HGV Migration Theory: Talking across Disciplines. New York: Routledge, pp.97-136. %UHWWHOO &% DQG -) +ROOL¿HOG ³0LJUDWLRQ WKHRU\ WDONLQJ DFURVV GLV- FLSOLQHV´ ,Q & %UHWWHOO DQG - +ROOL¿HOG HGV Migration Theory: Talking across Disciplines. New York: Routledge, pp.1-26 Cerrutti, M. and D.S. Massey. 2001. On the auspices of female migration from Mexico to the United States. Demography 38(2): 187-200. Chang, Garma C.C. 1977. The Hundred Thousand Songs of Milarepa. Boulder: Shambhala. Childs, G. 1999. Refuge and revitalization: hidden himalayan sanctuaries (sbas-yul) and the preservation of Tibet’s imperial lineage. Acta Orientalia 60: 126-58. —— 2000. Claiming the frontier: a note on the incorporation of Nubri within the borders of Nepal. Studies in Nepali History and Society 5(2): 217-26. —— 2001. A brief history of Nub-ri: ethnic interface, sacred geography, and historical migrations in a Himalayan locality. Zentralasiatische Studien 31: 7-29. —— 2004a. Tibetan Diary: From Birth to Death and Beyond in a Himalayan Valley of Nepal. Berkeley: University of California Press. —— 2004b. Culture change in the name of cultural preservation. Himalaya 24(1-2): 31-42, Special Issue—Tibetan Peoples and Landscapes. —— 2008. Tibetan Transitions: Historical and Contemporary Perspectives on Fertility, Family Planning, and Demographic Change. Leiden: Brill. Clarke, G. 1980. A Helambu history. Journal of the Nepal Research Centre 4: 1-38. De Jong, G.F. 2000. Expectations, gender, and norms in migration decision-making. Population Studies 54(3): 307–19. Ehrhard, F-K. 1997. “A “hidden land” in the Tibetan-Nepalese borderlands.” In A.W. Macdonald (ed.), Mandala and Landscape. New Delhi: D.K. Printworld, pp.335-64. —— 1999a. “The role of the ‘treasure discoverers’ and their search for Himalayan sacred lands.” In T. Huber (ed.), Sacred Spaces and Powerful Places in Tibetan Culture. Dharamsala: Library of Tibetan Works and Archives, pp.227-39. —— 1999b. “Political and ritual aspects of the search for Himalayan sacred lands.” In T. Huber (ed.), Sacred Spaces and Powerful Places in Tibetan Culture. Dharamsala: Library of Tibetan Works and Archives, pp.240-57. Goldstein, M.C. 1971. Serfdom and mobility: an examination of ‘human lease’ in traditional Tibetan society. Journal of Asian Studies 30(3): 521-34. TRANS-HIMALAYAN MIGRATIONS 31

Goldstein, M.C. and C.M. Beall. 1981. Modernization and aging: views from the rural, preindustrial hinterland in Nepal. Human Organization 40(1): 48-56. *RQ]DOH]1/6ROLHQGH)DPLO\RUJDQL]DWLRQLQ¿YHW\SHVRIPLJUDWRU\ZDJH labor. American Anthroplogist 63(6): 1264-80. ²²³&RQÀLFWPLJUDWLRQDQGWKHH[SUHVVLRQRIHWKQLFLW\LQWURGXFWLRQ´,Q N. Gonzalez and C. McCommon (eds.), &RQÀLFW0LJUDWLRQDQGWKH([SUHVVLRQ of Ethnicity. Boulder: Westview Press, pp.1-9. Granovetter, M. 1973. The strength of weak ties. American Journal of Sociology 78: 1360-80. —— 1982. “The strength of weak ties: a network theory revisited.” In P. Marsden and N. Lin (eds.), Social Structure and Network Analysis. Beverly Hills: Sage Publications, pp.105-30. Grieco, E.M. 1998. The effects of migration on the establishment of networks: caste disintegration and reformation among the Indians of Fiji. International Migration Review 32(3): 704-36. Hegmon, M., M.C. Nelson and M.J. Ennes. 2000. Corrugated pottery, technological style, and population movement in the Mimbres Region of the American Southwest. Journal of Anthropological Research 56(2): 217-40. Huber, T. 1997. “Guidebook to Lapchi.” In D. Lopez (ed.), Religions of Tibet in Practice. Princeton: Princeton University Press, pp.120-34. Jackson, D.P. 1978. Notes on the history of Serib and nearby places in the Upper Kali Gandaki. Kailash 6(3): 195-227. Lhalungpa, Lobsang P. (trans.). 1996. The Life of Milarepa. New Delhi: Book Faith India. Lim, Francis Khek Gee. 2004. Zombie slayers in a ‘hidden valley’ (sbas yul): sacred geography and political organisation in the Nepal-Tibet borderland. European Bulletin of Himalayan Research 27: 37-66. Malkki, L.H. 1995. Refugees and exile: from ‘refugee studies’ to the national order of things. Annual Review of Anthropology 24: 495-523. Margolis, M. 1995. Transnationalism and popular culture: the case of Brazilian immigrants in the United States. Journal of Popular Culture 29: 29-41. Massey, D.S. 1990. Social structure, household strategies, and the cumulative causation of migration. Population Index 56(1): 3-26. Massey, D.S., J. Arango, G. Hugo et al. 1993. Theories of international migration: a review and appraisal. Population and Development Review 19(3): 431-66. Oppitz, M. 1968. Geschichte und Sozialordnung der Sherpa. Innsbruck: Universitätsverlag Wagner. Ortner, S. 1989. High Religion: A Cultural and Political History of Sherpa Buddhism. Princeton: Princeton University Press. —— 1999. Life and Death on Mt. Everest: Sherpas and Himalayan Mountaineering. Princeton: Princeton University Press. Pignède, B. 1993 [1966]. The Gurungs. Kathmandu: Ratna Pustak Bhandar. Portes, A. 1995. “Economic sociology and the sociology of immigration: a conceptual overview.” In A. Portes (ed.), The Economic Sociology of Immigration: Essays on Networks, Ethnicity, and Entrepreneurship. New York: Russell Sage Foundation, pp.1-41. Ramble, C. 2008. The Navel of the Demoness: Tibetan Buddhism and Civil Religion in Highland Nepal. Oxford: Oxford University Press. 32 GEOFF CHILDS

Riley, N.E. and R.W. Gardner. 1993. “Migration decisions: the role of gender.” In Internal Migration of Women in Developing Countries. New York: United Nations, pp.195-206. Sardar-Afkhami, H. 1996. An Account of Padma-bkod: A Hidden Land in Southeastern Tibet. Kailash 18(3-4): 1-21. Schaeffer, K.R. 2004. Himalayan Hermitess: The Life of a Buddhist Nun. New York: Oxford University Press. Schuler, S.R. 1987. The Other Side of Polyandry. Boulder: Westview Press. Stark, O. and D.E. Bloom. 1985. The new economics of labor migration. American Economic Review 75: 173-78. van Spengen, W. 2000. Tibetan Border Worlds: A Geohistorical Analysis of Trade and Traders. London: Kegan Paul International. Watkins, J.C. 1996. Spirited Women: Gender, Religion, and Cultural Identity in the Nepal Himalaya. New York: Columbia University Press. Wangmo, Jamyang. 2005. The Lawudo Lama: Stories of Reincarnation from the Mount Everest Region. Somerville: Wisdom Publication. Wilson, T.D. 1998. Weak ties, strong ties: network principles in Mexican migration. Human Organization 57(4): 394-403. WHERE THE WATERS DRY UP – THE PLACE OF ORIGIN IN RAI MYTH AND RITUAL

MARTIN GAENSZLE

If one tries to understand the indigenous view of Rai social organisation in east Nepal and looks at their narrative traditions to see how they recount the history of their subgroups, it becomes evident that two notions are central in the conceptual order: origins and migrations. These two notions epitomise the basic concerns both in the domain of kinship and the domain of myth. Narratives of origin deal with the birth of people, their offspring and the beginning and perpetuation of social practices; narratives of migration deal with descent and diffusion from a homeland, and the ancestral lines and routes of transmission in time and space.1 This can be described and narrativised in terms of both genealogy and movement in geographical space; in other words, the ancestral lineages are mythically mapped onto the landscape. So stories about ‘origins and migrations’ are about the ancestral past, ancestral settlement and ancestral deeds. At the same time, however, they are nothing less than a legitimisation of the establishment of kinship-based groups, culture and territorial control—much as in Malinowski’s well-known charter-theory of myth (Malinowski 1948). Thus the concepts of origin and migration are crucial in Rai thinking about man and the world. Though there are several locations that can be seen as such places of origin, there is one particular place where everything started and which I will call the primeval Place of Origin. As I will try to show, this place is seen as immensely powerful and plays an important role in rituals (including healing rituals). At the same time, however, the place is also associated with danger, risk and death. This ambivalence, I will argue in the following, is no coincidence: rather, it is a crucial feature of origins that they are both revered and feared. The question I want to ask is: how are the places of origin depicted and imagined, and what accounts for their ambivalence? ,ZLOO¿UVWORRNDWWKHLQGLJHQRXVQRWLRQVRIWKHSODFHRIRULJLQDPRQJ the Mewahang Rai (Sankhuwa Sabha District, Nepal), relying on my

1 The spatial aspect of origins is retained in words like ‘provenance’ or German ‘Herkunft’. 34 MARTIN GAENSZLE earlier work on ritual and mythology in this group (Gaenszle 1999a, 1999b, 2002) whose language belongs to the Kiranti family of Tibeto- Burman.2%XW,ZLOODOVRLQFOXGHVRPHPRUHUHFHQW¿QGLQJVDPRQJWKH Puma, another Rai group in eastern Nepal, living in Khotang district, to the south of the Mewahang (Gaenszle et al. 2005).3 Most of the more than two-dozen Rai groups, who speak distinct languages, are rather small, often spreading only over a handful of villages. Nevertheless, there is a remarkable congruence in matters of myth and ritual. Places of origin are important in all the traditions, and one of them stands out.

THE “PLACE OF ORIGIN” IN MEWAHANG RAI MYTHOLOGY

In Mewahang myth and ritual there are several terms for the primeval place of origin, but they all refer to roughly the same place, located down VRXWKLQWKHSODLQV WKHWDUDLUHJLRQ 7KHULWXDOH[SUHVVLRQ0 ZDVѠSNKD ZDKܭSNKDEULQJVRXWQLFHO\WKHLGHDEHKLQG wa means ‘water’, VѠS- is WKHYHUEDOURRWµWRGU\¶DQG kha means ‘place’.4 In other words, it is the place where all water dries up. In traditional Rai thinking there is QRFRQFHSWRIDQRFHDQUDWKHUWKHULYHUVDOOÀRZGRZQWRWKHSODLQVRI QRUWKHUQ,QGLDZKHUHWKH\FRPHWRJHWKHULQFRQÀXHQFHVDQGWKHQGLVDS- pear in the ground. Looking down from the Siwaliks Hills to the Ganges EDVLQWKLVLVDSODXVLEOHLGHD ¿JXUH A second term, NKRZDOXƾDOVRFRQWDLQVWKHURRWIRUZDWHU wa), as ZHOODVWKHURRW OXƾ ‘stone’, ‘rock’.5 This indicates that the place is not simply a sandy basin but includes a rock which may highlight this place as a landmark (see below). :KHUHDVWKHVHWZRWHUPVUHIHUWRDQXQLGHQWL¿DEOHSODFHVRPHZKHUH down south, the name Barachetra (< N. T?PßF?IѪCRP? RU&KDWUDVLJQL¿HV

2)RUDFODVVL¿FDWLRQRI.LUDQWLODQJXDJHVVHH'H/DQFH\7KH.LUDQWLDVDQHWKQLF category include a variety of groups, mainly the numerous Rai groups and the Limbu, but also Yakkha and Sunuwar. 3 The project dealing with Chintang and Puma was funded by the Volkswagen Foundation under the DOBES program (Grant No. II/79 092, 2004-2008, PI B. Bickel). For further information on the “Chintang and Puma Documentation Project” see www.uni-leipzig. de/~ff/cpdp. I am grateful to the funding agency and to my colleagues in Leipzig and Kathmandu who made this work possible. 4 On the particularities of the ritual language see Allen 1978, Gaenszle 2002. In this essay I use M. for Mewahang, P. for Puma and N. for Nepali. In the case of ritual expressions LQ0HZDKDQJRU3XPD,DGGDQDVWHULVN 0 3 7LEHWR%XUPDQURRWVDUHPDUNHGE\D preceding asterisk. 5 The initial morpheme may be derived from < khokma ‘to dig’, ‘to clear’, but the etymology is not entirely clear. WHERE THE WATERS DRY UP 35

DVSHFL¿FSLOJULPDJHVLWHDELWVRXWKRIWKHFRQÀXHQFHRIWKHµ6HYHQ Rivers’ (Saptakoshi).

Figure 2.1. View south on the Koshi River from Chintang village in Dhankuta district (photograph courtesy CPDP).

It is an old WƯUWKD (‘passage’, ‘pilgrimage place’) going back at least to Licchavi times (fourth-seventh century A.D.),6 and though few Rai have ever been there, it is for them more or less identical with the primeval place of origin. As told in the origin myth cited below, it was this place from where important domestic species, the major crops, and mankind came: ...there, at the place called ZDVѠSNKD ZDKѓSNKD, in the PDGKHĞ [plain], ZKHUHYHUWKHZDWHUWKHULYHUVÀRZZKHUHYHUWKH\GU\XSWKHUHEHORZWKH clump of earth bakhapumma came to light. But then the siru reed pierced it - that is, it broke apart -, and rice came into being,... as there is rice only in the PDGKHĞ. Now, there was no one to tend to the rice. What to do? The chicken needed to be created. The chicken was created, but it, too, did not tend to the rice but only ate it! The pig was created, but the pig, too, only ate the rice and did not tend to it. Thereupon livestock - cows and

6 See Bhardwaj 1973. The place is also mentioned in an early account from Francis Buchanan Hamilton; Hamilton 1986 [1819]:152. 36 MARTIN GAENSZLE

cattle - were created, but these, too, only ate the rice, and no one, no one at all tended to it. Finally, though, a clump of ash [M. khapapumma] came down from above - we have all come from ash - came down, and in the end we, the humans arose…7 The basic idea here is that all the good things, the sources of life, have come into being in the south. It is interesting to note that according to another version of the myth, this genesis was only possible after the earth clump was coming down with the river water from above, so one could argue that everything started from above. But this is perhaps a somewhat abstract point, since for the Mewahang the origin is clearly seen as located down in the plains: this is the place which is associated with fertility and wealth. As will be argued below, the movement upward LVVHHQDVSRVLWLYHDQGEHQH¿FLDO$IWHUDOO²WKHJURZWKRISODQWVDQG animals is a process that generally moves upward. In any case, it should be stressed that in most other central Himalayan migration myths, the place of origin is in the north, so the Rai case is unique in this respect.8

SOMNIMA AND PARUHANG EPISODE

Another important episode that takes place down in the plains and brings out the cosmic dimension of this location involves Somnima and Paruhang. Somnima, the primeval mother and major Kiranti goddess, is looking desperately for a husband since there are no men on earth and her father, the wind, is not a suitable mate. Eventually she is given a KLQWDQG¿QGV3DUXKDQJXSLQWKHVN\%XWIDUIURPIDOOLQJLQORYHVKH rejects him because he strikes her as too ugly—he has a huge goitre. The mother, who acts as a kind of marriage broker and intermediary, advises Paruhang to trick her daughter by drying up the world. Then her mother said to Paruhang, “Well, Somnima doesn’t like you; go and make the world dry up [lit. ‘make U?QȡNIF?U?FȔNIF?’]. Leave the best part of you on a NćѢßJu leaf” - this urine of his, right? – “leave on a NćѢßJS leaf.” Thereupon Paruhang returned up to the sky. He made all the water dry up. Somnima almost died of thirst.9 In this way, Somnima is forced to drink the only liquid left in the world, Paruhang’s semen, and from this she becomes pregnant and eventually gives birth to the whole spectrum of species: thorny creeper, two kinds of 7 The full myth is given in Gaenszle 2000:233. 8 Compare, for example, the cases described by Blackburn, Huber, and Post in this volume, as well as those of the Gurung by Pettigrew 1999, or the Tamang by Höfer 1999. 9 Gaenszle 2000:236. WHERE THE WATERS DRY UP 37 bamboo—called Lalahang and Pakpahang, who later combine to become ERZDQGDUURZ²DVZHOODV7LJHU%HDU0RQNH\DQG²¿QDOO\²0DQ Thus, again, we have a contrast between water and drought, fertility and death. This contrast is linked to the vertical dimension: Paruhang is associated with the sky and water, and Somnima is associated with the earth and potential drought. The basic message is clear: terrestrial life originates only through the vitalising drop from above: only if the two forces, male and female, combine in creative union can fertile reproduc- tion occur.10

TIGER AND BEAR EPISODE

Eventually Somnima’s offspring begin to separate. Tiger wants to leave for the jungle and presents his mother some meat as a good-bye gift, telling her to keep her distance. But she, overcome by maternal affec- tion, disobeys him, and Tiger eventually kills his mother, who was again pregnant at the time. Then a fraternal quarrel ensues. Tiger orders his younger brother, Bear, to bury their mother, but the latter, lacking any ‘civilised’ behaviour, eats up her body instead. This scandalous action has to be stopped, and the two kinds of bamboo combine as bow and arrow and shoot the evil-doers. This inaugurates the division between wild animals and Man. But, above all, it is the origin of the different kinds of death spirits: There then arose three species (thari N): from Tiger (and Bear) the hillasi M, from their mother ma:maksi M, from her insides - the fetus - cha:nu M. The ma:maksi, the hillasi and the cha:nu thus came into being through Somnima; these evil beings have been around since that time...11 Here, too, the origin of death spirits is located down in the plains. Not only all the good things come from there, also certain dangerous beings originate at this place. It is due to this idea that in certain rituals the spirits are taken down to their original place. For example, the Ma:mangme FHUHPRQ\SHUIRUPHGIRUWKHEHQH¿WRIDSUHJQDQWZRPDQZKRVXIIHUV 10 As Alexander Macdonald writes in connection with the fertility aspect of megaliths: “…for [the populations in question] megaliths are linked with the periodic return of fertility to the world. In monsoon Asia, this fertility is, to some extent, considered everywhere, and with sound reason, as the result of the interaction between the rain and the earth, that is to say, of a harmonious relationship between the sky, the earth, and the subterranean world, between the living and the dead, between the celestial and earthy waters. In the last analysis, many religions envisage the periodic recreation of spring in terms of the human sexual act”; Macdonald 1984:20. This might be one explanation for the reference to a rock in the name of the place of origin. 11 Gaenszle 2000:243. 38 MARTIN GAENSZLE from pains and other disturbances, enacts a ritual journey down to the Place of Origin in order to take back the ma:maksi spirits who trouble the expecting mother.12

Figure 2.2. Ma:mangme journey.

The route of this travel follows the well-known path leading south. The ULWXDOH[SHUW¿UVWHQFLUFOHVWKHYLOODJHWHUULWRU\LQWKH6DQNKXZD9DOOH\ 12 Gaenszle 1994. WHERE THE WATERS DRY UP 39 and then proceeds down along the Arun River, past Tumlingtar,13 all WKHZD\WRWKHFRQÀXHQFHRI$UXQ6XQ.RVKLDQG7DPXU.RVKLDQG then to Chatra (or Barachetra). From there the route continues further south to Khowalung, which in ritual speech is also named 1MKLGQGK? 7?QGRCL—“the place where Somnima died”. Interestingly, the tarai is also referred to as @?HSFMRF?PSFM / @?HSFMRCLRF?PSFMRCL (M. RCL ‘village’), that is, the place where the Tharu live. In Rai imagination this place in the plains is rich for its extensive rice cultivation. This path can EHVHHQRQ¿JXUH

ANCESTRAL MIGRATION FROM KHOWALUƾ

7KHQRWLRQRINKRZDOXƾZKLFKLVPRVWO\XVHGLQRUGLQDU\VSHHFKLVDOVR of great importance in the context of early ancestral migration. After all the various species and eventually First Man originated, and after man became a cultural being, the descendants began to migrate towards the north and split up by walking uphill along different rivers. So the place of origin is also the place from where the ancestors began their search for a place to settle. This is recounted in the wide-spread Four Brother myth.14 In the Mewahang version of the story, these four brothers were Khambuhang, Mewahang, Limbuhang and Meche-Koche. Soon after leaving the place of origin, the youngest brother got left behind and eventually stayed in the plains. The descendents of this brother became the indigenous inhabitants of the tarai—the Meche (or Bodo), Koch (or today, Rajbamsi), Dhimal and also, for some the Tharu.15 After some quarrelling when crossing the river at Chatra, the eldest brother went up the Dudh Koshi, the second eldest brother went up the Arun Koshi, and the third brother settled up in the Tamur Koshi. The myth describes how the brothers travelled as individuals— along with their sisters, it is said—not as groups. After settling down, however, they became social groups (the present ‘subtribes’), and then divided up further into clans and lineages. This account of ancestral migration can be read as a story of descent DQG¿VVLRQPDSSHGRQWRWKHODQGVFDSH16 I call this the ‘drainage model’ RIJHQHDORJ\ ¿JXUH LWLVDVSDWLDOLVHGPRGHORIGHVFHQWLQZKLFKWKH

13 The present day airport town. 14 The story of four original brothers is found among most Rai groups, though their names usually differ. 15 Whereas all the other groups speak Tibeto-Burman languages, the Tharu speak an Indo- Aryan tongue. The Koche are related to the former and once important kingdom of Kocch. 16 Such stories are common across the Himalaya as well as among tribal groups in Central India. 40 MARTIN GAENSZLE focus is not blood relationships but rather links to a particular territory. The drainage model is similar to the tree model (German ‘Stammbaum’) in that it is characterised by natural divisions and subdivisions. But whereas the tree or root metaphor implies a direction of growth, the drainage model in fact has to be read “against the grain”: the migration fol- ORZVQRWWKHÀRZRIWKHZDWHUEXWWKHUHYHUVHGLUHFWLRQ+HUHLWVKRXOGEH remembered that the original earth clump was swept down from above, so in a way the migration can also be seen as a return—a return to the source.

RITUAL AS RETURN TO THE PLACE OF ORIGIN

Figure 2.3. The Kiranti drainage model of genealogy.

Coming back to the issue of ritual, it becomes apparent that the journey down to the place of origin is a “return into the myth” to use the words of András Höfer (1999:222).17 That is to say, in the Ma:mangme ritual, the priest goes back in both space and time, back into the times when Somnima was killed by her offspring. The interesting thing here is that this vertical symbolism of upward/ downwards is not only typical for rituals dealing with these death spirits (like Ma:mangme) but is in fact characteristic of most rituals performed for healing purposes. What is crucial about these rituals is the verticality enacted in the movements up 17 Höfer refers to the ritual journeys among the Tamang in which the shaman returns to Uiseme (< Tib. dBus-kyi bSam-yas), the mythic site of Tamang ethnogenesis (1999:222). WHERE THE WATERS DRY UP 41 and down the rivers and valleys. For example, the VLUXWKƗXQH ritual (M. saya po:kma), an important rite for strengthening the “head-soul” (well- being, power), also contains references to the various places of origin.

58 KGL?NM IF?&CIMK@? man rise-place (name) The place of origin of Man is called Hekomba, 59 NMU?NM IF?0ȡNAFCJCL chicken rise-place (name) the place of origin of the Chicken is called Rupchelen, 60 AFM@?NM IF? ?IQCJCL pig rise-place (name) the place of origin of the Pig is called Bakselen, 61 F?BGNM IF?2FSL?NR?ĩ cow rise-place (name) the place of origin of the Cow is called 2FSL?NR?ĩ 18

Here a further differentiation of names can be observed: all of these places are located down in Khowalung, but each species has its own particular place where it came into being. It is interesting to note that VHYHUDORIWKHQDPHVFRQWDLQWKHVXI¿[±len, which is derived from the verb lemmaµWRFRPHRXW¶7KLVVXI¿[VXJJHVWVWKHLPDJHRIDQRSHQLQJ in the ground from which all the various beings originated. In the course of the Q?W?NMIK?ULWXDODGLYLQDWLRQLVKHOGZLWKWKHVDFUL¿FHRIWZR FKLFNHQV$IWHUDEORZWRWKHQHFNHLWKHUEORRGÀRZVIURPWKHEHDN (which is a good sign) or it does not (which is a bad sign). In the latter case, the chicken is said in the ritual idiom to have “taken the blood back to U?QȡNIF?U?FȔNIF?” in other words, the “dried up blood” is associated with the place of origin “where all the waters dry up”. Here it shows clearly that this place is not auspicious, but dangerous. For ordinary humans it is a place to move away from, and only movement upward is what gives life. Here a certain ambivalence of the place of origin emerges: it is a place where one has to go in order to take evil spirits back to “where they come from”. Among some other Rai groups, it is also a place to which one leads the ordinary death spirits.19 But it is also a place where one can

18 Gaenszle 2002:263f. 19 This is, for example, the case among the Thulung Rai where the place of origin is called bebdu pabdu, the ‘place of the grandfathers and fathers’; Allen 1974:7. This place is located to the south-west of the settlement area. Another case are the Yamphu Rai of the upper Arun Valley described by Forbes 1998. Here, however, the place of the dead, called Lalu Tembe, is in the north, near Lhasa in Tibet. It is to this place that the soul of the dead is guided by the 42 MARTIN GAENSZLE

‘tap’ the powers of origin. However, this positive effect is not achieved by reaching the place; rather it is the return movement upwards that strengthens and increases health and fertility. It is the movement away from the origins that, as a mimetic and directional gesture in the vertical dimension, brings about a form of empowerment. Not only in the saya po:kma ritual but also in other ancestral rituals the performance is said to ‘raise the head soul’ by invoking a movement upwards. Bad things, on the other hand, are condemned to move downwards. Linguistic analysis also indicates that the vertical dimension is of cen- WUDOLPSRUWDQFH,WLVHYHQJUDPPDWLFDOLVHGLQWKHORFDWLYHVXI¿[V\VWHP that distinguishes three levels: up /neutral /down.20 A crucial question is: If the vertical dimension is so fundamental in grammar, is it of similar importance in the ritual symbolism? The examples discussed so far clearly indicate that ‘up-down’ movements are a basic feature of exor- cising and healing rituals. The following example from a neighbouring 5DLJURXSLVRILQWHUHVWDVLWSRLQWVWRDPRUHJHQHUDOVLJQL¿FDQFHRIWKH vertical axis.

PUMA GENRE OF HOPMACHAM

7KDWWKHSODFHRIRULJLQLVLPSRUWDQWQRWRQO\LQFDVHVRIDIÀLFWLRQDQG ritual therapy can be seen in another example, the Hopmacham songs among the Puma Rai (Khotang District). Among several genres of ritual speech, this is the most highly valued genre and seen to epitomise the power of the entire oral tradition.21 The Hopmacham is not a shamanic genre, and it is not sung for the purpose of healing. Yet it is regarded as SRVVHVVLQJDVSHFLDONLQGRIHI¿FDF\DNLQGRIPDJLF,WLVFRPPRQO\ explained, for example, that it can bring rain in the case of drought, it can bring life to dead trees, it can charm animals and trees, it can even OLJKW¿UHDQGDERYHDOOLWFDQHYRNHLQWHQVHHPRWLRQVERWKVDGQHVVDQG happiness, in the listeners. But at the same time, it is seen as dangerous for the performer because, if not sung properly, with adequate competence, the singer can become ill—or even die. :KDWLVLWWKDWPDNHVWKLVVRQJVRSRZHUIXO",WVSDUWLFXODUHI¿FDF\LV expressed in the name of the genre. &MNK?AF?K is derived from the verb FMNK?, ‘to drink,’ and the noun AF?K, ‘song’. So literally, it means ‘drinking song’, but this should not be understood as a song for entertainment and psychopomp, whereas death pollution is discarded in the south, in Kashi (Forbes 1998:115). 20,KDYHGHVFULEHGWKHVLJQL¿FDQFHRIWKLVDVSHFWLQULWXDOVSHHFKLQ*DHQV]OHD 21 The following account is based on Rai et al. 2008. WHERE THE WATERS DRY UP 43 merrymaking—though the drinking of beer is an essential part of ancestral rituals. Rather, it was explained by performers that it is the song which ine- briates and tends to make you drunk. It is sung on various occasions, not as a central ritual performance, but more as a special treat. For example, it is sung at the end of Spring Rituals (Sakewa), at Ubhauli/ Udhauli (Manggen) and in the context of weddings. One might say that whenever there is a traditional event in which the presence of the ancestors is invoked and there is a sense of community with the ancestral beings, the song highlights this state of mind. The Hopmacham songs invoke the ancestors, in particular the creator couple Somnima and Paruhang, and they refer to the local landscape, WKHZDWHUWKHWUHHVWKHÀRZHUV7KHFUXFLDOSRLQWKRZHYHULVWKDWLWLV also about the origins, and it includes a verbal journey to Khokwalung. In the following, I will present a few revealing lines of a text recorded in the Puma area. After the initial invocations in the house, where the journey starts from the hearth, the singer states the aim of his undertaking:22 U?JCLK? GÏU?QCNIF?U?K?IF? G *RLQJ GRZQZKHUHWKHZDWHUVÀRZZKHUHWKHZDWHUVDSSHDUZKHUHWKH waters disappear The term (U?JCLK?U?QCNIF?U?K?IF?) used to denote the place of origin is very similar to that described above for the Mewahang (U?QȡNIF? U?FȔNIF? ,W LV D WULQRPLDO DQG FRQ¿UPV RXU DQDO\VLV FRQFHUQLQJ LWV PHDQLQJLQWKH¿UVWWHUPwa ‘water’ is linked to lenma µWRÀRZ¶ LQWKH second term it is linked with QCNK? (‘to originate’) and in the third term it is linked with mama (mas-) ‘to get lost’. So here, it is also the place ZKHUHWKHZDWHUVµJHWORVW¶%XWLQWHUHVWLQJO\LWLVDOVRDWHUPWKDWUHÀHFWV WKHSURFHVVRIÀRZLQJDQGWKHQKLJKOLJKWVWKHFRQWUDVWRIZDWHUEHLQJYLV- ible and then disappearing. It is a highly poetic image, which plays with an ambivalent state of both being and non-being. Thus again ambivalence emerges as a crucial feature of the place of origin below. But before reaching the place of origin, the singer follows the course RIWKHULYHUDQGSDVVHVE\WKHFRQÀXHQFHVRIVPDOOHUULYHUV,QIDFWDW some point the text even refers to a boat that is used for the journey 3 L?U?BSĩACNF?BSĩ ‘boat’): Q?I?U?RSNKCPGU?RSN J?K C through the meeting place of pure waters, of sacred waters, hey

22 The quotes are taken from the session “hopmacham_01” (line 079b, 093, 117, 131, 132, 153). 44 MARTIN GAENSZLE

Eventually the singer reaches the Place of Origin, here called Khokwalung Manalung: IFMIU?JSĩK?L?JSĩGJ?KUCLĩ?ĩ? through down in Khokwalung Manalung I proceed :KLOH SURFHHGLQJ WKH VLQJHU DGGUHVVHV WKH DQFHVWRUV 3 K?K?LG ASLSLG) and requests his tutelary divinity (‘dream king’) to keep away enemies who might block the path. Thus, with the help of ancestral spir- LWVKHWUDYHOVWKHDQFHVWUDOZD\ 3 L?U?PGQSLRSKPG): L?U?PGQSLRSKPGJ?K I am proceeding on the way of the ancestors 7KHSODFHRIRULJLQLVKHUHHTXDWHGZLWKWKHFRQÀXHQFHRIWKHµ6HYHQ 5LYHUV¶ 6DSWDNRVKL ZKLFKLVWKHXOWLPDWHFRQÀXHQFHRIFRQÀXHQFHV $QRWKHUSODFHQDPHXVHGLQWKLVFRQWH[WLV3 A?R?P?U?B?P?, which refers to the famous pilgrimage site, Barachetra (see above). ?PSLRSNR?KMPRSNBSBIMQGRSNQ?I?U?KCPGU?RSN G DWWKHFRQÀXHQFHRIWKH$UXQWKH7DPXUWKH'XGK.RVLDWWKHSXUHZDWHU sacred water, A?R?P? U?BCP? J?K C A?R?P?G U?BCP?G C L?U?BSĩ ACNF?BSĩJ?K J?K UCLĩ?ĩ? By the way of Catara Wadara, I proceed through Catara Wadara on boat After voicing a number of requests, the singer returns, reversing the order of place names and locations he has recited. He eventually reaches the hearth of the house, from where he began his travel. Thus the journey achieves a link between the domestic hearth (P VXSWXOXƾ) and the Place RI2ULJLQ 3 NKRNZDOXƾPDQDOXƾ). This indicates one explanation for WKHFRPPRQURRW OXƾ ‘stone’ in the two terms: they are identical.23 By ritually linking the hearth, the seat of the ancestors, with their origin, the song revitalises the ancestral order.

CONCLUSION

To come back to my initial question concerning the ambivalence of the Place of Origin, why is it dangerous to go to the Place of Origin? Why is this link with origins associated with risk and requires extraordinary

23$QRWKHUSRVVLEOHUHDVRQIRUWKHVLJQL¿FDQFHRIDVWRQHRUURFNZDVLQGLFDWHGHDUOLHU VHH IRRWQRWH WKH ZLGHVSUHDG RFFXUUHQFH RI PHJDOLWKV 2QH GRHV QRW ¿QG WKHP LQ DOO Rai traditions, but in the Mewahang case, for example, each village has a VDNKHZDOXƾ representing the territorial spirit; Gaenszle 2010. WHERE THE WATERS DRY UP 45 priestly competence? I should stress again that the Place of Origin is not simply a homeland, a place of longing and belonging, but is marked by a fundamental emotional ambivalence. On the one hand, it is a place of power, where one can acquire strength and fertility, tapping the power of ancestral origins. But, on the other hand, it is a dangerous place which only ritual experts can reach safely, and even they should leave again soon. One doesn’t really want to stay there, and it is above all the movement away which has a positive effect in the ritual performance. This ambivalence of origins was perceptively brought out by Klaus Heinrich, a philosopher of religions.24 He regards genealogy—or genealogical derivation—as the central function of myth. Maintaining a genealogical chain is crucial for any identity, but a fundamental gap cannot or must not be avoided. Heinrich wrote: To derive from the origins means on the one hand to come from the origin, to carry on the power of origin. On the other hand it means to cut oneself off from the origin, to have escaped the origin. The escapee acquires independence from the origin—up to the point of fear and helplessness to lose the backing of the origin. (Heinrich 1966:15) [my translation] He calls this the dialectics of origins: it is important to have a link to the origin, but escape from it gives autonomy. And I think it is this inherent tension that accounts for its dangers. Ritual recitations are not simply a paean or praise of the origins; they are also a call to remember the gap, the distance—both geographical and existential—between oneself and the origin. After all, there are good reasons why one has moved away. The memory of the past is crucial for one’s self-identity, but the place of origin is no paradise, it is no place for living. Quite the opposite, it is a place associated with death. Though the variety of species and the sources of making a living—hunting as well as domestic animals and domestic crops—all originated down in the plains, it was also death which came into being there. And with death the natural cycle of growth and decay, of sowing and harvesting, in short: the basic constraints of life also originated. Through ritual the Rai commemorate these events and bring back to mind the creative powers of the origins. But the same ULWXDOVDOVRPDNHWUDQVSDUHQWWKH³¿VVXUH´ LQ+HLQULFK¶VZRUGV RUWKH existential displacement which characterises the human condition.

24 Heinrich was teaching at the Freie Universität at Berlin from 1971-1995. 46 MARTIN GAENSZLE

BIBLIOGRAPHY

$OOHQ1-³7KHULWXDOMRXUQH\DSDWWHUQXQGHUO\LQJFHUWDLQ1HSDOHVHULWXDOV޵ In C. von Fürer-Haimendorf (ed.), Contributions to the Anthropology of Nepal. Warminster: Aris and Philipps, pp.6-22. —— 1978. Sewala Puja Bintila Puja: Notes on Thulung Ritual Language. Kailash 6(4): 237-56. Bhardwaj, S.M. 1973. Hindu Places of Pilgrimage in India. Berkeley: University of California Press. DeLancey, S. 1990. “Sino-Tibetan languages.” In B. Comrie (ed.), The world’s major languages. New York, Oxford: Oxford University Press, pp.797-810. Forbes, A.A. 1998. “Sacred geography on the cultural borders of Tibet.” In A.-M. Blondeau (ed.), Tibetan Mountain Deities: Their Cults and Representations. Wien: Verlag der Österreichischen Akademie der Wissenschaften, pp.111-121. Gaenszle, M. 1994. “Journey to the origin: a root metaphor in a Mewahang Rai healing ritual.” In M. Allen (ed.), The anthropology of Nepal: peoples, problems, processes. Kathmandu: Mandala Book Point, pp.256-68. —— 1999a. “Travelling up and travelling down: the vertical dimension in Mewahang Rai ritual texts.” In B. Bickel and M. Gaenszle (eds.), Himalayan space: cultural horizons and practices. Zürich: Völkerkundemuseum der Universität Zürich, pp.135-63. —— 1999b. Gesungene Reisen zum Ursprung: Mytho-rituelle Topographien im Himalaya. Zeitschrift für Religionswissenschaft 7: 171-85. —— 2000. Origins and migrations: kinship, mythology and ethnic identity among the Mewahang Rai of East Nepal. Kathmandu: Mandala Book Point & The Mountain Institute. —— 2002. Ancestral voices: oral ritual texts and their social contexts among the Me-wahang Rai in east Nepal. Münster, Hamburg, London: LIT Verlag. —— 2010. “Political aspects of the territorial cult among the Mewahang Rai.” In M. Lecomte-Tilouine (ed.), Nature, Culture and religion. At the Crossroads of Asia. Delhi: Social Science Press, pp.190-221. Gaenszle, M., B. Bickel, G. Banjade, et al. 2005. “Worshiping the king god: a preliminary analysis of Chintang ritual language in the invocation of Rajdeo.” In Yogendra Yadav, Govinda Bhattarai, Ram Raj Lohani (eds.), Contemporary issues in Nepalese linguistics. Kathmandu: Linguistic Society of Nepal, pp.33-47. Hamilton, F.B. 1986 [1819]. An Account of the Kingdom of Nepal and of the territories annexed to this dominion by the house of Gorkha. New Delhi: Asian Educational Service (Reprint). Heinrich, K. 1966. “Die Funktion der Genealogie im Mythos.” In Parmenides und Jona. Vier Studien über das Verhältnis von Philosophie und Mythologie. Frankfurt: Suhrkamp, pp.9-28. Höfer, A. 1999. “Nomen est numen: preliminary notes on the verbal journey in some Western Tamang oral ritual texts.” In B. Bickel and M. Gaenszle (eds.), Himalayan space: cultural horizons and practices. Zürich: Völkerkunde- museum der Universität Zürich, pp.205-44. WHERE THE WATERS DRY UP 47

Macdonald, A.W. 1984 [1966]. “A note on Tibetan megaliths.” In Essays on the Ethnology of Nepal and South Asia. Kathmandu: Ratna Pustak Bhandar, pp.15-25. Malinowski, B. 1948. Magic, science and religion; and other essays. New York: The Free Press. Pettigrew, J. 1999. “Parallel geographies: ritual and political values of a shamanic soul journey.” In B. Bickel and M. Gaenszle (eds.). Himalayan space: cultural horizons and practices. Zürich: Völkerkundemuseum der Universität Zürich, pp.247-70. Rai, Shree Kumar, Arjun Rai, Vishnu S. Rai, Narayan P. Gautam, Martin Gaenszle, and Balthasar Bickel. 2008. “A comparative study of the Hopmacham tradition and ordinary songs in Puma”, paper presented at Himalayan Language Symposium 2007. Shimla (revised version).

WHERE DID THE QUESTION ‘WHERE DID MY TRIBE COME FROM?’ COME FROM?1

ROBBINS BURLING

,¿UVWYLVLWHGQRUWKHDVWHUQ,QGLDLQWKH¶VZKHUHDVD\RXQJDQWKUR- pologist, I lived among the Garos in what is now the state of Meghalaya, south of the Brahmaputra valley. My research had nothing to do with migration, but the topic was raised by Garos, now and then, most often in the form of a question: ‘Where did the Garos come from?’ By this, the questioners meant: ‘Where did the Garos live before they migrated to their present homeland? Where was their earlier homeland?’ People in other parts of the north-eastern hills ask parallel questions about the origin of their own tribes. 7KLV TXHVWLRQ KDV DOZD\V SX]]OHG PH ,W VHHPHG LQ WKH ¿IWLHV DV it seems to me now, exactly like asking ‘Where did the French come from?’ We know enough about the history of Western Europe to know that people have migrated into what is now France from every direction—from Rome with Caesar, from Scandinavia as Vikings, from Germany, from England, from Wales, and surely from many other places. The arrivals all contributed both to French culture and to the French gene pool, but they were not French before they arrived. French ethnicity, French nationality, and French culture all had their origins right where the French live now. Even the French language developed only in France. Roman soldiers brought their language, just as they brought their genes, but it was in France where Latin grew into French. This all seems so obvious, that to ask “Where did the French come from?”, as if the French had lived somewhere else before arriving at their present homeland, would be ridiculous. I believe that it makes no more sense to ask: “Where did the Garos (or any other north-eastern tribe) come

1 I am much indebted to my friends and colleagues, Thomas Trautmann and Thomas Toon, for many long and stimulating discussions about my ideas of migration and ethnicity. Among PDQ\RWKHUWKLQJVWKH¿UVW7RPOHGPHWRWKHZRQGHUIXOZULWLQJVRI6LU:LOOLDP-RQHV The second Tom showed me that it was no longer necessary to imagine that Anglo-Saxon migrations brought serious population displacement. I am also indebted to many friends who have lived and worked in northeast India and with whom I have debated these matters. In SDUWLFXODU0DUN3RVWE\KLVIULHQGO\EXW¿UPGLVDJUHHPHQWKDVKHOSHGPHJUHDWO\WRFODULI\ P\RZQWKLQNLQJ,WZDV5XDWD5HQJVLZKR¿UVWWROGPHDERXWWKHUHPDUNDEOHVWRU\RIWKH Mizo Jews. 50 ROBBINS BURLING from?” than to ask: “Where did the French come from?” If the question seems more reasonable when we ask it about a north-eastern tribe, that is only because we know so little about the actual history of those tribes WKDWZHUHVRUWWRP\WKVWR¿OOLQWKHJDSVLQRXUNQRZOHGJH In the following pages I will express great scepticism about migration stories, but I never mean to imply that I am sceptical about migration itself. Indeed, people migrate all the time and in every direction—back and forth and up and down—and surely people in north-eastern India, like people everywhere else, have been migrating for thousands of years. Just as people have migrated into (and also out of) what is now France, I presume that people have migrated from every direction into (and out of) what is now the Garo Hills. My scepticism is not about the migration of people, which goes on all the time, but about the migration of tribes and about the north-eastern rhetoric of migration that is expressed in terms of tribes. I do not believe that tribes very often pick themselves up and move to a new location. People migrate. Small groups migrate. Tribes, I believe, rarely do. If “Where did the Garos (or any other tribe) come from?” is not a reasonable question, we ought to ask a different question—a question about the question: Where did the question come from? Did it come from the local people? Is it their curiosity about their own origins that leads them to ask the question? Or is it a question that western visitors brought with them? It does not have to be just one or the other, and quite likely it is both, but I will emphasise the western roots of the question because I believe that side of it has been rather badly neglected. &RORQLDORI¿FHUVLQQRUWKHDVWHUQ,QGLDZHUHFHUWDLQO\LQWHUHVWHGLQ the origins of the people they administered, and they certainly looked for migrations. Between about 1910 and 1940 British administrators who had lived among the north-eastern tribes wrote and published a series of ethnographies, each dealing with a single tribe. The tables of contents of these books follow a standard format, and each book has, in LWV¿UVWFKDSWHUDGLVFXVVLRQRIWKHPLJUDWLRQVWKDWEURXJKWWKHWULEHWR its present location. J. H. Hutton gave a particularly detailed account of migration in The Angami Nagas (1921), and unlike other authors, he did not restrict himself to just one tribe but considered other Naga tribes as well. He wrote: “The history of how the Naga tribes came precisely to occupy their present position has, of course, passed into the dim obscurity of vague traditions. But enough of them remain to give some indication of the course which the migrations took” (1921:6). He goes on to describe the WHERE DID THE QUESTION 51 movements of the Angamis, and of several other tribes, in some detail, DQGKHHYHQSURYLGHVDPDS VHH:HWWVWHLQ¿JXUHLQWKLVYROXPH with variously coloured and variously directed arrows crisscrossing each other. The arrows show the routes of eleven Naga tribes, from the Pom and Konyak in the north to the Tangkhul and Kacha Naga in the south, as they are supposed to have migrated into and round about the hills. These paths, the map proclaims, are shown ‘according to their own traditions’, but Hutton clearly gives these traditions considerable credence. He discusses the Angami in the greatest detail, and he suggests that their knowledge of wet rice agriculture might have been acquired during “a sojourn in the lowlands of Imphal [Manipur] as they migrated north”. However, he then goes on to say, “Where the Angamis came from before WKH\UHDFKHGWKHFRXQWU\QHDU0DQLSXULVDPXFKPRUHGLI¿FXOWSUREOHP and one quite beyond the scope of this book” (1921:8). The presumption behind these words is that the Angami passed, as a group, through Manipur, and even if we cannot know exactly where they came from before that, we can be certain that they came from somewhere. J. P. Mills, in his book The Ao Nagas (1926), begins his discussion RI PLJUDWLRQV E\ VD\LQJ ³$R WUDGLWLRQ VWDWHV TXLWH GH¿QLWHO\ WKDW WKH ancestors of the tribe came out of the earth at Lungterok” (Mills 1926:6). Presumably Mills took that to be a myth rather than history, but he goes on to describe traditions of more recent events, including their arrival at their present homeland, and he treats these as more historical. He says: “I have been at pains to collect all the traditional information possible as to the people whom the Aos found in possession of their present country when they invaded it. These stories give us some of our very rare glimpses of the early history of the hills, and may help to throw welcome light on the complicated question of the origin and composition of the Naga tribes as we know of them today” (Mills 1926:8). Mills is very careful as he evaluates the weak evidence, but he clearly takes the migration of tribes for granted. The question was not ‘Did they migrate?’ but rather ‘How and from where did they migrate?’ Oddly, Mills also says: Naga invaders do not as a rule obliterate their foes. More usually, after reducing the village which is their objective to a suitable frame of mind by repeated raids, they come and live in it as overlords, take wives from it, and gradually absorb it into their own community (Mills 1926:8-9). This is an unusual acknowledgement of tribal mixture, and it recognises the failure of one tribe to completely displace another. In some way that Mills leaves unclear, the resulting mixed population is still considered 52 ROBBINS BURLING to be ‘Ao’, and it is still regarded as having migrated in, in spite of WKHIDFWWKDWDVXEVWDQWLDODOWKRXJKXQVSHFL¿HGSURSRUWLRQRIWKHSUHV- ent population’s ancestors had lived there before the migrations. Even by this colonial account, then, people have mixed ancestry, but this is not allowed to disturb the presumption of ethnic continuity. The mixed descendants are still Ao, and they migrated from somewhere else. In his book, The Garos (1909), A. Playfair is even more cautious about accepting traditions of migration. He gives a long and detailed story about migration from Tibet, but then admits serious scepticism about its historical veracity. He says that there is no way the people could have had any real memory of such an ancient migration, and he FRQFOXGHVµ,WLVGLI¿FXOWWRSODFHDQ\UHOLDQFHRQDOHJHQGZKLFKKDV been handed down by word of mouth from generation to generation’ (Playfair 1909:14). Nevertheless, he also suggests that some Garo customs, particularly the ceremonial use of yak tails, point to ties to the north, and he says: …the coincidence of a similar belief existing in Bhutan and on this side of the Himalayas, which is further supported by evidence of language, points to the possibility that in bygone ages the ancestors of the Garos and of the many tribes with which they are closely allied, did cross the Himalayas and settle in the plains at their foot (Playfair 1909:14). 6RHYHQZKLOHGLVPLVVLQJWKHVWRU\DVP\WKKH¿QGVRWKHUJURXQGVIRU suspecting migration. Hutton, Mills and Playfair were all commendably cautious about accepting the evidence for migrations at face value. They were all WKRXJKWIXODQGMXGLFLRXV%XWEHKLQGWKHLUFDXWLRQWKH\VKDUHGD¿UP SUHVXPSWLRQ WKDW SHRSOH PLJUDWH LQ VXI¿FLHQWO\ ODUJH DQG FRKHUHQW groups to make it reasonable to ask where a tribe, and not just individual wanderers, came from. They were not sceptical about tribal migration, but only about the particular stories that they were told. Did these men, who were writing almost a century ago, get their enthusiasm for migrations from the people they described? Or did the people whom they were describing have their interest aroused by the outsiders? Who persuaded whom to believe that tribes migrate? We have earlier and clearer evidence for the western interest in the migration of peoples in the Indian subcontinent. The interest goes back at least as far as the 1780’s when Sir William Jones was a magistrate in the British administration in Calcutta. Today, Jones is remembered primarily for his brilliant scholarship. He was one of the founders of The Asiatic Society of Bengal and was, for some years, its president and WHERE DID THE QUESTION 53 leading light. Linguists remember Jones most of all for his recognition of the relationship of Sanskrit to Latin and Greek. He described the relationship among these languages in strikingly modern terms, asserting, LQWKHVHQRZIDPRXVZRUGVWKDWWKHLUDI¿QLWLHVZHUHµVRVWURQJLQGHHG that no philologer could examine them all three, without believing them to have sprung from a common source, which, perhaps, no longer exists’ (Jones 1993 [1807]:34). With this sentence, Jones recognised what we now call the ‘Indo-European’ family of languages and thereby laid the groundwork for the comparative linguistics of the ninetheenth century. /LQJXLVWVDUHDSWWRSODFH-RQHVXSWKHUHLQWKH¿UPDPHQWRILPPRUWDOV along with the likes of Panini and Chomsky. What linguists do not usually realise is that Jones’s interest in Indo- European was only a small part of a much broader scholarly program. What Jones wanted to do was to group the peoples of Asia into ‘nations’. These were not ‘nations’ in the modern sense of ‘nation-states’, but more DEVWUDFWO\GH¿QHGJURXSVRISHRSOHZKRKDGFRPPRQRULJLQVDVKDUHG history, similar talents, and a common character. Language was only one of four areas that Jones believed could help him to group peoples into nations. He said: ‘…we seem to possess only four general media of satisfying our curiosity concerning [their ancient civil history]; QDPHO\¿UVWWKHLULanguages and Letters; secondly, their Philosophy and Religion; thirdly, the actual remains of their old Sculpture and Architecture; and fourthly, the written memorials of their Sciences and Arts’ (Jones 1993:32). Jones was avidly interested in all four of these ‘media’, not just in language. He used the evidence of the media to group the peoples of Asia (and sometimes those from other continents as well) into a few great ‘nations’. He listed the nations in 1786, in a lecture to The Asiatic Society. This is NQRZQDVµ7KH7KLUG$QQLYHUVDU\'LVFRXUVH¶EXWLWZDVWKH¿UVWLQDVHULHV of seven related lectures concerned with the nations. He said, ‘The ¿YH principal nations, who have in different ages divided among themselves, as a kind of inheritance, the vast continent of Asia, with the many islands depending on it, are the Indians, the Chinese, the Tartars, the Arabs, and the Persians: who they severally were, whence, and when they came, where they now are settled… will be shown, I trust, in ¿YH distinct essays; the last of which will demonstrate the connection or diversity between them, and solve the great problem, whether they had any common origin, and whether that origin was the same, which we generally ascribe to them’ (Jones 1993:27-28, all italics as in the original). This initial discourse was GHYRWHGWRWKH+LQGXVDQGIURPWKH¿UVWLQKDUPRQ\ZLWKKLVUHFRJQLWLRQ 54 ROBBINS BURLING of the relationship of Sanskrit to Latin and Greek, Jones joined most Europeans with the Hindus into one great ‘nation’. This was a grouping of peoples, not just of languages, but the language connection gave him vital evidence about the relationship among the people. Each of the other nations was the subject of a later discourse. -RQHV SUHVHQWHG WKH ¿QDO GLVFRXUVH RI WKH VHULHV LQ VL[ \HDUV DIWHUWKH¿UVWRQHDQGE\WKHQKHKDGUHGXFHGWKHµQDWLRQV¶IURP¿YHWR WKUHH)LUVWKHQRZIHOWFRQ¿GHQWDERXWJURXSLQJWKH3HUVLDQVWRJHWKHU with the Hindus, Romans, Greeks, and Goths, and also with the ancient Egyptians (he called them ‘Ethiops’) into one nation. In our terms, KHKDGMRLQHGWKH3HUVLDQVDQGWKH*RWKV¿UPO\WRWKH,QGR(XURSHDQ language family, which is exactly where we would still place them linguistically. No one, today, however, would group the ‘Ethiops’ with WKH,QGR(XURSHDQV-RQHVDOVRUHFRJQLVHGWKHLQÀXHQFHVWKDW+LQGXLVP and Buddhism had had on Southeast Asia, so he added both mainland and island Southeast Asia to the ‘Hindu’ nation. More cautiously, he suggested: ‘… that the settlers in China and Japan had a common origin with the Hindus is no more than highly probable’ (Jones 1993:186). A second nation included “… the Jews and Arabs, the Assyrians, or second Persian race, the people who spoke Syriack, and a numerous tribe of Abyssinians” (Jones 1993:186). The third nation was that of the Tartars, migrating peoples of north and central Asia. After describing WKHVHWKUHHµQDWLRQV¶-RQHVVDLG³&RXOGWKHVHIDFWVEHYHUL¿HGE\WKH best attainable evidence, it would not, I presume, be doubted, that the whole earth was peopled by a variety of shoots from the Indian, Arabian and Tartarian branches, or by such intermixtures of them, as in a course of ages might naturally have happened” (Jones 1993:186). :H VKRXOG QRW EODPH -RQHV IRU PRGLI\LQJ KLV FODVVL¿FDWLRQ DV WKH years passed. The job he had set himself was an empirical one. He wanted to use the evidence from his four areas of knowledge to group the peoples of Asia into ‘nations’, and we should expect some revision of his ideas as his investigations progressed. It was certainly convenient, however, to group the Persians, and even the Chinese and Japanese, with the Hindus, because in this way he reduced the ‘nations’ to three. 7KDWDOORZHGHDFKQDWLRQWREHLGHQWL¿HGZLWKRQHRIWKHWKUHHVRQVRI Noah, and Jones did indeed identify them in that way. The children of Japheth became the Tartarian branch. The children of Shem gave rise to the Arabs, Jews, and Abyssinians, while the Hindus (and thus the Europeans) along with what we would now call the southeast Asians, and, perhaps, even the Chinese and Japanese, were all descended from WHERE DID THE QUESTION 55 the children of Ham (Jones 1993:194-95). A considerable effort at cultural relativism is needed for the modern mind (or at least for my modern mind) to appreciate the worldview of an educated and urbane Englishman of the late eighteenth century. Jones simply took it for granted that the world was created four thousand years EHIRUH&KULVWDQGWKDWWKHÀRRGWRRNSODFHDFRXSOHRIPLOOHQQLDODWHU 2QO\ DIWHU WKH ÀRRG FRXOG SRSXODWLRQV JURZ GLYLGH DQG VSUHDGLQWR WKHLUSUHVHQWKRPHODQGV-RQHVFRQ¿GHQWO\DVVHUWVµ,I0RVHVWKHQZDV endued [sic] with supernatural knowledge, it is no longer probable only, but absolutely certain, that the whole race of man proceeded from Iràn,2 DVIURPDFHQWUHZKHQFHWKH\PLJUDWHGDW¿UVWLQWKUHHJUHDWFRORQLHV and that those three branches grew from a common stock, which had been miraculously preserved in a general convulsion and inundation of this globe’ (Jones 1993:196-97). He gives ‘less than three thousand years’ for the dispersal to be complete (Jones 1993:189). In reviewing the ideas of Sir William Jones, I do not mean to ridicule him. He truly was a brilliant man. He had vast knowledge of the languages, religions, architecture, and of the arts and sciences, and from all of these he attempted to infer the history of nations. Nevertheless, he ZDVORFNHGLQDZRUOGYLHZWKDWZH¿QGGLI¿FXOWWRFRPSUHKHQG-RQHV was active three quarters of a century before Darwin’s Origin of Species was published. Even the long geological chronology of James Hutton had, in Jones’s time, been shared only within a narrow circle in Scotland (Repcheck 2003). In the early part of the nineteenth century, however, the layers of rocks and their fossils would seriously challenge the short chronology of the bible. The short chronology then collapsed completely with Darwin, although it still took some time for the full immensity of geological time to become clear. As it did so, the investigations into the origins and migrations of nations that Jones had conducted were quietly, or perhaps deliberately, forgotten. Even a century ago, Jones’s short chronology must have seemed, as it still seems to us, to be just a WULÀHHPEDUUDVVLQJ,WPXVWDOZD\VKDYHVHHPHGNLQGHUWRJLYHFUHGLWWR the genius who discovered the Indo-European language family, than to dwell on a chronology from which he simply had no means of escape. Both Sir William Jones and the later colonial administrators who wrote about northeast Indian tribes show us how great an interest was taken

2 Jones did not, of course, use ‘Iràn’ in the modern sense. Rather he used it to name WKHXQNQRZQSODFHIURPZKLFKWKHPLJUDWLRQVEHJDQ+HZDVFRQ¿GDQWWKDWWKH\KDGWR begin somewhere in, or at least not far from, Mesopotamia, but he denied knowledge of the exact spot. ‘Iràn’ was a convenient term for that spot, whatever its exact location; Jones 1993[1807]:189. 56 ROBBINS BURLING in migrations by the more scholarly British visitors to India. The most brilliant of them took it for granted that the world was settled, relatively recently from our point of view, by great migrations. Others took it for granted that the modern distribution of tribes could be explained by even more recent migrations. 0LJUDWLRQVKDYH¿JXUHGLQPDQ\RWKHUZD\VLQERWKZHVWHUQDQGHDVW- ern, thinking, and they have been central to a standard view of Indian history, a view that has not yet been fully shaken. The history of the sub-continent has often been described as moulded by wave after wave of invasions from the northwest. The Aryans were followed by the Bactrian Greeks, and then by the Mongols, the Turks, and the Persians. The British GRQRW¿WVRHDVLO\LQWRWKLVVHTXHQFHRILQYDGHUVSDUWO\EHFDXVHWKH\ came by sea rather than across the deserts, but also because they came so recently that we know they did not come in waves. The history of the British period shows us that outsiders can bring major changes even when they do not come in waves. When we recognise the relentless spread of the English language in India today, we ought to be very cautious about assuming that Indo-Aryan needed a massive surge of migrants in order to get established. Language shift can take place for other reasons than migration. The more thoughtful historians of India no longer take seriously the familiar picture of wave after wave of migrants who, in VRPHP\VWHULRXVZD\ZHUHDEOHWR¿QGHQRXJKIRRGWRNHHSDOLYHDV they swarmed across the deserts from the west. The time has past when repeated migrations are needed to give a framework to Indian history. Another favourite migration myth has recently been called into question. It used to be taken as almost obvious that swarms of Anglo-Saxons crossed the English Channel following the decline of Roman power in Britain. These invaders were supposed to have brought the language that became English and to have pushed their Celtic predecessors to the western and northern fringes of the British Isles. Of course, people have regularly migrated in both directions across the English Channel, but recent scholarship has made the picture of swarms of migrants seem unlikely. Modern archaeology suggests far more continuity, both through and following the period of Roman power, than is compatible with massive migrations and population displacement (Pryor 2004). Still another strand in the history of migration myths has been the unending search for the lost tribes of Israel. In his book The Lost Tribes of Israel 7XGRU3DU¿WWGHVFULEHVKRZZLGHDQGHQGXULQJWKHVHDUFK for the lost tribes has been. Eager travellers have scoured every continent for them. Visitors have regularly managed to convince themselves that WHERE DID THE QUESTION 57 they have spotted a lost tribe, and a remarkable number of people have been persuaded that their own group is one of the lost tribes. One of the scores of observers who have believed in lost tribes was a man named Dr. Francis Mason. Mason belonged to the American Baptist Foreign Missionary Society and he arrived in Burma in 1814. In time, he became convinced that the Karen, a large minority people of Burma, were a section of the Lost Tribes, and in 1833 he wrote in a letter: ‘There can scarcely be a rational doubt that the Yuwah of the Karens is the Jehovah of the Hebrews… from the foregoing I am constrained to believe the Karens to be descendants of the Hebrews. Look at them VLULVQRWWKH-HZZULWWHQLQWKHLUFRXQWHQDQFH"¶ 3DU¿WW 7KH Francis Mason who wrote these lines belonged to the same American Baptist Mission Society that sent missionaries to north-eastern India, and it would be surprising if the missionaries to the northeast were not aware of the possibility that lost tribes might be found there too. The belief that the Karens were Jews persisted into the 1930’s, and the idea was taken seriously by Jews in India.3 Even today, some Mizos from northeast India claim to belong to a lost tribe, and some of them have ‘returned’ to Israel where they have been accepted as lost brethren 3DU¿WW5HQJVLPV 7KHVHH[DPSOHVDUHHQRXJKWRVXJJHVW WKDWZKHQWKH\FDPHWRQRUWKHDVW,QGLDERWK%ULWLVKFRORQLDORI¿FHUV and foreign missionaries would have brought along a presumption of migration. We see this presumption clearly expressed in the later ethnographies of Playfair, Mills, and Hutton. 0LJUDWLRQVWRULHVFRQWLQXHWREHWROGDQGP\¿QDOH[DPSOHFRPHV from an article by Randy LaPolla: ‘The Role of Migration and Language Contact in the Development of the Sino-Tibetan Language Family’ (2001). LaPolla gives many fascinating and astonishing examples of population movements back and forth across China over the course of its long history, and then he turns to Southeast Asia. The passage that I quote here is just one very small piece of a much longer and more complex argument, but it is representative of many passages by many writers that I have encountered over the years, about many groups of people. Here is a part of what LaPolla says about the Burmese: The people we have come to think of as the Burmese had been in Yunnan, under the control of the Nanzhao kingdom, and moved down into Burma from the middle of the ninth century. They came down from the northern

3 6KRXOG DQ\RQH LPDJLQH WKDW WKH VHDUFK IRU ORVW WULEHV KDV HQGHG WKH\ PLJKW ¿QG LW interesting to disabuse themselves by a look at such enthusiastic reports as those found in Primack 1998 and Anonymous 1991. 58 ROBBINS BURLING

Shan states into the Kyanksè area south of Mandalay, splitting the Mon in the north and south, and pushed the Karens east of the Irrawaddy. $ERXW$'WKH%XUPHVHFRQTXHUHGWKH0RQWRWKHVRXWKDQGWKH¿UVW Burmese kingdom, the Pagan kingdom, was founded in 1044. The court DGRSWHG PXFK RI 0RQ FXOWXUH LW EHFDPH WKH RI¿FLDO FRXUW FXOWXUH DQG the Mon language (or Pali) was used for inscriptions; the Mon script also became the basis of the Burmese writing system. This was the early period RIPDMRUFRQWDFWDQGLQÀXHQFHRIWKH0RQRQWKH%XUPHVHZKLFKODVWHG until the twelfth century (LaPolla 2001:237). I have no privileged knowledge about the rise of Burmese power a thousand years ago, and I cannot know exactly what happened. Still, I ¿QGLWOLNHO\WKDWWKHDFWXDOHYHQWVZRXOGEHEHWWHUGHVFULEHGLQDUDWKHU different way. I would put it like this: ‘In the ninth and tenth centuries, people led by speakers of an old form of Burmese established control over an area of upper Burma. Speakers of Mon and Karen, who had long occupied much of this area, gradually shifted to the language of their rulers, even while retaining much of their earlier culture. Mon culture FRQWLQXHGDVWKHRI¿FLDOFRXUWFXOWXUHDQGWKH0RQODQJXDJH RU3DOL was used for inscriptions; the Mon script also became the basis of the Burmese writing system.’ The difference between these two descriptions may not seem great, but LaPolla’s implies more population movement and more displacement of people, while mine implies more language shift. LaPolla, like so many others, seems to assume that language change implies population change. I believe, instead, that language can shift without the large-scale population replacement that seems to be implied when LaPolla writes about ‘splitting the Mon in the north and south’ or when he says that the Burmese ‘pushed the Karens east of the Irrawaddy’. Indeed, if people were displaced so dramatically, it is hard to understand how Mon culture could have survived so successfully. Less dramatic migrations, such as those invoked by Hutton, Mills, and Playfair to explain the distribution of tribes and languages in northeast India, seem, to me, to be no more likely. My view of migration has no doubt been coloured by my particular experiences. Most of my time in northeast India has been spent among the Garos, and while I was in the Garo Hills, people occasionally asked for my opinion about where they might have come from. They were curious, I was an expert, and they supposed that I might be able to enlighten them. Only once, during my years in the Garo Hills, has DQ\RQHWROGPHZLWKDQ\FRQ¿GHQFHZKHUHWKH*DURVFDPHIURP7KLV was a half century ago and my memory is not clear on the details, but I do remember the man who explained to me that the older home of WHERE DID THE QUESTION 59 his ancestors was Palestine. The Garos, he assured me, left Palestine about two thousand years ago, and he had drawn a map of the route they followed. These ancient Garos travelled through central Asia via Samarkand and Tashkent, across Tibet, down from the mountains, across the river and then, at last, up to their present homeland. He had written dates beside several of the cities along their route showing when they had passed. In all, the migration took many centuries. I did not, alas, make a copy of his map or record his dates, and I know of no Garos who have sought to ‘return’ to their Palestinian homeland. Except for this single occasion, Garos expressed curiosity, but little knowledge of any migrations that could have brought them to what we now call the Garo Hills. If my memory is reliable, it was most often HGXFDWHG*DURVWKRVHZKRKDGSUR¿WHGIURPWKHPLVVLRQVFKRROVZKR raised questions about migrations and origins. I do not remember much curiosity about migration among the people in villages where formal education had not yet penetrated, and I cannot avoid the suspicion that those whose curiosity had been piqued had been stimulated by the education that western missionaries brought to the northeast. Perhaps this new education had simply stimulated a broader curiosity, but students may also have heard explicit suggestions about migrations. Of course, I should not generalise from the Garos and I should not presume that everyone in the northeast is like them. Colleagues who have worked in other parts of the northeast have assured me that the people among whom they have lived have shown a more lively interest LQPLJUDWLRQVDQGH[SUHVVHG¿UPHURSLQLRQVDERXWWKHLURULJLQVWKDQ, found among the Garos. At the same time, we should not ignore the pos- VLELOLW\WKDWZHVWHUQLQÀXHQFHKDVVWLPXODWHGWKHLULQWHUHVWDQGVKDSHG WKHLUEHOLHIVMXVWDVZHVWHUQLQÀXHQFHKDGXQTXHVWLRQDEO\VKDSHGWKH ideas of the Garo man who traced his ancestry back to Palestine. People everywhere are interested in their origins, and north-easterners do have a special reason for their interest. The tangle of ethnicities found in the hills cries out for an explanation. From ‘Why are there so many tribes?’ and ‘How did all that diversity come about?’ it is but a small step to ‘Where did these all these different people come from? The question of migration is enormously complicated by a tendency to equate language with ethnicity, to presume that everyone who speaks the same language belongs to the same ‘tribe’, and to imagine that both lan- JXDJHDQGWULEDODI¿OLDWLRQDUH¿[HGDQGHWHUQDO %XUOLQJ%XUOLQJ In press). Migration stories offer an explanation for the diversity, but they also imply a longer continuity and much sharper boundaries between the 60 ROBBINS BURLING

HWKQLFGLYLVLRQVWKDQFDQUHDOO\EHMXVWL¿HG,IZHDUHDPXVHGE\WKHVWRU\ RI*DURPLJUDWLRQIURP3DOHVWLQHLWLVSDUWO\EHFDXVHLWLVVRGLI¿FXOWWR imagine an ethnic group maintaining its coherence and distinctiveness over such long stretches of time and space. How many ethnic groups can we name in Europe that have survived for so long? Even to suggest that the Angamis maintained their ethnic boundaries as they wandered into Manipur, and then wandered out again, surely stretches credulity. People construct ethnicity for many reasons: to be like others; to be different from others; to include; to exclude; to assert superiority; to make claims over territory; to forge unity in aggression or defence against their neighbours. One means of asserting common ethnicity is to share a name, and it may be that the label ‘Garo’ helped to create Garo ethnicity. ‘Garo’ is not, in origin, a Garo word. Rather it is the name used for Garos by their Bengali and Assamese neighbours and now by English-speaking foreigners. Garos know the term, but if they are willing to use it, that is only because no single Garo word is its exact equivalent. It helps to give unity to a wider group of people than any other ethnic term available. It allows people to wonder and to ask ‘Where did the Garos come from?’ It is the only term that unambiguously includes everyone whom they would like to include within their ethnic boundaries. 1RUWKHDVWHUQHUVHQJDJHLQHQGOHVVGHEDWHVTXDUUHOVDQG¿JKWVDERXW just where the boundaries among tribes should be drawn, and sometimes the boundaries change. We do not know whether or not the people now known as ‘Nagas’ recognised any common ethnicity before the British came along and grouped them together under a single name, but they certainly recognise a relationship today, even if exactly who is and who is not a Naga remains in contention and can change over time. A small group called the ‘Anal’, who speak a Kuki language, decided, some \HDUVDJRWKDWWKH\ZDQWHGWRDI¿OLDWHZLWKWKH1DJDVDQGWKH\DUHQRZ widely accepted as Nagas (Ruata Rengsi, personal communication). Some Mizos insist that the Lakhers are a kind of Mizo. Some Lakhers insist that they are a separate tribe. A (literally) incredible migration story has made it possible for some Mizos to ‘return’ to Israel. Garos describe the people they call ‘Megams’ as a subtribe of Garos. The Khasis call the same people ‘Lyngngams’ and say that they are a subtribe of Khasis. Whether anyone has thought to ask the Megam-Lyngngams what they think, I do not know. These are only a few examples of changing or disputed ethnicities, but they are enough to undermine any notion that ethnic boundaries are either permanent or unambiguous. Rather, the boundaries are contested, constructed, negotiated, imagined. This allows WHERE DID THE QUESTION 61 ethnicity to be changed. The very lack of clarity about ethnic boundaries may encourage people to seize upon, or to invent, migration stories as a way of justifying their own claims. When ethnic boundaries are unclear, migration stories can prop up whatever one wants to believe. Migration stories may reveal much about the aspirations and ethnic values of the people. They can hardly be taken as reliable historical memories. People migrate. That is not in dispute. But I have never seen evidence that persuades me that whole tribes or ethnic groups migrate, at least not in north-eastern India or in the adjacent mountainous regions of Tibet, China, or Burma. Migration is best seen as a metaphor by which both the indigenous people and outside visitors have tried to understand the history and distribution of peoples. Some northeast Indian hill people may have migration stories that owe nothing to foreign visitors, but the visitors certainly came well equipped with migration stories of their own. 62 ROBBINS BURLING

BIBLIOGRAPHY

Anonymous. 1991. Beyond the Sambatyon. Tel Aviv: The Museum of the Jewish Diaspora. Burling, R. 2007. “Language, Ethnicity and Identity in Northeastern India.” In E. de Maaker and Vibha Joshi (eds.), The Northeast and Beyond: Region and Culture. Special issue of South Asia: Journal of South Asian Studies, pp.391-403. —— 2010. “Three meanings of ‘language’ and ‘dialect’ in North East India.” In G. Hyslop, S. Morey and M.W. Post (eds.), North East Indian Linguistics, Volume III. New Delhi: Cambridge University Press India, pp.35-45. Hutton, J.H. 1921. The Angami Nagas. London: Macmillan and Co. Jones, Sir W. 1993 [1807]. The Collected Works of Sir William Jones, Volume III. New York: New York University Press. LaPolla, R.J. 2001. “The Role of Migration and Language Contact in the Development of the Sino-Tibetan Language Family.” In A.Y. Aikhenvald and R.M.W. Dixon (eds.), Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics. Oxford: Oxford University Press, pp.225-54. Mills, J.P. 1926. The Ao Nagas. London: Macmillan and Co. 3DU¿WW7The Lost Tribes of Israel: The History of a Myth. London: Weidenfeld & Nicolson. Playfair, A. 1909. The Garos. London: David Nutt. William Clowes and Sons. Primack, K. 1998. Jews in Places You Never Thought Of. Hoboken N.J.: KTAV Publishing House. Pryor, F. 2004. Britain A.D.: A Quest for Arthur, England and the Anglo-Saxons. London: HarperCollins. Rengsi, Ruata. ca. 2004. ‘Mizo Migration to Israel’ (unpublished). Repcheck, J. 2003. The Man who Found Time: James Hutton and the Discovery of Earth’s Antiquity. Cambridge MA: Perseus. Trautmann, T.R. 1997. Aryans and British India. Berkely: University of California Press. —— 2006. Languages and Nations. Berkely: University of California Press. COEVOLVING WITH THE LANDSCAPE? MIGRATION NARRATIVES AND THE ENVIRONMENTAL HISTORY OF THE NYISHI TRIBE IN UPLAND ARUNACHAL PRADESH

ALEXANDER AISHER

INTRODUCTION

In this chapter, I will present an accepted version of the account that Nyishi accomplished storytellers (nyejuk) and shaman-priests (nyubu) offer of the westward migration of the Gisum1 sub-clan of the Nyishi tribe through upland parts of the far eastern Himalaya to Talum village in present-day Arunachal Pradesh where they now reside. I will also explore key social and environmental dynamics that may have motivated such Nyishi storytellers and shaman-priests to retain ‘in the belly’ extremely detailed lists of forefathers reaching back over thirty- two generations or approximately eight hundred years. I will be asking what possible insights Nyishi accounts of migration might offer into the broader social and environmental history of the Nyishi tribe, in SDUWLFXODU KRZ VXFK DFFRXQWV KDYH LQÀXHQFHG SUHVHQW SHUFHSWLRQV RI the landscape in Talum village. Furthermore. I will raise the question of whether accounts of migration shed light upon how key institutions and practices of the people living in Talum village have ‘coevolved’ over time with ecological systems in which they are embedded. In order to construct a speculative timeline consistent with the historical understanding of my Nyishi respondents, throughout this paper I treat the oral historical materials offered to me in Talum village as true. In the second half of the essay, I try to imagine what purpose chronicling this migration might serve.

FROM DIVERSITY TO UNITY: RETURNING TO ORIGINS

Most migrations end somewhere, if only where a storyteller sits. For members of the Gisum sub-clan of the Nyishi tribe, with whom I con- GXFWHGDQWKURSRORJLFDO¿HOGZRUNLQWKHLUORQJPLJUDWLRQHQGV

1 A pseudonym has been used here to protect informants. 64 ALEXANDER AISHER in a small village on a steep ridge 12 kilometres outside the township of Koloriang in the upland district of Kurung Kumey in Arunachal Pradesh. 7R VRPH H[WHQW WKLV JHRJUDSKLFDO VSHFL¿FLW\ DULVHV RXW RI IRUZDUGV chronology. The evolutionary biologist Richard Dawkins (2004:6) notes succinctly how, in the realm of biological speciation, backwards and forwards chronology are each good for different purposes: “Go backwards and no matter where you start you end up celebrating the unity of life. Go forwards and you extol diversity.” If a child living in Talum village in the present day were to ask an accomplished storyteller or shaman-priest to list the human generations, the patriline, leading back into the past, as they travelled back through generations they would discover the following. Five generations back, approximately 125 years ago (circa 1875), they would come upon a man called Gisum, after whom their sub-clan is named. Eight generations back from the present day, some two hundred years ago (circa 1800), they would encounter the name of the man Suming2 who settled Talum village. Sixteen generations back (circa 1600), they would come upon Bengia, after whom the powerful upland Bengia clan takes its name. Twenty-eight generations back (circa 1300), they will encounter the three brothers Dodum, Dopum and Dolu, after whom the three main phratries of the Nyishi tribe take their name. The storyteller may note sagely that this marks the historical moment when the Nyishi tribe separated off from other Tani tribes. If the storyteller continued listing the names of forefathers, eventually, some thirty-two generations ago (circa 1200), they would arrive at the famed human ancestor Singtong-Tani, known popularly as Abu-Tani or ‘Father’ Tani, to whom all people in the Tani tribes of Arunachal Pradesh, including the Nyishi, trace their ancestry. Continuing further back towards unity, some thirty-six generations ago (circa 1100), they would come upon the name Kulu, a term many Nyishi villagers use as a general term denoting the distant past. Here the storyteller may slap his hands and end his (or her) exercise in genealogical memory. If that same child of Talum village were to ask the accomplished storyteller or shaman-priest to tell him more about these named forefathers he would learn of exploits of certain prominent forebears, and he would hear a great deal about the great trickster forefather Abu-Tani. But most interesting of all, especially if the child wanted to know from ZKHUHVSHFL¿FDOO\KHFDPHKHPD\KHDUDERXWYDULRXVVHWWOHPHQWVKLV forefathers established during their ‘migration’ westwards over several

2 A pseudonym has been used here to protect informants. COEVOLVING WITH THE LANDSCAPE? 65 hundred years in the direction of Talum village. The storyteller might recount certain events that occurred in key settlements. He or she might also describe some features of the landscape through which the forefathers travelled in a great stop-start migration that ended, for better or for worse, in Talum village. Nine hundred years of oral history ‘held in the belly’ is impressive— that it can be linked to concrete landscapes along a migration timeline is even more intriguing. For an anthropologist a series of questions arise: why remember so far into the past? What do such memories tell us about both the Nyishi tribe and the landscapes through which they travelled?

THE NYISHI TRIBE OF CENTRAL ARUNACHAL PRADESH

The Nyishi tribe, whose migration is the subject of this paper, is one of a group of tribes collectively known as Tani tribes, who inhabit the central region of Arunachal Pradesh. They refer to themselves as a Tani tribe after the legendary ancestor Abu-Tani, or Abo-Tani, from whom they trace their descent.37KH1\LVKLWULEHZKLFKKDVVLJQL¿FDQWSRSXODWLRQV in the districts of East Kameng, Papum Pare, Upper Subansiri, Lower Subansiri, and Kurung Kumey, is divided into strictly exogamous, patrilineal clans. A strong sense of kinship and loyalty binds members of each clan. Marriage involves the payment of bride price. Inheritance of property still tends to follow down the male line. The majority of Nyishi people still live in villages where the longhouse, constructed mainly out of bamboo with sloping layers of wild banana-leaf thatch, remains the main form of habitation. Between two and four families, and sometimes more, inhabit the longhouse, which serves as the centre of the social and economic life of the village. Shifting cultivation of dry rice and millet and animal husbandry, primarily pigs, chicken and semi-domesticated jungle oxen known across Northeast India as mithun (Bos frontalis), are central to the village economy. Due in part to the steep gradient of the rugged and sparsely populated hills and valleys of upland Arunachal Pradesh, Nyishi villagers tend to have access to extremely diverse forest types, from tropical forests at the bottom of valleys to subtropical and temperate forests above many villages. Fourteen species of bamboo and seven species of cane form a

3 See the chapters by both Blackburn and Huber in this volume. While linguists identify a group of related Tani languages throughout central Arunachal Pradesh and adjacent regions (see the chapter by Post in this volume), these are not always coincident with ethnic claims for the existence of Tani tribes. 66 ALEXANDER AISHER key component of village-level material culture. Villagers also procure from the forests a wide range of leafy green vegetables, fruits and edible bamboo shoots. At tropical elevations sago palms also grow in small, LQGLYLGXDOO\RZQHGSODQWDWLRQV+XQWLQJDQG¿VKLQJLQPDQ\YLOODJHV UHPDLQVLJQL¿FDQWFRPSRQHQWVRIWKHYLOODJHHFRQRP\

FROM UNITY TO DIVERSITY I: BIRTHING A WORLD

The migration begins where all things begin, in the deep pre-human past of Kirium-Kulu. Before the Earth and her sister Sky have separated, EHIRUHQLJKWDQGGD\EHIRUHWKH6XQZHVHHWRGD\¿UVWVKRQHXSRQWKH Earth, events occur that make way for the existence of all things. In the beginning there is no Earth and no Sky. There is neither Sun nor Moon nor Stars. Instead there is just a great sunless plain. This, it is said, is the ‘mother’ of all things. Some call it Kurium-Ngarngum-Kulu- Ngarngum-Ngarngumney. Here is the source of all things. This is the mother of Earth and Sky (Sechung-Nyedoh). All things come from this primordial plain. The entity is liquid, perhaps molten and unstable. The great, undivided entity begins to divide. Ngarkuph and Ngarteney, two sisters that will later become Earth and Sky, begin to separate. In this murky world before sunlight, the two sisters proto-Earth and proto-Sky remain close to each other, only partially distinct. They start to revolve, ¿UVWWRZDUGVWKH3ODFHRI6SLULWV 8\X1\RNX WKHQWRZDUGVWKH3ODFHRI Humans (Nyia-Nyoku). By so doing, they bring forth these places. Still there is no sunlight. $6XQFRPHVLQWRH[LVWHQFH7KLV¿UVW6XQFDOOHG-XQJNXSK-XQJPDK Jungturr-Ajungney, shines brightly upon the Earth. Seeds germinate and WKH (DUWK EHFRPHV JUHHQ ZLWK SODQWV %XW WKLV ¿UVW 6XQ VKLQHV GRZQ VR ¿HUFHO\ XSRQ WKH (DUWK WKDW LW GULHV LW RXW 3ODQWV ZLWKHU DQG GLH )URPWKHVDPHVRXUFHDVWKH¿UVW6XQDQRWKHUDULVHV7KLVVHFRQG6XQ FDOOHG-XQJNXQJ.DUQ\XQJVKRRWVRXWWKHH\HRIWKH¿UVW6XQ,WVDUURZ pierces deep into the eye and it explodes into countless fragments. These IUDJPHQWVRIWKH¿UVW6XQEHFRPHWKHVWDUV%RWK6XQVWKHQGLVDSSHDU A third comes into existence—the Sun we see today, who some call Jungur-Juyung-Ajungney. Under the light of this Sun life grows again upon the Earth. Earth becomes the place and shelter of living beings. Green plants grow richly upon the Earth. It is during this period also, at the time of the appearance of the third Sun, that the Moon, which is male, also comes into existence. %HIRUHWKH6XQZHVHH¿UVWVKLQHVGRZQXSRQWKH(DUWKWKHXQGLYLGHG COEVOLVING WITH THE LANDSCAPE? 67 being called Ngarkuph-Ngarteney gives birth to Water, Air and Fire. /RQJEHIRUHWKH¿UVWGDZQEHIRUHWKHVLVWHUV(DUWKDQG6N\KDYHIXOO\ separated, whilst they still ‘sit’ close to each other, they give birth to Hiyeeney, Water. From beneath the thighs of Ngarkuph-Ngarteney water ÀRZV7KHQ1JDUNXSK1JDUWHQH\DVNV³1RZWKDW+L\HHQH\KDVEHHQ born, what shall we now bear?”, and from her teeth she draws forth Emey, Fire. She then states, “We have created a source of heat. Let us now create a source of cold.” And from her mouth Ngarkuph-Ngartey GUDZV$MXSK$DQH\WKH0RWKHURI$LU,QWKHQLJKWEHIRUHWKH¿UVWGDZQ this Mother of Air, Ajuph-Aaney, enters the body of the insect Songsong- +DKUL\LQJ$VVKH¿OOVLWVKHDGDQGERG\LWVRXWHUVKHOOSHHOVDZD\DQGD large emerald-green insect with small wings, the yahniy insect, emerges. 7KHQWKH0RWKHURI$LU¿OOVLWVOHJVLWVVWRPDFKDQGLWVZLQJV6ORZO\ its wings unfurl and extend out, and the insect takes to the air. It beats its ZLQJVFDXVLQJ$LUDURXQGLWWRFLUFXODWHDQGPRYHDQGÀRZLQGLIIHUHQW GLUHFWLRQVDQGZKHUHYHULWÀRZVLWEULQJVOLIHDQGUHIUHVKHVOLIH/LIH ÀRXULVKHV-XVWEHIRUHGDZQWKH$LUHQWHUVWKHFRFNHUHO,W¿OOVLWVOXQJV and the cockerel crows. This is why, still today, the cockerel crows just before dawn. At some point in this distant time of beginnings, the sisters Earth and Sky grow ill. The two sisters are not born whole. As they settle into the form they have today large holes (ongopeh) develop. Because of these holes they suffer. Earth and Sky converse with each other, and from the FHQWUHRIWKHLUEHLQJWKH\GUDZIRUWKWKH¿UVWVKDPDQSULHVW*RQJWXQJ 7DMXUU7KLV¿UVWVKDPDQSULHVWSHUIRUPVDJUHDWVDFUL¿FHWRKHDOWKHPD ULWXDOWR¿OOWKHLUKROHV7KHVDFUL¿FHIRUVLVWHU6N\KHDOVKHUDQG¿OOVKHU holes. Then Gongtung-Tajurr performs a ritual to heal the Earth. Liquid or molten stone (takum) seeps out of the holes and creeps and ‘slithers’ RYHUWKHVXUIDFHRIWKH(DUWK7KLVOLTXLGVWRQHVWDUWVWR¿OOWKHKROHV in the Earth, thus giving birth in the process to Lungfungney, stone. However, as Gongtung-Tajurr conducts the ritual some of those ‘people’ JDWKHUHG DURXQG WKH DOWDU EHFRPH XQVDWLV¿HG ZLWK WKHLU VKDUH RIWKH VDFUL¿FH7KH\VSUHDGGDPDJLQJUXPRXUVDQGGRXEWVDERXW*RQJWXQJ Tajurr’s ability. As a result, he is unable to complete the ritual. The Earth LVQRWKHDOHGIXOO\7KLVLVZK\ODQGVOLGHVDQGÀRRGVVWLOORFFXUWRGD\ Although the Earth is not healed completely, this healing of the Earth and Sky made life possible. It is a root cause or ‘mother-reason’, for the existence of living things. So the shaman-priests say. $WVRPHSRLQWLQWKLVGLVWDQWWLPHRIEHJLQQLQJVWKH¿UVWVSLULW uyu) DQG¿UVWKXPDQEHLQJ nyia) are born. For a time, the trickster forefather 68 ALEXANDER AISHER

RUDOOKXPDQEHLQJV¿UVWLQWKHOLQHRISURWRKXPDQVNQRZQFROOHFWLYHO\ as Abu-Tani, walks the Earth. Half human and half-spirit at this time, Abu-Tani lives in a world populated by plants and animals. At this time they too have a human form, and he ‘marries’ many of them, thereby fathering many creatures and entities. He marries dry-leaves-and-twigs, Kosuk-Koruh, and the leech (tapik) is born. He marries broken bamboo, +RWXQJ+D\XQJDQGWKHÀ\ tanyik) is born. He fathers many creatures as the world grows ever more diverse. Tragically, the libidinous trickster Abu-Tani also marries the spirit Uyuney and this great mother of spirits bears the ‘spirit-mothers’ of diverse spirit lineages, including many that DIÀLFWKXPDQEHLQJV7KXVWKHOLQHVRIKXPDQEHLQJVDQGVSLULWVDUH established respectively. The implications of these events reach deep into the condition of humanity in the present day. With multiple beings come multiple perspectives and also multiple stakeholders and claims upon the world. Spirits grow jealous of Abu- 7DQL &RQÀLFW DULVHV EHWZHHQ KXPDQV DQG VSLULWV -XVW DV D KXPDQ arbitrator (gingdong-nyejuk VLWVEHWZHHQFRQÀLFWLQJKXPDQSDUWLHVVR in this time of beginnings a shaman-priest positions himself between humans and spirits. A settlement is reached. Laying a partition between KXPDQVDQGVSLULWVWKHVKDPDQSULHVWFDXWLRQVWKHFRQÀLFWLQJSDUWLHV “You humans should not go to the spirit side, and you spirits should not go to the human side. Only the dead will go to the side of spirits!” In this way a boundary (lungruk) is erected between humans and spirits. Through this they were separated, rendering spirits and humans only partially visible to each other.

