The Art of Symbolic Resistance

introduction i

# Brill's Inner Asian Library

ii introduction

*Editors*

Michael R. Drompp Devin DeWeese

VOLUME 30

*The titles published in this series are listed at brill.com/bial*

# The Art of Symbolic Resistance

introduction iii

# Uyghur Identities and Uyghur-Han Relations in Contemporary Xinjiang

*By* Joanne Smith Finley

LEIDEN • BOSTON 2013

This is an open access title distributed under the terms of the CC BY 4.0 license, which permits any non-commercial use, distribution, and reproduction in any medium, provided no alterations are made and the original author(s) and source are credited. Further information and the complete license text can be found at https://creativecommons.org/licenses/by/4.0/ *Cover illustration* iv : Lone figure on grasslands, rural Aqsu, 1996. Photo by author. introduction

The terms of the CC license apply only to the original material. The use of material from other sources (indicated by a reference) such as diagrams, illustrations, photos and text samples may require further permission from the respective copyright holder. Library of Congress Cataloging-in-Publication Data Smith Finley, Joanne. The art of symbolic resistance : Uyghur identities and Uyghur-Han relations in contemporary

Xinjiang / by Joanne Smith Finley. pages cm. -- (Brill's inner Asian library, ISSN 1566-7162 ; volume 30) *Cover illustration* iv : Lone figure on grasslands, rural Aqsu, 1996. Photo by author. introduction

 Includes bibliographical references and index. ISBN 978-90-04-25491-6 (hardback : alk. paper) -- ISBN 978-90-04-25678-1 (e-book) 1. Uighur Library of Congress Cataloging-in-Publication Data

(Turkic people)--China--Xinjiang Uygur Zizhiqu--Ethnic identity. 2. Uighur (Turkic people)-- Smith Finley, Joanne.

China--Xinjiang Uygur Zizhiqu--Government relations. 3. Government, Resistance to--China-- Xinjiang Uygur Zizhiqu. 4. Xinjiang Uygur Zizhiqu (China)--Ethnic relations. I. Title. The art of symbolic resistance : Uyghur identities and Uyghur-Han relations in contemporary Xinjiang / by Joanne Smith Finley.

 DS731.U4S65 2013 305.894'323--dc23 pages cm. -- (Brill's inner Asian library, ISSN 1566-7162 ; volume 30) Includes bibliographical references and index.

 2013023185 ISBN 978-90-04-25491-6 (hardback : alk. paper) -- ISBN 978-90-04-25678-1 (e-book) 1. Uighur (Turkic people)--China--Xinjiang Uygur Zizhiqu--Ethnic identity. 2. Uighur (Turkic people)-- China--Xinjiang Uygur Zizhiqu--Government relations. 3. Government, Resistance to--China-- Xinjiang Uygur Zizhiqu. 4. Xinjiang Uygur Zizhiqu (China)--Ethnic relations. I. Title.

 DS731.U4S65 2013 305.894'323--dc23

2013023185

This publication has been typeset in the multilingual "Brill" typeface. With over 5,100 characters covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more information, please see www.brill.com/brill-typeface.

ISSN 1566-7162 ISBN 978-90-04-25491-6 (hardback) ISBN 978-90-04-25678-1 (e-book)

Copyright 2013 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing, IDC Publishers and Martinus Nijhoff Publishers. covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more information, please see www.brill.com/brill-typeface. ISSN 1566-7162 Copyright 2013 by Joanne N. Smith Finley. Published by Koninklijke Brill NV, Leiden, The Netherlands. Smith Finley. Published by Koninklijke Brill NV, Leiden, The Netherlands.

This publication has been typeset in the multilingual "Brill" typeface. With over 5,100 characters

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. ISBN 978-90-04-25491-6 (hardback) ISBN 978-90-04-25678-1 (e-book) Copyright 2013 by Koninklijke Brill NV, Leiden, The Netherlands.

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. Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing, IDC Publishers and Martinus Nijhoff Publishers. 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,

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,

This book is printed on acid-free paper. photocopying, recording or otherwise, without prior written permission from the publisher.

This book is printed on acid-free paper.

Fees are subject to change.

222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA.

*For my mum and dad, who encourage me in all that I do*

contents v

vi contents

# Contents

contents vii


#### PART ONE

# THE ART OF SYMBOLIC RESISTANCE


### PART TWO

# SOUNDS OF DISCONTENT


#### viii contents


#### PART THREE

# 'SILENCE': AFTER THE 1997 GHULJA DISTURBANCES


#### contents ix


x contents Contents v List of illustrations ix Acknowledgements xi Preface xiii Structure of the Book xxi PART ONE xxvii THE ART OF SYMBOLIC RESISTANCE xxvii Introduction 1 Scholarship on China's Minority Nationalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Scholarship on the Contemporary Uyghurs 3 Identity from a Minority Perspective 6 Sartrean Notions of We-hood and Us-hood 7 The Art of Symbolic Resistance: From Cultural to Religious Symbols. . 9 A Brief Political History 16 Chapter One 23 Inequalities: The Socio-Economic Backdrop 23 1.1. Han In-migration 23 1.2. Language Policy 29 1.3. Education 36 1.4. Employment 44 1.5. Wealth Distribution 55 1.6. Resource Exploitation 60 1.7. Political Representation 70 Conclusion: Development – Equality = Conflict. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 PART TWO 79 SOUNDS OF DISCONTENT 79 Chapter Two 81 Stereotypes: talk breeds conflict 81 2.1. Social Stereotypes: Theoretical Frame 82 2.2. Theorising Han Chauvinism: The 'Civilising Project'. . . 84 2.3. Different Ways of Seeing: Counter-stereotypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 2.4. Common Heritage: Foundations of Counter-stereotypes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 2.5. Stereotypes relating to Uyghur Cultural Norms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 2.6. Stereotypes relating to Islamic Social Practices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 2.7. Stereotypes relating to Colonisation and Exploitation. . . 118 Conclusion: Talk Breeds Conflict (or the Power of Oral Complaint). . .123 Chapter Three 130 Making culture matter: symbolic, spatial, and social boundaries. . . 130 3.1. Barthian Theory: Ethnic Boundaries as Resistance. . . 131 3.2. Symbolic Boundaries 134 3.3. Spatial Boundaries 147 3.4. Social Boundaries 159 3.5. The Consequences of Boundary-crossing: Bus Stories and Street Fights . .163 3.6. Managed Interaction 164 Conclusion: The Roots of Constructed Segregation . .167 Chapter Four 173 Illuminists: Popular song and the waking of the Uyghur nation 173 4.1. The Musician-comprador 175 4.2. Popular Singers as Illuminists: The Advent of 'New Folk'. . . 184 4.3. Cassette Culture and Mass Mediation of Political Messages. . . 189 4.4. Ömärjan Alim: The People's Choice 191 4.5. Abdurehim Heyit: The Intellectuals' Choice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 4.6. Religious and Class Cleavages 221 Conclusion: New Folk as a Locus of Solidarity and Disunity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 PART THREE 233 'SILENCE': AFTER THE 1997 GHULJA DISTURBANCES. . . 233 Chapter Five 235 Reverts: sources and dynamics of Islamic renewal . . 235 Introduction 235 5.1. Islamic Renewal in Ürümchi since the Late 1990s . .236 5.2. Globalising Forces: Border Trade, Pilgrimage, Study Abroad and News Media. . .245 5.3. Islam as Symbolic Resistance against Perceived Oppression of Muslims. . . 260 5.4. Islam as a Response to Failed Development . .265 5.5. Islam as an Egalitarian Ethos 271 5.6. Islam as a Response to Frustrated Ethno-political Aspirations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274 5.7. Modernity and the Return to Cultural 'Purity'. . .276 5.8. Islam as a Vehicle for Personal and National Reform . .281 5.9. Internal Divisions within the Uyghur Muslim Population. . . 284 Conclusion: 'All Five Digits on one Hand are not Alike'. . . 287 Chapter Six 294 Endogamists: gender, nation and a selective taboo. . .294 6.1. The Statistics on Intermarriage 297 6.2. Parental Prohibition and Excommunication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 6.3. Community Supervision 305 6.4. Perceived Barriers: Racial, Cultural, Religious, Historical, Political. . . 314 6.5. Managing Intermarriage 324 6.6. Hierarchy of Intermarriage Preferences 334 6.7. Women as 'Culture Bearers' 340 6.7. Endogamy as Cultural Survival 345 Conclusion: Honour and Shame in a Context of National Impotence . . 347 Chapter Seven 349 Hybrids: Identity negotiations among the urban youth. . . 349 7.1. The First-generation Minkaohan: 'Internalised Oppression'. . . 350 7.2. The Second-generation Minkaohan: Product of Urban Instrumentalism. . . 354 7.3. Minkaohan Self-identity in Contemporary Xinjiang: 'Ethnic Anomalies'?. . . 359 7.4. The Minkaohan-minkaomin Divide 366 7.5. The Han Wannabe: Prototype for the Third-generation Minkaohan?. . .375 7.6. Cosmopolitan Multilingualism: Uyghur Youth as 'World Citizens' . . 380 7.7. The Beginnings of Rapprochement? 393 Conclusion: From 'Ethnic Anomaly' to Subaltern Cosmopolitan . . 402 Conclusions 408 Mainstreaming of Uyghur National Identity in the 1990s. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408 Eclipse of the Old and Foregrounding of the New. . . 410 Factors Affecting Uyghur Solidarity in the 1990s. . . 411 1997: Enter the Silent Era 413 From Independence to Federation 416 Policies towards Conflict Resolution 421 Bibliography 426 INDEX 445