INTERLUDE

Thus, the world that Talum villagers know is brought forth, driven by PDUULDJHVDQGELUWKVGLYLVLRQVFRQÀLFWVDQGUHVROXWLRQVLOOQHVVHVDQG healings. Through such processes the world transforms and moves from the past towards the present, from singularity to multiplicity, from sim- plicity to complexity. In these stories of beginnings, both parentage and causality intertwine richly; most entities are born rather than created. Processes that are as much ‘social’ as ‘natural’ drive this early history of the cosmos. +HUH LQ WKLV WLPH RI EHJLQQLQJV PDQ\ RWKHU VLJQL¿FDQW RULJLQDWLQJ events take place: the origin of human illness; the origin of poverty; the origin of divination, and many more. But we must leapfrog over these events, and turn instead to that time and place when all humans lived COEVOLVING WITH THE LANDSCAPE? 69 together, when the people of the plains and the plateau and all tribes and clans lived together. The migration of the Nyishi tribe has its source and origin in this place of beginnings, this place of unity. The migration began somewhere in the East, where the Sun rises, so storytellers from Talum village claim. The ancestral migrants travelled through a concrete, geographically describable landscape, establishing settlements as they went. Storytellers and shaman-priests claim these settlements still exist today. Accomplished storytellers with whom I spoke described the following events.

FROM UNITY TO DIVERSITY II: MIGRATION TO TALUM VILLAGE

Thirty-two generations ago, some eight hundred years before the present (circa 1200), around the time that Genghis Khan was uniting the nomadic WULEHV LQ QRUWKHDVW$VLD ZH ¿QG WKH IRUHIDWKHUV RI SHRSOH LQKDELWLQJ Talum village, the forefathers of all Tani tribes and nyipa-nemey—a category which includes Tibetans, Chinese and ‘plains people’—on the move. This as yet undivided group travels towards a valley called Sungdu-Numchung, literally ‘Sungdu-settlement’, situated in the east “where the Sun rises”. Storytellers describe how the ancestral migrants looked down from a high place upon that pleasant valley blanketed with grass-like herbaceous plants and broad medium-sized waxy-leaved shrubs and trees; a valley abundant with game. However, two enormous vine-strewn boulders block their way down into the valley. The ancestral migrants clamber down these vine-strewn boulders to UHDFKWKHYDOOH\ÀRRUEXWDVWKH\GRVRWKH\ORVHWKHLUJULSDQGIDOO +XPDQVRQFHKDGWDLOVEXWWKH\IHOOVRKDUGXSRQDÀDWERXOGHUWKHUH in Sungdu valley that they lost their tails. As one Nyishi shaman-priest IURP ,WDQDJDU QRWHG ³%HIRUH 1DUED >WKH ¿UVW KXPDQ VHWWOHPHQW@ ZH were just like monkeys. We lost our tails at Narba. We fell heavily upon that stone.” In Sungdu valley the ancestral migrants establish a settlement called 1DUED +HUH WKH PLJUDQWV DQG E\ LPSOLFDWLRQ KXPDQLW\ ÀRXULVK DQG multiply. This settlement, Narba, resonates powerfully in the imagina- WLRQRIPDQ\1\LVKLSHRSOH$OWKRXJKLWLVQRWWKH¿UVWKXPDQVHWWOH- ment—most storytellers claim it is a mid-point in the migration—many key institutions and practices of the Nyishi and other Tani tribes are said to have originated here. “This is how we did it in Narba”, is a common statement, and one that has normative value. In the Nyishi historical imagination Narba is more than just a place. It also represents an ideal: 70 ALEXANDER AISHER a settlement where the Nyishi and other tribes lived together peacefully DQGZKHUHWKH\ÀRXULVKHG,WLVDSODFHRIXQLW\SULRUWRWKHVHSDUDWLRQ RIWKHWULEHVEHIRUHFRQÀLFWDURVHEHWZHHQGLIIHUHQWWULEHVDQGFODQV)RU WKLVUHDVRQWUDGLWLRQDODUELWUDWRUVVLWWLQJEHWZHHQFRQÀLFWLQJLQGLYLGXDOV and groups refer them to Narba as a yardstick against which they should judge whether their behaviour is appropriate. In short, Narba serves as a template for best practice. )RU VHYHQ JHQHUDWLRQV²URXJKO\ RQH KXQGUHG DQG VHYHQW\¿YH \HDUV²WKH IRUHIDWKHUV OLYH DQG ÀRXULVK LQ 1DUED 'XULQJ WKLV SHULRG (circa 1200-1375), when the Italian traveller Marco Polo visits China, the bubonic plague kills a large proportion of the populations in Europe and Asia, and the Tai-Ahom establish the Ahom Kingdom in parts of present-day Assam, a series of critical environmental and historical events take place. By overlaying genealogical and migration timelines the following events seem to occur in Narba. Here the Abu-Tani known as Singtong-Tani acquires livestock. Here in Narba Abu-Tani also dis- covers rice, marries the ‘mother’ of rice, Duguhney, and learns how to cultivate it. Unable to return with Abu-Tani to the Place of Humans, Duguhney sends her seven children, ‘season birds’, to help human cultivators time their agricultural schedule. 2IPXFKVLJQL¿FDQFHIRU1\LVKLSHRSOHLWDSSHDUVWKDWLQ1DUEDWKH Abu-Tani known as Singtong-Tani marries the daughter of the Sun, Griungney, who ‘spreads human seed’. As a result, many spirits grow jealous of Abu-Tani, including the powerful land spirit Dohjung. By way RIFRPSHQVDWLRQ*ULXQJQH\RIIHUVWKHVSHFLHVRI¿JWUHHFDOOHGsangrik trees as ‘tree-brides’ for Dohjung. These spirit-trees (uyu-sangney) serve not only as the house and shelter of certain land spirits, but in many villagers—Talum village included—they mark the upper elevation at which human ownership of land terminates. Thereafter, such trees will mark the edge of the commons, beyond which Dohjung spirits own the land. Dohjung tells Abu-Tani, “You shall cultivate only up to this point. Anywhere above this is my land.” In temperate forest areas above where sangrik trees grow Talum and other villagers cultivate at their peril, and accomplished hunters (nyegum) tend to employ a different vocabulary to avoid alerting Dohjung spirits to their intentions. Here in the commons, with its boundary clearly marked by the sangrik tree-brides given by the daughter of the Sun, villages never utter the name of the spirit Dohjung. The consequences of these events that supposedly occured in Narba thus ripple down through generations of the Nyishi tribe, shaping the lives and practices of their descendants, marking the beginning of a new COEVOLVING WITH THE LANDSCAPE? 71 relationship between the Nyishi and powerful Dohjung land spirits, and by extension the land itself. Another critical event in the history of the Nyishi tribe also occurred in Narba, thirty-one generations ago (circa 1225), at around the time that 4XWEDO'LQ$\EDN¿QLVKHGHUUHFWLQJWKH4XWE0LQDULQ'HOKLWKHQWKH WDOOHVWPLQDUHWLQWKHZRUOG7KH¿UVWKXPDQKXQWHUV%RGXDQG%RPLK sons of Abu-Tani, are annihilated. In a spectacle of revenge, blood- coloured clouds—symbolic perhaps of Dohjung spirits—gather above the dense forest where Bodu and Bomih exuberantly, fanatically, attempt to hunt to extinction the last of the boar, bear, monkey and deer. The two die there in the forest, and remain for generations to come—including hunters living in Talum village—as examples of hunters who took too much from the Dohjung spirits who rear these forest animals, and who paid the price for their excess. Somewhat later, twenty-eight generations ago (circa 1300), around the time that the imperial census of China recorded roughly sixty million inhabitants, three brothers are born in Narba. Their names are Dodum, Dopum and Dolu, after whom the three main phratries or descent groups of the Nyishi tribe take their name. When chanting, shaman-priests often refer to the Nyishi tribe as “Dodum, Dopum, Dolu”. This, we could speculate, signals the emergence in Narba of what will become the Nyishi tribe. Here, perhaps, the Nyishi tribe split off from other Tani tribes. Following down the Dodum line—as we must to reach Talum village—two generations later (circa 1375), a man called Kiyfah leaves Narba. Twenty-four generations back (circa 1400), around the time that the 0XVOLP*UDQG(XQXFK=KHQJ+HLVSUHSDULQJWRVHWRXWRQWKH¿UVWRI seven naval expeditions during which he will visit South India, two Nyishi brothers, Fadeh and Fagah, initiate a new relationship with the land spirit Dohjung. As the story goes, these two brothers are out hunting in the forest. Far from their village they camp for the night in a cave. There they encounter two girls. They marry the girls, Siney and Miney, who turn out to be daughters of the land spirit Dohjung. Elements in this story suggest this is a forced marriage. As a marriage gift, the spirit Dohjung offers the two brothers various species of forest animals. As Fadeh and Fagah set off back to the village, their wives warn the two hunters not to look to the forest. The two hunters agree. But just as they are entering their village they hear a sound like thunder. They turn around and all the forest animals that were following behind, gifts from 'RKMXQJÀHHEDFNWRWKHIRUHVW Hunters claim that as a result of this event forest animals remain 72 ALEXANDER AISHER

IHDUIXORIKXPDQEHLQJVDQGGLI¿FXOWWRFDWFK7KLVVWRU\WKHUHIRUHPDUNV another pivotal moment in the Nyishi environmental history when a new relationship with the forest and forest animals is brought forth. For the descendants of Fadeh and Fagah, which include Talum villagers and other members of the Bengia clan, this event marks the beginning of an DI¿QDOUHODWLRQVKLSZLWK'RKMXQJODQGVSLULWVDQGDQHZUHODWLRQVKLSRI exchange with the forest and wider landscape. Hunting and cultivation ULWXDOVSHUIRUPHGLQ7DOXPYLOODJHLQWKHSUHVHQWGD\H[SUHVVWKLVDI¿QDO relationship. This event also marks the origin of an important Nyishi ceremony called Longtey Yuloh. Sixteen generations ago (circa 1600), around the time that the play- wright William Shakespeare was composing Hamlet, the forefathers of people living in Talum village established a settlement called Ridu- Numchung. Nyishi storytellers offer clear details of the landscape. Two rivers, the Sepih River—a tributary of the Siang River—and the Sulu 5LYHUÀRZWKURXJKWKLVYDOOH\VLWXDWHGFORVHWRWKHUHPRWHXSODQGDUHD of Taksing that borders the Tibetan Plateau. So Talum oral historians SURSRVH+HUHLWVHHPVVHYHUHFRQÀLFWHUXSWHGDQGPDQ\SHRSOHGLHGDV a consequence. The group of migrants divided up into several separate clans. As a result of this, the Bengia clan was born. A new ritual was initiated. The newly differentiated clans migrated out in different directions into the surrounding landscape. The forefathers of Talum villagers established a settlement called Rigu-Numchung, in a valley through ZKLFKWKH8SSHU6XEDQVLUL5LYHUÀRZV Eleven generations ago (circa 1725), around the time that Catherine I EHFRPHV(PSUHVVRI5XVVLDZH¿QGWZREURWKHUV7XPFKL\DQG7XPL\ forefathers of the inhabitants of Talum village, living near the Kurung River in latter-day Kurung Kumey District. Storytellers of Talum village describe how Tumchiy and Tumiy notice a group of Himalayan Goral (Naemorhedus goral)—long-haired mountain goats with backward curving horns—heading down through the thick forest that once blanketed that side of the Kurung valley. They follow the goral and notice from a distance that the group are drinking at a spring. They realise immediately that this is a saltwater spring. And so they settled there. Prior to road communications with the lower belt and plains, access to salt was highly valued by upland Nyishi. Rock salt was only available WKURXJKWUDGHDFURVVGLI¿FXOWPRXQWDLQSDVVHVZLWK7LEHW Eight generations ago (circa 1800), around the time that Napoleon %RQDSDUWHFURVVHGWKH$OSVDQGLQYDGHG,WDO\ZH¿QGWKHIRUHIDWKHUVRI Talum villagers living in a village called Papuph, just a half day’s walk COEVOLVING WITH THE LANDSCAPE? 73 from Koloriang. This village stands on the other side of the Kurung valley from Talum village, and is visible from there. Here in Papuph village the forefathers Suming and Sumsih4ÀRXULVKDQGWKHLURIIVSULQJ spread out into the hills surrounding Koloriang. Then, an event occurred that apparently forced the forefathers to cross the Kurung River and to establish the village that comes to be known as Talum. As the story goes, whilst hunting in the forests which once blanketed that side of the valley, and which swidden cultivators have since reduced to secondary re-growth, the brothers Suming and Sumsih encounter a large boar. They FKDVHWKHERDUZKLFKÀHHVGRZQWKURXJKWKHIRUHVWWRWKH.XUXQJ5LYHU some distance below. The boar leaps into the glacial waters, swims DFURVVVFUDPEOHVXSWKHRSSRVLWHEDQNDQGÀHHVLQWRWKHGHQVHIRUHVW on the other side. Unwilling to give up the chase the two brothers follow upriver to a place where they can cross. They run back down to where the boar emerged from the river and follow its tracks into the forest. There they locate it, kill it, and carry the boar back to the riverside. From there they call out to villagers in Papuph village. A group cultivating on the hillside above hear them and rush down to the riverside. In traditional fashion, they tie a length of cane to a tree, and then hurl it over to the brothers waiting on the other side. They secure the line and create a crossing. On a steep ridge some distance above the crossing the elder brother Suming settles the village now known as Talum. His wives MRLQKLPDQGWKHUHWKH\ÀRXULVK Jumping forward another eight generations, some two hundred years, ZHDUULYHEDFNEHVLGHD¿UHLQDORQJKRXVHLQ7DOXPYLOODJHDWWKHEHJLQ- QLQJRIWKHWZHQW\¿UVWFHQWXU\DVDFKLOGOLVWHQVWRDQDFFRPSOLVKHGVWR- ryteller recount the names of forefathers leading forward from Kirium- Kulu, where all things began, to Talum village. So ends one thread of the migration of the Nyishi tribe. Depending upon how far back one chooses to travel in the narrative, this is also the migration of humanity, of the Tani tribes, the Nyishi tribe, Bengia clan, and Gisum sub-clan.

WHY REMEMBER THE MIGRATION?

I would now like to explore what this migration narrative might actually mean to people living in Talum village. Whilst accomplished storytellers in Talum and many other upland villages often characterise the migration as a coherent event culminating in the settlement of their village, in

4 A pseudonym has been used here to protect informants. 74 ALEXANDER AISHER

DOOOLNHOLKRRGLWRFFXUUHGLQ¿WVDQGVWDUWVWULJJHUHGE\GLIIHUHQWW\SHV of events. If storytellers present it as a single coherent event this may be due to the pattern underpinning events occurring throughout the timeline: a continual progression from unity to diversity. As noted earlier, if our child of Talum village asked their interlocutor where the long migration of the Nyishi tribe began, the storyteller or shaman-priest might tell them the migration began where all things began, in Kirium-Kulu, and had its root cause or ‘mother-reason’ in the processes that gave birth to the world itself and to humanity. As I hope to have shown above, the migration of the Nyishi tribe is an extension of their creation myth, and not entirely separate from it. Like the earliest moments of creation, the migration of the tribe itself is marked by a series of distinctions and separations from an original unity. This process, referred to by Nyishi villagers as doging, takes place throughout the migration. Whether the event is the separation of Sky and Earth, or separation of the domains of spirits and humans, or division of the original stock of migrants into tribes, clans and sub-clans, this differentiation is effectively ongoing. A range of factors may have driven the forefathers of people living in Talum village ever westwards towards Koloriang. However, it seems FRQÀLFWEHWZHHQJURXSVDQGFODQZDUIDUHZHUHVLJQL¿FDQWIDFWRUV0RVW Nyishi storytellers identify one driver above all others, one that is closely WLHGWRFRQÀLFWDQGFODQZDUIDUHWKHVHDUFKIRUQHZODQGWRFXOWLYDWH and to settle. Storytellers state that the forefathers migrated across the uplands as part of a process of ‘land-claiming’ (nyoku-lahling). Some of the narratives that occur within the broader sweep of Nyishi migration may refer to concrete historical events, whilst others have a PRUHP\WKRORJLFDOÀDYRXU&OHDUO\WKLVPLJUDWLRQQDUUDWLYHGRHVQRWDGG up to a straightforward chronicle of events surrounding the movement of a people, as they migrated through the uplands of latter day Arunachal Pradesh. If these disparate stories, these fragments of a complex history, do not serve simply as a chronicle, then what purpose do they serve? Why invest energy in remembering them? Below I highlight several social and environmental dynamics that may have catalysed the process of remembering long genealogical lists reaching back several hundred years and the story of the migration of these forefathers through the far eastern Himalayas. COEVOLVING WITH THE LANDSCAPE? 75

POSSIBILITY I: COEVOLVING WITH A SPIRIT-INHABITED LANDSCAPE

Undoubtedly, the landscape Nyishi villagers came to inhabit at the end of their long migration from Kirium-Kulu was different to the landscape through which they initially travelled. Its character changed. As the tribe, clan and sub-clan moved through it, they accumulated stories and histories, and so did the landscape. Whilst some stories mark or justify the origin of a particular practice or tribal institution or even ritual, some portray new relationships with spirits, particularly land spirits, and by H[WHQVLRQZLWKWKHODQGLWVHOI,IZHLQFOXGH²VRPHZKDWDUWL¿FLDOO\EXW not unreasonably—in the historical timeline of the Nyishi migration the OLIHDQGH[SORLWVRIWKH¿UVWKXPDQEHLQJ$EX7DQLDQGDOVRFRVPRORJL- cal events that preceded the birth of humanity, several events powerfully inform the character of the landscape with which Talum villagers interact in the present day. An incomplete list of perceptions and events that inform them might include the following: i) The Earth is inherently unstable and susceptible to earthquakes and landslides. (YHQW 3UHKLVWRU\ 7KHULWXDOWRKHDODQG¿OOWKHKROHVLQWKH(DUWKLVQRW completed. ii) Humans cohabit the landscape with a host of malign spirits who threaten human life, wealth and property. (YHQW 3UHKLVWRU\ $EX7DQL WKH ¿UVW KXPDQ EHLQJ KDV VH[XDO LQWHU- course with the mother of all spirits, Uyuney, who gives birth to the malign spirits that threaten human beings today. iii) Spirits of all types, including malign spirits remain under normal circumstances invisible to human eyes. (YHQW FLUFD &RQÀLFWHVFDODWHVEHWZHHQKXPDQVDQGVSLULWVDQGLV ¿QDOO\UHVROYHGE\HUHFWLQJDSDUWLWLRQEHWZHHQWKHGRPDLQVRIKXPDQV and spirits. iv) Dohjung land spirits are legitimate owners of land (the commons); the sangrik tree marks the vertical extent of human ownership of land. Event: (circa 1200) The daughter of Sun compensates Dohjung land spirits for her marriage to Abu-Tani with a gift of sangrik ‘tree-brides’; Dohjung claims the land above the sangrik tree as his own. v) Successful human cultivation requires that human cultivators time their agricultural schedules according to the annual migration into the local landscape of seven ‘season birds’. Event: (circa 1200) Abu-Tani marries Duguhney, the ‘mother’ of rice cultivation, who bears seven children. She promises to send her children each year to aid cultivation. 76 ALEXANDER AISHER

vi) Rice cultivation implies exchange, both material and ritual, with land spirits. Event: (circa 1200) Duguhney, the mother of cultivation, advises humans, ZKHQWKH\VRZWKHLU¿HOGVWRPDNHDQRIIHULQJ µZLWKWKHOHIWKDQG¶ WR those jealous land spirits that may be offended by human cultivation. vii) Hunting risks revenge by land spirits. (YHQW FLUFD 7KH IRUHIDWKHUV %RGX DQG %RPLK WKH ¿UVW KXPDQ hunters, exuberantly hunt forest animals to the point of extinction, and are consequently annihilated by Dohjung land spirits. viii) 7DOXPYLOODJHUV²OLNHDOOPHPEHUVRIWKH%HQJLDFODQ²PDLQWDLQDQDI¿QDO relationship with Dohjung land spirits, with attendant ritual and material forms of exchange. Event: (circa 1400) brothers Fadeh and Fagah marry Siney and Miney, the daughters of Dohjung. ix) It is the fault of human beings that forest animals remain timid and elusive. Event: (circa 1400) Fadeh and Fagah fail to receive the marriage gift of 'RKMXQJFDXVLQJIRUHVWDQLPDOVWRÀHHEDFNWRWKHIRUHVW This list could be extended to include many more events. But we can see from the list that many events that occurred during the migration have come to inform the character of the landscape, as Talum villagers presently understand it. On this basis, a series of motivations for remembering the migration come into view: the landscape is inhabited by a vast number of spirits; these spirits inhabit diverse forest types and many features of the landscape, such as rivers, ridges, trees and boulders; these spirits render areas of the landscape sentient; cohabitation with spirits commits villagers to particular forms of conduct and modes of exchange based upon historical precedent. These four motivations appear at that juncture between humans, spirits and the landscape where ‘social’ transformations have ‘ecological’ implications, and vice versa. Historically, transformed relationships with spirits entail new ways of perceiving and interacting with the landscape. The migrants did not simply pass through an inert landscape, a passive backdrop to human action; they coevolved with it. Viewed from this perspective, Nyishi memories of migration, at some levels at least, offer Nyishi villagers a chronicle of several hundred years of coevolution of their tribe, clan and sub-clan with the landscape. We could even claim that, via the spirits, Nyishi villagers also credit the landscape and forests with a memory; the land also remembers the migration of their forefathers through it. Following this line of interpretation, Talum villagers’ memories of migration apparently contain useful and important information about KRZWRGHDOZLWKZKDWDPRXQWVWRDGLI¿FXOWFRPPRQO\XQVWDEOHDQG COEVOLVING WITH THE LANDSCAPE? 77 sometimes treacherous, landscape. Cast in terms of historical exchanges and agreements with land spirits, the migration timeline may serve as a store of knowledge based upon trial-and-error testing of ecological thresholds and the carrying capacity of the landscape over several hundred years. Complimenting this interpretation, Talum villagers have at their disposal a wide range of narratives that portray the actions and implications—often negative—of the activities of individual hunters and cultivators with regard to spirits over many generations. In such accounts, cases of failure to respect historical agreements with spirits, ZKLFK OHDG WR VWRUPV ODQGVOLGHV ÀRRGV DQG RWKHU GDPDJLQJ µQDWXUDO GLVDVWHUV¶ ¿JXUH SURPLQHQWO\ 7KHVH HYHQWV QRW RQO\ GHVWUR\ KXPDQ wealth but also transfer it back into the domain or ‘houses’ of dispos- sessed spirits. Viewed from the perspective outlined above, Nyishi memories of migration may have ‘ecological’ value for people living in Talum village, comparable to the ‘social’ value of memories of past intra-human FRQÀLFW %RWK PD\ RIIHU JXLGHV IRU KXPDQ FRQGXFW DQG D µURDGPDS¶ IRUVRFLDODQGHFRORJLFDOFRQÀLFWUHVROXWLRQ,QWKLVZD\WKHFROOHFWLYH memory of problems encountered by the forefathers, and the exchanges and agreements with spirits that they made, may help to prevent villagers having to repeat the mistakes of the past in the present day. This would include, amongst other things, over-hunting, cultivating in areas that are likely to trigger landslides, degrading common land, causing crop failure by neglecting to time cultivation accurately. Historically, this may have motivated generations of hunters and cultivators, and storytellers, to remember the migration timeline. However, before we get too carried away with this interpretation we PXVWFRQVLGHUWKHIROORZLQJ0DQ\RIWKHKLVWRULFDOHYHQWVWKDWVLJQL¿- FDQWO\LQÀXHQFHSUHVHQWGD\SHUFHSWLRQVRIWKHODQGVFDSHLQ7DOXPYLO- lage occurred some six hundred or more years ago. If we are to accept the chronology of Nyishi storytellers, most such events occurred long ago. This begs the question, why bother to remember the intervening six hundred years in such detail, during which the forefathers of Talum village covered most of the distance between Narba and Koloriang? Should we place greater emphasis upon the ‘social’, rather than ‘ecological’, XWLOLW\RIPHPRULHVRIPLJUDWLRQ",IVRZKDWVRFLDOEHQH¿WVPLJKWVXFK memories confer? 78 ALEXANDER AISHER

POSSIBILITY II: MAINTAINING EXCHANGE WITH THE DEAD

Observe the following example of forefathers leading forward chronologically to the man Bengia, after whom the Bengia clan takes its name: Dodum, Dumgurr, Garrkik, Kiyfah, Fagah, Garchoh, Chohpu, Poyoh, Yoyurr, Yarmugh, Murmiy, Mibing, Bengia. Note that throughout the VHULHVWKHODVWV\OODEOHRIWKHIDWKHU¶VQDPHEHFRPHVWKH¿UVWV\OODEOH of the son’s name. This is a strong mnemonic device, and one that is very useful for a shaman-priest performing the prestigious Yuloh clan ceremony and chanting for hours or even days long lists of names leading from the distant past through into the present day. In this central ceremony of the Nyishi tribe, to be named is to be included, and to be included allows the remains of a person’s departed soul, their aram dwelling with other clan members in the House of Aram, Aram-Namlo, to exchange wealth with the living; to receive its share of WKHVDFUL¿FHGMXQJOHR[HQDQGWREOHVVWKHOLYLQJLQUHWXUQ,QGHHGWR die and be omitted from this list of forefathers is a tragedy reserved only for those who have broken critical taboos, for example, by having sexual intercourse with a fellow clan member. To be excluded from this most powerful exchange between the living and dead is something no Nyishi person, at least no non-Christian Nyishi person, desires. Might the logic underpinning this central ceremony of the Nyishi tribe have motivated generations of storytellers and shaman-priests to accurately recall the names of previous generations? Did memories of where these forefathers settled, where they migrated, and various events surrounding their lives, ride on the back of this ritual requirement to recall such genealogies? Might this have energised the process?

POSSIBILITY III: MAINTAINING CLAN NETWORKS IN AN AGE OF VIOLENCE

December 2002. I am trekking with my friend and assistant Bengia Chongpi back through the forest from the remote township of Sarli to Koloriang. We pick our way along the steep forest paths. A porter from the Sulung tribe with a bag of books slung heavily across his forehead follows silently behind. It is getting late, but we are still half a day’s PDUFKIURP.RORULDQJ$VGXVNEHJLQVWRULVHXSIURPWKHIRUHVWÀRRUZH realise we cannot reach Koloriang before nightfall. We need shelter for the night somewhere along the way. Half an hour later, with dusk now rising up through the trees, insects ratcheting up their nighttime chorus COEVOLVING WITH THE LANDSCAPE? 79 and the path growing dim, we enter the outskirts of Buyung village. Neither Chongpi, nor I, nor our Sulung porter, knows anyone in the village. But we need somewhere, anywhere, to stay. Tonight we must rely on the hospitality of strangers. From the path we watch smoke rise through the thatch of a dozen hunched longhouses. No one sits outside and no one comes to greet us. Then a man emerges from a longhouse not far from the path and strides through the mud towards us. This stranger worries us. Sporting one good eye and a large scar across his face, his manner is gruff and somewhat threatening. He points at a single-hearth longhouse on the outskirts of the village and urges us to stay. His aggressive body language and tone trigger strong feelings that his invitation is not entirely sound. I look over to Chongpi uncertainly. His face also betrays a certain reticence. However, the air is growing cold and the sweat from the day’s trek is already making us shiver. With nowhere else to stay, we agree to follow him down to his hut. The dark, broken-down interior of the longhouse XQGHUFXWVDQ\FRQ¿GHQFHZHKDYH7KHVSOLWEDPERRÀRRULVEDGO\LQ need of repair and rat faeces mark most surfaces. Two unhealthy chick- ens hobble around in the dark underbelly of the longhouse. Grateful for WKHZDUPWKZHVLWE\WKH¿UH “We need somewhere else to stay”, I say. Chongpi knows the etiquette and refuses. “No, he will feel otherwise.” “Do you know him?” “No.” “But he is a Bengia man?” “Yes, but distant—very little connection.” $VZHVLWWKHUHDQGVWLUWKH¿UHRXUKRVWH\HVXVVLOHQWO\+HVHHPV ever more menacing. An unpleasant night with an unknown stranger approaches. We sit there for some time in silence as the air of unease and suspicion thickens around us. There seems to be no common ground, no shared context, and nowhere else to stay. &KRQJSLLV¿UVWWREUHDNWKHVLOHQFH+HOHDQVIRUZDUG¿[HVWKHPDQ¶V one good eye and asks him where precisely their ancestry converges. (DFK PDQ VORZO\ FRXQWV EDFN WKURXJK JHQHUDWLRQV²¿YH VL[ VHYHQ eight, nine—until they reach a common ancestor. Upon discovering this ancestor the atmosphere changes. Chongpi makes the most of the discovery. Both men visibly relax. Chongpi exaggerates his surprise. I smile and try to express how glad I am that Chongpi has found a clan member here in this unfamiliar place. Suddenly, we all have something to work with; a shared context and, perhaps more important still, a shared 80 ALEXANDER AISHER stake in each other’s well being. So we begin an evening of meager rice and silty millet beer in the hungry longhouse of a distantly related clan member. December 2003, Koloriang town. I am standing on the resting platform at the back of a longhouse high up on a hill overlooking the township. Its owner, a local politician, looks out over the longhouses of Koloriang, the occasional corrugated iron roof glinting like an eye beneath the morning 6XQ+HWXUQVWRPHDQGVWDWHVÀDWO\³7DOH[WKH$JHRI9LROHQFHLVRYHU For Nyishi the Age of Violence is over.” Why mention these two events above? Because they point to another set of factors that may have motivated generations of Nyishi storytellers and shaman-priests to remember complex genealogies and associated migration timelines. It boils down to this: in a society ravaged by clan ZDUIDUHRUDVWKH¿UVWDQWKURSRORJLVWWRZRUNLQWKLVUHJLRQ&KULVWRSK von Fürer-Haimendorf (1982:300), described it, “rent by unending feuds”, a little knowledge of ancestry can be useful. &RQÀLFWDQGFODQZDUIDUHPD\ZHOOKDYHGULYHQWKHIRUHEHDUVRISHRSOH living in Talum village westwards in search of more land. The trauma DVVRFLDWHG ZLWK FODQ FRQÀLFW DQG DVVRFLDWHG G\QDPLFV RI DEGXFWLRQ compensation and revenge, may also help us to understand the manner in which malign spirits come by surprise and abduct human souls (yaji-yalu), how shaman-priests negotiate ritually for their release, and how land spirits take revenge upon hunters and cultivators who steal too PXFKRIWKHLUZHDOWK&RQÀLFWDOVRUHQGHUVDVYDOXDEOHWKHPHPRU\RI how a person is connected with other people and other clans, however far back through the generations they have to travel. In a society in which allegiance was, and still is today—albeit in a very different political context—indexed to clan membership, knowledge of ancestry can be a valuable type of social capital. In this way, clan warfare may well have contributed to the collective preservation of memories of migration by the Nyishi tribe.

CONCLUSION

Embedded as they are within a genealogical timeline, Nyishi accounts of migration are committed to a logic that moves forward towards diversity and backwards towards unity. Nyishi storytellers and shaman-priests carry ‘in their bellies’ a long history. Depending on whether they choose to travel forwards or backwards through time, this history has a wide range of potential uses. COEVOLVING WITH THE LANDSCAPE? 81

This history of migration over several hundred years through the uplands of the far eastern Himalaya not only chronicles the movement of a people—a tribe, clan and sub-clan—through a landscape, but also offers insights into how that people changed over time. At the level of environmental history, these stories of migration may also offer a history of the forest and landscape, and how these too gathered stories and changed over time. At a profound level, this history of migration is a history of exchange, with spirits and with the landscape that spirits cohabit with humans. It is through stories such as these that a sub-clan of the Nyishi tribe living in a small village half a day’s walk from Koloriang add to a history of their tribe, the Tani tribes, and the Tibeto-Burman-speaking peoples. This history is also a history of spirits, a small thread in the enormous and rapidly unraveling web of stories that make up the environmental history of Arunachal Pradesh, and the tribes living there. 82 ALEXANDER AISHER

BIBLIOGRAPHY

Dawkins, R. 2005. The Ancestor’s Tale: A Pilgrimage to the Dawn of Life. London: Phoenix. von Fürer-Haimendorf, C. 1982. Tribes of India: The Struggle for Survival. Berkeley: University of California Press. Nabam, Rikam. 2005. Emerging Religious Identities of Arunachal Pradesh: A Study of the Nyishi Tribe. New Delhi: Mittal Publications. Roy, N.C. 2005. Arunachal Pradesh Human Development Report 2005. Itanagar: Department of Planning, Government of Arunachal Pradesh. MICRO-MIGRATIONS OF HILL PEOPLES IN NORTHERN ARUNACHAL PRADESH: RETHINKING METHODOLOGIES AND CLAIMS OF ORIGINS IN TIBET

TONI HUBER

INTRODUCTION

The topic of origins and migrations has frequently featured in descriptions of the hill peoples of the far eastern Himalaya. This interest not RQO\UHÀHFWVORRPLQJXQDQVZHUHGTXHVWLRQVSRVHGE\RXWVLGHREVHUYHUV about where many of these populations originated. It also reveals the currency of interest in, and claims about, these same issues among local communities throughout the region. Unfortunately, most existing writings on this topic are unsatisfactory in that they offer only very speculative and sweeping—in terms of time and space—reconstructions based primarily upon claims found in local oral narratives. As a counter to this approach, herein I will be advocating thorough investigation of exactly how and why hill populations actually move from place to place in this part of the Himalaya, as well as the use of data derived from a broad array of sources. A case study of local population movements within the northern Subansiri River catchment of Arunachal Pradesh is offered as an example of a non-speculative approach to understanding migrations in the far eastern Himalaya, one that opens up different types of questions and hypotheses from those suggested so far. I am introducing the term ‘micro-migrations’ here to describe the types of human movements in the region that my data demonstrate, and also as a way to break with the existing discourse on ‘migrations’ in the far eastern Himalaya. Local oral narratives and scholarly writings which discuss the origins and migrations of hill peoples of the far eastern Himalaya often share the same propositions. Firstly, both types of sources tend to plot routes of migration between an assumed original homeland area or origin place and a present-day dwelling location; direction of movement and itineraries DUHRIVKDUHGLPSRUWDQFHKHUH6HFRQGO\WKH\ERWKFODLPLGHQWL¿FDWLRQRI contemporary populations with their purported ancestors from past times and distant places, with implicit and explicit claims of ethnic continuity. The occurrence of these shared propositions is no mere coincidence. In most texts on the topic, local oral narratives and their discourses have 84 TONI HUBER simply been adopted and uncritically incorporated by scholars into their writing. In more elaborate cases, local oral narratives about origins and migrations are articulated with written histories and cultural or linguistic data. Instances of this can be found in various treatments of the Adi complex of central-east Arunachal Pradesh by Sachin Roy, Tai Nyori, and Jogendra Nath, or in Robert and Betty Morse’s study of the Rawang further to the east.1 Examples of a less sophisticated order abound in the literature on many of the hill peoples of Arunachal Pradesh.2 While it is not uncommon for scholars to incorporate local oral narrative data into their accounts, this process itself demands both theoretical considerations and methodological cautions; neither seems to be much in evidence in the writings I am drawing attention to here. The status of oral narratives as historical data is of course a crucial issue when discussing a region in which almost all the languages of the diverse hill populations lack a script, and where the penetration of organised states and their discourses, practices and institutions is an experience only of the past two or three generations in many places. Among scholars who study hill populations in India’s northeast, research approaches to RUDOQDUUDWLYHVWHQGWREHTXLWH¿UPO\GLYLGHG2QWKHRQHKDQGORFDO and regional scholars from India widely endorse oral sources as both valid and valuable for recording or reconstructing ‘historical facts’ with virtually no reservations.3 On the other hand, the few western scholars working in the same region emphasise ongoing theoretical debates about oral sources, the cautions needed when using them, and their inherent limitations. This latter view holds that while local oral narratives may in some cases be used to reconstruct past events, their main importance lies in understanding the contemporary social contexts in which narratives are spoken or circulated, and their role as a strategic and adaptable resource for certain types of societies, such as non-literate hill peoples.4 A second unsatisfactory aspect of existing writings on origins and migrations of the hill peoples in Arunachal Pradesh is the poverty of sources they employ beyond repeating local oral narratives. A minor exception is that some authors have also introduced particular readings

1 Roy 1960:11-7; Nyori 1993:chapt. 2, and his map on p.59; Nath 2000:11-27; Morse & Morse 1966. On the actual arguments put forward in these sources, see Blackburn 2003/4. 2 For recent examples, see Riddi 2008, Pegu 2008, Mibang 1998, and for earlier examples, Bhattacharjee 1972, Dhasmana 1979:21-23 and Shastri 1969. 3 For instance, see most recently Bath 2008, Bhatttacharjee 2008, Billorey 1997, Biswas 1997, Borang 2008, Rikan 2008, and Singh 2008. 4 For instance, see Blackburn 2003, Blackburn 2003/4, Huber 2010, and Scott 2009:chapt. 61/2. MICRO-MIGRATIONS 85 of regional history to frame oral narratives; the use of secondary sources on early Tibetan history and the Bön religion to support claims of the origins of local hill peoples in neighbouring Tibet being a case in point.5 Clearly, the evidence that has most often been ignored or omitted by writers is that which can demonstrate what happens on the ground, that LVUHOLDEOHDQGYHUL¿DEOHGDWDUHYHDOLQJKRZZK\ZKHQDQGZKHUHUHDO people actually move through the eastern Himalaya. Such information about human movement might come from a number of sources: the recording of direct observations made over time (in written reports, sta- tistics, photographs, maps, etc.); oral reports by eye-witnesses that can EHFRQ¿UPHGRUFURVVFKHFNHGDJDLQVWDOWHUQDWLYHGDWDPDWHULDOWUDFHV and so on. These types of sources that can inform us about population movements are more abundant than one might imagine for eastern Himalayan regions.

CASE STUDY: MICRO-MIGRATIONS IN NORTHERN SUBANSIRI6

People and Region

,ZLOOQRZEULHÀ\GLVFXVVPRYHPHQWVWKDW,WHUPPLFURPLJUDWLRQV7KHVH are movements that have been made by various small highland communities who have been living in the northernmost parts of the Subansiri River catchment of Arunachal Pradesh, India, and sometimes also just to the north of the de facto India-China border zone7 ¿JXUH I will have most to say about the Mra, an exogamous mono-clan community, but will also mention their near neighbours, the Na and Nilo, as well as the clan-cluster dwelling around the present-day Naba settlement, which includes small numbers of Ngoju, Bai, Puri, and Kyali clan members. With the sole exception of the Na, who partly reside in the administrative unit of Taksing Circle on Indian territory and partly in the Doyü (Doyou on Chinese maps) settlement of the Chayül region of southern Tibet on Chinese territory, all these clans are presently located

5 See, for example, Blackburn 2003/4:23-25 for discussion of the ‘Tibet hypothesis’. 6 Since 2006, the research upon which this case study is based has been conducted within the project Between Tibetanization and Tribalization: Towards a New Anthropology of Tibeto-Burman-Speaking Highlanders in Arunachal Pradesh, funded by the Deutsche Forschungsgemeinschaft, Bonn. 7 The post-1914 McMahon Line between Tibet/China and India has become highly contested and militarized since 1959, and is perhaps better described as a ‘zone’; the now common Indian term Line of Actual Control (LAC) points to the highly contingent nature of any border in the region. 86 TONI HUBER

Figure 5.1. Map of the northern Subansiri region. within Limeking Circle on the Indian side of the border. Each of these communities contains between 200-400 persons living primarily in small, scattered settlements. There is every indication that their populations have remained relatively stable over the period that living memory can report, or for which lineage reconstruction is possible and census data available. All the communities are speakers of closely related Tani languages/ dialects (see the chapter by Post in this volume). The Mra and their neighbours are typical of hill societies found throughout the far eastern Himalaya both before and after the advent of the modern Indian and Chinese states in highland areas during the 1950s. They were and still are mainly engaged in swidden cultivation, hunting and gathering wild food. Cross-border trade with Tibetan partners was very important to the Mra and Na prior to the Sino-Indian border war of 1962, after which it came to a halt. More recently, a few northern Subansiri households also have access to a cash income by way of employment in a limited number of government posts, or from temporary contracts with state agencies (building or maintaining local LQIUDVWUXFWXUH¿UHZRRGFROOHFWLRQHWF DVZHOODVVWDUWLQJVPDOOSULYDWH businesses. According to the Government of India, all the communi- WLHVOLYLQJLQQRUWKHUQ6XEDQVLULDUHRI¿FLDOO\FODVVL¿HGDVPHPEHUVRI a Scheduled Tribe known as Tagin, although the application of this identity label is an administrative invention dating from the 1950s only. MICRO-MIGRATIONS 87

In reality, local clan and phratry identities are all that count within the region today. The name ‘Tagin’ is used with outsiders for convenience of recognition and strategic positioning within the state system.8 2QHVLJQL¿FDQWFRPPRQIHDWXUHRIWKHVHQRUWKHUQ6XEDQVLULSHRSOHVLV that their own origin and migration claims differ from the closely related ethnolinguistic communities who live further downstream in the same region. These northern groups mainly stress their initial descent to earth from ancestors in the sky,9 followed by a migration off of the Tibetan Plateau southwards to their present locations. This contrasts with many RWKHUQHLJKERXULQJ7DQLVSHDNLQJJURXSVZKRVWUHVVWKHWULFNVWHU¿JXUH Abo Tani as their originating ancestor (see both Blackburn and Aisher in this volume). Mra and their neighbours also maintain the narrative theme of migration and settlement generated by the social dynamics between sets (often pairs) of siblings—often brothers, sometimes incestuous brothers and sisters—a theme which recurs throughout much of the extended eastern Himalyan region (see chapters by Aisher, Blackburn, Gaenszle, McKhann and Wellens in this volume).10

Sources

We are able to reconstruct a certain number of micro-migrations that the peoples of northern Subansiri made over a period of approximately the past one hundred years. This is possible because the upper catchment of the Subansiri and its main tributary the Tsari Chu, and the peoples living in and using the area, regularly attracted the attention of outside observers for a variety of reasons. A major Tibetan Buddhist pilgrimage, the Tsari Rongkor, regularly traversed a section of the areas occupied and used by both Mra and Na; due to this, until 1956, both groups usually

8 The Na are the only local group to have actively contested the Tagin label, having petitioned the Arunachal Pradesh state government several times for Na (or ‘Nah’) to be recognised as a separate ‘tribe’. This is now recognised at the state level of administration, but not formally at the national level. 9'HVFHQWIURPWKHVN\QDUUDWLYHVRIYDULRXVW\SHVDUHPDLQWDLQHGE\VSHFL¿FVRFLHWLHV in what appears to be a distinctive zone of the eastern Himalayas. In western Arunachal Pradesh, they are spread from the Mra and Na area of northern Subansiri westward among the Levai/Bangru and Pukoik/Sulung in northern Kameng, in groups within the Monyul Corridor including Hruso/Aka and Bugun/Khowa, and into eastern Bhutan among ancient Dung (gDung) peoples, such as the Ura; see Huber 2010:308-9, n.18, Aris 1979:125-26, Deuri 1982:47, Kennedy 1914:1, Pandey 1996:15. Notably, most of these peoples speak QRQ7DQLODQJXDJHVQDPHO\WKRVHSUHVHQWO\FODVVL¿HGZLWKLQ+UXVLVK.KR%ZDDQG(DVW Bodish by van Driem (2001, II:473-481, 908ff.) and others. 10 On such narratives among the Mra and elsewhere, see Huber 2010. 88 TONI HUBER received a form of tribute from the Tibetan government in Lhasa.11 The Tibetan border settlement of Migyitün used Mra land at Longju for cultivation and thus Tibetans paid taxes to the Mra until 1959, while the Longju area marked the approximate position of the McMahon Line, and so both Tibetan and British observers collected information on the area and its inhabitants. The area has a very high rainfall, with complex vegetation communities and stark ecological transitions between wet Himalayan and dry Tibetan Plateau zones. It thus attracted the visits of a number of British naturalists, who also reported what they saw. Northern Subansiri was one of the last politically autonomous regions within the claimed boundaries of independent India to be penetrated and incorporated into the new state. As a result, we have a long series of Indian UHFRUGVEHJLQQLQJZLWKUHSRUWVRIWKHµ¿UVWFRQWDFW¶H[SHGLWLRQVRIWKH mid-1950s up until the 2001 Census of India and the current Electoral Roles. When I surveyed all of these sources together with extensive oral history interviews that I conducted among local clans between 2002- 2008,12 I was able to discover and cross-check movements into, out of, and around the region as a whole. The results of my research are summarised chronologically in the sections to follow.

Movements

'LVSODFHPHQWGXHWR/RFDO&RQÀLFW 7KH ¿UVW VLJQL¿FDQW SRSXODWLRQ PRYHPHQW LQ WKH UHJLRQ DSSHDULQJLQ RXUUHFRUGVZDVGXHWRVHYHUDOFRQÀLFWVEHWZHHQ7LEHWDQVDQGWKH1DD people whom the Tibetans refer to as Lungtu Lopa or Khalo.137KH¿UVW FRQÀLFWZKLFKZDVWULJJHUHGE\GLVSXWHVRYHUWUDGHPRQRSROLHVWRRN place in 1906, when 146 Na were killed by local Tibetans from Chayül and troops from the Kurab Namgye Dzong. The Na had originally resided at Lung on the Subansiri River (Chayül Chu), about 8 km above WKH0F0DKRQ/LQHRQWKH7LEHWDQVLGH7KLVFRQÀLFWGLVSHUVHGWKH1D and by 1919 the survivors had settled around Taksing, approximately 15

11 See Huber 1997, Huber 1999. 12 Elsewhere (Huber 2010), I have dealt extensively with local oral narratives that explicitly report notions of origin and migration among peoples of northern Subansiri. My study concluded that such sources vary according to the social location and interests of the tellers, the intended recipients, and the context of delivery, and that without an intimate knowledge of this context and the various languages involved, local oral narratives have very limited value for understanding migrations. 136HH+XEHUIRUDFRPSOHWHGLVFXVVLRQRIWKHVRXUFHVRQWKH1D7LEHWDQFRQÀLFWV and Tsarong Sharpé’s activities in the area. MICRO-MIGRATIONS 89 km downstream from Lung along the Subansiri and technically on the British Indian side of the McMahon Line as it appears on the maps. A second Tibetan military action against the Na at Taksing led by Tsarong Sharpé Dasang Drandul (1888-1959), commander-in-chief of the Tibetan army, followed at the beginning of the year 1920 with subsequent actions several years later. These attacks on Na were partly related to Tibetan attempts to establish experimental tea plantations in northern Subansiri, DQGSDUWO\GXHWRYLROHQWFRQÀLFWVZKLFKDFFRPSDQLHGWKH Tsari Rongkor pilgrimage. Once again the Na were dispersed, and by the late 1920s or early 1930s, the remaining Na population had resettled as two separate groups, one at Raprang, about 7 km above Lung on the Char Chu River, and the other around the Taksing area once again. These two Na populations still exist around the same sites today, on Chinese and Indian territories respectively.14 During the late 1920s, the warrior Mra Pusing, who was then the local ‘big man’ among the Mra, and his clansmen from the Tapuk lineage, had a feud with the Migyitün Tibetans. When Mra Pusing and his party were ambushed at Longju, Pusing, two of his followers and some female servants were executed by the Tibetans. This set in train a series of retal- iatory killings of local Tibetans by Mra. Prior to the feud, Mra Pusing and his clansmen had maintained a small settlement at Longju just to the south of Migyitün, mainly to enable trading and to enforce their claims over land in the area for use of which the Migyitün Tibetans paid them an annual rent or tax. Following Pusing’s execution, the Mra at Longju were displaced back down around the area known as Gelling 6LQ\LNQHDUWKHFRQÀXHQFHRIWKH7VDUL&KXDQG6XEDQVLULDGLVWDQFHRI approximately 25 km.15 It is worth emphasizing here that such complex political relations between Tibetans and groups of hill peoples in northern Subansiri, and at other points along the frontier between the Tibetan Plateau and far eastern Himalayan highlands, can readily be traced back centuries prior to the examples given KHUHE\XVLQJ7LEHWDQKLVWRULFDOGRFXPHQWV7KXVZHFDQKDYHFRQ¿GHQFHWKDW these have been much older and ongoing processes throughout the region.16

14 Following Chinese occupation of Tibet, the Na who remained at Raprang after 1962 eventually moved upstream about another 7 km to Doyü (Doyou on the Chinese maps). 150UD3XVLQJ¶VVWRU\ZDV¿UVWUHFRUGHGIURPKLVGDXJKWHULQODZE\*LWD.ULVKQDWU\LQ 1956, about 30 years after the events it describes (see Krishnatry 1956:entry for 5 March, “Lemeking-Ging”). Almost identical oral versions are still maintained by senior Mra males (interviews: Nyamen Mra, Logam Mra, April 2006; Bekab Mra, Taser Mra, Decmber 2006), and Tibetan informants from Migyitün (interviews: Wangdu Dorje, Wangdu Gompo, February 2007). 16 Events in the Tsari-northern Subansiri area (see Huber 1997, 1999, 2011) offer an 90 TONI HUBER

2. Bamboo Flowering and Migration Around 1950,17 D PDMRU EDPERR ÀRZHULQJ HYHQW RFFXUUHG LQ WKH Subansiri gorges throughout the whole of the area inhabited by the Mra, the Nilo and the clans around Naba, as well as many areas further downstream. The bamboo species concerned was Schizostachyum arunachalensis,18 locally known as tok, which is particularly abundant around Naba and areas downstream. All tok plants in any given area ÀRZHURQO\RQFHEXWGRVRVLPXOWDQHRXVO\LQPDVVÀRZHULQJHYHQWVDIWHU long intervals of 40-50 years; the last recorded (and remembered) tok ÀRZHULQJ LQ /LPHNLQJ &LUFOH ZDV LQ 6XFK ÀRZHULQJV DUH well-known for their profound ecological consequences.19 Flowerings produce enormous quantities of bamboo seeds—and tok has large seeds compared with other local bamboos. Seeds are consumed by jungle rats, which then breed rapidly due to the sudden availability of high quality food. When the seeds germinate and are no longer available to rats as food, mass infestations of these rodents then migrate into adjacent areas where planted crops, granaries and human habitations are located, and they consume every type of edible foodstuff in an area, which leads to human famine. High rodent numbers can also result in the outbreak of plague-like illnesses that infect human populations. The tokEDPERRÀRZHULQJHYHQWDURXQGLQQRUWKHUQ6XEDQVLUL resulted in both famine and disease for many local communities between Siyum and Naba, causing deaths and migrations out of the area. Although all Mra settlements also experienced an infestation of rats, they were H[DPSOHUHÀHFWLQJSKDVHVRI7LEHWDQUHOLJLRXVSROLWLFDODQGHFRQRPLFH[SDQVLRQVDORQJWKH IURQWLHUZKLFKZHFDQGRFXPHQWEDFNIRXURU¿YHFHQWXULHVZLWKFHUWDLQW\DQGWKDWKLJKO\ OLNHO\JREDFNDWOHDVWHLJKWFHQWXULHVLIQRWPRUH7LEHWDQFRQÀLFWVZLWK'XQJ J'XQJ peoples in southern Tibet and subsequent Dung migrations into Bumthang, Tawang-Monyul and other places south of the Tibetan Plateau are events datable with certainty to the mid.- 14th century; see Ardussi 2004, Aris 1979:chapt.5. As historical scholarship develops, we are likely to have more such examples available. 17 $PRQJ HLJKW H\HZLWQHVV LQIRUPDQWV WKH GDWLQJ RI WKLV EDPERR ÀRZHULQJ YDULHG between 1946, 1949, 1950 (four reports) and 1951. Northern Subansiri peoples had no pre-modern system for numbering years, thus calculating Western calendar equivalents LVRIWHQLPSRVVLEOH%DPERRÀRZHULQJHYHQWVDQGHDUWKTXDNHVDUHWKHWZRPDLQµUHPRWH¶ chronological reference points used by my elderly Mra informants to calculate their ages. 18 I thank Dr. R.C. Srivastava, Botanical Survey of India, Itanagar, for kindly identifying my specimens. 19 %DPERR ÀRZHULQJ HYHQWV LQ QRUWKHDVW ,QGLD WDNH SODFH DFURVV FDOHQGDU \HDUV ÀRZHULQJRFFXUVDIWHUWKHPRQVRRQ 6HSW2FW VHHGVKHGFRQWLQXHVIURP'HFHPEHULQWR May the following year; seed germination then commences in June when the monsoon begins, at which time seed is no longer available to rodent seed predators, and they move en mass into adjacent areas of cultivation and human habitation to seek food; see John and Nadgauda 2002, Nag 1999, Kumar and Pathak 2000. MICRO-MIGRATIONS 91 buffered against its effects since they practiced little cultivation and instead derived their income mainly from trade with both the Tibetans and their partners in the Kamla Valley. The Kamla peoples, who experi- HQFHGQREDPERRÀRZHULQJDQGWKHIRRGFULVLVLWHQWDLOHGFRQWLQXHGWR supply the Mra with all their staple food grains. The higher altitude Na area of Taksing was not affected at all by the crisis. Thus, the upstream Mra and Na territories were seen by refugees from Nacho and Naba as the only possible place to try and survive, and over one hundred persons from downstream clans (including Puri, Bai, Ngoju, Kyali, Serrah, Rai, Chokkar, Nalo, and Singming) eventually migrated up the Subansiri and resettled around Nilo, in the Mra areas of Nyaré and upper Ging on the south bank of the Subansiri, at the abandoned Mra site of Longju on the Tsari Chu which they resettled, and in Na areas around Taksing. A few migrants also went higher up into Tibetan areas when recruited as porters and labourers.20 The resettlement distribution of these migrants during the mid- WZHQWLHWKFHQWXU\EDPERRÀRZHULQJZDVQRWUDQGRP,QGLYLGXDOVDQG households from downstream clans who already had established contacts through friendship, marriage or trade relations with Nilo, Mra and Na went directly to their respective partner families seeking permission to move into nearby jungle and live by swiddening, gathering and hunting, or to join the domestic labour force as porters ferrying trade goods back and forth to Tibetan settlements. All of these migrants were displaced back down to their home areas again during the Sino-Indian war of 1962 along with most of their Mra and Na hosts. However, their dozen years living among Mra and Na resulted in quite a number of new marriages WR0UDDQG1DSDUWQHUV7KLVZDVVLJQL¿FDQWVLQFH1DFKRDQG1DEDFODQV were the least favoured marriage partners among Mra and Na prior to this time. Additionally, a number of these migrant households returned back to Taksing after 1983 and were absorbed into the Na community where they still live today.

3. Movement via Slave Trading Slave trading was a normal part of economic life among the communities of northern Subansiri. The practice was only effectively brought to a halt in the region by the Indian administration during the mid-1960s. The trade throughout northern Subansiri gradually moved almost all

20 Interviews: Chabé Cheddar, Jay Cheddar, March 2004; Tamey Ngoju, Logam Mra, Yaming Cheddar, March 2005; Talok Mra, March 2006; Yayak Mra, November 2006; Tabin Puri, Bekab Mra, Taser Mra, Nyamen Mra, “Bai Tara” February 2007. 92 TONI HUBER enslaved persons northwards towards the Tibetan border. Downstream slaves generally entered Mra and Na possession in two ways: as part of economic transactions involving Tibetan goods being traded southwards/ downstream; or as part of negotiated settlements for marriage contracts and compensation payments to resolve disputes. Mra households tended to keep few slaves during the mid-twentieth century since they were viewed as a potential security liability in maintaining the travel and trade blockade against downstream communities. Thus, most slaves acquired by Mra were traded directly up to Tibetan buyers at Migyitün. The Na, on the other hand, not only traded slaves into Tibet but kept quite a few in their own domestic labour force. While slaves could originate in neighbouring communities, the vast majority were persons from more distant communities who had been initially acquired as kidnap victims or prisoners during raiding, or who had already been passed on one or more times between a series of adjacent groups as payment for trade deals and negotiated settlements. Detailed data is available on slave holdings by the Na settlements around Taksing IRU WKH SHULRG ZKHQ WKH ,QGLDQ JRYHUQPHQW ¿UVW FRQWDFWHG WKH DUHD21 and this provides a good impression of the distances that slaves had been forced to migrate away from their original home areas. In 1957, Na households kept a total of 63 slaves (28 male + 35 female) who ZHUHLGHQWL¿HGDV%DQJQL 6XOXQJ LH3XURLN 7DJLQ DQG Tibetan (1). At least 41 of these slaves (Bangni and Sulung) are recorded as originating in the Khru Valley. At its closest, the Khru is 50 km as WKHFURZÀLHVIURP7DNVLQJDSUHPRGHUQODQGMRXUQH\ZKLFKZRXOG entail crossing at least two major ranges and one major river valley and passing through numerous separate clan territories. While many slaves were integrated into the household structure and worked and lived alongside other household members, they were subject to various social and ritual restrictions. Low status22 slaves could not marry into local clans and could only marry other slaves. Certain slaves who were from highly regarded families or lineages (typically kidnap victims or prisoners of war) could become socially integrated via marriage and permanently join local communities. Slavery was also intergenerational since the offspring of established slave families within a household also served their owners. Thus, through such mechanisms, many slaves subject to forced migration into northern Subansiri became 21 Sailo 1957:Appendix “C”. 22µ/RZVWDWXV¶LQWKLVFRQWH[WUHIHUVVSHFL¿FDOO\WRSHUVRQVZKRZHUH6XOXQJ3XURLNRU from known poor families (and thus not worth having alliances with), or from the families or lineages of known enemies. MICRO-MIGRATIONS 93 long-term or permanent residents. While all Na domestic slaves were freed by the government after 1965, a separate community consisting entirely of former Sulung/Puroik and Bangni slaves still resides in the Na area today.23 One informant who resided at Taksing for 4 years during the early 1950s reported that Sulung/Puroik males from the Khru Valley voluntarily migrated to Na settlements where unwed Sulung/Puroik female slaves were being kept in order to marry them and to join the Na domestic workforce.24 Thus, slavery could also generate voluntary migration.

4. Patrilocal Marriage $QRWKHUVLJQL¿FDQWIRUPRIKXPDQPRYHPHQWWKURXJKRXWWKHUHJLRQKDV resulted from marriage practices. The Mra have always been an exogamous community, and almost every married woman in all Mra households has migrated in from elsewhere. In larger polyganous households, RQHFRXOG¿QGFRZLYHVIURPWKUHHRUIRXUFRPSOHWHO\GLIIHUHQWUHJLRQV Similarly, virtually all Mra girls who married would migrate out to new patrilocal households in other distant areas. Prior to the 1970s, preferred Mra marriage alliances were nearly all arranged with trading partner communities and immediate strategic allies. Brides were exchanged especially with clans in the Soreng-Lingpu area of the Upper Kamla River, with the Na clans settled around Taksing, and less so with the Nilo and the Naba clans immediately downstream. Soreng-Lingpu was three days travel southwards from Mra lands across the high Ketch Pass, Na ZDVWKUHHGD\VWUDYHOZHVWZDUGVXSWKHGLI¿FXOW6XEDQVLULJRUJHVZKLOH Nilo and Naba were only one day’s travel away down the Subansiri. Each of the distant clans with whom the Mra preferred to exchange brides also maintained several additional favoured communities with whom they in turn forged alliances via marriage, and who themselves all lived HYHQ IXUWKHU D¿HOG (YHU\ HOGHUO\ 0UD ZRPDQ , LQWHUYLHZHG NQHZ DW least the home place of her mother and grandmother—but occasionally also great-grandmother. It is thus easily possible to trace how arranged patrilocal marriages with constellations of favoured partners continually moved women through a series of adjacent areas and also mixed each community via matrilineages.25

23 This is Redding village in Taksing Circle, with a population of 63 persons in 14 households during 2001; see Directorate of Census Operations 2006: village no. 00185000. 24 Interview: Tamey Ngoju, March 2005. 25 Fürer-Haimendorf 1947:84 gives evidence of family and community migration between marriage partner groups in the Mengo-Panior region during the mid-twentieth century. 94 TONI HUBER

5. Labour Migration With the incorporation of Mra and Na territories into the Indian state during 1956-1957, administrative headquarters (H.Q.) were established DWERWK/LPHNLQJDQG7DNVLQJ(DFK+4FRQVLVWHGRIDQ$VVDP5LÀHV post, a civil administrator and his support staff, with various agencies to build and maintain mule and porter tracks and bridges up to the McMahon Line boundary zone, plus supply depots for airdrops, and RWKHU IDFLOLWLHV $W WKH WLPH RI WKH LQLWLDO µ¿UVW FRQWDFW¶ H[SHGLWLRQV into the area during 1956-57, and prior to any direct contacts with the northern Subansiri populations, the Indian state had to rely on recruiting large numbers of local labourers from settlements further downstream of Limeking Circle to work as porters, guides, track-cutters and runners. These peoples, from areas such as Nacho and Siyum, had never or only very rarely ever visited the upstream areas due to the travel and trade monopoly with Tibet that the Mra had enforced for their own economic EHQH¿W7KHUH ZDV WKXV D KLVWRU\ RI PXWXDO KRVWLOLW\ DQG FRPSHWLWLRQ between the Mra and their downstream neighbours in Nacho and Siyum. Yet government penetration into the Mra area now meant that downstream groups could travel freely into the upstream areas due to their employment as labour corps for the administration. At the same time, the Mra were resistant—and mostly refused—to be engaged as menial labourers for the government because they had been enjoying a superior economic and political status due to their relations with Tibet vis-à-vis RWKHUFODQV7KH\FRQVLGHUHGVXFKZRUN¿WRQO\IRUVODYHV7KLVPHDQW that government agencies came to rely almost entirely upon porters and labourers from the poorer downstream areas of Nacho and Siyum. Initially, all such local government staff were merely transient residents at Limeking H.Q. However, by the late 1960s, many Nacho- and Siyum- based workers moved to and settled in Limeking.26 They resided there in houses they erected on small plots of land loaned to the government by Mra belonging to the Runyu and Runya lineages upon whose swidden area the Limeking H.Q. was established. Many of these workers married and had children, and these families still live in Limeking today.27

26,QWHUYLHZGDWDLVFRQ¿UPHGLQWKH&HQVXVRI,QGLDVHH%DUWKDNXU 71, 104-05, 108, 112-13: The total Scheduled Tribe (ST) population of Limeking H.Q. was 97, with 78 males and 19 females (note that in Mra villages sex ratios are close to even); total ‘workers’ was 110 male and 11 female (including non-ST persons), of whom 5 male and 11 female were ‘cultivators’ (note that in Mra villages ‘cultivator’ is the only recorded occupation for all workers), and 105 males were listed as engaged in ‘other services’. The VXUSOXVRIPDOH67QRQFXOWLYDWRUVLQWKHVH¿JXUHVDUHQRQ0UDODERXUHUVVHWWOHGDW/LPHNLQJ 27 The extent and origins of these non-Mra Scheduled Tribe residents in Limeking H.Q. MICRO-MIGRATIONS 95

/DERXUPLJUDWLRQKDVKDGVSHFL¿FVRFLDOHIIHFWVXSRQWKH0UDSRSXODWLRQ One development has been due in part to the presence of the children of migrant labourers from Nacho and Siyum alongside Mra children in the only government school in the area established at Limeking H.Q. in 1973. Parallel to the existence of the school, and other social factors, a strong increase in preference for love marriages has developed since the 1970s. A number of Mra love marriages in recent decades have been with partners from Nacho and Siyum migrant labourer families whom the Mra partners met and got to know as co-students in the Limeking school. Previously, Nacho and Siyum clans had been among the least favoured marriage partners for Mra arranged marriages.

,QWHUQDWLRQDO&RQÀLFWVDQG7HPSRUDU\0LJUDWLRQV Chinese military occupation of the Tibetan Plateau throughout the 1950s eventually resulted in a mass migration of Tibetan refugees south across the McMahon Line into Indian territory during 1959. Refugees from Tibetan border villages adjacent to the Subansiri who where trading partner communities for the Na and Mra sought and were given refuge among them. Thus, Chayül Tibetans became settled with Na at Taksing and Tsari Tibetans were settled with Mra at Ging. The Indian government permitted this settlement, and the Tibetans desired it. All parties needed to wait and see how the situation would develop, and there was initial hope that it might be possible for the refugees to return home again before too long. Any hopes of return were dashed in 1962 when WKH6LQR,QGLDQERUGHUFRQÀLFWEHJDQWREUHDNRXWDORQJWKH0F0DKRQ Line, and all the refugees who had migrated into northern Subansiri and remained there for 3 years had to move southward ahead of a Chinese invasion. They never returned to the region again, being resettled in RI¿FLDOUHIXJHHFDPSVLQIDUGLVWDQWORFDWLRQV With the 1962 Chinese invasion south of the Himalayan watershed, including the Tsari Chu and northern Subansiri valley down through the whole of Taksing and Limeking Circles, many Na, Mra, Nilo and Naba inhabitants migrated out of the area either down the Subansiri Valley to Taliha and Darporijo, or southwards across passes into the Kamla Valley where there were no Chinese troop movements. The remaining Mra deserted their settlements and migrated upwards into the dense is clear from the clan names from Nacho and Siyum Circles on the current electoral role, HVSHFLDOO\ 6LQJPLQJ 6L\XP 5DL 1DFKR DQG (EL\D VHH (OHFWRUDO 5HJLVWUDWLRQ 2I¿FHU 2008:2-9. 96 TONI HUBER forests on the higher slopes of the main Subansiri Valley. There they ZHUHZHOOUHPRYHGIURPWKHDFWXDO¿JKWLQJEHWZHHQ,QGLDQDQG&KLQHVH troops, which mainly occurred along the strategic government trail and around administrative centres like Limeking next to the river. Messages dispatched by both armies had made it known that neither side wanted WRLQYROYHRUKDUPWKHORFDOSRSXODWLRQLQDQ\ZD\GXULQJWKHFRQÀLFW provided they did not take sides and stayed out of harm’s way. Thus, Mra refugees only needed to wait out the short war in the high hills, where they easily survived by harvesting wild sago palm and tree fern pith and by hunting game. These refugees returned back downhill to their settlements along the river within weeks of the Indian retreat from the area, to be warmly welcomed by the temporary Chinese occupation IRUFH7KRVH0UDZKRKDGÀHGGRZQWR'DUSRULMRZLWKWKH,QGLDQWURRSV RUFURVVHGLQWRWKH.DPOD9DOOH\WRVKHOWHUZLWKWKHLUDI¿QHVDQGWUDG- ing partners were slower to return and in some cases stayed away from Limeking Circle for up to 6 months before moving back to their homes. Almost all the movements by local individuals and families in and out of and around the area during this period were transitory, and residential status quo was restored once again within half a year.

7. Roads, Resettlement and Migration One of the most profound and accelerating population movements throughout the eastern Himalayas in recent decades has been due to infrastructure development in the form of new roads extending into KLJKODQGDUHDV:KLOHWKHLQFUHDVLQJVRXWKZDUGÀRZRIUXUDOPLJUDQWV out of middle hill villages and down to growing administrative and commercial townships in the valleys and lower hills is a well-known phenomenon in Arunachal Pradesh, such migration has to date not taken SODFHLQQRUWKHUQ6XEDQVLULWRDQ\VLJQL¿FDQWGHJUHH7KHPDLQUHDVRQ IRU WKLV ORFDO ODJ EHKLQG WKH UHJLRQDO WUHQG LV WKDW WKH ¿UVW URDG RQO\ reached Limeking H.Q. a decade ago, and it has yet to be connected to Taksing H.Q. Nevertheless, the presence of the new road caused immediate micro-migrations to occur throughout Limeking Circle in the form of settlement relocation. 8S XQWLO WKH ¿UVW ,QGLDQ JRYHUQPHQW DGPLQLVWUDWLYH SHQHWUDWLRQLQ 1956, the main Mra trade route through the area up to the Tsari Chu and Tibet followed along the north bank of the Subansiri River. Apart from a few households of the Runyu and Runya lineages, all Mra maintained their settlements on the north bank since the south-facing slopes there provide sunnier sites for dwellings and swidden plots, not to mention MICRO-MIGRATIONS 97 a series of rock bluffs which allowed for highly defensible settlement locations in case of surprise attack. While this last strategic advantage ceased to be important after the Indian administration took control of the region, most Mra swidden plots continue to be worked on the sunnier north bank even today. The Indian government chose Limeking, on the south bank, as its administrative H.Q. for two reasons: it was the only open site in the otherwise steep, heavily forested upper valley where airdrops could easily be made and recovered; and it possessed a small, ÀDWDUHDIRUFRQYHUVLRQLQWRDKHOLSDG7KXVWKHSRUWHUDQGPXOHWUDFN to supply Limeking and points north was constructed along the south bank of the Subansiri to reach Limeking without the need for expensive DQGGLI¿FXOWEULGJHEXLOGLQJDFURVVWKHPDLQ6XEDQVLUL7KHQHZURDG of today follows this original government supply track along the south bank. As soon as the road got near to Limeking during the late 1990s, all north bank Mra settlements began to relocate themselves across to new locations on the south bank.28 Today, Mra abandonment of the north bank is virtually complete, with only a few lone households remaining there, and each of these is now serviced by its own durable hanging bridge and thus better connected with the south bank than at any time in the past. The readiness and speed with which this movement of dwelling occurred is not surprising when one considers that Mra and their near neighbours have no history of constructing permanent dwellings RUXVLQJ¿[HGYLOODJHVLWHV7KHPDLQUHDVRQVIRU0UDPLJUDWLRQDFURVV the Subansiri to the south bank are—as in so many similar communities throughout the highlands—the local wish for direct and easy access to transportation possibilities, goods and services, and new economic opportunities which the road has brought with it.

Complex Micro-Migrations

In this attempt to move beyond speculative and narrowly sourced discussions of origins and migrations among hill peoples of the eastern Himalaya, the types of data I have employed will never allow us to probe very far back in time. Nevertheless, the data can tell us with certainty DERXWZK\ZKHUHDQGZKHQVSHFL¿FKLOOSHRSOHVKDYHDFWXDOO\PRYHG 28,QWHUYLHZGDWDLVFRQ¿UPHGE\WKHERWKDQG&HQVXVRI,QGLD7KHPDS of Limeking Circle still marks Muri and Mepu, the major Mra settlements on the north bank of the Subansiri, as ‘villages with a population size below 200’ (Senapati 1995: Limeking Circle) while the 2001 map of Limeking Circle marks them (along with ‘Orak Camp’, formerly also on the north bank) as ‘uninhabited villages’ (Directorate of Census Operations 2006:90, map 24 (II)). 98 TONI HUBER

ZLWKDKLVWRULFDOGHSWKRIDERXW\HDUV7KH¿UVWIRXUH[DPSOHVRI PRYHPHQW , SUHVHQWHG WKRVH LQYROYLQJ ORFDO FRQÀLFWV EDPERR ÀRZ- ering events, slave trading and patrilocal marriage practices, are all phenomena we can safely assume have a longer history of occurrence throughout the region. The remaining three examples, involving labour PLJUDWLRQLQWHUQDWLRQDOFRQÀLFWDQGURDGEXLOGLQJDUHDOOHIIHFWVRIPRG- ern state formation in the region and clearly have no direct parallels in WKH¿UVWJURXSRIH[DPSOHV+RZHYHULIZHFRQVLGHUWKHZD\VLQZKLFK the hill peoples under study here have responded to all of these different phenomena—whether pre-modern or modern—in terms of their movements, some commonalities are evident. What we see in responses to these various push and pull factors are generally small numbers of individuals or smaller-sized groups moving relatively short distances, that is, movements I characterise as ‘micro-migrations’.29:HDOVR¿QGWKDWVXFK micro-migrations are not only simple or unidirectional: some moves can be temporary, with return back to the original sites occurring; temporary movements can produce a social—not to mention a linguistic—‘residue’ in the form of persons remaining behind at a temporary residence when RWKHUVUHWXUQWRRULJLQDOVLWHVRIGHSDUWXUHDOVRSHUVRQVFDQÀRZLQWZR or more directions as part of processes of socio-economic exchange, such as marriage and slaving, and so on. The century of known movements of northern Subansiri peoples presented in our case study equates to a complex whole in terms of their number, causes, distances, directions and destinations, and participants. This complexity contrasts strongly with the longer distance, unidirectional, en masse types of movements of hill people that have been described or assumed so often under the heading of ‘migration’ in the literature on the eastern Himalayas. Furthermore, in social terms, the results of all the movements documented above indicate long-term processes of ongoing mixing and diffusion, rather than the maintenance of any essential and tightly bounded ethnic corporate, such as the monolithic ‘tribes’ who populate the recent literature on Arunachal Pradesh. Another departure that our results take from the existing literature concerns the nature of push and pull factors. When a reason has been put forward to explain the migration of a particular eastern Himalayan group, one of the most common proposals by various authors is ‘increase in population’ and perhaps a subsequent ‘search for new land’.30 At least for northern Subansiri (see 29 Here, one might compare Fürer-Haimendorf’s 1955:160-61 observations on small- VFDOHJUDGXDOPRYHPHQWVRISRSXODWLRQZKLFKKHWHUPHGµLQ¿OWUDWLRQ¶LQWKH0HQJR3DQLRU region during the mid-twentieth century. 30 For examples, see Nyori 1993:41, Bhattacharjee 1972:23, Bhattacharjee 1983:14, Roy MICRO-MIGRATIONS 99 also the Bokar of northern Siyom below), we know that these factors have not played any role in precipitating micro-migrations.31

Rethinking Claims of Origins in Tibet

Our case study of the known movements of northern Subansiri peoples in relation to the neighbouring Tibetan Plateau should also cause us to challenge a persistent claim in the existing literature that many of the highland ‘tribes’ of Arunachal Pradesh must have had their origins in, and migrated from, Tibet.32 Authors making this claim mostly base themselves upon local oral narratives collected from hill peoples in the region. As a clear example of many such claims in the scholarly litera- WXUHZH¿QGWKHUHVSHFWHGOLQJXLVW7LDQ6KLQ-DFNVRQ6XQVWDWLQJ³$ number of facts suggest that the Tani speakers represent relatively recent waves of Tibeto-Burman migrations to Arunachal Pradesh…First, the migration routes recorded in the oral traditions of many northern Tani tribes, such as Ramo, Bokar, Tagin, and Simong, point unambigously to southern Tibet as their original habitat.”33 Other writers also freely interpret such narratives of migration using speculative readings of an older generation of largely superceded scholarship on the early history of Tibet, the Tibetan Bon religion, and so on. Our data reveal that small groups of hill people managed, when necessary or to their advantage, to migrate back and forth between lower hill and valley areas and higher altitude sites along the southern margins of the Tibetan Plateau. Such migrants came from societies of swidden cultivators and hunter-foragers with a material culture based heavily upon bamboo and other sub-tropical forest products, and dwelling in an alti- tudinal zone of ca. 1300-2500m. However, they were able to relatively easily adapt to living at altitudes up to around 3500m by selectively borrowing from and adapting Tibetan material culture and productive systems—typically in terms of dwelling construction, clothing and

1960:17, Sarkar 1999:1-2. All such statements can only be speculation, since there was no usable demographic data for the entire eastern Himalayas prior to the 1960s. 31 All peoples of northern Subansiri regularly rotate their use of swidden plots around different areas; however these rarely result in residential migrations, and the rotation cycle LVIDLUO\¿[HGDQGSUHGLFWDEOHDVIDUEDFNDVZHFDQDVFHUWDLQ+HQFH,GRQRWFRQVLGHUWKHP as micro-migrations. 32 Blackburn 2003/04 reviews most of these sources. 33 Sun 1993:10. Here Sun refers to Sachin Roy’s oft-cited 1960 work Aspects of Padam- Minyong Culture, which seems to have played a key role is shaping the ‘Tibet origins’ discourse among scholars, based upon Roy’s particular presentation of oral narratives. 100 TONI HUBER

Figure 5.2. Partly Tibetanized Na living near Taksing, northern Subansiri (photograph by L.R. Sailo, 1957). animal husbandry. While we have only discussed the Mra and Na (see ¿JXUH KHUHLQRXUFDVHVWXG\RWKHUSDUDOOHOH[DPSOHVRIVXFKVPDOO adaptive migrant groups can also be found along the entire far eastern Himalayan-Tibetan Plateau interface. Like the Na population of today, many of these transient frontier groups became divided between India and China after 1962, and now live separated on either side of the contested international border. $QRWKHUVXFKFDVH,NQRZRI¿UVWKDQGLVWKDWRIWKH%RNDUSHRSOHRI northern Siyom, who developed small and often semi-permanent enclaves in neighbouring Tibet along the Neyü Phu Chu just north of the Dom La pass, mainly in order to trade with nearby Tibetan settlements.34 Bokar adopted certain aspects of Tibetan dress (e.g., heavy woolen cloth,

34 By the mid-1950s, there were 15 small Bokar hamlets in the Neyü Phu Chu across the Dom La pass; Haldipur 1957: ‘Political’ annex following p.52. These people are often depicted as examples of the ‘Lopa’ ethnicity (Chinese: Luobazu) in Chinese publications about Tibet (see the photographs in Cai Xiansheng 1981:144-7), although they only represent one of several quite different ethnolinguistic groups encompassed by this blanket EXW RI¿FLDO minzu FODVVL¿FDWLRQ 2Q WKH SROLWLFDO DQG HFRQRPLF UHODWLRQVKLS EHWZHHQ WKH Bokar of northern Siyom and Tibetans, see Huber 2011:264-65. MICRO-MIGRATIONS 101

Figure 5.3. Bokar man wearing Tibetan wool tunic and nalong earrings, Monigong, Siyom (photograph by Toni Huber, 2002). and jewelry), albeit worn according to their own particular style35 VHH ¿JXUH DQG PDQ\ VSRNH VRPH 7LEHWDQ ODQJXDJH +RZHYHU Tibetans maintained a strict pre-modern marriage bar towards Bokar, which—together with trading—ensured a regular circulation of people up and down between the southern Monigong and northern Neyü Phu Bokar settlements. As a result of their Tibetan outposts, the Bokar became rather unique among neighbouring Tani-speaking hill peoples in the region prior to the 1960s in adopting a simple form of Tibetan diary production. Dairy products are in fact a culturally proscribed (ari) food for the Bokar themselves, and not traditionally consumed; the butter they produced was solely for trade and tax payments to their highland Tibetan neighbours. In addition to material culture, linguistic evidence provides another enduring and typical marker of strategic adaptation by such forest-dwelling, swidden cultivating hill peoples of Arunachal Pradesh WR KLJKHU DOWLWXGH 7LEHWDQ 3ODWHDX HQYLRUPHQWV 'XULQJ ¿HOGZRUNLQ 356HH¿JXUH%RNDUPHQDOZD\VZHDUSDLUVRIWKH7LEHWDQJROGVLOYHUDQGWXUTXRLVH earrings known as nalong (rna long) in southern Tibet. Whereas Tibetan men only wear a single nalong in their left ear, it is Tibetan women who always wear a pair. Similarly, the Tibetan sleeveless tunics made of heavy, dark wool and belted at the waist which Bokar men wear are primarily a woman’s garment in neighbouring Tibetan regions to the north. 102 TONI HUBER northern Siyom (2002), I discovered that the Bokar, who have a Tani language highly cognate with those of all their hill-dwelling neighbours, and one rich in terminology for swidden cultivation and hunting, have borrowed virtually all of their vocabulary for diary production directly from neighbouring Tibetan.36 Between 1959 and 1962, due to political tension and military action, some Bokar living north of the Dom La pass migrated southwards back onto Indian-controlled territory, while others remained in Chinese-controlled territory. Other, very similar examples of such pre-modern, shifting frontier populations can also be found scattered along the Tibetan Plateau-Arunachal Pradesh interface north of the Kamla River, in the upper Siang River, and around the Mishmi Hills region. Thus, rather than scholars uncritically claiming (and accepting local claims of) the origins and past migrations of eastern Himalayan hill peoples from Tibet on the basis of a handful of oral narratives, we can now propose a much more sound hypothesis based upon our current knowledge. In the past, some groups could indeed have migrated from the southern fringes of the Tibetan Plateau to their present territories in neighbouring highland regions of Arunachal Pradesh. However, such moves would have been part of longer cycles of shifting back and forth between higher and lower sites in response to a range of changing conditions, including known economic, political and ecological factors of the kind we have presented in our examples above. We can historically prove that certain factors propelling these movements are many centuries old in the region (i.e., political relations), and safely assume that others (i.e., recurrent ecological events) are too. Such movements back and forth between higher and lower sites, as we have ample evidence for in northern Subansiri, northern Siyom and elsewhere, do not offer any explanation for the origins or “original habitat”—to use Tian-Shin Jackson Sun’s phase—of highland Tani-speakers and their hill neighbours on the Tibetan Plateau. Nevertheless, local memories of more recent movement phases down from Tibetan areas southwards may have indeed been preserved in the surviving oral traditions which various authors have exclusively focused upon when claiming origins in Tibet for highland peoples in Arunachal Pradesh.37

36 For example, Bokar have borrowed ba lang, o ma, mar, mar khal, mar phye, mdong mo, ’o mdong and other terms for dairy production from neighbouring Tibetan speakers. 37 For example, it is hardly surprising that peoples in the Siang River basin here reported they came ‘from the north’, given the known history of regular incursions down the river by Tibetans and their Tibetanzied agents, which usually triggered displacements; see Huber 2011. MICRO-MIGRATIONS 103

This single case-study of northern Subansiri hill peoples cannot be simply simply generalised for the entire the entire eastern Himalayas. However, it does demonstrate the value of a different and more sophisticated method of documenting and understanding origins and migration in the region. A focus on relatively recent history within a limited geography enables us to look at the detailed and textured pattern of population movements and their causes. As small populations living right along the frontier, groups such as the Mra, Na and Bokar might be viewed as VRPHZKDWDW\SLFDO\HWSDWULORFDOPDUULDJHEDPERRÀRZHULQJHYHQWV VODYHWUDGLQJDUPHGFRQÀLFWDQGQHZURDGVKDYHDOOEHHQFRPPRQWR the expereince of most groups throughout the region. Similar studies of micro-migration may also help us to build up a more accurate picture of distant events and identify their underlying causes throughout the extended eastern Himalayas.

Appendix: Romanized Tibetan Names and Proper Spellings

Char Chu byar chu Chayül bya yul Chayül Chu bya yul chu Dom La dung la Khalo kha klo Kurab Namgye Dzong sku rabs rnam rgyal rdzong Lhasa lha sa Longju lung mjug Lung klung Lungtu Lopa klung tu klo pa Migyitün mi khyim bdun Neyü Phu Chu gnas yul phu chu Raprang rab ’phrang Tsari tsa ri Tsari Chu tsa ri chu Tsari Rongkor tsa ri rong skor Tsarong Sharpé Dasang Drandul tsha rong zhabs pad zla bzang dgra ’dul 104 TONI HUBER

BIBLIOGRAPHY

Ardussi, J. 2004. “The Gdung Lineages of Central and Eastern Bhutan - A Reappraisal of their Origin, Based on Literary Sources.” In K. Ura and S. Kinga (eds.), The Spider and the Piglet. Thimphu: Centre for Bhutan Studies, pp.60-72. Aris, M. 1979. Bhutan. The Early History of a Himalayan Kingdom. Warminster: Aris & Phillips. Bath, N. 2008. “Genealogy as a Source of History – A Case Study of the Apa Tanis.” In S. Dutta and B. Tripathy (eds.), Sources of the History of Arunachal Pradesh. New Delhi: Gyan Publishing House, pp.91-94. Bhattacharjee, J.B. 2008. “Folk Materials as Alternative Indigenous Sources for History of Arunachal Pradesh.” In S. Dutta and B. Tripathy (eds.), Sources of the History of Arunachal Pradesh. New Delhi: Gyan Publishing House, pp.11-16. Bhattacharjee, T.K. 1972. The Adis—Origin and Migration. PINES—Academic and Cultural Quarterly (Arunachal Number) 1(3): 20-23. Bhattacharjee, T.K.M. 1983. The Idu Mishmis. Itanagar: Directorate of Research, Government of Arunachal Pradesh. Billorey, R.K. 1997. “Oral History in North East India.” In S. Dutta (ed.), Studies in the History, Economy and Culture of Arunachal Pradesh. Delhi/Itanagar: Himalayan Publishers, pp.1-8. Biswas, N.B. 1997. “Historical Research in Arunachal Pradesh: Oral Source and Method.” In S. Dutta (ed.), Studies in the History, Economy and Culture of Arunachal Pradesh. Delhi/Itanagar: Himalayan Publishers, pp.9-14. Blackburn, S. 2003. Colonial contact in the ‘hidden land’: Oral history among the Apatanis of Arunachal Pradesh. The Indian Economic and Social History Review 40(3): 335-61. —— 2003/4. Memories of migration: notes on legends and beads in Arunachal Pradesh, India. European Bulletin of Himalayan Research 25/26: 15–60. Borang, G. 2008. “Oral Tradition: An Important Source of Adi History.” In S. Dutta and B. Tripathy (eds.), Sources of the History of Arunachal Pradesh. New Delhi: Gyan Publishing House, pp.41-52. Barthakur, J.K. (ed.) 1972. Census of India 1971. Series No.24 Arunachal Pradesh, Part X. District Census Handbook, Siang District, Part A & B. Government of India. Cai Xiansheng. 1981. “Ethnic Minorities in Tibet.” In Tibet. London: Frederick Muller Limited, pp.132-55. Deuri, R.K. 1982. The Sulung. Shillong: Research Department, Government of Arunachal Pradesh. Directorate of Census Operations. 2006. Arunachal Pradesh Administrative Atlas (Census Survey 2001). Itanagar: Census of India, Government of India. Dhasmana, M.M. 1979. The Ramos of Arunachal. A Socio-Cultural Study. New Delhi: Concept Publishing Company. van Driem, G. 2001. Languages of the Himalayas, 2 vols. Leiden: Brill. MICRO-MIGRATIONS 105

(OHFWRUDO5HJLVWUDWLRQ2I¿FHUElectoral Roll, 2008: State-Arunachal Pradesh. 22-NACHO(ST), Part 6: Basic roll of intensive revision, 2006 with supplement up to Special Summary Revision, 2008, (1) LIMEKING H.Q., (2) SERABA, (3) DOLORING AND (4) RAW. Government of India. von Fürer-Haimendorf, C. 1947. Exploration in the Eastern Himalayas: Diaries of Travels in the Subansiri area 1944 and 1945. Shillong: Assam Government Press. —— 1955. Himalayan Barbary. London: John Murray. Haldipur, R.N. 1957. 7RXU 'LDU\ RI 6KUL 51 +DOGLSXU 3ROLWLFDO 2I¿FHU 6LDQJ Frontier Division for the month of February/March, 1957. Arunachal Pradesh State Archives, Itanagar, File R187/57 [=12/57]. Huber, T. 1997. “Ritual and politics in the eastern Himalaya: the staging of processions at Tsa-ri.” In S. Karmay and Ph. Sagant (eds.), Les habitats du Toit du monde. Nanterre: Société d’Ethnologie, pp.221-60. —— 1999. The Cult of Pure Crystal Mountain. Popular Pilgrimage & Visionary Landscape in Southeast Tibet. New York & Oxford: Oxford University Press. —— 2010. “Relating to Tibet: Narratives of Origin and Migration among Highlan- ders of the Far Eastern Himalaya.” In S. Arslan and P. Schwieger (eds.), Tibetan Studies: An Anthology. PIATS 2006: Tibetan Studies: Proceedings of the Eleventh Seminar of the International Association of Tibetan Studies, Königswinter 2006. Andiast: International Institute for Tibetan and Buddhist Studies, pp.297-335. —— 2011. “Pushing South: Tibetan Economic and Political Activities in the Far Eastern Himalaya, ca. 1900-1950.” In A. McKay and A. Balicki-Denjongpa (eds.), Buddhist Himalaya: Studies in Religion, History and Culture. Proceedings of the Golden Jubilee Conference of the Namgyal Institute of Tibetology, 2008, vol.1. Gangtok: Namgyal Institute of Tibetology, pp.259-76. John, C.K. and R.S. Nadgauda. 2002. Bamboo Flowering and Famine. Current Science 82(3): 261-62. Kennedy, Captain R.S. 1914. Ethnological Report on the Akas, Khoas and Mijis and the Monbas of Tawang 7KH %ULWLVK /LEUDU\ 2ULHQWDO DQG ,QGLD 2I¿FH Collection, Eur Mss F 157 324(h). Krishnatry, G. 1956. Manuscript Diary by Mrs. Gita Krishnatry of an expedition to Upper Subansiri during winter of 1955-1956, in company of Maj S.M. .ULVKQDWU\ 3ROLWLFDO 2I¿FHU =LUR 1()$ [158 handwritten sides, plus 5 additional loose pages]. New Delhi: Nehru Memorial Library and Museum. .XPDU '. DQG .$ 3DWKDN %DPERR ÀRZHULQJ DQG URGHQW RXWEUHDNLQ North Eastern hill region of India. Indian Journal of Hill Farming 13(1/2): 1-7. Mibang, T. 1997. “Origin and Migration of the Tribes of Tirap.” In S. Dutta (ed.), Studies in the History, Economy and Culture of Arunachal Pradesh. Delhi/ Itanagar: Himalayan Publishers, pp.45-51. Morse, R. & B. 1966. “Oral Traditions and Rawang Migration Routes.” In B. Shin, J. Boisselier and A.B. Griswold (eds.), Essays Offered to G.H. Luce by his Colleagues and friends in Honour of his seventy-Fifth Birthday, vol.1. Ascona: Artibus Asiae, pp.195-204. Nag, S. 1999. Bamboo, Rats and Famines: Famine Relief and Perceptions of British Paternalism in the Mizo Hills (India). Environment and History 5(2): 245-52. 106 TONI HUBER

Nath, J. 2000. Cultural Heritage of a Tribal Society (The Adis). New Delhi: Omsons Publications. Nyori, T. 1993. History and Culture of the Adis. New Delhi: Omsons Publications. Pandey, B.B. 1996. The Buguns. A Tribe in Transition. Itanagar: Himalayan Publishers. Pegu, L. 2008. “Oral Sources of the Mising - A Study of Origin and Migration.” In S. Dutta and B. Tripathy (eds.), Sources of the History of Arunachal Pradesh. New Delhi: Gyan Publishing House, pp.53-69. Riddi, A. 2008. “Myth and Oral Tradition: Sources for Reconstructing Origin and Migration of the Tagins.” In S. Dutta and B. Tripathy (eds.), Sources of the History of Arunachal Pradesh. New Delhi: Gyan Publishing House, pp.71-81. Rikan, N.T. 2008. “Historical Value of Indigenous Oral Tradition: A Study of Nishi Tribe.” In S. Dutta and B. Tripathy (eds.), Sources of the History of Arunachal Pradesh. New Delhi: Gyan Publishing House, pp.31-40. Roy, S. 1960. Aspects of Padam-Minyong Culture. Shillong: North-East Frontier Agency. Sailo, L.R. 1957. Reports and Tour Diary on an Exploratory Tour Undertaken in the Upper Subansiri Area and the Tsari Chu Valley during January-March, 1957 (Secret). Arunachal Pradesh State Archives, Itanagar, File 47/57. Sarkar, N. 1999. The Tagins. Itanagar: Directorate of Research, Government of Arunachal Pradesh. Scott, J.C. 2009. The Art of Not Being Governed. An Anarchist History of Upland Southeast Asia. New Haven: Yale University Press. Senapati, T. (ed.) 1995. Census of India 1991. Series-3, Arunachal Pradesh, Part XII-A & B District Census Handbook (Maps Only) Upper Subansiri. Itanagar: Government of India. Shastri, B. 1969. “Note on the Tour of the Philologist in North-Eastern Siang in October-November 1968.” In Research on the Migration of the People of NEFA. Arunachal Pradesh State Archives, Itanagar, File 105/69, pp.1-9. Singh, N.J. 2008. “Reconstruction of Tribal History from Oral Tradition.” In S. Dutta and B. Tripathy (eds.), Sources of the History of Arunachal Pradesh. New Delhi: Gyan Publishing House, pp.25-30. Sun, T.J. 1993. A Historical-Comparative Study of the Tani Branch of Tibeto- Burman. Ph.D. dissertation, Department of Linguistics, University of California, Berkeley. APATANI IDEAS AND IDIOMS OF ORIGINS

STUART BLACKBURN

My starting point for this essay is a perception that, while a good deal has been written about the origins and migrations of various peoples in the eastern Himalayas, we know comparatively little about their own ideas on these topics. As Robbins Burling makes clear in his essay in this volume, those ideas may have been altered or indeed created by the interest of outsiders keen on establishing the provenance of populations over which they sought civil or spiritual dominion. Other essays in this book explore the accuracy of claims of origins in the eastern Himalayas, the historical, VRFLDODQGHFRQRPLFFDXVHVDQGFXOWXUDOVLJQL¿FDQFHRIWKHVHFODLPVDVZHOO as their role in the construction of identities and assertion of rights in the context of the nation-state. These various dimensions of current thinking on origins and migration enrich our understanding of the broader, earlier and ongoing debate about traditions reinvented, rediscovered or revived. As I began to think about the Apatani material in these terms, however, something seemed missing. I could write about Apatani claims of a Tibetan origin and the long migration of the ancestors before settling in the Apatani valley, at about 5,000 feet in the Indian state of Arunachal Pradesh. I could then go on to describe Abo Tani, the common ancestor for all the ‘Tani’ tribes in central Arunachal Pradesh and whose name is the Apatani autonym (tanii). These imagined origins and narratives of the ancestral past play a pivotal role in current identity politics within the state. This would be useful, possibly interesting, but would still not answer the question that I had not yet even been able to articulate. Unsure of my own misgivings, I then came across a collection of essays on origins in Austronesian languages and cultures. In his introductory chapter, James J. Fox writes: One of the perennial preoccupations in Austronesian studies has been with the tracing the origins of the Austronesians. Archaeologists, linguists and historians have all been concerned with this task. A less prominent concern has been to examine indigenous ideas of origin and how they function within Austronesian societies.1

1 Fox 1996:5. 108 STUART BLACKBURN

Fox goes on to say that while he and his colleagues have written about origins and migrations in this vast region—stretching from Malaysia and Indonesia to Micronesia, Melanesia and Polynesia—they had little knowledge of the local meanings of these ideas. Do Austronesian languages even have equivalents of ‘origin’ and ‘migration’? If so, how are they used? English-speakers talk of genealogical ‘trees’ and migration ‘routes’. What metaphors do Austronesians use? In looking through the work of his colleagues, Fox found that the concept of ‘origin’ is articulated through a variety of themes—ancestry, alliance, spirit contact, sacred objects—and in a variety of forms: dreams, chants, legends. Underneath this diversity, however, he identi- ¿HG D µFRPELQDWLRQ RI HOHPHQWV SKUDVHG LQ FRPPRQ PHWDSKRUV DQG cognate terms…[a] discourse on origins that is distinctly Austronesian.’2 A central element in this discourse, Fox concluded, is precedence, both the language and the practice of precedence. A similar isolation of the idioms of origins that are characteristically eastern Himalayan is the long-term goal toward which this essay takes a small step. Already, several colleagues have succeeded in identifying a few elements that are surely part of this complex. The ritual journey, for example, has now been shown to play a central role in the cosmology of many Tibeto-Burman-speaking populations in the region.3 Another shared idea of the past is an oral narrative about (typically two) brothers who compete, separate and settle in different locations as a migration story that explains relations to other—usually dominate and literate— groups.4 Similarly, a myth of multiple suns (and moons), often explaining the origin of mortality, and a legend of lost writing are also told by many populations in Nepal, Arunachal Pradesh, Nagaland, upland Southeast Asia and southwest China.5 Another kind of common element in a possible eastern Himalayan discourse on origins, and one which recurs in the essays in this book, is the symbolism of water. Other similarly widespread ideas and expressions undoubtedly exist, and a proper comparative analysis is needed in order to identify a common language and practice of origins and migration in the region, if indeed one exists. As a contribution to that comparative task, I present here my understanding of Apatani ideas and idioms of origins. I will say little about migration because, to Apatanis, it is secondary to more fundamental concepts about origins. While ideas about geographical migration are 2 Fox 1996:5. 3 Fürer-Haimendorf 1952; Allen 1980, 1997; Gaenszle 2000, 2002; McKhann 2003. 4 Oppitz 2006; Wellens this volume. 5 Blackburn 2008:215-18. APATANI IDEAS AND IDIOMS 109 not unimportant, I believe that claims of Tibet as the homeland of Apatanis are the result of political, economic and cultural changes set in motion in the mid-nineteenth century. In brief, increasing domination from Assam during this period enhanced a pre-existing but weak legend about migration from Tibet. Responding to this unwelcome advance of RXWVLGHLQÀXHQFHVIURPWKHVRXWKHVSHFLDOO\+LQGXLVPDQG&KULVWLDQLW\ Apatanis began to invest the largely unknown north with a greater role in their past.6 Apatanis, I found, mainly think of themselves not as people who came from a certain place but as people who are related to others; not as migrants, but as descendants. As a result, Apatani ideas of their own origins, as a human population, are expressed in terms of genealogy and not geography. Apatanis also speak about origins in a second idiom, which explains the emergence of the natural world as well as material objects. For lack of a better word, I will call this second idiom ‘cosmogony’. Finally, underlying both these expressions is a single idea, or process, of differentiation. This underlying idea of differentiation is expressed through the two distinct but related idioms of cosmogony and genealogy. Apatani ritual texts use a variety of images—division of a woman’s body, separation of earth-sky, biological birth, splitting of human populations—but all involve the break up of a prior whole into parts. Together, these idioms and images are the imaginative language that Apatanis rely on again and again when attempting to explain the beginnings of the world they see around them. These are tentative conclusions, an interpretation based largely on Apatani ritual texts and interviews with ritual specialists. Something as conceptual, almost ineffable, as ideas of origins is unlikely to be transparent in indigenous thinking, let alone an outsider’s attempt at analysis. Still, I hope that my interpretation stimulates others to look more closely into local ideas and idioms, so that we are able to build up a collective and more accurate understanding of origins and migrations across the eastern Himalayas. Although they are an upland population, Apatanis practice wet-rice agriculture in their small but fertile valley. Their population of about 35,000 is divided into roughly 80 patrilineal clans living in a handful of densely-settled villages. Apatani is a Tibeto-Burman language, closely related to others in central Arunachal Pradesh.

6 For more details, see Blackburn 2008:35-40, 228-30, 244-46. 110 STUART BLACKBURN

COSMOGONY

To go straight into that language (in translation), here is a composite Apatani account of the beginning of the world.7 (text 1) $W¿UVWWKHUHZDVQRWKLQJRQO\kolyung-kolo, or formlessness. This was the time of kami-kamo, or darkness. A power called Pinii Siyo brought forth a procreative female power called Nguntii Anii, and she gave birth to Earth-Sky, to a formless earth and to a formless sky. Many spirits emerged. Some of these were harmful (arii) and turned against Earth-Sky. Since the earth and sky were not yet fully formed, they were vulnerable to these bad spirits, who gave them diseases. Hirii obstructed the birth and growth of Sky-Earth. Giirii and Gyopu made them weak and lonely. Wi gave them stomach ulcers. Yachu made them suffer mental problems. Milya and Dopung gave them headaches. And Taisime infected them with venereal disease. This child, which we call Earth-Sky, was imperfect, born without proper form. Then Ami Dinchi Banyi was born as a sister to Earth-Sky. She tried to cure the Earth-Sky, but she was infected with venereal disease by Taisime. She married Kotu Butang-Korda Horming, the source of rats and squirrels, but her husband was also infected with disease through the feet. Next a priest named Kolyung Bumya Nyikang was born to mediate between Earth-Sky and the bad spirits. Popi was also born as an advisor, DQGKHGLUHFWHGWKHSULHVWWRGRDVDFUL¿FHWREDGVSLULWVLQRUGHUWRFXUH WKH(DUWK6N\$SLJZDVVDFUL¿FHGWR+LULLDGRJWR*LLULLDFKLFNHQWR Yachu; a bamboo piece to Milya and Dopung; a thread to Wi; a bamboo VKDYLQJWR*\RSXDELWWHUOHDIWR7DLVLPH7KHVDFUL¿FHVZHUHVXFFHVVIXO and Earth-Sky was free from disease. Now Earth-Sky was able to separate and assume two proper shapes. But there were no pillars to support them, and so the pillars were formed. The earth was formed from Nikun. Her hair became the nests for birds, and her buttocks became the bottom of the sea. Her eyes became clear lakes and her shoulders the broad horizon. Her chest became the roof of the world, and her heart became rocks and mountains. Her breastbone turned into forests, and her stomach turned into weeds and grasses. From another spirit, Kolyung Pinii, emerged the Sun, the Moon and the stars. Still another spirit, Chaha, brought forth the plants and vegetables on earth, and Doha did this for the sky. Finally, from the Earth-Sky came Chantung, a female force, who protects human beings and their souls. From her came the tanii people, or humans. She had many children, and WKHODVW1HKD7DQLZDVWKH¿UVWKXPDQEHLQJ+HLVDOVRFDOOHG$ER7DQL and we are his children, his descendants.

7 Based on oral accounts from Hage Hiiba, Hage Tapa, Padi Tasan and Mudan Donyi (recorded between 2002-2005) and a written account in Takhe Kani 1996/1997. APATANI IDEAS AND IDIOMS 111

This, let me emphasise, is not ‘the Apatani creation myth’. No such text exists. Apatanis do not tell a single story of the beginning of the world. They speak, instead, of the formation of the world as an episode in different ritual texts, using different names and events. Nevertheless, there is enough consistency in them to identify common images and themes. These are the recurring features of Apatani cosmogony, which we can see in text 1 above. Taken together, these images and themes tell a kind of story, but not of creation ex nihilo. Instead, Apatani speak of an emergence from a pre-existing ambiguity, or formlessness, called kolyung-kolo.8 Kolyung-kolo is the starting point for Apatani ideas of origins. It is the inchoate state that precedes the world, and it is the source of everything within it. It is also, as the text says, a time of kami-kamo, a fertile darkness that holds the seeds of life. The unfolding of that formlessness is the central idea in Apatani thinking about origins. More than a narrative theme, it is an active process, an exfoliation into birth, growth and maturation, best captured by the Apatani word doging, which I translate as ‘differentiation’.9 Looking back at the origin story above (text 1), we see several images of cosmogonic differentiation. Out of the inchoate emptiness (kolyung- kolo) comes a female power, a spirit (wi), here named Nguntii Anii. She has various names, and even in this brief text she—or the power she represents—reappears as Nikun and Chantung. Sometimes, as here, she is preceded by a spirit of unknown gender, Siyo Pinii, often a collective term for spirits, who is quickly ignored as the story continues. The important point, however, is that this procreative female power gives birth to earth-sky. Various spirits also emerge, including dangerous ones (arii) that bring diseases to the imperfectly formed earth-sky (siican-mudo). When other, more benevolent, spirits, especially Popi, WKH HPERGLPHQW RI ZLVGRP SHUIRUP VDFUL¿FHV RI SLJ DQG GRJ WR WKH VSLULWVWKHPDOIRUPHGHDUWKVN\¿QDOO\GLYLGHVDQGLVFXUHG A similar separation of earth and sky as a primary act of creation is a commonplace in world mythology (Egyptian, Maori and Chinese traditions are well-known). A few examples are reported in central Arunachal

8 Kolyung-kolo has cognates in Tani languages: kol/korium/kirium-kulu among Nyishi/ Bangni (Bora 1995:1; Aisher 2006:147, this volume; Tob Tarin Tara 2005:2; Taring 2006); keyom kero among Adi (Mibang and Abraham 2002:15-18); kerium kelu and kolo kolung among Tagin (Mitkong et al. 1999:75); and korom among Galo (Dutta et al. 1999:47). 9 The Nyishi word dogging apparently means both ‘place to rest’ and movement in search of a ‘secure place’ (Taring 2006). 112 STUART BLACKBURN

Pradesh, and better documentation would surely reveal more.10 In fact, for Apatanis this separation of upper and lower spheres appears to hold VSHFLDOVLJQL¿FDQFHVLQFHLWH[HPSOL¿HVWKHEURDGHUSURFHVVRIGLIIHUHQ- tiation that underlies all origins. Moreover, this separation is linguistic as well as cosmogonic. ‘Earth- sky’ is a compound, or what I call a ‘noun-pair’, a kind of parallelism that is found in the ritual speech of other Tibeto-Burman-speakers in the eastern Himalayas.11 Apatani ritual speech is dominated by these noun- pairs. We have already come across a few abstract ones (kolyung-kolo, kami-kamo) and will see others below that refer to spirits, rivers and human-animal pairs (Pyokun-Pyopa, Supyu-Gyayu, Nikun-Sukun). The distinctive features of this noun parallelism in Apatani ritual speech are not important here. What we need to understand is that (like all these noun-pairs) ‘earth-sky’ is a whole containing two fused elements that later separate. Once the earth and sky have separated, a further unfolding of formlessness takes place. A female spirit or power known as Nikun appears, and from her body come various parts of the natural world. While a pri- PHYDOPRWKHU¿JXUHLVQRWXQFRPPRQLQFRVPRJRQLHVLQPDQ\UHJLRQV including the eastern Himalayas, this explicit image of her divided body EHFRPLQJVSHFL¿FSDUWVRIWKHQDWXUDOZRUOGLVOHVVZLGHVSUHDGWKDQRQH might think.12 It is, however, the most frequently repeated image of differentiation in Apatani descriptions of origins. It is heard in nearly every ritual text, often more than a dozen times, and usually in greater detail WKDQLQWH[WDERYH2FFDVLRQDOO\WRRKHUELUWKLVFRQÀDWHGZLWKWKH separation of earth-sky, as hinted in the same text. In other texts, the association is more explicit. In the Subu Heniin chant, for instance, at birth she is malformed and beset with maladies. She grows too quickly, in the wrong directions, and cannot be cured. After a shaman heals her, her two halves separate out into the upper world of spirits and the human world below: (text 2) Nikun lay down and part of her became the world of Pyokun-Pyopa spirits; She lay down and

10 The separation of earth-sky (with cognate words for siican-mudo) is reported in Nyishi/ Bangni (Aisher, this volume), Galo (Dutta et al. 1999:48-53), Hill Miri (Baruah 1999:103) and Miji, on the western edge of the region (Elwin 1958:14-15). 11 More details are found in Blackburn 2008, 2010. 12 I have found clear parallels reported only among the Thulung Rai (Allen 1980:2), Tagin (Mitkong et al. 1999:95) and Bangni/Nyishi (Aisher, personal communication, 2008). APATANI IDEAS AND IDIOMS 113

KHUXSSHUERG\EHFDPHRXUÀDWHDUWK Her long hair became soft nests for wasps; Her nose turned into large nests for kites. Her buttocks became WKHRFHDQÀRRU Her arms turned into FKDQQHOVZKHUHZDWHUÀRZV And her neck became the wide horizon of the rising sun. Her eyes became kar simi, the white lake in the mountains which turns viewers blind. Her heart became duri yalang, the stone near the lake which swallows those who sit there. Her knees became the pillars of earth, KHUKDQGVLWVÀDWVXUIDFH Her chest became the roof of the sky, KHUEUHDVWERQHWKH¿UPDPHQWDERYH Her hair became green grasses covering the earth, her sinews the creepers in the forest; Her stomach became vegetation and plants, her liver the even surface of earth. This image of the female body as a protean source is deeply rooted in Apatani thinking about origins. These lines, or a version of them, are a formula used whenever the Apatani shaman wishes to explain the origin of almost everything, from a winnowing basket to swift rivers. Even in prose accounts, Apatanis use this image to explain the origins RIHYHU\GD\REMHFWVWKH¿UVWZHDYLQJWKUHDGVIRUH[DPSOHZHUH+LQWL Anii’s hairs, and the dyes were originally mixed in a pot made from her head. The female body is also the source for most objects of ritual VLJQL¿FDQFH7KHZRPDQ¶VOXQJIRULQVWDQFHEHFRPHVWKHEURDGÀDW leaf (niji) placed in front of a shaman while chanting; her toe turns into the special bamboo (tajer) that he holds in his hand; and her forearms transform into the wooden posts (gyadi WRZKLFKVDFUL¿FLDODQLPDOVDUH tied in the land of the dead. In these examples, we can see that some attention is paid to matching the shape of body parts to objects. The full power of this idiom, however, is unleashed when shamans describe the formation of the natural ODQGVFDSH$JDLQZKLOHRXUWH[WDERYHEULHÀ\PHQWLRQVWKHJUHHQLQJRI 114 STUART BLACKBURN the earth (by Chaha, yet another multiform of the procreative female), other chants are more expansive. In the long Subu Heniin text, after Nikun gives rise to mortars and pestles, and other domestic implements, the shaman lists the various parts of the visible landscape that formed from her: (text 3) Rivers and valleys, hills and streams appeared on earth: The Siya Balyi and Ngiira Piisa streams, the Padu Tiike and Supyu Gyayu rivers. Fields and forests, ravines and hollows formed. They appeared at Karin Parin, Hambo Halo and Ngenci Peji, Liri Lilya and Patu Page, Panyu Pare and Yacu Yaja, Siiro Sango and Soka Petii, ….. [many more similar lines are omitted]

GENEALOGY

While the Apatani story of cosmological origins has no linear narrative, the sequence of events is more or less consistent in all tellings. Kolyung- kolo is followed by the separation of earth-sky, concurrent with or followed by the birth/appearance of one or more female spirits, whose body separates out into domestic and ritual objects, the sun, moon and stars, as ZHOODVGR]HQVRIQDPHGULYHUV¿HOGVIRUHVWVDQGPRXQWDLQV7KH¿QDO stage of this differentiation is the appearance of humans, who usually (as in our text) come from Chantung, another of the various names for the procreative female. Chantung, unlike other women in Apatani mythol- RJ\ LV D ¿JXUH LQ KHU RZQ ULJKW ZLWK D WRXFK RI DQWKURSRPRUSKLVP She plays a key role in the popular narrative of Abo Tani’s life, and she is believed to protect human and animal souls by keeping them in a bamboo basket held in the right hand corner at the back of an Apatani house. Each year, that protection is renewed by a shaman who chants and sprinkles chicken blood on the basket. Newly-wed couples are also blessed with a Chantung ritual. In the Apatani origin story (text 1), Chantung arises from the separation of earth-sky and then gives birth to humans. After a number of FKLOGUHQ XQPHQWLRQHGLQWKHWH[W ZKRZHUHXQ¿WWREHFRPHKXPDQ VKH¿QDOO\JLYHVELUWKWRDIXOO\KXPDQ¿JXUH+HLV1HKD7DQLL RU$WR APATANI IDEAS AND IDIOMS 115

Neha) but is also called Abo Tani. Here is the genealogical expression of the differentiation that we have just seen articulated in cosmogonic terms. This second idiom of origins, however, is more complex than our abbreviated text suggests. Apatanis describe the process of genealogical differentiation in three distinct phases. First comes the actual emergence of human beings, which like all beginnings in Apatani thinking, has a pre-existing form. Abo Tani did not spring fully-formed from Chantung, as Athena from the head of Zeus. He was, instead, part of a prior compound: in ‘the time of the ancestors’, Abo Tani was half-human, half-spirit. His double- LGHQWLW\LV¿[HGLQWKHIDFWWKDWKLVIDWKHULQODZ EURWKHURUEURWKHULQ law) and arch rival is the leader of the spirit world. And as a legacy of this ancestry, he retains the ability to see the spirits, who are invisible to ordinary humans (except shamans). Abo Tani’s life-story, which is the best-known narrative across the whole of central Arunachal Pradesh, is a transformation from this protean ambiguity, in which he wanders without home or family, to a VHWWOHGH[LVWHQFHDVWKH¿UVWDQFHVWRURI$SDWDQLVDQGDOOWKH7DQLWULEHV13 It is the story of how humans competed with and then differentiated themselves from their spirit brothers. Like the cosmogonic pair earth- sky (siican-mudo), this spirit-man pair (wi-myu) separates out into two distinct entities, each with a discreet location: spirits in the forest, man in villages. Another story, not found in the origin text (text 1), explains a similar split between man and mithun. Nor is this random since the mithun is the principal medium of exchange between humans and spirits. At one level, this is a version of the widespread narrative in the eastern Himalayas, mentioned earlier, about competition between two brothers.14 Apatanis, however, do not tell it to explain why two, closely related groups divided and live in separate terrains. Instead, they tell it to explain how man and DQLPDOVHSDUDWHGDQGDVVXPHGWKHLUUROHVLQWKHVDFUL¿FH Nikun and Sukun are sisters, who compete to see who can prepare the EHVWULFHEHHUNQRZLQJWKDWWKHORVHUZLOOEHFRPHWKHVDFUL¿FLDODQLPDO One sister is indifferent, uses wild grass to make her beer and becomes the mithun. The other sister ferments rice and becomes the man who will now kill that animal. With this story, part of a day-long chant that pre- FHGHVWKHVDFUL¿FHRIPLWKXQV DQGFRZV WKHVKDPDQDVNVWKHDQLPDOV

13 Abo Tani also gives his name to the ‘Tani’ branch of the Tibeto-Burman language family, spoken in central Arunachal Pradesh. 14 See footnote 4 above. 116 STUART BLACKBURN not to blame him or the ritual’s sponsor for their deaths. Another detail is that the two sisters are referred to as Nikun-Sukun, a noun-pair that is simultaneously whole and yet divided. As with our previous examples (earth-sky, spirit-man), man-mithun must separate. The second phase of genealogical differentiation is the division of the ancestors into separate peoples. This is the story of Abo Tani’s descendants, who, from an Apatani point of view, include all known human populations. Apatanis place these groups into three categories: tanii (Apatani autonym), misan (other tribes but primarily Nyishis and Hill Miris, the Apatanis’ nearest neighbours) and halyang (non-tribal outsiders, principally Assamese). Apatanis also recognise the other tribes of central Arunachal Pradesh as ‘children of Abo Tani’.15 Finally, Apatanis also speak of ‘Nyime’, which refers to the region near the high Himalayas, and possibly beyond (Tibet). However, Apatanis have no ethnic term for ‘Tibetans’. Like the story of cosmogonic origins, Abo Tani’s genealogy is seldom told as a self-contained narrative and is usually embedded within ritual texts. In fact, Abo Tani’s ancestry is more than a story and is far more widely known than the division of a procreative female or the separation of earth-sky. His life-history is a piece of cultural knowledge, an accepted version of the past, that is implicit in popular prose tales and ritual texts alike. At the same time, this shared ancestry is described, or summarised, in three ritual texts. Two texts16 are accounts of the ancestors’ migration from ‘up north’, or Nyime (Tibet), to the Apatani valley. The ancestors travel across uniden- WL¿DEOHKLJKPRXQWDLQVDQGVZLIWULYHUVVWRSSLQJDWDVHULHVRISODFHV (called supung), sometimes for generations. At the most important of these places, the original stock of ancestors divides into separate groups: (text 4) 7KHOLIHRIRXUDQFHVWRUVÀRXULVKHGDW0XGR6XSXQJIRUPDQ\JHQHUDWLRQV But some people decided to migrate elsewhere to start a new life. So they again conducted the turi tunii divination ritual on a lapang [ritual platform] in order to decide which direction to take. The ritual, performed by an old woman and an old man, advised them to take the Landu and the Lacho Leyu paths, and then the Chiilyang and Kiipu Pingo paths. Now the ancestors split into smaller groups and followed separate routes.

15 To Apatani, the Tani group includes Apatani, Nyishi, Hill Miri, Tagin, Sulung, Aka, 0LML*DORDQG$GL,Iµ7DQL¶LVGH¿QHGE\WKRVHZKRWHOOVWRULHVRI$ER7DQLWKHQ$NDDQG Miji are excluded. Curiously, Sulung, Miji and Aka are the three languages excluded from WKH SURYLVLRQDO FODVVL¿FDWLRQRIWKH7DQLJURXSE\OLQJXLVWV 16 Rego Ayu and Nyime Ayu. APATANI IDEAS AND IDIOMS 117

Abo Tani divided things among his descendants. The people of Nyime got bitter leaves and wild animals; the Sulung got the sago plant; the misan got millet and wild bamboo. The halyang got betel nut, salt, brass plates and special cloth.17 We tanii people [Apatanis] got rice, the pine tree and domesticated bamboo. When the halyang left the forest and went down the mountains toward the plains, they cut down a tree to block the path of the misan and the tanii people. That’s why we stayed in the hills. Later, the teller explains how Apatanis themselves separated and travelled along different paths to reach their villages in the Apatani valley. (text 5) Here our ancestors again split into smaller groups and followed the paths that eventually took them to the villages in the Apatani valley. Our people of Hao took the Silo path, which led up from the river bank to the high mountain, and later they continued on the Pyutu route. The people of Hong chose the lower route, the Siike path, which followed course of the river and led to another route known as Supyu Gyayu. The people of Diibo crossed the river and followed the Chiilyang route. A few of these people, the sons of Aba Pigu, lost their way but eventually found the Chiilyang route. The people of Hari, travelling on the Pyutu route, faced resistance from the people of Tabyang-Talyang, but our ancestors defeated them. A more complete statement of Abo Tani’s ancestry is found in the Subu Heniin chant performed during the winter feasts. After describing the ancestry of the spirits, the shaman comes to a section called the ‘Genealogy of Abo Tani’ (tani diiri bunii), which opens with his birth: (text 6) Chanii Chantung married Chanka Riman and gave birth to Yama Tani; This was the man also known as Abo Tani. Abo Tani married Diilyang Diibu, who had many children, Tarin and Tati, Khatii and Khotii, and Nici and Nica. The shaman then describes the life of Abo Tani (his many wives and deprivations), which culminates in the birth of his son, Ato Neha (Neha Tani): (text 7) A child was born to Abo Tani,

17 This association of brass plates with the halyang is not consistent with local belief that they are Tibetan. However, many of these plates are manufactured in Assam, and their high value is naturally linked with the ‘wealthy’ halyang. 118 STUART BLACKBURN

and he was named Ato Neha. +HZDVWKH¿UVWKXPDQ WKH¿UVWGHVFHQGDQWRI$ER7DQL Ato Neha married Lorii Biinyi of Tangu, who gave birth to the eagle, the tiger and the jungle cat.18 Then Ato Neha married Lorii Binyi of Cha, whose son was the buru.19 But none of these children was truly human. Next, the shaman describes the generations that descend from Ato Neha. The ancestors of the halyangDQG6XOXQJDUHFRYHUHGEULHÀ\LQDIHZ verses, while those of the tanii (Apatani) and the misan (their neighbours) require many more lines: (text 8) I am the Aba Nibo shaman, standing on this lapang; I am the Ato Doni priest, chanting in the early morning. I will explain the full genealogy of the Dolo and Doding people, of the misan and the tanii. The ancestor Ato Pombo was the son of Ato Neha and the grandson of Abo Tani. He married Donyi Lota, whose daughter was Lota Lobya. But she was not suitable to become a Doding [Apatani] ancestor. Then Ato Pombo married Anii Khentu, who gave birth to Niitu; Ato Pombo married Anii Khempu, who gave birth to Niipo; Ato Pombo married Anii Yalang, who gave birth to Ayen Tacang. After more than an hour of chanting similar lines, he brings each set of ancestors along a separate route to settle in a village (as in text 5). The third and last phase of the genealogical differentiation described by Apatanis is a village-by-village listing of ancestors. Here is a sample, 18 The tiger, jungle cat, and eagle are considered close relatives of humans; when one is killed, a ritual is held to placate its soul, just as the soul of man is placated when he is prematurely and/or wrongly killed. 19 The buruLVVDLGWREHD¿HUFHDTXDWLFDQLPDOVLPLODUWRDFURFRGLOH APATANI IDEAS AND IDIOMS 119 again from the Subu Heniin text: (text 9) I invite the ancestors [khalo] of Tajang village to attend this feast; I summon the ancestors of Ditii Latii20 to come to this Murung. I call Diyu Solo of the Tabyu clan, Nembu Nenda of the Tage clan, Doni Hiki of the Tage clan, Nembu Nenda of the Tage clan, Miso Halyi of the Miso clan, Shaha Habung of the Radhe clan, Kani Tari of the Ngilyang clan, Pai Che of the Millo clan, Taker Bumer of the Mihin clan, and Rubu Chigin of the Rubu clan. Listen, all you ancestors; come and taste our rice-beer; Come, all of you, and make the jilo lyayu spirits happy. When the shaman recites the full list of clans in the seven villages, he comes to the end of the genealogy recognised by most Apatanis.21 Moving backward through the series of texts given above, we can see that these approximately 80 clans, about 35,000 Apatanis, belong to the common stock of ancestors that divided along the migration route. Those DQFHVWRUV DUH WUDFHG EDFN WKURXJK$WR 3RPER DQG RWKHUV WR WKH¿UVW human being, Abo Tani, who separated out from a half-human, half-spirit form, which came from Chantung, who came from the separation of earth- sky, which divided the primal formlessness, with which the world began.

CONCLUSIONS

Apatanis speak about origins in terms of cosmogony and genealogy, with comparatively little interest in geography. Even the migration legend is less about spatial movement from a place of origin than a process of descent from a common source. Moreover, these two idioms—the cosmogonic and the genealogical—are similar. Recalling the accounts of the beginning of the world (text 1) and of human beings (text 6), we

20 Ditii Latii is a sacred grove associated with this village. 21 Genealogical knowledge required to determine marriage partners stretches back only ¿YHRUVL[JHQHUDWLRQV 120 STUART BLACKBURN

VHHWKDWWKH\DUHOLQNHGWKURXJKWKH¿JXUHRI&KDQWXQJ6KHLVERWKD procreative female and the mother of Abo Tani. Human (and animal) birth, the emergence of two from one, underlies both idioms. Running through all Apatani expressions of origin is the division of a prior unity into separate elements. From the protean formlessness earth- sky splits in two, and a female body divides into the various parts of the natural, human and material worlds. That female (usually Chantung) then gives birth to Abo Tani, who separates from his spirit-half, just as humans split off from their mithun-half. Abo Tani’s children are the ancestors, who break off from a common stock to form the individual tribes in the Tani group. Apatanis then break up into clan ancestors, who settle in each village in the Apatani valley. The origins of everything, from rivers to ethnic groups, from baskets to clans, is imagined as a process of differentiation. Taken as a group, these expressions are the Apatani response to the paradox of creation and origination. The enigma that something comes from nothing is answered by the pre-existing formlessness (kolo-kolyung), which is both nothing and something at the same time.22 A related puzzle—that two come from one—is addressed by the various images of birth and genealogical division, as well as by the noun-pair that dominate Apatani ritual speech. This noun parallelism is not well understood, let alone fully documented, but its ambiguous form—a single entity that conceals a pair—appears to be another attempt to explain the riddle of origination. Returning to the Austronesian material that was the catalyst for this essay, Fox argued that the core idea of origins in that region was precedence. In this ‘founder-focused ideology’, as one of his colleagues put it, rank and rights are determined by a person’s or group’s relation to founding ancestors, their actions and ancestral objects.23 A similar emphasis on founders and ancestor cults has also been noted in upland Southeast Asia, among populations related to the peoples of central Arunachal Pradesh by language, oral tradition and ritual practices.24 Among Apatanis, however, similar ideas and practices of precedence are not prominent. Although, as we have seen, ancestors are important, there is no founders cult and virtually no ritual commemoration of ancestors. Ancestors are remembered but merely as generic, mythic ¿JXUHV DQG QRW DV QDPHG KLVWRULFDO ¿JXUHV ZKR LQLWLDWHG D FODQRU

22 For a recent discussion of this paradox, see Doniger 2008. 23 Bellwood 1996. 24 See, for instance, Tannenbaum and Kammerer 2003. APATANI IDEAS AND IDIOMS 121 lineage. In addition, no clan claims special authority based on prior status or settlement. True, Apatani clans are divided sharply between high status and low status, largely on the perception and partial reality that the low status clans contain neighbouring tribesmen and women who assimilated into local society as slaves, servants or bondsmen and women. However, none of the high status clans—about 75% of the total population—asserts rights or privileges with reference to the past. Senior men, in all clans, sometimes enjoy preference at rituals, but even this is informal and often absent. Also, ancestor worship is relatively weak among Apatanis. Ancestors are summoned, by generic names, to the large feasts, but Apatanis do not attempt to contact them or represent them in physical form. Their overriding attitude to historical ancestors is to keep them at a safe distance, in the land of the dead. Most of this discussion has been conceptual and textual, so I will end E\ VXJJHVWLQJ VRPHWKLQJ RI WKH VRFLDO VLJQL¿FDQFH RI$SDWDQL LGLRPV and ideas of origins. While founder ideology and precedence are weak, genealogy is central to Apatani thinking. Most Apatanis know little if anything about doging and kolyung-kolo, and few know the myths about the beginnings of the world. However, virtually everyone knows the story of Abo Tani’s genealogy. It tells Apatanis everything they need to know about who they are and how they are related to others. It tells them that they are distantly connected to powerful outsiders (halyang) and closely related to but distinct from neighbours (misan). From that genealogy, they know that they are Apatanis (tanii), and members of DVSHFL¿FFODQDQGYLOODJH7KHVHERXQGDULHVRIHWKQLFLW\DUHVKLIWLQJ as Apatanis reposition themselves within the new context of statehood politics. Still, Abo Tani’s genealogy remains the framework for the formation of local identity. %RXQGDU\GH¿QLWLRQRIFRXUVHLVWKHNH\VWRQHWRHWKQLFLGHQWLW\LQ most societies, and the Apatani use of genealogy reveals little that is distinct about local culture. More, I believe, may be learned by pushing our analysis back to the ideas from which such identity springs. In the Apatani case, the near-universal idiom of inclusion and distinction is LQÀHFWHGWKURXJKWKHLGHDRIGLIIHUHQWLDWLRQIURPDSULPDOIRUPOHVVQHVV That idea is evident even in the language (earth-sky, spirit-man, man- mithun) used to describe origins, which Apatanis think of as a process of separation, but not as a separate-ness. They know that the various things and people in their world are related through a common source or a shared ancestry. The emphasis is not on precedence but on belonging. 122 STUART BLACKBURN

BIBLIOGRAPHY

Aisher, A. 2006. Through spirits: Tribal cosmology and landscape ecology in Arunachal Pradesh, northeast India. PhD dissertation, University of London. Allen, N.J. 1980. “Tibet and the Thulung Rai: towards a comparative mythology of the Bodic speakers.” In M.V. Aris and Aung San Suu Kyi (eds.), Tibetan Studies in Honour of Hugh Richardson. Warminster: Aris and Phillips, pp.1-8. —— 1997. “Animal guides and foundation myths.” In S. Karmay and P. Sagant (eds.), Les Habitants du Tote du Monde. Nanterre: Société d’ethnologie, pp.375-90. Baruah, S.K. 1999. “Oral literature of the Hill Miris; an appraisal on some selected myths of creation of the universe.” In B.B. Pandey (ed.), Oral Literature of Arunachal Pradesh: Creation of Universe. Itanagar: Government of Arunachal Pradesh, pp.101-07. Bellwood, P. 1996. “Hierarchy, founder ideology and Austronesian expansion.” In J.J. Fox and C. Sather (eds.), Origins, Ancestry and Alliance. Explorations in Austronesian Ethnography. Canberra: Department of Anthropology, Australia National University, pp.18-40. Blackburn, S. 2008. Himalayan Tribal Tales: Oral Tradition and Culture in the Apatani Valley. Leiden: Brill. Blackburn, S. 2010. The Sun Rises: A Shaman’s Chant, Ritual Exchange and Fertility in the Apatani Valley. Leiden: Brill. Bora, D.K. 1995. Myths and Tales of the Nishings of Arunachal Pradesh. Itanagar: Government of Arunachal Pradesh. Doniger, W. 2008. “‘You can’t get here from there’: The logical paradox of ancient Indian creation myths.” In M.J. Geller and M. Schipper (eds.), Imagining Creation. Brill: Leiden, pp.87-102. 'XWWD'.7%DGXDQG-(WWH³6LJQL¿FDQFHRIWKHK\PQRQWKHFUHDWLRQ of the universe among the Gallongs.” In B.B. Pandey (ed.), Oral Literature of Arunachal Pradesh: Creation of Universe. Itanagar: Government of Arunachal Pradesh, pp.43-71. Elwin, V. 1958. Myths of the North-East Frontier of India. Shillong: North-East Frontier Agency. Fox, J.J. 1996. “Introduction.” In J.J. Fox and C. Sather (eds.), Origins, Ancestry and Alliance. Explorations in Austronesian Ethnography. Canberra: Department of Anthropology, Australia National University, pp.1-17. von Fürer-Haimendorf, C. 1952. The after-life in Indian tribal belief. Journal of the Royal Anthropological Institute of Great Britain and Ireland 83(1): 37-49. Gaenszle, M. 2000. Origins and Migrations: Kinship, Mythology and Ethnic Identity among the Mewahang Rai of East Nepal. Kathmandu: Mandala Book Point/ The Mountain Institute. —— 2002. Ancestral Voices: Oral Ritual Texts and their Social Contexts among the Mewahang Rai of east Nepal. Hamburg: LIT Verlag. McKhann, C. 2003. Sacred trails: genealogical mapping and the creation of historical space among the Naxi of southwest China. Histoire et Anthropologie Asies 2: 29-47. Mibang, T. and P.T. Abraham. 2002. Indian Folk Tales of North-East. Delhi: Farsight. APATANI IDEAS AND IDIOMS 123

Mitkong, N., A. Megu and B. Paul. 1999. “Oral literature of the Tagins (origin of the universe).” In B.B. Pandey (ed.), Oral Literature of Arunachal Pradesh: Creation of Universe. Itanagar: Government of Arunachal Pradesh, pp.72-100. Oppitz, M. 2006. Die Geschichte der verlorenen Schrift. Paideuma 52: 27-50. Takhe Kani. 1996/1997. Kolyung-Kolo Ayu: Apatani folklore. Arunjyoti 9/10: 13-6. Tannenbaum, N. and C. Kammerer. 2003. Founders’ Cults in Southeast Asia: Ancestors, Polity, and Identity. New Haven: Yale Southeast Asia Series. Taring, O. 2006. A journey of ancestors of the Nyishi community for prosperity. Journal of the Anthropological Society of India 41(2): 147-61. Tayeng. A. 1976. Two creation myths of Adis. Resarun 2: 163-70. Tob Tarin Tara. 2005. Nyishi World. Itanagar: Eastern Horizon.

MIGRATION NARRATIVES, OFFICIAL CLASSIFICATIONS, AND LOCAL IDENTITIES: THE MEMBA OF THE HIDDEN LAND OF PACHAKSHIRI

KERSTIN GROTHMANN

INTRODUCTION

This chapter examines oral and written origin and migration narratives of the Memba, a little-known eastern Himalayan highland society inhabiting the present-day Mechukha Circle area of West Siang (Arunachal Pradesh, India).1 These narratives mainly concern Memba memories of previous migration to, and settlement in, the so-called ‘hidden land’ (Tib. sbas yul)2 of Pachakshiri (Tib. sBas chags shing ri),3 being the valley now known as Mechukha (Tib. sMan chu kha). Based upon recent oral data recorded in Mechukha, and a survey of all the available documents and accounts about the area and its population, this study will demonstrate two points: that migration narratives are still very relevant to contemporary Memba identity and their claims to being the legal and rightful owners and occupants of their territory; and that the view long-held by outside observers and today by the Indian state, that Memba are a single population of common origins who actually live in two different locations—Mechukha Circle and Tuting/Geling Circles to the east—is problematic and in need of reassessment.

THE MEMBA OF PACHAKSHIRI

The territory known by local Memba inhabitants of Mechukha Circle as

1 Fieldwork data on the Memba of Mechukha and the inhabitants of Tuting area was gathered during 2007-2009 as part of the project Between Tibetanization and Tribalization: Towards a New Anthropology of Tibeto-Burman-Speaking Highlanders in Arunachal Pradesh, directed by Prof. Toni Huber (Humboldt University, Berlin) and funded by the Deutsche Forschungsgemeinschaft, Bonn. 2 All proper Tibetan spellings in the main text are romanized using the Wylie system and preceded by the abbreviation Tib. 3 There are at least four Tibetan variations of the name: sBad lcags shing yul, sBas chags shing gi ri, sBas chags shri, and sBa lcags sher ri ljongs. English language sources from India also give it mixed treatment, with Billorey 1998:64 having ‘Rajashiri’, which he translates as ‘hidden heaven’, and the ‘Vheza Shingiri’ of Dutta 2000. 126 KERSTIN GROTHMANN Figure 7.1. Map of sites related to the Memba in eastern Himalaya. MIGRATION NARRATIVES 127

Figure 7.2. Mechukha valley view north (photograph by Kerstin Grothmann, 2007). the ‘hidden land’ of Pachakshiri is located at an altitude of approximately 1900m in the north-western corner of West Siang District of Arunachal 3UDGHVK7KHZLGHULYHUYDOOH\LVÀDQNHGE\WZRVQRZFDSSHGPRXQ- tain ranges, the Damchen La (Tib. Dam can la) to the northeast and the Shinjong La (Tib. Zhing skyongs la) to the southwest. Memba settlements are located mainly along the sunnier northern bank of the Yargyab Chu (Tib. Yar rgyab chu) river and are composed of various clusters RIKRXVHVHDFKEHDULQJDFRPPRQQDPH:KLOHRI¿FLDOµYLOODJH¶QDPHV are nowadays applied by the state, among themselves Memba villagers still use the individual house cluster names to identify their place of residence or birth. Although Memba territory fell south of the 1914 McMahon Line that was agreed in principle and represented upon both British and Tibetan maps, the administrative power of the British colonial government in India never reached as far as Pachakshiri prior to 1947. In 1951, the ¿UVWUHSUHVHQWDWLYHVRIWKHQHZO\LQGHSHQGHQW,QGLDQVWDWHDUULYHGLQWKH YDOOH\ DQG HVWDEOLVKHG DQ DGPLQLVWUDWLYH SRVW DQG WKH\ FODVVL¿HG WKH local inhabitants with the name Memba. At that time, according to all local accounts, the valley dwellers referred to themselves as either the 128 KERSTIN GROTHMANN

Nänang (Tib. gNas nang) or Pachakshiriba (Tib. sBas chags shing ri ba). While Pachakshiriba simply means ‘one from Pachakshiri’, Nänang can mean something like ‘inside (Tib. nang) the holy place (Tib. gnas)’, DQGWKHVLJQL¿FDQFHRIWKLVZLOOEHFODUL¿HGE\WKHGLVFXVVLRQWRIROORZ Regardless of meaning or the connotations which the generic Tibetan term Memba (i.e., a phonetic variant of Tib. Mon pa) carries (see below), LW ZDV WDNHQ E\ WKH JRYHUQPHQW DQG WUDQVIRUPHG LQWR D FODVVL¿FDWRU\ referent for a population under new administration. The name for the new administrative centre, and the entire region, was adopted from the ¿UVW0HPEDVHWWOHPHQW0HFKXNKDZKLFKWKHJRYHUQPHQWSDUW\DUULYHG at upon initially entering the valley. Up until the mid-1950s, the Memba continued to pay taxes to the Lhasa-based, Tibetan aristocratic Lhalu (Tib. lHa klu) family who held the area as an estate,4 and such payments were made via the administration of the adjacent Tibetan district of Gacha (Tib. sGar chags) to the north. The Memba had extensive trade relations with both the neighbouring Tibeto-Burman-speaking highland societies to the south and Tibetans to the north, and acted to some extent as intermediaries between them. With the development of Indian administration in their region, Memba relations with Tibet declined sig- QL¿FDQWO\DQGFDPHWRDQDEUXSWHQGDVGLGDOOWKHLUSUR¿WVIURPWUDGH GXULQJDQGDIWHUWKH6LQR,QGLDQERUGHUFRQÀLFW,Q-XO\WKH Mechukha Subdivision was declared autonomous with its headquarters in Mechukha township. Its population of approximately 10,000 persons5 RI¿FLDOO\FRQVLVWVRI%RNDU0HPED5DPRDQG3DLOLERµWULEHV¶RIZKLFK the Memba make up at least one quarter.6 The Memba act as an exogamous, polygynous society in relation to other Tibetan Buddhist societies but are endogamous in relation to non-Buddhists. Arranged marriages are common. However, these traditions are all in transition among the younger generation. A range of features, including ‘primitive’ traits, geographical isolation, distinct culture, shyness of contact with outside agents, and economic ‘backwardness’ were all used by the Indian state to classify local

4 Both Shing sdong 1988:67 and bDe rab Tshe rdor 1988:73 mention that Pachakshiri belonged to the 12th Dalai Lama’s paternal family and was later assigned to Lhalu. 5 See Statistical Handbook of West Siang District – 2005'LVWULFW6WDWLVWLFDO2I¿FH Along, West Siang District. 6 There is no precise population data available for the Memba of Mechukha. The numbers ZH¿QGLQSXEOLVKHGOLWHUDWXUHFRYHUERWKJURXSVRI¿FLDOO\FODVVL¿HGDV0HPEDZKRGZHOOLQ 0HFKXNKDDQG7XWLQJ7KHRI¿FLDO(OHFWRUDO5ROOIRU0HFKXNKDRIOLVWVDSSUR[LPDWHO\ 2000 Memba who are entitled to vote, and to these must be added an unknown number of PLQRUVDQGWKRVH0HFKXNKD0HPEDRI¿FLDOO\UHVLGLQJRXWVLGHWKHUHJLRQLQKLOOWRZQVVXFK as Along and Itanagar. MIGRATION NARRATIVES 129

FRPPXQLWLHVVXFKDVWKH0HPEDDVµWULEHV¶7KLVFODVVL¿FDWLRQV\VWHP paid little or no attention whatsoever to the self-perceptions of these groups. The entries on the Memba in several encyclopaedias published in India are indicative of this, and repeat more or less the same short assertions made about language, migration, group or clan division, eating habits, marriage customs, major festivals and relation to neighbouring tribes or the Tibetans to the north.77KHDEVHQFHRIVXI¿FLHQW¿HOG studies, and no critical and comparative analysis of the available reliable data, has led to the assumption in India that there are two populations living in separate places who together constitute Memba society, that is, the Memba of Mechukha and the Memba of the region of Pemakö (Tib. Padma bkod) which is some 100 km to the northeast in the upper 6LDQJ9DOOH\2I¿FLDOGRFXPHQWVPDNHQRGLVWLQFWLRQEHWZHHQWKHPDQG label the populations of both areas with the common name Memba.8 Published studies by Indian scholars9 who focus only upon the Memba of Mechukha, fail to make any clear distinction between them and the population in Pemakö. In what follows, I will give a preliminary argument for the existence of two different societies in Mechukha and Pemakö based largely upon their migration histories. Further research in WKHUHJLRQRI3HPDN|LQWKHIXWXUHZRXOGEHGHVLUDEOHWRFRQ¿UPPRUH details of what I will set forth below.

THE MEMBA OF MECHUKHA AND PEMAKÖ

The places where the Memba of Arunachal Pradesh settled are Mechukha valley in the north-western corner of West Siang District and the area around Tuting (Tib. Tu lding) and Geling (Tib. dGe gling, or dGe ring), which is also known as Pemakö, in the north of Upper Siang District. In Tibetan societies in general, place of residence is almost DOZD\VDKLJKO\VLJQL¿FDQWDVSHFWRILGHQWL¿FDWLRQDQGVHOILGHQWLW\DQG with this in mind we can note that Memba is not an autonym used by either the Mechukha or Pemakö populations. They refer to themselves as Pachakshiriba and Pemaköpa respectively. The name Memba10 is actually a phonetic variant of the generic Tibetan ethnonym Mönpa

7 See, for example, Singh 1994:769-70, Choudhury 1995:195-99, Choudhury 2004:1036-40. 8 See, for example, Choudhury 1994:3, 73. 9 See Billorey 1998:63-7, Dutta 1998:36-40, Dutta 2000:45-64, Dutta 2006 and Badu 2002. 10 Variant forms in the European language literature include Menba, Moinba, Mönba and Mempa. 130 KERSTIN GROTHMANN

(Tib. Mon pa), which simply means “one from Mön”, and which is widely used to designate very different societies and groups across time and space.11 The name has often been associated with the notion of being non-Buddhist, in the sense of lacking the high culture or civilisation associated with Buddhism by Tibetan elites and thus it carries the strong connotation of ‘barbarian’. Mönpa has frequently been applied to different groups living on the southern and south-eastern slopes of the Himalaya, especially in parts of eastern Bhutan and the neighbouring West Kameng District—including Tawang, Dirang and Kalaktang—of present-day Arunachal Pradesh. Even though these ‘Mönpa’ groups were, in the course of their history, converted (Tib. ’dul ba) to and LQÀXHQFHGE\%XGGKLVP²DQGWKHUHIRUHLQWKH7LEHWDQGLVFRXUVHUDLVHG at least onto the scale of ‘civilisation’—from the central Tibetan point of view they often appear to have never been fully included in what was considered to be the properly civilised world. When Nem Singh, one of the ‘Pandit’ explorer-spies in the employ of the British colonial government of India, visited the Tibetan areas of Orong and Gacha just north of Pachakshiri in 1878/79, he reported of the people visiting there from the south that, “They are called Mönbas by the Tibetans, who give the same name to the Lepchas of Sikkim. […] They call themselves Pachakshiriba.”12 In more recent times, among themselves the people of Pachakshiri rather prefer to call themselves Nänang, as do neighbouring Tibetan communities: “The people from Molo, Orong and Gacha call us Nänang. Some from Lhasa call us Pachakshiri. Some call us Mijim Thangba or Tsari Mijim Thangba. The other tribal people call us Memba or Nyema. This is Lopa language.”13:H¿QGGLIIHUHQWLD- tions within these local designations, and it seems that the further the place the more precise is the usage of the toponym. For example, Tsari Mijim Thangba (Tib. Tsa ri Mi khyim bdun ba) is the name of a village in the very south of the Tsari region to the west.14 But the closer the place, for instance Orong, Gacha and Molo which all belonged to the same former Tibetan administrative district as Pachakshiri, the less it seems necessary to use the exact location as reference. Instead characteristics of the place such as its perceived sacredness as a holy place or nä (Tib. gnas) are favored. Central Tibetans collectively labeled another group inhabiting the same region as Lopa (Tib. Klo pa), which emphatically

11 See Pommaret 1999, Aris 1979 and Bellezza 2008:21-2. 12 Harman 1915:211. 13 Interview with Tashi Naksang from Dechenthang in March 2008. 14 Also known as Lo Mikyimdün (Tib. Klo Mi khyim bdun), see Huber 1999:131ff. MIGRATION NARRATIVES 131 Figure 7.3. Map of Pachakshiri and adjacent Tibetan sites. Tibetan Figure 7.3. Map of Pachakshiri and adjacent 132 KERSTIN GROTHMANN means ‘barbarian’.15 This blanket designation covers all non-Buddhist populations in the southeastern Himalaya whom the Tibetans characterise as savage or wild, living in the jungle in a hostile and untamed environment, and as subsisting on hunting wild animals and gathering wild plants. The so-called ‘Lopa’ who are neighbours of the Pachakshiriba describe the latter as Nyema, a variation of the ethnonym (also Nyime or Nyimak) which is commonly in use by various Tani-speaking hill peoples throughout central Arunachal to refer to ‘Tibetans’ and ‘Tibet’ to the north more generally.16 But they also adopted the generic Tibetan term Mönpa or Memba to designate their Buddhist neighbours. It is probably this fact, and the acquisition by different generations of outside explorers and administrators of this identity referent from neighbour- LQJQRQ%XGGKLVWSHRSOHVWKDWOHGWRWKHLGHQWL¿FDWLRQRIRQHVRFDOOHG ‘Memba’ society in two different locations. Even though most of the early western explorers to visit the far eastern Himalaya and the British colonial administration knew of the local autonyms that different hill communities used to identify themselves, they frequently did not take these into account in their descriptions and FODVVL¿FDWLRQV *HRUJH 'XQEDU IRU H[DPSOH ZKR YLVLWHG WKH 3HPDN| region in 1913, stated that he adopted the name Memba for the people in Pemakö from the neighbouring non-Buddhist Tangam tribe. He also noted that these same Memba were to be found in Bhutan and were settled in Pemakö at the beginning of the nineteenth century.17 Similarly, many of the other neighbouring non-Buddhist tribes use the term Memba, although they have their own word, Mimat,18 to collectively label the Buddhist populations who “[…] live close to the snow-line, dress in skins, and are spoken of by the Abors with disgust and contempt”, as Arthur Bentinck reported.19 Indeed, the common use by non-Buddhist tribes of the terms Mönpa or Memba to label their Buddhist neighbours seems to be a reaction to the degrading Tibetan term Lopa being applied to themselves. Concerning the migration history of the Buddhist population of Pemakö whom outsiders like the Tangam and Dunbar labelled as Memba, we know from Frederick Marshman Bailey that:

15 For some further elaborations on the term Lopa, see Huber 2011, Huber 1999:179-81, Huber 1997:226. 16 See, for examples, Blackburn 2003/2004:49, n.46 on the Apatani, and Huber 2010 on the Mra. 17 Dunbar 1916:93. 18 The term Mimat is still in use, whereas Memba is used with outsiders from the region. 19 Bentinck 1913:105. MIGRATION NARRATIVES 133

7KH GHVFHQGDQWV RI WKHVH ¿UVW LPPLJUDQWV QRZ IRUP WR D ODUJH H[WHQW the population of the valley; they are called Mönbas or Drukpas indis- criminately: the former name means an inhabitant of the Tibetan district of Mönyul near Tawang, and the latter means Bhutanese. They still speak a dialect of Mönba, the language spoken near Tawang.20 This assertion corresponds to some extent with the present-day statements of the Pemaköpa inhabitants of Tuting area about their own migration and identity. At the beginning of the twentieth century, only a few Buddhist families were settled in Tuting. However, the place became the main settlement for Pemakö inhabitants south of the McMahon Line as a result of migration between 1959 and 1962 following the Chinese PLOLWDU\ RFFXSDWLRQ RI 7LEHW DQG WKH 6LQR,QGLDQ ERUGHU FRQÀLFW 6XEVHTXHQWO\WKH%XGGKLVWSRSXODWLRQRIWKLVUHJLRQZDVRI¿FLDOO\FODV- VL¿HGE\WKH,QGLDQDXWKRULWLHVLQWRWKUHHµWULEHV¶QDPHO\.KDPSD/DPD DQG0HPED7KHFODVVL¿FDWLRQ0HPEDLVORFDOO\UHMHFWHGE\PDQ\VLQFH for them it clearly designates the people of the Tawang region, i.e., the ‘Mönpa’. Khampa refers to the non-Tshangla speaking Buddhist population that migrated mainly from the Metog Dzong area, although the term is also often synonymously used for ‘Tibetans’ or Böpa (Tib. Bod pa). Lama, Drukpa and Pemaköpa are the autonyms which are used as terms of self-reference by the Tshangla-speaking population of the region who trace their origins back to Bhutan but who distance themselves clearly from the people of the Tawang region. In 1968, the Indian government philologist B. Shastri stated in his notes of a tour in the Siang valley that the dialect of the Memba in the Tuting area is virtually the same as that of the Mönpa in Kameng District.21 During eastward migrations from Bhutan and Tawang about two centuries ago,22 a small group of migrants appear to have entered the Pachakshiri valley and settled at its northwestern corner at a place named Lhalung. Frederick Marshman Bailey wrote of these migrants that they “…are known as Pachakshiribas. They speak Mönba among

20 Bailey 1914:2, whose reporting is to be viewed as reliable since he spoke Tibetan fairly ÀXHQWO\DQGKDGYLVLWHGWKHUHJLRQVLQFOXGLQJ%KXWDQZKLFKDUHEHLQJGLVFXVVHGKHUH 21 See Shastri 1968:3. Any information on the language spoken by the Khampa in Tuting is missing in the report. 22 According to the 1878/79 report by Nem Singh, “[…the Pachakshiriba] had many villages near the line of route from Tsetang to Tawang and Odalguri in Assam”; see Harman 1915:211. With the help of Nem Singh’s statement, some of these places might be tentatively LGHQWL¿HG VHHPDS DVIROORZV7DVKLNKDQJ.DUWKHQJ/XPOD7DZDQJZLWK8UJ\HOLQJDQG Sangyeling in the immediate vicinity, Zemithang, Tsona, Ongbagang (Yum bu bla mkhar),Ye Gongmo Khangsar or Agom Khangsar (E Yul, Gong khang gsar), Lhagyari (lHa rgya ri), Dakpnang (Dwags po Nang rDzong) and Lelung. 134 KERSTIN GROTHMANN

Figure 7.4. View of Lhalung (photograph by Kerstin Grothmann, 2007). themselves, but those we met could all speak good Kongpo Tibetan.”23 One of the prominent clans in Mechukha is Naksang, and according to their migration narrative they came from Urgyeling and Sangyeling near Tawang via Kharteng, Lumla, Zemithang and Tsona,24 and from there onwards were accompanied by the Tsona clan which was on pilgrim- DJHWR7VDUL7KH1DNVDQJPLJUDQWV¿UVWVHWWOHGLQWKH/KDOXQJDUHDRI Pachakshiri, and their ‘Mönpa’ language from the Tawang area would not have been well understood in the region. Indeed, Nem Singh reported that at Orong just to the north an interpreter was kept for those who visited from Pachakshiri.25 The language of the Naksang migrants was also not understood by the other clans of Pachakshiri and is almost lost nowadays. Thus, contemporary informants report that “The people of Molo and Gacha speak the same language as we do. At Lhasa they speak a little different but we understand each other.”26 This is because the common dialect in Mechukha nowadays is that of the adjacent Tibetan region of Kongpo, as are the local dress and hairstyle. In fact, in 1913 Frederick Marshman Bailey observed that “they wear Tibetan clothes except for their foot-gear […]. They do their hair in two queues in the Kongpo fashion; the Mönbas of Pemakö as well as those of Mönyul […] cut their hair short.”27 The presumed migration from parts of Bhutan and Tawang at the beginning of the nineteenth century, and the languages from that same region spoken by most of these migrants, has led to the assumption that 23 Bailey 1914:18. 24 Urgyeling and Sangyeling are about 3 km south of Tawang monastery. Kharteng and Lumla are further to the south-west of Tawang, while Zemithang is to the north-west and Tsona to the north. 25 Harman 1915:211. 26 Interview with Tashi Naktsang from Dechenthang, March 2008. Similar statements were made by all informants who had visited Tibet prior to 1962. 27 Bailey 1914:18. MIGRATION NARRATIVES 135 the inhabitants of Mechukha and Pemakö are all the same people only VHWWOHGLQGLIIHUHQWSODFHV([SORUHUVDQGDGPLQLVWUDWLYHRI¿FHUVHQWHULQJ from the south uncritically adopted the generic term Memba which they heard from neighbouring communities and then transformed it into a tribal label. Seemingly trivial statements on clothing and hairstyle in earlier reports, and the known local autonyms and ethnonyms, were not seriously considered or they were neglected, and further research by independent anthropologists had not taken place until my own study RI WKH DUHD 2XWVLGH FODVVL¿FDWLRQV DQG LQVXI¿FLHQW GHVFULSWLRQV ZHUH thus used to classify local populations, and despite many differences between some communities, they were nevertheless grouped together in order to administer and incorporate these societies into the modern Indian nation state. Thus, the peoples of Pachakshiri and Pemakö were listed by the Indian administration as one tribe, the Memba, during the 1950s, and went on to share the same fate as other independent societies in Arunachal Pradesh, even though they themselves have a clear idea where they come from, who they are and who they are not.

THE IDEA OF BEYÜL — ‘HIDDEN LANDS’

Unlike other tribes in Arunachal Pradesh, the Memba of Mechukha do not have a single narration explaining their migration process or route to Pachakshiri, although we can reconstruct it to a certain point using other sources. Rather, of much larger importance for the Memba sense of origin and identity is a narrative about the exploration of the ‘hidden land’ of Pachakshiri and its transformation into a habitable place. According to later Tibetan historiographies, Buddhism was estab- OLVKHGDVDVWDWHUHOLJLRQLQ7LEHWGXULQJWKHQLQWKFHQWXU\E\WKH¿JXUH of Padmasambhava, and this beginning is closely associated with the construction of the Samye monastery by Emperor Trisong Detsan. The Memba are followers of the Nyingmapa tradition of Tibetan Buddhism, a movement which places particular emphasis upon the narratives DERXW3DGPDVDPEKDYDDVDIRXQGHU¿JXUH2QHVXFKVHWRIQDUUDWLYHV relate Padmasambhava concealing several valleys as ‘hidden lands’ or beyül (Tib. sbas yul) on the southern slopes of the Himalaya during his journey to Tibet. There are written Tibetan texts that describe routes to these beyül, and the means by which such places will be discovered and opened in the future, a task only to be undertaken by accomplished Buddhist masters. The hidden lands were meant to be sanctuaries in times of chaos and disorder where people from every strata of Tibetan 136 KERSTIN GROTHMANN

VRFLHW\FRXOG¿QGUHIXJHDQGIRUPDQLGHDOLVHGVRFLHW\ZLWKDNLQJDV the legitimate ruler to preserve moral and political values.28 Beyül are not only a refuge; they are also places of Buddhist realisation, where advanced practitioners gain greater insight for their own and others’ EHQH¿WDQGZKHUHVDFUHGWUHDVXUHVZLOOEHUHYHDOHG3HUFHLYHGDVEHLQJ situated right at the border between the civilised Tibetan world and the uncivilised tribal world, these places not only have to be discovered, they also have to be opened and transformed into a suitable and fertile territory to shelter a certain number of people. As it is represented in the Tibetan texts, this opening and transformation phase has to be read as a civilising process. A powerful master has to subdue all local deities and non-human forces and bind them by oath to protect Buddhism, the religious system that must be introduced to the local non-civilised population who may already be inhabiting the place. This notion of an ideal society that re-establishes law and order in a predetermined land and which preserves and protects moral and religious values is an important part of Memba identity, and in particular one used for drawing distinctions with their tribal neighbours. For instance, one local narrative has it that all those who now live in the region were once without a written script and a religion, but they all received these from Tibetan Buddhist Lamas. The Memba wrote the religious teaching down upon stones, while their ‘Lopa’ neighbours wrote them down on animal skins. One day the Lopa got very hungry, but there was no food left. They boiled the skins and ate them, and so both script and religion were lost. Thus, the Memba consider themselves superior to the neighbouring tribes since they are the sole inhabitants of the region with an indigenous script, an organised religion, and therefore a ‘civilised culture’.

WRITTEN ACCOUNTS OF PACHAKSHIRI’S ESTABLISHMENT

Being literate, the Memba have two textual sources29 relating aspects of Pachakshiri as a beyül and how it came to be inhabited. One of these WH[WVFRQWDLQVVLJQL¿FDQWPDWHULDODERXWPLJUDWLRQLQWRDQGVHWWOHPHQWRI

28 On the concept of Tibetan Buddhist hidden lands, see Childs 1999, and Childs in this volume. 29 According to Memba informants, both of these texts (see the details given below and in the Bibliography) were taken by the Chinese in 1962. However, according to Tibetan refugees who dwelt in Mechukha for some years, the texts were voluntarily delivered to the Chinese at the time, as a kind of symbolic gesture towards the possible new ruling power in the area. MIGRATION NARRATIVES 137

WKHYDOOH\7KH¿UVWWH[WLVRIWKHµLWLQHUDU\¶JHQUH30 and is entitled The Itinerary of Pachakshing valley called ‘Self-liberation on Seeing’. The text was revealed by one Dudjom Tsal from the Ketshel cave in Kongpo.31 The narrative is composed as a dialog between Padmasambhava and his consort Yeshe Tsogyal, who is said to have concealed the text at Ketshel. ,WGHVFULEHVZKHUHDQGZKHQWKHGHVLJQDWHGPDVWHUZLOO¿QG3DFKDNVKLUL and by what means he will open this ‘principal’ sanctuary.32 However, it contains no information about a possible migration. The meaning of the name Pachakshiri—given in the form sBas lcags shing yul—is glossed in the text as follows: Sbas pa means that, in future, when evil armies from beyond the frontiers invade, it is the place where living beings will go. Chags33 pa means that in future the spread of the explanation and accomplishment of the [Buddhist] teachings will be established. Shing yul means that red and white Sandalwood, Aloe and other [precious trees] are there in their HQWLUHW\%HLQJHTXLYDOHQWWRWKHSDUDGLVHVRIWKH¿YHLGHDO%XGGKDVLWV advantageous qualities are beyond description. This text is only known to a few religious specialists in Mechukha, and it is not used in ritual performances or daily life. To fully understand it, one has to have advanced knowledge of particular religious traditions. The second written account belongs to the “regulations for public guidance” (Tib. bca’ yig) genre, and is in the form of a long, handwritten scroll without title.347KH¿UVWKDOIRIWKHWH[WSUHVHQWVLQIRUPDWLRQRQ the discovery of Pachakshiri, its opening and the migration process, and names the religious masters and others involved. The numerous spelling and grammatical errors and fragmentary sentences due to physical

30 The Memba themselves refer to the text as gnas yig, even though its written title includes the word lam yig; see the Bibliography for details. Thanks to Matthew Akester who translated it. 31 Bdud ’joms rtsal kyis kong gi rke tshal phug pa nas ston pa’o. Treating this name as an abbreviated form, there are two immediate possibilities as for his identity: bDud ’joms rol pa rtsal (born nineteenth cent.); or bDud ’joms gling pa (1835-1904), whose secret initiatory QDPHZDVE'XG¶MRPVUGRUMHJURORGUWVDO)RUDPRUHSUHFLVHLGHQWL¿FDWLRQIXUWKHUUHVHDUFK will be necessary. 32 Pachakshiri is in fact one of the minor and less well-known Himalayan beyül, for which no other Tibetan language guidebook-style text has currently come to light, to my present knowledge. 33 This element is spelled lcags (meaning ‘iron’) on the title page, and is typical of the variation in spelling of the whole name in different sources. 34 This scroll of local handmade paper is approximately 3m long and 40cm wide. When I recorded it, it was in poor condition with its upper part and margins torn. It is in the possession of the Gompa Secretary, P.C. Kigar. I am grateful to Tsewang Norbu who worked through the text with me. 138 KERSTIN GROTHMANN

GDPDJHPDNHDGHSHQGDEOHWUDQVODWLRQRIWKHWH[WGLI¿FXOWDQGLQVRPH parts even impossible. The following passage is thus a summary focusing on data that help illustrate the text’s particular version of the migration process. The text begins with a description of a past era when religion in Tibet declined and epidemics occurred. Then: .KDP.DWKRJ/DPD6RJ\DOUHFHLYHGDYLVLRQDQGZDVWKH¿UVWZKRZHQW via Samgarong to Pemajeling,35 where he stayed in meditation for some time. Later, Lama Lodre Gyamtsho went on pilgrimage to the Sogyal cave where he had many extraordinary visions and was able to stay in meditation for almost eight years. After that he went to Par, and together with Rumdo Abo and others he went on pilgrimage. But something caused them great fear, so they couldn’t proceed further and had to stop. There was the question of whether the time for the countless beyüls has arrived or QRW ³:LOO LW EH KHOSIXO WR WKH 'KDUPD RU EHQH¿FLDO IRU WKH SHRSOH LI , [Lodre Gyamtso], with all my responsibility, were to declare the place as a beyül, a holy place where people will settle and build roads?” Rumdo Abo, Tsetan Norbu […] and I, together with 40 labourers, took off for pilgrimage. Drubchen Chözang Namgyal and Changchü Lingpa wanted to stay there during winter and build a Stupa at Dechen[thang]36 to cast out all evil from that region. Two Lamas and ten other people stayed near Pemajeling. In the Wood Dragon Year (=1724/25), the Samye Chökyong told me to invite Tagsham Chogtrül on the tenth day of the Monkey Month. Everything was arranged, but he was not able to come. Rumdo Abo and Tsetan Norbu, together with 20 labourers, were sent to the west entrance via Pälungthang37 for the construction of roads. Norbu constructed the ¿UVWEXLOGLQJDVPDOOWHPSOHDW3HPDMHOLQJ,QDXWXPQKHDGPDQ7HQ]LQ Norbu, accompanied by 50 labourers, came to construct roads. Again in the Fire Sheep Year (=1727/28), people constructed and repaired roads as well as bridges. In the Earth Monkey Year (=1728/29), Menchu Lama, Tashi Dorje and 20 labourers came and their work was successful. In the Earth Bird Year (=1729/30), the great government gave the order to the people of Tsegang and Tongshongog, that whatever labour would be necessary should be provided. Already prior to the Water Hare Year (=1723/24), it was evident that the place was special. Irrespective of the subjects, all necessary things and favourable conditions should be provided for this pilgrimage site. This is how the place was inhabited. This section gives us some clear indications of a migration process being carried out in successive waves headed by accomplished Buddhist masters and rooted in the local landscape of the valley whose names and features are still current today. The climax of this migration must

35$SODFHDWWKHULYHUEDQNZKHUHWKH¿UVWWHPSOHRI3DFKDNVKLULZDVEXLOW 36 The largest Stupa of the valley is near the village Dechenthang. 37 The village of Pälungthang is on the northern side of the Lola pass. MIGRATION NARRATIVES 139 have been under the guidance of Lama Lodre Gyamtso. According to the scroll’s colophon, it was in fact he who wrote the text in the Iron Pig Year (=1731/32) of the twelfth sixty-year cycle.38 Within a period of about seven years the place was explored and transformed by groups who carried out labour service. Several persons from these groups remained in Pachakshiri, without returning to their native places. After constructing the necessary infrastructure, religious buildings were established DQGSHRSOHEHJDQWRWLOO¿HOGVDQGUDLVHOLYHVWRFN7KLVEDFNDQGIRUWK movement of labour groups implies that a part of the migrants came from places not too far away from Pachakshiri, probably places to the north across the mountains within the former Tibetan administrative district of Gacha. If the text is to be believed, this migration during the third decade of the eighteenth century would have been about a century earlier than the migration of Mönpa reported by Frederick Marshman Bailey and Nem Singh. Any Mönpa coming from the greater Tawang region with this later migration must have become almost completely assimilated, or DWOHDVWZLWKLQWKHODVW\HDUVVLQFHZH¿QGRQO\YHU\IHZSHRSOHRI that age in Mechukha who still speak a Mönpa language from the west. We could assume that they only ever formed a minor part of the total Pachakshiriba community. Interestingly, their tradition of migration from the west, while absent from the written texts, forms an important part of the oral migration tradition, as we will see below. Following the section on exploration and settlement of Pachakshiri that we have summarised above, the text of the scroll comments on correct moral and religious behaviour, the relationship between the king and his subjects, and the establishment of law and social and religious order. It also gives directions how the society should deal with new- FRPHUVRUWULEDOQHLJKERXUV,WLVDQH[SODQDWLRQDQGMXVWL¿FDWLRQRIWKH Pachakshiribas’ rights as the chosen community, those who have been assigned to preserve and protect the standards of an ideal society in WKHµKLGGHQODQG¶7KLVLVZULWWHQGRZQLQDGH¿QLWLYHIRUPZKLFKFDQ KDUGO\UHVSRQGWRRXWVLGHLQÀXHQFHV7KLVVDPHNLQGRIGLVFRXUVHGRHV not occur overtly in the oral tradition but is present as a kind of subtext.

ORAL ACCOUNTS OF PACHAKSHIRI’S ESTABLISHMENT

Local oral narratives of origins and migration found among many non- Buddhist highland societies of Arunachal Pradesh often begin by relating 38 The colophon goes on to mention that the text was later completed by Sonam Gelek Rabtan Lhawang in the Earth Horse Year of the twelfth sixty-year cycle (=1738/39). 140 KERSTIN GROTHMANN their descent from the sky, or the creation of a place where life starts to ÀRXULVKRUULYDOULHVDQGFRQÀLFWVDPRQJVLEOLQJVUHVXOWLQJLQVHSDUDWLRQ and migration, and so on.39 Memba oral narratives of migration, on the other hand, closely resemble the Tibetan literary genre of guidebooks to holy places. They are descriptions of the religious geography of a holy place and function as an aid for pilgrims, and therefore enjoy great popularity. They not only provide interesting and useful information but also emphasise and revitalise events of religious and historical import and articulate them with local landscapes so that pilgrims can participate in the epic past and its sacred traces in various ways. Guidebooks are rarely if ever read or heard in their entirety but function rather more as DVRXUFHRIUHIHUHQFHYHUL¿FDWLRQRUDGYLFHWREHFRQVXOWHGIRUVSHFLDO occasions or at particular places, and their content is very often elabo- rated by additional oral stories at such sites. Whenever Memba saw my informant P.C. Kigar and me at some holy site around the Pachakshiri valley, be it during a pilgrimage or just while travelling past en route to their villages, they quickly rushed over to listen to the stories and explanations being expounded.40 7KH0HPEDSHUIRUPVL[SLOJULPDJHVWR¿YHPDMRUKRO\VLWHVDURXQG the Pachakshiri valley during each twelve-month period.41 These pilgrimages are organised by the Gompa Committee,42 with the involvement of volunteers providing refreshments for the public at the beginning of each pilgrimage and also preparing the sites where rituals will be performed during the journey. At the different holy sites on any pilgrimage route, the events believed to have happened in the past are recalled there by presenting prayers and offerings. The story of the exploration and transformation of Pachakshiri into an inhabited place is not usually a part of ritual performances or found in daily life among the Memba. Only a few people like Lamas or lay persons with strong interests in local history and religion are familiar with the details of the story. Therefore, 39 For example, see the discussion of narratives of origins and migration among the Mra by Huber 2010, and the Apatani by Blackburn 2003/2004, as well as the chapters by Blackburn, Huber and Aisher in this volume. 40 P.C. Kigar is one of the most knowledgeable persons regarding the religious life of the Memba community. As Gompa Secretary, he represents approximately 50 religious specialists called Lama, and is involved in almost all decisions concerning the religious, social or political life of the community. 41$FFRUGLQJWRWKH7LEHWDQOXQDUFDOHQGDUWKHVHSLOJULPDJHVEHJLQZLWKWKH¿UVWDQQXDO HYHQWRQWKHWHQWKGD\RIWKHWZHOIWKPRQWKDQGWKHQIROORZLQJRQWKH¿IWHHQWKGD\RIWKH ¿UVWPRQWKWKH¿IWHHQWKGD\RIWKHVHFRQGPRQWKWKHWZHQW\¿IWKGD\RIWKHVHFRQGPRQWK WKH¿IWHHQWKGD\RIWKHVHYHQWKPRQWKDQG¿QLVKZLWKWKH¿QDODQQXDOHYHQWRQWKH¿IWHHQWK day of the eighth month. 42 The Gompa Committee consists of eight elected members from the Memba community. MIGRATION NARRATIVES 141

Figure 7.5. Picnic atmosphere during pilgrimage (photograph by Kerstin Grothmann, 2007). the annual pilgrimages are most welcome occasions since they offer opportunities to refresh former knowledge of the story or to add new details to it. It appears that nobody, aside from a few specialists, feels the need to be familiar with all the details or the chronological order of HYHQWVLQWKLVVWRU\5DWKHUZKDWLVPRVWVLJQL¿FDQWLQLWWRWKH0HPED is the overall idea of establishing themselves and living as the chosen people in a beyül, a special type of holy valley surrounded by hostile, non-Buddhist tribal neighbours. Oral traditions are inherently dynamic, and they are often crafted according to the needs and circumstances of the present context of WKHLUWHOOLQJ7KXVZHGRQRW¿QGDVLQJOHFRPSUHKHQVLYHRUDOYHUVLRQ of the story of Pachakshiri as a ‘hidden land’ of Memba settlement. The following is a version of the Memba oral story, one with a particular IRFXVXSRQGH¿QLQJWKHERXQGDULHVRIWKHµKLGGHQODQG¶DWRSLFWKDWLVRI KHLJKWHQHGVLJQL¿FDQFHIRUWKH0HPEDQRZDGD\VDQGLQWKHUHFHQWSDVW ,WFDQEHUHDGLQSDUWDVDUHVSRQVHWRWKHDGYHQWRIWKH¿UVWURDGLQWRWKH YDOOH\DIWHUDQGVXEVHTXHQWLQ¿OWUDWLRQE\PRUHRXWVLGHUVGXHWR ease of access. Against the spectre of potential territorial competition, the story functions as proof of the more ancient claims of the Memba to 142 KERSTIN GROTHMANN Figure 7.6. Map of the Mechukha valley. MIGRATION NARRATIVES 143 have been the original migrants into, and settlers of, the valley: On his way back to India, Padmasambhava placed hidden treasures in all places such as Tawang, Bomdila and Tuting that he had visited earlier, and told his disciples that it would be important to have a holy place in Pachakshiri. One Lama from Kham, Kathog Lama Sogyal Rinpoche, visited Pachakshiri and opened the place. He meditated at Pemajeling for many years. It is a part of Tsari.43 That is why the place is also called Tsari Nänang [“Within the holy place of Tsari”]. On his way back to Tibet, he told many people about the place. Later, Gyalpo Yeshe Dorje44 came to the region to open a new holy place. When he reached Tsari Chözam, he opened up a new pilgrimage. After his return to Tibet, he instructed Lama Lodre Gyamtso and Thangtong Gyalpo45 to go to Pachakshiri. Both took different routes to enter the “hidden land”. Thangtong Gyalpo crossed the Dom La in the Manigong area and Lama Lodre Gyamtso crossed the Lo La into Pachakshiri. They went down following respectively the rivers of each area and were supposed to meet downstream at Tato, the conjunction of the two rivers. Lama Lodre Gyamtso was not able to proceed further and after hiding his key inside a rock at Künse Lhakhang, which will enable other Lamas in future to open further holy places, he went back and stopped at Karte where he met Thangtong Gyalpo. Together they built and consecrated a Stupa. At Karte we can see the footprints of Lama Lodre Gyamtso, his dog and horse. Thangtong Gyalpo was not successful in his mission to convert the people of the Manigong area to Buddhism and Lama Lodre Gyamtso only reached up to Künse Lhakhang. That is the reason why the Lopa don’t believe in Buddhism. Up to Künse Lhakhang the area belongs to Tibet. On the way back to Tibet both Lamas stopped at Pemajeling and from there Lama Lodre Gyamtso went to Lhalung. He saw that the place was not suitable to settle at. He put his walking stick made of VDQGDOZRRGLQWRWKHJURXQGDQGZLVKHG³7RQLJKWWKHUHVKRXOGEHDÀRRG ZKLFKÀDWWHQVWKHSODFHVRSHRSOHFDQVHWWOHKHUH´7KHQH[WPRUQLQJWKH SODFHZDVÀDWDQGKHFDOOHGLW/KDOXQJWKHµ9DOOH\RIWKH*RGV¶7KHVWLFN grew into a tree and is still to be seen there. He also constructed a Chörten IRU WKH EHQH¿W RI WKH SHRSOH$IWHU KH GLG DOO WKLV KH UHWXUQHG WR7LEHW He designated Yabme Pawo Dorje as his successor. The king of Gacha in Kongpo had a daughter who was very ugly and no one wanted to marry her. But suddenly she got pregnant and people were gossiping about how this could have happened. She delivered a son. When the boy reached the age of two years he went to the king and said, “Don’t accuse my mother of sex before marriage, I’m without father and my name is Yabme46 Pawo Dorje.”