# List of illustrations

preface xi


xii preface

### Acknowledgements

acknowledgements xiii

I could not have completed this book without the support of key academic institutions and many individuals both within and without Xinjiang. Funding for my postgraduate study, including the initial field trip in 1995-6 and a 'hard language grant' to learn Uyghur, was provided by the Economic and Social Research Council, UK, and permitted me to complete my thesis 'Changing Uyghur Identities in Xinjiang in the 1990s' (Leeds University, 1999). Mr. Guo Taishan and Ms. Minäwär Abdurehim at the Xinjiang Academy of Social Sciences, Ürümchi, were instrumental in smoothing the way to a one-year institutional affiliation in 1995-6, and provided me with comfortable on-site accommodation and support. Two Uyghur scholars, Tahirjan, an esteemed specialist in Uyghur literary studies, and Äli, a talented young Uyghur historian, showed patience and humour during the many hours they spent helping me grasp Uyghur phonetics, orthography and grammar. I owe a deep debt of gratitude to my PhD supervisors, Professor Gregor Benton (now at Cardiff University) and Professor Ray Pawson (now, as then, at Leeds University), who introduced me to relevant concepts in sociology and anthropology, and provided incisive and often painfully honest comments on my work; also to Professor Jennifer Mason (now at Manchester University) for showing me how to think reflexively in social research. I am equally indebted to my scholarly peers for, variously, their sharp insights, collaborative spirit, encouragement, empathy and companionship in the field. These individuals include Dr Ablimit Baki, Dr Nimrod Baranovitch, Mr Nicolas Becquelin, Dr Ildikó Bellér-Hann, Dr Gardner Bovingdon, Dr Trine Brox, Dr Cristina Cesàro, Dr Chen Yangbin, Dr Jay Dautcher, Dr Michael Dillon, Dr Arienne Dwyer, Dr Moshe Gammer, Professor Dru C. Gladney, Professor Chris Hann, Dr Rachel Harris, Professor Raphael Israeli, Dr Tamara Jacka, Mr Thierry Kellner, (the late) Dr Vanessa Knights, Professor Colin Mackerras, Professor James Millward, Dr Laura Newby, Dr Barry Sautman, Professor Yitzhak Shichor and Professor Naomi Standen. They also include the anonymous reviewer of this manuscript, whose meticulous, stimulating and enlightening comments were invaluable at the revisions stage. I would like to express my warm appreciation of recent times spent in lively discussion with the 'new blood' in the field of Uyghur studies, including the following postgraduate students: Mr Tom Cliff, Ms Janina Feyel, Mr Joshua Freeman, Mr Timothy Grose, Mr Eric Schluessel, Mr Mamtimyn Sunuodula and Mr David Tobin. Most of all, I thank my Uyghur respondents and their families (who remain anonymous) for their trust, support and willingness to disclose, even in an environment where communication with foreigners is increasingly outlawed as 'subversive'. Finally, I pay tribute to my family members and friends for their interest, enthusiasm, insights (academic and human), painstaking reflection on draft chapters, and unfailing support: my parents Victor and Daphne Smith, my husband and life partner Luke Finley, our beautiful, affectionate and patient son Harvey Gene, my parents-in-law Terence and Diana Finley, and my lifelong friends Emma Stone, Keith Tyndall, Jon Seagrave, Steve Longworth, Matt Thomas and Frazer Marshall.

# Preface

preface xv

I first came across the Uyghurs in 1988 while studying Chinese in Beijing. Across the road from the Beijing Foreign Studies University1 (then *Beijing waiguoyu xueyuan*北京外国语学院) was a Uyghur migrant worker community at Weigongcun. Stumbling into a Uyghur-owned restaurant one evening, I found myself surrounded by people who spoke in an unintelligible tongue and owned faces strikingly different to the Han Chinese. I was fascinated: who were they, and how had they come to be in China? Further investigations were carried out with difficulty since neither I nor the restaurateurs had yet become especially fluent in Mandarin. It emerged that these people were a Turkic-Islamic minority nationality known as the Uyghurs. Four years later, in June 1992, I held a conversation with a young Uyghur migrant worker, once more in Beijing. Under cover of darkness, he whispered excitedly that Uyghurs were being provided with weapons 'from the outside' with which they would fight the Han Chinese. August 1991 had seen the break-up of the former USSR and the establishment of independent Central Asian states adjacent to Xinjiang. There then followed considerable academic and media speculation about a potential 'contamination effect' (Bøckman 1992: 192): would movements for national independence spread across the border into Northwest China? International news had been flooding into Xinjiang since its opening in the mid-eighties, and many Uyghurs had relatives in the newly independent states. It was in this context and climate that I first went to Xinjiang to research Uyghur identities.

In contemporary China, conditions for doing fieldwork are increasingly dependent on locality, with greater possibilities available in the coastal areas than in the less developed western regions. Xinjiang, like Tibet and, to a lesser extent, Inner Mongolia, is a sensitive border region dominated by peripheral peoples who historically resisted incorporation within the Chinese polity.2 Although researchers gained unprecedented access to the

<sup>1</sup> This institution changed its name in 1996 to *Beijing waiguoyu daxue* (北京外国语 大学).

<sup>2</sup> In response to growing Chinese influence in the region from the 1800s on, three independent states were established: the Kashgar [hereafter, Qäshqär] Emirate of Yaqub Beg (1864-1877), the Turkish-Islamic Republic of Eastern Turkestan (capital Qäshqär, 1933- 34) and the Eastern Turkestan Republic (capital Ghulja, 1944-49). See Kim (1986; 2004), Forbes (1986) and Benson (1990). I use the transliteration 'Qäshqär' for the ancient silk road

#### xvi preface

region following the re-emergence in 1979 of the social sciences as officially sanctioned disciplines, that access has been under continual threat. A series of domestic and external events, including the 1989 pro-democracy campaign in China, culminating in the Tian'anmen incident (6.4 事件), the collapse of Marxist-Leninist parties in Eastern Europe in the same year, the Baren uprising of April 1990 in south Xinjiang, and the collapse of the USSR and formation of the CIS in 1991, impacted in turn on local conditions, fuelling both the desire for independence among Xinjiang's Uyghurs and official wariness towards researchers and journalists.3 Ethno-nationalist aspirations climaxed in the Ghulja [Ch. Yining] disturbances of 5th-6th February 1997,4 then waned in the aftermath of their military suppression. Now there followed a period of 'silence', characterised by repeated Strike Hard anti-separatist campaigns and vastly reduced civil rights for Uyghurs. Following the attacks of 11th September 2001 (9.11), the US-led war against terror provided in the view of many observers a convenient means to repackage legitimate Uyghur grievances as expressions of 'international Islamic terrorism' (Millward 2004: 10-14). Levels of state repression heightened, and the atmosphere in Xinjiang gradually came to resemble that inside a pressure cooker. The state's campaigns against what it viewed as Uyghur disloyalty utterly failed to eliminate Uyghur discontent; instead they simply drove it underground (cf. Dillon 2004; Bovingdon 2010: 83). For a time all was relatively quiet until, in July 2009, Uyghurs' shared sense of social injustice found an outlet in the Ürümchi protest against state handling of the

centre 'Kashgar' throughout the book, since it is closest to the original pronunciation in Uyghur.

<sup>3</sup> Following the suppression of the 1989 pro-democracy movement, Chinese officials pointed to the adverse influence of 'bourgeois' social sciences on the student movement, arguing that social research had identified negative social trends and fuelled calls for reform (Curran and Cook 1993: 72).

<sup>4</sup> For an authoritative account of this event, see Millward (2004: 16-17). Several hundred Uyghurs demonstrated after two *taliplar* (religious students) were arrested for alleged 'illegal religious activities.' Protestors called for ethnic equality and Uyghur independence, and expressed religious sentiments such as 'We have only one God, and he is Allah.' Antiriot police reportedly deployed dogs, tear gas, fire hoses, beatings and firepower against the demonstration. Riot targets included vehicles, police and Han residents. See also *Ming pao*, Hong Kong, 10 February 1997 in BBC Summary of World Broadcasts (Asia Pacific) hereafter SWB (AP), 11 February 1997, FE/2840 G/1; *Ming pao*, Hong Kong, 11 February 1997 in SWB (AP), 12 February 1997, FE/2841 G/1; *Sing tao jih pao*, Hong Kong, 11 February 1997 in SWB (AP), 12 February 1997, FE/2841 G/1; *Sing tao jih pao*, Hong Kong, 15 March 1997 in SWB (AP), 18 March 1997, FE/2870 G/6; Kyodo News Service, Tokyo, 2 May 1997 in SWB (AP), FE/2909 G/7; *The Guardian* (1997).

# preface xvii

Shaoguan factory incident.5 Under these erratic political conditions, field research in the region has come under intense state scrutiny.

I spent one year in the region in 1995-6 conducting fieldwork in the regional capital Ürümchi (nine months), and in the southern oases of Kucha, Aqsu, Qäshqär and Khotän (three months). My research questions at that time were:


On subsequent trips in 2002 (Ürümchi and Ghulja, six weeks) and 2004 (Ürümchi, six weeks), research questions focused on poles of resistance and accommodation among Uyghurs in the post-1997 period. I wanted to know:


Many of these issues were highly sensitive; the first three because they investigated the growth of Uyghur nationalism in the 1990s, the emergent independence movement, and Uyghur-Han conflicts, thus calling into question the official line of *minzu tuanjie* (nationality unity); the fourth because it touched upon the potential link between Islam and politics in a region where the state has outlawed so-called 'illegal religious activities'. As I have argued elsewhere, in investigating sensitive issues surrounding identity politics and ethnic relations in Xinjiang, formal interviewing was an unrealistic option (Smith 2006: 135-136). Following my arrival in Ürümchi in 1995, cadres at the Xinjiang Social Sciences Academy (*Xinjiang shehui* 

<sup>5</sup> For detailed accounts of this incident, in which at least two Uyghur workers were killed by Han co-workers in a toy factory in Guangdong, see Millward (2009); Hess (2009); Hu (2009); Buckley (2009); Watts (2009); Leow and Fairclough (2009); Jacobs (2009); and Mudie (2010). For an analysis of how the relationship between civil rights and civil duties underpins state handling of Uyghur grievances, see Smith Finley (2011a).

#### xviii preface

*kexueyuan* 新疆社会科学院) agreed that I might conduct a formal social survey; yet it quickly became clear that interviews would be supervised and thus carry little validity. Respondents would likely be briefed on what to say, and would almost certainly be afraid to answer frankly in front of Academy employees. In this way, power imbalances can sometimes render the official interview in China meaningless. Even where respondents are forthcoming, they will often give the interviewer a 'nice rounded picture of reality *as it ought to be*' rather than providing real insights (Thøgersen 1991: 31, 42; my emphasis). This problem was encountered by Thomas B. Gold while studying *getihu*—urban private entrepreneurs—during the 1980s; because this social group remained stigmatised at that time (being associated with petty capitalism), respondents were evasive in supervised interviews (1989: 180). Meanwhile, state-sponsored surveys conducted by Han social scientists suffer from a serious lack of impartiality, as their authors seek to distort or refute altogether the views of Uyghur respondents (Bovingdon 2010: 87).

In Xinjiang, the most viable method of enquiry was unstructured interview (in the shape of informal conversation),6 followed up by semi-structured interview. During my initial field trip, I selected key informants strategically to yield views indicative of different social, age and gender groups (Tremblay 1982: 98-99). This approach allowed for an initial patterning of identity ascriptions. In terms of my 1995-6 aims, the comparison across generations was important in order to determine the extent to which Uyghur national identity had strengthened over time (see Smith 2000). Choice of additional informants and content of subsequent conversations was steered by information already gathered in a process of 'progressive restructuring' (Tremblay 1982: 99). This enabled me to build up the number of key informants within each group while introducing partial structure to subsequent interviews in order to probe emerging themes. In order to triangulate the data, themes emerging from interviews were tested against direct observations of practices and interactions among Uyghurs, and interactions between Uyghurs and other groups, including Han Chinese. A basic assumption was made that meanings are sustained in processes of social interaction (Blumer 1969). Interview and observation data were further compared with textual analyses of Uyghur song content. In order to achieve inter-subjectivity of interpretation (Scheff 2006: 35, 40-41), all translations and lyrical interpretations were prepared in consultation with at least two Uyghur respondents.

<sup>6</sup> Of relevance here are the concepts of 'guerrilla interviewing' (Gold 1989: 180) and 'ad-hoc street interviewing' (Rudelson 1997: 11).