43 For a comprehensive study on Tsari, see Huber 1999. 447KLVSHUKDSVUHIHUVWR.\HZR

By the age of 17, he left for Tsari accompanied by many other people. After crossing the Tsari Chu River and Lo La pass he came to Pachakshiri. He opened many new holy sites. He discovered Do Pemakö, the ‘map-stone’ of Tsari region that was made by the Buddha himself and actually belongs to Tsari, although nobody knows how it came to Pachakshiri. There is another stone called Shinjong Saptra, which is a map of Pachakshiri. Yabme Pawo Dorje meditated for a long time at Pemajeling and visualised the entire place: in the east there is a pilgrimage area called Shar Dorjeling; in the south there is a holy place called Lho Rinchenling; in the west there is Nub 3HPDOLQJ ZKLFK UHDFKHV XQWLO7VDUL ,Q WKLV DUHD ZH ¿QG WKH SLOJULPDJH sites Pema Shelphug, Tseriphug and Nä Sarma. And in the North there is Chang Gawaling. It reaches up to the border in the Manigong region. In between these main pilgrimage areas there exist other pilgrimage sites, and in total there are eight. Yabme Pawo Dorje performed many rituals at Pemajeling and many disciples and caretakers stayed with him. They all came together with him from Tibet. At that time, many problems occurred in Tibet. People didn’t have enough food and whenever people went from here to Tibet, Yabme Pawo Dorje told them to spread the message that there is a place called Pachakshiri which is a good place to settle and who- ever is willing to come is most welcome. Most of the people who settled at Pachakshiri came from Tibet. Only the Naksang clan came from Bhutan. First there was the oldest brother who came. After some time, when he didn’t return to his native place, the second oldest brother left in search for KLP%XWKHDOVRGLGQRWUHWXUQDQGVRWKH\RXQJHVWEURWKHUWRRNRIIWR¿QG his elder brothers. He found them both settling in Pachakshiri, and since the place was nice he decided to settle down there as well. Other clans like Dorsom came from Tibet. The Kigar and Sharjo clans came from Ngabi, the Dabo clan from Dakpo, the Tsona clan from Tsona, the Tsugla clan from Lhagyari, and many other people came from Ye Gongmo Khangser and Kongpo. They all got the message about Pachakshiri and came here. The time for Yambe Pawo Dorje’s return to Tibet came, but before he left he thought that the Tertön Chöje Lingpa Dakpo Kogi Lama should come to this place. He should be in charge. Then Chöje Lingpa came to Pachakshiri. Taksam Rinpoche was his father and Orgyen Chögyur Dechen Lingpa his younger brother. They came together, accompanied by 16 disciples, and VWD\HGDW3HPDMHOLQJZKHUH&K|MH/LQJSDFRQVWUXFWHGWKH¿UVWPRQDVWHU\ and named it Samden Yangchag. The 16 disciples were divided into two groups. One group was not allowed to marry and have children. Their duty ZDVPHGLWDWLRQDQGWKHSHUIRUPDQFHRIULWXDOVIRUWKHEHQH¿WRIWKHSHRSOH The eight caves where they meditated are still there, and the place is called Drub Khang. The other group was allowed to marry and have children. Their duty was to provide food and clothes to the former group. They had to cultivate the land and perform rituals in people’s houses. Chöje Lingpa ZDVWKHRQHZKRJDYHWKH¿QLVKLQJWRXFKWRWKHSODFHWRDOOSLOJULPDJH places, and he was the one who taught the people. When Guru Rinpoche MIGRATION NARRATIVES 145

stayed in Pachakshiri he established all pilgrimage places, but these places had to be discovered by a different Lama. This is a list of the Lamas who came to Pachakshiri: Kham Kathog Lama Sogyal Rinpoche, Lama Lodre Gyamtso, Yabme Pawo Dorje, Taktsan Nulden Dorje, Orgyen Chögyur Dechen Lingpa, Chöje Lingpa from Powo with his father and brother, Chöje Lingpa from Kham, Taklung Ngawang Trakpa, Lama Thondrup, and Lama Pasang. He was the last from outside. After him, and in between the other Lamas’ visits, local Lamas were in charge. Chöje Lingpa wished, ³%HFDXVHLWLVGLI¿FXOWWRFRPHWRWKHSODFHVRPHRQHZKRLVERUQLQWKH year of the monkey should come and be in charge of the place.” This person was Kunsang Dechen Rangdrol from Bhutan. After he passed away, his son Trinley Norbu Chöwang Naksang is now responsible. 7KLVYHUVLRQRIWKHVWRU\IRFXVHVXSRQWKHDGYHQWXUHVDQGDFWVRIVSHFL¿F Lamas who, by means of their power as accomplished Buddhist masters, GH¿QHGWKHERXQGDU\RIWKHBeyülEXW\HWQRWLQ¿QDOLVHGIRUPWKHNH\ hidden at Künse Lhakhang promises the option of expansion to the south in a future time.

Figure 7.7. Pemajeling with meditation caves to the right (photograph by Kerstin Grothmann, 2009).

It might be possible to throw some additional light upon aspects of the migration of the Memba by identifying several historical Lamas ZKRDUHNH\¿JXUHVLQWKHVWRU\DQGVWLOOUHOHYDQWWRWKH0HPEDWRGD\ namely Lama Lodre Gyamtso and Tertön Chöje Lingpa. Because of his founding deeds in the narrative, Lama Lodre Gyamtso has a primary VLJQL¿FDQFH LQ WKH RUDO WUDGLWLRQ7KHUH DUH WZR LQWHUHVWLQJ FDQGLGDWHV 146 KERSTIN GROTHMANN

ZKRPLJKWEHLGHQWL¿HGZLWKKLP,IZHDVNWKH0HPEDWRGD\DERXWWKH identity of Lama Lodre Gyamtso, they answer that he was the Merag Lama and founder of the Tawang monastery in Mönyul to the west. In some accounts, Merag Lama and his disciples are forced to leave their religious properties in northeast Bhutan, and retreat to the Tawang area where they establish their new monastery.477KLVLGHQWL¿FDWLRQUHÀHFWV Memba traditions about earlier migrants to Pachakshiri coming from the 7DZDQJUHJLRQ$QRWKHU/RGUH*\DPWVRZDVWKHKHURZKR¿UVWRSHQHG up the famous Rongkor pilgrimage route at Tsari, not far to the west of 3DFKDNVKLUL+LVMRXUQH\ZDVDUHVXOWRIÀLJKWIURPKLVKRPHDUHDGXH to a murder.48 He is of interest because Tsari is clearly assimilated to Pachakshiri in the beginning of the Memba narrative, and his role at both Tsari and Pachakshiri were essentially the same: a Lama who ventures down into a wild valley south of the Himalayan divide where the Lopa live and who successfully traverses and opens the area for others to come in and consolidate. Beyond any speculations about his exact identity, we note that tradition places both Lodre Gyamtso candidates as being active sometime between the 1640s and the beginning of the eighteenth century and thus contradicts the claim in the text that he was a contemporary of 7KDQJWRQJ*\DOSRZKROLYHGGXULQJWKHIRXUWHHQWKDQG¿IWHHQWKFHQWX- ULHV7KHRWKHUVLJQL¿FDQWHDUO\DFWRULQWKH0HPEDVWRU\RI3DFKDNVKLUL is Chöje Lingpa, who was active around the late seventeenth and early eighteenth centuries. Various sources place his activities at the three Tibetan Buddhist sanctuaries that extend south of the Himalayas in the eastern zone, namely Tsari, Pachakshiri and Pemakö, all of which are LGHQWL¿HGDVµKLGGHQODQGV¶RIUHIXJHDQGUHDOLVDWLRQE\YDULRXVSUDFWLWLR- ners and schools of Tibetan Buddhism. Some sources have it that Chöje Lingpa went to Pemakö during the Dzungar invasion of Central Tibet in 1717, where he eventually converted local Lopa to Buddhism.49 We currently lack adequate historical support to be certain about the

47 The Merag Lama, Lodre Gyamtso, born at Berkhar village near Tawang, was a contemporary of the Fifth Dalai Lama (1617-1682), and also of the Zhabdrung Ngawang 1DPJ\DO " RI%KXWDQ0LFKDHO$ULVLGHQWL¿HGWKLVVDPH/RGUH*\DPWVRDVWKH Lama Nagseng (Bla ma Nag seng) of Merag mentioned in an important Bhutanese historical document dating from the second half of the seventeenth century; see Aris 1997:18-19, Aris 1986:99-101, 107, 119, n.36, Sarkar 2006:7-12. 48 See Huber 1999:155. 49 Lazcano 2005:47 writes of Chöje Lingpa: “In 1717, in the middle of the invasion of Lhasa by the Dzungar Mongols, he headed for Padma bkod by way of sPo yul, where he passed on his teachings to the Klo pa and converted many of them.” A small statue of Chöje Lingpa exists in the old monastery in Mechukha valley and is said to have miraculously survived two destructions of the site. MIGRATION NARRATIVES 147

SUHVHQFHDQGDFWLYLWLHVRI¿JXUHVVXFKDV/RGUH*\DPWVRDQG&K|MH/LQJSD LQWKH0HPEDVWRU\1HYHUWKHOHVVWKHHOHPHQWVDVVRFLDWHGZLWKWKHVH¿J- ures are consistent on two points. Firstly, they are all connected, in one way or another, with the idea of taking refuge and founding new religious communities and institutions in marginal places south of the main eastern Himalayan watershed. Secondly, they suggest migration into Pachakshiri may be dated to the late seventeenth and early eighteenth century.

THE SOCIAL LIFE OF ORIGIN AND MIGRATION STORIES

As I pointed out above, texts in the form of guidebooks provide useful information for pilgrims who are unfamiliar with the holy places they visit. But this doesn’t really apply to the Memba who are all familiar with their whole valley and its religious aspects. We have to ask why the oral tradition about origin and migration into Pachakshiri is still a vibrant part of contemporary life in the Memba community? One answer to the above question is that, in spite of Memba stories about cultural superiority due to their possession of a script that their southern neighbours lack, there is a real absence of proper literacy in Tibetan. Awareness of this, especially on the part of a few literate persons who function as local ‘cultural experts’, makes the transmission of oral traditions more acutely important for the sake of maintaining the past as a viable resource for the community. Storytelling also is still a much appreciated form of entertainment, as I witnessed when out in the valley with older, knowledgeable persons who could tell stories about the Pachakshiriba’s history to younger, enthusiastic hearers who were HQJDJHGLQWKHLUGDLO\EXVLQHVVEHWZHHQYLOODJHDQG¿HOGV The on-going negotiation of local and external identities is also associated with oral stories about Pachakshiri’s past. This is partly in terms of statements of self-consciousness and distinctiveness towards Tibetans, whom the Memba feel have always looked down upon them to some degree as dwellers in a border zone between civilised and uncivilised worlds. The Memba do not think of themselves as the people who, according to the label Mönpa, were subjugated and converted to Buddhism. They are rather the chosen people who, despite their diverse origins, form a new and ideal society within which threatened religion will be preserved. This is only one component of a much more complex, and often also positive, identity relationship with Tibetans today.50 Closer 50,QUHFHQWWLPHVWKHDWWLWXGHWRZDUGV7LEHWDQVLQH[LOHKDVFKDQJHGDVWKH0HPEDEHQH¿W from the support of the Tibetan Government-in-Exile. For example, there are quotas for 148 KERSTIN GROTHMANN to home, the oral traditions are a statement of superiority over the neighbouring non-Buddhist groups and of local rights in relation to outsiders. They act as a resource against potential territorial bids by neighbouring communities who claim to have been originally settled in Pachakshiri during earlier times before integration by the modern Indian state. Thus, WKHERXQGDULHVGH¿QHGLQWKHVWRULHVE\WKHDFWLYLWLHVRI/DPDVDQGIHD- tures in the landscape such as foot prints, meditation caves or religious monuments, are revitalised by the oral tradition and also now by new pilgrimage processions associated with them. Furthermore, Mechukha has become the site of a very large Indian military facility, which has also brought with it certain religious claims over Memba sites within the valley.51 The lively local oral tradition about Pachakshiriba origins and history is ¿QDOO\DUHVSRQVHWRWKHRI¿FLDOWULEDOODEHODSSOLHGWRWKHFRPPXQLW\E\WKH Indian administration. To a certain degree, the Pachakshiriba do pragmati- FDOO\LGHQWLI\WKHPVHOYHVZLWKWKHLURI¿FLDOWULEDOLGHQWLW\ZKHQWKH\GHDO with the administration or with other tribal groups. But the name Memba is perceived by them as a degradation, and among themselves the older QDPH1lQDQJLVVWLOOLQXVH,QVSLWHRIWKHYDULRXVEHQH¿WVWKH\VKRXOGDQG VRPHWLPHVGRUHFHLYHGXHWRWKHLUVWDWXVDVDQRI¿FLDO6FKHGXOHG7ULEH resentments against the Indian state are just below the surface in relation to WKHLURI¿FLDOLGHQWL¿FDWLRQ$VRQHLQIRUPDQWSXWLW³67>6FKHGXOHG7ULEH@ and SC [Scheduled Caste] are there to uplift the backward people. Even I don’t feel like backward, but the government made us backward. Even the CM [Chief Minister, Dorjee Khandu, a Mönpa from Tawang] is backward; we have to accept that we are backward.”52

Memba students to attend Tibetan exile schools and monasteries in India. During the Chinese LQYDVLRQLQVRPH0HPEDHYHQWRRNWKHFKDQFHDQGMRLQHGWKHZDYHRI7LEHWDQVÀHHLQJ to India and proclaimed themselves to be refugees. 51 After 1987, a Gurudwara was established there by the Sikhs serving in the Indian Army. The meditation cave called Pemashelphug were Padmasambhava once meditated, now also houses an image of Guru Nanak. The surroundings of the place are well maintained by the Sikhs, and the cave is strikingly advertised with a local notice board, and also in publications as a meditation place of Guru Nanak. See also Huber 2008:245-46. 52 Interview with Dawa Tashi Naksang from Tachingphanga in April 2008. MIGRATION NARRATIVES 149

BIBLIOGRAPHY

Aris, M. 1979. “Notes on the History of the Mon-Yul Corridor.” In M. Aris and Aung San Suu Kyi (eds.), Tibetan Studies in Honour of Hugh Richardson. Warminster: Aris & Phillips, pp.9-20. —— 1986. Sources for the History of Bhutan. Wien, Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien (Wiener Studien zur Tibetologie und Buddhismuskunde, 14). Badu, Tapoli. 2002. Memba Language Guide. Itanagar: Directorate of Research, Arunachal Pradesh. Bailey, F.M. 1914. Report on an Exploration on the North-East Frontier 1913. Simla: Government Monotype Press. Bellezza, J.V. 2008. Zhang Zhung. Foundations of Civilization in Tibet. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Bentinck, A. 1913. The Abor Expedition: Geographical Results. The Geographical Journal 41(2): 97-109. Billorey, R.K. 1998. “The Memba of Arunachal. A brief note on their migration.” In S. Karotemprel (ed.), The Tribes of Northeast India. Shillong: Centre for Indigenous Cultures. pp.63-67. Blackburn, S. 2003/2004. Memories of Migration: Notes on legends and beads in Arunachal Pradesh. India European Bulletin of Himalayan Research 25/26: 15-60. Childs, G. 1999. Refuge and Revitalization: Hidden Himalayan Sanctuaries (Sbas-yul) and the Preservation of Tibet’s Imperial Lineage. Acta Orientalia 60: 26-58. Choudhury, S.S.D. 2004. “Memba. Buddhist Community of Arunachal Pradesh.” In N.S. Bisht and T.S. Bankoti (eds.), Encyclopaedic Ethnography of the Himalayan tribes, vol.3. Delhi: Global Vision Publishing House, pp.1036-40. —— 1995. “Memba”, In K.S. Singh, P. Dutta and S.I. Ahmad (eds.), People of India. Vol. XIV. Arunachal Pradesh. Calcutta: Anthropological Survey of India/Seagull Books, pp.195-99. Choudhury, S.D. 1994. Arunachal Pradesh District Gazetteers. East Siang and West Siang Districts. Itanagar: Government of Arunachal Pradesh. Dunbar, G.D-S. 1916. “Abors and Galongs: Part III. Personal Narrative of a Visit to Pemakoichen.” In Memoirs of the Asiatic Society of Bengal. Extra No. February, pp.93-114. Dutta, D.K. 1998. A Visit to Mechukha: Land of the Membas. Resarun 24: 36-40. —— 2000. An ethnographic note on the Membas of Arunachal Pradesh. Resarun 26(1-2): 45-64. —— 2006. The Memba of Mechukha Valley. Arunachal Pradesh. Itanagar: Himalayan Publishers. Erhard, F-K. 1999. “The Role of “Treasure Discoverers” and Their Writings in the Search for Himalayan Sacred Lands.” In T. Huber (ed.), Sacred Spaces and Powerful Places in Tibetan Culture. Dharamsala: Library of Tibetan Works and Archives, pp.227-39. Gyatsho Döndrub (rGya-mTsho Don-Grub). 1989. An Account of the History of Spo-smad Ka-gnam Sde-pa. Tibet Studies 1: 83-9. 150 KERSTIN GROTHMANN

Harman, H.J. 1915. “Report on the Exploration of in Eastern Tibet, 1878-79, Survey of India.” In Records of the Survey of India, Vol. VIII (part 1): Exploration in Tibet and Neighbouring Regions, 1879-1892 'HKUD 'XQ 2I¿FH RI WKH Trigonometrical Survey, pp.209-13. Huber, T. 1994. “When What You See is Not What You Get: Remarks on the Traditional Tibetan Presentation of Sacred Geography.” In G. Samuel, H. Gregor, and E. Stutchbury (eds.), Tantra and Popular Religion in Tibet. New Delhi: Aditya Prakashan, pp.39-52. —— 1997. “Ritual and politics in the eastern Himalaya: the staging of processions at Tsa-ri.” In S. Karmay and Ph. Sagant (eds.), Les habitats du Toit du monde. Nanterre: Société d’ethnologie, pp.221-60. —— 1999. The Cult of Pure Crystal Mountain. Popular Pilgrimage and Visionary Landscape in Southeast Tibet. New York: Oxford University Press. —— 2008. The Holy Land Reborn. Pilgrimage and the Tibetan reinvention of Buddhist India. Chicago: The University of Chicago Press. —— 2010. “Relating to Tibet: Narratives of Origin and Migration among Highlanders of the Far Eastern Himalaya.” In S. Arslan and P. Schwieger (eds.), Tibetan Studies: An Anthology. PIATS 2006: Tibetan Studies: Proceedings of the Eleventh Seminar of the International Association of Tibetan Studies, Königswinter 2006. Andiast: International Institute for Tibetan and Buddhist Studies, pp.297-335. —— 2011. “Pushing South: Tibetan Economic and Political Activities in the Far Eastern Himalaya, ca. 1900-1950.” In A. McKay and A. Balicki-Denjongpa (eds.), Buddhist Himalaya: Studies in Religion, History and Culture. Proceedings of the Golden Jubilee Conference of the Namgyal Institute of Tibetology, 2008, vol.1. Gangtok: Namgyal Institute of Tibetology, pp.259-76. Lazcano, S. 2005. Ethnohistoric Notes on the Ancient Tibetan Kingdom of sPo ERDQGLWV,QÀXHQFHRQWKH(DVWHUQ+LPDOD\DVRevue d’Etudes Tibétaines 7: 41-63. Pommaret, F. 1999. “The Mon-pa revisited: In search of Mon.” In T. Huber (ed.), Sacred Spaces and Powerful Places in Tibetan Culture, Dharamsala: Library of Tibetan Works and Archives, pp.53-73. Sarkar, N. 2006. Buddhism Among the Monpas and Sherdukpens. Itanagar: Directorate of Research, Government of Arunachal Pradesh. Schwieger, P. 2002. “A Preliminary Historical Outline of the Royal Dynasty of sPo- bo.” In K. Kollmar-Paulenz and C. Peter (eds.), Tractata Tibetica et Mongolia. Festschrift für Klaus Sagaster zum 65. Geburtstag. Wiesbaden: Harrassowitz, pp.215-29. Shastri, B. 1968. Notes on the Tour by the Philologist in the Siang District in January and February 1968. Itanagar: Govt. of Arunachal Pradesh Archives. Directorate of Research. Singh, K.S. (ed.) 1994. The Scheduled Tribes: The People of India. (National Series. Vol. III, Anthropological Survey of India). New Delhi: Oxford University Press. Statistical Handbook of West Siang District – 2005. 2006. Along: District Statistical 2I¿FH$ORQJ:HVW6LDQJ'LVWULFW MIGRATION NARRATIVES 151

Tibetan Language sources bDe rab Tshe rdor. 1988. Nga sgar chags gzhis sdod las thog skabs hin rdu’i btsan ‘dzul dmag gis sbas chags shri btsan bzung byas pa’i gnas tshul dngos [The real story of the occupation of Pachakshiri by Indian invaders at the time when I was at the headquarters of the estate]. In: Bod kyi lo rgyus rig gnas dpyad gzhi’i rgyu cha bdams bsgrigs [Collection of Articles on the History of Tibet], vol. 10. Beijing: pp.73-77. Shing sdong. 1988. Sbas chags shri ni rang rgyal gyi mnga‘ khongs bsnyon med yin [Pachakshiri is an undeniable territory of our country]. In: Bod kyi lo rgyus rig gnas dpyad gzhi’i rgyu cha bdams bsgrigs [Collection of Articles on the History of Tibet], vol. 10. Beijing: pp.65-72. Sbas lcags shing yul gi lam yig mthong ba rang grol [The Itinerary of sBas lcags shing valley called ‘Self-liberation on Seeing’]. Revealed by bDud ‘joms rtsal from the rKe tshal cave in Kong po. Handwritten dbu can manuscript in dpe cha format, 8 folios (r/v).

THE LANGUAGE, CULTURE, ENVIRONMENT AND ORIGINS OF PROTO-TANI SPEAKERS: WHAT IS KNOWABLE, AND WHAT IS NOT (YET)

MARK W. POST

RECONSTRUCTING THE SHADOWS OF HISTORY

This chapter represents a linguist’s effort at pre-historical reconstruction—an inherently interdisciplinary enterprise—in an area where there is not only no well-established interdisciplinary framework within which to couch one’s research, but where there also exists no more than a pale fraction of the data (from all relevant disciplines, certainly including linguistics) that would normally be required in order to form solid, defensible hypotheses of any appreciable strength. So what’s the point? :HOOWKHUHDUHUHDOO\WZRWKH¿UVWSRLQWLVWRVXPPDUL]HZKDWLVLQIDFW known (at least, to linguists), so that we can begin to assess what kind of “picture of the past” the facts suggest. The second is to help initiate a conversation between linguistics and other disciplines in the area, so that we can begin to understand what kinds of data and hypotheses we’re currently able to offer one another, and/or may eventually be capable of offering one another, given additional data, time, resources, and so on. Ultimately, my goal is not to form a grand, sweeping hypothesis of the type that would make a good newspaper headline (e.g., Johnson 1996). Rather, my hope here is simply to take whatever small steps the data FXUUHQWO\SHUPLWPHWRZLWKLQWKHFRQ¿QHVRIP\PHWKRGRORJLHV:LWK any luck, this may be able to stimulate further thinking either in similar directions or, just as welcomely, in opposing ones.

THE TANI CULTURAL-LINGUISTIC AREA TODAY

My main interest here is in “Tani”, being a name for both a set of related languages and a set of related cultures found primarily in the central part of the modern North East Indian State of Arunachal Pradesh, and to a lesser H[WHQWLQ7LEHWDQG$VVDP ¿JXUH /LQJXLVWLFDOO\WKH7DQLODQJXDJHV form a basically coherent branch of the Tibeto-Burman (TB) language family, about which more below, while the majority of traditional Tani 154 MARK W. POST

SRSXODWLRQV DUH VWUDLJKWIRUZDUGO\ LGHQWL¿DEOH DV 0DLQODQG 6RXWKHDVW Asian hill tribal in terms of culture and socio-economic organization (cf. Burling 1965: chapt. 4). Most Tani peoples are strongly exogamous and patrilineal, traditionally practice swidden cultivation, and observe animist/shamanist ritual practices. However, many changes to these latter two traditional norms have long been underway.1

Figure 8.1. Map of the cultural-geographical context of the Tani languages today (approximate).

On the face of things, the Tani tribes would appear to be living in a busy cultural-linguistic neighbourhood. Tibetans and other Bodic groups are found to the north and west, Indo-Aryans are found in huge numbers to the south together with TB plains tribals, non-Tani TB hill tribals are found to the east, west, and even inside the Tani area in a few pockets, and large numbers of “temporary” residents and other migrants enter the Tani area from every corner of modern India, Bangladesh and even Tibet, Nepal and Bhutan. The vast majority of today’s Tani tribespeople speak a genetically Tani language as their mother tongue, although most are also competent in at least one or two other languages of whatever type. $ VLJQL¿FDQW DQG JURZLQJ QXPEHU RI WULEDOO\ 7DQL SHRSOHV DUH

1 See von Fürer-Haimendorf 1980:chapt.10-11, and Blackburn 2010 for description of ongoing changes in Apatani society. LANGUAGE, CULTURE, ENVIRONMENT 155

Figure 8.2. Riyu (Minyong) village, East Siang District (photograph by Mark W. Post, 2008).

Figure 8.3. Galo ѪLEy (ritual specialist) and bòo (assistant) atop VDFUL¿FLDODOWDU'DO଎ғܺYLOODJH SKRWRJUDSKE\0DUN:3RVW 156 MARK W. POST

)LJXUH*DORHOGHU1\DDGRR5ܺEDD 'DDU଎ғܺYLOODJH SKRWRJUDSKE\ Mark W. Post, 2005).

Figure 8.5. Galo bride, )LJXUH0LODQJHOGHU$DPܺQ 'DDU଎ғܺYLOODJH SKRWRJUDSKE\ 0RRGԥ0LODQJYLOODJH SKRWRJUDSK Mark W. Post, 2007). by Mark W. Post, 2008). LANGUAGE, CULTURE, ENVIRONMENT 157 however, in fact native speakers of a genetically non-Tani language – most often, Hindi or Assamese. Throughout most of central Arunachal Pradesh, Hindi is used as a lingua franca among Tani tribespeople and when speaking to non-Tani (Modi 2005), while Assamese is more commonly spoken among the Mising of Upper Assam (as well as in many Assam-bordering areas of Arunachal Pradesh). English is used in several types of government functions and in education (to a certain extent), but plays only a minor role in day-to-day communication among most Tani peoples living in India. Little information of any reliability is currently available regarding Tani tribespeople living in modern-day Tibet, although one imagines that most will be at least competent in some variety of Chinese. So, the question that we want to ask is: how did it get to be this way? Or, to put the question somewhat more concretely: given the modern-day cultural-linguistic makeup and distribution of Tani peoples—geographically and in relation to one another and to their neighbours—what types of historical circumstances are most likely to have given rise to the set RIFRQGLWLRQVWKDWZHQRZ¿QG"

VIEWS ON THE PAST

There is a near-universal belief among Tani people themselves that they have not “always” occupied the area currently occupied, but have rather “migrated” to those various places. The most commonly held Tani view is one of coming “from the north”, i.e., from somewhere in modern-day Tibet or even beyond (see Huber in this volume).2 Some people—often, though not always, people who are also endowed with special spiritual powers—are able to recount lengthy and detailed histories of extended migrations which are supposed, in some cases, to have happened over hundreds of years.3 Ultimately, most Tani tribespeople have traditionally traced all of their lineages to a single individual, Abo Tani, literally, the ‘father (of) mankind’ (see both Aisher and Blackburn in this volume).4 Tani accounts of migration with which I am familiar are typically

2 See Blackburn 2003/2004 for detailed discussion along these lines, primarily with reference to the Apatani. 3 A detailed account of the “Adi” migrations via the Subansiri and Siang Rivers based on such oral histories is provided by Nyori 1993. 4 The single exception of which I am aware is that of the Milang, whose traditional legends make little or no reference to Abo Tani (see §4.1, as well as Sun 1993:§3.4.2 and Post and Modi (in press) for additional discussion of Milang exceptionalities). On the exception of the Mra, see Huber in this volume. 158 MARK W. POST

IUDPHGVXFKWKDWWKHGLUHFWDQFHVWRUVRIDVSHFL¿FFXUUHQWO\LGHQWL¿DEOH population—usually, a tribe, subtribe, clan grouping, or even a single clan—are viewed as having shifted, en masse, from one geographical DUHDWRDQRWKHU'LYHUVL¿FDWLRQLVXVXDOO\YLHZHGDVKDYLQJRFFXUUHGDV a direct outcome of migration; for example, many of my consultants on WKLVWRSLFKDYHH[SODLQHGGLDOHFWGLYHUVL¿FDWLRQLQWHUPVRIWKH³LQÀX- ence” of different “climactic conditions” encountered as populations shifted from the cooler northern highlands to the warmer, lower south. According to one respected view, the ancestors of today’s Galo tribespeople are thought to have migrated southward, as a group, from Tibet via the Siang River, to have continued south-westward via the Siyom River, to have settled for some time in a sort of earthly paradise called Chibo-Nyobo, and to have then dispersed—clan by clan – to the different locations in which Galo clans are in fact concentrated today (Geiyi Undated: 131-132). Dialectal differences—as well as certain cultural differences – which may be found among the Galo today are by this account attributed to the initial group dispersal. In some cases, migrations are a matter of living memory. For example, most of the Galo villages now found along the Arunachal border with Assam were founded in the mid-to-late 20th century by individuals who are either still alive or who have expired so recently that their accounts were able to be recorded by their descendants. In almost all cases of which I am aware, what is recounted by them is the migration of a spe- FL¿FJURXS²XVXDOO\PHPEHUVRIWKHVDPHFODQKDLOLQJIURPWKHVDPH village—to an area previously uninhabited (or inhabited only by non- Tani), due to the need for additional cultivable land. Once established, the populations of such villages increased primarily by means of further migrations of families and individuals hailing from the pioneers’ home village(s) (see Childs in this volume). Popular accounts of “mass migrations” have been problematised by Burling (2007a, and this volume), among others, and I do not intend to take up the topic in any great detail here. What I will say is simply that population movements certainly have occurred in the Tani area within historical times, and are believed by many Tani people to have occurred in pre-historical times as well. Whether one assumes that pre-historical population movements would have more closely resembled the “micro- migrations” we typically observe today (see Huber in this volume) or might have been something on a grander scale (as oral histories often suggest) is to an extent unimportant for present purposes. What is important to me here is that Tani cultural memories include the concepts LANGUAGE, CULTURE, ENVIRONMENT 159 of 1) an original unity 2) population splits 3) population movements and FRQVHTXHQW GLYHUVL¿FDWLRQV$UH VXFK YLHZV ZHOOVXSSRUWHG E\WKH linguistic evidence?

AN EFFORT AT RECONSTRUCTION

Genetic Perspective

There are many extravagant claims in circulation regarding the “relatedness” or “ancestry” of various world languages and the populations which are supposed to have spoken them. Without going into any great detail here,5 it is important to understand that, in the sense in which linguists have traditionally used the term, a hypothesized “relatedness” between two languages explicitly entails a claim that they were once one ODQJXDJHZKLFKHYHQWXDOO\VSOLWDQGGLYHUVL¿HG,QWKHLGHDOFDVHWKH GDWDDYDLODEOHWRWKHUHVHDUFKHUDUHVXI¿FLHQWWRVXSSRUWDK\SRWKHWLFDO reconstruction of the common ancestral language, which is in essence what remains after all of the systematic differences between the two or more related languages are “undone” by working backwards in time.6 The key words here are “data available” and “hypothetical”. In absence of written records (and sometimes even with them) there is no way our methods can uncover “facts” about the past. What there are, then, are hypotheses which are more or less well-supported by the data available to the researcher, which, in turn, may be more or less conducive in nature to the methods s/he is trained to apply. Ultimately, every statement of “relatedness”, every reconstructed “ancestral language” and every linguistic “family tree” must be evaluated with these facts in mind. On the basis of a survey of all Tani language data available at the WLPH7LDQ6KLQ-DFNVRQ6XQ ZDVDEOHWRVKRZDVGH¿QLWLYHO\DV our methods allow that the Tani languages are related, in the abovemen- tioned sense that they share at least one common ancestor which can be 5 See Heggarty 2007, particularly §2, for an excellent and succinct discussion of the principles of historical linguistics from a general perspective. See also Dixon 1997 for extended discussion on the relationship between linguistic “events” and their correlates in terms of the histories of human populations. 6 For example, a t in languages A and B may correspond to cࡊLQODQJXDJH&LQHYHU\ word in which these sounds are followed by a vowel i. Since t becoming cࡊEHIRUHi is a very common type of change in the world’s languages, and one which obeys well-understood phonetic principles, we suppose that the common ancestor of all three languages – which PLJKW EH FDOOHG ³3URWR$%&´ ± KDG WKH VHTXHQFH ti in these words, just as in modern languages A and B. Language C would have undergone a sound change, presumably after the community which spoke C split off from the other ABC-speaking populations. 160 MARK W. POST systematically reconstructed to some extent. This ancestral language, whose actual historical name is unknown to us, is then referred to as Proto-Tani. By comparing correspondences between modern Tani languages in his data, partially reconstructing a common ancestor, and then determining at which points different ancestral Tani languages would have separated from one another (and thus began to change), Sun was DEOHWRGHULYHWKHSURYLVLRQDOIDPLO\WUHHJLYHQLQ¿JXUH

Figure 8.7. Provisional Tani family tree (based on Sun 1993:272).

As shown, the most important division is between a Western and an (DVWHUQEUDQFK,QWKHRU\WKLVZRXOGUHÀHFWDVLQJOHHDUO\GLYLVLRQRI Proto-Tani-speaking people into two groups, one which presumably stayed in or moved toward the west (and subsequently underwent further VSOLWVDQGGLYHUVL¿FDWLRQ DQGDQRWKHUZKLFKVWD\HGLQRUPRYHGWRZDUG the east (and did the same). This basic division can be straightforwardly LOOXVWUDWHGE\ORRNLQJDWKRZWKHZRUG³¿VK´LVUHSUHVHQWHGDFURVVPDQ\ RIWKH7DQLODQJXDJHVRIWRGD\ ¿JXUH $VZHFDQVHHLQWKH¿JXUHDOORIWKH:HVWHUQ7DQLODQJXDJHVKDYH a form ending in -i, while all of the Eastern Tani languages have a form beginning in a vowel a-, Ω or o-. For reasons which I won’t go into, LWLVSRVVLEOHWRGHWHUPLQHWKDWWKHRULJLQDO7DQLURRWIRUµ¿VK¶ZRXOG have been ƾy, but that there would have been variation in the proto- ODQJXDJHEHWZHHQWZRGHULYHGZRUGV DƾRғµ¿VK¶DQG ƾyLғµ¿VKLH¶7KH ¿UVWIRUPUHÀHFWVDSUH¿[DWLRQLQ a-, signalling the word’s status as a LANGUAGE, CULTURE, ENVIRONMENT 161

Western Lg. Form Eastern Lg. Form Apatani ƾѠ-i Bori ΩƾR Bengni ƾX-i Bokar o-ƾRR Galo ƾR̗t Damu a(a)-ƾR Hills Miri ƾX-i Milang a-ƾX̗ Nishing ƾX-i Mising o- ƾ ƾR Nyisu ƾR-i Padam ΩƾR Tagin ƾX-i Minyong ΩƾR Yano ƾD-i Pasi ΩƾR

)LJXUH0RGHUQUHÀH[HVRI3URWR7DQL ƾRғµ¿VK¶ DGDSWHGIURP6XQ ³EDVLFQRXQ´ZKLOHWKHVHFRQGIRUPUHÀHFWVD'LPLQXWLYHVXI¿[DWLRQ something like English -y/-ie (Post 2007:259, 844). The Proto-Eastern- 7DQLVSHDNLQJ FRPPXQLW\ SULYLOHJHG WKH ¿UVW IRUPDWLRQ ZKLOH 3URWR Western-Tani speakers privileged the second. The daughter languages which later split off from these two ancestral languages inherited and consolidated ancestral speech habits, and the “alternative” form in both cases was lost. But the data are not always so kind. Although dozens of examples similar to the above may be adduced in support of a basic Western- Eastern split, there are also examples of sound changes or shared vocabulary and/or grammar among languages which span the Western/ Eastern divide. The best examples of this are often found in Galo, the most geographically central of the Tani languages. Figure 8.9 illustrates VRPHPRGHUQUHÀH[HVRIWKH3URWR7DQLQRXQ a-rjék ‘pig’ and verb root OjƾµWDNH¶LQYDULRXVPRGHUQ7DQLODQJXDJHV,QWKH¿UVWFDVH*DOR VKDUHVDQ³,QLWLDOFOXVWHUVLPSOL¿FDWLRQ´VRXQGFKDQJHZLWKWKH(DVWHUQ Tani languages, in which the reconstructed Proto-Tani consonant cluster rj- LV VLPSOL¿HG WR EHFRPH RQO\ RQH FRQVRQDQW 7KLV PLJKW VXJJHVW ±GHI\LQJWKHLQIRUPDWLRQSUHVHQWHGLQ¿JXUH±WKDW*DORLVLQIDFW an Eastern Tani language, sharing a common ancestor with Bokar and 0LVLQJDPRQJRWKHUV,QWKHVHFRQGFDVHKRZHYHU*DORKDVORVW¿QDO ƾDQGUHSODFHGLWZLWKDOHQJWKHQLQJRIWKHYRZHO -a-, like (other) Western Tani languages but unlike languages of the Eastern branch. 7KLVZRXOGVXSSRUWWKHVFHQDULRRXWOLQHGLQ¿JXUHLQZKLFK*DORLV aligned with the Western branch. So what’s going on here? Well, the actual distribution of the modern 162 MARK W. POST

Branch Western Transitional? Eastern Language Bengni Apatani Lare Pugo Bokar Mising Galo Galo PT *a-rjék ‘pig’ Ȓ PHȡI ? JHGɍ ȑ PȑʴI C H¤I ȑ HȑI C HCI PT Ojƾ‘take’ laa- laa- làa- làa- JMĩ J?ĩ

Figure 8.9. Cross-branch innovations, contact and the genetic position of Galo. Tani languages, and the kind of historical picture that this distribution suggests, is considerably more complex than the neatly-branching dia- JUDPLQ¿JXUHPLJKWOHDGRQHWREHOLHYHKRZHYHUZHOOVXSSRUWHG it is in its basic outlines. The Tani languages in fact largely form what is sometimes called a “dialect chain”, in which nearly every population is able to converse with its neighbours, but in which distant populations cannot generally understand one another. The best illustration of this is again Galo, whose speakers are, by virtue of their geographically central location, in contact with speakers of Nyishi, Hills Miri, Tagin, Bokar, Ramo, Pailibo, Minyong, and, nowadays at least, Mising. With the possible exception of Mising, Galo speakers are able to converse to a degree with speakers of all these languages, irrespective of the branch of origin (depending, of course, on the Galo dialect in question and the VSHFL¿F ODQJXDJH H[SHULHQFHV RI WKH LQGLYLGXDOV LQYROYHG ,Q RWKHU words, genetic relatedness is not the only dimension we need to consider when attempting to characterize the distribution and history of the Tani languages. We also need to consider contact and areal diffusion. With a handful of exceptions to be discussed shortly, the type of picture that emerges from extended study of Tani dialects is one of long-term stability. Generally lacking many of the abrupt “punctuations” (sharp structural divergences, discontinuity in the data, etc.) which are often thought to UHÀHFWSHULRGVRILQVWDELOLW\DQGORVVRISRSXODWLRQFRQWDFWV 'L[RQ Tani languages on balance tend to exhibit a kind of “equilibrium” (structural convergences, continuity in the data, etc.) which suggest a long period of inter-group contact, communication and exchange.7 Before pressing further with this characterization, I want to discuss

7 Jacquesson 2008 argues the Tani and “Naga” areas to be perfectly opposed in terms of language diversity (Tani low, Naga high), despite not being obviously different in terms of relative time-depth within Tibeto-Burman. He explains this in terms of the relatively peaceful relations traditionally found among most Tani groups, leading to frequent population exchange and the emergence of a broad overall stability. This is contrasted to the relatively hostile relations traditionally found among many “Naga” groups, which led to the formation of insular communities and sharp linguistic differentiation. LANGUAGE, CULTURE, ENVIRONMENT 163 two examples which contrast with the general picture being described: WKH¿UVWLV$SDWDQLDQGWKHVHFRQGLV0LODQJ $SDWDQLDQG0LODQJZHUHLGHQWL¿HGE\6XQ DVUHODWLYHO\³DEHU- rant” Tani languages, in the sense that they exhibit more differences from their neighbours than is the overall Tani norm. Apatani is spoken in an island-like plateau surrounded on all sides by Nyishi tribespeople and exhibits a great number of lexical and grammatical differences from other Western Tani languages. Although some Nyishi and Apatani appear to be able to converse to a degree, this would appear to owe PRUHWRVSHFL¿FODQJXDJHH[SHULHQFHVRIWKHLQGLYLGXDOVLQYROYHGWKDQ it would to any particular genetic closeness among Nyishi and Apatani within the Western Tani branch. The example of Milang is even more striking, in that it is not mutually intelligible with any other Tani language (nor, indeed, with any other known language at all). So different is Milang from all “other” Eastern Tani languages – most of which are mutually intelligible to a noticeably large extent, even by Tani standards – that a legend has come about in which it is asserted that Milang is in fact a “code language”, purpose- fully distorted in order to confuse enemies in times of war.8 As is argued in more detail in Post and Modi (in press), the character RIVSHFL¿FJUDPPDWLFDODQGSKRQRORJLFDOGLIIHUHQFHVEHWZHHQ0LODQJ and what we are currently able to reconstruct for “Proto-Tani” suggests that Milang – while most likely a genetically “Tani” language – in fact lacks many of the post-Proto-Tibeto-Burman (PTB) innovations which characterize the Tani branch. Notable examples of Milang PTB retentions include a third person singular pronoun (replaced elsewhere in 7DQL DVKRUW aYRZHO VKLIWHGWR o elsewhere in Tani), and a ‘same OHYHO¶ WRSRJUDSKLFDOGHLFWLF GHPRQVWUDWLYH LQ yo/yu UHSODFHG E\ a elsewhere in Tani) (Post and Modi in press). :KDWGRHVWKLVLQGLFDWH":HOOWZRWKLQJV7KH¿UVWLVWKDW0LODQJPD\ not in fact be a descendant of Sun’s “Proto-Tani” as such; rather, it may share a common ancestor with Sun’s “Proto-Tani” at a higher level (that is to say, at an earlier stage). This does not mean that the coherence of the Tani branch or the past existence of a set of Proto-Tani speakers are now in doubt. What becomes doubtful is rather that reconstruction of an ancestral

8 This view is neither shared nor particularly appreciated by the Milang themselves, who generally attribute it to the misunderstandings of outsiders. The view espoused by many Milang is that the peculiarities of their language derive from an ancient merger with a distinct tribe referred to as the Soi-Sotem. No corroborating information concerning the Soi-Sotem (nor any potential descendant population of the Soi-Sotem other than the Milang themselves) has yet been discovered. For additional discussion, see Modi 2008. 164 MARK W. POST language on the basis of data considered by Sun (1993) alone will enable us to characterize the status of this post-PTB population; a more complex and possibly not-fully-reconstructible set of languages and events may well have preceded Sun’s “Proto-Tani” stage at a post-PTB level. 7RUHFDSLWXODWHZKDWZH¿QGZKHQZHORRNDFURVVWKHPRGHUQ7DQL languages is a fairly homogenous and areally-continuous group char- acterized by similar sets of relatively-recently-innovated morphology, punctuated by a small handful of relatively unusual (though clearly related) languages whose more diverse structures offer hints of an earlier and more diverse past. One way of understanding the type of scenario the Tani data seem to be pointing to is provided by Burling (2007b), who argues for the linguistic effects of lingua franca use among TB-speaking populations. 9HU\EULHÀ\%XUOLQJDUJXHVWKDWXVHRISDUWLFXODUTB languages as lingua IUDQFDVKDVOHGWRWKHLUVWUXFWXUDOVLPSOL¿FDWLRQ±LQFOXGLQJWKHORVVRI earlier morphology – and the eventual innovation of new structures and PRUSKRORJ\ DV WKH OLQJXD IUDQFD LWVHOI VSUHDGV DQG GLYHUVL¿HV DV WKH “native language” of a new population. Following the Burling view, Sun’s “Proto-Tani” might be broadly characterisable as such a lingua franca. It would have shared a common ancestor with the relatively more conservative Milang language, but ZRXOGKDYHKDGDUHODWLYHO\VLPSOL¿HGDQGSUREDEO\VRPHZKDWYDUL- able structure due to use by a diverse population with different original language backgrounds; the later structural complexities of languages like Galo would have been – as was partly argued in Post (2007: chapt. 2) – largely secondary. Does this view have any plausibility? Well, in a word, yes; even to this day Padam (Adi) is used as a lingua franca in the Siang River valley (including by the Milang), and minor “lingua francisations” can, in effect, be seen nearly every time speakers of different Tani dialects come into contact and adjust their speech to meet the comprehension needs of their interlocutors.9 Indeed, the relative structural homogeneity of the Tani languages, the areal sharing of innovations and other features, and the readiness with which Tani language speakers make adjustments to maximize comprehension, are all suggestive of a history in which a premium has long been placed on inter-group cooperation and compre- hensibility; in short, an ideal set of conditions for the development and spread of a lingua franca.

9 This traditional situation is of course changing now with the spread of Hindi and Assamese, use of which has the effect of supplanting local languages entirely; Modi 2005. LANGUAGE, CULTURE, ENVIRONMENT 165

In sum, genetic evidence adduced to date presents a picture of a very early “Proto-Tani” population, whose language or languages remain to be fully reconstructed, but which no doubt has a primarily Tibeto- Burman background. Possibly through use as a lingua franca, a sim- SOL¿FDWLRQ RI LWV VWUXFWXUH RU VWUXFWXUHV RFFXUUHG ODWHU JLYLQJ XV WKH “Proto-Tani” partially reconstructed by Sun (1993). Modern structural complexities in the descendent Tani languages were probably the result RIODWHUGHYHORSPHQWVDVSRSXODWLRQVVSOLWDQGWKHODQJXDJHVGLYHUVL¿HG At the same time, however, ongoing contacts and exchange also led to structural convergence and long periods of slow, areally-shared changes and an overall cultural-linguistic stability.

Environmental Perspectives

,QWKHSUHYLRXVVHFWLRQZHLGHQWL¿HGDJHQHWLFXQLW\WRWKH7DQLEUDQFK and discussed the type of ancestor language(s) and population(s) of speakers that may have existed in early times. Assuming, then, that there were indeed one or more “Proto-Tani” populations, where might they have lived? One type of evidence which is often adduced in favour of linking posited proto-languages to particular geographical areas is sets of words found in the reconstructed proto-lexicon. If there are words in a particular proto-language for particular types of natural objects which occur in one area but not in another, it is only natural to suppose that the proto-language would most likely have been spoken in the area where the named objects are represented (Campbell 2001 [1998]:351); for instance, the prevalence of words denoting seasonal winds and other meteorologi- cal phenomena in Proto-Oceanic gives strong evidence in support of the quite natural assumption, under the circumstances, that Proto-Oceanic ZDV VSRNHQ VRPHZKHUH LQ WKH 3DFL¿F ,VODQG HQYLURQPHQW ZKHUH WKH named seasonal winds are in fact to be found (Ross, Pawley et al. 2007 [2003]:§4.2). There are of course dangers in making assumptions of this kind, as the meanings of words tend to shift over time, and existing words can quickly and easily be adapted to suit a novel object which may be notionally quite different from the originally-denoted object from some perspectives (consider English “mouse”). But as one type of evidence, it can be of considerable value. Unfortunately, the scale and variety of Tani language data published WR GDWH LV LQVXI¿FLHQW WR HQDEOH D ULFK FKDUDFWHUL]DWLRQ RI DQ\ K\SR- thetical Proto-Tani environment on the basis of lexical reconstructions. 166 MARK W. POST

However, we can note that most modern Tani languages which have been studied to date contain large sets of cognate, seemingly indigenous WHUPVGHQRWLQJKLJKODQGÀRUDDQGIDXQDDVZHOODVPRXQWDLQWRSRJUDSK\ and hillside cultivation. While space prevents me from providing large lists of such forms here, several dozen Galo forms of these types may be found in Post (2007:Appendix A), together with Proto-Tani reconstructions where available. Just as important as their existence, however, is the fact that such lexical sets tend to be structured on the basis of native morphological PDWHULDO)ROORZLQJDEDVLFDOO\FODVVL¿FDWRU\FRPSRXQGLQJV\VWHP 3RVW DOPRVWDOOWHUPVGHQRWLQJKLJKODQGÀRUDDQGIDXQDLQDGGLWLRQWR many other types of items, are formed from a native, morphologically LUUHGXFLEOH¿QDOURRWGHQRWLQJDW\SHRILQGLYLGXDO DQ³H[HPSODU´ SOXV D QDWLYH IRUPDWLYH SUH¿[ RU LQLWLDO URRW ZKLFK GHVLJQDWHV WKH RYHU- all semantic class to which the individual belongs (a “type”). So, for example, in Galo Ω]z ‘variety of bamboo (Bambusa tulda)’, the initial Ω̖ - formative denotes ‘bamboo’ in general (cf. GaloΩ̖ Ω ‘bamboo’), while the second formative zò- effectively denotes the species tulda. Moreover, FODVVL¿FDWRU\RUJDQL]DWLRQSURFHHGVYHUWLFDOO\DVZHOODVKRUL]RQWDOO\ That is to say, in addition to relationships among a set of terms (say, “higher animals”) which all bear a particular semantic-class-denoting IRUPDWLYH LQWKHODWWHUFDVHDUHÀH[RIPTB *s(y)a- ‘meat/animal’), there are relationships among semantic classes too, as an exemplar-denoting formative gives rise to the type-denoting formative of a daughter seman- WLFFODVV ¿JXUH 7KHH[LVWHQFHRIVXFKKLJKO\VWUXFWXUHG³IDPLOLHV´ of conceptually-related terms – not simply isolated, arbitrarily-shaped lexical items – suggests a long period of intimacy and co-evolution with the items represented. Another highly structured aspect of Tani languages which bears on geographical and/or environmental conditions is topographical deixis (TD). Found widely in TB (though not universally) TD is a means of grammatically and/or lexically encoding the orientation of a referent vis-à-vis a shifting reference point (usually, the place of speaking) in terms of one of three possible planes: 1) upward 2) downward and 3) on the same or an unknown level. In Tani languages, TD is represented in four linguistic VXEGRPDLQVPRWLRQYHUEVGLUHFWLRQDOVXI¿[HVGLVWDOGHPRQVWUDWLYHVDQG demonstrative postpositions (Post 2010). A sample set of terms illustrating TDIRUWKH/DUHGLDOHFWRI*DORLVJLYHQLQ¿JXUH As I have argued in Post (2011), systems of TD appear to evolve in the context of a frequent and intimate interaction with a topographically varied LANGUAGE, CULTURE, ENVIRONMENT 167

Figure 8.10. Taxonomic structure in the Tani lexicon (Pugo dialect of Galo).

Motion Directionals Distal Locative dem. verbs demonstratives postpositions UP/NORTH/UP RIVER càa- -càa WΩ̖ tolò DOWN/SOUTH/DOWN ìi- -lòo EΩ̖ bolò RIVER SAME-UNKN. LEV./E-W/ áa- -áa àa DOy NON-RIVER-ORIENTED

Figure 8.11. Representative set of terms employing topographical deixis in Lare Galo. HQYLURQPHQWVXFKDVWKDWIRXQGWKURXJKRXWPRVWRIWKH7DQLDUHD ¿JXUH 8.12). From this topographical origin, the system may develop certain other extensions (such as compass and riverine orientation), according to the salience of such features in the environment and/or their de facto ability WR¿WZHOOZLWKLQWKHIUDPHZRUNRIWKHV\VWHP10 As was suggested in Post (2008), the systems of TD found in most modern Tani languages – and which are reconstructible to the Proto- Tani stage – ultimately have a genetically much deeper TB (possibly PTB) origin. That is, it would seem that TD itself did not evolve within

10 The usual associations in Tibeto-Burman are XS ĺ QRUWKXSULYHUGRZQ ĺ VRXWK GRZQULYHUDQGVDPHXQNQRZQOHYHOĺHDVWZHVWXQNQRZQGLUHFWLRQQRWDORQJULYHUFRXUVH, although they may of course be different according to the particularities of the environmental context in which a language is spoken. 168 MARK W. POST

Figure 8.12: Environmental source of topographical deixis

Proto-Tani or any of its descendants. However, ongoing research into the maintenance and loss of TD in Tani and in other TB languages has suggested that TD is maintained in languages which continue to be spoken in an environment of saliently uneven terrain, and/or in areas which UHWDLQDQDGGLWLRQDOHQYLURQPHQWDOIHDWXUHRIVXI¿FLHQWVDOLHQF\WRDQFKRU the system (such as a major river, as in the case of the Brahmaputra- adjacent Mising). TD is lost in languages which cease to be spoken in hilly areas and lack an alternative environmental anchor (such as, quite likely, Proto-Bodo-Garo, as well as – to an extent – the only long-term “plateau” Tani language, Apatani). TD is also lost in languages which undergo intense contact and/or population mixture with a cosmopolitan and/or plains/non-TB language (such as Indo-Aryan languages in the case of several TB languages of Nepal). The implication would then be that, inasmuch as there was a pre- Proto-Tani ancestor which exhibited a TD system (as seems to be the case) and to the extent that this was retained in Proto-Tani itself and in most of the daughter languages, there is a relatively higher likelihood that Proto-Tani would have been spoken by hill populations who were not in intense contact with cosmopolitan/plains-based/non-TB populations than in any situation resembling the reverse (about which more in the next section). In sum, evidence from lexical items and lexical structuring, taken together with evidence from topographical deixis, suggests that the environment of Proto-Tani speakers was broadly similar to that in ZKLFKPRVW QRWDOO 7DQLVSHDNLQJSRSXODWLRQV¿QGWKHPVHOYHVWRGD\ ±QDPHO\DKLOO\DUHDZLWKÀRUDDQGIDXQDQRWDJUHDWGHDOXQOLNHWKDW found in central Arunachal Pradesh. In addition, while two Tani languages – namely Apatani and Mising – eventually underwent develop- ments due to the establishment of corresponding populations in areas which probably contrasted in some ways with the proto-environment, LANGUAGE, CULTURE, ENVIRONMENT 169 the majority of Tani languages did not undergo such changes and can in turn be assumed to have been continuously spoken within an environment which resembled that in which ancestral languages were spoken (i.e., in the hills).

Cultural-Contextual Perspectives

As I mentioned above, the modern Tani languages of North East India would appear to be spoken in a culturally busy neighbourhood, sharing borders (and sometimes villages or even domiciles) with speakers of TB languages from at least two major branches, in addition to several XQFODVVL¿HG TB ODQJXDJHV SRWHQWLDOO\ UHSUHVHQWLQJ DV PDQ\ DV ¿YH or six additional branches), and, of course, at least two enormously LQÀXHQWLDO ,QGR(XURSHDQ ODQJXDJHV +LQGL DQG $VVDPHVH11 Contact between Tani language speakers and speakers of other languages of the same area – particularly Hindi and Assamese – is having strong and undeniable effects both on modern Tani cultures and on the structures of their languages. How closely might pre-historical circumstances have UHVHPEOHGWKHVLWXDWLRQZH¿QGWRGD\" Mainland Southeast Asian languages in general, and TB languages in particular, are frequently subdivided into those with a relatively more Indospheric or a Sinospheric W\SRORJLFDO SUR¿OH12 First labelled as such by Matisoff (1990:n.17) and discussed in more detail in Matisoff (1991b:485-486), “Indospheric” languages are spoken by populations ZKLFKDUHWKRXJKWWRKDYHEHHQFXOWXUDOO\DQGOLQJXLVWLFDOO\LQÀXHQFHG by their proximity to and contact with Indic language speakers. They would tend to show characteristically Indic-area typological features as a result, such as having relatively long and morphologically complex words, having a structurally well-coded main/subordinate clause distinction and having word rather than syllable tone systems (or else having no tones), and so on. “Sinospheric” languages are in turn spoken by populations whose cultures and languages are thought to have been LQÀXHQFHGE\SUR[LPLW\WRDQGFRQWDFWZLWK6LQLWLFODQJXDJHVSHDNHUV They would tend to exhibit characteristically Sinitic-area typological features, such as relatively shorter, morphologically simpler words, a 11 The relatively smaller number of Tani languages spoken in Tibet are in turn presumably in close contact with Mandarin Chinese, in addition, potentially, to local Bodic languages; again, however, little information of certain reliability is available. 12 A well-known, recent compilation of papers by leading Tibeto-Burmanists takes the basic possibility of this bifurcation as a starting point for enquiry; see Bradley, LaPolla et al. 2003. 170 MARK W. POST less-well-coded main/subordinate clause distinction – or no such distinc- WLRQLHYHUEVHULDOLVDWLRQ±DQGDVPDQ\DV¿YHRUVL[V\OODEOHWRQHV among other traits. Tani languages, together with most other relatively well-established North East Indian languages (excluding, for example, more recent arrivals such as the Tai languages of Upper Assam and the %XUPDERUGHULQJDUHDVRI$UXQDFKDO3UDGHVK DUHGHVFULEHGDV³¿UPO\´ Indospheric by Matisoff (1991b:485). Typologically speaking, there is no doubt that this is an accurate designation. Tani languages are morphologically synthetic and aggluti- nating (i.e., they have often large and morphologically complex words, VRPHWLPHV LQFOXGLQJ DV PDQ\ DV VL[ RU VHYHQ VXI¿[HV WKH\ KDYH numerous subordinate clause types and, so far as I am aware, no true verb serialisation. Most also exhibit word tone or typologically similar systems which, while complex in terms of overall organization, have an extremely small and stable inventory (typically exhibiting only two primary lexical “tonemes”). The following set of examples from Galo (Indospheric TB), Assamese (Eastern Indo-Aryan), Lahu (Sinospheric TB) and Mandarin Chinese (Sinitic) will roughly illustrate the case (for abbreviations in example sentences 1-9, see list at chapter end):

ܺK଎ҒܺRyPiOjDU଎ҒJԥUԥғ-làa arúm=ԥҒPiDUԥғ-kú wood vegetable search.for-NF do-ACNC-NF evening=ACC come-IRR-CMPL µ$IWHUVHDUFKLQJIRU¿UHZRRGDQGYHJHWDEOHV WKH\¶OO UHWXUQLQWKH evening.’ Galo, Indospheric TB

ta-i ula-i dza-i ula-i go-i pela-i bhab-e. 3-F emerge-NF go-NF emerge-NF go-NF throw-NF think-3.SUB ‘She goes ahead on out and thinks.’ Assamese, Eastern Indo- Aryan (Data from Joana Jansen, gloss adjusted by this author) LANGUAGE, CULTURE, ENVIRONMENT 171

ࡂ ݦ FKqݦFk SԥҒ ãƝYHFrܧOkS WLJHUMXPS ELWHHDW ¿QLVK ADVS NZR QUOT ‘The tiger jumped (out) and bit (into them) and ate (them) all up!’ Lahu, Sinospheric TB (Matisoff 1991a:411)

WƗMLMu[SiGjRVKVKjQJT]KӽLOt 3 SFOC continue climb arrive tree top go pick pear ‘So he continued climbing up the tree to pick pears.’ Mandarin Chinese, Sinitic

The implication is that Lahu, being a Sinospheric language, would either have been, or have had an ancestor which was, in contact with Sinitic languages (or else with other languages which were in turn in contact with Sinitic languages). Galo, an Indospheric language, might in turn be presumed to have had such a contact history with Indic. Are there other linguistic facts available which could be brought to bear on the question? Although I do not have space here to recapitulate the arguments in any great detail, Post (2006; 2007:§2) has argued that Proto-Tani, or a near ancestor, was likely to have been a basically Sinospheric language.13 If it is the case, then, that modern Tani languages are relatively more Indospheric, does this indicate that Tani languages (or any of their ancestors) would have once been in contact with one or more Sinitic languages, but then came into contact with Indic languages at a later stage? There is very little that can be said regarding the possibility of early

13 The basic arguments concern the transparency with which much Tani grammatical morphology is reconstructible to lexical source forms (suggesting a recent origin), the lack of paradigmatic irregularities or stem alternations (which would be indicative of older grammatical morphology) and a complex predicate word-structure of a type which suggests the earlier occurrence of verb serialisation, in addition to other more general facts. 172 MARK W. POST

Sinitic contact, since it is not currently possible to distinguish between features held in common among Tani and Sinitic languages due to common ancestry (Proto-Sino-Tibetan) and those which might have entered the language at some later stage due to contact. Regarding the second possibility, however, we can say something. When language speakers have close cultural contacts, bilingualism, borrowing and structural convergence almost always occur (Thomason DQG.DXIPDQ 7KH¿UVWDQGSHUKDSVPRVWLPSRUWDQWWKLQJWRSRLQW out, then, is that there is in fact only a tiny handful of well-assimilated forms of Indic origin in all Tani languages with which I am at all familiar.14 Figure 8.13 illustrates the complete set of well-assimilated words of Indic origin which I have attested in the Lare dialect of Galo. This is an extremely small number. TB languages with truly palpable Indic contact histories, such as Newari, exhibit literally hundreds of words of Indic origin which have been incorporated at various histori- FDOWLPHV *HQHWWL $GGLWLRQDOO\¿JXUHVKRZVWKDWDOPRVWDOO well-assimilated Indic forms found in Galo represent items which are not traditional to the Tani natural and/or cultural world, and which would either have been introduced via administration and trade or via an increase in proximity on the Galo part to a previously unfamiliar natural and cultural world (as in the case of ‘sand’). Let us contrast this situation with that found in modern-day Karbi- Anglong (K-A) Mising, a Mising dialect spoken on the south bank of the Brahmaputra and currently experiencing a high degree of contact with $VVDPHVH ¿JXUH $V ¿JXUH VKRZV WKHUH LV D ODUJH QXPEHU RI EDVLF HYHU\GD\ vocabulary items which are referred to by K-A Mising speakers using words of Assamese origin, and which Lare Galo speakers refer to using a native Tani word. Although K-A Mising presents a somewhat extreme case due to its especially high modern contact situation, it is nonetheless the case that Indic loanword use is in general much higher in Mising than in any other Tani language with which I am familiar, even to the H[WHQWRILQÀXHQFLQJWKHQXPHUDODQGFODVVL¿HUV\VWHPV 'ROH\DQG3RVW in press). Given the likelihood that the ancestors of many if not all Mising speakers were originally to be found in central Arunachal Pradesh – presumably, 14 “Well-assimilated” in this case would mean forms which entered the language (or an ancestral language) long ago, and which underwent any applicable regular sound changes which have occurred in the language since then and/or which can be demonstrated to have integrated well into the native language structure, such that a foreign origin can no longer be readily discerned by at least some speakers. LANGUAGE, CULTURE, ENVIRONMENT 173

Galo Gloss Source Form Gloss azár ‘thousand’ Asm hezar ‘thousand’ pohàa ‘money’ Asm poisa ‘money’ dukàn ‘shop’ Asm dukan ‘shop’ bozár ‘market’ Asm bozar ‘market’ NXOt ‘open (for Asm kuli ‘open (in business)’ general)’ EyQG ‘close (for Asm EэQGў ‘close (in gen- business)’ eral); strike’ gám ‘village headman’ Asm gaõ ‘village’ QDKyU ‘Ceylon ironwood Asm QDƫRU ‘Ceylon iron- tree’ wood tree’ XQWΩUiD ‘orange (citrus Asm sumtra ‘orange’ fruit)’ UΩEiS ‘pomelo Asm UэEэS ‘pomelo’ (grapefruit)’ kurìi ‘cat’ Asm mekuri ‘cat’ tamúr ‘betelnut’ Asm tamul ‘betelnut’ WѠNΩ̗ ‘contract’ Asm tika ‘contract’ potáa ‘paper; letter’ Asm SэWUэ ‘paper’ SRWΩ̗ ‘license’ Asm SэWUэ ‘paper’ gurée ‘horse’ Asm JўRUD ‘horse’ hàa ‘tea’ Asm sa ‘tea’ balìi ‘sand’ Asm balu ‘sand’ NѠUNtL ‘window’ Asm kirki ‘window’ ODJt ‘want/need’ Asm lag- + -i ‘want/need; attach + NF’ porì ‘study; read’ Asm porh- + -i ‘study; read + NF’ Figure 8.13. Well-assimilated Indic loanwords in Lare Galo (Asm = Assamese). 174 MARK W. POST

Gloss Proto-Tani Lare Galo K-A Mising Assamese ‘cloud’ GRƾPѠW GRRPΩ̖ daor GD Z эU ‘tongue’ *arjo DUy diba zibև a µ¿QJHU¶ ODNNHƾ ODNFΩ̗ Ω DƾROL DƾXOL ‘bean’ *peeron peerén PэWэU PэWэU ‘mountain’ PURƾGL moodìi pahar pahar ‘cow’ *sa-??? KRΩ̖ gooruu goru µ¿HOG¶ UѠN UѠNΩ̗ SэWDU SэWhar ‘stick’ VѠƾ""" KѠѠGjD lakoti lakhuti ‘sky’ WDOΩƾ WDOΩ̗ Ω akak DNDƫ Figure 8.14. Assamese loanwords in Karbi-Anglong Mising. in the Siang River valley where the Tani languages to which Mising is genetically closest are also spoken (and which local migration legends also point to as the Mising place of origin) – it would appear that most words of Indic origin entered Mising following migrations of Mising SHRSOHWRWKH$VVDPÀRRGSODLQZKHUHVSHDNHUVRI,QGLFODQJXDJHVKDYH come to vastly outnumber speakers of non-Indic languages in recent centuries (Kakati 1995). In other words, the incorporation of many Indic IHDWXUHVLVDUHODWLYHO\UHFHQW0LVLQJVSHFL¿FGHYHORSPHQWDQGLVQRW something we can project into the general pre-history of Tani languages. Given how quickly and extensively Indic contact has affected the struc- WXUHRI0LVLQJLWLVLQGHHGVWULNLQJKRZOLWWOHHYLGHQFHRI,QGLFLQÀXHQFH lexical borrowing offers in the cases of most other Tani languages. Before we draw conclusions, however, let us look at a construction which I refer to as the Modal of Necessity/Obligation for Galo, some form of which is also found in many if not most other Tani languages. In this FRQVWUXFWLRQDPDLQLQÀHFWLQJYHUERIFRQVSLFXRXV,QGLFRULJLQWDNHV DVLWVFRPSOHPHQWDFODXVHVXERUGLQDWHGE\PHDQVRIDSUHGLFDWHVXI¿[ SΩ̖ , a form of seemingly indigenous origin which is probably traceable to the well-known PTB nominalizer *pa (LaPolla 2003). Comparing (5) and (6) from Galo and Assamese respectively, note the high structural similarity between the Galo construction and a parallel construction found in Assamese. Assuming that the Assamese construction is indeed native and not obtained through contact with Tani – as would seem to be the case – it looks very much as though the Tani construction has been calqued; that is, modelled on the Assamese one. LANGUAGE, CULTURE, ENVIRONMENT 175

(5) E࠴ը Ѡ ƾyP kâapΩODJt dù. E଎Ғܺ ƾyP NiDSԥҒ lagíGX s/he me see-COMPLEMENTIZER must-IMPERFECTIVE ‘She has to see me.’ Lare Galo (Tibeto-Burman, Tani)

(6) tai muk VDEэODJH tai muk sa-Eܧ lag-e she me see-COMPLEMENTIZER must-3.PRESENT ‘She has to see me.’ Standard Assamese (Eastern Indo-Aryan)

Normally, calquing of constructions – particularly when non-native lexical or morphological material is also used, as it is here – would usually be taken as evidence of strong population contacts and a resulting bilingualism (Aikhenvald 2007). The reason is that while borrowing of a common noun such as pajama or sushi is a relatively trivial matter, and can be accomplished without any real knowledge or understanding of the language and culture one is ultimately borrowing the word from, calquing of grammatical constructions requires comprehension of the structures of another language at an advanced level, intimacy with the cognitive “contents” underlying these structures, and an inclination to reproduce both by re-structuring the calquing language along these lines. So can we argue for strong, early Tani-Indic contacts on the basis RIVXFKGDWD":HOOQRWZLWKRXWTXDOL¿FDWLRQ As it happens, there is no attested native verb with the meaning ‘want’ in modern Tani languages. That is to say, there is no independent word which is dedicated to expressing an actor/experiencer’s desire or need, to possess a particular item for example, or to have a particular experience.15:KDWWKHUHLVLQVWHDGLVD'HVLGHUDWLYHVXI¿[ZKLFKZKLOHLW may derive historically from a verb meaning ‘want’, must in the modern languages be attached to a predicate root which expresses some sort of action. The upshot of this is that one cannot express the idea of wanting/ needing a thing without mentioning an activity that one wants or needs to do to it (7)-(8):

15,QVRPH7DQLODQJXDJHVLWLVSRVVLEOHWRXVHDUHÀH[RIWKH37YHUE PѠƾ ‘think’ with a sense very much like ‘want’, although this meaning has come about by way of a semantic extension and carries overtones of the original sense; namely, ‘think about something’ becomes ‘wish that one had something’. Accordingly, a better English translation than ‘want’ for this sense of PѠƾ may be ‘pine (for/after)’. 176 MARK W. POST

(7) ƾyvVWѠѠO࠴ը ѠG ƾy LVu W଎ғܺO଎ҒܺGX I water imbibe-WANT/NEED-IMPERFECTIVE ‘I want/need to drink water.’ (Lare Galo) (grammatically well- formed sentence)

(8) ƾyvVOѠѠG ƾy LVu O଎ҒܺGX I water WANT/NEED-IMPERFECTIVE µ, want/need water.’ (Lare Galo) (grammatically unacceptable sentence)

In terms of everyday village life, there is nothing particularly crippling about this situation. Unless, that is to say, one exists in a cultural situation in which accumulation of goods is regularly done for its own sake, with no immediate purpose or use, one does quite well to express desires in terms of events one wishes to bring about, with the objects mentioned only as participants. However, if one begins trading with non-intimates, this sort of situation becomes unmanageable. It is both unrealistic to expect an interlocutor to learn one’s entire inventory of verbs denoting the various actions one eventually might wish to do with items in an array of objects and unrealistic to learn such an inventory in another language oneself. Rather, it becomes necessary to simply and straightforwardly express the concept “I want/need x”. Accordingly, the Assamese verb lagZDVERUURZHGLQLWVQRQ¿QLWHIRUPLQi – the most “versatile” form in the sense that it carries no additional contextual baggage such as tense or agreement (Post 2008) – into a large number of Tani languages, or, more likely perhaps, into one or more of their ancestors (9).

(9) ƾyvVLJzODJtG ƾyLVu=go lagíGX I water=INDEFINITE want/need-IMPERFECTIVE ‘I want/need some water.’ (Lare Galo)

Following importation of a verb meaning ‘want/need’, development of a Modal of Necessity/Obligation via language-internal means would then have become straightforward. A desired/needed activity can be expressed in place of a desired/needed object by inserting an appropriately- structured clause (most often, a clause which is somehow nominalized) into the appropriate syntactic slot. Note that this is precisely the type LANGUAGE, CULTURE, ENVIRONMENT 177 of development which also occurred in English (and dozens of other well-known languages): I want/need some waterĺ,ZDQWQHHGto go ¿JXUH

Figure 8.15. Development of Galo modal of necessity via language-internal means following borrowing of an Indic lexical verb meaning ‘want/need’.

So what can we conclude from this? Well – tentatively of course – two WKLQJVWKH¿UVWLVWKDWWKH7DQL0RGDORI1HFHVVLW\2EOLJDWLRQPD\QRW have in fact been calqued from Assamese, despite initial appearances. Rather, it may well have arisen language-internally following assimilation of an Assamese lexical verb meaning ‘want (something)’. Although the “calquing” hypothesis would have pointed toward strong early Indic- Tani population contacts and bilingualism, the “borrowing plus internal innovation” hypothesis rather points toward the likelihood that early Indic-Tani contacts were in fact of the relatively looser nature of trade. This second hypothesis is more in line with the overall lack of well- assimilated Indic loanwords in Tani languages, which would be hard to explain if contact and bilingualism had indeed been high. So, if early contacts among Tani and Indic cultures cannot be clearly demonstrated on lexical and grammatical grounds, then what can explain WKHPDQLIHVWO\,QGRVSKHULFW\SRORJLFDOSUR¿OHRI7DQLODQJXDJHV" As Epps (2007:272) has shown, prosody (rhythm and intonation) is more readily diffusible across languages than is lexicon or grammar. In fact, prosody has the potential to diffuse over great geographic distances to link cultures who are in “contact” only tenuously, through multiple intermediaries. This is because prosodic “borrowing” requires only imi- tation – not comprehension. Spreading from language to language and area to area, prosodic diffusion can create vast areas of “prosodic convergence” in which mutually-unintelligible languages come to “sound alike” LQVLJQL¿FDQWUHVSHFWV,WZLOOEHP\FRQWHQWLRQKHUHWKDWSURVRGLFFRQYHU- JHQFHLVDOOWKDWLVUHTXLUHGWRH[SODLQWKH³,QGRVSKHULF´SUR¿OHRI7DQL languages, and, indeed, the existence of characteristically Indospheric 178 MARK W. POST

DQG6LQRVSKHULFSUR¿OHVLQJHQHUDO7KHFRQFOXVLRQWKDW,ZLOOGUDZIURP WKLV LV WKDW DQ ,QGRVSKHULF W\SRORJLFDO SUR¿OH FRXOG KDYH EHHQ ± DQG probably was – obtained by Tani languages largely in absence of direct FXOWXUDOFRQWDFWVZLWKVSHDNHUVRI,QGLFODQJXDJHVRIDQ\VLJQL¿FDQFH from earliest times up until only a handful of decades ago. 7KH DUJXPHQW ODUJHO\ UHVWV RQ WKH ¿QGLQJV RI 'RQHJDQ DQG 6WDPSH (1983; 2004; also discussed in Post in press-b), who argue, mainly with reference to Austro-Asiatic languages, that languages which develop an overall iambic rhythmic preference (a rising or weak-strong rhythm, as in English retain >UܼޖWHܼQ@ JHQHUDOO\ DOVR GHYHORS SUHSRVLWLRQV DQG SUH¿[HV UHODWLYHO\ LVRODWLQJ PRUSKRORJLFDO VWUXFWXUHV FRPSOH[ V\OODEOH structures including complex clusters, diphthongal vowels, and complex contour tones on syllables. Languages which develop an overall trochaic rhythmic preference (a falling or strong-weak rhythm, as in English ݦQࡦ @ JHQHUDOO\DOVRGHYHORSVXI¿[HVSRVWSRVLWLRQVV\QWKHWLFܭșUޖ

Figure 8.16. Rhythm and the opposite typological drifts of Munda and Mon- Khmer (adapted from Donegan and Stampe (1983:346); eighth notes basically indicate moras, a sixteenth note a brief, semi-syllablic and destressed onset; sometimes called a “sesquisyllable”).

Figure 8.17. Rhythm and typological drift in Tibeto-Burman (PTB reconstruction by Matisoff 2003). language speakers, and not obviously in contact with Sinitic speakers in post-Proto-Tani times (because the languages became progressively more “Indospheric”, not the reverse), is it possible to say anything about what kinds of cultural-linguistic contacts Tani speakers did have? Well, yes and no. It does seem likely that there were certain “neigh- ERXUV´ZKLFKPD\HLWKHUKDYHKDGFRQWDFWLQÀXHQFHVRQ7DQLODQJXDJHV or, perhaps even more likely, were either partially or fully incorporated into Tani-speaking populations as time went on. This is because we ¿QG VLJQL¿FDQW QXPEHUV RI IRUPV LQ 7DQL ODQJXDJHV PDQ\ RI ZKLFK are reconstructed by Sun for Proto-Tani, which have scarce or non- RFFXUULQJFRJQDWHVHOVHZKHUHLQ7LEHWR%XUPDQ ¿JXUH 7KHUHLV 180 MARK W. POST

DSDUWLFXODUO\KLJKFRQFHQWUDWLRQRIVXFKIRUPVLQ0LODQJ ¿JXUH (also see Post and Modi (in press) and Sun (1993:§3.4.2)).

PTB PT Galo Milang Gloss ? *ko kò- UX ‘request’ *gwap/krap *g(j)am gàm- ƾRW ‘bite’ OD\OΩ\UH\ UΩΩ UΩ̗ - ]DN ‘buy’ OLƾORZUDP UѠN U࠴̗NΩ DSX µ¿HOG¶ *grip/krep/rwak *taruk/tarup tarùk SDƾNHU ‘ant’ *rak *pVrok SRUyN DFX ‘fowl’ Figure 8.18. Selection of Tani forms with scarce or unattested PTB cognates (based on Sun 1993: Appendix 3; PTB reconstructions based on Matisoff 2003)

PTB PT Gloss PTB PT Gloss NU M DƾGDW WXU ‘alive’ *nja/sja GѠQ µPHDWÀHVK¶ *da/dzaan/la/ SXN ‘arrow’ *mruk/mjuk/ruk/ EHH ‘monkey’ tal woy/yuk *tak FDƾ ‘ascend’ *dan/dik/t(j) NRQ ‘one’ DNW M LNҌLW *bja/daw/dow/ WDƾ ‘bird’ *ka UMDS ‘door’ ZDƾDN Figure 8.19. Milang forms with scarce or non-occurring TB/Tani cognates (based on Sun 1993:§3.4.2).

If there were “others” in the Proto-Tani area, who were they? The simple fact is that we don’t know, primarily because extensive and reliable data for most area languages are not yet available. First of all, there DUHVHYHUDOXQFODVVL¿HG RUEDUHO\FODVVL¿HG TB languages of Arunachal Pradesh and, to a lesser extent, Tibet, such as Puroik (Sulung), Bugun, Sherdukpen, and others. These are currently spoken in relatively small enclaves often surrounded by and/or interspersed with speakers of Tani or other area languages. It is entirely possible, given the current distribution of these languages, that they may once have had a wider distribution but were later displaced by and/or partially incorporated into Tani-speaking (and/or other) populations. It may be that speakers of these languages – that is to say, their ancestors – formed one or more LANGUAGE, CULTURE, ENVIRONMENT 181 substrates of Proto-Tani and/or particular Tani languages after the Proto- Tani stage. However, so long as we continue to lack reliable data of any scale from the vast majority of Arunachali languages, there is little more on these points that can be said. Finally, we might think to look beyond TB when seeking sources of past Tani contacts. However, without any modern non-TB languages in the Tani area other than the recently-arriving Indo-Aryan languages, there are no clear candidates, nor have any VXJJHVWLRQV\HWEHHQDGGXFHGRIOLQNVIXUWKHUD¿HOG KRZHYHUVHH3RVW and Blench 2011). In sum, it is entirely possible that Proto-Tani speakers or their ancestors had contacts with certain neighbours which were FORVHHQRXJKWRLQÀXHQFHWKHGHYHORSPHQWRI7DQLODQJXDJHV+RZHYHU it is just as possible that we may never be able to say very much about precisely who these people may have been.

SUMMARY

In the preceding sections, I have reviewed a fairly diverse set of linguistic data which I have argued can be brought to bear on questions of the culture, origins, and environment of Proto-Tani speakers, and/or their near ancestors. The picture that emerges is as follows. There existed at least one pre-historical population who spoke a language of primarily Tibeto-Burman ancestry, from which the modern Tani languages primarily descend: “Proto-Tani”. This language, or a near ancestor, was probably spoken by a hill tribal population or populations not very different from most Tani of today, in an environment not very different from modern-day central Arunachal Pradesh. At some point in the history of this language, it appears likely that Proto-Tani – or possibly, several varieties thereof, at different places and/or times – may have been used as a lingua franca, potentially in the Subansiri and/or Siang River valleys. Subsequent to this period, there appear to have been several population splits, beginning with a single split (resulting in two ZHOOGH¿QHG³:HVWHUQ´DQG³(DVWHUQ´EUDQFKHVDJDLQSRVVLEO\UHSUH- senting Subansiri and Siang River valley population concentrations) and IROORZHGE\DWOHDVWIRXURU¿YHIXUWKHUVSOLWVDQGTXLWHOLNHO\PDQ\ more. Despite these splits and the emergence of a small number of relatively isolated cultural-linguistic “islands” (notably in the Apatani and Milang areas), intra-Tani population contacts remained constant in most cases, leading to a broad overall cultural-linguistic equilibrium which has endured more or less up to and including the present day. Although there appear to have been relatively widespread, close and 182 MARK W. POST constant intra-Tani contacts, evidence for extra-Tani contacts is slim at best in the case of well-known candidate languages (such as Indic languages) at any time at or after the Proto-Tani era, until relatively recent historical times. Although it does appear likely that extra-Tani contacts did occur – probably at different times and in different areas, given the different dis- tributions of non-universally-shared forms which have been incorporated into the Tani lexicon – it is not currently possible to say much about the identities of these populations other than that they were probably also, like the Tani, hill tribes inhabiting more or less the same geographical area.

LOOKING AHEAD

These admittedly sketchy remarks will, I hope, at least stimulate some WKLQNLQJUHJDUGLQJWKHW\SHVRIHYLGHQFHIURPRWKHU¿HOGVZKLFKPLJKW be brought to bear on the same broad questions, and which may either correlate well with and enable development of or contrast with and motivate revision of the hypothetical account outlined above. I am shamefully ignorant regarding the methods and accomplishments of my colleagues in cultural anthropology, archaeology, population genetics and so on, and far be it from me to try to direct anyone’s research; that said, some of the things it would be nice to see emerge over the coming years would include:

1) large compendia of well-transcribed, analysed and translated legends/folktales (which are of potential value both as approximate historical records (Blackburn 2003/2004) and as “vertically-transmitted”—and therefore potentially reconstructible —structures (Blackburn 2007)); 2) detailed comparative accounts of material culture, including evidence bearing on transmission of technologies (such as weaving) in pre-modern times; 3) comprehensive descriptions of area ritual traditions and other folk/ festival practices; 4) comprehensive ethnographies of individual cultures/ communities; ODUJHVFDOHLQYHVWLJDWLRQVE\SRSXODWLRQJHQHWLFLVWVRIVLJQL¿FDQW area retentions and mutations 6) good topographical maps of the area which may be legally published in India and abroad; PDSSHGDFFRXQWVRIWUDGHPLJUDWLRQURXWHVERWKH[LVWLQJFRQ¿UPHG and legendary/plausible. LANGUAGE, CULTURE, ENVIRONMENT 183

There’s a lot of work to be done, and a dwindling amount of time available in which to do it. So above all, here’s hoping it can become a cooperative enterprise!

Abbreviations:

3 Third person ACC Accusative ACNC Additive concessive ADVS Adverse C Consonant CMPL Completive F Feminine IRR Irrealis NF 1RQ¿QDO NZR Nominalizer PT Proto-Tani PTB Proto-Tibeto-Burman QUOT Quotative SFOC Sequential focus SUB Subject VVowel 184 MARK W. POST

BIBLIOGRAPHY

Aikhenvald, A.Y. 2007. “Grammars in contact: A cross-linguistic perspective.” In A.Y. Aikhenvald and R.M.W. Dixon (eds.), Grammars in contact: A cross- linguistic typology. Oxford: Oxford University Press, pp.1-66. Blackburn, S. 2003/2004. Memories of migration: Notes on legends and beads in Arunachal Pradesh, India. European Bulletin of Himalayan Research 25/26: 15-60. —— 2007. Oral stories and culture areas: from Northeast India to Southwest China. South Asia: Journal of South Asian Studies 30(3): 419-37. —— 2010. The Sun Rises. A Shaman’s Chant, Ritual Exchange and Fertility in the Apatani Valley. Leiden: Brill. Bradley, D., R.J. LaPolla, et al. (eds.) 2003. Language Variation: Papers on Variation and Change in the Sinosphere and in the Indosphere in Honour of James A. Matisoff&DQEHUUD3DFL¿F/LQJXLVWLFV Burling, R. 1965. Hill Farms and Padi Fields: Life in Mainland Southeast Asia. Englewood Cliffs, NJ: Prentice-Hall. —— 2007a. Language, ethnicity and identity in Northeastern India. South Asia: Journal of South Asian Studies 30(3): 391-404. —— 2007b. The lingua franca cycle: Implications for language shift, language change, DQGODQJXDJHFODVVL¿FDWLRQAnthropological Linguistics 49(3-4): 207-236. Campbell, L. 2001 [1998]. Historical Linguistics: An Introduction. Cambridge, MA: MIT Press. Dixon, R.M.W. 1997. The Rise and Fall of Languages. Cambridge: Cambridge University Press. 'ROH\6.DQG0:3RVW,QSUHVV&ODVVL¿HUVLQ0LVLQJ,Q*+\VORS60RUH\ and M. W. Post, (eds.). North East Indian Linguistics Vol. 4. New Delhi: Cambridge University Press. Donegan, P.J. and D. Stampe. 1983. Rhythm and the holistic organization of linguistic structure. Chicago Linguistic Society 19(2) [Papers from the Parasession on the Interplay of Phonology, Morphology and Syntax]: 337-53. —— 2004. “Rhythm and the synthetic drift of Munda.” In R. Singh (ed.), The Yearbook of South Asian Languages and Linguistics. Berlin: Mouton de Gruyter, pp.3-36. (SSV3³7XFDQRDQLQÀXHQFHRQ+XS´,Q$<$LNKHQYDOGDQG50:'L[RQ (eds.), Grammars in Contact: A Cross-Linguistic Typology. Oxford: Oxford University Press, pp.267-89. von Fürer-Haimendorf, C. 1980. A Himalayan Tribe: From Cattle to Cash. Berkeley: University of California Press. Geiyi, N. (ed.). N.d. Mirror of K.A.W.A. [Descendants of Karko Welfare Association]. Naharlagun, Arunachal Pradesh: Memkee Publishers. Genetti, C. 2007. A Grammar of Dolokha Newar. Berlin: Mouton de Gruyter. Heggarty, P. 2007. Linguistics for archaeologists: Principles, methods and the case of the Incas. Cambridge Archaeological Journal 17(3): 311-40. LANGUAGE, CULTURE, ENVIRONMENT 185

Jacquesson, F. 2008. “The speed of language change, typology and history: Languages, speakers and demography in North-East India.” In A. Sanchez- Mazas, R. Blench, M. Ross, I. Peiros and M. Lin (eds.), Past Human Migrations in East Asia: Matching Archaeology, Linguistics and Genetics. London: Routledge, pp.288-310. Johnson, G. 1996. New family tree is constructed for Indo-European. The New York Times January 2, B Section, Late Edition: 9. Kakati, B. 1995. $VVDPHVHLWVIRUPDWLRQDQGGHYHORSPHQW$VFLHQWL¿FWUHDWLVHRQ the history and philology of the Assamese language. Gauhati: Lawyer’s Book Stall. LaPolla, R.J. 2003. “Overview of Sino-Tibetan morphosyntax.” In G. Thurgood and R.J. LaPolla (eds.), The Sino-Tibetan Languages. London: Routledge, pp.22-42. Matisoff, J.A. 1990. On Megalocomparison. Language 66(1): 106-20. —— 1991a. “Areal and Universal Dimensions of Grammatization in Lahu.” In E.C. Traugott and B. Heine (eds.), Approaches to Grammaticalization, I: Focus on Theoretical and Methodological Issues. Amsterdam/Philadelphia: John Benjamins, pp.383-453. —— 1991b. Sino-Tibetan linguistics: Present state and future prospects. Annual Review of Anthropology 20: 469-504. —— 2003. Handbook of Proto-Tibeto-Burman: System and Philosophy of Sino- Tibetan Reconstruction. Berkeley: University of California Press. Modi, M. 2008. The Millangs. Itanagar/New Delhi: Himalayan Publishers. Modi, Y. 2005. Emergence and impact of Hindi as a Lingua Franca in Arunachal Pradesh. M.Phil thesis, Department of Linguistics, Jawaharlal Nehru University, New Delhi. Nyori, T. 1993. History and Culture of the Adis. New Delhi: Omsons Publications. Post, M.W. 2006. Compounding and the Structure of the Tani Lexicon. Linguistics of the Tibeto-Burman Area 29(1): 41-60. —— 2007. A Grammar of Galo. Ph.D. dissertation, Research Centre for Linguistic Typology, La Trobe University, Melbourne. —— 2008. “Verbs of posture, existence, location and possession and their grammaticalization pathways in the Tani languages.” In S. Morey and M.W. Post (eds.), North East Indian Linguistics. New Delhi: Cambridge University Press India, pp.127-52. —— 2008. “Grammaticalization and the discourse distribution of serial verbs in Assamese.” In W. Khanittanan and P. Sidwell, (eds.), SEALS XIV(2): papers from the 14th meeting of the Southeast Asian Linguistics Society (2004). &DQEHUUD3DFL¿F/LQJXLVWLFVSS —— 2008. Topographical deixis in Tani and beyond. Paper presented at the 3rd International Conference of the North East Indian Linguistics Society, Gauhati University, Guwahati, Assam, India. —— 2009. The phonology and grammar of Galo “words”: A case study in benign disunity. Studies in Language 34(4): 931-971. —— 2011. “Topographical deixis and the Tani languages of North-East India.” In G. Hyslop, S. Morey and M.W. Post (eds.), North East Indian Linguistics Volume III. New Delhi: Cambridge University Press India, pp.137-154. 186 MARK W. POST

Post, M. W. and R. Blench. 2011. Siangic: A new language phylum in North East India. Paper presented at the 6th International Conference of the North East Indian Linguistics Society, Tezpur University, Tezpur, Assam, India. Post, M.W. and Y. Modi (in press). Language contact and the genetic position of Milang in Tibeto-Burman. Anthropological Linguistics. —— in press-b. Prosody and typological drift in Austroasiatic and Tibeto-Burman: $JDLQVW³6LQRVSKHUH޵DQG³,QGRVSKHUH޵Mon-Khmer Studies. Ross, M., A. Pawley and M. Osmond. 2007 [2003]. The Lexicon of Proto-Oceanic: The Culture and Environment of Ancestral Oceanic Society. Volume 2: The Physical Environment. Canberra: Australian National University Press. Sun, T.J. 1993. A Historical-Comparative Study of the Tani Branch of Tibeto- Burman. Ph.D. dissertation, Department of Linguistics, University of California, Berkeley. Thomason, S.G. and T. Kaufman. 1988. Language Contact, Creolization and Genetic Linguistics. Berkeley: University of California Press. GLIMPSES OF THE ETHNOLINGUISTIC PREHISTORY OF NORTHEASTERN INDIA

GEORGE VAN DRIEM

OLD AND NEW LINGUISTIC PHYLA IN THE NORTHEAST OF THE SUBCONTINENT

The world’s two most populous families of languages meet in the Himalayas. These are Tibeto-Burman, which includes Cantonese and Mandarin, and Indo-European, to which languages such as English and Bengali belong. In addition to these two great linguistic phyla, Kra-Dai alias Daic, Austroasiatic and Dravidian language communities skirt the eastern Himalayan region. For example, the Austroasiatic language .KDVLLVVSRNHQLQWKH,QGLDQVWDWHRI0HJKƗOD\D$KRPDQRZH[WLQFW Kra-Dai tongue, was once a prominent cultural language in northeastern India, where scattered Daic language communities are still settled today. The Dravidian tongues Dhangar and Jhangar, which are basically dialects of Kurukh or Uraon, are spoken in Nepal’s eastern Terai. Kurukh is also spoken in scattered communities throughout northeastern India by people displaced by British colonial policies in India involving resettlement, not unlike the policy of transmigrasi in the former Dutch East Indies. Yet the Himalayas would appear to be peripheral to our understanding of the prehistory of Dravidian. The crux of the ethnolinguistic prehistory of the eastern Himalayan region are the language families Tibeto-Burman, Austroasiatic, Indo-European and, to a lesser extent, Kra-Dai. The advent of Indo-European and Kra-Dai languages to the northeastern portion of the Indian subcontinent is relatively recent and to some extent historically attested. Yet whilst the impact of Kra-Dai today is marginal, the intrusion of Indo-European into the region has been more robust and is represented by expansive languages such as Assamese, Hindi, Bengali and English. At the same time, historical, ethnographic, archaeological, anthropological and linguistic data all point to an Austroasiatic and Tibeto-Burman presence in northeastern India which must date back to some hoary period of ethnolinguistic prehistory. Austroasiatic and Tibeto-Burman, therefore, hold the key to understanding the population prehistory of northeastern India and the Indo-Burmese borderlands. The deeper phylogeny of the Indo-European language family was 188 GEORGE VAN DRIEM once conceived by August Schleicher as a branching oak tree, but the phylum has increasingly assumed a more rake-like appearance in more recent literature. The currently best-informed family tree structure of the Tibeto-Burman phylum is likewise essentially rake-like, a situation for ZKLFK,LQWURGXFHGWKHPHWDSKRURIIDOOHQOHDYHV ¿JXUH

Figure 9.1. The fallen leaves diagram for Tibeto-Burman. Some subgroups are well-establish ed, whilst others are less so. Brahmaputran may include Kachinic and Dhimalish. For the sake of argument, this diagram breaks up the WUDGLWLRQDOFDWFKDOOµ4LƗQJLF¶LQWRWKHċUVǌFOXVWHUDQGDWUXQFDWHGµ4LƗQJLF¶ not to posit a robust phylogenetic hypothesis but to challenge, thus emphasising that crucial work in this area has been left undone. The precise phylo- JHQHWLF UHODWLRQVKLSV EHWZHHQWKH GLYHUVH U*\DOURQJ ODQJXDJHV ċUJǀQJ 4LƗQJ0LxDJ 0\ӽ 7DQJXWċUVǌ/ԉVǌ7RVX 'Xǀ[ 1jP\u6Kӿ[ƯQJ *XuTLyQJ&KR\R 4Xq\ =KƗEjDQG3ULQPL 3ԃPӿ KDYH\HWWREHGHPRQ- VWUDWHG:KDWLVVRPHWLPHVFDOOHGµ1RUWKHUQ4LƗQJLF¶LVVXSSRVHGWRLQFOXGH WKHU*\DOURQJLFJURXSUHFRJQLVHGE\-DFNVRQ6XQ 6ǌQ7LƗQ[ƯQ DQG+XiQJ %IiQ+RSHIXOO\VFKRODUVZRUNLQJRQWKH7LEHWR%XUPDQODQJXDJHVRI6uFKXƗQ and Yúnnán provinces will in the coming years shed light on the structure of this portion of the Tibeto-Burman family tree (van Driem 2001, 2006). GLIMPSES OF THE ETHNOLINGUISTIC 189 Each diamond represents not a language, but major subgroup. Figure 9.2. Geographical distribution of the major branches of Tibeto-Burman. Tibeto-Burman. Figure 9.2. Geographical distribution of the major branches 190 GEORGE VAN DRIEM

The geographical distribution of the major Tibeto-Burman subgroups saliently shows the densest concentration in the northeastern portion of the Indian subcontinent, with most major subgroups being exclusive WR WKLV DUHD ¿JXUH 7KH PRVW DXWKRULWDWLYH SK\ORJHQHWLF WUHH IRU $XVWURDVLDWLFDWSUHVHQWLVWKHPRGHOGHYHORSHGE\*pUDUG'LIÀRWK ¿J- ure 9.3).

)LJXUH $XVWURDVLDWLF ZLWK *pUDUG 'LIÀRWK¶V WHQWDWLYH FDOLEUDWLRQ RI WLPH GHSWKVIRUWKHYDULRXVEUDQFKHVRIWKHODQJXDJHIDPLO\ PRGL¿HGIURP'LIÀRWK 2001, 2005). The precise phylo genetic propinquity of Pearic, after Khmeric loan OD\HUVKDYHEHHQVWULSSHGRIIUHPDLQVXQFHUWDLQH[FHSWWKDW'LIÀRWKREVHUYHV that Pearic is Mon-Khmer and not ‘une espèce de vieux khmèr’, as some scholars once maintained. This diagram arranges in a tree-shaped phylogeny the fourteen recognised branches of Austroasiatic, i.e. North Munda, South Munda, Khasian, Paka nic, Palaungic, Khmuic, Vietic, Katuic, Bahnaric, Khmeric, Pearic, Monic, Aslian and Nico bar ese. GLIMPSES OF THE ETHNOLINGUISTIC 191

Some novel insights into the phylogeny of Kra-Dai have been put forward by Edmondson and Solnit (1988, 1997) and Ostapirat (2005). The location of the Austroasiatic ancestral homeland can be argued from a purely linguistic point of view principally on the basis of linguistic palaeontology and the geographical centre of gravity of the family. The latter is ascertained from the distribution of modern Austroasiatic language communities and the deep phylogenetic divisions in the family. The distribution of the modern language communities and the geography of the deepest divisions in linguistic phylogeny would put the geographical centre of the family somewhere between South Asia and Southeast Asia, in the area around the northern coast of the Bay of Bengal. Whether we assume that the deepest division in the family lies between Munda and the rest, as an older generation of scholars used WRVXVSHFWRUDVVXPHWKHYHUDFLW\RI'LIÀRWK¶VQHZWULSDUWLWHGLYLVLRQ then the geography of deep historical divisions in linguistic phylogeny would compel us to look for a homeland on either side of the Ganges and Brahmaputra delta, although we would be unable to say whether this homeland would have to have lain to the east or to the west of the delta.1 When linguists look beyond what linguistic phylogeny can tell them, they must ask which archaeological transition or modern genetic gradient FDQ EH UHODWHG ZLWK FRQ¿GHQFH WR DQ DQFLHQW OLQJXLVWLF LQWUXVLRQ RU WR the prehistorical spread of a language family. When linguists resort to OLQJXLVWLFSDODHRQWRORJ\WKH\PXVWKDYHUHFRXUVHWRWKH¿QGLQJVRISDO- aeoclimatologists and, more particularly, palaeobotanists. Most conceiv- able theories about the homelands of Tibeto-Burman and Austroasiatic have already been put forward. The idea of a Tibeto-Burman homeland VLWXDWHGLQRUQHDUE\SUHVHQWGD\6uFKXƗQKDVEHHQHQWHUWDLQHGVLQFHWKH ninetheenth century, especially by British scholars in India. Sinocentrists favour a northern Tibeto-Burman homeland in the lower Yellow River basin on the plains of northern China, whereas some have proposed a provenance within the Himalayan region itself. Scholars have sought to situate the Austroasiatic Urheimat as far west as the Indus valley and as far east as the Yangtze delta or insular Southeast Asia. However, the main contenders today for the Austroasiatic homeland are the Indian subcontinent, mainland Southeast Asia and the middle Yangtze.

1 The presence of many speakers of Munda languages in northeastern India is a legacy of resettlement to Assam orchestrated by the East India Company, an economic policy mentioned above in connexion with speakers of Northern Dravidian languages. 192 GEORGE VAN DRIEM

ARCHAEOLOGY AND LINGUISTIC PALAEONTOLOGY

The fundamental epistemological question that will continue to haunt us is whether the spread of a recognisable Neolithic and Bronze Age DVVHPEODJH FDQ DFWXDOO\ HYHU EH WDNHQ ZLWK FHUWDLQW\ WR UHÀHFW WKH VSUHDGRIDODQJXDJHDQGVRRIDODQJXDJHIDPLO\$UFKDHRORJ\UHÀHFWV what we have been able to glean about the material culture of past communities. In fact, how often can we be certain which language was spoken by ancient stone knappers or by the potters behind a particular ceramic culture in some archaeologically attested pre-literate society? Indeed, we must ask whether the modern geographical distribution of the Tibeto-Burman language family correlates with the mute testimony of any single portion of the archaeological record that happens to have been preserved, discovered and studied by archaeologists. A more general issue is time depth. Archaeological transitions are reconstructed at very different times in the past, e.g., the palaeontologically attested spread of anatomically modern humans, the spread of agriculture, DQG WKH VRPHWLPHV ZHOOGH¿QHG SDWWHUQV RI GLVSHUVDO RI LGHQWL¿DEOH cultural assemblages in the Neolithic and Bronze Age. At the same time, many known historical transitions and conquests with linguistic consequences have left little or no clear-cut traces whatsoever in the archaeological record. Therefore, we are free to speculate, for example, whether the expansion of early Bodic language communities into the Himalayas was DVVRFLDWHG ZLWK WKH VXGGHQ DSSHDUDQFH RI H[SRQHQWV RI WKH 0ӽMLƗ\iR Neolithic in eastern Tibet at mKhar-ro and in Kashmir at Burzahom at WKHVDPHWLPHWKDWWKHFRUHDUHDRIWKH0ӽMLƗ\iRFXOWXUDODVVHPEODJHLQ *ƗQVVKUDQNGXULQJDSHULRGRIFOLPDWHFKDQJHEHWZHHQWKH0ӽMLƗ\iR SKDVH %& DQGWKH%jQVKƗQSKDVH %& RIWKH 0ӽMLƗ\iRVHTXHQFH7KLVDWOHDVWLVDVFHQDULRZKLFK,DUJXHGLQVHYHUDO earlier publications (van Driem 1998, 2001, 2002). Recent palaeoecological evidence indicates that the vast and once heavily forested Tibetan Plateau underwent large-scale deforestation precisely during this period of projected Bodic expansion. The palaeobotanical evidence indicates that this deforestation occurred at the hands of human settlers (Kaiser et al. 2006, Miehe et al. 2006, Wu et al. 2006, Kaiser et al. 2007). :HUHWKHVHSHRSOHSHUKDSV%RGLFFRORQLVWVIURPWKH0ӽMLƗ\iR1HROLWKLF core area who introduced to the Tibetan Plateau a new lifestyle with GHOHWHULRXVHFRORJLFDOUDPL¿FDWLRQV"5HFHQWO\,KDYHSUHVHQWHGVHYHUDO alternative scenarios which differently relate the traceable patterns of GLIMPSES OF THE ETHNOLINGUISTIC 193 dispersal of cultural assemblages in the archaeological record with the present geographical distribution of Tibeto-Burman language communities (van Driem 2006). Linguistic palaeontology, a term introduced by Adolphe Pictet in 1859, is an attempt to understand the ancient material culture of a language family on the basis of the lexical items which can be reliably reconstructed for the common ancestral language. Linguistic palaeontology, like any attempt to give an ethnolinguistic interpretation to the archaeological record, invariably raises complex issues. Elsewhere, I have discussed at some length the arguments relating to what we can glean about ancient Tibeto-Burman culture and the role played by broomcorn millet Panicum miliaceum and by foxtail millet Setaria italicaWKHODWWHUUHÀHFWHGLQODQJXDJHVDVIDUÀXQJDV2OG&KLQHVHで btsïk in the Yellow River basin and Lhokpu2 AáɹIRM ‘foxtail millet’ in modern southwestern Bhutan (van Driem 2006). At the present time, the earliest archaeologically attested domestic millet dates from before 6000 BC at㠸䲶Ⓧ;ƯQJOǀQJJǀXQHDU䎔ጠ&KuIƝQJZKHUHD1HROLWKLF FXOWXUHZLWKRXWVLFNOHVRQFHÀRXULVKHG =KjR /LQJXLVWLFSDODHRQWRORJ\VWURQJO\TXDOL¿HVWKHDQFLHQW$XVWURDVLDWLFV DVWKHPRVWOLNHO\FDQGLGDWHVIRUWKH¿UVWFXOWLYDWRUVRIULFH0RUHRYHU 'LIÀRWKKDVVKRZQWKDWWKHUHFRQVWUXFWLEOH$XVWURDVLDWLFOH[LFRQSDLQWV WKHSLFWXUHRIDIDXQDÀRUDDQGHFRORJ\RIDWURSLFDOKXPLGKRPHODQG environment. He adduces three salient isoglosses diagnostic for the faunal ecology of the Proto-Austroasiatic homeland that are reconstruct- LEOHDOOWKHZD\EDFNWRFRPPRQ$XVWURDVLDWLFDQGDUHUHÀHFWHGLQDOO branches of the family, i.e., KP?ʈI ‘peacock Pavo muticus’, RƲPISƲR ‘tree monitor lizard Varanus nebulosus or bengalensis’ and RƲLWSʈɐ ‘binturong’ or the ‘bear cat Arctitis binturong’, a black tropical mammal WKDW LV WKH ODUJHVW RI WKH FLYHW FDWV 'LIÀRWK 1RQH RI WKHVH species are native to areas that currently lie within China, and, to our present knowledge, none were ever native to the area that is today China. Such linguistic palaeontological evidence therefore appears to render the middle Yangtze homeland hypothesis less likely. More reconstructible Proto-Austroasiatic roots indicative of a tropical RUVXEWURSLFDOFOLPDWHDUHDGGXFHGE\'LIÀRWK LH@ƲLHMʈJ\

2 The Lhokpu are an inbred and genetically highly distinct group within the Himalayan region as a whole; Kraaijenbrink et al. 2006a, Parkin et al. 2006a. The impact of matrilocality and cross-cousin endogamy is clearly discernible in the genetic signature of this language community. Many of the ancient Tibeto-Burman groups may have been matrilineal, matrilocal societies with uxorilocal marriage such as the modern Lhokpu and Gongduk of Bhutan. 194 GEORGE VAN DRIEM

HƲPKMʈJ ‘ant eater, Manis javanica’, BƲI?L ‘bamboo rat, Rhizomys sumatrensis’ (an Austroasiatic root which has found its way into Malay as a loan), I?AG?ĩ ‘the Asian elephant, Elephas maximus’, IG?ȑ ‘mountain goat, Capricornis sumatrensis’, PƲK?ʈQ ‘rhinoceros, Dicerorhinus sumatrensis’ and RƲLPG?I ‘buffalo, Bubalus bubalus’. $GGLWLRQDOO\'LIÀRWK SRLQWVRXWDIDFWORQJQRWHGE\VFKRO- ars of Austroasiatic linguistics, e.g., Osada (1995), namely that a rich repertoire of reconstructible roots representing ancient rice agriculture LVUREXVWO\UHÀHFWHGLQDOOEUDQFKHVRI$XVWURDVLDWLFYL]IƲȏ?ʈɐ ‘rice plant’, PƲĩIMʈɐ ‘rice grain’, AƲĩI?ʈK ‘rice outer husk’, IƲLBƲI ‘rice inner husk’, NFCʈɐ ‘rice bran’, RƲKN?J ‘mortar’, HƲLPCɐ ‘pestle’, HƲKNGƲP ‘winnowing tray’, ESʈK ‘to winnow’, HƲPKSƲJ ‘dibbling stick’ and IƲLRSʈɐ ‘rice complement’, i.e., accompanying cooked food other than rice. Nicole Revel (1988) contributed one of the most elaborate ethnobo- tanical studies on rice, rice cultivation practices and rice terminology in various Asian language communities. Her work points towards where we should look when considering the origins of domesticated rice. The other main candidate for early cultivators of rice are the ancestral Hmong-Mien. Great strides have been made in our understanding of Hmong-Mien historical phonology (Haudricourt 1954, Purnell 1970, Wáng and Máo 1995, Niederer 1998), although the reconstructible lexi- FRQVSHFL¿FWRULFHFXOWLYDWLRQLVOHVVLPSUHVVLYHWKDQWKH$XVWURDVLDWLF repertoire. The three Hmong-Mien etyma relating to rice cultivation that appear to be original to the linguistic phylum are LRQȕʈG ‘husked rice, Ȯ?ʈĩ ‘cooked rice’ and L˜HCĩ ‘rice head, head of grain’. On the other KDQGWKH+PRQJ0LHQWHUPVIRUJOXWLQRXV ULFH SDGG\ ¿HOGVLFNOH rice cake and (rice) seedling ‘are likely to have had a Chinese origin’ (Ratliff 2004:158-59).

A TALE OF RICE

The rice story is complex, and the plot of the story has changed more than once in recent decades. Whereas the origin of rice cultivation was once held ‘incontestably’ to have lain in the Indian subcontinent (Haudricourt and Hédin 1987:159-61, 176), subsequent scholarship moved the homeland of rice agriculture from the Ganges to the Yangtze. For years conventional wisdom in archaeological circles dictated that rice was domesticated in the Middle Yangtze, perhaps as early as the sixth millennium BC. More recently, scholars have increasingly begun GLIMPSES OF THE ETHNOLINGUISTIC 195

WRWDNHQRWHRI¿QGLQJVWKDWZRXOGPRYHWKHRULJLQDOKRPHODQGRIULFH cultivation back to the Indian subcontinent. Against the background of older datings of domesticated rice and ceramic culture from Gangetic basin and Doab sites such as Koldihawa and Mahagarha, reportedly dating from the seventh millennium BC (Sharma et al. 1980, Pal 1990, Agrawal 2002), there are now newer sites with more reliable dates at /DKXUDGHZD /DKXUƗGHYƗ ৫RNXYƗDQG6DUƗƯ1DKDU5ƗƯ At the Lahuradewa site (26°46’ N, 82°57’ E), the early farming phase, corresponding to period 1A in the site’s clear-cut stratigraphy, has radiocarbon dates ranging from ca. 5300 to 4300 BC. Carbonised material IURPSHULRG$ZDVFROOHFWHGE\WKHÀRWDWLRQPHWKRG\LHOGLQJSetaria glauca and 2U\]DUX¿SRJRQ as well as a morphologically distinct, fully domesticated form of rice ‘comparable to cultivated Oryza sativa’ (Tewari et al. 2002). More recently, accelerator mass spectroscopy dates were obtained on the rice grains themselves, corroborating the antiquity of rice agriculture at the site. Most recently, new radiocarbon dates for rice agriculture have been FRPLQJ IURP WKH *DQJHV EDVLQ ZLWK WKH ৫RNXYƗ VLWH QHDU$OODKDEDG now yielding dates comparable in antiquity to those of the Lahuradewa VLWH 9DVDQW6KLQGH>9DVDQWĝLYDUƗPĝLQGH@SHUVRQDOFRPPXQLFDWLRQ November 2007), and exciting new dates for ancient rice agriculture are DOVRHPHUJLQJIURP6DUƗƯ1DKDU5ƗƯ 0DQMLO+D]DULNDSHUVRQDOFRP- munication 7 March 2008). Of course, we are living at a time when a more reliable calibration of radiocarbon dates in general has become a matter of great urgency. )XUWKHUHDVWDWOHDVW¿YHVSHFLHVRIZLOGULFHDUHQDWLYHWRQRUWKHDVWHUQ India, viz. 2U\]DQLYDUD2U\]DRI¿FLDQDOLV 2ODWLIROLD 2U\]DSHUHQ- nis (O. longistaminata), Oryza meyeriana (O. granulata) and Oryza UX¿SRJRQ, and reportedly over a thousand varieties of domesticated rice are currently in use in the region (Hazarika 2005, 2006a). The different varieties of rice in northeastern India are cultivated in three periods by distinct cultivation processes. In the process of ƗKX NKHWL, the rice is sown in the months of 3KƗJXQ and Sot, i.e., mid February to early April. 7KHVHHGOLQJVDUHQRWWUDQVSODQWHGEXWULSHQLQMXVWIRXUPRQWKVLQ¿HOGV that must be constantly weeded. In EƗXNKHWL, the rice seedlings are sown IURPPLG0DUFKWRPLG$SULOLQSORXJKHGZHW¿HOGVDQGOLNHZLVHGRQRW need to be transplanted. In ĞƗOLNKHWL, the rice is sown from mid May to mid June, and the seedlings are transplanted. ĝƗOLNKHWL rice varieties are suspected to derive from the wild RI¿FLDQDOLV rice still widely found in swampy village areas. The wild UX¿SRJRQ rice cannot be used for human 196 GEORGE VAN DRIEM consumption because the plants shed their seeds before they ripen, so it is used in Assam and other parts of northeastern India as cattle feed (Hazarika 2006b). Whilst claims have been published dating the earliest rice cultivation in East Asia to as long ago as 10,000 BC, the currently available evidence indicates that immature morphologically wild rice may have been used by foragers before actual domestication of the crop, e.g., at the ޛॱ᫻ %ƗVKtGjQJ VLWH %& EHORQJLQJ WR WKH ᖝ九ኡ 3pQJWyXVKƗQFXOWXUHLQWKH0LGGOH

Twelve wild forest-margin rice species are known and found mostly in 6RXWKHDVW$VLDDVZHOODVDWROGVLWHVRIKXPDQKDELWDWLRQHJ-LӽK~ RQWKH+XiL5LYHULQ+pQiQLQWKHVHYHQWKPLOOHQQLXP%&RU+pPԃG LQWKH

POPULATION GENETICS AND TWO OLD LINGUISTIC PHYLA

Very often language seems to be less ambiguously correlated with the geographical distribution of genetic markers in the populations speaking the languages in question than either genes or languages can be 198 GEORGE VAN DRIEM correlated and contrasted with the fragments of material culture that happen to have resurfaced unscathed from the sands of time. So, can genes and languages generally be correlated and contrasted with each other in a more meaningful way? On the one hand, the linguistic ancestors of a language community were not necessarily the same people as the biological ancestors of that community. On the other hand, the genetic picture often shows sexual dimorphism in linguistic prehistory. Some languages appear to be mother tongues, whereas others show up as father tongues. In Baltistan, in what today is northern Pakistan, the phonologically highly conservative local Tibetan dialects appear to correlate with the SUHGRPLQDQWO\7LEHWR%XUPDQPLWRFKRQGULDO'1$ZKLFKUHÀHFWVWKH Balti community’s maternal ancestry (Poloni et al. 1997, 2000; Zerjal et al. 1997; Quintana-Murci et al. 2001; Qamar et al. 2002), whilst the intrusive paternal Y haplogroups from the Near East appear to correlate ZLWKWKHIRUFLEOHFRQYHUVLRQRIWKHDUHDWR,VODPLQWKH¿IWHHQWKFHQWXU\ By contrast, the ‘Father Tongue hypothesis’ may apply to the spread of Indo-Aryan into the Indian subcontinent and, further east, to the spread of Sinitic during the Hàn demic expansion. In fact, a likely correlation EHWZHHQ WKH OLQJXLVWLF DI¿QLW\ DQG WKH< KDSORJURXS RI D SRSXODWLRQ appears to be a more widespread phenomenon. At many times and in many places in prehistory, the father tongue may have been the guiding mechanism in language shift. The dynamics of a process whereby mothers passed on the language of their spouses to their offspring also has major implications for our understanding of language change. If the language shift giving rise to the Sinitic languages and perhaps also the eastward spread of Indo-Aryan speech across northern India took place in this way, then such languages may have begun as languages belonging to another phylum until they reached the stage currently attained by Michif. In origin at least, Michif is genetically an $OTRQTXLDQODQJXDJHWKDWZDVVSRNHQE\ZRPHQZKRUHOH[L¿HGWKHODQ- guage with the French spoken by their husbands to such an extent that the genetic affnity has nearly been obscured (Bakker 1992, 1994; van 'ULHP ,IWKHSURFHVVRIUHOH[L¿FDWLRQZHUHWRFRQWLQXH beyond the stage attained by Michif, then a language could conceiv- DEO\FKDQJHLWVJHQHWLFDI¿QLW\HYHQWKRXJKWKHG\QDPLFVRIWKHSURFHVV would introduce a discontinuity with its past. If such a process took place, could it ever be reconstructed linguistically? $GGLWLRQDOO\ WKRXJK WKHVH ¿QGLQJV KDYH QRW \HW UHDFKHG D ZLGHU public, geneticists have increasingly been identifying single nucleotide GLIMPSES OF THE ETHNOLINGUISTIC 199 polymorphisms on the autosomes that are diagnostic for geographical DQGUDFLDODI¿QLW\DQGSUREDEO\UHÀHFWVDOLHQWHSLVRGHVRIRXUFROOHF- tive population prehistory. When we turn now to Tibeto-Burman and Austroasiatic, however, to the present state of our knowledge the pre- dominant Y haplogroups in the populations tested to date correlate best ZLWKWKHOLQJXLVWLFDI¿QLW\RIWKRVHFRPPXQLWLHV,QRWKHUZRUGVERWK Austroasiatic and Tibeto-Burman may, on the grander scale over time, have preponderantly, albeit not exclusively, spread by the Father Tongue mechanism. By this interpretation, the shared Y haplogroup is assumed WRUHÀHFWWKHIRXQGLQJGLVSHUVDORIWKHODQJXDJHIDPLO\ A few of the pioneering genetic assays of Tibeto-Burman populations inside China ventured some plausible claims. The reduced polymor- phism of northern populations of East Asia, which represent a subset RI WKH KDSORW\SHV IRXQG LQ VRXWKHUQ SRSXODWLRQV ZDV WDNHQ WR UHÀHFW WKHSHRSOLQJRIWKHQRUWKDIWHUWKH,FH$JH 6HWDO %\FRQWUDVW the high frequency of H8, a haplotype derived from M122C, was seen DVUHÀHFWLQJDJHQHWLFERWWOHQHFNHIIHFWWKDWRFFXUUHGGXULQJDQDQFLHQW southwesterly migration about 10,000 years ago, suggesting a demic dif- IXVLRQDWWKHRXWVHWRIWKH1HROLWKLF 6HWDO'LQJHWDO6KL et al. 2005). Another study suggested that Hàn Chinese did not originate in the Yellow River basin but had more recently migrated to this area IURPVRXWKZHVWHUQ&KLQD &KԃHWDO Comparison of various haplogroup frequencies exhibited by Tibetans YV7ԃMLƗ%iLDQG/ROR%XUPHVHJURXSVVKRZHGDOOWKHVH7LEHWR%XUPDQ groups to have a high frequency of the Y-chromosomal haplogroups O3e DQG2 ZLWKWKHDYHUDJHKRYHULQJDSSUR[LPDWHO\DURXQG7KHVH ¿QGLQJVZHUHLQWHUSUHWHGDVVXSSRUWLQJDPDOHELDVHGLQ¿OWUDWLRQIURP the Bodish area in Amdo into Yúnnán and Húnán about two and a half millennia ago. However, ‘the less drastic bias between male and female lineages’ suggested that these putative southward migrations ‘likely occurred with the involvement of both sexes rather than as conquests involving expedition forces primarily consisting of male soldiers’ (Wen et al. 2004). Interestingly, genetic studies often appear largely to corroborate the long-standing intuitions of linguists, historians and ethnographers working in the region and even to support their previously published models of ethnolinguistic prehistory. Yet these Chinese studies are limited by the fact that most Tibeto-Burman language communities and even most branches of the language family are exclusively represented outside of China. The picture of the Tibeto-Burman past has been rendered far 200 GEORGE VAN DRIEM

Figure 9.4. The portion of the Y chromosome phylogenetic tree relevant to the Father Tongue hypothesis, with regard to the Austroasiatic and Tibeto-Burman linguistic phyla, provided by Mark Jobling and Emma Parkin. PRUHFRPSOHWHE\¿QGLQJVRIRXURZQUHVHDUFKWHDPZKLFKKDVFRQ- ducted the most extensive sampling of Tibeto-Burman populations in the Himalayan region (Kraaijenbrink et al. 2006a, 2006b; Parkin et al. DE 7KHVH¿QGLQJVKDYHDOORZHGXVWRPDNHQRYHOLQIHUHQFHV about the population prehistory of Tibeto-Burman, and has also uncov- HUHGVRPHZKROO\XQH[SHFWHG¿QGVVXFKDVWKHJHQHWLFDI¿QLWLHVRIWKH %ODFN 0RXQWDLQ 0|QSD$GGLWLRQDO ¿QGLQJV IURP QRUWKHDVWHUQ ,QGLD Tibet and Burma will enable us to identify the possible molecular correlates corresponding to more episodes in the spread in Asia of ancient Tibeto-Burman language communities. The Y-chromosomal haplogroup O3e (M134), which seems to tag Tibeto-Burman language communities within and outside of the Himalayan region, may very well have a time depth of at least 10,000 \HDUV ¿JXUH ,IVRWKLVZRXOGSXWXVLQDWLPHIUDPHWKDWFRP- pels us to consider in which localities the ancient Tibeto-Burmans may KDYHGZHOWGXULQJWKHODVWJODFLDOPD[LPXP&RXOGWKHVRXWKHUQÀDQN of the Himalayas have served as a vast refuge area during the last Ice Age, or did the early Tibeto-Burmans at this time dwell in some area to the northeast or to the east of the Himalayas? Palaeoclimatologists remain divided into rival camps on the question of the extent of the glaciation in the Himalayas in recent prehistory, e.g., Kuhle (1985, 1986, 1987, 1988a, 1988b, 1990a, 1990b, 1991, 1995, 1997, 1999, 2005), Thompson et al. (1989), Lehmkuhl (1995), Schäfer et al. (2002), Owen et al. (2002), Owen et al. (2003), Jin et al. (2005), Lehmkuhl and Owen (2005), Vandenberghe (2007). By virtue of the sheer scale and diversity of the topography, the Himalayas harbour a panoply of GLIMPSES OF THE ETHNOLINGUISTIC 201

FOLPDWRORJLFDO HQFODYHV DQG VKHOWHUHG DUHDV ZLWK WKHLU RZQ VSHFL¿F microclimate. Did the Himalayas offer hospitable ice-age refugia to the ancestral Tibeto-Burmans? In seeking an answer to this question, northeastern India still remains scantily documented and poorly understood in archaeological terms. Much of the major work in this region was carried out over a generation ago and has been discussed and referenced in my handbook (van Driem 2001). Yet most such work pertains to the Neolithic, a later period which does not help shed light on the issue of possible ice-age refuge habitats. By contrast, a partial answer emerges from the far more complete picture that we have for Nepal thanks to the posthumously published work RI*XGUXQ&RUYLQXVZKRWUDJLFDOO\ZDVPXUGHUHGLQKHURZQÀDWLQ 3X৆HE\WKHKHQFKPDQRIDORFDOUHDOHVWDWHFULPLQDO,URQLFDOO\VKHKDG FKRVHQWROLYHLQ3X৆HWRDYRLGWKHPD\KHPH[WRUWLRQDQGPXUGHUWKDW 0DRLVWVZHUHLQÀLFWLQJLQGLVFULPLQDWHO\RQWKHLUIHOORZFRXQWU\PHQLQ the regions where she conducted her work. On the basis of a lifetime of palaeontological and archaeological research in the sub-Himalayan tracts of Nepal, Gudrun Corvinus (2007) developed the theory that the Early and Middle Palaeolithic and micro- OLWKLFWUDGLWLRQVLQWKHZHVWHUQ7HUDLDQGĝLYƗOLNVGHULYHIURPFRQWHPSR- raneous cultures further south on the Indian subcontinent in a period still FKDUDFWHULVHGE\ORZSRSXODWLRQGHQVLW\+RZHYHUWKH3Ɨ৬XFXOWXUHLQ WKHHDVWHUQĝLYƗOLNVDQGWKH%UƗNKX৬ƯFXOWXUHLQWKHZHVWHUQ7HUDLVKRZD ODWHULQÀXHQFHHPDQDWLQJZHVWZDUGDFURVVWKH+LPDOD\DQIRRWKLOOVIURP Southeast Asia in the Late Pleistocene and Holocene, with the archaeological record suggesting an increase in population density at the end of WKHODVWJODFLDOPD[LPXPURXJKO\FRLQFLGHQWZLWKWKHÀRXULVKLQJRIWKH %UƗNKX৬ƯFXOWXUH Corvinus’ comprehensive pioneering work in Nepal has yielded good stratigraphies and optical and infra-red stimulated luminescence datings RI NH\ VHGLPHQW OD\HUV EDVHG RQ ZRUN DW VLWHV DORQJ WKH 5ƗWR .KROƗ VRXWKRI6LQGKXOƯLQ0DKRWWDUƯGLVWULFWLQWKHHDVWHUQĝLYƗOLNVVLWHVDWDQG QHDU6ƗWSDWƯKLOOHDVWRI%KDLUDKDYƗLQ/XPELQƯGLVWULFWDQGYDULRXVVLWHV LQWKH'ƗৄJDQG'HXNKXUƯEDVLQ7KHDUFKDHRORJLFDOUHFRUGVXJJHVWVD long prehistory of human habitation. For example, the alluvial and col- OXYLDOKLOOZDVKGHSRVLWVFRPSRVLQJWKH*LGKLQL\ƗDQG%DEDLIRUPDWLRQV LQWKH7XƯEDVLQLQ'ƗৄJGLVWULFWKDYH\LHOGHGDEXQGDQWOLWKLFPDWHULDO from the Early, Middle and Late Palaeolithic periods as well as from the Mesolithic and Neolithic periods. An Early Palaelothic presence of the Acheulian tradition in South Asia in the early Middle Pleistocene 202 GEORGE VAN DRIEM

LV LQGLFDWHG E\$FKHXOLDQ ELIDFLDOV DQG ÀDNH WRROV LQGXVWU\ DW*DGDUƯ LQ'ƗৄJLQWKHZHVWHUQ7HUDLDQGDW6ƗWSDWƯLQFHQWUDO1HSDOMXVWZHVW RIWKH1ƗUƗ\D৆Ư5LYHU$OOXYLDOWHUUDFHGHSRVLWVDWWKH$UMXQVLWHLQWKH badlands on the left bank of the Arjun River show Levallois technology appearing in the Middle Palaeolithic. At the same time the foothills of the central Himalayas show a complex cultural prehistory. Nepal straddles the so-called Movius line and represents a transition zone between two traditions of lithic H[SUHVVLRQ6LWHVLQWKHZHVWHUQ1HSDOHVH7HUDLVKRZDI¿QLW\ZLWKRWKHU more Occidental technocomplexes as represented by artefacts of the $FKHXOLDQDQG/HYDOORLVWUDGLWLRQVFXOPLQDWLQJLQODWH3DODHROLWKLFÀDN- LQJDQGPLFUROLWKLFLQGXVWULHV7KH3Ɨ৬XFXOWXUHLQ0DKRWWDUƯGLVWULFWLQ eastern Nepal, however, stands outside of the main subcontinental context and shows clear connexions with coetaneous technocomplexes in WKHIRUHVWHGKDELWDWVRI6RXWKHDVW$VLD3Ɨ৬XWHFKQRORJ\LVFKDUDFWHULVHG E\FREEOHWRROVDQGOHVVGLVWLQFWO\UHWRXFKHGVPDOOÀDNHWRROVDQGLQWKH Mesolithic context, by macroliths such as adzes and cobble tools as well DV WKH XQLIDFLDO ÀDWEDVHG DQG VWHHSHGJHG WRROV FDOOHG VXPDWUDOLWKV ZKLFKFOHDUO\VXJJHVWDQDI¿QLW\ZLWKWKH+zDEuQKLDQ

Indo-European. Rather the divide runs roughly through the sub-Hima- layas or the Terai. Similarly, there is a marked discontinuity between Neolithic and Bronze Age traditions up in the hills and those down on the plains. Yet such later archaeological assemblages appear, to our current state of knowledge, to be younger than the population genetic divide, and perhaps also to the linguistic one. The genetic divide between Tibeto-Burman and Austroasiatic in the region shows a far more complex structure than the clear line demarcating Tibeto-Burman from Indo-European. To the south, in the Brahmaputran basin and the Indo-Burmese borderlands, some of the spread of Tibeto-Burman may have been at the expense of indigenous Austroasiatic populations whom the Tibeto-Burmans assimilated linguistically. The Y haplogroup O2a (M95) is represented at a frequency of 77% in Austroasiatic groups in India and 47% in Tibeto-Burman groups of northeastern India (Sahoo et al. 2006). This pattern could suggest that Tibeto-Burman paternal lineages partially replaced indigenous Austroasiatic lineages in the northeast of the Indian subcontinent in the distant past, and that Austroasiatic populations preceded the Tibeto- Burmans in this region, as linguists and ethnographers have speculated for over a century and a half (van Driem 2001). The geographical extent of the Bodo-Koch languages and the shallow time depth of this sub- branch of Tibeto-Burman might represent the linguistic corollaries of such partial genetic replacement. My earlier arguments regarding the locations and geographical proximity of a more littoral Austroasiatic homeland and a more montane Tibeto-Burman homeland, both within or near the northeastern portion of the Indian subcontinent (van Driem 2006, 2007), continue to be supported by more recent genetic studies involving the Y-chromosomal haplogroups O2a and O3, e.g., Kumar et al. (2007). Austroasiatic is an old language family, and we would expect the population history of this family to be at least as complex as that of Tibeto-Burman, if not more so. In future, more detailed and careful FRUUHODWLRQRIOLQJXLVWLFDQGSRSXODWLRQJHQHWLF¿QGLQJVEDVHGRQPRUH ¿QHPHVK SRSXODWLRQ JHQHWLF VDPSOLQJ PD\ HQDEOH XV WR UHFRQVWUXFW early language contact situations and ancient cases of language shift and linguistic intrusions that might, for example, account for the phe- notypical differences readily observable between Munda speakers as opposed to Khasi-Khmuic and Mon-Khmer language communities as well as between Aslian negrito populations, Aslian non-negrito populations and the Nicobarese. Somewhat in parallel with such somatological 204 GEORGE VAN DRIEM observations by ethnographers, linguists have long observed corresponding typological differences between various branches of Austroasiatic. Donegan and Stampe (1993, 2004) suggest that Austroasiatic spread from the Indian subcontinent to Southeast Asia but argue that para- GR[LFDOO\WKHV\QWKHWLFKHDG¿QDOW\SRORJ\RI0XQGDODQJXDJHVUHVXOWHG from an innovative process of drift which unfolded within South Asia after the linguistic ancestors of modern Khasi-Khmuic and Mon-Khmer language communities migrated towards Southeast Asia. In their view, the typological change in Munda was triggered by a prosodic shift to a falling rhythm, whereas the analytic head-initial typology observed LQ.KDVL.KPXLFDQG0RQ.KPHUODQJXDJHVUHÀHFWVWKHPRUHRULJLQDO Austroasiatic state of affairs. By contrast, Zide and Anderson (1999, 2003) have argued that Munda verbal morphology is a conservative reten- tion, and that older Austroasiatic grammatical systems were secondarily lost in the Khasi-Khmuic and Mon-Khmer languages of Southeast Asia. Yet if the Father Tongue hypothesis holds true for the spread of Austroasiatic into South Asia (van Driem 2007), then this outcome would vindicate Robert von Heine-Geldern’s view of the Munda as the result of the ‘Einwanderung mongolider austrasiatischer Stämme in Vorderindien’ and of their ‘Mischung mit Dravida und Urbevölkerungselementen’ (1928, 1932). In that case, the typological divergence between Munda as opposed to Khasi-Khmuic and Mon-Khmer, lucidly discussed by Donegan and Stampe, may be the result of the adoption of an intrusive paternal tongue by indigenous pre-Austroasiatic populations of the Indian subcontinent. The apparent Munda penchant for a falling prosodic rhythm might then be just one residue of a far-reaching action de substrat. If, however, the Father Tongue hypothesis holds true for Austroasiatic, and Zide and Anderson are correct, then the Munda descendants of the linguistically assimilated South Asian indigenes may have more faith- fully preserved the original Austroasiatic morphology and grammar of the early bearers of the Y-chromosomal haplogroup O2a than is now seen UHÀHFWHG LQ PRGHUQ .KDVL.KPXLF DQG 0RQ.KPHU ODQJXDJHV ZKLFK XQGHUZHQWGLYHUJHQWDUHDOGHYHORSPHQWVVSHFL¿FWR6RXWKHDVW$VLD On the other hand, the Father Tongue hypothesis may very well not apply in all cases for the biological ancestry of Austroasiatic language communities, just as language spreading solely via the paternal line cannot account for the linguistic identity of all Tibeto-Burman populations, e.g., maternal Balti vs. paternal Hàn. More decisive insights into both the historical linguistics and historical typology of Austroasiatic and 7LEHWR%XUPDQ QHZO\ LGHQWL¿HG VLQJOH QXFOHRWLGH SRO\PRUSKLVPV RQ GLIMPSES OF THE ETHNOLINGUISTIC 205

WKH DXWRVRPHV HWKQROLQJXLVWLFDOO\ LQIRUPHG ¿QHPHVK JHQHWLF DVVD\V of Aslian negrito populations, Aslian non-negrito populations, the Nicobarese and peoples of Tibet and Burma, and insights from other disciplines may help us to retrieve more undiscovered bits and pieces of prehistory that may not have been irretrievably lost. 206 GEORGE VAN DRIEM

BIBLIOGRAPHY

Agrawal, D.P. 2002. “The earliest pottery and agriculture in South Asia.” In Yoshinori Yasuda, (ed.), The Origins of Pottery and Agriculture. New Delhi: Lusre Press and Roli Books, for the International Center for Japanese Studies, pp.81-8. Bakker, P. 1992. ‘A Language of Our Own’, the Genesis of Michif, the Mixed Cree- French Language of the Canadian Métis. Ph.D. dissertation, Universiteit van Amsterdam. —— 1994. “Michif, the Cree-French mixed language of the Métis buffalo hunters in Canada.” In P. Bakker and M. Mous (eds.), Mixed Languages, 15 Case Studies in Language Intertwining. Amsterdam: Instituut voor Fundamenteel Onderzoek naar Taal en Taalgebruik (Studies in Language and Language Use, 13), pp.13-33. Blench, R.M. 2001. “From the mountains to the valleys: Understanding ethnolinguistic geography in Southeast Asia.” In L. Sagart, R. Blench and A. Sanchez- Mazas (eds.), The Peopling of East Asia: Putting Together the Archaeology, Linguistics and Genetics. London: Routledge, pp.31-50. &Kԃ-LƗ\zX:+XDQJ64.XDQJ-0:DQJHWDO*HQHWLFUHODWLRQVKLSRI populations in China. Proceedings of the National Academy of Sciences of the United States of America 95: 11763-68. O’Connor, R.A. 1995. Agricultural change and ethnic succession in Southeast Asian states: A case for regional anthropology. Journal of Asian Studies 54(4): 968-96. Corvinus, G. 2007 [posthumous], G. Freund and L. Reisch (eds.), Prehistoric Cultures in Nepal: From the Early Palaeolithic to the Neolithic and the Quaternary Geology of the Dang-Deokhuri Dun Valleys, 2 vols. Wiesbaden: Harrassowitz Verlag. 'LIÀRWK*³7HQWDWLYHFDOLEUDWLRQRIWLPHGHSWKVLQ$XVWURDVLDWLFEUDQFKHV´ paper presented at the Colloque Perspectives sur la Phylogénie des Langues d’Asie Orientales at Périgueux, 30 August 2001. —— 2005. “The contribution of linguistic palaeontology to the homeland of Austroasiatic.” In L. Sagart, R. Blench and A. Sanchez-Mazas (eds.), The Peopling of East Asia: Putting Together the Archaeology, Linguistics and Genetics. London: Routledge Curzon, pp.77-80. Ding Yuan-Chun, S. Wooding, H.C. Harpending, Chi Han-Chang et al. 2000. Population structure and history in East Asia. Proceedings of the National Academy of Sciences of the United States of America. 97: 14003-06. Donegan, P.J. and D. Stampe. 1983. “Rhythm and holistic organization of language structure.” In J.F. Richardson, M. Marks and A. Chukerman (eds.), Papers from the Parasession of the Interplay of Phonology, Morphology and Syntax. Chicago: Chicago Linguistics Society, pp.337-63. —— 2004. “Rhythm and the synthetic drift of Munda.” In R. Singh (ed.), The Yearbook of South Asian Languages and Linguistics 2004. Berlin: Mouton de Gruyter, pp.3-36. van Driem, G. 1998. “Neolithic correlates of ancient Tibeto-Burman migrations.” In R. Blench and M. Spriggs (eds.), Archaeology and Language II. London: Routledge, pp.67-102. GLIMPSES OF THE ETHNOLINGUISTIC 207

—— 2001. Languages of the Himalayas: An Ethnolinguistic Handbook of the Greater Himalayan Region, containing an Introduction to the Symbiotic Theory of Language, 2 vols. Leiden: Brill. —— 2002. “Tibeto-Burman phylogeny and prehistory: Languages, material culture and genes.” In P. Bellwood and C. Renfrew (eds.), Examining the Farming/Language Dispersal Hypothesis. Cambridge: McDonald Institute for Archaeological Research, pp.233–49 —— 2006. The diversity of the Tibeto-Burman language family and the linguistic ancestry of Chinese. Bulletin of Chinese Linguistics 1(2): 211-70. —— 2007. Austroasiatic phylogeny and the Austroasiatic homeland in light of recent population genetic studies. Mon-Khmer Studies 37: 1-14. Edmondson, J.A. and D.B. Solnit (eds.) 1988. Comparative Kadai: Linguistic Studies beyond Thai. Arlington: Summer Institute of Linguistics of the University of Texas. —— 1997. Comparative Kadai: the Tai branch. Arlington: Summer Institute of Linguistics of the University of Texas. Fuller, D.Q. 2005a. Ceramics, seeds and culinary change in prehistoric India. Antiquity 79: 761-77. —— 2005b. Formation processes and palaeothenobotanical interpretation in South Asia. Journal of Interdisciplinary Studies 2(1): 93-115. ²²F 7KH *DQJHV RQ WKH ZRUOG 1HROLWKLF PDS 7KH VLJQL¿FDQFH RI UHFHQW research on agricultural origins in northern India. 3UƗJGKƗUƗ 16: 187-206. —— 2006a. Agricultural origins and frontiers in South Asia: A working synthesis. Journal of World Prehistory 20: 1-86. —— 2006b. “Dung mounds and domesticators: Early cultivation and pastoralism in Karnataka.” In C. Jarrige and V. Lefèvre (eds.), South Asian Archaeology 2001: Proceedings of the Sixteenth International Conference on South Asian Archaeology, European Association of South Asian Archaeologists, Paris, 2-6 July 2001. Paris: Éditions Recherche sur les Civilisations, pp.117-127. —— 2006c. “Silence before sedentism and the advent of cash crops: A status report on early agriculture in South Asia from plant domestication to the development of political economies (with an excursus on the proble of semantic shift among millets and rice).” In O. Toshiki (ed.), Proceedings of the pre-Symposium of the Research Institute for Humanity and Nature and 7th Ethnogenesis of South and Central Asia Round Table Harvard-Kyoto Roundtable. Kyoto: Research Institute for Humanity and Nature, pp.175-213 —— 2007a. Contrasting patterns in crop domestication and domestication rates: Recent archaeobotanical insights from the Old World. Annals of Botany 2007: 1-22. —— 2007b. “Non-human genetics, agricultural origins and historical linguistics in South Asia.” In M.D. Petraglia and B. Allchin (eds.), The Evolution and History of Human Populations in South Asia: Interdisciplinary Studies in Archaeology, Biological Anthropology, Linguistics and Genetics. Dordrecht: Springer, pp.393-443. Fuller, D.Q., E. Harvey and Ling Qin. 2007. Presumed domestication? Evidence for ZLOGULFHFXOWLYDWLRQDQGGRPHVWLFDWLRQLQWKH¿IWKPLOOHQQLXP%&RIWKH/RZHU Yangtze region. Antiquity 81: 316-31. 208 GEORGE VAN DRIEM

Harvey, E.L., D.Q. Fuller, R.K. Mohanty and B. Mohanta. 2006. Early agriculture in 2ULVVD6RPHDUFKDHRERWDQLFDOUHVXOWVDQG¿HOGREVHUYDWLRQVRQWKH1HROLWKLF Man and Environment 31(2): 21-32. Haudricourt, A.-G. 1954. Introduction à la phonologie historique des langues miao- yao. Bulletin de l’École Française d’Extrême-Orient 44: 555-74. Haudricourt, A.-G. and L. Hédin. 1987. L’homme et les plantes cultivées. Paris: Éditions A.-M. Métailié. Hazarika, M. 2005. Neolithic culture of Northeast India with Special Reference to the Origins of Agriculture and Pottery3X৆H 3RRQD 0$WKHVLV'HFFDQ&ROOHJH —— 2006a. Neolithic culture of northeast India: A recent perspective on the origins of pottery and agriculture. Ancient Asia 1: 25-43. —— 2006b. Understanding the process of plant and animal domestication in northeast India: A hypothetical approach. Asian Agri-History 10(3): 203-12. von Heine-Geldern, R. 1928. “Ein Beitrag zur Chronologie des Neolithikums in Südostasien.” In W. Koppers (ed.), Festschrift Pater Wilhelm Schmidt: 76 sprachwissenschaftliche, ethnologische, religionswissenschaftliche, prähistorische und andere Studien. Vienna: Mechitaristen-Congregations- Buchdruckerei, pp.809-43. —— 1932. Urheimat und früheste Wanderungen der Austronesier. Anthropos 27: 543-619. Jin Liya, A. Ganopolski, Fahu Chen, M. Claussen and Huijun Wang. 2005. Impacts of snow and glaciers over Tibetan Plateau on Holocene climate change: Sensitivity experiments with a coupled model of intermediate complexity. Geophysical Research Letters 32: L17709. Kaiser, K., G. Miehe, W.H. Schoch, A. Zander and F. Schlütz. 2006. Relief, soil and lost forests: Late Holocene environmental changes in southern Tibet (under human impact). Zeitschrift für Geomorphologie, Neue Folge 142: 149-73. Kaiser, K., W.H. Schoch and G. Miehe. 2007. Holocene paleosols and colluvial sediments in northeast Tibet (Qinghai province, China): Properties, dating and paleoenvironmental implications. Catena 69: 91-102. Kraaijenbrink, T., G. van Driem, Karma Tshering of Gaselô and P. de Knijff. 2006a. Allele frequency distribution for 21 autosomal STR loci in Bhutan. Forensic Science International 170: 68-72. Kraaijenbrink, T., G. van Driem, J.R.M.L. Opgenort, Nirmal Man Tuladhar and P. de Knijff. 2006b. Allele frequency distribution for 21 autosomal STR loci in Nepal. Forensic Science International 166(2-3): 176-81. Kuhle, M. 1985. Glaciation research in the Himalayas: a new ice age theory. Universitas 27: 281-94. —— 1986. The upper limit of glaciation in the Himalayas. GeoJournal 13: 331-46. —— 1987. “The problem of a Pleistocene inland glaciation of the northeastern Qinghai-Xizang Plateau.” In J. Hövermann and W. Wenying (eds.), Reports of the Qinghai–Xizang (Tibet) Plateau. Peking: Science Press, pp.250-315. ²²D*HRPRUSKRORJLFDO¿QGLQJVRQWKHEXLOGXSRI3OHLVWRFHQHJODFLDWLRQLQ Southern Tibet and on the problem of inland ice. GeoJournal 17: 457-512. —— 1988b. Topography as a fundamental element of glacial systems. GeoJournal 17: 545-68. —— 1990a. The cold deserts of high Asia (Tibet and contiguous mountains). GeoJournal 20: 319-23. GLIMPSES OF THE ETHNOLINGUISTIC 209

—— 1990b. “Ice marginal ramps and alluvial fans in semi-arid mountains: Convergence and difference.” In A.H. Rachocki and M. Church (eds.), Alluvial Fans: A Field Approach. Chichester: Wiley & Sons Ltd, pp.55-68. —— 1991. Observations supporting the Pleistocene inland glaciation of High Asia. GeoJournal 25: 131-231. —— 1995. Glacial isostatic uplift of Tibet as a consequence of a former ice sheet. GeoJournal 37(4): 431-49. ²²1HZ¿QGLQJVFRQFHUQLQJWKH,FH$JH /DVW*ODFLDO0D[LPXP *ODFLHU cover of the East-Pamir, of the Nanga Parbat up to the Central Himalaya and of Tibet, as well as the age of the Tibetan inland ice. GeoJournal 42(2-3): 87-257. —— 1999. Reconstruction of an approximately Quaternary Tibetan inland glaciation between the Mt. Everest and the Cho Oyu massifs and the Aksai Chin: A new JODFLRJHRPRUSKRORJLFDO6(1:GLDJRQDOSUR¿OHWKURXJK7LEHWDQGLWVFRQVH- quences for the glacial isostasy and Ice Age cycle. GeoJournal 47: 2-276. —— 2005. The maximum Ice Age (Würmian, Last Ice Age, LGM) glaciation of the Himalaya: A glaciogeomorphological investigation of glacier trim-lines, ice thicknesses and lowest former ice margin positions in the Mt. Everest-Makalu- Cho Oyu massifs (Khumbu and Khumbakarna Himal) including informations on late glacial, neoglacial and historical glacier stages, their snowline depres- sions and ages. GeoJournal 62: 193-650. Kumar, V., A.N.S. Reddy, J.P. Babu, T.N. Rao et al. 2007. Y-chromosome evidence suggests a common paternal heritage of Austro-Asiatic populations. BioMed Central Evolutionary Biology 7: 47. Lehmkuhl, F. 1995. Geomorphologische Untersuchungen zum Klima des Holozäns und Jungpleistozäns Osttibets. Göttinger Geographische Abhandlungen 102: 1-184. Lehmkuhl, F. and L.A. Owen. 2005. Late Quaternary glaciation of Tibet and the bordering mountains: A review. Boreas 34: 87-100. Miehe, G., S. Miehe, F. Schlütz, K. Kaiser et al. 2006. Palaeoecological and experimental evidence of former forests and woodlands in the treeless desert pastures of southern Tibet (Lhasa, A.R. Xizang, China). Palaeogeography, Palaeoclimatology, Palaeoecology 242: 54-67. 1DVX+LUǀ$UDWD0RPRKDUD7KH 5LFH DQG )RRG &XOWXUH RI 0XQGD LQ (DVWHUQ India: An Ethnolinguistic Study].\ǀWR.RNXVDL1LKRQ%XQND.HQN\ǌ6HQWƗ Ostapirat, W. 2005. “Kra-Dai and Austronesian: Notes on phonological correspondences and vocabulary distribution.” In L. Sagart, R. Blench and A. Sanchez- Mazas (eds.), The Peopling of East Asia: Putting Together Archaeology, Linguistics and Genetics. London: Routledge Curzon, pp.107-31. Owen, L.A., R.C. Finkel and M.W. Caffee. 2002. A note on the extent of glaciation in the Himalayas throughout the global Last Glacial Maximum. Quaternary Science 21: 147-57. 210 GEORGE VAN DRIEM

Owen, L.A., R.C. Finkel, Ma Haizhou, J.Q. Spencer et al. 2003. Timing and style of Late Quaternary glaciation in northeastern Tibet. Geological Society of America Bulletin 115(11): 1356-64. Pal, J.N. 1990. The early farming culture of northern India. Bulletin of Deccan College Post-Graduate and Research Institute 49: 297-304. Parkin, E.J., T. Kraayenbrink, Karma Tshering of Gaselô, G. van Driem et al. 2006a. 26-locus Y-STR typing in a Bhutanese population sample. Forensic Science International 161(1): 1-7. Parkin, E.J., T. Kraayenbrink, J.R.M.L. Opgenort, G. van Driem et al. 2006b. Diversity of 26 Y-STR haplotypes in a Nepalese population sample: Isolation and drift in the Himalayas. Forensic Science International 166(2-3): 176-81. Poloni, E.S., O. Semino, G. Passarino, A.S. Santachiara-Benerecetti et al. 1997. +XPDQ JHQHWLF DI¿QLWLHV IRU < FKURPRVRPH 3DITaqI haploptypes show strong correspondence with linguistics. American Journal of Human Genetics 61: 1015-35 (cf. the erratum published in 1998 in the American Journal of Human Genetics 62: 1267). Poloni, E.S., N. Ray, S. Schneider and A. Langaney. 2000. “Languages and genes: 0RGHV RI WUDQVPLVVLRQ REVHUYHG WKURXJK WKH DQDO\VLV RI PDOHVSHFL¿F DQG IHPDOHVSHFL¿F JHQHV´ ,Q -/ 'HVVDOOHV DQG / *KDGDNSRXU HGV Proceedings: Evolution of Language, 3rd International Conference 3-6 April 2000. Paris: École Nationale Supérieure des Télécommunications, pp.185-86. Purnell, H.C. 1970. Toward a Reconstruction of Proto Miao-Yao. Ph.D. dissertation, Cornell University, Ithaca. Qamar Raheel, Qasim Ayub, Aisha Mohyuddin, A. Helgason et al. 2002. Y-chromosomal variation in Pakistan. American Journal of Human Genetics 70: 1107–24. Quintana-Murci, L., C. Krausz, T. Zerjal, S.H. Sayar et al. 2001. Y-chromosome lineages trace diffusion of people and languages in southwestern Asia. American Journal of Human Genetics 68: 537–42. Ratliff, M. 2004. “Vocabulary of environment and subsistence and in the Hmong- Mien protolanguage.” In N. Tapp, J. Michaud, C. Culas and G. Yia Lee (eds.), Hmong/Miao in Asia. Chiang Mai: Silkworm Books, pp.147-65. Revel, N. 1988. Le riz en Asie du sud-est, 3 vols. Paris: Éditions de l’École des Hautes Études en Sciences Sociales. Sahoo, S., A. Singh, G. Himabindu, J. Banerjee et al. 2006. A prehistory of Indian Y chromosomes: Evaluating demic diffusion scenarios. Proceedings of the National Academy of Sciences 103(4): 843–8. 6FKlIHU-067VFKGL==KL]KRQJ:;LKDRHWDO7KHOLPLWHGLQÀXHQFH of glaciations in Tibet on global climate over the past 170,000 years. Earth and Planetary Science Letters 194: 287-97. Sharma, G.R., V.D. Misra, D. Mandal, B.B. Misra et al. 1980. Beginnings of Agriculture: From Hunting and Food Gathering to Domestication of Plants and Animals. Allahabad: Abhinav Prakashan. Shi Hong, Dong Yong-li, Wen Bo, Xiao Chun-Jie et al. 2005. Y-chromosome evidence RIVRXWKHUQRULJLQRIWKH(DVW$VLDQVSHFL¿FKDSORJURXS2±0American Journal of Human Genetics 77: 408–19. GLIMPSES OF THE ETHNOLINGUISTIC 211

6%ƯQJ;LDR-XQKXD38QGHUKLOO5DQMDQ'HNDHWDO<FKURPRVRPHHYL- dence for a northward migration of modern humans into eastern Asia during the last Ice Age. American Journal of Human Genetics 65: 1718–24. 6%ƯQJ&KXQMLH;LDR5'HND076HLHOVWDGHWDO3DODHRERWDQ\DQGLWVUHFHQW progress], .iRJ· 2005(7): 42-9. Zide, N.H. and G.D.S. Anderson. 1999. “The Proto-Munda verb system and some connections with Mon-Khmer.” In P. Bhaskararao (ed.), Working Papers of the International Symposium on South Asian Languages: Contact, Convergence and Typology, Tokyo, December 6-9, 1999. Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa, pp.400-19. Zide, N.H. and G.D.S. Anderson. 2003. “Towards an analysis of the South Munda verbal system.” In K.L. Adams, T.J. Hudak and F.K. Lehman (eds.), Papers from the Seventh Annual Meeting of the Southeast Asian Linguistics Society. Tempe, Arizona: Arizona State University Program for Southeast Asian Studies, pp.243-59. =RQJ<=&KHQ-%,QQHV&&KHQHWDO)LUHDQGÀRRGPDQDJHPHQWRIFRDVWDO VZDPSHQDEOHG¿UVWULFHSDGG\FXOWLYDWLRQLQHDVW&KLQDNature 449: 459-62.

ORIGIN AND MIGRATION MYTHS IN THE RHETORIC OF NAGA INDEPENDENCE AND COLLECTIVE IDENTITY

MARION WETTSTEIN

“See!” exclaimed the gentleman in his 50s while we were enjoying our meal in a small Sumi Naga village, “Even the rice is different, the meat is different, the vegetable is different. This clearly proves that we Nagas DUHQRW,QGLDQV´+HFRQWLQXHGWREULHÀ\H[SODLQWKDWWKH\KDGDOZD\V been, and still are, a free people and therefore need not ask for independence; they always were independent. The conversation then took the turn it often does when encountering one of the representatives active for the ‘Naga cause’: a lecture in history informing one that the Nagas migrated from somewhere in Thailand or Mongolia—where all Mongol peoples came from—to their present location in the hills between Assam and Burma. Among many politically active Nagas, migration stories form an integral part of their rhetoric for constructing a collective Naga identity and for claiming an independent Naga nation. The same is true of local intellectual discourse, which is mainly lead by Nagas of sociological, anthropological, or theological background, and mostly—and not surprisingly—in favour of the nationalistic cause. The aim of this paper1 is to discuss the use of origin tales and migration myths in this local discourse on Naga collective identity. Thus, the following discussion is not intended as a political statement but as a critical analysis of local recourses in the debate in question.

THE QUEST FOR A COLLECTIVE NAGA IDENTITY

The hill peoples commonly referred to as ‘the Nagas’ dwell in the northern hill ranges between the Brahmaputra and Chindwin Rivers on both sides of the present India-Myanmar (Burma) international border. Their number on the Indian side is estimated at approximately two million (Census of India 2001, projection 2008). Since colonial times, and

1 Data for this study was gained within the project Material Culture, Oral Traditions and Identity among the Naga in Northeast India, funded by the Swiss National Science Foundation, and lead by Prof. Michael Oppitz, former Director of the Ethnographic Museum of Zürich University. 214 MARION WETTSTEIN especially since the beginning of the twentieth century, the Nagas were known to Western observers for their outstanding cultural features of headhunting, feasts of merit and a spectacular material culture. To this day, illustrated published accounts tend to emphasise these aspects of Naga culture (Stirn and van Ham 2000 and 2003, Arya and Joshi 2004). The popularity of these topics has its roots in British colonial ethnography and museum collections,2 primarily those contributed by James Phillip Mills (1922, 1926, 1937), John Henry Hutton (1921a, 1921b), and the Austria-born anthropologist Christoph von Fürer-Haimendorf (1939). While these features were an important testimony to the ‘traditional’ Naga culture, they have virtually vanished today in the wake of Christian conversion by American Baptist Missionaries, along with other introduced changes and local agency. It is widely accepted among both Naga intellectuals and the general public that in former times—and even to some extent today—the essential element of identity has been the village. One often hears that the Nagas lived in self-sustained village-republics, which had bilateral relations and contracts with other, at times quite distant, villages. A second important element of identity was and is the clan (Jamir and Lanunungsang 2005:37ff.; Kejong 2008; Kumar 2005:67-72). Before the arrival of the British during the mid-nineteenth century, hill peoples RIWKH1DJDUHJLRQGLGQRWLGHQWLI\WKHPVHOYHVZLWKVSHFL¿FWULEHV2QO\ a few groups, such as the Ao, formed larger entities with a collective name. Usually, if a person was asked who they were, they most probably replied using their village name. In most regions it was British administrators who categorised local populations into tribes, and many of today’s tribal borders date back to colonial times (von Stockhausen 2008). Nowadays, the notion of the tribe is commonly accepted and WULEDOLGHQWL¿FDWLRQLVTXLWHVWURQJDPRQJWKH1DJDV Furthermore, in the past the Nagas did not use the generic term ‘Naga’ to refer to themselves; this designation derives from agents outside of their region. There are different opinions about the origin and meaning of the word. One of the most widely accepted refers to different forms of Assamese ‘noga’ or Sanskrit/Hindi ‘nanga’ meaning ‘naked’ (van Driem 2008; Kumar 2005:23-24). Most probably the usage dates back to the period of Ahom rule in Assam, beginning around the thirteenth century.

26LJQL¿FDQWFROOHFWLRQVRI1DJDDUWLIDFWVDUHKRXVHGDWWKH3LWW5LYHUV0XVHXPRI2[IRUG University, the Museum of Ethnology in Vienna, the Ethnographic Museum of Zürich University, the Museum der Kulturen in Basel, the Völkerkundemuseum Berlin-Dahlem, and the Cambridge Museum of Archaeology and Anthropology. For object catalogues, see Jacobs 1990 or Kunz and Joshi 2008. ORIGIN AND MIGRATION MYTHS 215

The Ahom were a Tai-speaking people who migrated from Burma into the upper Assam valley and settled on the plains. Their kingdoms always maintained contacts with hill peoples on both sides of the Brahmaputra, as their chronicles record (Barua 1985). When the British conquered the Ahom kingdom, they adopted the term ‘Naga’ for the tribes living in the hills to the east, which came to be named the Naga Hills. As the Nagas had no common term for themselves as a larger entity, they adopted the foreign name for convenience of communication with outsiders (von Stockhausen 2009). However, there are certain regions of the Naga Hills where, according to Kumar (2005:24-25), up until recently local populations still resented being referred to as ‘Nagas’. Under colonialism, the Nagas did not feel integrated into British India. Experiences of the First World War can be seen as the initial trigger that awakened a collective Naga identity. A group of between one and two thousand Nagas were taken by the British to Europe to be employed as labour corps during the war effort, and while building roads in France they became aware of a greater political context. Some of those who UHWXUQHGIRXQGHGWKH1DJD&OXEWKH¿UVWSROLWLFDORUJDQLVDWLRQFODLPLQJ to represent the Nagas as a people (Nuh and Wetshokhrolo 2002:111; :HVW DQGE\WKHVWKH¿UVWSHWLWLRQVIRU1DJDLQGHSHQ- dence began to appear. A second important juncture for Naga identity appears to be the period of British withdrawal from South Asia and subsequent modern state formation in the region. The awareness that only as a united people would the Nagas have a chance to withstand the newly forming Indian and Burman nation states was most probably the initial motivation for attempting to bring the very diverse and often mutually hostile villages and tribes together under a single umbrella. Since then, incorporation within a collective Naga identity remains an incomplete SURFHVVDQGLQPDQ\UHVSHFWVLGHQWL¿FDWLRQZLWKRQH¶VYLOODJHDQGWULEH UHPDLQVVWURQJHUWKDQLGHQWL¿FDWLRQZLWKWKH1DJDFROOHFWLYH0RUHRYHU the question of exactly who is a ‘Naga’ and who is not continues to generate quite heated debate at times. Shortly before Indian independence in 1947, the Naga National Council (NNC) was founded and formerly demanded independence from the British in the name of the Naga people. The Governor of Assam for the British administration, Sir Akbar Hyder Ali, negotiated a nine-point- agreement with the NNC which foresaw that the Nagas should decide their political future by themselves. Nevertheless, the Naga Hills were then annexed by independent India. Based upon the previous British ‘inner line regulation,’ the region was declared a ‘restricted area’ and 216 MARION WETTSTEIN foreign visitors were not permitted to enter until the year 2000. A plebi- scite among the Nagas during 1951 revealed that 90% wanted to have their own state.3 Since India failed to accept any Naga independence, the NNC went underground and formed a rebel army to drive Indian troops from the region. Many bloody battles later, one faction of the NNC signed the so-called Shillong Accord accepting Naga territorial integration within India, and the Indian union state of Nagaland was FUHDWHGLQ$QRWKHUIDFWLRQRI1DJD¿JKWHUVIRXQGHGWKH1DWLRQDO Socialist Council of Nagaland/Nagalim (NSCN) and continued to wage campaigns against the Indian Army well into the 1980s. According to Naga sources, an estimated 100,000-200,000 Nagas lost their lives in this war, which remained largely unnoticed outside of South Asia (Oppitz 2008:26). During the late 1980s, the NSCN itself split into two factions, HDFKRIZKLFKFRQWLQXHVWR¿HOGLWVRZQXQGHUJURXQGDUPLHV:KLOHERWK JURXSVKDYHDFHDVH¿UHDJUHHPHQWZLWK,QGLDVLQFHDQGQRZUDWKHU ¿JKWDJDLQVWHDFKRWKHUWKHLUFRPPRQGHPDQGUHPDLQVIRUDQLQGHSHQ- dent Naga Nation. While support for the underground among the general public is not as enthusiastic as it used to be, the number of sympathisers remains high, with many political activists, political journalist and intellectuals supporting the agenda of the underground. During the course of talks between the Indian Government and the underground leaders, a central point of the Naga rhetoric has been that the Nagas are one people, distinct from the Indians. This same point has been picked up by the general public. Formation of such a collective identity is considered crucial for the ‘we-feeling’ of the new Nation. The rhetoric in the local discourse is mainly concerned with justifying the demand for a ‘Naga Homeland’. In my opinion, four quite dominant statements can be isolated at the core of this Naga rhetoric. They may be simply stated as follows: (1) the Nagas are not Indians; (2) the Nagas are one people; (3) the Nagas demand nothing other than the land upon which they have lived since time immemorial; (4) the Nagas were always independent and were never conquered by anybody (at least until the British came, and then only partially). These four statements appear to be the basis for the notion of the ‘uniqueness of the Nagas’, an expression which is often referred to—directly or indirectly—in today’s newspapers, in speeches by politicians and as a standard expression in general discourse (e.g.,Vero 2005, in the local daily Morung Express). A close look at the arguments supporting these four statements will

3 Nuh and Wetshokhrolo 2002:67, Prakash 2007:1930, IWGIA 1986:21ff. The actual conduct of this vote and exactly who it included remain disputed points until today. ORIGIN AND MIGRATION MYTHS 217 reveal that origin and migration myths play a crucial part in the rhetoric. :HDOVR¿QGWZRW\SHVRIDUJXPHQWVXVHGRQHLVYDJXHLQLWVQDUUDWLYH evidence, but is accepted by a large majority of the Nagas and therefore intensely instrumentalised; the other is precise and detailed in narration although not representative for all Naga groups, and when sometimes put forward can create considerable controversy at times.