#### preface xix

During the first six months in Xinjiang, I studied the Uyghur language intensively so that I might hold conversations with respondents in their mother tongue. The process of learning a Turkic-Altaic language for the first time was hindered by three conditions:


Changes in local dialect between oases brought new problems, although interrogation of local vocabulary sometimes produced synonyms learned in Ürümchi; at other times, I relied on translation by an urban companion. Such companions were helpful in terms both of obtaining introductions to rural relatives and of rephrasing statements in standard Uyghur. As Cortazzi et al. have pointed out, English can be problematic as a language choice for interviewing in neo-colonial contexts owing to its 'negative associations of power and dominance' (2011: 507). Since local peoples in Xinjiang almost unanimously consider their homeland to have been 'invaded' by Han Chinese, learning to speak the regional *lingua franca* was clearly of paramount importance. Speaking to respondents in the second language imposed on them by a perceived coloniser would have created unhelpful distance. Addressing them in their mother tongue, on the other hand, could be expected to aid disclosure of personal thought, emotions and beliefs (Chiu, cited in Cortazzi et al. 2011: 517). There were also other practical considerations. While most educated, urban Uyghurs speak good Mandarin, many rural Uyghurs—including some migrants to urban centres know little Chinese. Finally, communicating in respondents' second language brings difficulties of interpretation as there may be no term in Chinese that exactly corresponds to the Uyghur, and vice versa.

<sup>7</sup> I learned Uyghur via a set of language textbooks which targeted Han Chinese students in higher education (Razzaq and In'amshah 1991). The language learning situation is slightly improved these days, since many Uyghur-English and English-Uyghur dictionaries are now available, as well as a number of English-language Uyghur grammar references. See for example Engesæth (2002).

Figure 1: The author, 3 months into the first field trip (1995). Photograph by Emma Stone.

Yet state politics are only one facet of the difficulties faced in Chinese Central Asia. Personal politics of gender, culture, religion and values also impact on the negotiation of research roles, necessitating a willingness and ability to operate in shifting margins (Smith 2006: 142-147). As a woman studying an Islamic society, I encountered certain field problems absent in the experiences of male peers, although they faced different gender-related problems (see Dautcher 2009; Bovingdon 2010: 17). Keen to avoid being channelled into women-only activities, I strove to re-negotiate my research role afresh depending on the gender, educational and religious (nominal or observant) background of each of my companions. With rural men and most women, I was the epitome of female modesty. With educated urban males, I could play up my status as an independent Western woman and academic. This allowed me to assume a neutral role vis-à-vis men and to be included in the 'male fraternity' closed to the society's female members. I thus gained perhaps greater access than a male researcher, who would have had access to male domains but only limited access to female ones. In 1996, I donned a baggy *ätläs* dress, *közlük halqa* [earrings with 'eyes' or semi-precious stones], a headscarf and thick nylon tights in a bid to safeguard feminine modesty and gain acceptance in a rural village in

#### preface xxi

Aqsu.8 Later, in 2004, it was advantageous to dress in (modest) Western clothes for some of the time since one focus of my enquiry was the hybridised identities of Uyghur urban youth (men and women), many of whom had embraced Western musics and fashions.

Participant observation constitutes a commitment to adopt the perspective of those being studied, at least for a temporary period (Denzin 1970: 185). It would be dishonest to pretend that, after hearing countless Uyghur accounts of injustices suffered under Han rule, I did not develop political sympathies. In fact, I was at constant risk of over-identifying with respondents and, as a result, ceasing to problematise their perspectives. As sociologist Howard S. Becker famously wrote: '…we fall into deep sympathy with the people we are studying […] because of this, we do not give a balanced picture' (1970: 16). In Xinjiang, almost every conversation would begin with a Uyghur respondent trying to 'test my position' within what quickly emerged as a heavily demarcated ethnic dichotomy. Only after respondents had satisfied themselves that I was a sympathetic interlocutor were they inclined to speak frankly. In this way, the question of whether to take sides in research became central, leading eventually to a commitment to *being seen* to take sides, while struggling to 'maintain marginality' and manage the inevitable partisanship that resulted (Hammersley and Atkinson 1983). My personal values, based on social justice and equality, made the prospect of 'siding' with the Uyghurs a very real threat. During the initial months, I was thus sometimes guilty of asking 'loaded questions', designed to encourage particular responses, i.e. the ones I wanted to hear (*ibid*: 23). It took a few unanticipated replies to remind me that the views and values of different individuals would not always or necessarily mirror my own, nor indeed those of other Uyghurs. Subsequently, I strove to manage this potential for bias in two ways:


When collecting data, I learned to guard against selective interviewing talking only with people I found politically sympathetic—and against the

<sup>8</sup> *Ätläs* (*Khan ätläs* or 'King's silk') is characterized by bold, multi-coloured patterns, created by a tie-dye or dye-resist process. It is popular in Xinjiang, Tajikistan and Uzbekistan.

#### xxii preface

selective recording of material (Hammersley and Atkinson 1983: 84). At the analysis stage, it was important not to ignore hostile data (material which did not accord with my personal views, or fit in to an established pattern), since that could result in selective presentation of data and distortion. Each account had to be explained *within the context in which it was produced*, always allowing that another respondent (or group of respondents) might provide a different account of the same circumstance or event (*ibid*: 112). While doing everything possible to grasp social meanings as experienced by respondents (become an 'insider'), it was crucial then to step outside their perspectives; after all, it is the sociologist's job to 'look at human situations in ways enabling them to see things that are not ordinarily seen by the participants in them' (Gouldner 1973: 57). While the battle to maintain the dual roles of insider and outsider was constant, this process nonetheless maximised opportunities to participate and reflect in a cyclical fashion. Furthermore, the longitudinal nature of this study provided the time and space to develop my arguments and hone conclusions concerning continuities and change in Uyghur identities and Uyghur-Han interactions over time (see Magnusson, Bergman and Rudinger 1994; Menard 2002; Ruspini 2002).