VAGUELY COMING FROM MONGOLIA

The notion that the Nagas are not Indians is a dominant perception among Nagas, at least when talking in terms of culture. ‘Indianness’ is generally associated with an entirely different set of religiosity, language, material culture and physical features. Naga experiences when in ‘mainland’ India—if we may call most areas to the west of Nagaland VXFKIRUFRQYHQLHQFH²FHUWDLQO\FRQ¿UPWKLV+DYLQJPRUH%XUPDQRU Tibetan features, Nagas are often perceived as foreigners by Indians: they may be charged tourist prices because they are taken for Japanese, or they may be accused of lying when insisting that they are not Tibetan immigrants. Thus, one can say the notion that ‘Nagas are not Indians’ is DFWXDOO\JHQHUDOO\VKDUHGDQGWKLVSRLQWLVFHUWDLQO\FRQ¿UPHGIRU1DJDV through experience. However, for many Nagas this does not automati- cally mean that politically they would not wish to be Indian. The younger JHQHUDWLRQLQSDUWLFXODUHQMR\WKHEHQH¿WVRIEHLQJ,QGLDQFLWL]HQV0DQ\ have absolved their education in one of the big Indian cities and rely upon comforts the Indian welfare state and the Indian economy provides for them. So they would rather prefer a better form of integration than a complete separation. However, other Nagas would argue that this is exactly how the Indian state ‘buys’ the Nagas, and subjects them not by war but by their own accommodating nature. Thus, in pro-independence thinking the Nagas should not be Indian in any political sense. One of the main arguments for the Nagas not being Indians is based upon oral traditions that place Naga origins either somewhere to the southeast of their current homeland, or in the South Seas, Mongolia or China. From the narrative point of view, such myths are not very precise and not necessarily what one may encounter in an evening ses- VLRQ RI VLQJLQJ RU SRHWLF UHFLWDWLRQ DURXQG WKH KHDUWK ¿UH²WKH XVXDO form in which Naga oral tradition is preserved. The stories may be summarised in a discussion or a speech and they may be recounted quite vaguely. They may, but rather seldom do, mention village names, most of which are nonexistent today and nobody knows exactly where they 218 MARION WETTSTEIN were once situated. Such village names are often associated or even equated with similar sounding village names somewhere in Mongolia, like Mongkhuma/Mongkonyu, to which I will return below. In order WRSURYLGHWKHFODLPVZLWKDVFLHQWL¿FEDFNLQJDUJXPHQWDWLRQE\ORFDO intellectuals is enriched with speculations originally taken from the monographs of Mills and Hutton or from other authors of the early twentieth century. Evolutionist or diffusionist ideas often come into play. These are, in my opinion, to a large extent reinterpreted from, and based upon a misunderstanding of, the ‘Introduction’ in Jacobs’ work The Nagas (1990:10-14), albeit that he explicitly warns readers “[i]t is HDV\EXWSUREDEO\XQKHOSIXOWRLPSO\WKDWVRFLDORUOLQJXLVWLFLQÀXHQFHV are a matter of the physical intrusive migration”. A remarkable example for this type of rationale can be seen in the works of Reverend V.K. Nuh, former secretary of the Council of the Baptist Churches in Nagaland and mediator in the peace process between WKHXQGHUJURXQG¿JKWHUVDQGWKH,QGLDQ*RYHUQPHQW,QWKHVHIXQFWLRQV KH KDV EHHQ D PRVW LQÀXHQWLDO ZULWHU DQG DQ LQWHOOHFWXDO DXWKRULW\ LQ Nagaland. He is not, however, what one would call a political activ- ist and therefore his argumentation may be a suitable example for the general public to explain who the Nagas are. Apart from a collection of historical and political documents called The Naga Chronicle (Nuh and Wetshokhrolo 2002), his more than a dozen locally edited books centre on theological questions: on the morality of the Nagas; on the Christian history of the Nagas; or on an indigenous theology (e.g., Nuh 1986, 1996a, 1996b, 1998, 2001, 2002, 2006). In some of his works KH SXEOLVKHV D SUHIDFH RU ¿UVW FKDSWHU H[SODLQLQJ DERXW WKH 1DJDV LQ general, including their origins. For example, …there is a general consensus that has been agreed upon by all scholars that the Nagas are Mongoloid by race. […] It is believed that the earliest home of the Mongoloid people was on the upper reaches of the Hwang-Ho River in China and further moving down to South-East Asia through the mountains. The Nagas belong to the great Mongolian race, which spread all over the world as far as China and South America. […] History unveils WKDWWKH1DJDVDUHIURPDKLJKHUFLYLOL]DWLRQZKLFKÀRXULVKHGVRPHZKHUH in South East Asia from time immemorial. (Nuh 2002:2-3) He even goes as far as to suggest that The Nagas were probably the original megalith builders and many of the megaliths that were erected over the world from Europe to America to South-East Asia and Middle East, were in all likelihood erected by them. The reason for this conclusion is not that the raising of megaliths ORIGIN AND MIGRATION MYTHS 219

is immune to independent origin, but that these megaliths are connected ZLWKDGH¿QLWHDQGXQLWHGFXOWXUHFDOOHGWKH1DJDFXOWXUH 1XK In another of his introductions employing virtually the same wording, Nuh emphasises that the Nagas he is discussing are not the present Nagas, but an ancient race—which he likes to write as Naga in italics—that is to say, a primary culture dispersing through the whole world of which the current Nagas are descendants. Its origins “could be anywhere from Sumeria and India to Southeast Asia and Oceania” (Nuh and Wetshokhrolo 2002:9), and they were a seafaring people—with a reference to Thor Heyerdahl—who spread to all coasts of America, Africa, Europe and Asia (Nuh and Wetshokhrolo 2002:11-12). Finally Nuh comes to the conclusion that It would not be unreasonable if Naga people are tempted to think they had Hebrew ancestors. For according to the Apocryphical II Andross, the Ten Tribes of Israel made miraculous crossing of the Euphrates after the exile imposed by the Assyrian empire. They ventured for a year and a half beyond the river to a far off country known as Arshaut. There is a great likelihood that they retreated [sic.] Malay Archipelago. (Nuh and Wetshokhrolo 2002:13) When talking of the current Naga groups, Nuh admits that …there are many theories propounded so far about, and on the origin and progress of the Nagas as a people. Along with these theories of origin are GH¿QLWLYHO\YLHZVRIKRZWKH\FDPHWREHLQWKHODQGWKH\SUHVHQWO\GZHOO Both Nagas and non-Nagas have put these theories forward after serious consideration and researches. They are mostly based on superstitions, though. (Nuh 2006:14) Here ‘superstitions’ does not mean oral tradition, which Nuh defends as being a most valuable source. He asserts that the tracing back of the origins of the Nagas has to be done using folktales and oral tradition, and to the defence of this method he insists repeatedly that these oral traditions can be trusted as sources and are invaluable for the reconstruction of Naga history (Nuh and Wetshokhrolo 2002:14). However, Nuh gives no example of such a folktale in the form of a citation or a song text. It seems that ‘superstitions’ does not refer to the British colonial monographs either, which do refer to oral traditions or to linguistic and technological comparison hinting at an origin in China or Southeast Asia, which Nuh agrees with. As in many local publications, no references are given for many statements. Nevertheless, it can be assumed that the basis for his argumentation is taken from Hutton’s and Mills’ 220 MARION WETTSTEIN monographs. In The Angami Nagas, for example, the second migration wave from northwest China originating between the upper Yangtze and the Ho-ang-ho river is mentioned for the Tibeto-Burman-speaking peoples of the region (Hutton 1921a:6-9), while Mills’ Introduction to The Lotha Nagas (1922:xvi-xxi) suggests at length four directions of migration into the greater region of what is today called northeast ,QGLDD¿UVWIURPWKHQRUWKWKDWLV7LEHWDQG1HSDOIRUWKHQHLJKERXULQJ Singpho, Aka, Mishmi and Bodo groups; a second from south China via the Irrawaddy River for the neighbouring Shan, Ahom and Taman groups; a third from the south for the neighbouring Lushai and Kuki groups; and a fourth—and probably very early one—from the Kol-Mon- Annan region by people occupying the territory before the other groups arrived. While Mills insists that today’s Nagas are probably a mix of all these migration waves, the Tibetan or Nepalese origins are overlooked for the most part in local Naga accounts. And, while Mills points to VWURQJDI¿QLWLHVZLWKLPPHGLDWHQHLJKERXUV1XKSUHIHUVWRHPSKDVLVH DI¿QLWLHVZLWKWKHPRUHGLVWDQW3DSXDRU3KLOLSSLQHJURXSV 7KXV9. 1XK QHYHU VSHFL¿HV ZKDW KH DFWXDOO\ PHDQV E\ µVXSHU- stition’. In summary, he leaves us with the impression that the early origins and migrations of the Nagas—from somewhere in the East, in China, Mongolia, or in the South Sea—are mainly reconstructed from oral traditions. And it is this provenance that proves that the Nagas are QRW,QGLDQV%XWFRQ¿UPDWLRQVRIVXFKRUDOWUDGLWLRQVLQSUHFLVHP\WKV songs or poems are neither given by the author himself nor derivable from his most probable sources, the British monographs. A slightly more precise picture is given by Jamir and Lanunungsang (2005:11-12). Jamir is a freelance author writing about his own tribe, the Ao, after a career serving in the state administration. Lanunungsang is a professor of sociology at the University of Nagaland. In their work the pair trace the origin of the Nagas back to a Mongolian village named Mongkhuma and, according to Burmese oral tradition, give a migration route passing through Yunnan into today’s Kachin area and the foundation of a village named Khankha. From there the migratory group is said to have moved over the hill regions of Melikha and Khamung to the upper Chindwin River and into Hwakong valley to reach a place called Sukai. Up until this point the entire migratory group stayed together, but then split into two: one group moved southwards over the Patkai Hill ranges until they reached the plains of Manipur. From there they dispersed towards the north into today’s Naga Hills; the other group crossed the Chindwin River in the north and spread over the same hills ORIGIN AND MIGRATION MYTHS 221 from the northern and eastern sides. This account might sound like a reasonably funded oral tradition, even supplying village names and giving their geographical locations. 7KHH[WHQWWRZKLFK-DPLUDQG/DQXQXQJVDQJ¶VQDUUDWLYHKDVEHHQLQÀX- HQFHGE\DFFRXQWVIURPHDUOLHU%ULWLVKOLWHUDWXUHLVGLI¿FXOWWRGHWHUPLQH The authors do not cite any songs in support of their information, but they give two primary sources in a footnote. The interesting background PHDQLQJRIWKHVHUHIHUHQFHVPLJKWHDVLO\EHRYHUORRNHGWKH¿UVWVRXUFH refers to ‘Burmese elders’, namely S.S. Khaplang; the second is a book- let by Isaac Chichi Swu. To understand the meaning that resonates for Naga readers in these references, it is necessary to take a short look into the recent political situation once more. The sources each refer to the name of a leader of the two main opposing armed underground factions, the NSCN-K led by Shangwang Shangyung Khaplang, and the NSCN-IM led by Thuingaleng Muivah and Isaac Chichi Swu. The narrative is thus directly linked to the most prominent representatives of the claim for an independent Naga Nation. Since the splitting of the NSCN, FRQÀLFWVEHWZHHQWKHVHWZRIDFWLRQVGXULQJWKHSDVWIHZGHFDGHVKDYH resulted in thousands of Naga civilian deaths. Local authors cannot cite one faction and omit reference to the other without getting into severe trouble, but in my opinion there is more to these references than merely maintaining an equilibrium of factional representation. Leaving aside the strong political bias of the sources as such, in my view Jamir and Lanunungsang created an interesting possibility for political criticism by citing both factions in the same footnote for the same migration story: 2QHRIWKHPDLQPXWXDODFFXVDWLRQVH[FKDQJHGEHWZHHQWKHWZR¿JKW- ing factions is that each denies the other as being Naga. A feature of this accusation is that one of the leaders (Muivah) is a Tangkhul Naga from Manipur, while the other (Khaplang) is a Heimi Naga from the Burmese side. One can often hear the objection aimed in both directions that the other is not even ‘from Nagaland’. On the other hand, we could say that the Tangkhul and the Sumi (Isaac Chichi Swu’s tribal DI¿OLDWLRQ DFFRUGLQJWRWKHDERYHUHFRXQWHGVWRU\ZRXOGEHORQJWRWKH JURXSZKRPLJUDWHG¿UVWVRXWKDQGWKHQQRUWKZKLOHWKH+HLPL1DJDV settled along the northern route without entering today’s Nagaland. This would mean that the two factions, via their leaders, represent the two migration groups in Jamir and Lanunungsang’s narrative. Seen from this perspective, the migration story suggests not only that both factions are originally of the same blood, but also that they both even tell the same story and thus suggest themselves that they are brothers. By means of 222 MARION WETTSTEIN a migration story the authors not only succeed in indicating that the Nagas are one people coming from Mongolia, but also make a political statement about the oneness of all Nagas that goes beyond the defence against India by hinting at the inner disruption of today’s Naga society without criticising either of the armed factions directly.

PRECISE BUT NOT COMMONLY ACCEPTED

The opinion that the Nagas are one people has recently been stressed by one of the most important and locally renowned Naga anthropologists, Abraham Lotha. In his 2008 article, he sets out his opinions about which features constitute Naga unity, for example, clan organisation, the system of morung or men’s dormitories, headhunting, feasts of merit, and recent Naga history which generated a ‘we-feeling’. But when Lotha mentions migration, he formulates with great caution, observing how certain migration myths have become the focus of a popular topos which states that the Nagas are one people, regardless of the fact that there is no place to which all Nagas tribes can trace back their origin. In contrast to the vague ‘Mongolian-origins’ or ‘eastern-origins’ myths reviewed above, there are many migration myths of a precise and detailed narration, and which mention origin places situated within Naga territory itself. One such place of importance is Chungliyimti village, ZKLFKLVKHOGWREHWKHRULJLQRIWKH$R1DJD2I¿FLDO$RPLJUDWLRQ myths do not recognise any migration prior to Chungliyimti, stating that the Ao arose out of six stones—three male and three female—near the village and which represent each of the three clans of Pongen, Longkum and Jamir. These clans married mutually and this was the beginning of the Ao tribe. So the argument that the Nagas are one single people because WKH\KDYHRQHFRPPRQRULJLQZLWKRWKHUWULEHVZRXOGQRW¿QGIDYRXULQ RI¿FLDO$RH\HV6HYHUDO\HDUVDJRDQ$RVFKRODU3XUWRQJ]XN/RQJFKDU (2002), suggested a successive migration history from the south to the north, and which preceded the time of Chungliyimti. His interpretation FRVW KLP ¿QHV LPSRVHG E\ PDQ\$R YLOODJH FRXQFLOV 6LQFH WKHQ$R authors have become somewhat cautious in disputing the Chungliyimti myth. The myth in its recited form is expressed in a number of songs which differ from village to village but which commonly focus upon the VL[VWRQHVDQGWKH¿UVWYLOODJHZKHUHDOOFXOWXUDODQGVRFLDOIHDWXUHVRI Ao society were founded. Ao songs are highly poetical and composed in an archaic language, thus their meaning is often not transparent at ¿UVWVLJKW7RGD\¶V\RXQJJHQHUDWLRQRI$RUHTXLUHWKHLQWHUSUHWDWLRQRI ORIGIN AND MIGRATION MYTHS 223 elder singers in order to understand lyrics. Additionally, grasping the connotations of a song usually requires a large amount of ethnographic background information. Here is an example of one such song referring to the six stones (longterok LQ&KXQJOL\LPWL ¿JXUH Oh formed from six stones In the land of Chungliyimti a head-dance was performed A great warrior Shiluti was born During the time of Konem and Arangba The Mongsen enemies were attacked and lost.4 Chungliyimti is not only reckoned as the origin place of the Ao Naga, but is also an important site featuring in the migration of other tribes. The village is presently situated in Sangtam area, and the current head gaonbura DYLOODJHHOGHUZLWKDGPLQLVWUDWLYHIXQFWLRQ H[SODLQHGWKDW¿YHWULEHV the Ao, Sangtam, Konyak, Phom and Chang claim it as important for their migration. He gives an account of a migration starting in Thailand at a place named Chamai or Chaiem (a name he is sure the British mis- spelled as Shemshen) following a war with the Burmese. The migratory group crossed over several rivers until they reached a place named Mükur (‘skull’) near the mountain of Saramati. His account continues, From there we stayed in Khozama and left the Chakhesang people and then we came to Yimphir, there we left the Yimchungr and also the Sangtam part of Tukumi district. We sent the Khiamnungan from Yimphir on and some of the Yimchungr. People from the Kephire district of Tukumi were also sent from there. From there we went to Mao and from there the Tangkhul, Kuki and Zeliang left. After that we came to Philimirutomi and from there the Lotha were sent. We left Phek area and when the Angami and Sema followed us, we told them to go back again. The people who came to Chungliyimti from Phek area were the Rutu, Luma, Pangrong, Jimilomi and Liten.5 But the Ao, Sangtam, Konyak and Phom came directly from Yimphir.6 While the narrator of the story admitted that, from the perspective of the tribes mentioned in it, the story might look different for each, as far as the Sangtam are concerned he had learned from his father that there 4 See Kaiser 2004: O longterok ko oker Chungliyimti lima saku semer yartina, Shiluti nokzenketer soka Konem, Arangba soyimla Mongsen sariertsü nenoksem ko mezüjar. 5 These are not names of present-day tribes, but most probably of single clans. 6 Hotingse Sangtam, head gaonbura of Chungliyimti, interviewed on 16 November 2008 at Chungliyimti/Longterok by Alban von Stockhausen of the Material Culture, Oral Traditions and Identity among the Naga in Northeast India Project. 224 MARION WETTSTEIN were three periods in the development of the place later to be known DV&KXQJOL\LPWL7KH¿UVWZDVFDOOHG0HWLGHPLQ$RODQJXDJHDQDPH which means ‘animals and humans living together’. The narrator clearly dated the arrival of humans in the area to between 550 and 400 BCE. Where this dating comes from is unclear; he just heard it, according to tradition. The second period was called Semetem meaning ‘settled together’, because many more people arrived at the place. And the third period was then called Chungliyimti, which means ‘shield place’, the name explained by the residence there of a man of Longkumer clan who was so excellent at manufacturing shields that people came even from Angami area to buy them. The Sangtam lived there together for 900 years and once numbered about 9000 people, but a part of them decided to leave the place between 81 and 115 CE, a dating which the narrator FRQ¿GHQWO\ RIIHUHG )XUWKHU KH H[SODLQHG WKDW WKH SHRSOH ZKR ZRXOG become called the Ao—actually ‘Aor’ (‘those who went’)—crossed the Dikhu River westwards, destroying the bridge behind them, and left the rest of their community behind because they feared overcrowding. Those people who were left behind became the Phom and the Konyak. They took another route ‘sideways’ along the river terrace and were therefore called the ‘merir’. The Ao believe that they came out of the stones and WKHVWRQHVZHUHWKHUHLQWKH¿UVWSODFHEXWWKH6DQJWDPgaonbura has another story about this: It was during the early Chungliyimti period that the people brought together the stones to commemorate the big leaders (‘Ong’) of the village. That is why the Sangtam themselves call the stones Ongterok (‘Six Leaders’), and not Longterok (‘Six Stones’) as do the Ao. He also refers to a recently conducted archaeological excava- tion (compare Jamir and Vasa 2008), which showed signs of holes for house poles under the stones. To him this represented a clear proof that the settlement, and therefore people, existed before the stones. Thus, he concluded: “If somebody says that the Ao don’t come from the six VWRQHVKHZLOOEH¿QHG>«@,NQRZWKHVWRU\EH\RQG&KXQJOL\LPWLEXW the Ao don’t appreciate it. So what can I do?” The Chang, who according to the Sangtam gaonboura also lived in Chungliyimti, do not however refer to this place when asked for their origin. They clearly state that their origin place is Mongkonyu. It is also called Changsang Mongko and was the origin of all cultural invention (Kumar 2005:56-57). One of the songs about the village runs: ORIGIN AND MIGRATION MYTHS 225

Figure 10.1. Longterok/Ongterok stones near the village of Chungliyimti, origin of the six clans of the Ao and, in the background to the right of the WUHHPHPRULDOVWRQHVIRUWKH¿UVWYLOODJHOHDGHUVRIWKH6DQJWDP,QWKH foreground, the stone of the Longkumer clan of the Ao. (Photograph by Alban von Stockhausen, 2008). 226 MARION WETTSTEIN

Besi bou pakbu ayingsi loubü kaoen It is believed that our forefathers nyen Mongkonyu to nyenkei, originated from Mongkonyu, müt kho chungnyu boka pakei to tet it is believed that man originated from kei, a rubber tree, honglau bükho Mongkonyu ka pakei Foreigners also originated at lauta. Mongkonyu.

Mütsi saonyu wan sat yubü mütla )RUPDNLQJWKH¿UVW¿UHWKHUHZDVD wanlapkei, wan mütei jukei. contest between tiger and man, but PDQLQYHQWHG¿UHPDQLQYHQWHG¿UH

Lei tobu kho Mongkonyu a to Writing was invented in Mongkonyu, khangkei, in yau bü kho Mongkonyua yaukei, PDFKHWHVZHUH¿UVWPDGHLQ Mongkonyu, jak mongkonyua yaukei, Earthen pots were made in Mongkonyu, nei mongkonyua takei, Clothes were woven in Mongkonyu, wüt si yang youk kei. KDQGLFUDIWVZHUH¿UVWGRQHWKHUH

Shou kho Mongkonyu sanga pok Babies were born in Mongkonyu, khangkei, phang jankei, dynasties were divide from there, kumsi lei janyu kei, works was divided clanwise, lao lam yak nga yu kei, marriages were made there, Lümsi lang lei kei, worships were held, chisi chia tom la Mongkonyu folk-dances originated from khangkei, Mongkonyu, sekmou khangbü kho Mongkonyu cultivation was learned from sanga khangkei. Mongkonyu.

Müt si saonyu hoba ngou to son Man, tiger and bison lived together, lamkei, Mütei kok kok shou la lei kei, people adopted worms as their children, mütsi saonyu jakei hoba la yeian enmity and quarrels broke out between lenkei, man and tiger, Mongkonyu sanga yaksa chei DJLUOZDVVDFUL¿FHGGXULQJDÀRRGDW teilakukei, Mongkonyu, jümtei kho Mongkonyu sanga hüpkei. salt-water was found in Mongkonyu.

Mongkonyu sanga yembu longei ko The dead resurrected from kekei, Mongkonyu, Nakshou naknyu lotkei, Mongkonyu was surrounded by darkness,

ORIGIN AND MIGRATION MYTHS 227 müt pakbu chungnyu Mongkonyu the men-bearing tree was cut down by kabu müt shongei a kan apkei, the people of Mongkonyu head-taking chang hongto wapyukei, took place with the eastern foreigners, Mongkonyu sangalang At Mongkonyu a mortar was made shemlakumkei, of rock, oningei tei lenkei. the Cicada showed drinking water. (Kaiser 2007, Chang 2005:3-5)

The elder singers in Tuensang recount that there are some ‘youngsters’ who think that Mongkonyu is the same village as Mongkhuma, a place in Mongolia. They had even showed them the village-name on a Mongolian map. But in reality, the old men say, Mongkonyu is a patch of forest some kilometres away from Hakchang village; one can go there by foot in three hours and it is situated in the very heart of Chang territory itself. So in their eyes there is no proof in the many songs about Mongkonyu that the Nagas came from Mongolia. Usually the origin myths of Chungliyimti or Mongkonyu (as situated in Chang territory) are not instrumentalised for nationalistic discourse, but another origin place is referred to every now and then in public, for example in the daily newspapers. In a recent issue of the Nagaland Post, one can read: This tree is known as the oldest tree in the history of the Nagas. It is said that once all the Nagas lived at Makhel. But when [the] population increase[d], all people gathered at the foot of this tree and departed to different directions for new settlements. This tree still stands as a symbol of unity and oneness of the whole Naga tribes. When a branch of the tree is broken, all the Nagas observe genna7 for one day. (Nagaland Post, Sunday Post, April 15, 2007:1) The place of Makhel (also written Maikel) mentioned in this passage is one of the locations to which many tribes trace back not their actual origin, but the point in history and geography from where they dispersed over the Naga Hills. From Makhel a whole group migrated to the village of Khezhakenoma. One narrative relates that there was a magic stone ZKLFKGRXEOHGWKHJUDLQODLGRQWRLW%XWWKHVWRQHZDVEURNHQLQD¿JKW and lost its magic power, which caused the split of the tribes. Lotha (2008) avoids naming these tribes, knowing that the perception of the different tribes as to which other groups split from there might differ in their respective accounts. Mills (1922:4) and Hutton (1921a:6-7) suggest

7 Genna is a form of a ‘taboo’. In earlier times it included the closing off of a village, rituals, and abstinence from certain foods and sexual intercourse. Today in its original form it no longer exists, but the notion of a special day is still attached to it in many regions. 228 MARION WETTSTEIN

Figure 10.2. One of the trees near Makhel said to be the origin of the southern Naga groups (photograph by Alban von Stockhausen 2004). ORIGIN AND MIGRATION MYTHS 229

Figure 10.3. Detail of a tree near Makhel said to be the origin of the southern Naga groups (photograph by Alban von Stockhausen, 2004). that at least the Angami (which in their time included sections of the Chakhesang), Rengma and Lotha are clearly linked to Khezhekenoma and maybe also part of the Sumi (Sema). But many other Naga tribes do not recognise these places as either their point of origin or a station on their migration route. So the statement that ‘once all Nagas lived at Makhel’ is highly controversial. Among the Nagas there have been many origin stories of the kind recounted above.8 They are usually very detailed and exact until they reach the starting point of the migration somewhere in an earth hole or cave, in a stone or on a mountain peak, features which can be located inside today’s Naga territory or not far distant from it. Hutton attempted to link as many of these detailed accounts as possible and plotted them on DPLJUDWLRQPDS ¿JXUH LQKLVPRQRJUDSKDERXWWKH$QJDPL1DJD The map shows some continuous movements but also the fact that there is not one single origin place for all the tribes which can be deduced from the oral tradition alone. This fact is also discussed at length by B.B. Kumar (2005:54-66), who acknowledges that there might be some oral traditions reaching fairly far into Burma or even Thailand, but who clearly concludes “that not a single legend tradition of origin and

8 See, for examples, those recorded in the monographs by Mills 1922, Hutton 1921a, or Kumar 2005:54-66. 230 MARION WETTSTEIN

Figure 10.4. J.H. Hutton’s Migrations of Naga Tribes map (source: The Angami Nagas 1921:opp. p.7). ORIGIN AND MIGRATION MYTHS 231 migration is acceptable to or has currency among all the Naga tribes. In many cases, a single tribe has more than one legend of origin and tradition of migration” (Kumar 2005:65). B.B. Kumar was the Principal of Science College in Kohima, the capital of Nagaland. His work is highly esteemed locally, and perhaps this is one reason why he can afford to clearly advance a critique concerning topics about which the fear of possible reaction from certain political forces hinders others. With the publication of Kumar’s work, the hope shining through in earlier local writings, such as that by Chasie (2005 [1999]:28), that when thoroughly researched the Naga migration stories will provide the basis for a common origin, slowly fades.

INDEPENDENT SINCE TIME IMMEMORIAL

The ‘precise’ type of migration myths discussed above show that a common origin of the Nagas cannot be deduced from oral tradition. However, they would be more than suitable to support the third important statement in the rhetoric of the Naga identity discourse: that the Nagas demand nothing other than the land on which they lived since time immemorial. Nearly every Naga tribe or subgroup would surely be able to produce a migration or origin story showing that as far as their memory goes back they lived roughly on the land where they are settled now. Of course, some stories may include encounters with peoples who lived on the same terrain previously. But these people may have left and found a homeland somewhere else, or they may have been integrated into the approaching ‘tribe’. A narrative example for such a process can be seen in the case of the Ao. A version of it may be read in Mills (1926:8-11), and it is still related today in many variations: When the Ao reached what is today’s Mokokchung district after crossing the Dikhu River, they encountered two groups of people already living there, the Molungr and the Nokrangr. With the Molungr they established peaceful agreements; they bought the land to settle from them and largely integrated them into their clan-system. In some villages of the Asetkong KLOOUDQJHRQHFDQVWLOOWRGD\¿QGFODQVRULJLQDWLQJIURPWKHP:LWKWKH Nokrangr, however, the encounter was rather harsh. A war between the expanding villagers of Longkum and the Nokrangr people drove the latter away in all directions. Mills suggests (1926:11) that a section of Nokrangr merged with the people living in today’s Changki, an Ao vil- ODJHRIVSHFLDOSRVLWLRQLQWKHWULEHEHFDXVHLWLVRQHRIWKHIHZRI¿FLDOO\ FODLPLQJQRWWRKDYHRULJLQDWHGIURP&KXQJOL\LPWL$QRWKHUJURXSÀHG 232 MARION WETTSTEIN north and turned east again to merge with the Phom and Konyak. And a third group even crossed the Bramahputra valley to enter the hills on the north side in today’s Arunachal Pradesh and became a part of the VRFDOOHG'DÀD 0LOOV All in all, precise origin myths might be able to stress the claim on WKHWHUULWRU\ZLWKDOOMXVWL¿FDWLRQ,QWKLVUHVSHFWWKH\ZRXOG¿WSHUIHFWO\ into the discourse. However, they are not often used and the question is, why? For one, these stories carry a high risk of factionalism. The tendency of tribalism is already quite strong in Naga society and the breach between the two opposed underground armies mostly runs along tribal lines. We might acknowledge that the exponents of the discourse are sensitive enough not to support this tendency any further. On the other hand, some of the stories are indeed used to deduce a (non-existing) common origin of the Nagas, as the Makhel case shows. Thus, we are not dealing with a question of sensitivity towards possible tribalism but with a question of power: the group fostering its own origin as that of all Nagas is symbolically the group in power. There is a general problem with both types of migration stories, the vague Mongolian provenance and the precise origin tales, within today’s Naga territory: they often apply to many other neighbouring tribal populations as well. Therefore there is no real point in claiming the provenance from the East as being exclusively Naga. How then, for example, to explain to the outside world and the Indian Government why the Kukis do not belong to the ‘Naga’ tribes on that ground? This is an idea, by the way, which the Kukis themselves would also reject at all costs, even if, until the 1990s (i.e., until the Naga-Kuki war), the Kukis were counted as a ‘Naga tribe’ (Lotha 2007:3). The claim that the Nagas lived in situ ‘since time immemorial’ and were always independent is a recurring argument in Naga identity discourse. As a prominent exponent of this argumentation, we can take as an example Kaka D. Iralu, a local free-lance political journalist who stresses the point at length in his most popular book Nagaland and India: The Blood and the Tears (2000:3 ff.). Rather than offering the precise origin myths, he uses a set of stories which may be situated between historical account (for the sources) and myth (for their interpretation). Like many ORFDODXWKRUV,UDOXWDNHVIRUJUDQWHGWKDWWKH¿UVWHYHUPHQWLRQRIWKH Nagas in written sources is found in Ptolemy’s geography of 150 CE, the original reference likely taken from a footnote by Hutton (in Mills 1926:1). Ptolemy mentions the people of the Nangalothae (‘the naked’) at a geographical position quite possibly near the present-day Naga ORIGIN AND MIGRATION MYTHS 233 territory (cf. also Stevenson 1932). In local Naga discourse, this date is often taken as an unquestioned reference for the antiquity of the Nagas living where they are today. However, it is now not possible to know exactly which people Ptolemy meant when using this term. ‘The naked’ might refer to any regional group, even the peoples of Assam, depending upon how one wants to interpret the account. The next written mentions of the Nagas are said to occur in the accounts of a Chinese traveller of the seventh century and thereafter mostly in the chronicles of the Ahom kings, the Ahom Buranji, commencing from the thirteenth century. These are the three dated sources most often cited when the question of the antiquity of the Nagas turns up. And rather than oral tradition it is these accounts which are taken as evidence that the Nagas have lived where they are since ‘time immemorial’. To prove that the Nagas had always been independent and were never under any foreign control up to British times, the chronicles of the kings of Manipur and of the Ahom are referred to by Naga commentators. However, if we take the case of the Ahom Buranji, one soon sees that the relation between the Nagas— more precisely the Ao and Konyak—and the Ahom had not always been as peaceful as Kumar would like us to believe (2005:20). In the Ahom Buranji, one can read of wars, of slavery and of the slaughter of whole Naga villages conducted by the Ahom. The Ahom had no interest in the hills in terms of conquering them. But from the Ahom perspective, Nagas border villages were subject to them, and there were times when they had to pay tribute (Barua 1985). From the point of view of the Nagas, the Ahom were invaders coming long after the Nagas lived in the territory. There are indeed oral accounts among the Ao and Konyak to describe WKH ¿UVW PHHWLQJV ZLWK WKH$KRP DQG WKH UHODWLRQV WKH\ KDG WRJHWKHU In these accounts, one can even hear of marriage relations between an Ahom Kings and a Naga princess. Here the Nagas would have a point in claiming equality with the Ahom their longterm presence in the region accounting to written and oral sources alike. The nationalistic discourse, however, preferably focuses much further back in time, on Ptolemy, even though it cannot be taken for granted that Ptolemy’s Nangas referred in any way to today’s Nagas (cf. Stockhausen 2009).

AN ALTERNATIVE MYTH-MODEL

As a conclusion, one can state that the discourse of independence and collective identity among the Nagas does indeed strategically employ myths of origin and migration. The precise and detailed myths of oral 234 MARION WETTSTEIN

WUDGLWLRQZKLFKDUHVXQJDURXQGWKHKHDUWK¿UHDQGDUHFRQVLGHUHGWKH mythological heritage of the Nagas, would be a brilliant argument for the claim that the Nagas demand nothing other than the land they have inhabited since time immemorial. But they are not so suitable as an argument for the unity of Nagas because they tell of very diverse origins. Thus, they carry a high risk of tribalism and factionalism, which is most XQZHOFRPHLQDSURFHVVRIXQL¿FDWLRQ,WVHHPVWKDWDPRQJWKH1DJDV the notion of a collective identity is still in the making and the assertion that the Nagas are one people has priority at this point in time. For this process another type of migration mythology is much more suitable: vague tales of an origin in Mongolia, or somewhere to the east. The advantage of these tales is precisely their vagueness, as they offer a possibility for everybody to identify with little or no contradiction, such as in the case of the Ao we examined. The notion that the Nagas are one people seems to be a crucial argument for the claim of independence and statehood, and thus Naga unity must EHDFFRPSOLVKHGE\DOOPHDQV2IFRXUVHWKHIDFWLRQDO¿JKWLQJEHWZHHQ the armed underground groups works to the very contrary of this endeav- our. But if we consider for a moment the nation-building myths of other countries, such as the United States of America or Switzerland, we will ¿QGWKDWWKHQRWLRQRIEHLQJRQHSHRSOHLVQRWPDQGDWRU\IRUWKHFODLPRI being a nation. In the case of Switzerland, the story runs that three men from three villages came together on a meadow called ‘the Rütli’ and made a vow that they would stand together against enemies and would not attack each other under any circumstances. By and by, other regional groups—including those with different languages—joined them and thus ‘Switzerland’ was created. The story of the original Rütli-vow is as unproven and moot as a Naga migration route from Mongolia, and even its veracity is contested by historians (Sablonier 2008). But this does not change the fact that the Swiss nation-building myth is ‘federalistic’ and not a myth of ‘oneness’. Why do the Nagas insist on the idea of oneness based upon migration myths rather than the idea of federalism? If we consider once more that in earlier times the Nagas lived in self-sustained village-republics with bilateral relations and contracts amongst each other, we can speculate that in theory the emphasis on a federalistic model for the Naga myths of nation-building would have been a possible alternative. I think, however, in praxis we are dealing with an aspect of the colonial legacy. The Nagas, perhaps like many other groups aspirating to their own state following the end of colonial rule, were directed towards a myth-model of oneness. Most indigenous peoples in the colonies were ORIGIN AND MIGRATION MYTHS 235 generally grouped and termed as ‘tribes’ by their colonisers, and this was also true for the Nagas—in spite of the warnings of some colonial commentators, such as Thomas Callan Hodson (1911:81). The classical colonial notion of a tribe includes a common origin, kinship and territory, sometimes also a common language (Gingrich 2001). The concept of ‘tribe’ was institutionalised especially in postcolonial Indian administration in order to develop so called ‘backward’ peoples. According WR 3DQGD WKH GH¿QLWLRQ RI µWULEH¶ DFFHSWHG PRVW ZLGHO\ LQ Indian context is offered by Majumdar and Madar (1980[1956]:241), who, based on colonial concepts of the term, emphasise that ‘[a] tribe is (...) above all, conscious of a homogeneity of ethnic and territorial integration’. With such notion of oneness in the back of one’s mind, an alternative, federalistic model for nation-building myths among the Naga ‘tribes’ became unlikely. Thus, we can observe that, among local peoples like the Nagas, their traditional myths of origin and migration, and especially their use in the rhetoric of independence and a collective LGHQWLW\DUHKHDYLO\LQÀXHQFHGE\FRORQLDOKLVWRU\ 236 MARION WETTSTEIN

BIBLIOGRAPHY

Anonymous. 2007. A Trip back to the Ancestral Land. Nagaland Post, Sunday Post, April 15: 1. Arya, A. and J. Vibha. 2004. The Land of the Nagas. Ahmedabad: Mapin. Barua, G.C. 1985 [1930]. Ahom Buranji: From the Earliest time to the end of Ahom Rule. Guwhati: Spectrum Publications. Census of India. 2001. http://censusindia.gov.in/ (accessed 08.10.2009) Chang, C. 2005. Mongkonyu Nguhli / Mongko Nyu Chia [History of Mongkonyu / Folk Songs of Mongkonyu]. Tuensang (Private Edition). Chasie, C. 2005 [1999]. The Naga Imbroglio: A Personal Perspective. Kohima: Standard Printers and Publishers. van Driem, G. 2008. “The Naga Language Groups within the Tibeto-Burman Language Family.” In M. Oppitz et al. (eds.), Naga Identities: Changing Local Cultures in the Northeast of India. Gent: Snoeck Publishers, pp.311-22. von Fürer-Haimendorf, C. 1939. The Naked Nagas: Head Hunters in Assam in Peace and War. London: Methuen & Co. Gingrich, A. 2001. “Tribe.” In N.J. Smelser and P. Baltes (eds.), International Encyclopaedia of the Social & Behavioural Sciences. Oxford: Pergamon, pp.15906-09. Hodson, T.C. 1911. The Naga Tribes of Manipur. London: Macmillan. Hutton, J.H. 1921a. The Angami Nagas. London: Macmillan. —— 1921b. The Sema Nagas. London: Macmillan. Iralu, K.D. 2003 [2000]. Nagaland and India: The Blood and the Tears. A historical Account of the Fifty-Two Year Indo-Naga War and the Story of those who were never allowed to tell it. Kohima (Private Edition). IWGIA Document 56. 1986. The Naga Nation and its struggle against Genocide. Copenhagen: IWGIA. Jacobs, J., A. Macfarlane, S. Harrison and A. Herle. 1990. The Nagas. Hill Peoples of Northeast India: Society, Culture and the Colonial Encounter. London: Thames and Hudson. Jamir, N. Talitemjen and A. Lanunungsang. 2005. Naga Society and Culture: A Case Study of the Ao Naga Society and Culture. Kohima: Nagaland University: Tribal Research Centre. Jamir, T. and D. Vasa. 2008. “Archaeology of Local Cultures: New Findings and Interpretations in Nagaland.” In M. Oppitz et al. (eds.), Naga Identities: Changing Local Cultures in the Northeast of India. Gent: Snoeck Publishers, pp.323-40. Kaiser, T. 2004. The Naga Music Archive. Zürich. (unpublished Archive of the Project ‘Material Culture, Oral Traditions and Identity among the Naga in Northeast India’ funded by the Swiss National Science Foundation. Archive Number <040430_ao_ningsashiba_lon>. Sung and translated by Ningsashiba, Longjang). ORIGIN AND MIGRATION MYTHS 237

—— 2007. The Naga Music Archive. Zürich. (unpublished Archive of the Project ‘Material Culture, Oral Traditions and Identity among the Naga in Northeast India’ funded by the Swiss National Science Foundation. Archive Numbers: <071219_chang_tuensang_51> and <071219_chang_tuensang_78>. Translated by T. Impongsoted, Tuensang). Kejong, C. 2008. “Interview.” In M. Oppitz et al. (eds.), Naga Identities: Changing Local Cultures in the Northeast of India. Gent: Snoeck Publishers, pp.107-10. Kumar, B.B. 2005. Naga Identity. New Delhi: Concept Publishing Company. Kunz, R. and V. Joshi (eds.). 2008. Naga: A Forgotten Mountain Region Rediscovered. Basel: Museum der Kulturen Basel / Christoph Merian Verlag. Longchar, P. 2002. Historical Development of the Ao Nagas in Nagaland. Dimapur: Print Home. Lotha, A. 2007. History of Naga Anthropology (1832-1947). Dimapur: Chumpo Museum Publication. —— 2008. “Naga Identity Enduring Heritage”. In M. Oppitz et al. (eds.), Naga Identities: Changing Local Cultures in the Northeast of India. Gent: Snoeck Publishers, pp.47-56. Majumdar, D.N. and T.N. Madan. 1980 [1956]. An Introduction to Social Anthropology. Calcutta: Asia Publishing House. Mills, J.P. 1922. The Lhota Nagas. London: Macmillan. —— 1926. The Ao Nagas. London: Macmillan. —— 1937. The Rengma Nagas. London: Macmillan. Nuh, V.K. 1986. Nagaland Church and Politics. Kohima: V. Nuh & Bro. —— 1996a. Crusade on Naga Morality. Kohima: Council of Naga Baptist Churches. —— 1996b. $7KHRORJLFDO5HÀHFWLRQRQ1DJD6RFLHW\. Kohima: Council of Naga Baptist Churches. —— 1998. Revive Us Again. Calcutta: Research Department, Council of Naga Baptist Churches. —— 2001. Struggle for Identity in North-East India. Guwahati: Spectrum Publications. —— 2002. My Native Country: The land of the Nagas. Delhi & Guwahati: Spectrum Publications for Research Department, Council of Naga Baptist Churches. —— 2006. 165 Years History of Naga Baptist Churches. Kohima: MEK Computers for Research Department, Council of Naga Baptist Churches. Nuh, V.K. and L. Wetshokhrolo. 2002. The Naga Chronicle. New Delhi: Regency Publications. Oppitz, M., T. Kaiser, A. von Stockhausen and M. Wettstein. 2008. Naga Identities: Changing Local Cultures in the Northeast of India. Gent: Snoeck Publishers. Panda, N.K. 2006. Policies, Programmes and Strategies for Tribal Development. Delhi: Kalpaz Publications. Prakash, Col. V. 2007. Encyclopedia of North-East India, vol. 5. New Delhi: Atlantic Publishers and Distributors. Sablonier, R. 2008. Politik und Gesellschaft in der Innerschweiz um 1300. Baden: Hier und Jetzt. Stevenson, E.L. 1932. Geography of Claudius Ptolemy. New York: New York Public Library. 238 MARION WETTSTEIN

Stirn, A. and P. van Ham. 2000. The Seven Sisters of India: Tribal Worlds Between Tibet and Burma. Munich: Prestel. —— 2003. The hidden world of the Naga: living traditions in Northeast India and Burma. Munich: Prestel. von Stockhausen, A. 2008. “Creating Naga: Identity between Colonial Construction, Political Calcuation, and Religious Instrumentlisation.” In Michael Oppitz et al. (eds.), Naga Identities: Changing Local Cultures in the Northeast of India. Gent: Snoeck Publishers, pp.57-80. —— In press. Naga: Lineages of a term. Presented at the Conference Writing the Northeast: New Perspectives, CHS, Jawaharlal Nehru University, Delhi. January 14-16, 2009. Vero, C. 2005. “Hornbill Festival”. Morung Express. Saturday, December 3: 11. West, A. 1999. The Most Dangerous Legacy: the development of identity, power and marginality in the British transfer to India and the Nagas. Hull: The University of Hull, Centre for South-East Asian Studies. ORAL HISTORIES AND THE ‘ORIGINS’ OF CURRENT PEOPLES: DYNAMIC ETHNOGENESIS, WITH REMARKS UPON THE LIMITATIONS OF LANGUAGE-FAMILY SUBGROUPING

F. K. L. CHIT HLAING

ABSTRACT

I shall examine the current controversies in the theory of ethnic categories, focusing upon the lability of categorization in the context of changing inter-group relations. My main point will be to demonstrate that, in general at least, the farther back in time we go the less sensible it is that we can even ask where ‘this particular people’ was, even given migration evidence. I shall also go into the arguments about whether various upland groups really did, as their oral traditions claim, once have a kingdom of their own—examining the actual reasons for such persistent claims. Finally, I shall try to examine effects of these considerations upon evidence used in linguistic sub-grouping. My materials will come largely from work, much of it my own, on Chin, Kachin, and Kayah.

INTRODUCTION

What I want to do here is look at the matter of ethnic identity and supposed ethnic history from a particular point of view, namely, the argument that one cannot in general ask usefully where such and such a people in their current identity and under their present ethnonym were very far back in time. I shall do this simply by examining several cases from the Burma-China-India part of mainland Southeast Asia, not only to LOOXVWUDWHP\SRLQWLQJHQHUDOEXWDOVRIRUWKHSXUSRVHRIGUDZLQJVSHFL¿F conclusions bearing upon ethnicity and ethnogenesis. Almost all of what I shall use here comes form my own research except the Kachin/Jingpo material, the data for which is the result of my work with my one-time pupil and subsequent colleague, the Kachin anthropologist-linguist Dr. Maran LaRaw, and subsequently, the result of my work with my current &KLQHVHSXSLO=KDQJ:HQ\LDQGKLVH[WHQVLYH¿HOGZRUNDPRQJVWWKH Jingpo of southwestern Yunnan, China.1

1 See now Chit Hlaing 2007a. 240 F. K. L. CHIT HLAING

THE CASE OF THE KAYAH

I want to begin with the case of the Kayah (Red Karen/Karenni2). As I showed in my monograph on these people (Lehman 1967a), as a self- designated ‘people’, their existence goes back no further in time that the late eighteenth century. I need not recapitulate the details here, but because one of the themes here is the common dependence of ethnic identities upon inter-group relations and politics,3 I have to say something about the circumstances of this extraordinarily well-documented instance of ethnogenesis. The world shipbuilding demand for teak wood had risen sharply after the sixteenth century because of the presence of European shipping in the India Ocean-South China Sea region. It happens that the most exten- VLYHIRUHVWVZLWK¿UVWFODVVWHDNZHUHLQWKHPRXQWDLQVRQZKDWLVQRZ the Thai-Burma border within the drainage of the Salween River, down ZKLFKWKHORJVFRXOGEHÀRDWHGWRSRUWVRQWKH*XOIRI0DUWDEDQ6R the Shan of the Shan States of Burma, a system of trade-based polities as I have shown (Chit Hlaing 2007c, 2007d) with many Shan merchant agents in Burmese port cities such as Rangoon, began to exploit these forests in a big way. To this end they took an especial interest in the forests in what is now the Kayah/Karenni State in Burma, just south of the Shan area proper, and on the left bank of the Salween, partly in what is now Mae Hongsorn Province in Northwest Thailand, no doubt because here there are numerous usable streams feeding directly into the Salween. For this work they needed elephants, and so they employed elephant drivers as contractors from Central Karen communities in the region. This employment put such men in more direct contact with the larger cultural world of the Shan and of the Mon and Burman population of the lower Salween with whom they worked and amongst whom they had on occasion to sojourn, This in turn seems to have enabled those men to build upon the widespread Karen millenarian traditions4 by laying claim to mystical-supernatural powers and knowledge; they became what is called in Kayah phrei-phrow, that is, a person with mystical powers who appears magically from far away. At the same time,

2 Karenni is the anglicisation of the Burmese kayinni and, either way, means ‘Red Karen’. 7KLVUHÀHFWVWKHDXWRQ\PKayah-li. The word kayah cognate with ‘Karen’ and with, say, the Sgaw Karen word [pra]kanyo, is simply the word for ‘a person’, regardless of ethnicity (the SUH¿[praEHLQJMXVWWKHQXPHUDOFODVVL¿HUIRUKXPDQEHLQJVprei in Kayah); for the Kayah language, see now Solnit 1997. 3 See now Lehman 1967b and Chit Hlaing 2007a, 2007b. 4 For details and references see Chit Hlaing 2007b, 2007c. ORAL HISTORIES 241 though, they were increasingly involved in and knowledgeable about the teak trade. Using their knowledge and their claim to the status of phrei-phrow, they organised the overthrow of their Shan masters.5 They began managing the exploitation of the teak forests on their own behalf and they set themselves up as princes of small states modeled on Shan polities and given authority ideologically and ritually by a new cult enriched by a good deal of Buddhist symbolism recast in non-Buddhist, Karen-millenarian terms. Furthermore, at least in part, these new princes undertook a recurrent campaign of raiding and slave-trading, in order to expand their control of the teak trade and their presence as a power in the trading system more generally. This enabled them to control more fully the resources of the Salween drainage, and to establish external recognition and trade into the Southernmost Shan states and into the borderlands of the then Northern Thai (Yuan, Lanna Thai) territories along the Salween. As part of this campaigning system (annually, during the dry season), they adopted the custom of having their men wear characteristic short trousers in red as a marker of identity, and it was in this context that they began to adopt the identity and name Red Karen! Thus was born ‘a people’. Note in particular that there was no one-to-one connection of this ethnicity with language. For instance, there is not the slightest distinction between Western Kayah (the language of the state of Kyebogyi) and that of the people known as Manumanaw in Burmese, who live in the mountains on the west. The name comes from the Kayah phrei manoe (phanoe, in Eastern Kayah, meaning ‘west’), literally ‘people of the west’. Yet these latter do not lay claim to Kayah ethnicity nor do the Kayah claim them amongst themselves. .D\DKRULJLQP\WKVDOVRPDNHFOHDUWKHH[WHQWWRZKLFKWKH\GH¿QH themselves relatively to the civilisations of the lowlands. The myth I

5 Indeed, at least in the case of the largest Kayah State, Kyebogyi, where I worked in 1961, the need for Shan artisans for making ironwork and other metal-craft was met by literally enslaving Shan and keeping them in a settlement then known as phre lave, literally ‘forced settlement of Shan’, now known as Phruso (Shweidaung in Burmese). It is also worth noting here that the Shan term (itself at once a royal title and a titular reference to the Lord Buddha) cao pha or the cognate Northern Thai (Kham Mueang) term cao phraya, signifying ‘celestial lord’, was adopted as at once the title of Kayah rulers and the name of the supreme god in the new state cult, in the form of Kayah Sophrya. More accurately, ph(r)a is the Tai word for ‘sky’, but phraya in Northern Central Thai is an obscure Indic loanword (vara in Pali, vrahLQWKHLPPHGLDWH2OG.KPHUVRXUFH VLJQLI\LQJP\VWLFDOSRZHU7KHGLI¿FXOW\LV that either way, the Shan form comes down to pha, except that in some Shan dialects the meaning ‘celestial/sky’ surfaces as fa. This obscures the exact source of the Kayah Sophrya EHFDXVHHLWKHUWKHODVWV\OODEOHUHÀHFWVD7KDLVRXUFHphrayaRUUHÀHFWVDQ2OG6KDQVRXUFH as phra, with subsequent palatalisation of the medial ‘r’, as in Burmese. 242 F. K. L. CHIT HLAING have in mind, which I have set forth elsewhere recently (Chit Hlaing 2007b), features a sort of prototype Phrei Phrow. He had appeared suddenly some years before and, seeing a beautiful young women, has fallen in love with her, married her, and taken up residence with her and her mother. During those years he taught the villagers all sorts of things to ensure their well-being and prosperity. Eventually, however, he grew bored with having to do this and so he decided to leave. He ascends into the sky and they beg him to remain with them, saying so to his wife DQG PRWKHULQODZ LQ WKH YLOODJH RI WKH µ¿UVW .D\DK¶ DV KH DVFHQGV into the sky, so they can know how to live. ‘How will we know how to live without you?’, they ask. He responds by advising them to pray to God for knowledge, but they ask him how one prays to God, To this he answers, ‘The trouble with you is just that; you don’t even know how to pray to God. So, just set up a high pole and pray to that!’ That myth is intended to explain the origin of the state cult aforementioned, centred as it is on a ritual object, a tall pole erected annually and decorated in the form of the Burmese Buddhist tagundaingRUÀDJSROHRIWKHYLFWRU\RI Buddha over Mara (a model which Kayah systematically refuse to admit they can even recognise). What stands out here is the fact that, in effect, they are claiming that their own ancestors were, relatively to the wider world at least, hopelessly ignorant and incapable.6 From this we can see at least two things for purposes of the present discussion. First, the extent to which ethnic identity can be (I say generally is) a function of the politics and economics of inter-group relations. Second, that to ask, as one might, where ‘the ancestral Kayah’ were at some distant past time would be a misplaced question. In addition, let me merely point out that any attempt to place ‘the language of The Kayah’ in a subgrouping framework within Karennic overlaps entirely with subgrouping Manumanaw; the Kayah language simply does not branch neatly within a Stammbaum of Karennic within Tibeto-Burman.

THE CASE OF THE CHIN

7KLVVHFRQGFDVHZLOOEHEULHIHUIRULWFRQFHUQVFKLHÀ\WKHODEHOZo, the nearly universal autonym for this group of linguistically closely related

6 I have written elsewhere of this sort of myth/legend, altogether widespread amongst XSODQGSHRSOHVRI6RXWKHDVW$VLDZKHUHWKHµ¿UVWDQFHVWRU¶LVOHIWDWWKHPHUF\RIDµ¿UVW lowlander’ and indeed duped by the latter so that ‘we’ end up forever in our mountains without the amenities of civilisation—legends with the theme I have called ‘where was our ancestor when God gave out the brains?’ ORAL HISTORIES 243 peoples of the mountains between Burma and India (Lehman 1963). The word ‘Chin’, by the way, is from the Burmese pronunciation of what is in Written Burmese/Old Burmese khrang, and, as I have written elsewhere (Chit Hlaing 2007b) it is the Old Burmese attempt to transcribe the Old Kuki-Chin word for ‘a people’, which exists currently in Southern Chin languages (e.g., in k’Cho)7 as [khxang] written as k’Chang. Apparently the Burmans of about 1100 AD/CE may have imagined that it was an ethnonym, and it was used in stone inscriptions by the then king at Pagan, Kyanzitha, in the Chindwin drainage to record a possibly informal arrangement to leave these people alone beyond the west bank of the river. However, their proper auto-ethnonym is everywhere zo, or cognates such as aforementioned (K’)Cho [kxo]; it surfaces in all Northern and Central Chin languages as either Zo mi or (as in Lushai/Mizo) as Mi Zo with mi being the widespread Tibeto-Burman etymon for ‘person(s) or ‘one(s)’. What is important for our purposes is that the word clearly VLJQL¿HVVRPHWKLQJOLNHµFUXGHEDFNZDUGPDUJLQDO¶,KDYHVKRZQWKLV HOVHZKHUH HJ &KLW +OLDQJ F DQG ¿QG QR QHHG WR GHPRQVWUDWH this etymology here, save to point out that it is also used to describe VZLGGHQ ¿HOGV lo) higher up on the slopes that are less well cleared and cultivated, less productive and used only for a single year (zo lo as contrasted with more elaborately worked swiddens lower down, lai lo, lai = ‘central’). Discussions with knowledgeable Chin informants have usually produced additional support for this etymology. Nothing shows more clearly than this autonym the way these people identify themselves relatively to the dominant peoples of the lowlands. The latter are called vaai, signifying anything thought to be more elaborate than what they have themselves in the way of technical knowledge, production or amenities of life. For Chin Vaai mi are Kawl, Burmans; for Mizo/Lushai Vaai = Indians, or more narrowly Bengalis/Assamese. Again, as with the Kayah, to ask where ‘The Chin’—ancestral, of course—were prior to settling where they are now, say before about 1000 &($'LVUDWKHUDPLVOHDGLQJTXHVWLRQ&KLQDVVXFKZHUHQRWGH¿QHG prior to their apparent early association with the present lowlanders. This may seem problematical, in as much as it may well have been that they had similar relations with other lowlanders there or elsewhere, and comparative linguistic work shows that at some time prior to, perhaps, WKH¿UVWPLOOHQQLXP$'&(WKH\KDGWRKDYHEHHQOLNHPRVW7LEHWR Burman peoples except maybe for Pyu and Karen, somewhere north of Mainland Southeastern Asia. But more to the point is the altogether

7 See now van Bik 2006 and Mang 2006. 244 F. K. L. CHIT HLAING widespread variety of Chin origin legends, of which I shall say only a bit here. On the one hand, at least Northern and Central Chin and 0L]R/XVKDLKDYHRULJLQP\WKVVD\LQJWKDWWKHµ¿UVWDQFHVWRUV¶HPHUJHG IURPURFNVRUFDYHVLGHQWL¿HGTXLWHVSHFL¿FDOO\ZLWKORFDOSODFHV VHH Wettstein in this volume). On the other hand, they have legends that clearly and unambiguously serve as symbolic claims upon the amenities of the lowlands and plains adjacent. I need not and shall not recapitulate any of these legends—often in Southern Chin related in the form of songs. But they invariably say that of old, ‘before these Vaai pushed WKHP RXW RI WKH SODLQV¶ WKH\ OLYHG LQ VSHFL¿F FXUUHQW SODFHV LQ WKRVH YHU\SODLQV$QGWKH\DUHVXI¿FLHQWO\IDPLOLDUZLWKVXFKUHJLRQVWRKDYH migration legends detailing supposed journeys from one such actual place to another on the way into the present mountains. What this means is that they identify themselves quite explicitly with the present Vaai neighbourhoods. So, once again, we can see that as ‘The Chin’/Zo they relate strictly to their present8 larger world; and how their ancestors may KDYHLGHQWL¿HGWKHPVHOYHVHWKQLFDOO\LVGRXEWOHVVYHLOHGLQP\VWHU\,W has now to be clear that these sorts of myths and legends serve to identify the peoples of the uplands as dependents of the civilised lowlands and to lay symbolic claim to a right to the amenities of civilisation in IDFWGHQLHGWRWKHP%XWPRUHWRWKHSRLQWLVWKDWLWLVDOZD\VDVSHFL¿F civilisation: Burma-and-Assam for the Chin, Shan-and-Mon for the Kayah, China for Akha, Lahu and other Southern Yi speaking peoples, who, after all, came into Southeastern Asia from China only from the eighteenth century. And, as with the Southern Yi speaking peoples most particularly, there is the associated legend that, until they were ‘pushed out into the hills’, they did indeed have a civilisation of their own, or were at least given the things of civilisation by God, but robbed of it by their ancestral ‘elder brother’, the original Burman or whatever, as amongst the Chin. Thus once more we see that the historical upland ethnic entities in our region more or less invariably identify themselves QRW RQO\ UHODWLYHO\ WR ORZODQG FLYLOLVDWLRQ JHQHUDOO\ EXW WR D VSHFL¿F hegemonic lowland civilisation.

8,FDQQRWVD\WKDWWKH\GH¿QHWKHPVHOYHVin situ. For Chin, Kayah, Karen and so on, this maybe almost the case, but for Akha and other Southern Yi uplanders (and Mien and Hmong, though not Tibeto-Burman-speakers) who have in fact recent and well documented histories of migration, the prototype civilisation is always where they started from, China as the arch hegemon; in this connection, on Akha see now Althing von Geusau 1982—who seemed WRWDNHOLWHUDOO\OHJHQGVRIDIRUPHUNLQJGRPRQ0LHQ

ON THE KACHIN

As is well understood, the Kachin9 are at once an ethnically composite ‘people’ and a political system dominated by the unitary ethnic people, the Jinghpaw/Jingpo10, and it is this puzzle of ethnicity I want to deal with here. The languages of the various included peoples, are all Tibeto- Burman, but of rather different branches of this family. Of Jinghpaw itself, let me postpone consideration of its subgrouping for now. Other Kachin are Maru, a language closely related to Burmese within the Yi/ Lolo-Burmese branch, Atsi, also Burmish, and others that won’t concern me here, such as Nung-Rawang—provisionally close to Jinghpaw according to Matisoff 2003, but see now, on ethnological grounds, Gros 2004—and even one variety of Lisu, of the Southern Yi/Lolo branch of Yi-Burmese. What is important here is that, for instance, Maru and Atsi indeed consider themselves to be Kachin in some ethnic sense; their cosmology, traditional religion, and so on, as well as their system of politics are shared with the Jinghpaw and the ritual language of all this is a somewhat archaic form of the Jinghpaw language. So, to this extent it can be said that they are ethnically Kachin, though not Jinghpaw (of course Atsi DQG0DUXRQWKH&KLQDVLGHDUH¶-LQJSR¶RI¿FLDOO\ $IWHUDOOFXUUHQW theory of the relation between ethnicity and culture recognises that ‘a culture’ is what any set of people attributes to itself as their way-of- life without regard to any objective uniformity of custom and practice, variation being taken to be ‘mere’ variation on a common core. So, we can say that Maru and Atsi are indeed culturally Kachin, and they also talk this way, both to ethnographic researchers and to Kachins in the modern rhetoric of ethnic politics. None the less, when one probes further with Atsi or Maru they will say that they are in actuality Zaiwa and Lawngwaw, respectively. More precisely, to be Maru or Atsi is to be a ‘kind of’ Kachin, if only for the fact that these two labels are Jinghpaw words and represent how Jinghpaw speakers proper talk about these other Kachin. But Zaiwa and 9 For details, and the foundation of all my remarks here, see Chit Hlaing 2007c and the numerous references cited there. 10:KDW LV FDOOHG .DFKLQ LQ %XUPD LV FDOOHG -LQJSR RI¿FLDOO\ LQ &KLQD -LQJSR LV WKH Putonghua Pinyn transcription of what otherwise is written Jinghpaw. This word is the name of the dominant ethnic and language component of the complex. Kachin is a word based on the Burmese name for these people(s), which is taken, perhaps out of ignorance of the Jinghpaw language, from the Jinghpaw name of the heartland of traditional political chieftaincy, ga khyen, ‘the Red [Earth] Land’, roughly the region of the Hukawng Valley and the Triangle. 246 F. K. L. CHIT HLAING

Lawngwaw are their respective autonyms in their own home languages. And so it turns out in effect that being Zaiwa or being Lawngwaw is not being ‘a kind of Kachin’! Discussions with knowledgeable members of WKHVHJURXSVFRQ¿UPWKLVSURSRVLWLRQ In the matter of Jinghpaw proper, however, things are more complicated, having regard to the question of linguistic subgrouping within Tibeto-Burman. It is well known, see now Matisoff (2003:3-5), that the subgrouping of Jinghpaw itself has been recurrently problematical. Benedict (1972) puts Kachin at the centre of a sort of star-shaped subgrouping diagram of Tibeto-Burman on the grounds that it seems to show shared innovations with all sorts of other branches almost indis- criminately. Others have stressed its shared innovations with Bodo-Garo within a Northern Naga branch, and Matisoff now provisionally classes it within his Jinghpo-Nungish-Luish ‘branch’, which seems a catch- all ‘other’ or residual branch apart from the more clear-cut branches North Assam/Tani, Himalayish including Bodish/Tibetan, Qiangic, Baic Karennic, and Lolo/Yi.

CONCLUSION

,ZLVK¿UVWWRGUDZDFHUWDLQWHQWDWLYHFRQFOXVLRQIURPWKLV-LQJKSDZ SX]]OH7KHOLQJXLVWLFVLWXDWLRQLHWKHYHU\GH¿QLWLRQRIµWKHODQJXDJH¶ is not unrelated to the ethnic one, namely, that it is partly a function of current inter-group relations; indeed, this is not unlike Benedict’s point about Kachin. Second, more particularly with regard to Jinghpaw, the language is, as a sort of classical/ritual language, the property so to say of the whole Kachin complex. On the one hand it is a hegemonic language for the complex; but, on the other hand, it is inconceivable that this situation, which has a history (cf. Wang 1997:54ff.; on the relevant history of the Tai/Shan/Ahom, cf. Fernquest n.d.) reaching back well into the fourteenth century, has not involved a great deal of persistent bilingualism amongst communities of different Tibeto-Burman branches. This cannot KHOSEXWKDYHDIIHFWHGWKH-LQJKSDZODQJXDJHGHHSO\$QG¿QDOO\KDYLQJ regard to origins and migrations, the Jinghpaw11 who seem to have come along as a client people with the Ahom Tai in their fourteenth century movement from Yunnan (from the old Tai state of Meng Mao) onward to North Eastern India cannot be said to be ‘the same people’ as the histori- FDO-LQJKSDZLQDVPXFKDVWKH\ZHUHQRWEDFNWKHQ.DFKLQQRWGH¿QHG 11 Note that the word Jinghpaw now takes the form Singhpo amongst those communities that still live in North East India. ORAL HISTORIES 247 as the ethnic centre of an ethnic complex and political order. They were only the ancestors of today’s Jinghpaw. Thus, taking all the above cases into consideration, we see that ‘the language’ of ‘a people’ has in great PHDVXUH WR EH D IXQFWLRQ RI WKDW SHRSOH¶V GH¿QLWLRQin situ within its system of inter-group relations, itself a product of particular historical circumstances, bearing on our understanding of the nature of what is reconstructed by historical-comparative linguistics. In fact we simply WHQGWRDFWZULWHDVLIµD¶SHRSOH E\WKHLURZQRUE\DQ\RQH¶VGH¿QLWLRQ is bound to have a distinct language of its own. However, in reality, there simply is no one-to-one relation between an ethnic category, however GH¿QHGDQGDVHSDUDWHODQJXDJH%XWWKDWKDVWREHOHIWWRDQRWKHUSDSHU because, on technical linguistic grounds, there is no purely linguistic (as against socio-political) meaning to the distinction between ‘language’ and ‘dialect’; as someone or other once said, a dialect is a language without its own army. In fact, again technically, the number of languages is equal to (a multiple of) the number of speakers of human languages alive at any moment, and what, by useful convention, we generally call ‘the such-and-such language’ is simply that collection of effectually mutually intelligible individual languages of any set of communities of speakers that consider themselves to be speaking ‘the same language’.12 Consider again what I said about Kayah. There is no difference of VLJQL¿FDQFH EHWZHHQ WKH VSHHFK RI :HVWHUQ .D\DK DQG WKDW RI WKHLU Manumanaw neighbours. And yet we speak easily about ‘the Kayah language’ because the Kayah people are in some historical sense prominent and possessed of a state, whilst the Manumanaw are a ‘hill tribe’. The import of my remarks on the Jinghpaw language is more complicated because, what is at issue in that case is not what makes it ‘a language’ of ‘a people’ but rather, owing to its apparent creolisation, how it can be assigned to a distinct branch of Tibeto-Burman on the grounds that it is basically the direct descendent of what we might call a proto-Jinghpaw existing historically prior to that apparent creolisation and thus belonging to a distinct branch. But these are very technical linguistic considerations that cannot be pursued in the present paper. Let me turn secondly to the general matter of the way in which ethnic 12,QDFWXDOLW\µDODQJXDJH¶LVGH¿QHGE\WKHPHQWDOJUDPPDUWKDWJRYHUQV GH¿QHVDV well-formed) actual speech, and it is trivially well known that even in a single community almost any two individuals differ as to whether, say, a given syntactic construction is or is not acceptable. If one needs evidence for the more conventional idea of what constitutes ‘a language’, one need only consider, say, how it is that English people, Americans, Australians, Lowland Scots, and even many people in India all speak English. For after all there are all sorts of differences between, say, Indian English, Standard British English, thick outback Australian and so on. Indeed, they are only ‘effectually’ mutually intelligible. 248 F. K. L. CHIT HLAING identity is a function of complex considerations (political, economic, HFRORJLFDO DERXWLQWHUJURXSUHODWLRQV,PXVW¿UVWFODULI\WKHVHQVHLQ ZKLFK XSODQG µWULEDO¶ SHRSOHV GH¿QH WKHPVHOYHV DV GHSHQGHQW XSRQ ORZODQG VWDWHV &OHDUO\ WKH ORZODQG VWDWHV GH¿QH XSODQGHUV HLWKHU DV subordinate, or else, as the Burmese kingdoms treated most of the Chin traditionally, as hardly worth dealing with. And it is also obvious that the upland peoples have often quite ambivalent relationships with the ORZODQGVWDWHV)RUUDWKHUWKDQGH¿QLQJWKHPVHOYHVDVsimply dependent, they commonly understand themselves to be treated as dependent but construct narratives and legends to account for this and to make claims on the lowland amenities. These legends frequently assert that they were originally given such amenities but lost them for all sorts of reasons.13 Other legends claim even that they formerly themselves had a state but were pushed off into the uplands; and, of course, considering how many upland peoples have such legends about former states, and considering the symbolic function of such tales as justifying a claim to lowland amenities, it is impossible to take any of such claims as actual history. In any event, they remain ambivalent about themselves in the sense that, again as with the Chin, they do not at all look kindly upon their lowland KHJHPRQVDQGRIWHQ¿QGWKHLURZQXSODQGHQYLURQPHQWPRUHEHDXWLIXO and pleasant than the hotter, more fever-ridden lowlands.14 For example, Karen peoples are especially well-known for considering themselves the subjects of endless oppression by lowlanders. And certainly it is not uncommon for upland communities to accept an identity imposed upon them by a powerful lowland state, especially where there is a legal or political advantage to be gained from doing so, as in the case of the Yi identity imposed by the People’s Republic of China on the collection of peoples formerly called Lolo.15 However, it must not be imagined that ethnic identity of uplanders is determined solely in this way, for an important factor is also relations amongst upland groups themselves. For these considerations one must see my paper on Chin identity (Chit Hlaing 2008).

13 See my earlier remarks about myths of silly or incompetent primordial ancestors (‘where were out ancestors when God gave out the brains?’), as in both the Chin and Kayah cases. These tales are far from uniform in their particular content of course. For instance, while the Chin, like so many tribal peoples over the whole region, say they had writing but it was lost because their ancestor ate the animal hide on which it was given, the Kayah do not say they ever were given writing. 14 There are numerous Chin songs and poems that praise the beauties of the mountain country, whilst the very word kawl for the Burmans (and the Burman state) is a pejorative epithet. 15 I have discussed this case in Chit Hlaing 2007. ORAL HISTORIES 249

I want to close this paper by making brief reference to a theme that is found in more than one of the other papers in this volume. I mean the rituals for sending off the spirits of the dead. There are two sorts of systems of sending away the dead who have died a ‘good death’. One is emphasised in several essays here especially concerning peoples in the Tani region of Arunachal Pradesh, but also found in much of upland Southwestern China and amongst, for instance, the Akha people of China and Northern mainland Southeast Asia proper. In this version the dead go ‘back’ to a supposed earthly point of origin of the ethnic people or of a lineage or clan. In this version and its associated chants and myths, the implicit theme is a claim on a former more important political order and a claim, likewise, upon the whole region through which the spirit is made to pass. And in this case we see it as a sort of reversal of a supposed migration history that is at once a migration in the ordinary geographical sense and a migration, so to say, from a former better situation to their present socio-political condition. The other version sends the dead directly ‘up’ to the land of the dead or to the part of it where the particular community’s ancestors are. This correlates with the idea that there is an asymmetrical mutual dependency between the two ‘worlds’. Sending ‘our’ dead to ‘our’ land of ancestors gives them the power/ prosperity to ensure ‘our’ welfare/prosperity; just as does our sending the spirits of slain heads of animals and enemies at Feasts-of-Merit-and- Celebration! Keep it in mind that it is such a Feast that turns a bad-death spirit/soul into something tractable, which would otherwise be a virulent wandering ghost. In both versions of this myth and ritual of sending away the spirits of the dead the purpose is the same, namely, to link up the living community as a collectivity, via a continuing connection with ancestors in ‘a better place’, as a means of tapping into sources of well- being and prosperity. The two versions are not always or necessarily mutually exclusive; most if not all Northern and Central Chin groups, for instance, have ‘origin myths’, claiming to have arisen from some local rock, cave or whatever, whilst some Southern Chin recite migra- WLRQ WDOHV ZLWK RULJLQV LQ WKH %XUPD ORZODQGV 7KLV FHUWDLQO\ UHÀHFWV the fact that Northern Chin historically were less immediately adjacent to Burman civilisation than were those of the South (k’Cho, Dai and VRRQ VRWKDWIRUWKHODWWHU%XUPDQFLYLOLVDWLRQDQGDFWXDOVSHFL¿HG Burman places of note mentioned in these songs and recitations were an immediate presence, which was never the case for the Northern and Central Chin. I discuss this in my Chin monograph (Lehman 1963). 250 F. K. L. CHIT HLAING

BIBLIOGRAPHY

Althing von Gesau, L. 1982. “The dialectics of Akhazang: Interiorization of a minority group.” In J. McKinnon and W. Bhruksasri (eds.), The Highlanders of Thailand. Kuala Lumpur: Oxford University Press, pp.241-278. Benedict, P.K. 1972. Sino-Tibetan: A Conspectus. Cambridge and New York: Cambridge University Press. van Bik, K. 2006. Proto-Kuki-Chin. Ph.D. dissertation, University of California, Berkeley. Chit Hlaing. 2007a. “Introduction: Notes on Edmund Leach’s Analysis of Kachin Society and its Further Applications.” In M. Sadan and F. Robinne (eds.), Social Dynamics in the Highlands of Southeast Asia: Reconsidering Political Systems of Highland Burma by E. R. Leach. Leiden: Brill Handbook of Oriental Studies, Section 3, South East Asia, pp.xxi-lii. —— 2007b. “Remarks upon Ethnicity Theory and Southeast Asia, with Special Reference to the Kayah.” In M. Gravers (ed.), Burma: Ethnic Diversity Past and Present. Copenhagen: NIAS Press, pp.107-48. —— 2007c. “The Central Position of the Shan/Tai as ‘Knowledge Brokers’ in the Inter-ethnic Network of the China-Burma (Myanmar) Borderlands.” Paper presented to the First International Conference on Shan Religion and Culture. London: School or Oriental & African Studies. ²²G5HYLHZRI+MRUOHLIXU-yQVVRQMien Relations. Journal of Southeast Asian Studies 38(2): 401-03. —— 2008. “Ethnicity Theory, Linguistics and Chin Identity.” Paper presented the international seminar Exploring the History, Culture and Identity of the Chin Groups, Aizawl Mizoram, India, September 2008. Fernquest, J.N.D. Crucible of War: Burma and the Ming in the Tai Frontier Zone (1382-1454) 77. http://slipperybannanapeel.blogspot.com. *URV67KH3ROLWLFVRI1DPHV7KH,GHQWL¿FDWLRQRIWKH'XORQJ 'UXQJ RI Northwest Yunnan. China Information 18(2): 275-302. -yQVVRQ+Mien Relations: Mountain People and State Control in Thailand. Ithaca: Cornell University Press. Lehman, F.K. 1963. The Structure of Chin Society. Urbana: University of Illinois Press (Illinois Studies in Anthropology, 3). —— 1967a. “Kayah Society as a Function of the Shan-Burma-Karen Context” In J.H. Steward (ed.), Contemporary Change in Traditional Societies. vol.1. Urbana: University of Illinois Press, pp.1-104 —— 1967b. “Ethnic Categories in Burma and the Theory of Social Systems.” In P. Kunstadter (ed.), Southeast Asian Tribes, Minorities and Nations, vol.1. Princeton: Princeton University Press, pp.93-124. Matisoff, J.A. 2003. Handbook of Proto-Tibeto-Burman. Berkeley: University of California Press. Mang, Kee Shein [Kee Shing Mang]. 2006. An Explanation of Verb Stem Alternation LQ .¶&Kz D &KLQ /DQJXDJH MA thesis, Payap University, Chiang Mai, Thailand. Solnit, D. 1997. Eastern Kayah Li: Grammar, Texts, Glossary. Honolulu: University of Hawai’i Press. ORAL HISTORIES 251

Walker, A.R. 2003. Merit and the Millennium: Routine and Crisis in the Ritual Lives of the Lahu People. New Delhi: Hindustan Publishing Co. Wang, Zhusheng. 1997. The Jingpo Kachin of the Yunnan Plateau. Tempe: Arizona State University Program for Southeast Asian Studies.

CORDS AND CONNECTIONS: RITUAL AND SPATIAL INTEGRATION IN THE JINGHPAW CULTURAL ZONE

MANDY SADAN

INTRODUCTION: RELATIONS AND RELATIVITY1

The Jinghpaw peoples residing in what is today northern Burma/ Myanmar, under their multi-group exonym Kachin, have long been FRQVLGHUHGDQHWKQRJUDSKLFH[HPSODURIÀXLGLGHQWLWLHVDQGPDOOHDEOH FRQVWUXFWHGHWKQLFDI¿OLDWLRQV(5/HDFK¶VZHOONQRZQDQGVRPHWLPHV controversial monograph on Kachin ‘Political Systems’ (1954), alongside his less well -known but equally important article on the ‘Frontiers of Burma’ (1960), established a benchmark against which the relational dimension of Kachin ethnicity was proclaimed its key ontological feature. Following Leach, this concentration on the relational has usually focused on the Kachin juxtaposition with the Shan people, themselves a heterogeneous social grouping when viewed across the full geographical range of Shan historical habitation from northeast India to the Mekong sub-region. One problematic aspect of Leach’s model of ethnic relativity is that the internal features of Tibeto-Burman ritual language practices, and the ways in which Kachin identities were negotiated through them, are usually seen as subsidiary to the Kachin-Shan dynamic, which is assumed to be the primary ethnographic fact. This paper will attempt to re-focus attention on internal relational identities within the Jinghpaw (‘Kachin’) cultural sphere, and in this respect Tibeto-Burman ritual ODQJXDJHSUDFWLFHVDUHVLJQL¿FDQW The criticism could be levelled that the term Kachin is an etic, quantitatively-driven notion through which an ethno-demographic constituency has been produced within the national system of ethnic structuring in Burma/Myanmar. In this context, the functions of origin and migration narratives, through which claims to common ancestry of all the Kachin sub-groups are made, are typically misunderstood by WKHLUPRGHUQGHWUDFWRUV$FFXVDWLRQVRIDUWL¿FLDOLW\DERXQGERWKRIWKH 1 A large number of people in the Kachin region have assisted me with this research over many years but I am particularly grateful to Pungga Ja Li and members of the Yup Uplift Committee for their support, and especially for allowing me to work with them on their DUFKLYHRIULWXDOSUDFWLFHVIURPZKLFKPDQ\RIWKH¿QGLQJVRIWKLVSDSHUKDYHEHHQGHULYHG All errors of interpretation are my own. 254 MANDY SADAN narratives themselves and, by implication, of the ethnic identities and the claims that appear to be based upon them. One reason why these narratives are prone to such criticism today is that they are typically dissociated from their original ritual idiom and are circulated in colloquial prose form. This paper will attempt to elucidate some of the different understandings that can be developed when the context of ritual performance is restored. Unfortunately, however, indigenous spirit practices in Jinghpaw ritual dialects have almost disappeared inside Burma/Myanmar today because of the widespread conversion of Kachin groups to various forms of Christianity.2 On the other hand, however, local research groups have attempted to document these practices, and this paper draws on the access to these cultural artefacts that has been granted to me over a number of years. Before considering these further, the broader political environment in which all of these forms currently H[LVWQHHGVWREHEULHÀ\RXWOLQHG

JINGHPAW ‘KACHIN’ ORIGIN MYTHS AND THEIR CONTEMPORARY POLITICAL CONTEXTS

The contemporary understanding of the identity Kachin in Burma/ 0\DQPDULVWKDWWKH.DFKLQDUHRQHRIWKHRI¿FLDOO\UHFRJQLVHG1DWLRQDO 5DFHV RI WKH %XUPHVH VWDWH 7KLV LGHQWL¿FDWLRQ ZDV HQVKULQHG LQWKH ¿UVW FRQVWLWXWLRQ RI PRGHUQ %XUPD LPSOHPHQWHG DW WKH WLPH RI LQGH- pendence from Britain in 1948. The story of independence, and the constitutional arrangements by which that independence was effected, LVRQHWKDWEULQJVOLWWOHJORU\WRWKH¿QDOGD\VRIHPSLUH)RUPDQ\RI the ethnic minority communities who found themselves bound to the Burmese state in structured political arrangements, in many cases for WKH¿UVWWLPHDJUHHPHQWWRHQWHULQWRWKHQHZFRQVWLWXWLRQDOIUDPHZRUN was something of a leap of faith. Many prefaced their agreement with reservations about the kind of autonomy that they anticipated would be implemented in their respective areas, and some, notably the representatives of the Karenni and Shan States, enshrined these reservations in the constitutionally-approved caveat that they could withdraw from the 8QLRQRI%XUPDDIWHUWHQ\HDUVLILWIDLOHGWRIXO¿OWKHLUH[SHFWDWLRQV Indigeneity and nationhood were inextricably combined in these con- VWLWXWLRQDODUUDQJHPHQWV7KLVZDVLQÀXHQFHGWRDJUHDWH[WHQWE\ERWK

2 Complex relationships between Buddhism, Christianity, state secularism and indigenous spirit practices also pertain in contiguous Jinghpaw regions in northeast India and Yunnan but will not be discussed in this paper. CORDS AND CONNECTIONS 255 the perception and reality of large scale Indian and Chinese immigration into Burma during the period of British colonial rule. The need to GH¿QHQDWLRQDOLQGLJHQHLW\WKURXJKWKHDQWLFRORQLDOZKLFKZDVHTXDWHG with the historically pre-colonial, was readily established in the structures of the new state. Indigenous races of the new Burma were those who were deemed to have been settled in the country prior to 1823, MXVWSULRUWRWKH¿UVW$QJOR%XUPHVHZDUDQGWKHEHJLQQLQJVRIIRUPDO British colonial expansion in the region. This decision was made despite the fact that ‘Burma’ did not exist as a border-regulated, territorially- delineated space at this time in many of the regions that the modern state now claimed as its own. The explicit nature of the proofs by which LQGLJHQHLW\ZDVWREHFRQ¿UPHGKRZHYHUZDVQRWFOHDUO\GHOLQHDWHG As independent Burma entered its long slide into military rule, which became entrenched following a military coup under General Ne Win in 1962, national constitutions were and are still being progressively reinvented as blueprints for the consolidation of centralised power in the face of opposition. Decisions about national indigeneity became the preserve of the Council of State, which could decide whether or not any ethnic group was deemed to be ‘national’. ,QDFRXQWU\LQZKLFKVLJQL¿FDQFHZDVDWWDFKHGWRWKHORFDOLVDWLRQRI an historical moment as a watershed by which the right to citizenship for all subsequent generations was predetermined, new importance was attached to the ability of minority ethnic groups to substantiate their own historical knowledge. Localised ethno-histories needed to be resituated as readable by the conventions of a statist history, both to establish the historical presence of the source communities who otherwise often lacked extensive own-language historical documentation and to counteract the powerful national narrative being produced by elites at the national centre as ethnic antagonisms developed into full-blown FRQÀLFWV,QWKLVFRQWH[WRIKHLJKWHQHGSROLWLFDOYLROHQFHQHZPHDQLQJV undoubtedly accrued to narratives of origin and migration. The principal ethnic groups around whose identities modern South East Asian states and nations have developed (the Burmans, the T(h)ai, WKH.KPHUWKH9LrWHWF WUDFHWKHLURULJLQVWRRWKHUGLVWDQWSODFHVIURP which they migrated to their present sites of permanent settlement and from where the modern nation developed. Origins and migration had a distant historical function as a starting point for the trajectory of modern state development. The value of migration as an ongoing, contemporary discourse is negated by this notion of national development and is even deemed to stand in opposition to the notion of settled civilisation. In 256 MANDY SADAN creating a constitutionally demarcated historical marker of settlement for all ‘National Races’ in Burma/Myanmar, migration was relegated to a historically transient or pre-national feature. For communities such as the Jinghpaw, however, for whom the notion and reality of migration was a critical aspect of their (ongoing) social formulation evolving over time and space, this interpretation of the ‘value’ of migration created an RQWRORJLFDOGLI¿FXOW\LQUHODWLRQWRWKHLURZQFRPPXQLW\KLVWRULHV7KH QHHGWRUHFRQ¿JXUHWKHPHDQLQJVRIPLJUDWLRQWRJLYHLWDµUHDOWLPH¶ historical validity by which the actual origins of the ‘Kachin’ peoples FRXOGEHWUDFHGWRDVSHFL¿FSODFHDQGWLPHZHUHXQGRXEWHGO\VLJQL¿- cant. Migration stories needed ideally to be rendered comparable and compatible with the migration narratives of national majorities in the modern (colonial and postcolonial) nation state. In the case of Kachin myths and narratives (and those of many other FRPPXQLWLHVWRR RQHRIWKHPRVWVLJQL¿FDQWLPSDFWVRIWKHHPHUJHQFH of the modern ethno-state has been the need to determine the common origins of all the principal sub-groups of the modern Kachin National 5DFHDQGWR¿[WKLVDVDOLQHDUQDUUDWLYH7KH.DFKLQVXEJURXSVWRGD\ are the Jinghpaw, the Lawngwaw (Maru, Lhaovo), the Zaiwa (Atzi, Atsi), the Lisu (Yawyin, Li-su), the Nung-Rawang (Nung, Rawang) and the Lachik (Lashi, Lacid). All of these communities, as currently GH¿QHGDUHFRPSOH[HQWLWLHVDQGWKHLUGLVWLQFWODQJXDJHV ZKLOVWUHFRJ- nising that polyglottism is the norm in this region) come from various branches of the Tibeto-Burman language family.3 At the level of ritual SUDFWLFHDQGVRFLDOLQWHJUDWLRQWKLVOLQJXLVWLFGLYHUVLW\LVVLJQL¿FDQWDV will be discussed below. Most modern accounts of the origins of the Kachin in Burma/ Myanmar iterate as an essential component the common ancestry of the different sub-groups. These are deemed to have separated from each other at various points along ancestral migration paths, reconnecting sometimes only in very recent history. For example, recently the discourse about the integration of the Lisu into the wider Kachin category has focused on 1963 and the coming together of Lisu groups in a large dance festival, or manau, at the outset of the Kachin civil war against the central government, this marking the return of the lost Lisu brother to the Kachin fold. This festival coincided with the widespread migration

3 Jinghpaw is placed in the Jinghpaw-Luish branch of the Jinghpaw-Konyak-Bodo EUDQFKRI7LEHWR%XUPDQWKH/DFKLN/DZQJZDZDQG=DLZDODQJXDJHVDUHFODVVL¿HGDV 1RUWKHUQ%XUPLVKLQWKH/ROR%XUPHVHEUDQFK/LVXLVFODVVL¿HGDV1RUWKHUQ/RORLVKLQWKH Lolo-Burmese branch; Nung and Rawang are considered to be part of the Nungish branch of Tibeto-Burman. CORDS AND CONNECTIONS 257 of Lisu communities from China entering the Kachin region as a result RIFRQÀLFWDFURVVWKHERUGHUDQGWKHVWDUWRIODUJHVFDOHDUPHGRSSRVL- tion to the Burmese regime in the Kachin region. As with many other communities, these transformations of local understandings of identity within new national systems that both privileged and marginalised non- national ethnicities also led to attempts to identify in a literal sense the site of common ancestral origins. In Jinghpaw myth, the geographical location of the mythic place from where the migration of all human EHLQJVFRPPHQFHGLV0DMRL6KLQJUD%XP2WKHUDWWHPSWVWR¿QGKLV- torical connections with archaeological sites in Tibet and Yunnan are also ongoing in the effort to trace routes of migration and to justify distinctive, indigenously produced historical frameworks that can coun- terbalance the discourse of ‘the nation’.4 Discourses of migration that LGHQWLI\VSHFL¿FJHRJUDSKLFORFDWLRQVDVSRLQWVRIRULJLQFRUUHODWHZLWK the interpretation of migration in conventional statist representations as a process undertaken by soon-to-be national groups and have been taken to imply (erroneously in many cases) the large scale movement RI SHRSOHV IURP RQH VHWWOHG SRLQW WR DQRWKHU DV D ODUJH VFDOH ÀRZ RI discrete populations. Yet these origin narratives were not recited with these intentions in the past, as will be demonstrated, and attempts to streamline them into linear projections of history have not surprisingly largely failed to reduce divergences in detail across different versions. This apparent uncertainty has led to assumptions that these accounts, at best, are confused or, at worst, are nothing less than products of modern political DUWL¿FHLQWHQGHGWRPRELOLVH.DFKLQQDWLRQDOLVWIHHOLQJ2QHRIWKHPRVW obvious examples of this was the production of a textbook by the Kachin Independence Army (KIA) in the mid-1980s, which was intended to be used in KIA schools in territory that they controlled and which explicitly set out to explain current political alliances of the Kachin sub-groups WKURXJKDJHQHDORJ\RIDQFHVWUDODI¿OLDWLRQV 1DZ$ZQ 7RFRXQ- ter Kachin nationalist claims and reassert the primacy of the Burmese nation state, writers within the Burmese establishment (especially during the time of the military regime, when styled as the Burma Socialist Programme Party (BSPP) under General Ne Win, 1962-1988) have tried to ‘prove’, for example, that the Lawngwaw or the Rawang were variously ‘descendants of the Pyu’ (considered by some to be progenitors 4 In 2008, for example, a large photographic display was exhibited at the manau festival held at the Headquarters of the Kachin Independence Army (KIA) in Laiza on the Burma- China border in which images were displayed that claimed to be of historic sites of Jinghpaw settlement and of the Tibetan Plateau as the locale of Majoi Shingra Bum. 258 MANDY SADAN of later Burmese society) or were ‘White Burmans’, thus splintering the Kachin nationalist discourse of common descent, repositioning these Kachin sub-groups as branches of the Bamar, as Burman peoples refer to themselves. Rather than being a tangential debate to more pressing political and socio-economic concerns, such discourses are important because their outcome is to challenge the demographic claims of Kachin nationalism. This impacts on the validity of Kachin claims for autonomous political structures in the Kachin region, especially as today the Kachin State retains hardly a Kachin majority demographic as a result of population transfer and movement, especially from lower Burma. When framed in this way, such discourses in the ethnic politics of contemporary Myanmar no longer seem trivial or a diversion from more pressing QHHGVDQGUHDOLWLHV5DWKHUWKH\UHÀHFWWKHORQJVWDQGLQJQHHGWRUHVLVW the cultural prejudices of the national majority against the languages and histories of many minorities. This is considered by minority nationality elites as an essential prerequisite for establishing fair systems of gover- nance in the future, as well as for consolidating in the present a strong body of evidence justifying claims for political rights.

CENSORSHIP AND MINORITY HISTORIES IN THE MODERN BURMESE STATE

Producing and circulating alternative discourses to the prevailing narrative of ‘the nation’ is no straightforward matter in contemporary Burma. Although the Burmese state recognises certain identities as indigenous categories (the Kachin, Kayah, Karen, Chin, Burman, Mon, Rakhine and Shan) and has a complex list of some 135 ‘National Races’ and/or linguistic subgroups arranged through thirteen ‘ethnic families’ (Gravers 2007; Smith 2002:15), this privilege has not in recent years extended to the right to publish what might be termed ethno-histories of these JURXSV2QHFRXOGQRWDQGLQGHHGRI¿FLDOO\VWLOOFDQQRWSXEOLVKDZRUN that presents itself explicitly as a ‘Kachin History’. However, within this system of constraint there are some important caveats which create gray areas that the various censorship committees seem badly equipped to manage where these relate to literary output in minority languages. One of the gray areas pertains to the distinction that the censor board makes between ‘factual’ history, the stuff of dates and ‘real’ events and places, which are heavily controlled, and ‘family’ histories, which require much OHVVULJRURXVIRUPVRIUHSUHVHQWDWLRQEHIRUHWKHRI¿FLDOFHQVRU,QWKH .DFKLQ FDVH WKH IRUPHU ZRXOG XOWLPDWHO\ OHDG WR UHÀHFWLRQV RQ WKH CORDS AND CONNECTIONS 259 origins of the post-colonial nationalist movement in the Kachin region and its opposition to the central regime. In the latter case, less stringent censorship criteria apply. For example, in 1999, a full volume text of ‘family history’ written in Jinghpaw language needed only to be pre- VHQWHGDVDWKLUW\SDJHVXPPDU\LQ%XUPHVHLILWFRXOGEHMXVWL¿HGDVD text of ‘family’ narrative; the more essential accompaniment would be the habitual payment of money to loosen the censor’s erratic grip. The traditional narratives have an advantage in that they can place themselves within the framework of family accounts, being the narratives of clans, lineages and family groups. This has been an important route by which these narratives have established a discursive presence in the relatively new secular public sphere of printed Jinghpaw language text in Burma. There are two main genres for ‘mythic’ histories published in this way as prose texts. In Jinghpaw they are Labau, which equates to something equivalent to ‘Old Story’, and the more traditional couplet phrase Ginru Ginsa, which equates broadly to ‘Origins and Migrations’. Many of these accounts are colloquial derivations of traditional recitations that would have been performed at various rituals but which have been removed from this context and stripped down to make them accessible as non-ritual texts. As a body, these works are also considered important for the strengthening of Jinghpaw language and literacy within %XUPDE\.DFKLQQDWLRQDOLVWV3ULRUWRWKHFHDVH¿UHEHWZHHQWKH.DFKLQ Independence Army (KIA) and the central regime in 1994, a number of Jinghpaw language publications were printed in the Kachin region using WKHVHJHQUHWHUPVLQWKHLUWLWOHV6RPHZHUHSURGXFHGE\WKHRI¿FLDOO\ VDQFWLRQHGEXWORFDOO\RUJDQLVHG&XOWXUDO&RPPLWWHHV¿UVWHVWDEOLVKHG XQGHUWKH%633DQGZKLFKKDYHIRXQGUHVXUJHQFHLQWKHSRVWFHDVH¿UH environment since 1994. Many small, hand-produced booklets detailing personal interpretations of various aspects of Labau have also been produced and circulated across the region through various kinship and other networks. Most important recently has been the reinvention of the public space of the manau performance across the Kachin region. These festivals have provided a forum that seems to have spurred local writers to produce booklets on origin and migration narratives as they are easily sold and distributed widely at these large public gatherings. A Jinghpaw language bookshop in Myitkyina, Laika NauraKDVDOVRÀRXULVKHGDQG encouraged this steady stream of self-published works. Frequently, young educated Kachin people within the Burmese higher education system and within the Kachin theological institutes, ZKLFKIRUPWKHSULPDU\URXWHLQWRKLJKHUHGXFDWLRQIRUPDQ\¿QGWKDW 260 MANDY SADAN these narratives fall far short of what one would expect of a ‘true’ history. Foreign colonial and contemporary local missionaries have also attempted to control their analytical frameworks within a discourse of &KULVWLDQLVDWLRQ WKH WDOH RI WKH ÀRRG EHLQJ DQ REYLRXV SRLQW RI GLV- SXWHLQFODLPVRIµDXWKHQWLFLW\¶7KHVXVSLFLRQWKDWSROLWLFDODUWL¿FHLV HPEHGGHGLQWKHVHFRQWHPSRUDU\WH[WVKDVDOVREHHQLQWHQVL¿HGE\WKH continued work of various Culture Committees and other groups since WKHFHDVH¿UH&RPPLWWHHVKDYHDWWHPSWHGWRVWDQGDUGLVHDFFRXQWVPDQ\ of their members being from an older, Burma Socialist Programme Party-controlled generation who wanted to minimise the appearance of differences that surfaced in the self-published, hand-duplicated texts. Kachin nationalism also has insisted on covering these discursive cracks for fear of splintering the Kachin nationalist movement. The resultant representations of origins and migrations are therefore often composites derived from a number of different oral models related as a homogenised whole (Duleng Labau Komiti 1994; Jinghpaw Wunpawng Htunghking hte Laili Laika Hpung Ginjaw, undated). By emphasising manau sites, funeral sites, village foundation sites, and so forth, the committees try to totalise the ritual geographic claims of the respective lineages and to emphasise their connectivity with the ritual space of what is now Kachin State. The composite nature of these accounts seems to indicate their lack of ‘authenticity’ and, again, the tendency is to dismiss their wider relevance in the dispute between the VXUIDFHDUWL¿FLDOLW\RIFDWHJRULHVOLNHµ.DFKLQ¶DQGWKHGHVLUHWRSURYH that such identities are vectors of historical identity. For communities lacking other forms of indigenously produced historical textual documentation, origin myths take on the intellectual burden of ‘proving’ the GLVFUHWHKLVWRULHVRIHWKQLFJURXSVDQGRISURYLGLQJMXVWL¿FDWLRQIRUWKH claims that are made for autonomy and rights within the modern state. During the period 1996-9, I was privileged to be able to engage with this kind of material, to study the ritual language of the dumsa ‘spirit priests’ and to compare multiple recordings of rituals conducted across the Kachin region in Jinghpaw. It is as a result of this study that it is possible to make the following comments.

MOVEMENT, MIGRATION AND ORIGINS – JINGHPAW CONCEPTS

Myths of origin were recounted through a series of oral recitations of which the most important were performed at manau festivals. The manau was the most prestigious rite of a Jinghpaw chief, and the right to hold the CORDS AND CONNECTIONS 261 festival was itself derived at a mythic level from the claims of a chief, as part of a particular lineage, to have had an ancestral kinship relationship with spirits in the upper levels of the Jinghpaw pantheon (what might be termed ‘sky’ spirits). The complexities of this are manifold, both in terms of the cosmological complications and the socio-political dimensions. Because of the nature of the ritual, in particular its intention to consolidate the mythic socio-ritual claims of particular lineages and the claims of particular individuals to lay claim to such high status, the main thrust of the narrative of origins is to explain the origin of the festival and its linear progression through many generations to the current, very particular site of local performance. This is perhaps best understood by considering one element of this ritual, the naura majai. This ritual, which is seldom performed anymore, was intended to SXULI\DQGFRQ¿UPWKHFLUFXODUGDQFLQJDUHQD naura) where sut (fecundity, prosperity) would accumulate through the coming together of different kinship groups, in various relations with that of the chief, to dance/‘step’ together.5 In the centre of the circle would be a set of posts, the shadung, which were embedded in the ground and were a receptacle for these ‘blessings’ from the spirits. These posts would be connected by strips of cane to the front roof post of the chief’s house, which in turn would be connected with another receptacle for the receipt of sut at the FHQWUDOULWXDO¿UHSODFHZKLFKZDVLQVLGHWKHKRXVHDQGZKLFKZDVWKH sole preserve of the chief. Large manau festivals took place only rarely, but the presence of the madai dap (dapLVD¿UHSODFHmadai is the main spirit honoured through the festival) was a permanent ritual structure LQVLGH WKH FKLHI¶V KRXVH WKDW FRQ¿UPHG WKH ULJKW WR KROG WKH IHVWLYDO When individuals wished to acquire the right to hold manau (which was possible in certain circumstances), what they in fact obtained was the right to have a madai¿UHSODFH6 At the end of the manau, the various strings and cords would be cut symbolically above a special, close- ZRYHQEDVNHWZLWKDWLJKW¿WWLQJOLGVRWKDWWKHHQGVGLSSHGLQVLGHDV they did so, all the sut that had accumulated would be packed up in the

5 This description comes from the extensive documentation of a ‘traditional’ manau that was performed in 1992 in KIA controlled territory in the northern Kachin region. The performance was held explicitly with the intention of creating an archival record as such events had not been carried out since the 1970s and prior to that only rarely since WKH HDUO\ WZHQWLHWK FHQWXU\7KH KLVWRULFDO DQG FXOWXUDO VLJQL¿FDQFH RI WKH IHVWLYDO WR WKH Kachin peoples enabled the festival to be held as a ‘one-off’ event for archival purposes (albeit in a remote area) despite contemporary objections amongst urban Kachin elites about the potential of spirit practices to make a resurgence and cause fragmentation of the nationalist cause. 6 Personal communication, Pungga Ja Li, May 1998. 262 MANDY SADAN basket and stored above the madai ¿UHSODFH7KLVUHSUHVHQWHGWKHDFFX- mulated prosperity and fecundity of the chief’s line both in the present and for future use. 7KHSXUL¿FDWLRQRIWKHGDQFLQJDUHDUHYHDOVPXFKDERXWWKHSXUSRVHRI origin stories in the Jinghpaw cultural sphere. The recitation would take the form of an account of the origins of the earth, the spirits and then human beings as a means of explaining the origins of the festival. The narrative would inevitably slip into what we could consider a migration narrative in that its purpose was to legitimise the particular claims of the current host to sponsor the ritual. This had to be established along a genealogical route so that the ritual and cosmological authenticity of the current performance could not be brought into question. The form that this genealogy took, however, was not a rigid generation-to- generation family tree. Rather, the genealogy was expressed through the line of those in the same spiritual and ritual position as the current host in the family, that is, the line of those who had previously held manau festivals, and generations would be omitted where no festival had been held.7 Its purpose was to consolidate contemporary claims as much as to give an (incomplete) account of the lineage. Origins in this context were not intended to answer the question ‘Where are we from?’, rather they were intended to answer the question ‘How are my rights to hold WKLVVLJQL¿FDQWULWXDOGHULYHGDQGOHJLWLPDWHG"¶)XUWKHUPRUHPRYHPHQW (‘migration’) was expressed not as a generation-by-generation linkage through all sites of residence (although it would be hoped that this would be the case as it would indicate that each generation of a lineage had VXI¿FLHQWSURVSHULW\WRPDLQWDLQWKHLUULWXDOREOLJDWLRQVWKURXJKWKHSHU- formance of manau), but rather through the spatial points at which these ritual obligations between humans and higher spirits had been maintained and consolidated. The critical aspect of these narratives of origin relating to the Jinghpaw cultural sphere is that when human beings separated from spirits, there developed a particular kind of cultural evolution resulting in a mature -LQJKSDZFXOWXUHZKLFKGH¿QHGKRZWKHFRQWLQXHGOLQNVEHWZHHQVSLULWV and humans were to be expressed. The markers of movement towards this cultural development are concepts known as Majoi, as referenced DOVRLQWKHWHUP0DMRL6KLQJUD%XP7ZRLQSDUWLFXODUDUHJLYHQVLJQL¿- cant cultural status, which are geographical sites rooted in the northern Kachin region: Chyai Hku Majoi and Hkrang Hku Majoi, the watersheds

7 Npawt La, Naura Majai, oral performance, April 1992, Mali Nmai Walawng, Kachin State. CORDS AND CONNECTIONS 263

RIWKH&K\DLDQG+NUDQJ5LYHUV$WHDFKRIWKHVHVLWHVDVSHFL¿FDOO\ Jinghpaw cultural form is deemed to have arisen: the development of the mayu-dauma marriage system, the development of the manau, and so RQ7KHVHVSHFL¿FFXOWXUDOIRUPVDUHLPSRUWDQWEHFDXVHWKH\GHWHUPLQH how connections are maintained not only between humans and spirits, but also between migratory groups in a large and potentially expanding environment in which there was constant pressure to move and relocate. The important issue is not the fact of movement per se, but the links and connections that give that movement meaning and maintain its integration into a social and ritual system. The mapping of distant paths was not the issue; rather, it was how the Jinghpaw cultural sphere was FUHDWHGDQGPDLQWDLQHGLWVHOILQDFRQWH[WRIFRQWLQXDOPLJUDWRU\ÀRZ and interaction. Another ritual in which myths of origin were described took place at PDUULDJHZKHQWZRNLQJURXSVZRXOGDI¿UPWKHLUUHODWLRQVKLSZLWKHDFK other through the giving and taking of brides. After the bride had entered the new home, it was common for a ritual (lanyi) to be performed for the daughters of chiefs or the wealthy. This was a truncated version of the creation story told at the manau festival.8 However, the purpose in this case was to educate the bride about her role and the mythic origins of her various duties and responsibilities. In this rendition of the origins of all things, the legitimation took the form of ancestral couplings: the origin RIDOOÀRUDDQGIDXQDZDVGHPRQVWUDWHGWKURXJKWKHSDLULQJRISDUWLFXODU plants and animals, and the origins of cultural artefacts such as pots and textiles were explained. The origin story was adapted to the particular circumstances that it was being asked to address; it was not asked to answer a narrow question of ‘Where are we from?’ When a married ZRPDQ KRVWHG RWKHU ULWXDOV LQ ZKLFK VRPH JHQHDORJLFDO YHUL¿FDWLRQ of her right to sponsor the ritual was required, the genealogy against which the narrative was structured would be done in a similar way as described for the manau. The main difference was that her lineage would be related through a similar ritual-cosmological position to herself, that is, the wives of those who had married into her husband’s lineage; her genealogy could not be related directly through the husband’s line. Each individual symbolically encapsulated the genealogical status of those who were situated in similar social positions through time, as a form of

8 See Pungga Ja Li and Chyakyi Brang 1999. This text was produced as a result of archival work engaged in by the author and local research groups during 1996-9 and is based on a recitation of the Lanyi performed by Chyahkyi Brang in 1992 in Monyin, Burma. Comments on the performance in this paper are derived primarily from this source and a number of smaller recitations performed by Npawt La and others in the early 1990s. 264 MANDY SADAN collective individuation. Another form of movement, a ‘reverse’ migration, was utilised and described when sending the spirit away after death.9 Again, lineage and geography both come into play as the spirit would be directed along a path, but not one that connected ancestral sites as is often the case in these kinds of reverse genealogies, but rather a route outlining the most direct way back to the site where the main ancestor spirit of the clan resides, and for the most part avoiding human habitations. This last geographical marker, often a mountain top at which the lineage ancestor VSLULWGHVFHQGHGWRWKHKXPDQUHDOPLGHQWL¿HVWKHSRLQWRIVHSDUDWLRQ between the living and the spirit world. From this point, the spirit is directed through a cosmic geography that in many cases involves the arrival at a junction of ten roads, the passing of ten houses and so on, until the ancestral village is reached. Again, one needs to consider the nature of this ‘reverse migration’ to better understand its purpose. A spirit would have to be sent back to the land of the ancestors no matter where the death occurred. For H[DPSOH .DFKLQ VROGLHUV ZKR GLHG LQ WKH 0HVRSRWDPLDQ FRQÀLFW RI the First World War had these rituals performed on their behalf. Their bags, which formed part of their military equipment in the Indian army, were taken back to their home villages which ensured that the spirit was present and could then be separated; full separation could only be achieved with the coming together of different kin groups to perform WKHLUVSHFL¿FWDVNVLQWKHSURFHVV,IWKLVGLGQRWKDSSHQWKHUHZDVJUHDW fear about the potentially malevolent power of those spirits who had not been despatched, particularly those who had died violent deaths. For H[DPSOHLQDFHUHPRQ\IRUWKHPRWKHURIRQH.,$RI¿FHUKHOGLQ&KLQD in 2000, a subsidiary ritual was incorporated for Maran Brang Seng, one of her kin group, who was leader of the KIA until he died in 1994.10 Because he was Christian, these rites had not been performed, but some HOGHUVUHWDLQHGDIHDUDERXWKLVSRWHQWLDODVDQXQVDWLV¿HGVSLULWZKLFK prompted the symbolic performance some years after the fact. The purpose in this ritual was to send the spirit back to the ancestral homeland

9 Comments made in relation to funeral recitations or tsu dumsa have been developed IURPDQXPEHURIVRXUFHVEXWRIPRVWVLJQL¿FDQFHZDVWKHGRFXPHQWDWLRQRIDSHUIRUPDQFH conducted by dumsa Waze Naw conducted in northern Kachin State in 1991 in the possession of the Yup Uplift Committee, Myitkyina, and of recordings made on behalf of the Green Centre for World Art in 1998 of a funeral in Shwebo region undertaken by Hkanhpa Tu Sadan, Saw Law Eh Soe and Kareng Naw Dawng. 10 See Pungga Ja Li 2001. These statements are also supported by consideration of the video and photographic documentation collated by Pungga Ja Li’s research group during this ceremony. CORDS AND CONNECTIONS 265 in a way that was likely to interfere as little as possible with the living, to whom it might become attached. It might thus refuse to depart and UHPDLQLQGH¿QLWHO\LQWKHKXPDQUHDOPDVDPDOHYROHQWIRUFHGHPDQGLQJ constant supplication. The purpose of the route, therefore, was a quick, secure despatch. The habitations of the living do not form any particular part of this route; rather it is tracked through rivers, mountain tops and other natural features, with the intention that the process could be completed without error and in good speed. Much of this geographical routing is developed by the spirit priest himself from his own knowledge of the local environ- PHQWDOWKRXJKLWFDQIROORZORFDOFRQYHQWLRQV0RVWVLJQL¿FDQWO\WKH routes tend to become truncated at the point at which the knowledge of the spirit priest—whether real or acquired from listening to other performances—becomes thin. In some cases, committed believers in indigenous spirit beliefs, who are concerned that dumsa priests today have inadequate knowledge of the routes that should be followed by spirits of a particular lineage, have been known to write them down on paper to avoid any mistakes being made in the route.11 Mistakes could result in the spirit being left to wander helplessly and never reach the point of separation. The human migration history is cut short as a linear, temporal narrative by the desire to merge it as quickly as possible with a point of historical separation and to place the spirit on a mythic trajectory. These points of separation rarely extend beyond the borders of the Jinghpaw language and cultural zone, in distinction to the kinds of historic migrations described by national groups discussed above. A similar feature of directing spirits along clearly described and demarcated routes to minimise their potential for wandering off and causing trouble to the living can be found in other forms of recitation.12 Spirits invited to the manau are summoned to the site of the performance along routes; those that are not wanted are invited initially so that they can be placated and then sent away, but this would always be done in such a way that their route should be contained along particular paths. However, these routes did not have to be very detailed; one spirit journey recited at a wedding empowered the spirits to take the train, even 11 This was seen at a tsu dumsa performed in 1996 at a ritual in Shwegu as well as in the naura majai recitation referenced earlier conducted by Waze Naw, where the concern similarly was that spirits in this case should be brought to the ritual site in a controlled fashion minimising the possibility of their making demands on the living away from the site. 12&RPPHQWVLQWKLVSDUDJUDSKDUHGHULYHGIURPVWXGLHVVSHFL¿FDOO\RIWKHmanau, lanyi and tsu dumsa as already detailed but also incorporate understandings gleaned from other recordings of many different spirits made by the Yup Uplift Committee in Myitkyina in the early 1990s. 266 MANDY SADAN complimenting the guard on his careful work as they handed him their tickets (Pungga Ja Li and Chyahkyi Brang 1999). :HVKRXOGEHZDUHRIWU\LQJWR¿[VWDWHPHQWVIURPDQ\RQHSDUWLFXODU recitation as a prescriptive geography without considering the way these features function across a range of different performance types. Many of the ritual journeys, however, are very important for establishing an embedded societal knowledge of the spatiality of the Jinghpaw cultural zone beyond the site of immediate habitation. Their purpose, nonetheless, is not to trace a geographical point of historical origin across continents, as has become the intention of such migration discourses in relation to national majorities in modern South East Asian states, as described above. Most geographical accounts of routes become truncated, and all of them seem to diminish in clarity when the outer limits of the Jinghpaw cultural zone are reached. This, however, is true because the narratives were never intended to map out routes in the way that is sometimes assumed of them, that is, to provide a detailed, literal knowledge of, for example, where Majoi Shingra Bum, the origin place of humans, was, beyond the Jinghpaw cultural sphere. They were not answering this question because this was not a question that needed to be asked. From these spirit journeys, however, it becomes easier to understand how these various forms of migration and origin narrative could coalesce WRIRUPFRPSRVLWHVSDWLDOXQGHUVWDQGLQJVWKHPRVWVLJQL¿FDQWRIZKLFK recently has been undertaken by the main Jinghpaw Literature and Culture Committee in Kachin State (Jinghpaw Wunpawng Htunghking hte Laili Laika Hpung Ginjaw ca. 2001). This ‘draft’ has attempted to bring together multiple routes and to juxtapose them in the representation of Kachin historical geography.13 The determination to produce in the future many comparative renditions of routes and paths presents considerable potential for generating a very detailed account of localised histories; however, this cannot be achieved until current political fears over fragmentation and disintegration of the nationalist narrative are overcome. The notion of movement is embedded in all these narratives and forms part of their meaning. But movement in this case has a different ontological basis to that of ‘migration’ conceived as the historically- framed, large scale movement of peoples. In this latter understanding of

137KHIDFWWKDWWKHPDQXVFULSWLVFLUFXODWHGDVDµGUDIW¶UHÀHFWVDQRWKHUZD\LQZKLFK people attempt to circumvent the censorship boards. Drafts for private circulation are not subject to submission to the censor. In this case, the text pushes the boundaries of what could EHMXVWL¿HGDVµIDPLO\KLVWRU\¶WRWKHFHQVRUVKLSERDUGDQGFRPHVFORVHWREHLQJVHHQDVD ‘Kachin’ national historical representation. The Committee has avoided this possibility by retaining ‘draft’ in the printed form, thus avoiding the need to submit it for censorship CORDS AND CONNECTIONS 267

‘migration’, movement itself is the underlying concept, often as a form of permanent separation from a previous habitation. In Jinghpaw models, in which origins and genealogical movement are described, other concepts come into play and movement can be differentiated into that which is meaningful and contributes to the production of meaningful cultural space and that which is not and is culturally unincorporated. The genre term Ginru Ginsa, used for origin myths, can best be translated as meaning ‘old, established cords’. This term and the notion RIFRUGVDUHNH\LQGHWHUPLQLQJDORQJKLVWRULFDOÀRZRIPHDQLQJIXO movement. As has been noted in relation to the manau, the linking of the ritual structures at these events with cane strips is used to visualise in a literal way that the cords are deemed to connect humans within the kinship network, which is ultimately the emic, ontological bearer of Jinghpaw and latterly ‘Kachin’ identity. People within this indigenous network system are deemed to be intimately connected to each other and to spirits in benevolent communication with them, through invisible life cords. The umbilicus is planted at the base of the house post as a means of embedding this connectivity; illness is caused by malevolent spirits ELWLQJRUSXOOLQJDWWKHOLIHVWULQJVDWVLJQL¿FDQWULWXDOVWKHZLIHWDNLQJ groups (dauma) frequently bring special spears made of a cane that is GHHPHGWRKDYHIDUUHDFKLQJURRWV\VWHPVDVDPHDQVRIFRQ¿UPLQJWKH RQJRLQJFRQQHFWLRQEHWZHHQNLQJURXSVRQO\DWWKH¿QDOULWXDOVZKHQ the spirit is sent away to the land of the ancestors, does the wife-giving group (mayu FRPHZLWKDVSHFLDOVZRUGWRPDNHWKH¿QDOFXWRIVHSDUD- tion, severing the life strings between the living and the dead. Similarly, at the manau, the sut descends and is transmitted through cords, visualised in a literal way through those connecting the manau post, the house posts, and the manau ¿UHSODFH DQG DUH GHSRVLWHG LQ the lineage when the cords are cut and the sutLVDOORZHGWRÀRZLQWR the basket. The language of the spirit priests is also full of images of connectivity, particularly through the symbol of cane root systems that cover wide areas; the umbilicus at the house post ensures a constant connectivity between those who are travelling away from the house to other places, until such occasion arises for separation. It is these cords DQGFRQQHFWLYLWLHVWKDWDUHFRQ¿UPHGLQWKHSURFHVVRIUHFLWLQJP\WKVRI origin. In this recitation, movement without connection, the notion of movement which does not consolidate the connectivity in a new space, is not ginru ginsa, is not part of a narrative of origins and migration. Indeed, such movement does not even have a name as it marks a point of separation from the Jinghpaw cultural sphere. 268 MANDY SADAN

There are, of course, other forms of movement than these spatially DQGULWXDOO\VLJQL¿FDQWIRUPVEXWWKHVHPRYHPHQWVDUHRQO\PHDQLQJ- ful where they invoke the network of connectivities across this space, essentially through the kinship system. If one wants to learn about micro-patterns of movement (whether for trade, cultivation etc.) over the Jinghpaw cultural area, a different notion is relevant, a notion of networks. This notion is based on the concept of matut mahkai, which, rather than ‘cord’, can be translated as ‘links’, as in a chain. It is also a facilitator of meaningful interaction as one moves through space in the ordinary conduct of life, but it does not involve the establishment of new demarcations of ginru ginsa. These two terms, ginru ginsa and matut mahkai, have symbolic form on many manau shadung or posts, where curvilinear, intertwining cords represent the migrations and origins of the various kin groups. In these symbolic forms, interconnected lines denote the indigenous network system by which movement leads to the consolidation of socio-cultural and ritual space across a diverse geographical span. These concepts of connection over space and time, through life strings and umbilical cords, through chains and links, underpin Jinghpaw notions of cultural and ritual space. These notions have a geographical referent, but we should not expect local narratives of origin and movement to be able to answer some of the bigger social science questions that we may have concerning the large scale movement of ethnicities over time and space. More importantly, if we anticipate that these narratives can provide us with these answers, derived from our own preconceived notions of what they should ‘mean’, then we set them up to fail in the task, whilst perhaps missing some of the more important insights that they can give. In the ‘Kachin’ case, insights are gleaned on the complexity of the social and cultural formulations that have been developed in this region over a historically indeterminate but clearly extensive period of time. They reveal the ways in which migratory kin groups become integrated in a broader indigenous network system, which still continues to evolve and be refor- mulated through the discourses of contemporary Kachin nationalism.

RITUAL AND THE TIBETO-BURMAN BOUNDARIES OF THE JINGHPAW CULTURAL SPHERE

With these new understandings, the value of these narratives to historians, as well as anthropologists, takes on a new light. By exploring processual dimensions of Jinghpaw ritual language and performance CORDS AND CONNECTIONS 269

FRPSDUDWLYHO\ UDWKHU WKDQ MXVW DV VSHFL¿F HWKQRJUDSKLF DQG ORFDOLVHG examples, one can address some of the ways in which multi-group LGHQWLWLHVDUHLQWHJUDWHGLQWHUQDOO\QRWDVDV\VWHPRIDUWL¿FHEXWDVD necessary means by which complex socio-political environments could be negotiated through an indigenised social process over time. One of the features that can be subjected to historical examination is the ways in which the Tibeto-Burman linguistic boundaries may be VLJQL¿FDQW LQ SURGXFLQJ WKLV V\VWHP RI LQWHJUDWLRQ$V VWDWHG HDUOLHU critical in this kind of analysis is the need to move away from colloquial renderings in prose accounts of origin and migration narratives and to UHVLWXDWHWKHPLQWKHLUULWXDOFRQWH[W:KHQWKLVLVGRQHWKHVLJQL¿FDQFH of particular linguistic relations can be seen more clearly. For example, the Jinghpaw ritual idiom is conducted in an elaborate, creative but ultimately prescriptive euphonic structure of couplet phrases. Where genealogies, and therefore names, are used in these rituals, the need to DGKHUHWR-LQJKSDZQDPLQJSDWWHUQVLPSRVHVDFRQVWUDLQWRQWKHÀH[LELO- ity of the form. One might compare the traditional genealogical tradition of groups, such as the Lawngwaw (Lhaovo or Maru) and the Lachik (Lashi, Lachid), whose primary language (and naming systems) are Lolo-Burman. It is common in these Lolo-Burman (Northern Burmish) language communities for names to incorporate part of the father’s name IURPWKHSUHYLRXVJHQHUDWLRQVRWKDWWKHODVWQDPHEHFRPHVWKH¿UVW name in the next generation. This system of naming is distinctive and does not map well onto Jinghpaw rituals. There are ways round this, but they involve giving the individual a Jinghpaw name. Furthermore, Lawngwaw ritual language is performed in triplet sentences, and cannot be simply re-encoded in a Jinghpaw form, which depends upon pairs. Clearly, when a community using Jinghpaw ritual speech codes is in contact with a community using primarily Lawngwaw speech codes, a good deal of negotiation must take place if the two groups are to engage in common ritual practices, which are a necessity in the establishment RI DI¿QDO NLQ UHODWLRQVKLSV DQG WR H[WHQG WKH FRQWRXUV RI PHDQLQJIXO connectivities across new geographical and socio-cultural space. This is exactly what takes place. Built into Jinghpaw ritual practice is an intensive system of negotiation on form and practice. It is critical that before any ritual performance, local elders, the group hosting the event and the ritual practitioners meet together to decide upon correct practice and how to manage any amendments or adaptations that might be necessary. This is usually rationalised as a process that will ensure that no errors are made that might incur the wrath of the spirits, but 270 MANDY SADAN its outcome is the constant social negotiation of status between groups and a management of the local political environment. This occurs as much between groups within the Jinghpaw kinship system as between Jinghpaw groups and those from outside this network; it also helps to explain many apparent incongruities in local areas, for example, the fact that many Zaiwa communities of Yunnan conduct their ritual practices in Jinghpaw and not in Zaiwa: it has historically been the case that Jinghpaw chiefs were able to dominate socio-ritual relations in this area, which reveals much about local transformations of authority and power when placed in historical context. Furthermore, when approached in this way, this system also helps to establish a clearer delimiter of the Jinghpaw cultural sphere: where the local socio-political environment is such that negotiations cannot be made, or where the Jinghpaw ritual speech form is displaced in favour of others, boundaries are established. In many cases, dual systems develop where groups retain distinctive practices performed in non-Jinghpaw languages but add others that may blend the two. In these situations, cosmological parallels are established between lineages so that, for example, the elder line of a particular Lisu group may be made equivalent with that of a Jinghpaw kin group. This LVSDUWLFXODUO\QHFHVVDU\ZKHQQHZDI¿QDOUHODWLRQVDUHEHLQJIRUPHG but the parallels enable kinship relations to be established with any other group within the Jinghpaw sphere with whom they might in future come into contact. The surface-level relativity of identities in this system exists to facilitate the integration of difference where the local dynamics of power relations enable this to take place, but it also creates coherency where there are constant forms of migration and movement over historical time and space. Indeed, these rituals rationalise the historical experience of movement rather than seek to normalise the ideal of static settlement; WKH\ UHSUHVHQW D GLIIHUHQW FRJQLWLYH FRQ¿JXUDWLRQ RI WKH PHDQLQJ RI space and movement through it to that expressed by the discourse of migration as a necessary precursor to settlement and state formation. These rituals thus embed a different meaning for the notion of ‘origins’. The production of Kachin myths of origin seems to have been a constantly rearticulated and negotiated process, which retained vitality because of the notion that at an epistemological level these temporary negotiations did not involve permanent displacements of meaning for local groups. They were tools by which value systems were correlated for the purposes of establishing effective communication, although that did not make the adaptations any less ‘true’ in a spiritual or cosmological CORDS AND CONNECTIONS 271 sense. Leach commented more negatively on the great diversity he found in ritual practice and myth across the Kachin region, suggesting that it was a sign of the innate volatility of Kachin political life. The need constantly to re-negotiate meanings for each performance of a ritual was taken as a sign of fragmentation and weakness, and of the potential disintegration of the Kachin social sphere; its obverse, the assertion of DXWKRULWDULDQµFODVVLFDO¶FKLHÀ\UXOHZDVWDNHQDVWKHQHPHVLVRIFRQVHQ- sus and thus equally damaging of social stability. The colonial archive FRQWDLQVGRFXPHQWHGFDVHVRIVXFKQHJRWLDWLRQVDQGWKHLQWHQVHFRQÀLFW RI:RUOG:DU,,LQ%XUPDZKLFK/HDFKZLWQHVVHG¿UVWKDQGSURGXFHG an environment in which the negotiation of meaning was particularly hard fought.14 Understanding this process of constant negotiation helps us again to explore the historical process through documentation that is in more traditional historical thinking considered a-historical, if not avowedly anti-historical. What has been overlooked, or inadequately appreciated, has been the innate processual engagement with discourses of commonality that rituals (such as the tsu dumsa and the naura majai) constantly recon- ¿JXUHG,QWKHSRVWFRORQLDOZRUOGWKHVHSURFHVVHVKDYHFRQWLQXHGWR underpin a good deal of Kachin symbolic interaction with the modern nation state. These negotiated rituals demonstrate how internal models have been effective in integrating a common socio-cultural sphere as a Jinghpaw domain, delineated by contact not with the Shan but with other Tibeto-Burman-speaking groups who share a similar need to rationalise the experience of constant movement and engagement with other migratory groups. The outcome in the Kachin case has in recent years been different to what we hear of among the Naga, or the Chin, for example. Wettstein (this volume) describes persuasively how Naga usages of myths of origin have developed in the opposite direction, producing extreme levels of socio-cultural and political fragmentation. That there should be such different responses to common pressures produced by the emergence of modern states and forces of globalisation requires deeper, comparative analysis, not just at the political level but also at the societal and ritual level. It also suggests new ways in which the study of origin myths has relevance to our historical knowledge of minority ethnicities in national systems. In sum, complexity and negotiation were essential tools of process in the rendering of Kachin myth and ritual; when stripped down in

14 For a discussion of one example of these negotiations at a manau festival in the early 1920s, see Sadan 2002. On Leach’s wartime experience, see Anderson 2007. 272 MANDY SADAN contemporary narratives oriented towards the discourse of the modern state, these functions become invisible. As I have tried to argue, the original questions these myths were intended to address were probably not the literal ones: ‘Where are we from? How did we get from there to here, our present settlement?’ and ‘When did we arrive here?’ Rather, the myths were intended to answer the question: ‘How do we explain and control the relations we have in the space in which we live and move?’ Revising the question also helps to explain why knowledge of geographical space and routes generally diminishes at the boundaries of the language-cultural sphere in which the dominant idiom of the ritual prevails. Certainly, Jinghpaw narratives of origin and migration lack all clarity beyond the geographical delimitations of the space in which Jinghpaw exists as a primary medium of communication. This is true because these myths were intended to detail the creation of meaningful social environments, not just linear trans-continental routes; migration narratives were intended to reinforce collectively individuated, indigenous network systems, not impersonal paths. Yet, the comparatively blunt conceptual tools of the static, bounded nation-state have constructed new meaning frameworks for the notion of ‘origins’. With new, evolutionary and developmental connotations for such concepts, communities such as the Kachin today have to prove their historical settlement within the pre-1823 Burmese state as a precursor to claims for contemporary citizenship. Only in this light can we better appreciate, too, some of the GLI¿FXOWLHVIDFLQJWKHVHFRPPXQLWLHVLQUHQGHULQJWUDGLWLRQDOV\VWHPVRI knowledge within contemporary value systems and thus better understand why local contemporary discourses on the value and meaning of these origin and migration narratives can seem so problematic according to the conventions of historical knowledge in the western tradition. CORDS AND CONNECTIONS 273

BIBLIOGRAPHY

Anderson, R. 2007. “The biographical origins of ‘Political Systems of Highland Burma’.” In F. Robinne and M. Sadan (eds.), Social Dynamics in the Highlands of Southeast Asia: Reconsidering Political Systems of Highland Burma by E.R. Leach. Leiden: Brill, pp.3-29. Dureng Labau Komiti. 1994. Dureng Labau (Draft). Myitkyina (typescript). Gravers, M. 2007. “Introduction: Ethnicity against State–State against Ethnic Diversity?” In M. Gravers (ed.), Exploring Ethnic Diversity in Burma. Copenhagen: Nordic Institute of Asian Studies Press. Jinghpaw Wunpawng Htunghking hte Laili Laika Hpung Ginjaw. N.d. Jinghpaw Wunpawng Amyu Sha Ni A Ginru Ginsa Yu Hkrat Ai Lam, Ningpawt Ninghpang, Labau Kadun (First Draft). Myitkyina (typescript). Kuper, A. 1986. An Interview with Edmund Leach. Current Anthropology 27(4): 375-82. L. Zau Mun. 1995. Gauri Krung hte Jinghpaw Gauri Ni (Draft). Myitkyina (typescript). L. Zau Mun. 1997. Jinghpaw Wunpawng Sha Ni A Npawt Nhpang (Draft). Myitkyina (typescript). Leach, E.R. 1961. The Frontiers of ‘Burma’. Comparative Studies in Society and History 3(1): 49-68. Naw Awn, H. 1986. Wunpawng Labau Ginshi (publisher and place of publication unknown). Pungga Ja Li and Chyakyi Brang. 1999. Shanhpyi Laika: Lanyi. Yangon: A Z Offset. Pungga Ja Li. 2001. Jan Ga De Sa Dat Ai Lam. (publisher and place of publication unknown). Sadan, M. 2002. “The Kachin Manau and Manau Shadung: The Development of an Ethno-cultural Symbol.” In R. Blurton and A. Green (eds.), Burma: Art & Archaeology. London: British Museum Press, pp.119-31. Smith M. 2002. Burma (Myanmar): The Time For Change. London: Minority Rights Group.

ORIGIN AND RETURN: GENESIS AND THE SOULS OF THE DEAD IN NAXI MYTH AND RITUAL

CHARLES F. MCKHANN

Along with several other Tibeto-Burman peoples in southwest China, the Naxi have an anthropogenic myth that emphasizes migration from an ancestral homeland (located vaguely to the north) by pairs of brothers who became the progenitors of named clans that spread out across the landscape. Many of these same groups also perform funerary rituals that involve sending the souls of the dead back along these ‘roads’, to that original place ‘where the ancestors dwell’. In this paper, I focus on the symbolism of origin and division, on the one hand, and of return and UHXQL¿FDWLRQ RQ WKH RWKHU ZLWKLQ WKH 1D[L UHOLJLRXV WUDGLWLRQ$ORQJ the way, I look at parallels and differences in the Yi (mainly Nuosu) tradition, to suggest the scope of regional variation, and possible links to social organization. Of particular interest in the Naxi case is the way in which Buddhist ideas, mainly from Tibet, have been grafted onto an earlier Naxi cosmology. This is most evident in Naxi funerary symbolism. The Naxi myth of anthropogenesis is recorded in the dongba texts Coqbbert’v and Coqbbersal. Dongba are traditional Naxi ritual specialists who have much in common with bimo among the neighboring Yi (Nuosu). Both types of specialist chant from texts written in their own scripts—one pictographic (dongba) and the other syllabic (bimo)—but the religions are diffused, and there is no church or formal institutional structure that binds the practitioners together, either hierarchically (as in ranks) or architecturally (as in temples or monasteries). In this respect they are different from (Han) Chinese, Prmi, Naxi, Na/Mosuo, and Tibetan Buddhist communities also present in the region—i.e., throughout Khams, and at the temple complexes at Mt. Jizu and Mt. Emei—and more similar to the indigenous daba and anji (Ch. hangui) among the Na/ Mosuo and Prmi, respectively, although the latter lack written scripts.1 1 Two other interesting cases include the Pumi (Prmi) of Ninglang County, whose hangui (anji) have texts written in Pumi, but using Tibetan script (Wellens 2006), and Lisu ritualists in northern Weixi County, who consciously invented a script for their chants in the early 20th century. The Lisu script is syllabic, like the Yi bimo script and the lesser-used Naxi dongba syllabic script, called ggeqbbaq. In general, possessing ritual scripts has been a matter of prestige throughout the region. Whereas the Lisu created one less than a hundred years ago, many other groups display a kind of ‘script envy’ in legends about former scripts that somehow got lost. Na/Mosuo daba say they used to have one, written on animal hides, but 276 CHARLES F. MCKHANN

Coqbbert’v and Coqbbersal are myths recounting the origin and migration of Coqsseileel’ee (Coq), the earth-born male progenitor of the KXPDQ UDFH DQG KLV FHOHVWLDO ZLIH &HLOKHHTEEXEHT &HLO 7KH\¿W together as two parts of a story:

The Migration of Coq2

In the beginning heaven and earth comingled. The god Dduq and the goddess Seiq had not yet given order to things. Trees could walk and rocks could talk. Before the things appeared, there were the shadows of things. The true and the real appeared, and they became a white hen. It had ZLQJVEXWFRXOGQRWÀ\XSWRKHDYHQDQGIHHWEXWLWFRXOGQRWMXPSGRZQ to earth. It laid nine pairs of white eggs and they became the male and female deities, the greater and lesser headmen, the (male) priest (dongba) and the (female) shaman (sainii), the father and son, the white clan and the black clan. The false and the unreal appeared, too, and they became a black hen. She laid nine pairs of black eggs. They hatched to become nine pairs of demons. ,QWKH¿UVWJHQHUDWLRQWKH1LQH%URWKHU*RGVWULHGWRRSHQWKHKHDYHQV but could not. The lofty heights would not stay up. The Seven Sister Gods tried to extend the earth, but could not, and the earth remained wet.3 In the next generation, the Nine Brother Gods studied to become the artisans who opened the heavens. The Seven Sister Gods studied to be the crafts persons who extended the earth. And at that time they opened the heavens and extended the earth. Next, they supported Mt. Jjuqnalsheello’s east side with a white conch prop; they supported the south side with a green turquoise prop; they supported the west side with a black carnelian prop; they supported the north side with a yellow golden prop; in the middle between heaven and earth they erected a white iron prop. Oh, Mt. Jjuqnal-sheelloq! Its head holds the heavens in place; its feet stabilize the they were forced to eat the hides during a time of famine. 2 7KLV VLPSOL¿HG YHUVLRQ RI WKH P\WK LV EDVHG RQ +H

earth.4 Where the heavens were not vaulted, they stuck in boulders of jade WRPDNHLWURXQG:KHUHWKHHDUWKZDVQRWÀDWWKH\SUHVVHGLWGRZQZLWK boulders of gold. Before the myriad things appeared, there came on high a beautiful sound, which begat a sweet breath, and together they produced a vapor and three drops of dew, which became three lakes. From the lakes emerged WKH ¿UVW PDQ +HHOVKHHTKHHOJJXT ZKRVH HLJKWK JHQHUDWLRQ GHVFHQGDQW was Coqsseileel’ee. &RTKDGIRXUEURWKHUVDQGVL[VLVWHUVQRQHRIZKRFRXOG¿QGVSRXVHVDQG so the brothers and sisters—all except for Coq—had sex with one another. This caused heaven and earth and everything in between to become polluted. Day and night became indistinguishable, the rivers rose up out of their banks. The god Dduq said “In three days, the mountains will crash down, and the YDOOH\VZLOO¿OOXSDQGWKHUHZLOOEHDJUHDWÀRRG´ Coq and one of his (incestuous) brothers begged Dduq to help them. 'GXTLQVWUXFWHG&RTWREXLOGDWLJKWGUXP ERDW DQG¿OOLWZLWKXVHIXO WKLQJVDQGKHWROGWKHEDGEURWKHUWRPDNHDORRVHGUXPDQGWR¿OOLWZLWK useless things. On the third day the skies opened and the rains poured down. A blast of lightening smashed the bad brother and his leaky boat, but &RTÀRDWHGVDIHO\DZD\RQWKHULVLQJZDWHUV After many months the waters receded, and Coq landed on a mountain- top. He longed for a wife, and Dduq said to him “There are two celestial sisters: one is beautiful and has vertical eyes; the other is capable and has horizontal eyes. Don’t go for the one with vertical eyes.” But he did choose beauty, and when they returned to earth she gave birth to the pine and the oak, the bear and the boar, the monkey and the pheasant, the frog and the snake. When Coq again approached Dduq for help, Dduq said “You don’t listen so well. Now you need to send all these creatures to the woods and meadows, the streams and lakes, where they belong.” This he did, and in this way the half-siblings of men came to inhabit all the wild places. Coq then went back for the other sister, horizontal eyes, who is also known as Ceil. They met at the Wish-granting Tree between the lands of black and white, where she transformed herself into a crane, and carried him off to her father’s house in heaven.5

4 The image of Mt. Jjuqnalsheello as the square center of the universe and a bridge between heaven and earth by which the ancestors descended is not unique to Naxi cosmology. The mountain appears also in Hindu, Buddhist and Tibetan Bön mythologies under a variety of names, the most common of which are Meru, Kailas and Ti si. Its presence in Naxi FRVPRORJ\XQGRXEWHGO\UHÀHFWVDKLVWRU\RIFXOWXUDOGLIIXVLRQIURP,QGLDDQG7LEHWDQG most directly from the Bön religion, for it is there associated with the demigod, gShen rab mi bo, the Naxi Dibba [= dobbaq] Sheello; see Tucci 1980. As we shall see, however, the image of Jjiuqnalsheello is inscribed with a unique set of meanings in the Naxi cultural context. 5 The Nisu (Yi) of southern Yunnan also speak of a tree on the border between the human and spirit worlds, Dimi and Ngomi, respectively. Along with the Lolopo (Yi) of north central Yunnan, they sometimes also speak of the human and spirit world as two sides of a (market) street ; see Li 2001 and Mueggler 2001. 278 CHARLES F. MCKHANN

Ceil’s father, Zzee’laq-eqpv, did not welcome Coq, and agreed to marry Ceil to him only if he could perform a series of tasks. These appeared impossible—cut down 99 forests in one night, plant them in crops the next, harvest the crops on the third—but with Ceil’s magical intercession, he succeeded. After some further wrangling, during which Coq manages to convince Ceil’s father that his just completed labors should count as bride service, Ceil’s father agrees to the marriage and sends them back down to earth, bearing a sizeable dowry: cattle and sheep, riding horses and pack horses, priests and shamans, and all the other domestic animals and plants, save two (which they stole). They crossed a golden bridge and climbed down a silver ladder and eventually made their way down to the summit of Mt. Jjuqnalsheelloq. From there they descended further, to the foot of the mountain and the place where humans dwell. Coq set up a tent, and Ceil burned incense. They herded the animals and planted the crops. For a long time they tried to conceive children, but could not. Finally, they sent a bat and a dog to heaven to ask advice from Ceil’s father and mother. The two old parents refused to tell the secret, however, and sent the dog and bat on their way. But the animals did not leave, and instead hid, one behind the door, and the other in the rafters. The next day they overheard the father say to the mother: “It is because they do not know how WRVDFUL¿FHWRKHDYHQDQGWRHDUWKWRHUHFWWKHMXQLSHUDQGWKHRDNWRGULYH off the pollution with smoke and water. That is why they cannot conceive.” Hearing this, the dog and bat returned to earth and told all to Coq and Ceil, who promptly performed the rites. Nine months later, Ceil gave birth to three boys, but after three years, they still could not talk. Coq and Ceil again sent the dog and bat to heaven, DQGWKLVWLPHKHUSDUHQWVLQVWUXFWHGWKHPLQRQH¿QDOULWH²WKH6DFUL¿FHWR Meezzei-kolo, Ceil’s mother’s brother.67KLV ¿QDO ULWXDO FRPSOHWHG RQH morning the boys began to speak. They were playing together in the turnip patch in front of the house when a white horse walked over and started to eat. “Hey,” they cried out, “the white horse is eating the turnips,” (Naxi: Rua perq ekeeq kai), in three different languages—Naxi, Tibetan, and Bai. Or as the texts say: “From one mother comes three peoples; from one jug comes three kinds of wine.” The eldest son was Tibetan. He settled above, near Lhasa, and worshipped Tibetan gods in the Tibetan way. The youngest was Bai. He lived below (near Erhai Lake) and worshipped Bai gods in the Bai way. The middle son was Naxi, and he lived in the middle, and worshipped the Naxi gods in the dongba way. The Naxi son was Eeheeq-nol, and his son was No-bbeiqpiu, and his son

6 For many Naxi, the mother’s brother is considered to have preferential rights to marry his sister’s daughter to his own son—a system of delayed (patrilateral cross-cousin) exchange. In the myth, these rights are overridden, and compensation is required. ORIGIN AND RETURN 279

was Bbeiqpiu-oq, and his son was Oq-ga’leiq, and his son was Ga’leiq- cuil. Ga’leiq-cuil and his wife had four sons, and they became the founders of the four primordial Naxi partriclans: Seel, Ieq, Meiq and Hoq. The Seel and the Mei stayed together and migrated to the cool highlands (godiuq) around Lijiang. The Meiq and the Hoq likewise did not separate, and they settled in the hot lowlands (rerdiuq) along the Golden Sand River. “As the father walks, so follows the son. As the mother walks, so follows the daughter. It is not the custom for the woman to climb up on the roof of the house. May the descendants of Coqseei-leel’ee live long and prosper.” The Naxi myth of Coq and Ceil has close parallels among neighboring Tibeto-Burman groups. David Graham (1961:84-88) recorded two 1XRVX Lugu Lake man (2-3,000 B.P.)—and the present day Naxi. This kind of all-encompassing QDWLRQDOLW\KLVWRU\¿QGVLWVPRVWH[WUHPHH[SUHVVLRQLQWKHZULWLQJVRIWKH “Yi Culture School”, whose members (e.g., Liu Yaohan and Liu Zhiyi) have argued a similar thesis regarding Yuanmou Man but go on to suggest ORIGIN AND RETURN 281 that Han culture, religion, writing, and calendar systems may also spring from Yi antecedents, and that Yuanmou Man is the ancestor of all East Asians and Native Americans (Harrell and Li 2003).

ANCESTRAL ROADS

Geographical space is an important component of most dongba rituals. Mountains and rivers, cliffs and valleys, meadows and forests, rocks and springs all have deities associated with them, so that the entire landscape is conceived as inhabited by an array of positive and negative supernatural forces. Every village has an associated mountain god, who is WKHVXEMHFWRIDQDQQXDOVDFUL¿FHZKLOHORFDOZDWHUVRXUFHV HVSHFLDOO\ VSULQJV DUHVLWHVIRUVDFUL¿FHVWRWKHshvq serpent spirits (= nagas, Tib. klu, Prmi lwejabu [Wellens 2006:165-70]). The conception of space associated with ancestors differs from this general picture in that it is profoundly linear: the place where the ancestors dwell and the place where a family now lives constitute opposing poles, between which are the dwelling sites of dozens or even a hundred or more ancestral generations. Similar ideas regarding ancestral roads are found also among neighboring Tibeto-Burman groups, including the Yi, Pumi, and Lisu.7 The most obvious physical representation of the linearity of Naxi roads is the Gods’ Road Map (he reel pi) traditionally used in funeral rituals. The structure and content of funeral rituals varies according to the status of the deceased—age, sex, wealth, rank, and manner of death are among the several variables—but all include some variant of the Gods’ Road Map to mark out the road by which the deceased’s soul is sent off. Most maps are painted hempen cloth scrolls, about a third of a metre wide and 10-12 metres long, depicting various levels of the underworld, human ZRUOGDQGKHDYHQV ¿JXUH Except for its linear structure and the absence of the surrounding ¿JXUHRI

7 Yi myths posit a point of origin somewhere in the vicinity of the Yunnan-Sichuan- Guizhou border, with three pairs of brothers striking out in different directions to found the differing Yi (Ni) subgroups; see Wu 2001, Ma Changshou [in Harrell 1996], and Fan n.d.. The Nisu Yi of southern Yunnan recognize ancestral roads that head north toward Kunming and beyond; see Li 2001. Some Yi groups recognize three roads, of which only the middle one leads to the place of the ancestors; see Fan n.d.. I have also heard of the three roads being associated with different ‘castes’ in Nuosu society. 282 CHARLES F. MCKHANN

Figure 13.1. Naxi Gods’ Road Map, showing demon pole (foreground) and god pole (back) (photograph by Charles McKhann, 1986).

In the Buddhist conception, there are six main modes of existence—as demons, hungry ghosts, animals, human beings, demi-gods and gods. The Naxi scrolls contain representations of these same six categories, but WKHLU OLQHDU SRUWUD\DO DV VHTXHQWLDO µUHDOPV¶ UDWKHU WKDQ µPRGHV¶ UHÀHFWV an indigenous notion of existence in which the idea of reincarnation is downplayed or absent: when a Naxi person dies, his or her soul is sent to WKHODQGRIWKHDQFHVWRUVDQG LQSULQFLSOH VWD\VWKHUH ¿JXUHV 8 8 Exceptions to the general idea that the dead stay put in the ancestral land include annual rituals when ancestors are ‘called back’ to receive offerings from their descendants, and FDVHV ZKHUH WKH VRXOV RI WKH GHFHDVHG DUH GLVVDWLV¿HG IRU VRPH UHDVRQ²XVXDOO\ EHFDXVH WKH\ DUH QRW UHFHLYLQJ DGHTXDWH VDFUL¿FHV RU EHFDXVH WKH\ KDYH GLHG DQ XQSURSLWLRXV death (e.g., by suicide, or away from home)—and take it out on the living. The former resonates with Han ancestor worship involving ancestral tablets and grave sites (hearth and FRXUW\DUGVDUHWKHORFLIRU1D[LDQFHVWUDOVDFUL¿FHV DQGWKHODWWHUZLWK+DQ DQG%XGGKLVW notions that unpropitiated spirits will become ‘hungry ghosts’. Some of the differences from Han practices, include a lack of concern for the material bodies of the deceased— Naxi traditionally cremated their dead, and have only taken up Han ideas of fengshui in the ORIGIN AND RETURN 283

Figure 13.2. Tibetan Wheel of Existence (photograph by Charles McKhann, 1996).

In some places, dongbas use a length of plain hempen cloth in place of the painted scroll, suggesting that the Buddhist iconography of the painted scrolls is a later addition.9 We will look in detail at one of these rituals below. During a traditional Naxi funeral, the Gods’ Road Map is laid out ZLWKWKHGHPRQSROHDI¿[HGWRWKHKHDGRIWKHFRI¿QDQGWKHJRGSROH H[WHQGLQJ DZD\ IURP WKH FRI¿Q LQ WKH GLUHFWLRQ RI WKH DQFHVWUDO URDG that the soul of the deceased must follow. That is, the deceased at once travels from hell ‘up’ to heaven, and from the present ‘back’ to the past. As mentioned above, for most Naxi this also means traveling in a ‘north- erly’ direction, and there is a general opposition in dongba rituals and cosmology between gods and ancestors in the north, and demons and the living in the south. The basic idea of ancestral roads is uniform throughout the Naxi last couple of hundred years—and the idea that the ancestors come to receive offerings by following a particular ‘path’. 9 The diffusion of Buddhism throughout the region—mainly from Tibet—was relatively weak until the early seventeenth century, when the Gelugpa sect achieved prominence under the powerful Fifth Dalai Lama. 284 CHARLES F. MCKHANN territory. The ancestors of every lineage are believed to have migrated from the common ancestral land to the village where the group now resides by stages that are recorded as a series of place names, sometimes numbering a hundred or more.10 When the dongba or daba sends off the soul of the deceased he tells it to listen carefully to his recitation of the list, and to keep its eyes open along the way, so that it does not get lost. After a few preliminary remarks—assuring the deceased that he has everything he needs (clothes, food, a horse, etc.), and instructing him not to be concerned any more about this place, and to not be afraid in his travels—the dongba/daba then recites the list. The place names in the lists are often descriptive of local geographical features (e.g., rivers, mountains, a cliff of a particular color, etc.), so that they call up distinct visual and aural images.11 In his translation of a dongba funeral text, Joseph Rock (1955:198, n.3) notes also the idea that the soul traveling to the ancestral land will revisit all of the places (and events) that the living person experienced. In the text, this idea is presented in stylized form: One day when you were 12 or 13 years old you quarreled with other boys [a different text is used for women], played wild animals, or dogs chasing wild animals. You played with white and black pebbles, and rode a hobby horse made of bamboo. When you were 25 years old you went to cut the green bamboo in the valley, you went to fetch water. Before the summer UDLQV FDPH \RX EXLOW D KRXVH ZHQW WR FXW ¿UHZRRG VSOLW WKH ZRRG IRU shingles and boards; you laid traps for wild animals and herded your sheep RQWKHDOSLQHPHDGRZV

10 Dongba treat the geographical stages as if each represented the movement of a different ancestral generation, but in fact the places named are often quite close together, making this unlikely. This is especially true of places near the current residence, i.e., at the beginning of the list. 11 When in 1996 I made a month-long trip with a 74 year old dongba into the northern reaches of Naxi territory (where all of the ancestral roads ultimately meet), he remarked several times how important it was for him to remember his sensory impressions of the places we visited because he would soon be visiting them again. ORIGIN AND RETURN 285

Figure 13.3. Demon pole of the Gods’ Road Map (photograph by Charles McKhann, 1988). with a great mountain, located somewhere to the north. In the dongba tradition, it is called Jjiuqnalsheel’lojjuq (the mountain of Sheel’lo); the daba of Yongning call it Sibu’alawa (Guo and He 1994:54; Zhan et al. 1980:32; Yan and Song 1983:32). Two of the founding western scholars of Naxi religion—Joseph Rock and Anthony Jackson—identify this (mythological?) mountain with the famous Tibetan holy mountain, Kailas (the Indian [Su]meru), in southwestern Tibet (e.g., Rock 1963:342 and Jackson 1979:95). Contemporary Naxi and Han scholars suggest a place much closer to hand—namely, Gongga Mountain (7556m), located about 250km southwest of Chengdu in Garze Tibetan Autonomous Prefecture in western Sichuan Province (Lin 1970:187; Guo and He 1994:52-56). Both are probably correct. Rock and Jackson argue convincingly that much of the dongba belief and ritual system derives from Tibetan Buddhist and pre-Buddhist (Bön) traditions. 286 CHARLES F. MCKHANN

Figure 13.4. Human realm on the Gods’ Road Map (photograph by Charles McKhann, 1988).

Certainly, the name Jjiuqnalsheel’loq supports this claim: Sheel’loq is the personal name of the man said to have founded the Naxi dongba UHOLJLRQ²'RQJED6KHHO¶ORT5RFNLGHQWL¿HV6KHHO¶ORTZLWKWKH7LEHWDQ ston-pa (teacher) gShen-rab mi-bo, the recognized founder of the Bön tradition, who is thought to have come from someplace in southwestern Tibet—where Mt. Kailas is located (Rock 1972:202). On the other hand, there is considerable evidence that Naxi dongba and Moso daba treat Gongga Mountain as metaphorical of (if not identical to) Mt. Kailas. While many of the place names that constitute the ‘recent’ ends of ancestral roads correspond to known localities, the roads in all cases get fuzzy the further ‘back’ in time (north) one goes, and ultimately they converge on the Shuiluo River valley in southwest Sichuan Province. A third candidate that has emerged from my own research is Yading Mountain (6032m)—also known locally as ‘Gongga’—on the border ORIGIN AND RETURN 287

Figure 13.5. Xiuqiu (the garuda) in the Wish-granting Tree (Haiqyi- bbaddaq-zzer) devouring Shv (nagas). Shv are mountain deities and the half-siblings of humans. They control the wild domain (animals, birds, IRUHVWVHWF IURPZKLFKKXPDQVKDELWXDOO\SRDFK6DFUL¿FHVDUHPDGH to compensate the angry shv, but Xiuqiu also helps to suppress them (photograph by Charles McKhann, 1988). between Muli and Daocheng counties. The mountain is a well-traveled pilgrimage site, in close proximity to known portions of Naxi and Moso ancestral roads (see below). Naxi and Tibetan (Prmi) informants throughout the lower reaches of the Shuiluo and Dongyi river valleys identify the mountain’s three main peaks as representing three ethnic brothers—Tibetan, Naxi, and Han—similar to the Naxi origin myth outlined above.12 Several Naxi and Han scholars have tried to reconstruct the tracks of 12 The relationship between Gongga and Yading mountains in the dongba tradition is unclear. The name ‘Gongga’ appears to be applied locally to a number of mountains regarded as holy sites in southwest Sichuan Province. 288 CHARLES F. MCKHANN

Figure 13.6. God Pole of the Gods’ Road Map (photograph by Charles McKhann, 1988).

Naxi and Moso migration, on the basis of the ancestral roads iterated in funeral and related ceremonies. Figure 13.7 is a composite map based on these sources13DQGP\RZQ¿HOGZRUN DQG Moso or Na of the Yongning region have a slightly different version of the anthropogenic myth. Most Moso reckon descent in the matriline, and identify six primordial (matri)clans (er)—Ci, Hong, Yan, Ngua, Bu and Chu—the last two of which are said to have died out (Shih 1992:103). While Moso agree on the idea that their ancestral homeland is a mountain to the north called “Sibu’alawa” (Guo and He 1994:54; Zhan et al. 1980:32; Yan and Son 1983:32), there is disagreement as to how and when the six clans got from that place to their present locales in the vicinity of Yongning. In one version (Zhan et al. 1980:30-31) they migrated to the Yongning region in the order just given. According to 13 See Li Lin-ts’an 1970, Li Jinchun 1981 [cited in Guo and He 1994:53], Wang Chengxu 1981 [cited in Guo and He 1994:53], Yan et al. 1983:32, n.3, Zhan et al. 1980:32. ORIGIN AND RETURN 289

Figure 13.7. Map of Naxi and Moso ancestral roads. 290 CHARLES F. MCKHANN another (Yan and Song 1983:31-32), the six clans are said to have split into three pairs—Ci and Hong, Yan and Ngua, and Bu and Chu (much like the Naxi story)—which arrived in the Yongning area at different times, and by different roads, which are still recognized in the sending RIIRIWKHVRXOVRISHUVRQVDI¿OLDWHGWRWKHWZRUHPDLQLQJFODQSDLUV The question of primordial ‘clans’ in Naxi and Moso societies is the subject of much debate. Today, these groups exist as names only. Dongbas, dabas and many elders know the clan with which particular OLQHDJHVDUHLGHQWL¿HGEXWGRQRWDWWULEXWHDQ\IXQFWLRQDOYDOXHWRWKHP i.e., marriage, property rights, etc. Rather, clan identity serves mainly to establish a link between the recorded genealogies and ancestral roads of particular lineages, and the mythic past. In this sense, kinship ideology and geography (in the form of roads) appear as parallel tropes, whose main function is to explain human diversity in the present as emergent from a divine unity in the past. The interconnectedness of these tropes is clear also in the association of clans with particular funeral grounds. According to the daba tradition, the four primordial Moso matriclans of the Yongning region, each has its own communal funerary ground, where bones and ashes of the deceased DUHSODFHGDIWHUFUHPDWLRQ7KHIRXUVLWHVDUHORFDWHGRQWKHÀDQNVRIIRXU local mountains, a practice that parallels the traditional Naxi practice of removing the charred remains of cremated dead to particular mountain caves in the direction of the ancestral road from existing villages (Zhan et al. 1980:32; McKhann 1992:283-88)

A TRADITIONAL NAXI FUNERAL TODAY

The vast majority of Naxi live in districts of northwest Yunnan that have EHHQKHDYLO\LQÀXHQFHGE\+DQFXOWXUDOWUDGLWLRQV%XULDOLQH[WHQGHG family plots, fengshui for grave siting, white mourning garb, music to accompany the dead, all are signs of the incorporation of the Naxi into a Han cultural sphere—a process that has been ongoing for hundreds of years. There are a few places where relatively ‘traditional’ funerary practices persist, including the remote valleys of southern Muli County in 6LFKXDQ,GLG¿HOGZRUNLQWKHUHJLRQLQDQGDQGZDV fortunate in 2006 to have been invited to a funeral in Eagleback Village.14 Eagleback is an exception to the rule of ‘sending the ancestors north’. The village was established at the end of the Ming Dynasty

14 Eagleback is a pseudonym. ORIGIN AND RETURN 291

(1368-1644)—during the rule of the Naxi (Lijiang) native chief (tusi) Mu Zeng (1587-1646)—as a military outpost for the protection of Lijiang’s northern frontier. All of the original inhabitants came from the Lijiang basin area (150 km south), and still today they speak a dialect that is closer to the Lijiang dialect than to the northern Naxi dialects that surround them.15 Because of this historical migration, Eagleback villag- HUV¿UVWVHQGWKHLUGHDGEDFNVRXWKWR/LMLDQJZKHUHWKH\³WXUQDURXQG´ and head north again (by the next valley over).16 Wema Bbusso (a pseudonym) was 57 years old when he fell off a FOLIIRQWKHZD\KRPHIURPFROOHFWLQJ¿UHZRRGLQWKHKLJKFRXQWU\QHDU Eagleback in March 2006. As a violent death it was inauspicious, doubly so because his family was poor and he was the only adult male laborer. 1RUPDOO\,ZDVWROGDPDQLQKLVSRVLWLRQZRXOGUHFHLYH¿YHIXQHUDO ULWHVRYHUWKHFRXUVHRI¿YHGD\V:HPD%EXVVRKDGDQH[WUDRQHDWWKH beginning, because he was considered a troubled and possibly trouble- some spirit (due to the manner of his death), and the dongbas compressed two others because the family could not afford the full treatment. As it was, the funeral lasted four days. Along with rice, homemade beer DQGKDUGVSLULWVIRXUFRZVDQGPRUHWKDQDGR]HQJRDWVZHUHVDFUL¿FHG and fed to the scores of mourners who attended the rites. Three senior dongbas organized the rites, and several junior ones helped out. They were paid in meat, wine, tobacco and hand-spun hempen cloth. It was a very expensive undertaking. The six rites performed were: Day 1: Derq Zee (for an violent death) Day 2: Ddaiq Ngvl (for a capable leader) Day 3: Zeesherq Ngvl (for long life) and Neineiq Sher (wealth of the dead) 'D\5HH*Y%HL RSHQWKHURDG DQG6HHO.Y VDFUL¿FHWRWKHKRXVHJRG The morning after Wema Bbusso’s body was found, dongbas went to that place and sent his anguished spirit back to the village (the corpse had been retrieved the day before). The body was washed, rubbed with \DNEXWWHUEXQGOHGLQWRDÀH[HGSRVLWLRQDQGVHZQLQWRDKHPSHQVDFN &RYHUHGZLWKKLV¿QHVWUREHVPRQH\EHOWDQGKDWKHZDVSODFHGRQWKH seating platform near the hearth—in the same place that a mother and 15 Of course, they probably weren’t the original inhabitants. Those were probably Xifan (Prmi), or possibly Kham Tibetans. 16 Although the funeral I attended was ‘traditional’ by comparison to what one would see today in the Lijiang basin, it has some obvious Buddhist elements that are probably attributable to the interactions between the Eagleback Naxi and neighboring Xifan/Prmi and Kham Tibetans. For the last 300 years before 1949, Eagleback was ruled by the Muli tusi, who was Xifan/Prmi, under a Tibetan Buddhist theocracy. To the west and north were Kham Tibetan polities, similarly administered. 292 CHARLES F. MCKHANN her newborn are made to lie during the naming ceremony.17 Beside him were other of his personal items—his pipe, gun and walking stick—and DWDEOH¿OOHGZLWKRIIHULQJVRIPHDWZLQHIDWJUDLQEXWWHUEXWWHUODPSV and incense. Tied to the leg of the table was a clump of braided strings. (YHU\WLPHDQDQLPDOZDVVDFUL¿FHGDQRWKHUVWULQJZDVDGGHGDQGE\ the end of four days there were more than two dozen strings. Rising from the corpse was a long strip of plain hempen cloth, a simpler version of the Gods’ Road Scroll described above. It wound through the rafters, out through the smoke hole and up onto the roof, where it led to a shrine symbolizing Mt. Jjuqnasheel’loq, and thence up to the heavens and an LPDJHRIWKH%XGGKD ¿JXUH DerqDUHKHDGOHVVGHPRQVDVVRFLDWHGZLWKYLROHQWGHDWKDQGWKH¿UVW ULWHLQ%EXVVR¶VIXQHUDOLQYROYHGGULYLQJWKHPDZD\7RWKDWHQGD¿JXUH of a derq and his horse were made of plaited willow and erected out on the terrace. After offerings of blood and burnt grain were made to it, each member of the household symbolically rid himself of the pollution (chel) of this violent death by pressing it on to the demon with his hands, and laying on the demon a plaited willow hoop. A dongba then washed his head and hands with water from a ladle. When all household members had performed this rite, two young men in the role of Iemaq’s (the warrior deity’s) soldiers drove the derq out of the village. A small SLJZDVVFDSHJRDWHGLQWKHSURFHVV$ORQJZLWKWKH¿JXUHRIWKHderq, it was taken down to the river by the tail end of the village, where it was mercilessly stabbed and smashed with boulders. The following day, Ddaiq Ngvl (the funerary rite for a capable leader), was performed. A more detailed model of the Gods’ Road was constructed on the roof of the house. At the demon pole (north) was the image of the yak-headed leiq chel demon. He binds the soul of the deceased with chains and locks him in an iron prison. The dongba secures his release and then chants the soul up through other levels of hell, each overseen by a different leiq chel demon—the cow, goat, ox, tiger, musk deer, frog, crane, and many others. The soul then passes through the human realm, ZKHUH WKH JRGV RI WKH ¿YH FDUGLQDO GLUHFWLRQV UXOH DQG RQ XS WR WKH divine realm, Mt. Jjuqnalsheel’lo and the Wish-granting Tree (Hayi bada ]]HU ¿JXUH The next day’s rituals included Zeesherq Ngvl (for long life) and

17 The seating platform is called ggeqgvlle. From one corner rises up the meeldv (Heaven’s prop), which holds up the central roof beam. An axis mundi iconic of Mt. Jjunalsheel’lo, +HDYHQV3URSLVWKHREMHFWRIQXPHURXVVDFUL¿FHVLQ1D[LKRXVHKROGULWXDO6\PEROLFDOO\ it would seem that the newborn descend by it from heaven, while the dead ascend from the same place. ORIGIN AND RETURN 293

Figure 13.8. Bundled corpse on heath platform, with earthly possessions, offerings of food and wine, and Gods’ Road (hempen cloth strip) leading out through the ceiling to rooftop altar (photograph by Charles McKhann, 2006).

Neineiq Sher (sharing the wealth of the dead). The wealth of the dead was symbolized by plain hempen cloth, hempen garments, colorful woven puttees, and other articles of clothing belonging to the deceased. A dongba made gifts of them to all the immediate descendants, and then blessed them with dew from the Wish-granting Tree. That afternoon, DWDOOSROHZDVHUHFWHGEHVLGHWKHWHUUDFH'XULQJWKLVWLPHRIÀX[DQG for several weeks to come, the soul of the deceased will move back and forth between earth and the heavens, signifying its reluctance to depart WKHZRUOGRIWKHOLYLQJ7KHSROHLVWRSSHGZLWKDÀDJE\GD\DQGDQRLO ODPSE\QLJKWWRKHOSWKHVRXO¿QGLWVZD\EDFNDQGIRUWK 2QWKH¿QDOGD\WKHERG\ZDVSODFHGLQDVTXDUHFRI¿QRXWRQWKHWHU- race, with the offerings beside it, just next to the pole leading to heaven. A model of the God’s Road was again set up, only this time the deceased was positioned much further along on the journey. More offerings were 294 CHARLES F. MCKHANN

Figure 13.9. Gods’ Road depicted in rooftop altar (photograph by Charles McKhann, 2006). made, and the dongbas chanted the texts for Ree Gv Bei (to open the URDG ZKLFKLQFOXGHWKHSODFHQDPHVRIWKHDQFHVWUDOURDG ¿JXUH )LQDOO\WKHFRI¿QZDVFDUULHGWRWKHFUHPDWLRQJURXQGZLWK:HPD Bbusso’s daughters and other close descendants wailing at the rear. They did not enter the cremation ground but turned back at the gate and let the dead in Eagleback’s cremation ground, and Wema Bbusso was taken to the lesser of these, because his death had been so inauspicious. His body ZDV UHPRYHG IURP WKH FRI¿Q SDFNHG LQ ¿UHZRRG VRDNHG LQ RLO DQG burned. At one edge of the cremation ground stands a gnarled old tree (the Wish-granting Tree between earth and heaven), draped in weather worn clothes, blankets and other personal effects of the dead who have SDVVHGWKLVZD\ ¿JXUH 7KH¿QDOULWHZKLFK,GLGQRWDWWHQGZDV SHUIRUPHGODWHWKDWHYHQLQJEDFNDWWKHKRXVH6HHO.YLVWKH6DFUL¿FH to the House God and it is performed on many occasions, but especially when members join or leave the household, by birth, marriage or death. ORIGIN AND RETURN 295

)LJXUH&RUSVHLQFRI¿QRQWHUUDFHZLWK*RGV¶5RDG (photograph by Charles McKhann, 2006).

When performed in conjunction with a funeral, its purpose is to protect the other members of the family. Their love for the deceased puts their own souls in jeopardy. As his soul departs for the land of the ancestors, theirs may be drawn along as well. The House God holds them back, protecting and nourishing them, until their own time has come.