I am writing this book now for two reasons. First, I believe that its publication is timely. The ethnic riots which occurred in Ürümchi in July 2009 highlighted the urgent need for a current and in-depth understanding of social, economic and political conditions in Xinjiang, and of corresponding developments in Uyghur-Han relations. *The Art of Symbolic Resistance* is the product of a study of Uyghur identities and patterns of Uyghur-Han conflict and accommodation across two decades of socio-economic development and political change (1991-2011). Second, there is a scholarly need for a book of this type. In a review of the edited volume *Situating the Uyghurs between China and Central Asia* (Bellér-Hann et al. 2007), Professor Nicholas Tapp noted the 'striking omission' in the book of a detailed account of Uyghur-Han relations, which he described as crucial in Xinjiang: '…some account of Han migration into the region, and the resulting conflicts and accommodations that have been made, would have rounded this collection off more completely' (2008: 265). It is hoped that the present book can meet this need by providing a detailed analysis of the social, cultural, economic and environmental impacts of Han migration to Xinjiang, and of the ways in which these have influenced patterns of ethnic interaction and segregation.

#### preface xxiii

### *Structure of the Book*

The book comprises an introduction and seven chapters of original scholarly research, arranged in three parts. Part 1 (The Art of Symbolic Resistance) comprises an introductory chapter, in which I situate my arguments in relation to key approaches in the field and give a brief political history of the region, followed by a background chapter, which provides a detailed account of the socio-economic backdrop to the current conflict.

In the Introduction, I show how the book expands the spatial framework adopted in Bellér-Hann et al. (2007) by positioning the Uyghurs not merely between China and Central Asia, but also between China and the Middle East, and indeed China and the West. I set out a theoretical framework to support my core argument that Uyghur national identity involves a complex interplay between pre-existing intra-group identity and relational identities in contemporary times. I argue that culture has been politicised and deployed by Uyghurs as a means of symbolic resistance against both the Chinese state and the Han people. At the same time, however, I recognise mutual accommodations and, to this end, build on the concept of hybridity (Bellér-Hann et al. 2007) in a second sense, by illustrating the diversity between—as well as within—Uyghur individuals.

Chapter 1 (Inequalities: the socio-economic backdrop) sets out the conditions against which contemporary Uyghur identities and Uyghur-Han relations have developed. Here I combine secondary literature with my own field data to outline the extent of Han migration to Xinjiang and the impacts of state development. These include escalating inequalities in the spheres of language use, education, employment and wealth distribution; accelerated resource exploitation; environmental damage; and an absence of true indigenous political representation. In conclusion, I ask to what extent resistance among the Uyghurs may be linked to social (read: ethnic) inequalities, and whether it is sustainable to bring minority nationalities limited development without granting them equal empowerment.

In Part 2 (Sounds of discontent), I characterise the period from 1991 to 1997 as a run-up to an 'imagined independence', punctuated by the 'sounds of discontent'. I show how Uyghur actors from many different walks of life employed three powerful forms of symbolic resistance as a means to loudly and firmly reject state discourses of 'nationality equality' (*minzu ping-* *deng*) and 'nationality unity' (*minzu tuanjie*).9 In thus exercising their social agency, they challenged the legitimacy of Chinese rule.

Chapter 2 (Stereotypes) considers the first of three forms of symbolic resistance operating among Uyghurs prior to 1997: daily repetition of negative ethnic stereotypes of Han people. I consider China's 'civilising project' and the concepts of sinocentrism and Han chauvinism, and these provide the background against which to explore the counter-stereotypes deployed by Uyghurs to oppose Chinese hegemony and the Han people. Negative stereotypes are divided into three categories. Some reflect the perceived failure of Han people to conform to social and cultural practices taken for granted by Uyghurs, but not necessarily linked to Islam; others are linked to ethno-religious differences; and still others draw on Uyghur perceptions of Han territorial invasion and exploitation. I conclude with a discussion of how 'talk' bred hostility, segregation and conflict during this period.

In Chapter 3 (Boundaries), I consider the ways in which culture was manipulated by Uyghurs in the mid-1990s as a means to enforce symbolic, spatial and social segregation from Han people. I focus particularly on those criteria for cultural difference that were 'made to matter'. Symbolic boundaries were constructed around language use and language domain; time; intermarriage; and state birth control policy. Spatial boundaries ranged from Han-populated 'New Towns'/Uyghur-populated 'Old Towns' to segregated restaurants and university canteens. Social boundaries, based ostensibly on the religious need to keep separate from those with *haram* (unclean) diets, seemed in fact to be based on a broader disinclination to interact with Han people. While arguing that ethnic boundaries were designed to limit interactions solely to the workplace, I nonetheless recognise that interaction during this period could be managed *within limits*. I conclude that 'Us and Them' community dynamics served to create an (almost) blanket taboo on Uyghur-Han interaction in the mid-1990s, and that this caused an environment already fraught with cultural misunderstandings to deteriorate further.

Chapter 4 (Illuminists) deals with the third means of symbolic resistance operating in the mid-1990s: namely, alternative representations in Uyghur popular song. I focus on two Uyghur male folk singers, each characterised as the 'Voice of the Uyghurs'. Analysing their respective roles as

<sup>9</sup> Even after the Ürümchi riots of July 2009, which demonstrated serious social cleavages along ethnic lines, Xinjiang Regional Chairman, Nur Bäkri, insisted that the 'solidarity' amongst all ethnic groups in the region was 'as solid as a rock' (BBC Monitoring, 8 July 2009).

#### preface xxv

'illuminists', I ask to what extent each succeeded in their goal to 'awaken' the Uyghur people from their political stupor. After a consideration of 'new folk' music and the possibilities for mass mediation of ideas, I interpret key songs chosen from among the repertoire of two rival 'Voices of the Uyghurs': Ömärjan Alim and Abdurehim Heyit. I show how lyrical metaphor re-cast the fraternal relationship (as represented by the state) between the majority Han and the Uyghurs as one of coloniser to colonised; condemned the ethnic 'treason' committed by indigenous 'collaborators'; and urged the Uyghur community to unite against the outside threat. Finally, I explore regional, social class and religious bases for rivalry between the two singers, and suggest that new folk in 1990s Xinjiang was a locus of both ethnic solidarity and discord.

In Part Three ('Silence': after the 1997 Ghulja disturbances), I pinpoint the Ghulja disturbances of February 1997, their military suppression and the ensuing state crackdowns as a turning point: both a climax and an anti-climax in ethno-political terms. This event ushered in an 'era of silence', characterised by noiseless rather than vocal resistance, and a corresponding shift in means of symbolic resistance.

In Chapter 5 (Reverts), I outline how opposition since the latter half of the nineties has among a substantial portion of Uyghurs taken the form of religious renewal. Beginning with a discussion of the ways in which globalising forces (border trade, pilgrimage, study abroad, news media) aided the flow of traditionalist Islamic ideologies into Xinjiang during the eighties and first half of the nineties, I identify and analyse six sources of Islamic renewal. These include Islam as, variously, a form of local opposition to perceived national (and global) oppression;10 a response to failed development; a desire to return to social egalitarianism; a response to frustrated ethno-political aspirations;11 a return to cultural 'purity' (in reaction to modernity); and a vehicle for personal and national reform. In conclusion,

<sup>10</sup> Respondents frequently implicated intensified religious repression as a catalyst, describing the period prior to 1995 as *erkinlik* (free) and that since 1995 as *erkinsiz* (lit. freedom-less). In Xinjiang, the Muslim beard has become a visual sign of rebellion since the late nineties, attracting arrests and fines if sported by employees in state work units. State-controlled imams, meanwhile, are pronounced to have 'sold their beards' to the Chinese (for beard, read ethnic and religious conscience).

<sup>11</sup> Some respondents pointed to Islam as the sole remaining source of faith. Such remarks from within Xinjiang evoke views also espoused on émigré mailing lists, which assert that international organisations like the UN offer no real support to 'stateless nations.' Most Uyghurs now accept that they are totally alone, in stark contrast to the pre-1997 certainty that help would come from external sources.

#### xxvi preface

I evoke the Uyghur saying: 'All five digits on one hand are not alike' to emphasise the diversity of reasons underlying the phenomenon, and note that, in all the above senses, Islamic renewal in Xinjiang is currently symbolic, peaceful and cathartic rather than deriving from violent fundamentalist ideology.

In Chapter 6 (Endogamists), I examine the second form of symbolic resistance vigorously promoted since 1997: the intermarriage taboo. This is a boundary that was already in force back in the mid-1990s but which, following the demise of oral forms of resistance, has become paramount in the fight to ensure Uyghur linguistic and cultural survival. I illustrate how a mixture of parental prohibition and excommunication and 'community pressure' (social pressure exerted by ethnic peers) forcefully guarded against inter-ethnic courtship and marriage with Han Chinese. In exploring the dynamics of this taboo, I consider in turn racial, religious, cultural, historical and political barriers to intermarriage as perceived and articulated by Uyghur respondents. Couples who insisted on their bond frequently experienced community rejection and were forced to remove themselves to a 'neutral' context, such as China proper. However, I nonetheless recognise that in certain instances intermarriage (like ethnic interaction) could be managed if, for example, the Han partner submitted to religious conversion, learned the Uyghur language, and so on. Finally, I present a hierarchy of intermarriage preferences in contemporary Xinjiang, before concluding with a discussion of how marriage choices reflected ethnic honour or shame in a context of Uyghur national impotence.

Finally, Chapter 7 (Hybrids) focuses on *minkaohan* (Chinese-educated Uyghurs)12 and, more broadly, Uyghur urban youth. I look first at the selfidentities of first-generation *minkaohan,* showing how the experiences of these individuals often led to behaviour characterised by 'internalised oppression.' I then contrast these with the more positive self-identities maintained by second-generation *minkaohan*, illustrating how the latter have learned to negotiate hybrid identities to their advantage according to time, place and necessity. I propose that young Uyghurs may be located at various points on an 'accommodation spectrum' ranging from the pole of 'pure' (adhering strictly to orthodox Islamic practice) to the opposite pole of 'acculturated' (unskilled in mother tongue or culture), depending on the degree to which they have accommodated to a Han-dominated society.

<sup>12</sup> *Minkaohan* denotes a minority nationality individual who received a Chinese-medium education and whose first language is Chinese..

#### preface xxvii

Somewhere in the middle is located the high-school student, fluent in both Uyghur and Chinese, who flouts state prohibition by slipping to the mosque on Friday lunchtimes. Finally, I consider the empowering effects of education and globalisation on Uyghur youth, with reference to pop music production and consumption. I conclude that, given the imposition of Chinese as sole language of instruction since 2002, and accelerated globalisation of youth culture via diasporic movement and the internet, we might expect studies of multi-lingual and multi-cultural Uyghur youth to emerge as a major future trend.