CONCLUSION

Taken together, the myth of Coq and Naxi funerary rituals posit a process of origin and return that is metaphorically linked to the relations between gods and humans, heaven and earth, ancestors and the living, wife- givers and wife-receivers, and physically inscribed on the landscape of northern Yunnan and southern Sichuan. In the idea of ancestral roads, time is expressed in a spatial metaphor, and death is not a journey into the 296 CHARLES F. MCKHANN

Figure 13.11. Wish-granting Tree at village cremation ground (photograph by Charles McKhann, 2006). unknown, but a return to the familiar. The road is long and dangerous, and every effort is made by the living to provide for the deceased’s safety and comfort. The process also involves grave risks for the living, however, and care is taken to insure that the souls of the dead are detached from those of the living and return to the ancestral homeland alone. ORIGIN AND RETURN 297

BIBLIOGRAPHY

Fan Xiuli. N.d. “A Consideration about the Function of Funeral Ceremonies for Human-building in the Society of the Yi in China” Unpublished paper presented at the Second International Conference on Yi Studies, Trier University, Trier, Germany, June 19-23, 1998. Graham, D.C. 1958. The Customs and Religion of the Ch’iang. Smithsonian Misc. Coll., vol.135, no.1, Washington, D.C. —— 1961. Folk Religion in Southwest China. Smithsonian Misc. Coll., vol.142, no.2, Washington, D.C. Guo Dalie and He Zhiwu. 1994. Naxizu Shi. Chengdu: Sichuan Nationalities Press. Harrell, S. 1996. “The History of the History of the Yi.” In S. Harrell (ed.), Cultural Encounters on China’s Ethnic Frontiers. Seattle: University of Washington Press, pp.63-91. Harrell, S. and Li Yongxiang. 2003. The History of the History of the Yi, Part II. Modern China 29(3): 362-96 He Yuncai. 1986: La Leeq Keeq Coq Bber Co Zzoq [The Migration of Coq, from the Wind Ceremony Texts]. YSSMGZCGB 1986:151-279. He Zhiwu. 1983. Naxi Dongba Jing Xuanze [Selections from the Naxi Dongba Classics]. Lijiang, Yunnan: Yunnan Sheng Shehui Kexue Yuan Dongba Wenhua Yanjiu Shi [Yunnan Province Academy of Social Sciences, Dongba &XOWXUH5HVHDUFK2I¿FH@ Jackson, A. 1979. Na-khi Religion. The Hague: Mouton. Levi-Strauss, C. 1963. “The Structural Study of Myth.” In Structural Anthropology. New York: Basic Books, pp.206-31. Li Lin-tsan. 1970. Mosuozu de Gushi. Asian Folklore and Social Life Monographs. vol.3. Taipei: Dongfang wenhua gong ying she. Li Yongxiang. 2001. “The Cold Funeral of the Nisu Yi.” In S. Harrell (ed.), Perspectives on the Yi of Southwest China. Berkeley: University of California Press, Berkeley, pp.135-43. McKhann, C. 1992. Fleshing Out the Bones: Kinship and Cosmology in Naqxi Religion. Ph.D. dissertation, University of Chicago. —— 1995. “The Naxi and the Nationalities Question.” In S. Harrell (ed.), Cultural Encounters on China’s Ethnic Frontiers. Seattle: University of Washington Press, pp.39-62. —— 2003. Sacred Trails: Genealogical Mapping and the Creation of Historical Space Among the Naxi of Southwest China. Histoire et Anthropologie Asies 2: 29-47. Mueggler, E. 2001. The Age of Wild Ghosts: Memory, Violence, and Place in Southwest China. Berkeley: University of California Press. Rock, J.F. 1935. The Story of the Flood in the Literature of the Mo-so (Na-khi) Tribe. West China Border Research Society Journal 7: 64-80. —— 1948. The Muan Bpo Ceremony. Monumenta Serica, vol.13. —— 1952. The Na-khi Naga Cult and Related Ceremonies (2 parts). Roma: Istituto Italiano per il Medio ed Estremo Oriente (Serie Orientale Roma, 4:1-2). —— 1955. The Zhi ma Funeral Ceremony of the Na-khi of Southwest China. Studia Instituti Anthropos, vol.9. 298 CHARLES F. MCKHANN

—— 1963. A Na-khi—English Encyclopedic Dictionary. Roma: Instituto Italiano per il Medio ed Estremo Oriente (Serie Orientale Roma, 28). —— 1972. A Na-khi—English Encyclopedic Dictionary, Part 2. Roma: Instituto Ilaliano per il Medio ed Estremo Oriente (Serie Orientale Roma, 28:2). Shih Chuan-kang. 1992. The Yongning Moso. Ph.D. dissertation, Stanford University. Wellens, K. 2006. Consecrating the Premi House: Ritual, community and the state in the borderlands of East Tibet. Ph.D. dissertation, University of Oslo. Wu Jingzhong 2001. “Nzymo as Seen in Some Yi Classical Books.” In S. Harrell (ed.), Perspectives on the Yi of Southwest China. Berkeley: University of California Press, pp.35-48. Yan, Ruxian and Z.L. Song. 1983. Yongning Naxizu de Muxi Zhi. Kunming: Yunnan People’s Press. Zhan, Chengxu, C.Q. Wang, J.C. Li and L.C. Liu. 1980. Yongning Naxizu de Azhu Hunyin he Muxi Jiating. Shanghai: Shanghai People’s Press. MIGRATING BROTHERS AND PARTY-STATE DISCOURSES ON ETHNIC ORIGIN IN SOUTHWEST CHINA

KOEN WELLENS

INTRODUCTION

This essay will demonstrate that the categorisation by the majority of several ethnic minority groups in Southwest China also brings with it certain versions of the histories of these groups. These histories stress ancient north-south migrations and a strong process of ‘ethnic amalga- PDWLRQ¶OHDGLQJWRµ7LEHWDQLVDWLRQ¶&RQVWLWXWLQJWKHRI¿FLDOGLVFRXUVHRI the modern Chinese Party-state, these versions of history acquire dominant roles in reshaping these ethnic groups’ own myths of origin. This is SRVVLEOH¿UVWRIDOOEHFDXVHORFDOPRYHPHQWVRILQGLYLGXDOVDQGVPDOO scale migrations do indeed play a role in the transmitted histories of the different communities. Secondly, this study will argue that traditional narrative themes or mythologising templates, such as that of the ‘migrat- LQJEURWKHU¶IXQFWLRQDVHIIHFWLYHPHFKDQLVPVIRULQWHUQDOLVLQJRI¿FLDO discourses within the corpus of myths told and transmitted by members of these ethnic minority communities. 7KH 3HRSOH¶V 5HSXEOLF RI &KLQD 35& WRGD\ LV RI¿FLDOO\ D PXOWL QDWLRQDO VWDWH ZLWK ¿IW\VL[ QDWLRQDOLWLHV RU PtQ]~. Besides the Hàn &KLQHVHPDMRULW\WKH3DUW\VWDWHUHFRJQLVHV¿IW\¿YHPLQRULW\QDWLRQDOL- ties or VK΁RVKPtQ]~$QHWKQLFFODVVL¿FDWLRQSURMHFWLQWKHHDUO\35& period constituted the basis for identifying ethnic groups and determin- LQJZKLFKRQHVZRXOGEHUHFRJQLVHGDVRI¿FLDOPLQRULW\QDWLRQDOLWLHV $OWKRXJKWKHFODVVL¿FDWLRQSURMHFWZDVLQIRUPHGE\DVXEVWDQWLDODPRXQW of serious ethnographic studies, the top-down nature of this exercise LQQDPLQJDQG¿[LQJHWKQLFFDWHJRULHVQDWXUDOO\UHVXOWHGLQQXPHURXV LQVWDQFHVZKHUHORFDOFRQFHSWVRIHWKQLFLGHQWLW\ZHUHLQFRQÀLFWZLWK WKHQHZRI¿FLDOODEHOVDVVLJQHGE\WKH3DUW\VWDWH These national minority labels did not stand on their own but were HTXLSSHGZLWKRI¿FLDOYHUVLRQVRIWKHKLVWRU\DQGFXOWXUHRIWKHSHRSOH WKH\ UHIHUUHG WR 5HFXUULQJ HOHPHQWV LQIXVLQJ WKHVH RI¿FLDO YHUVLRQV were, for example, attributing nationalities with a long-standing history of integration with the imperial predecessors of the modern Chinese state and thereby legitimating the inclusion of the area inhabited by 300 KOEN WELLENS border nationalities within the territory of the PRC. Other characteristics were less politically motivated and more the result of epistemological traditions within Chinese intellectual history such as an extreme reli- ance on written Chinese sources. Such traditions became paired with the Marxist conviction that the laws governing the development of human society were as strict and predictable as those of the natural sciences. $VDFRQVHTXHQFHZKHQWKH3DUW\VWDWHGH¿QHGHWKQLFERUGHUVLWZDV more prone to rely on ‘objective’ factors such as linguistic similarities or LGHQWL¿FDWLRQRIDSRSXODWLRQZLWKDFDWHJRU\GHVFULEHGLQDQFLHQWWH[WV WKDQRQµVXEMHFWLYH¶IDFWRUVVXFKDVDORFDOJURXS¶VVHOILGHQWL¿FDWLRQ A recurring conclusion in several recent studies on ethnic groups in &KLQDLVWKDWHOHPHQWVRIWKHRI¿FLDOYHUVLRQRIWKHKLVWRU\RIDPLQRULW\ nationality have a tendency to become part of the common perceptions and cosmologies of the people carrying the nationality label. Such ele- PHQWVFDQEHFRPHHDVLO\LQWHJUDWHGEHFDXVHWKHRI¿FLDOYHUVLRQLVLWVHOI often selectively using elements of local oral history to substantiate its claims. Not a few mythological stories of origin and migration told among the Namuyi and Premi, two Tibeto-Burman-speaking groups in Southwest China, contain references to Tibet or Tibetan ancestors. When WKH&KLQHVH3DUW\VWDWHGHFLGHGWRRI¿FLDOO\FODVVLI\WKH1DPX\LDQGD large part of the Premi as Tibetans, these people were provided with new histories—histories that reconnected with traditional orally transmitted stories and became integrated with them. The Namuyi and Premi, together with several other more or less distinct ethnic groups in Southwest China, were known by the local Hàn &KLQHVH DV ;ƯIƗQ 7KHVH JURXSV DUH QXPHULFDOO\ VPDOO ZLWK SUHVHQW populations ranging from only a few hundred to a few tens of thousands members. This renders their further survival as culturally distinct entities within the PRC highly unlikely. Present policies of forced resettlement and voluntary labour migration are quickly eroding their possibilities to act concertedly in order to try to adapt their cultural traditions to a mod- ernising society. Unlike other groups—such as the Naxi in Yúnnán—who have been successful in utilising some of the limited advantages offered E\EHLQJHTXLSSHGZLWKDQRI¿FLDOQDWLRQDOPLQRULW\ODEHOVPDOOXQUHF- ognised groups such as the Namuyi have no such possibility. Claiming space for the use of their language or the practice of their non-Buddhist UHOLJLRQIRUH[DPSOHZRXOGHQWDLODUHMHFWLRQRIWKHLUDVVLJQHGRI¿FLDO status as Tibetans, something which is presently not an option in China. In this way PtQ]~FODVVL¿FDWLRQFDQQRWEHUHGXFHGWRDPHUHPDWWHURI authoritarian naming. MIGRATING BROTHERS 301

',63(16,1*+,6725,(6)520;Ʈ)Ɩ172=¬1*

As to the Zàng Nationality within the territory, originally there were all NLQGVRIDSSHOODWLRQV=jQJµ1DPX\L¶DQGWKHH[RQ\Pµ;ƯIƗQ¶$IWHUD WKRURXJKLQYHVWLJDWLRQE\WKHµ;ƯIƗQ,GHQWL¿FDWLRQ:RUNLQJ*URXSRIWKH (WKQLF$IIDLUV&RPPLWWHHRI6uFKXƗQ3URYLQFHFDUULHGRXWEHWZHHQWKHWK RI$XJXVWDQGWKHWKRI2FWREHULQ6KtPLiQ;ƯFKƗQJ²LQDOOQLQH counties—and based on the evidence collected, it was decided that they ZHUH=jQJ1DWLRQDOLW\ ;ƯFKƗQJVKu]KuELƗQ]XӽQZHӿ\XiQKXu The Zàng Nationality is the same ethnic label utilised for those people within the borders of the People’s Republic of China who call themselves Bod-pa, the appellation that has been translated as ‘Tibetan’ in :HVWHUQODQJXDJHV7KH=jQJ1DWLRQDOLW\LVQRZRQHRIWKH¿IW\¿YH RI¿FLDOO\UHFRJQLVHGPLQRULW\QDWLRQDOLWLHVLQ&KLQDZKLOHWKHVRFDOOHG Hàn Chinese constitute the default ethnic majority. The above quota- tion suggests that the project of dispensing ethnic labels was a rather undemocratic exercise in administration with limited room for dis- senting opinions at the grassroots level. This quote can be found in the FKDSWHURQWKH=jQJ1DWLRQDOLW\LQWKH 1HZ &RXQW\$QQDOVRI;ƯFKƗQJ &LW\'LVWULFW,QWKLVFRQWH[WWKHWHUP;ƯIƗQUHIHUUHGWRDFROOHFWLYHODEHO used by local Hàn Chinese to designate several distinct groups of ethnic PLQRULW\SHRSOHLQZKDWDUHQRZWKH&KLQHVHSURYLQFHVRI6uFKXƗQDQG Yúnnán. To the local Chinese—and other ethnic groups such as the Nuosu—these numerically small peoples resembled Tibetans in their outward appearance and cultural practices but were at the same time XQGHUVWRRGWREHGLIIHUHQWIURPµUHDO¶7LEHWDQV2I¿FLDOHWKQRJUDSKHUV of the new communist regime seemed more interested in the similarities WKDQLQWKHGLIIHUHQFHVWKRXJKDQGWKHFDWHJRU\RI;ƯIƗQZDVGXPSHG on the scrap heap of defunct historical terminology. As a consequence, ZLWKIHZH[FHSWLRQVDOOWKH;ƯIƗQLQWKLVFRUQHURIVRXWKZHVWHUQ&KLQD became Tibetans or, more precisely, members of the Zàng Nationality. Rather than replacing a little nuanced Chinese exonym with several dis- WLQFWFDWHJRULHVUHÀHFWLQJWKHHWKQLFLGHQWLWLHVDQGWKHFXOWXUDOGLYHUVLW\ RIWKHGLIIHUHQWSHRSOHVVKDULQJWKHH[RQ\PRI;ƯIƗQWKHQHZUHJLPH opted for an even less nuanced category of ethnic representation. 6R DQ REYLRXV TXHVWLRQ ZRXOG EH GR WKH GLIIHUHQW RI¿FLDO ODEHOV granted by the PRC to members of an ethnic minority population have consequences? While discriminatory practices or preferential treatment SROLFLHVFDQEHSRVLWLYHRUQHJDWLYHRXWFRPHVRIEHLQJRI¿FLDOO\UHF- ognised as an ethnic minority, one would expect that the exact name 302 KOEN WELLENS which the state has decided on should be of less importance. There are IHZ GLUHFW SUDFWLFDO FRQVHTXHQFHV RI EHLQJ RI¿FLDOO\ UHFRJQLVHGDVD minority nationality: your group has a representative in the National People’s Congress, and although this has limited practical consequences it has a certain symbolic value. If the minority nationality you belong to has been granted an autonomous minority area, one of the few tangible advantages is the possibility for the recognised group of the autonomy area to organise education in its own language. Nevertheless, there are several consequences that are more indirect but no less important. In an authoritarian one-party state such as China, state discourses on ethnic identity are very powerful and pervasive when conveyed via state education, the media, and—even to some extent—the social sciences. While LQUHFHQW\HDUVGLYHUJLQJLGHDVDQGRSLQLRQVDUH¿QGLQJVRPHOHHZD\ in an emerging civil society, the many decades during which the Party- state has had a near monopoly on proclaiming its version of China’s ethnic make-up have left a deep impression. One only has to consider the uniformity with which Chinese people have expressed their views of Tibetan people, Tibetan culture or Tibetan history following the March 2008 protests to understand the widespread and penetrating impact of WKHRI¿FLDOGLVFRXUVHLQ&KLQD Ethnic minorities themselves have also been strongly affected by the hegemonic views of the Party-state concerning history and ethnic identity in China. A considerable component of the Western scholarship on ethnic minorities in China has been concerned with the relationship between the labels given by the state and the way such labels and their connotations KDYHLQÀXHQFHGFRQFHSWVRIHWKQLFLGHQWLW\DPRQJWKRVHZKREHDUWKHP1 One general conclusion that can be drawn from these studies is that the RI¿FLDO FODVVL¿FDWLRQ KDV SURPRWHG RU HYHQ FUHDWHG PRGHUQ IRUPVRI ethnic identity which did not previously exist or had a different content, and which now acquire a dynamic existence of their own. Unfortunately for those concerned, the hegemonic manner in which the Chinese Party- VWDWHHQIRUFHVWKHRI¿FLDOPDSRIFXOWXUDOGLYHUVLW\ZLWKLQWKHWHUULWRU\ can turn the local expression of alternative versions into a politically VHQVLWLYHXQGHUWDNLQJ&RPPXQLWLHVPLJKW¿QGLWKDUGWRUHYLYHSURPRWH or develop aspects of their own local culture if these activities imply a GHQLDORUFULWLTXHRIWKHRI¿FLDOYHUVLRQRIWKHLUHWKQLFLGHQWLW\ This chapter is concerned with a number of different ethnic categories,

1 Important contributions include the monograph by Kaup 2000 and the different chapters found in two major publications on Chinese ethnic minorities; see Brown 1996 and Harrell 1995. MIGRATING BROTHERS 303

VRPH ZKLFK DUH FRQWHVWHG E\ SHRSOHV WKH\ FRXOG DSSO\ WR 7KH ¿YH FDWHJRULHVEHORZFHUWDLQO\GRQRWFRLQFLGHZLWKWKHRI¿FLDO3DUW\VWDWH version. Rather, they are based on existing autonyms and local concepts of common descent and identity, which—not surprisingly—coincide largely with prevalent linguistic boundaries: 1. The Premi number around 60,000 and live in an area stradling WKH&KLQHVHSURYLQFHVRI<~QQiQDQG6uFKXƗQ3UHPLOLYLQJLQ6uFKXƗQ DWWKHWLPHRIHWKQLFFODVVL¿FDWLRQEHFDPHFODVVHGDV=jQJRU7LEHWDQV OLNHWKHRWKHU;ƯIƗQ7KRVH3UHPLOLYLQJLQ<~QQiQRQWKHRWKHUKDQG FRQVWLWXWHGWKHPDLQH[FHSWLRQDPRQJWKH;ƯIƗQIRUEHLQJUHFRJQLVHG DV D VHSDUDWH PLQRULW\ QDWLRQDOLW\ WKH 3ԃPӿ]~ RU 3ԃPӿ 1DWLRQDOLW\2 As I have argued elsewhere, a separate ethnic label proved to be a great asset for the Premi living in Yúnnán, not only in maintaining, but also in promoting cultural distinctiveness.3 This not only provided a certain positive self-consciousness for a relatively poor minority population, it also created opportunities for participating in the economic fruits of ethnic tourism and functioned as a social glue for local communities in a rapidly developing Chinese modernity. 2. The contrast with the Premi’s close neighbours, the Namuyi,4 FRXOG QRW EH PRUH H[WUHPH )ROORZLQJ WKH RI¿FLDO VFULSW WKH\ LQVLVW that they are Tibetans, stressing myths that they originate from Lhasa. Living in a small isolated area stradling the counties of Muli, Jiulong, Mianning and Xichang—all in Sichuan—many of the approximately 5,000 Namuyi hold that they have lost many of the important attributes of Tibetan culture and that there are consequently few reasons to stick to their local traditions. A recent policy of resettlement in Namuyi areas, and the resulting dispersal of local communities because of hydropower development, constitutes an added threat to the survival of the unique Namuyi culture. 3. Another ethnic category mentioned here are the Na or Naze. In 6uFKXƗQ WKH +jQ PRVW RIWHQ OXPSHG WKH 1D WRJHWKHU ZLWK WKH RWKHU ;ƯIƗQJURXSVZKLOHLQ<~QQiQWKH\ZHUHNQRZQDV0yVXǀZKLFKLV\HW another Chinese exonym.5 They became rather well-known within the

27KDWLVLIZHGRQRWFRQVLGHUWKHPRGHUQ4LƗQJRI¿FLDOQDWLRQDOPLQRULW\OLYLQJLQWKH QRUWKRI6uFKXƗQDQGWKHVPDOOJURXSRIWKH1DERWKRIZKLFKVRPHWLPHVZHUHUHIHUUHGWR DV;ƯIƗQE\WKH+jQ&KLQHVH 3 See Wellens 2009. 4 ‘Namuyi’ is the most commonly used way of writing the name of the group in both :HVWHUQ DQG &KLQHVH ZULWLQJV ,Q ;LӽQJVKXӿ7RZQVKLS LQ ;ƯFKƗQJ 0XQLFLSDOLW\ ZKHUH , FDUULHGRXW¿HOGZRUNLQDQGWKHDXWRQ\P1DPX]LZDVXVHG 5/yQJ;ƯMLƗQJOLVWVWKH1DLQ6uFKXƗQDV0H[LƝ;ƯIƗQDFRPELQDWLRQRIWKHWZRH[RQ\PV see Long 1991:38. 304 KOEN WELLENS anthropology of Southwest China for the matrilineal system prevalent in some Na communities. The Na live in close contact with their other ;ƯIƗQ QHLJKERXUV VXFK DV WKH 3UHPL DQG 1DPX\L VRPHWLPHV VKDULQJ villages and frequently intermarrying. 2WKHU ;ƯIƗQ JURXSV EULHÀ\ PHQWLRQHG KHUH DUH WKH 0LQ\DN RU 0X\DWKH-LDURQJDQGWKH5PDRU4LƗQJOLYLQJWRWKHQRUWKRIWKHPDLQ focus in this chapter. )LQDOO\ WKHUH DUH WKH QRQ;ƯIƗQ HWKQLF FDWHJRULHV WKDW FRQVWL- WXWH WKH PDMRULW\ QHLJKERXUV RI WKH GLIIHUHQW ;ƯIƗQ JURXSV WKH+jQ the Kham-Tibetans and the Nuosu.6 The Nuosu are also speakers of a 7LEHWR%XUPDQODQJXDJHEXWZKLOHGLIIHUHQW;ƯIƗQJURXSVRIWHQKDYH close relations and intermarry, the Nuosu rarely marry members of other ethnic groups. The Nuosu population is relatively large—it counts several million—and they have been recognised by the Chinese state as WKH

Chinese ethnic group or/and VWDWHFODVVL¿FDWLRQ Nuosu appellation/ autonym PtQ]~ appellation exonym 3ԃPӿ]~ Premi, Ch’ruame ;ƯIƗQ Ozzu (in Yúnnán) Zàng-zú — LQ6uFKXƗQ Namuyi, Namyi, Zàng-zú ;ƯIƗQ Ozzu Namuzi 1j[Ư]~ Na, Naze 0yVXǀ Ozzu (in Yúnnán) 0ČQJJԃ]~ — 0yVXǀ;ƯIƗQ LQ6uFKXƗQ Minyak, Muya Zàng-zú ;ƯIƗQ Ozzu Jiarong Zàng-zú ;ƯIƗQ Ozzu (Rgya-rong) Rma 4LƗQJ]~ ;ƯIƗQ4LƗQJ ? Nuosu

)LJXUH&KDUWRIWKHGLIIHUHQW;ƯIƗQDXWRDQGH[RQ\PV throughout Chinese history when populations became assimilated, exterminated or displaced in the westward expansion of Chinese agrarian society. This ethnic boundary continued to shift until the Chinese met an ‘ecological frontier’ where they encountered people who subsisted in an environment most suitable for herding and nomadism and who KDG D PRUH ÀH[LEOH VRFLDO RUJDQLVDWLRQ :DQJ ,W LV WKHUH- IRUHYHU\GRXEWIXOWKDWWKHPHPEHUVRIWKHRI¿FLDOO\UHFRJQLVHG4LƗQJ PLQRULW\QDWLRQDOLW\OLYLQJWRGD\LQWKHQRUWKRI6uFKXƗQ3URYLQFHDUH GLUHFWGHVFHQGDQWVRIWKH4LƗQJIURPWKH+jQRU6KƗQJG\QDVW\VRXUFHV )XUWKHUPRUHFRQVLGHULQJWKHGH¿FLHQWNQRZOHGJHZHKDYHWRGD\DERXW WKHHWKQLFFXOWXUDODQGOLQJXLVWLFPDNHXSRI:HVWHUQ6uFKXƗQLWLVYHU\ unlikely that the historical categorisations by the Chinese, at times when this region was considerably less accessible, constitute a sound basis for DFFXUDWHLGHQWL¿FDWLRQRIHWKQLFDQFHVWRUVRISUHVHQWGD\SRSXODWLRQV 7KHWHUP;ƯIƗQLVVLPLODUWRWKHWHUP4LƗQJLQWKDWLWFRQVWLWXWHVD Chinese exonym for very different populations at different places in what is now Western China and during different periods of Chinese history. In VRPHFDVHVLWKDVEHHQXVHGLQDVLPLODUZD\WR4LƗQJWRGHQRWHDVKLIWLQJ ethnic border to the west of an expanding Chinese civilisation. While at 306 KOEN WELLENS

WLPHV ;ƯIƗQ DOVR LQFOXGHG WKH %RGSD RU 7LEHWDQV GXULQJ WKH HWKQLF FODVVL¿FDWLRQSURMHFWRIWKHHDUO\3HRSOH¶V5HSXEOLFRI&KLQDWKHDSSHO- ODWLRQ;ƯIƗQIXQFWLRQHGLQ:HVWHUQ6uFKXƗQDQG1RUWKZHVWHUQ<~QQiQLQ DPRUHVSHFL¿FZD\,WZDVXVHGDVDQH[RQ\PE\WKHORFDO+jQ&KLQHVH to designate peoples who were living wedged in between the Tibetans, WKH+jQDQGWKHSRSXODWLRQVZKREHFDPHFODVVL¿HGDVWKHGelug sect of Tibetan Buddhism. Besides being the head of the three major PRQDVWHULHVFRQVWLWXWLQJ0OӿKHZDVDOVRWKHZRUOGO\UXOHURIWKLVWHUUL- tory. While his family and other members of the local elite were highly LQÀXHQFHG E\ 7LEHWDQ FXOWXUH WKH PDMRULW\ RI KLV VXEMHFWV FRQWLQXHG to speak the Tibeto-Burman Premi language. Although these subjects were forced to send sons to become monks at the local Tibetan Buddhist Gelug monastery, in their villages they continued to practice their own shamanic religion based upon the worship of ancestral spirits and nature deities (see Wellens 2010). The Namuyi, on the other hand, were for the most part living in areas MIGRATING BROTHERS 307 ruled by non-Buddhist Nuosu W·VƯ who had been outside of the Tibetan Buddhist sphere for a long time. Although using Tibetan Buddhist texts in their rituals, the Namuyi were not able to understand their meaning and they did not participate in monasticism. Except for the ritual use of Buddhist texts and some other Buddhist ritual paraphernalia, their religious tradition was clearly distinct from any of the prevalent Tibetan Buddhist schools. ,IZHLPDJLQHDFRQWLQXXPRIGHJUHHVRILQWHJUDWLRQE\;ƯIƗQFRPPXQL- ties within the larger sphere of Tibetan culture, the Namuyi would be situated at the end of least integration and the Premi would be somewhere in WKHPLGGOH$WWKHRSSRVLQJHQGRIVWURQJHVWLQWHJUDWLRQZHZRXOG¿QGWKH speakers of Minyak and Jiarong (Rgyal-rong). Unlike the Premi people, ZKRVHFRQWDFWZLWK7LEHWDQ%XGGKLVPGDWHVRQO\IURPWKH¿IWHHQWKFHQ- WXU\%XGGKLVPZDV¿UPO\HVWDEOLVKHGDPRQJWKH0LQ\DNDOUHDG\VHYHUDO centuries earlier. Although the Minyak language is today only spoken by around 10,000 people, it has survived and is clearly different from the surrounding Kham dialects. The Jiarong language, spoken to the north of the 0LQ\DNDUHD VHH¿JXUH KDVEHHQPXFKPRUHLQÀXHQFHGE\7LEHWDQ8 ,QVSLWHRIVWURQJ7LEHWDQLQÀXHQFHERWKWKH0LQ\DNDQGWKH-LDURQJKDYH nevertheless survived until today as separate cultures. $OWKRXJKWKHHWKQLFFODVVL¿FDWLRQSURMHFWZDVLQJHQHUDOGUDZLQJKHDY- LO\RQWKHUHVXOWVRIOLQJXLVWLFVWXGLHVLQWKHFDVHRIWKH;ƯIƗQWKLVFOHDUO\ ZDVQRWWKHFDVH7KHPDLQUHDVRQLVWKDWDWWKHWLPHRIWKHFODVVL¿FDWLRQ such linguistic data were not available.9 As one of the last white spots on WKHOLQJXLVWLFPDSRI&KLQDWKH7LEHWR%XUPDQODQJXDJHVRIWKH;ƯIƗQ became an El Dorado for linguists only during the last two decades of the twentieth century. But even today there still exist controversies over ODQJXDJH FODVVL¿FDWLRQ RI WKH GLIIHUHQW ;ƯIƗQ ODQJXDJHV PRVW DJUHH WKDWDOO;ƯIƗQODQJXDJHVEHORQJWRWKH4LDQJLF*URXSZLWKLQWKH5XQJ or Jiarong/Drung Branch of Tibeto-Burman.10 While we can of course 8 See, for example, Nicolas Tournadre’s Sino-Tibetan hierarchical scheme at the Tibetan DQG +LPDOD\DQ 'LJLWDO /LEUDU\ ZHUH *\DURQJLF LV WRJHWKHU ZLWK %DLPD FODVVL¿HG DV D µ4LDQJLFODQJXDJHKHDYLO\LQÀXHQFHGE\7LEHWDQ¶KWWSZZZWKGORUJ[POVKRZSKS"[PO collections/langling/languages/index.xml&l=7 97KH;ƯIƗQRQO\UHFHLYHGWKHLUµ¿QDO¶FODVVL¿FDWLRQLQWKHEHJLQQLQJRIWKHV7KLV ZDV MXVW EHIRUH WKH PDMRU ODQJXDJH VXUYH\ RQ :HVWHUQ 6LFKXDQ ZDV SXEOLVKHG E\ 6ǌQ +yQJNƗL 10 Most studies list the following languages: Qiang, Ergong, Ersu (containing the three distinct dialects of Ersu, Duoxu and Liru), Guiqiong (sometimes Daofu), Shixing, Zhaba, 1DPX\L0X\D RU0LQ\DN DQG-LDURQJ RU*\DURQJ VHH6ǌQ+yQJNƗL'jL4uQJ[Lj 0DWLVRII'jL4uQJ[LjYLHZV-LDURQJDVDVHSDUDWHODQJXDJHXQGHUD different group (the Jiarong Group) in regard to the Qiangic Group, but places both groups under the Jiarong, Drung Branch of the Northern Language Family of Tibeto-Burman. There 308 KOEN WELLENS

Figure 14.2. Map showing the Qiangic languages (adapted from 6ǌQ+yQJNƗL not equate separate languages or dialects with single bounded cultural JURXSV WKH IHZ HWKQRJUDSKLF VWXGLHV LQ VRPH RI WKH ;ƯIƗQ LQKDELWHG areas seem to suggest that language often coincides with other cultural PDUNHUVVXFKDVUHOLJLRXVSUDFWLFHWUDGLWLRQDOGUHVVRUHQGRJDP\ /yQJ ;ƯMLƗQJ/L~/yQJFKǌ/DNKL+HIULJKWDQG6WXDUW Linguistic studies clearly demonstrate that including the Qiangic (and maybe even some Lolo-Burmese, see note 10) speakers in this corner of the PRC within the category of Tibetans was a rather meaningless exer- FLVHLIWKHFODVVL¿FDWLRQZDVPHDQWWRUHSUHVHQWFOHDUO\GLIIHUHQWFXOWXUDO HQWLWLHVZLWKLQ&KLQD(YHQWKRXJKWKHORFDOHWKQLFODEHORI;ƯIƗQZDVDQ externally ascribed identity marker, ignoring important cultural differences and concepts of ethnic identity, it is still not entirely clear why the Party-state choose to discard it in favour of an even less representative FDWHJRU\7KH XVH RI ;ƯIƗQ DW OHDVW KDG WKH PHULW RI PDNLQJ YLVLEOH D diverse but numerically small collection of separate cultures.

is some disagreement on the Namuyi Language, with some linguists placing it with the Lolo- Burmese Group, together with Naxi; see Matisoff 1991:483. MIGRATING BROTHERS 309

MIGRATION AS MYTH AND HISTORICAL EXPERIENCE

Migration—the process of movement and resettlement of individuals, families or larger groups of people—is a very prominent concept in both WKHKLVWRULRJUDSK\DQGHWKQRJUDSK\RIWKHGLIIHUHQW;ƯIƗQJURXSV7KHUH are the grand theories of large scale migration of proto Tibeto-Burman peoples along the north-south and south-east corridor of Tibet’s southeastern borderlands—the so-called Tibeto-Burman corridor; there is the recurring use of migration stories in popular myth to explain different aspects of the origin or ethnicity of local communities; there are the ceremonies for guiding the souls of the deceased back to a mythical homeland of the community; and—arriving into the new millennium—there DUHPLJUDWLRQVUHVXOWLQJIURPRI¿FLDOUHVHWWOHPHQWSROLFLHVDQGIURPWKH quest for economic opportunities in more developed parts of China. The variety of contexts in which migration is prominent in the southern section of the Tibeto-Burman corridor does not make it an easy concept to deal with, neither as a phenomenon to explain nor as an explanatory factor in other sociological or historical phenomena. It would be tempting to ascribe, for example, the importance of migration in popular P\WKRORJLVLQJWRWKHVSHFL¿FKLVWRULFDOH[SHULHQFHRIVRPHRIWKH;ƯIƗQ groups of constantly migrating from the northern parts of the corridor to the south. But there exists very little concrete evidence that physical groups of people actually broke up residence in one place and moved over a considerable distance to another where they then settled down. As mentioned earlier, Wang Ming-k’e has, for example, pointed to the KHUHV\RIOLQNLQJWKHSUHVHQFHRIDFDWHJRU\RI4LƗQJLQWKHQRUWKHUQSDUW RIWKH4ƯQJKӽL7LEHW3ODWHDXLQWKHZULWWHQVRXUFHVRIWKH+jQ'\QDVW\ more than 2000 years ago to the present existence of an ethnic group FDOOHG4LƗQJPRUHWKDQDWKRXVDQGNLORPHWUHVWRWKHVRXWK11 What seems most likely from ethnographic evidence is that among all the inhabitants RIWKHVRFDOOHGFRUULGRUWKHGLIIHUHQW;ƯIƗQHWKQLFJURXSVDUHWKHROG- est inhabitants. Today, remnant communities live spread out in the most mountainous parts of the area, having been driven out of the plains or assimilated by the now vastly more numerous Hàn Chinese and Nuosu. What we do know with some certitude from historical records is that ;ƯIƗQLGHQWL¿DEOHZLWKSUHVHQWGD\SRSXODWLRQVZHUHDOUHDG\VHWWOHGLQ

11 It is a received wisdom in Chinese ethno-history that not only the present ethnic PLQRULW\JURXSWKH4LƗQJDUHGLUHFWGHVFHQGDQWVIURPWKHDQFLHQW4LƗQJEXWDOVRPRVWRI the Tibeto-Burman-speakers in Southwest China, including the Tibetans. In spite of Wang’s convincing argument, this ‘fact’ is repeated in more recent and critical publications; see, for example, Shih 2001:383. 310 KOEN WELLENS

WKH DUHD RI ZKDW LV QRZ :HVWHUQ 6uFKXƗQ DW WKH WLPH RI WKH 0RQJRO conquest of Southwest China by Kublai Khan in the thirteenth century A.D. From Mongolian Yuán Dynasty sources we do learn that Kublai’s campaigns in the southwest heavily upset local population patterns. In their successful war against the Dali Kingdom in Yúnnán, the Mongol DUP\ KDG LWV UDQNV VXSSOLHG ZLWK ;ƯIƗQ VROGLHUV ZKRP WKH\ KDG HQOLVWHGLQ6RXWKZHVW6uFKXƗQ7KHVHVROGLHUVFDUU\LQJWKHLUIDPLOLHV with them—or who were later joined by them—were rewarded by the Mongols with land and titles in the newly acquired territory. This initial migration created the start of a continuous trickling movement of other ;ƯIƗQLQGLYLGXDOV²RUPD\EHDOVRHQWLUHIDPLOLHV²UHVXOWLQJLQWKHSUHV- ent population of the Tibeto-Burman-speaking Premi people in north- western Yúnnán Province. Today the 30,000 Premi in Yúnnán live in small tight clusters of a few villages thinly spread over a large area. One reason for this settlement pattern has been explained by the fact that the ancestors of today’s Premi arrived from different directions as part of the multi-pronged Mongol attack on the Dali Kingdom. The association ZLWKWKH0RQJROFRQTXHURUVDOVRDFFRXQWVIRUWKH;ƯIƗQRULJLQRIVRPH of the families ruling other ethnic populations such as the Na of the <ԁQJQuQJDUHDLQ<~QQiQ12 Among the Premi in north-western Yúnnán, as well as among other Tibeto-Burman-speakers in the region such as the Na and the Naxi, these historical events have also found their way into the oral history, where the Mongol factor has become an important ingredient of the mythologisation process. %XWPLJUDWLRQLQRUDOKLVWRU\LVDOVRSURPLQHQWDPRQJ;ƯIƗQJURXSV that did not join the Mongol campaigns and who remained settled in their ORQJWLPHKRPHODQGVRIVRXWKZHVWHUQ6uFKXƗQ,IWKHVHVWRULHVRIPLJUD- tion in their oral traditions can be linked to the transmission of historical events, they are most likely related to the smaller scale movements of IDPLOLHVDQGFRPPXQLWLHVZLWKLQWKHORFDOFRQWH[W,QWKH/LiQJVKƗQDUHD RI 6uFKXƗQ WKHUH DUH PDQ\ VWRULHV DPRQJ WKH 1DPX\L UHFRXQWLQJ WKDW long ago they used to live in the plains. During my visit in 2006, the eighty year old p’atse²RU1DPX\LULWXDOVSHFLDOLVW²RI;LӽQJVKXӿLQWKH IRRWKLOOVRQWKHQRUWKZHVWHUQHGJHVRIWKHODUJHDQGIHUWLOH;ƯFKƗQJ3ODLQ told me that the Namuyi only moved up into the mountains twelve or thirteen generations ago. This was the result of refugees arriving from the LQWHULRURI&KLQDIRUFLQJWKHPRIIWKHLU¿HOGVLQWKHSODLQ7KH\KDGEHHQ

12 Shih Chuan-kang 2001 convincingly demonstrates that the claim by the local W·VƯ ruling IDPLO\RIWKH0yVXǀRU1DRI<ԁQJQuQJWKDWWKH\DUHRI0RQJROGHFHQWLVLQFRUUHFWEXWWKDW WKH¿UVWW·VƯPXVWKDYHEHHQIURP;ƯIƗQ²LQWKLVFDVH3UHPL²RULJLQ MIGRATING BROTHERS 311 living there for a very, very long time, but before that, he insisted, they FDPH IURP /KDVD7RGD\ WKH /LiQJVKƗQ DUHD LQ VRXWKZHVWHUQ 6uFKXƗQ is—except for the major cities—mainly populated by the Nuosu people. But this is a relatively new development. According to the Chinese DQWKURSRORJLVW/yQJ;ƯMLƗQJPDQ\1XRVXSODFHQDPHVLQQRUWKZHVWHUQ /LiQJVKƗQ3UHIHFWXUHFDUU\PHDQLQJVVXJJHVWLQJWKDWWKH\ZHUHLQKDE- LWHGE\;ƯIƗQEHIRUHWKH1XRVXDUULYHG)RUH[DPSOHWKH1XRVXQDPH 0DFKDQOLQIRUWKHWRZQRI

131RZWKHFRXQW\RI<OyQJ 312 KOEN WELLENS

SURFHVVHV ZKHUH LQGLYLGXDOV PRYH ¿UVW HVWDEOLVKLQJ µEULGJHKHDGV¶ IRU other individuals to follow in a ‘trickling’ movement. They are motivated ¿UVW DQG IRUHPRVW E\ D TXHVW IRU ODQG UHVRXUFHV DV LQ WKH FDVH RI WKH population pressure among the Nuosu. However, certain aspects of social RUJDQLVDWLRQDPRQJWKH;ƯIƗQWKHPVHOYHVDOVRLQGXFHPLJUDWLRQ:KHQ, studied the origins of the different clans in a small cluster of Premi vil- ODJHVLQ0Oӿ&RXQW\LQ6uFKXƗQ,ZDVVXUSULVHGWRGLVFRYHUWKDWWKUHHRI WKH¿YHFODQVLQWKHDUHDZHUHFRQVLGHUHGQRWWREHRI3UHPLRULJLQ7ZR were established by men belonging to the neighbouring Naxi people mar- U\LQJLQWRWKHYLOODJHV7KH3UHPLDQGPDQ\RIWKHLU;ƯIƗQQHLJKERXUV practice clan exogamy and also intermarry frequently with members of other Tibeto-Burman groups, with the exclusion of the endogamous Nuosu. Since women become part of the clan of their husbands, the existence of different clans in close proximity of each other provides good opportunities to maintain a system where women are exchanged as marriage partners between the different clans. Such exchanges are closely linked to a preference for cross-cousin marriage. In the case of the Naxi FODQVWKLVPHDQWWKDWDIWHUWKH¿UVW1D[LHVWDEOLVKHGDKRXVHKROGLQWKH Premi villages, other Naxi women married the male descendants of the ¿UVW1D[LKRXVHKROGZKLOHWKHIHPDOHGHVFHQGDQWVZHUHPDUULHGRIIWR WKHRULJLQDO1D[LYLOODJH7KLVFRQVWDQWLQÀX[RIZRPHQEULQJLQJWKHLU cultural practices with them resulted in an actual migration of culture, language and people between the two areas. In one of the two Naxi clans in what was otherwise culturally a Premi village, Naxi language was used inside the house and the important religious rituals were conducted by Naxi ritual specialists. But there were also other ways in which social organisation generated small scale local migrations. In these two Premi villages there was a third clan, called Gonga, that was considered to be of non-Premi origin. This FODQZDVHVWDEOLVKHGE\DPHPEHURIWKH*DPLSHRSOH7KH*DPLLQ0Oӿ are considered to be ‘real Tibetans’, speaking a dialect close to the Dege dialect of Kham Tibetan. They live close to Gonga Mountain, which explains how the new clan obtained its name. The Gami follow the practice of primogeniture where the oldest son is the sole heir to the house and the land. The following story was told on the origin of the Gonga: 6HYHQJHQHUDWLRQVDJRWKHUHZHUH¿YHEURWKHUVLQ6KXӿOXz>DVPDOOWRZQLQ WKHHDVWRI0OӿRQWKHZD\WR*RQJD0RXQWDLQ@$VZDVFXVWRPDU\DPRQJ the Gami, the oldest brother inherited everything and consequently the four other brothers had no place to live after their father died. As a result they OHIW6KXӿOXzDQGVSUHDGWRGLIIHUHQWDUHDV2QHRIWKRVHIRXUZDV'DGUH MIGRATING BROTHERS 313

P’intsu, the ancestor of [the founder of the Gonga clan in the village]. In WKHEHJLQQLQJWKHEURWKHUVFDPHEDFNWR6KXӿOXzRQFHD\HDUDWWKHWLPH of the New Year. But after some time, Dadre P’intsu stopped going so it is not known what happened to the other brothers. Dadre P’intsu moved to <ԁQJQuQJLQ<~QQiQ)LYHJHQHUDWLRQVODWHU'DGUH3¶LQWVX¶VGHVFHQGDQW came to the village here. The protagonist here married a Premi girl, settled down, reclaimed new land and provided the local Premi with a new clan in the village with which they could practice marriage exchange. Although in this particular case, the migration of an outsider did not lead to continuing exchanges of women with his ancestral village, it is clear that a combination of VFDUFLW\RIDUDEOHODQGDQGVRFLDOSUDFWLFHVDPRQJWKH;ƯIƗQVXFKDV exogamous clans and primogeniture, have stimulated local migrations. The example also suggests how easily recognisable elements of social organisation, such as the brother that has to migrate because of the system of primogeniture, can become sources for the construction of mythologisation templates.

THE MIGRATING BROTHER AND OTHER TEMPLATES OF MYTHOLOGISATION

Lévy-Strauss recognised that myths are related to given facts, but are not representations of them (Lévi-Strauss 1968). I would like to suggest that WKHUHFXUULQJLPSDFWRIPLJUDWLRQLQ;ƯIƗQKLVWRULHVKDVPDGHWKHFRQ- cept of migration and linkage to origin into a template or mechanism for explaining that which needs to be, but cannot easily be, explained, such as ethnic origin, rules for clan endogamy, the lack of a script, gender roles, etc. Migration is but one such template. Another one—probably of a more archetypal and widespread nature—is that of different brothers constituting the origin of different communities. And the two templates can easily be combined. A striking illustration of how such templates are used is the story explaining the origin of the fourth clan in the cluster of Premi villages in 0Oӿ&RXQW\7KLVFODQWKH0HVpZDVWKHROGHVWLQWKHYLOODJHDQGRQH of Premi origin. The interesting point with this clan is that its exogamous group also includes the large Premi clan of Tsuop’i. They were the same clan, and yet, they were not. How so? People were not happy to tell me and intimated that the reason was related to the name of their clan: Mesé in Premi literally means ‘dead person’: A long time ago there were two brothers in [a village in Yúnnán] belonging 314 KOEN WELLENS

to the Tsuop’i clan. After partitioning the house, the younger brother, Monmon, did very well and had many animals. The oldest one was not that smart, and therefore very poor. Every night he went out to steal some DQLPDOV IURP KLV \RXQJHU EURWKHU 0RQPRQ FRXOG QRW ¿QG RXW ZKR WKH thief was, so one night he hid himself among his sheep, carrying a big NQLIH:KHQWKHWKLHI¿QDOO\FDPHLWZDVDOUHDG\SLWFKGDUNDQGWKHUHIRUH Monmon did not see who he killed. When his father found Monmon had killed his own brother he became so angry that he chased him out of the house and out of the clan. That is how Monmon came here and started WKH¿UVWKRXVH>LQWKLVYLOODJH@DQGRXWRIVKDPHKHNHSWWKHFODQQDPHRI Mesé. (see Wellens 2009). The story did not entirely match the genealogies according to which the village had only existed for six generations, while the break between Mesé and Tsuop’i was supposed to have occurred nineteen generations ago. The fact that there exist different variations of the story further suggests its mythical character. Its main function is to provide an explanation for something that has to be explained: How is it that our clan can’t intermarry with the Tsuop’i clan, even though we have a different clan name? I recorded another rather illustrative example of how narratives about migrating brothers mythologise the origin of the Premi, this time explaining the pattern of settlement spread of the Premi in Yúnnán. This example DOVRVKHGVOLJKWRQKRZPRGHUQHGXFDWLRQDQGWKHSUHYDLOLQJRI¿FLDOGLV- courses of the communist state are internalised and mythologised. In local text books used in junior middle school in Yúnnán, Premi children learn that they are descendants of soldiers who served in the army of Kublai .KDQ 3HRSOH ZHUH TXLFN WR WHOO PH WKLV LQ WKH YLOODJH RI 6zQJ]ӿ\XjQ KLJKXSLQWKHPRXQWDLQVDERYHWKHIDPRXVROGWRZQRI/uMLƗQJDQGVDLG WKDWWKHUHH[LVWHGDPSOHHYLGHQFHLQWKHLUORFDORUDOWUDGLWLRQFRQ¿UPLQJ WKHRI¿FLDOKLVWRU\2QHVWRU\H[SODLQVWKHIDFWWKDWWKHUHDUHWKUHHPDMRU VHWWOHPHQWDUHDVIRUWKH3UHPLWRWKHZHVWRI/uMLƗQJ /RQJDJRWKUHH0RQJROLDQEURWKHUVFDPHWRRXU6zQJ]ӿ\XjQ7KHROGHVW who was the strongest, did not want to stay but continued the whole way WR/iQStQJ>WKHZHVWHUQPRVWDUHDRI3UHPLVHWWOHPHQWLQ<~QQiQ@ZKHUH he settled down. The second eldest brother was not that strong and made it RQO\WR/ԃGLjQ>DQDUHDZLWKDFRQFHQWUDWHG3UHPLSRSXODWLRQWRWKHZHVW RI/uMLƗQJEXWQRWDVIDUDV/iQStQJ@7KH\RXQJHVWEURWKHUZDVWRRWLUHG WRJRDQ\ZKHUHVRKHVHWWOHGKHUHLQ6zQJ]ӿ\XjQ>DQGEHFDPHWKHDQFHVWRU of the people who told me the story] (Wellens 1998:28). This story is clearly a newly emerged myth embedded in the traditional template of migration and different brothers. Only in recent decades, MIGRATING BROTHERS 315

LHDIWHUEHLQJFODVVL¿HGE\WKHVWDWHDVRQHRI¿FLDOHWKQLFPLQRULW\ have local Premi communities been developing a sense of common identity and a conception of a common ancestry with each other. In their geographically isolated position, there would be no felt need to explain the existence of other clusters of people speaking a similar language since there was no contact between the more remote clusters and probably no awareness of their existence. 2I¿FLDOGLVFRXUVHLQWKH35&LVLQWXUQDOVRLQIXVHGZLWKWKHRULJLQ and migration mythology of local communities. The Na people are close neighbours of the Premi and the two groups often intermarry. Similar to the Premi, Na live in a mountainous region stradling the provincial ERUGHUV RI <~QQiQ DQG 6uFKXƗQ 7KH HWKQLF FODVVL¿FDWLRQ SURMHFWRI the 1950s and 1960s had a confusing result for the Na since it reached GLIIHUHQWFRQFOXVLRQVLQ<~QQiQDQG6uFKXƗQ7KH1DOLYLQJLQ Yúnnán province became lumped together with their neighbours, the Naxi, with whom they share some linguistic and other cultural features but towards whom they also clearly express cultural distance. This did QRWFRQYLQFHSURYLQFLDOOHYHOHWKQLFFODVVL¿FDWLRQWHDPVWRJUDQWWKHP a separate national minority status. Just across the provincial border in 6uFKXƗQWKHFODVVL¿FDWLRQWHDPVUHDFKHGDUDWKHUGLIIHUHQWFRQFOXVLRQ Basing themselves on genealogies and the convictions of some of the 1DUXOLQJIDPLOLHVWKHWHDPVRI¿FLDOO\FODVVL¿HGWKH1DLQ6uFKXƗQDV 0ČQJԃ]~PHDQLQJ0RQJROLDQ VHHHJ+DUUHOO 7UDFLQJ WKHLUDQFHVWU\WR0RQJRORI¿FHUVRUHYHQWR.XEODL.KDQKLPVHOIZDV a means for the ruling elites to legitimate their positions and therefore constitutes rather dubious historical proof (McKhann 1998:33). There LV QR RWKHU HYLGHQFH FRUURERUDWLQJ FODLPV WKDW WKH 1D LQ 6uFKXƗQ DUH related to Mongols of Kublai Khan’s era and thus warranting their ethnic FODVVL¿FDWLRQDV0RQJROV$QGLWFHUWDLQO\GRHVQRWMXVWLI\WKHIDFWWKDW their ethnic brethren just across the provincial border did not get classi- ¿HGDV0RQJROV 1HYHUWKHOHVVWKHVWDWHFODVVL¿FDWLRQHTXLSSHGWKH1DLQ6uFKXƗQZLWK the same minority nationality label as the Mongolian-speaking herders of the grasslands of Inner Mongolia, and this was clearly the result of WDNLQJ ORFDO P\WKRORJ\ DW IDFH YDOXH 1RW XQH[SHFWHGO\ WKH RI¿FLDO discourse performed its mythologising function: people from a Buddhist 1DFRPPXQLW\LQ6uFKXƗQH[SODLQHGWRPHWKDWWKHLUIRUPRI%XGGKLVP was different from that of their Premi neighbours because it originated in 0RQJROLDQRWLQ7LEHW7KHFRQVHTXHQFHVRIWKHVHRI¿FLDOGLVFRXUVHVDUH very strong because they do connect to existing elements in traditional 316 KOEN WELLENS mythology: it does make sense for the Na people to be associated with Mongols. Mythology becomes rationalised by the operations of the Chinese state and as such is remythologised to explain what needs to be explained. While a common minority nationality label for the Premi in Yúnnán has created a legitimate platform for expressing, developing and even re-creating a diverging cultural identity from mainstream Hàn culture, for the Na, being numerically even smaller, the Mongolian label does not offer such an opportunity. The fact that one part of the Na KDVEHHQFODVVL¿HGDV0RQJROLDQZKLOHWKHRWKHUEHFDPH1D[LPDNHVLW YHU\GLI¿FXOWWRGHYHORSIRUH[DPSOHFXOWXUDODVVRFLDWLRQVWRSURPRWH language use or revive traditional religious institutions. Such actions might even be met with suspicion by the Party-state since they question RI¿FLDOGLVFRXUVH The same mechanism of remythologisation can be observed among VRPHRIWKHRWKHUVPDOO;ƯIƗQJURXSVWKDWKDYHVXUYLYHGLQWRWKHQHZ millennium as separate cultural entities but who might be on the brink RI GLVDSSHDUDQFH /LNH DOO RWKHU ;ƯIƗQ JURXSV H[FHSW WKH 1D ZKR EHFDPH0RQJROLDQWKH1DPX\LLQ6uFKXƗQEHFDPHFODVVL¿HGDV=jQJ RU7LEHWDQV,XQGHUWRRN¿HOGZRUNLQD1DPX\LFRPPXQLW\LQ;LӽQJVKXӿ WRZQVKLS RI ;ƯFKƗQJ &LW\ /LiQJVKƗQ

14 There exists empirical evidence that goes against taking such association for granted. In one Premi community I managed to identify the entire trajectory mentioned in the ritual for guiding the souls of the deceased of one clan. But rather than being a migration route, the ritual constituted a very detailed local map leading to a sacred mountain cave a few days walk from the village where the urns with ashes and bone remains of the clan were placed after cremation. In this case at least, the ancestral spirits were guided to be reunited with the material left-overs of their bodily existence, as well as with the other ancestral spirits of their clan members. 318 KOEN WELLENS various ceremonies and rituals. Most of the texts are hand-written copies of Tibetan Buddhist texts, but I also observed several texts used by the p’atse written in Nuosu script. Although the p’atse are not able to understand the meaning of the Tibetan texts, the ritual power—or magic—lies LQUHFLWLQJWKHVDFUHGV\OODEOHV8QOLNHWKH3UHPLLQ0OӿZKRZHUHWR some extent integrated into the sphere of monastic Buddhism, nobody in ;LӽQJVKXӿKDGHYHUYLVLWHGD7LEHWDQ%XGGKLVWPRQDVWHU\RUEHHQRQD pilgrimage. They also had no recollections of such a tradition in the past. Having been informed by the authorities that they were Tibetans, they ZHUHVWUHVVLQJZKDWWKH\LGHQWL¿HGDV7LEHWDQLQWKHLUWUDGLWLRQ

CONCLUDING REMARKS

This essay has suggested that in an ethnographic study focusing on migration and origin among ethnic minority populations, the context in which knowledge is collected, processed and produced cannot be under- estimated. Minorities are precisely that: diverging cultural communities living in the context of a politically dominant cultural universe. It is LPSRVVLEOHWRVWXG\PLJUDWLRQDQGRULJLQDPRQJGLIIHUHQW;ƯIƗQJURXSV without taking into account China’s long historiographical tradition and the sanctity placed on the written word, the majority Hàn Chinese bias WRZDUGVWKLVWUDGLWLRQDQG¿QDOO\WKHDOOHQFRPSDVVLQJGLVFRXUVHSUR- duced and maintained by the modern Party-state. The essay has focused on some of the mechanisms at work in the process of internalising this discourse by the people it refers to, i.e., the mythologising template of the migrating brothers. The dominating cultural centre of the Hàn Chinese Party-state has, in a hegemonic way, established the only valid version of ethnic history within the territory of the PRC. Although this version, in several cases, LVUDWKHULQFRQÀLFWZLWKORFDOXQGHUVWDQGLQJRIWKLVKLVWRU\LWFDQQRW be challenged. There is no room for having recognised group identities WKDW DUH LQ RSSRVLWLRQ WR RI¿FLDO DQG VWDWLF PtQ]~ categories—or that DUHSDUDOOHOWRRURYHUODSSLQJZLWKWKHP2QFHFODVVL¿HGDV7LEHWDQV 1DPX\LFRPPXQLWLHVZLOO¿QGLWLPSRVVLEOHWRREWDLQDSSURYDODQGVXS- port for developing a writing standard for their language or for teaching local history focusing on a distinct Namuyi culture. But, as this chapter has shown, in view of how successful the Party-state has been in having the Namuyi internalise its version of their ethnic origins, maybe most of them will not object anymore in the near future. MIGRATING BROTHERS 319

BIBLIOGAPHY

Brown, M.J. (ed.). 1996. Negotiating Ethnicities in China and Taiwan. Berkeley: Institute of East Asian Studies. 'jL 4uQJ[Lj =jQJ0L΁Q \·]~ \·\iQ \iQMLǌ .ǌQPtQJ <~QQiQ PtQ]~ FKǌEӽQVKq Harrell, S. 1995. Cultural Encounters on China’s Ethnic Frontiers. Seattle: University of Washington Press (Studies on Ethnic Groups in China). —— 2001. Ways of being Ethnic in Southwest China. Seattle: University of Washington Press (Studies on Ethnic Groups in China). Kaup, K.P. 2000. Creating the Zhuang: Ethnic Politics in China. Boulder: Lynne Rienner Publishers. Lakhi, L., B. Hefright and K. Stuart. 2007. The Namuyi: linguistic and cultural features. Asian Folklore Studies 66: 233-53. Lévi-Strauss, C. 1968. Structural Anthropology. Harmondsworth: Penguin. /L~ /yQJFKǌ 1jPӽUpQ GH VKƗQOtQJ FKyQJEjL MtTt \ӽQELjQ 6KuMLq ]ǀQJMLDR \iQMLǌ 1: 106-14. /yQJ;ƯMLƗQJ/LiQJVKƗQ=KǀXMuQJQqLGH³;ƯIƗQ´Mt\XƗQ\XiQWjQWӽR;Ư]jQJ \iQMLǌ 1 & 3: 26-39 (vol. 1) & 56-65 (vol. 2). Matisoff, J. 1991. Sino-Tibetan linguistics: present state and future prospects. Annual Review of Anthropology 20: 469-504. McKhann, C.F. 1998. “Naxi, Rerkua, Moso, Meng: kinship, politics and ritual on the Yunnan-Sichuan frontier.” In M. Oppitz and E. Hsu (eds.), Naxi and Moso Ethnography. Kin, Rites, Pictographs. Zürich: Völkerkundemuseum Zürich, pp.23-45. Shih, Chuan-kang. 2001. Genesis of marriage among the Moso and empire-building in late imperial China. The Journal of Asian Studies 60(2): 381-412. 6ǌQ+yQJNƗL/LMLƗQJOL\GHPtQ]~\ԃ\iQMtTt[uVKԃIƝQOqL0tQ]~[XpEjR 3: 99-273. Wang, Ming-ke. 1992. The Ch’iang of ancient China through the Han dynasty: Ecological Frontiers and Ethnic boundaries. Ph.D. dissertation, East Asian Languages and Civilizations, Harvard University, Cambridge, Massachusetts. Wellens, K. 1998. What’s in a name? The Premi in Southwest China and the con- VHTXHQFHVRIGH¿QLQJHWKQLFLGHQWLW\Nations and Nationalism 4(1): 17-34. —— 2010. Religious revival in the Tibetan Borderlands: the Premi of Southwest China. Seattle: University of Washington Press. ;ƯFKƗQJVKu]KuELƗQ]XӽQZHӿ\XiQKXu HG ;ƯFKƗQJVKu zhì, =KǀQJKXiUpQPtQ JzQJKpJXyGuIƗQJ]KuFyQJVKǌ&KpQJGǌ6uFKXƗQUpQPtQ FKǌEӽQVKq

CONTRIBUTORS

ALEXANDER AISHER is a British Academy Postdoctoral Research Fellow and Lecturer at the Anthropology Department, University of Sussex. During 2002 and 2003 he carried out doctoral research with Nyishi communities inhabiting the remote uplands of Arunachal Pradesh. A detailed ethnographic account of his experiences with the Nyishi tribe will be published in 2012.

STUART BLACKBURN is a Senior Research Associate in the Department of South Asian Studies at the School of Oriental and African Studies, University of London. His research interests are in oral traditions and their ritual performance contexts. His most recent books are Himalayan Tribal Tales: Oral Tradition and Culture in the Apatani Valley (Brill, 2008) and The Sun Rises: A Shaman’s Chant, Ritual Exchange and Fertility in the Apatani Valley (Brill, 2010).

ROBBINS BURLING is Professor Emeritus of Anthropology and Linguistics DW WKH 8QLYHUVLW\ RI 0LFKLJDQ +H ¿UVW YLVLWHG 1RUWKHDVWHUQ ,QGLD LQ 1954, where he made an ethnographic study of the Garos, and following a long hiatus when the area was inaccessible, returned in 1996 as Visiting Lecturer at North-Eastern Hill University in Shillong. He is the author of several books on Garo society and Garo language in both India and Bangladesh, most recently The Strong Women of Modhupur (Dhaka, 1997), The Language of the Modhupur Mandi (New Delhi, 1997) and (with U.V. Joseph) Comparative Phonology of the Boro-Garo Languages (Central Institute of Indian Linguistics, 2006).

GEOFF CHILDS is Associate Professor at the Department of Anthropology, Washington University in St. Louis. His research uses anthropological approaches to understand social, cultural, and demographic changes in historical and contemporary Tibetan societies. His book, Tibetan Transitions (Brill, 2008), discusses the causes and consequences of fertility transitions that are occurring across the Tibetan world.

F.K.L. CHIT HLAING (F. K. LEHMAN) is Professor of Anthropology and of Linguistics in the University of Illinois at Urbana-Champaign. For over KDOIDFHQWXU\KHKDVGRQHH[WHQVLYH¿HOGZRUNDPRQJVWYDULRXVXSODQG and lowland peoples of Northern Southeast Asia, in Burma, Thailand, China and India, and has written monographs and numerous technical 322 CONTRIBUTORS articles on these peoples and their languages. His theoretical work is mainly in cognitive science and formal syntax. He grew up in a gem- trading family in Burma near the China border.

GEORGE VAN DRIEM is Professor of Historical Linguistics at the University of Berne, where he directs the Linguistics Institute and the Himalayan Languages Project. He has worked in the greater Himalayan region for 27 years and written grammars of Dzongkha, Limbu, Bumthang and Dumi and shorter descriptive accounts of other languages. He authored the ethnolinguistic handbook Languages of the Himalayas (Brill, 2001) and is completing grammars of several additional endangered languages. His research on population prehistory involves collaboration with geneticists, archaeologists and anthropologists. His theoretical work focuses on language evolution and semiosis.

MARTIN GAENSZLE is Professor in Cultural and Intellectual History of Modern South Asia at the Institute of South Asian, Tibetan and Buddhist Studies, University of Vienna. His research interests are in ethnicity, local history, oral traditions and religious pluralism in North India and Nepal. His book publications include Ancestral Voices: Oral Ritual Texts and their Social Contexts among the Mewahang Rai of East Nepal (LIT Verlag, 2002) and (with Karen Ebert) Rai Mythology: Kiranti Oral Texts (Harvard Oriental Series, 2008).

KERSTIN GROTHMANN is a Ph.D. Candidate in Tibetan Studies at the Humboldt University, Berlin. Since 2007 she has been conducting research in Arunachal Pradesh among the Memba community of Mechukha and Pemakö with a focus on issues of ethnicity and identity. She is author of the book “Wie der süße Duft der Blumen” - Die Arshe. Eine Untersuchung von Arbeitsliedern aus dem traditionellen tibet- ischen Bauhandwerk (Harrassowitz, 2011).

TONI HUBER is Professor for the Anthropology and Cultural History of Tibet and Himalayan Areas at the Humboldt University, Berlin. His long-term interest in highland peoples of the far eastern Himalaya has focused upon their relationships with neighbouring Tibetan Plateau societies. His most recent publications include The Holy Land Reborn: Pilgrimage and the Tibetan Reinvention of Buddhist India (Chicago, 2008), and (with Rinzin Thargyal) Nomads of Eastern Tibet. Social Organization and Economy of a Pastoral Estate in the Kingdom of Dege (Brill, 2007). CONTRIBUTORS 323

CHARLES MCKHANN is Professor of Anthropology and Director of Asian Studies at Whitman College (Walla Walla, Washington State). His research interests include religion, social history, ethnicity, kinship, art, and tourism among the Naxi and neighboring ethnic groups in southwest China (Yunnan and Sichuan). A list of his papers and publications can be found at: www.whitman.edu/anthropology/mckhann.html.

MARK W. P OST is a Research Fellow at The Cairns Institute, James Cook University, Cairns. He specialises in description and analysis of the Tani (Tibeto-Burman) languages of North East India, has completed a comprehensive grammar and dictionary of the Galo (Tani, Western) language of south-central Arunachal Pradesh and is currently working on grammars and dictionaries of the Eastern Tani languages Minyong (Adi) and Milang. He is Co-founder and Secretary of the North East Indian Linguistics Society, based at Gauhati University in Assam.

MANDY SADAN is Lecturer in the History of South East Asia at the School of Oriental and African Studies, University of London. Her research focuses on northern mainland South East Asia, especially Burma and the borderlands of North East India and Yunnan. She is currently completing a PRQRJUDSKRQWKHKLVWRU\RIHWKQLFFRQÀLFWLQWKH.DFKLQUHJLRQRI%XUPD

KOEN WELLENS is a Senior Researcher at the Norwegian Centre for Human Rights at the University of Oslo. He has been conducting DQWKURSRORJLFDO¿HOGZRUNDPRQJHWKQLFPLQRULWLHVLQ6RXWKZHVW&KLQD for many years. He is author of the book Religious Revival in the Tibetan Borderlands: The Premi of Southwest China (University of Washington Press, 2010).

MARION WETTSTEIN is a Postdoctoral Researcher in Anthropology at the University of Vienna. She undertook research in Nagaland from 2003 to 2010, mainly focusing on Naga textiles, and was co-curator of the Exhibition Naga—Beads and Ashes (Naga—Schmuck und Asche, Zürich: August 2008-March 2009, Vienna: Spring 2012) and co-editor of the volume Naga Identities (Snoeck, 2008). Since 2001, her research is mostly concerned with the extended Eastern Himalaya, especially Nagaland and Eastern Nepal, and her thematic interests include the anthropology of material culture and arts, performance and dance, religion, ritual and mythology, as well as identity processes.

INDEX

Aba Nibo, 118 Assamese, 60, 116, 157, 169, 170, Aba Pigu, 117 172, 174-177, 187, 214, 243 Abo Tani, also Abo-Tani, Abu-Tani, Ato Doni, 118 64, 65, 68, 70, 71, 75, 87, 107, Ato Neha, 117, 118 110, 115-120, 157 Ato Pombo, 118, 119 Acheulian 201, 202, Atsi, also Atzi, 245, 256 Adi 84, 164 Austroasiatic, also Austro-Asiatic, 2, Ahom 70, 187, 214, 215, 220, 233, 5, 178, 187, 190, 191, 193, 194, 246 199, 203, 204 Ahom Buranji, 233 Austronesian, 107, 108, 120 Ajuph-Aaney, 67 Ayen Tacang, 118 Aka, 220 Akbar Hyder Ali, 215 Bacot, Jacques, 14, 28 Akha, 244, 249 Bai, also Bái, 85, 91, 199, 278 Amdo, 199 Baic Karennic, 246 American Baptist Foreign Missionary Bailey, Frederick Marshman, 28, Society, 57 132-134, 139 Ami Dinchi Banyi, 110 Balti, 198, 204 Anal, 60 Baltistan, 198 Angami, see Naga Bangladesh, 154, 197 Anglo-Burmese war, 255 Bangni, 92, 93, Anii Khempu, 118 Barachetra, also Chatra, 34, 39, 44 Anii Khentu, 118 Barpak, 18, 20, 24 Anii Yalang, 118 Bay of Bengal, 191, 197 Ao, see Naga Bengali, 60, 187 Apatani 7, 8, 107-121, 163, 168, 181 Bengalis, 243 Arctitis binturong, 193 Bengia, 64, 72, 73, 76, 78, 79 Arun Koshi, also Arun River, 39 Bentinck, Arthur, 132 Arunachal Pradesh, 1, 2, 4, 5, 7, 8, %KDLUDKDYƗ 63-65, 74, 81, 83-85, 96, 98, 99, Bhutan, 28, 52, 87, 130, 132-134, 101, 102, 107-109, 115, 116, 120, 144-146, 154, 193 125, 127, 129, 130, 135, 139, 153, Bimo, 275 157, 168, 170, 172, 180, 181, 232, Black Mountain Mönpa, 200 249 Bodic, also Bodish, 87, 122, 154, Asetkong, 231 192, 199, 246 Asiatic Society of Bengal, 52 Bodo, 39, 220 Assam, 1, 70, 109, 153, 157, 158, 170, Bodo-Garo, 168, 246 174, 196, 213-215, 233, 244, 246 Bodo-Koch, 203 $VVDP5LÀHV Bod-pa, also Böpa, 133, 301, 306 326 INDEX

Bodu, 71, 76 Changter School, 18, 19, 29 Bokar, 99-103, 128, 161, 162 Chanii Chantung, 117 Bomdila, 143 Chanka Riman, 117 Bomih, 71, 76 Chantung, 110, 111, 114, 115, 117, Bon, also Bön, 85, 99, 285, 286 119, 120 Bos frontalis, 65 Char Chu River, 89 Boudhanath, 26 Chayül, 85, 88, 95, 103 Brahmaputra, 5, 28, 49, 172, 191, Chayül Chu, 88, 103 197, 213, 215 Chengdu, 285 Brahmaputran, 197, 203 Chibo-Nyobo, 158 %UƗNKX৬Ư &KuIƝQJ Bu, 288, 290 Chiilyang, 116, 117 Bubalus bubalus, 194 Chin, 5, 239, 242-244, 248, 249, 258, Buddha, 26, 144, 242, 292 271 Buddhism, also Buddha Dharma, 19, China, 1, 2, 7, 54, 57, 61, 70, 71, 85, 26, 54, 130, 135, 136, 143, 146, 100, 108, 191, 193, 199, 217-220, 147, 306, 307, 315, 318, 239, 240, 244, 245, 248, 249, 257, Bugun, 180 264, 275, 280, 299-310 Burma, 5, 7, 57, 58, 61, 170, 197, Chindwin River, 220 200, 205, 213, 215, 229, 239, 240, Chohpu, 78 243, 244, 249, 253-260, 271 Chöje Lingpa, 144-147 Burma Socialist Programme Party, Chokkar, 91 257, 260 Chongpi, 78, 79 Burmese, 57, 58, 178, 197, 203, 220, Christianity, 109, 254 221, 223, 240-245, 248, 254, 257- Chu, 288, 290 259, 272 Chumin, 18, 19, 21, 23, 26 Burmish, 254, 269 Chungliyimti village, 222 Burzahom, 192 Chyai, 262, 263 Buyung, 79 Chyai Hku Majoi, 262 Ci, 288, 290 Calcutta, 52 Coq, also Coqsseileel’ee, 276, 277, Cantonese, 5, 187 278, 279, 280, 295 Capricornis sumatrensis, 194 Coqbbersal, 275, 276 Ceil, also Ceilheeqbbubeq, 276-280 Coqbbert’v, 275, 276 Census of India, 88, 213 Corvinus, Gudrun, 201 Central Chin, 243, 244, 249 Council of the Baptist Churches in Chaha, 110, 114 Nagaland, 218 Chaiem, 223 Chakhesang, 223, 229, 'DÀD Chamai, 223 Dai, also Daic, Tai, 2, 170, 187, 215, Chang, 22, 144, 223, 224, 227 246, 249 Changki, 231 Dalai Lama, 20 Changsang Mongko, 224 Dali, 310 INDEX 327

Damchen La, 127 Dzongga, 20 'ƗৄJ Dzungar, 146 Daocheng, 287 Darjeeling, 13 East Kameng, 65 Darporijo, 95, 96 Eastern Kayah, 241 Dawkins, Richard, 64 Eastern Tani languages, 160, 161, 163 Ddaiq Ngvl, 291, 292 Elephas maximus, 194 Dege, 306, 312 Delhi, 9, 71 Fadeh, 71, 72, 76 Derq Zee, 291 Fagah, 71, 72, 76, 78 'HXNKXUƯEDVLQ Father Tongue, hypothesis, 198, 199, Dhangar, 187 204 Dhimal, 39 First World War, 215, 264 Dicerorhinus sumatrensis, 194 Fox, James J., 107, 108, 120 'LIÀRWK*pUDUG Fürer-Haimendorf, Christoph von, Diibo, 117 80, 214 Diilyang Diibu, 117 Dikhu River, 224, 231 Gacha, 128, 130, 134, 139, 143 Dirang, 130 Galo, 158, 161, 162, 164, 166, 170, Ditii Latii, 119 171, 172, 174, 175, 176, 178 Diyu Solo, 119 Gami, 312, 317 Doding, 118 Ganges, 34, 191, 194, 195, 197 Dodum, 64, 71, 78 *ƗQV Dohjung, 70, 71, 72, 75, 76 Garchoh, 78 Dolo, 118 Garo, 52, 59, 60 Dolu, 64, 71 Garo Hills, 3, 50, 58, 59 Dom La, 100, 102, 103, 143 Garrkik, 78 Domar Migyur Dorje, 22 Garze, 285 Dongba Sheel’loq, 286 Geling, 125, 129 Dongyi, 287 Gelug, 306 Doni Hiki, 119 Genghis Khan, 69 Donyi Lota, 118 Ghale, 18, 20 Dopum, 64, 71 *LGKLQL\Ɨ Dopung, 110 Giirii, 110 Doyü, 85 Ging, 89, 91, 95 Dravidian, 187 Ginru Ginsa, 259, 267, 268 Drukpa, 133 Golden Sand River, 279, 280 Drung, 307 Gonga, clan, 312, 313 Dudh Koshi, 39 Gonga Mountain, also Gongga Dudjom Tsal, 137 Mountain, 285, 286, 312 Duguhney, 70, 75, 76 Gongtung-Tajurr, 67 Dumgurr, 78 Gorkha, 17, 20, 27 Dunbar, George, 132 Graham, David, 279 328 INDEX gShen-rab mi-bo, 286 Imphal, 51 Gungtang, 19, 20, 23 India, also Indian subcontinent, 1, 2, Gurung, 20 5, 21, 34, 49, 50, 52, 56, 57, 58, Gushri Khan, 20 61, 65, 71, 84, 85, 86, 88, 100, Gyopu, 110 125, 127, 129, 130, 143, 154, 157, 169, 182, 187, 190, 191, 194, 195, Hakchang, 227 196, 197, 198, 200, 201, 203, 204, Hambo Halo, 114 213, 215, 216, 217, 219, 220, 222, Han, also Hàn Chinese, 198, 199, 232, 239, 240, 243, 246, 253 204, 275, 279, 281, 285, 287, 290, Indian Army, 216, 264 299, 300, 301, 303, 304, 306, 311, Indo-Aryan, 2, 56, 168, 170, 175, 316, 317, 318, 181, 198 Hàn Dynasty, 304, 305, 309 Indo-Burmese, 5, 187, 197, 203 Hao, 117 Indo-European, 5, 53, 54, 55, 169, Hari, 117 187, 202, 203 Heimi, see Naga Indonesia, 108 Heine-Geldern, Robert von, 204 Inner Mongolia, 315 Helambu, 13, 22, 23, 24, 28 Iran, 55 +pPԃG Irrawaddy, 58, 220 Hénán, 197 Isaac Chichi Swu, 221 Hill Miri, also Hills Miri, 162 Islam, 198 Himalayan Goral, 72 Itanagar, 9, 69 Himalayish, 246 Hindi, 157, 169, 187, 214 Jackson, Anthony, 285 Hinduism, 54, 109 Jamir, 214, 220, 221, 222, 224 Hindus, 53, 54 Jhangar, 187 Hiyeeney, 67 -LӽK~ Hkrang, 262, 263 Jiarong, 304, 307 Hkrang Hku Majoi, 262 Jinghpaw, 7, 245, 246, 247, 253, 254, Hmong-Mien, 194, 197 256, 257, 259, 260, 261, 262, 263, +zDEuQKLDQ 265, 266, 267, 268, 269, 270, 271, Hodson, Thomas Callan, 235 272 Höfer, András, 40 Jinghpaw Literature and Culture Holocene, 196, 201 Committee, 266 Hong, 117, 288, 290 Jinghpo-Nungish-Luish, 246 Hopmacham, 42, 43 Jingpo, 239, 245 Hotung-Hayung, 68 Jiulong, also-LԃOyQJ Huái River, 197 Jjiuqnalsheel’lojjuq, 285 Húnán, 199 Jungkung-Karnyung, 66 Hutton, John Henry, 50, 51, 52, 55, 57, Jungkuph-Jungmah-Jungturr- 58, 214, 218, 220, 227, 229, 232 Ajungney, 66 Hwakong, 220 Jungur-Juyung-Ajungney, 66 INDEX 329

Kacha, see Naga Khowalung, 39, 41 Kachin, 5, 7, 188, 220, 239, 245, 246, Khru, 92, 93 253-272 Khumbu, 21-23, 28 Kachin Independence Army, 257, 259 Kiipu Pingo, 116 Kachin State, 258, 260, 266 Kiranti, 34, 36 Kachin-Shan, 253 Kirium-Kulu, 66, 73-75 Kailas, Mount, 286 Kiyfah, 71, 78 Kalaktang, 130 Kohima, 231 Kameng, 65, 130, 133 Koldihawa, 195 Kamla, 91, 93, 95, 96, 102 Kol-Mon-Annan, 220 Kani Tari, 199 Koloriang, 64, 73, 74, 77-81 Kapu, 28 Kolyung Bumya Nyikang, 110 Karbi-Anglong, 172 Kolyung Pinii, 110 Karen, 57, 58, 240, 241, 243, 248, 258 Kongpo, 134, 137, 143, 144 Karenni, 242, 246 Konyak, 51, 223, 224, 232, 233 Karin Parin, 114 Kosuk-Koruh, 68 Kashmir, 192 Kotu Butang-Korda Horming, 110 Kathmandu, 21, 26, 27 Kra-Dai, 187, 191 Katog Rigzen Tsewang Norbu, 20 Kuàhúqiáo, 196 Kayah, 5, 239-244, 247, 258 Kublai Khan, 310, 314, 315 Ketch Pass, 93 Kuki, 60, 220, 223, 232, 243 Ketshel, 137 Kulu, 64, 66, 73-75 Khalo, 88 Künse Lhakhang, 143, 145 Kham, 22, 28, 133, 138, 143, 145, Kurab Namgye Dzong, 88, 103 275, 281, 304, 306, 307, 312 Kurium-Ngarngum-Kulu-Ngarngum- Kham Dege, 306 Ngarngumney, 66 Kham Tibetan, 304, 312 Kurukh, 187 Khambuhang, 39 Kurung Kumey, 64, 65, 72 Khampa, 133 Kurung River, also Kurung valley, Khamung, 220 72, 73 Khankha, 220 Kutang, 20 Khaplang, 221 Kyali, 85, 91 Kharteng, 134 Kyanksè, 58 Khasi, 60, 187, 190, 203, 204 Kyanzitha, 243 Khasi-Khmuic, 203, 204 Kyebogyi, 241 Khatii, 117 Kyika Ratö, 18 Khenbalung, 22 Kyimolung, 18-20, 23 Khezhakenoma, 227 Kyirong, 14, 19, 24 Khmer, 178, 190, 203f, 255 Khokwalung, 43, 44 Labau, 259, 260 .KROƗ Lachik, 256, 269 Khotang, 34, 42 Lacho Leyu, 116 Khotii, 117 Lahu, 170, 171, 244 330 INDEX

Lahuradewa, 195 Lotha, see Naga Laika Naura, 259 Lotha, Abraham, 220 Lakhers, 60 /XPELQƯ Lalahang, 37 Lumla, 134 Lama, 18, 20, 22-29, 133, 136, 138- Lung, 88, 89 140, 143-146, 148 Lungfungney, 67 Lamzha, 25 Lungterok, 51 Landu, 116 Lungtu Lopa, 88, 103 Langtang, 21-23, 28 Lushai, 220, 243, 244 Lanna Thai, 241 Lyngngams, 60 LaRaw, Maran, 239 Lare Galo, 172, 175, 176 Machanlin, 311 Lawngwaw, 245, 246, 256, 257, 269 Mahagarha, 195 Lepcha, 130 0DKRWWDUƯ Levallois, 202 0ӽMLƗEƗQJ Lhalu, 128 0ӽMLƗ\iR Lhalung, 22, 133, 134, 143 Majoi, 257, 262, 266 Lhasa, 88, 103, 128, 130, 134, 278, Majoi Shingra Bum, 257, 262, 266 303, 306, 311, 317 Makhel, 227, 229, 232 Lho, 25, 144 Malaysia, 108 Lhokpu, 193 Ma:mangme, 37, 40 /LiQJVKƗQ Manalung, 44 Libu Akhi, 317 Manau, 256, 259-268 Lijiang, also/uMLƗQJ Mandalay, 58 311, 314 Mandarin, 5, 170, 187 Limbuhang, 39 Manipur, 51, 60, 220, 221, 233 Limeking, 86, 90, 94-97 Manis javanica, 194 Liri Lilya, 114 Manumanaw, 241, 242, 247 Lisu, 245, 256, 257, 270, 281 Mara, 242 Lodre Gyamtso, 138, 139, 143, Maran Brang Seng, 264 145-147 Maru, 245, 256, 269 Lolo, 199, 245, 246, 248, 269, 279, Mason, Francis, 57 308, 316 Mayu, 263, 267 Lolo-Burmese, also Lolo-Burman, McMahon Line, 6, 88, 89, 94, 95, 199, 245, 269, 308 127, 133 Longju, 88, 89, 91, 109 Meche-Koche, 39 Longkum, 222, 224, 231 Mechukha, 125, 128, 129, 134, 135, Longkumer, 224 137, 139, 148 Longterok, 223, 224 Megam-Lyngngams, 60 Longtey Yuloh, 72 Megams, 60 Lopa, 88, 103, 130, 132, 136, 143, 146 Meghalaya, also0HJKƗOD\D Lorii Biinyi, 118 Mekong, 253 Lota Lobya, 118 Melikha, 220 INDEX 331

Memba, 6, 125, 127-136, 140, 141, Mon-Khmer, 187, 190, 203, 204 145-148 Mönyul, 133, 134, 146 Meng Mao, 246 Morse, Robert and Betty, 84 0ČQJԃ]~ Moso, 279, 284, 286-288, 290 Merag Lama, 146 Mosuo, also0yVXǀ Mesé, 313, 314 Movius line, 202 Metidem, 224 Mra, 85-97, 100, 103 Metog Dzong, 133 Mra Pusing, 89 Mewahang, 33, 34, 36, 39, 43 Mt. Emei, 275 Mewahang Rai, 33, 34, 322 Mt. Jizu, 275 Mianning, 303, 317 Mt. Jjuqnalsheel’loq, 280 Mibing, 78 Mt. Kailas, 286 Michif, 198 Mt. Meru, 280 Migyitün, 88, 89, 92 Mu Zeng, 291 Mihin clan, 119 Mudo Supung, 116 Milang, 163, 164, 180, 181, 323 Mükur, 223 Milarepa, 18, 19, 21, 22 Muli, also0Oӿ Millo clan, 119 311-313, 317, 318 Milya. 110 Munda, 187, 191, 203, 204 Miney, 71, 76 Murmiy, 78 Ming, 290 Murung, 119 Minyak, 304, 307 Muya, 188, 304 Minyong, 162 Myanmar, 213, 253, 254, 256, 258 Mipi, 28 Myitkyina, 259 Mishmi, 102, 220 Mising, 154, 161, 162, 168, 172, 174 Na, 85-95, 100, 103, 275, 279, 284, Miso, clan, 119 288, 303, 304, 310, 315, 316 Miso Halyi, 119 Naba, 85, 90, 91, 93, 95 Mizo, 57, 60, 178, 243, 244 Nacho, 91, 94, 95 mKhar-ro, 192 Naga, 2, 6, 7, 50, 51, 60, 213-223, Mokokchung, 231 227, 229, 231-235, 246, 271, 281; Molo, 130, 134 Angami, 50, 51, 60, 220, 223, Molungr, 231 224, 229; Ao, 51, 52, 214, 220, Mon, 58, 128, 130, 178, 190, 203, 222-224, 231, 233, 234; Heimi, 204, 240, 244, 258 221; Kacha, 51; Lotha, 220, 222, Mönba, also Mönpa, 129-134, 139, 223, 227, 229, 232; Rengma, 229; 148, 200 Sema, 223, 229; Sumi, 213, 221, Mongkhuma, 218, 220, 227 229; Tangkhul, 51, 221, 223 Mongkonyu, 218, 224, 226, 227 Naga Hills, 215, 220, 227 Mongolia, 213, 217, 218, 220, 222, Naga National Council, 215 227, 232, 234, 310, 314-316 Nagaland, 1, 2, 108, 216, 217, 218, Mongolian, 218, 220, 222, 227, 232, 220, 221, 231, 232 310, 314-316 Naksang, 134, 144, 145 332 INDEX

Nalo, 91 Nubri, 17-23, 25, 27 Namgo Dagam, 22 Nung-Rawang, 245, 256 Namuyi, 7, 188, 300, 303-307, 310, Nuosu, 275, 279, 301, 304, 306, 307, 316-318 309, 311, 312, 316-318 Nänang, 128, 130, 143, 148 Nyaré, 91 Nangalothae, 232 Nyema, 130, 132 Nanzhao, 57 Nyia-Nyoku, 66 1ƗUƗ\D৆Ư5LYHU Nyime, 116, 117, 132 Narba, 69-71, 77 Nyingmapa, 18, 19, 23, 27-29, 135 Nath, Jogendra, 84 Nyishi, 8, 63-65, 69-78, 80, 81, 116, National People’s Congress, 302 162, 163 Naxi, 8, 275, 278-288, 290, 291, 295, 300, 310, 312, 315, 316 Oceania, 219 Naze, 303 Old Burmese, 243 Ne Win, 255, 257 Old Chinese, 193 Neineiq Sher, 291, 293 Old Kuki-Chin, 243 Nem Singh, 130, 134, 139 Ongterok, 224 Nembu Nenda, 119 Orong, 130, 134 Nepal, 1, 4, 7, 8, 12, 13, 17, 20-28, Oryza, indica, 196; japonica, 196; 33, 34, 108, 154, 168, 187, 201, meyeriana, 195; nivara, 195, 196, 202, 220 197; RI¿FLDQDOLV, 195; perennis, Newari, 172 195; UX¿SRJRQ, 195, 196, 197; Neyü Phu Chu, 100 sativa, 195 Ngadag, 18, 20-27 Ozzu, 311 Ngarkuph and Ngarteney, 66 Ngedön Öseling, 27 Pachakshiri, 125, 127, 128, 130, Ngenci Peji, 114 133- 148 Ngiira Piisa, 114 Pachakshiriba, 128-130, 132, 133, Ngilyang, 119 139, 147, 148 Ngoju, 85, 91 Padam, 164 Ngua, 288, 290 Padmasambhava, 18, 21, 22, 135, Nguntii Anii, 110, 111 137, 143 Nici and Nica, 117 Padu Tiike, 114 Nicobarese, 203, 205 Pagan, 58, 243 Niipo, 118 Pai Che, 119 Niitu, 118 Pailibo, 128, 162 Nikun, 110-112, 114, 116 Pakistan, 198 Nikun-Sukun, 112, 116 Pakpahang, 37 Nilo, 85, 90, 91, 93, 95 Palbar Monastery, 19 1uQJOjQJ Palaeolithic, 201, 202 Nokrangr, 231 Panicum miliaceum, 193 Northern Chin, 191, 249 Panini, 53 Northern Naga, 246 Panyu Pare, 114 INDEX 333

Papum Pare, 65 Pyutu, 117 Papuph, 72, 73 3DU¿WW7XGRU Qiang, also4LƗQJ Paruhang, 36, 37, 43 304-309, 316 Patkai Hill, 220 Qiangic, 188, 246, 307, 308, 316, 3Ɨ৬X 4ƯQJKӽL7LEHW3ODWHDX Patu Page, 114, Pavo muticus, 193 Radhe, 119 Pemakö, 28, 129, 132-135, 144, 146 Rai, 8, 33-36, 39, 41, 42, 45 Pemaköpa, 129, 133, Rajput, 20 People’s Republic of China, 248, 280, Rakhine, 258 299, 306 Ramo, 99, 128, 162 Persians, 53, 54, 56 Rangoon, 240 3KƗJXQ Raprang, 89 Phom, 223, 224, 232 5ƗWR.KROƗ Phrei Phrow, 240-242 Rawang, 84, 245, 256, 257 Pictet, Adolphe, 193 Red Karen, 240, 241 Pinii Siyo, 110 Ree Gv Bei, 291, 294 Pleistocene, 201 Rengma, see Naga Pom, 51 Repcheck, 55 Pongen, 222 Revel, Nicole, 194 Pönzang, 18, 20, 21, 24, 26 Rhizomys sumatrensis, 194 Popi, 110, 111 Ridu-Numchung, 72 Poyoh, 78 Rigzen Gödemchen, 18, 19, 22, 23 Premi, 7, 300, 303, 304, 306, 307, Rma, 304 310-318 Rongkor, see Tsari Rongkor Prithvi Narayan Shah, 20 Ruata Rengsi, 60 Prmi, 275, 279, 281, 287 Rubu, 119 Proto-Oceanic, 165 Rubu Chigin, 119 Proto-Sino-Tibetan, 172 Rung, 307 Proto-Tani, 6, 160, 161, 163-168, Runya, 94, 96 171, 179-182 Runyu, 94, 96 Ptolemy, 232, 233 Puma, 34, 42, 43 Sakya, 19 Puma Rai, 42, Salween River, 240 Pumi, also3ԃPӿ3ԃPӿ]~ Sama, 12, 17-28 303 Sangtam, 223, 224 3X৆H Sangye Paljor, 22 Puri, 85, 91 Sangyeling, 134 Puroik, 92, 93, 180 Sankhuwa Valley, 38 Purtongzuk Longchar, 222 Sanskrit, 53, 54, 214 Pyokun-Pyopa, 112 Saptakoshi, 35, 44 Pyu, 243, 257 6DUƗƯ1DKDU5ƗƯ 334 INDEX

Saramati, 223 Siya Balyi, 114 Sarli, 78 Siyom River, 158 6ƗWSDWƯ Siyum, 90, 94, 95 Schizostachyum arunachalensis, 90 Solu-Khumbu, 21-23, 28 Schleicher, August, 188 Somnima, 36, 37, 39, 40, 43 Sechung-Nyedoh, 66 Songsong-Hahriying, 67 Seel Kv, 291, 294 Soreng-Lingpu, 93 Sema, see Naga Sot, 195 Sengtzet’u, 317 Southeast Asia, 1, 2, 54, 57, 108, 120, Serrah, 91 154, 169, 191, 196, 197, 201, 202, Setaria glauca, 195 204, 219, 239, 249 Setaria italic, 193 Southern Chin, 243, 244, 249 Shaha Habung, 119 Southern Yi, 244, 245 Shan, 58, 220, 240, 241, 244, 246, Subansiri, 65, 72, 83, 85-99, 102, 181 253, 254, 258, 271 Subansiri River, also Subansiri 6KƗQJ Valley, 72, 83, 85, 88, 95, 96 Shangwang Shangyung Khaplang, 221 Subu Heniin, 112, 114, 117, 119 Sheel’loq, 286 Sukai, 220 Shem, 54 Sukun, 112, 115 Sherdukpen, 180 Sulu River, 72 Sherpa, 13, 22 Sulung, see Puroik Shillong, 216 Sumi, see Naga Shinjong La, 127 Suming, 73 Shuiluo River valley, 286 Sumsih, 73 Shvq, 281 Sun Koshi, 39 Siang, 72, 102, 125, 127, 129, 158, Sungdu valley, 69 174, 181 Sungdu-Numchung, 69 Sibu’alawa, 285 Supyu Gyayu, 112, 114, 117 Sichuan, also6uFKXƗQ Sut, 261, 267 286, 290, 295, 301, 303-306, 310- 312, 315, 316 Tabyang-Talyang, 117 Siike, 117 Tabyu, 119 Sikkim, 130 Tage, 119 Silo, 117 Tagin, 86, 87, 92, 161, 162 6LQGKXOƯ Tai, 2, 170, 215, 226 Siney, 71, 76 Tai-Ahom, 70 Singming, 91 Tai Nyori, 84 Singpho, 220 Taisime, 110 Singtong-Tani, 64, 70 Tajang, 119 Sinitic, 169-172, 179, 198 Taker Bumer, 119 Sino-Tibetan, 57, 172 Taksing, 72, 85, 88, 89, 91-96, 100 Siwaliks Hills, alsoĝLYƗOLNVKLOOV Taliha, 95 ĝLYƗOLNV Talum, 63-65, 68-77, 80 INDEX 335

Taman, 220 Tsangpo River, 28 Tamur Koshi, 39 Tsari, 95, 103, 130, 134, 143, 144, 146 Tangam, 132 Tsari Chu, 87, 89, 91, 95, 96, 103, Tangkhul, see Naga 144 Tangu, 118 Tsari Rongkor, 87, 89, 103 Tani, 5,6, 64, 65, 69, 71, 73, 81, 86, Tsarong Sharpé Dasang Drandul, 89, 87, 99, 101, 102, 107, 115, 132, 103 153, 154, 157-171, 174-182, 246, Tshangla, 133 249 Tsoknyi Rinpoche, 27 Tapuk, 89 Tsona, 134, 144 Tarin, 117 Tsuop’i, 314 Tati, 117 Tuensang, 227 Tawang, 130, 133, 134, 139, 143, 146, 7XƯEDVLQ 148 7XƯYDOOH\ Terai, 5, 187, 201-203 Tumchiy, 72 Thailand, 213, 223, 229, 240 Tumiy, 72 Thangtong Gyalpo, 143, 146 Tumlingtar, 39 Tharu, 39 Tuting, 125, 129, 133, 143 Thuingaleng Muivah, 221 Tibet, 13, 15, 19, 20, 25-28, 52, 59, Ubhauli, also Udhauli, 43 61, 72, 85, 92, 94, 96, 99, 100, Uraon, 187 102, 109, 116, 128, 132, 133, 135, Urgyeling, 134 138, 143, 144, 146, 153, 154, 157, Urgyen Tulku, 27 158, 180, 192, 200, 205, 220, 257, 275, 285, 286, 300, 306, 215 Varanus nebulosus or bengalensis 193 Tibetan Buddhism, 135, 146, 306, 307 Wema Bbusso, 291, 294 Tibetan Plateau, 4, 20, 72, 87-89, 95, West Kameng District, 130 99-102, 192, 304 West Siang, 125, 127 Tibeto-Burman, family, language(s), Western Kayah, 241, 247 phyla, 5, 34, 109, 153, 163, 165, World War II, 271 175, 178, 179, 181, 187-193, 199, 200, 202-204, 242-247, 253, 256, ;LӽQJVKXӿ 268, 269, 306, 397, 316; popula- Xichang, 301, 303, 317 tions, speakers, 1, 2, 5, 7, 8, 9, 81, ;ƯFKƗQJ&LW\ 99, 108, 112, 128, 193, 198-200, ;ƯFKƗQJ3ODLQ 203, 204, 220, 243, 271, 275, 279, Xifan, also;ƯIƗQ 281, 300, 309, 310, 312, 316 303-313, 317, 318 Tingri, 14 ;ƯQJOǀQJJǀX ৫RNXYƗ Tradumtse Monastery, 20, 23 Yachu, 110 Trisong Detsan, 135 Yacu Yaja, 114 Tsang, 19, 20 Yading Mountain, 286 336 INDEX

Yama, 281 Yoyurr, 78 Yama Tani, 117 Yu zzu mber-mei gko, 280 Yan, 288 Yuán Dynasty, 310 Yangtze, 191, 193-197, 220