My reflections in the Conclusion, based upon ethnographic material gathered during the three field trips I conducted between 1995 and 2004 and 'virtual' internet interviews conducted between 2010 and 2012, are tempered by recent discussions held with a long-term Uyghur respondent in 2013.

xxviii preface

PART ONE

preface xxix

THE ART OF SYMBOLIC RESISTANCE

xxx preface

# IntroductIon

introduction 1

In this Introduction, I aim to situate *The Art of Symbolic Resistance* in relation to key arguments and approaches in the study of china's minority nationalities more broadly, and within Xinjiang/uyghur studies in particular. General studies of china's minority nationalities tend to consider them from the perspective of the majority Han, be that the chinese state or ordinary Han people. In this book, I consider ethnic identity formation from a subjectivist point of view (Eriksen 1993: 37), as a deliberate process of 'self-ascription' (Barth 1969: 14; de Vos 1975: 16). In analysing the perspectives of in-group members, pieced together via interviews, observations, and textual analysis, I seek to question assertions that the uyghurs are to a large extent a 'creation of the modern chinese state' (Gladney 1990; Ben-Adam 1999). In exploring the social, cultural and religious resources embodied by contemporary uyghur identities, I illustrate how identity formation involves a complex Sartrean interplay between intra-group socio-cultural commonalities ('We-hood'), based on largely Islamic cultural values, assumptions and practices shared for over 500 years, and a common sense of enmity towards Han chinese ('us-hood') (Sartre 1943). I demonstrate too how uyghur national identity has partially—though not wholly—transcended earlier oasis-based, urban-rural and social cleavages; no longer is it limited to urban, intellectual circles (rudelson 1997). With this study, I hope to push the discipline forward on four levels: from state representations of uyghur identities to alternative 'local' representations; from a conceptualisation of 'uyghur as victim' to one of 'creative social agent'; from a focus on violent opposition to one on symbolic resistance; and from Xinjiang Studies (with their inevitably sinological framework) to uyghur studies, a field which recognises that contemporary uyghurs draw upon resources far pre-dating the advent of the unitary chinese nation-state in 1949, and reaching well beyond contemporary national borders.

# *Scholarship on China's Minority Nationalities*

Earlier scholarship on china's minority nationalities was aimed at a general readership and focused primarily on state policies towards the 55 minority groups (dreyer 1976; Heberer 1989; Mackerras 1994, 1995) or else a

substantial portion of them (Schwarz 1984). These works, while carefully researched and widely read, could not provide more than a broad survey of the population trends, stage of economic development, linguistic and cultural background, and level of social integration of each group. More recent works narrowed their focus to one minority group, for example, the Hui, the Uyghurs, the Zhuang, the Miao, the Mongols or the Naxi (Gladney 1991; Rudelson 1997; Kaup 2000; Schein 2000; Gillette 2000; Bulag 2002; Bellér-Hann et al. 2007; Yu 2010), or to one minority-populated region such as Yunnan, Xinjiang or rural south China (Halskov Hansen 1999; Starr 2004; Dillon 2004; Rack 2005). This allowed for a more in-depth analysis of the specific conditions of that group or region. A parallel but contrastive approach tackled the question of minority nationalities en masse but through examination of one specific social issue, for instance, education (Postiglione 1999), globalisation (Mackerras 2003) or migration (Iredale et al. 2003).

In terms of theory, the pioneering work on the relationship between China's minority nationalities and the state was Stevan Harrell's *Cultural Encounters on China's Ethnic Frontiers* (1994). In its Introduction, he showed how the Chinese state characterises the relationship between the centre and China's peripheries in familial and hierarchical terms: '… the juvenile image of the peripheral is presented in official discourse in which the Han and minorities are described as *xiongdi minzu*, or big-brother/little-brother *minzu*, with the Han (the socially and economically more advanced group) representing the big brother' (1994: 14). This concept has been reaffirmed and developed in the work of scholars such as Dru Gladney (1994; 2004), whose approach to majority-minority relations in contemporary China privileges state representations, and illustrates how minorities are exoticised and feminised so as to clearly define the modern [Han] Chinese nation.1 Susan Blum's *Portraits of Primitives* likewise explores how the majority group in China characterises those it regards to be 'other', although she extends the analytical framework to focus not only on scholarly and state discourses but also on portraits of minority nationalities painted in conversations and essays by 'people who are not experts or policy makers' (2001: 10-11; 13).

<sup>1</sup> See Ma Rong (1999) for an analysis of how the Han nationality, far from being homogenous, is an ethnic community formed through thousands of years of interaction and integration with other ethnic groups.

#### *Scholarship on the Contemporary Uyghurs*

Turning to academic treatments of the Uyghurs, the key proposition that has circulated for some time asserts that Uyghur national identity was largely 'created by the Chinese state', and that prior to 1949 the peoples of the region subscribed rather to an oasis-based or broad *musulman* (Muslim) identity (Gladney 1990). This argument hinges principally on the notion that, during the staged conversion to Islam which took place in the region, the ethnonym 'Uyghur', being associated with the Buddhist kingdom of Khocho [Ch. Gaochang, 850-1250] in Turpan, fell into disuse among the southern converts. By the mid-1400s, the Buddhist Uyghurs of Turpan had also converted to Islam, so that the ethnonym 'Uyghur' was abandoned altogether. For the next 500 years, Uyghurs identified themselves using terms denoting social group (e.g. 'merchant') or oasis origin, a tendency that many have argued reflects the geographical isolation of the region's disparate oases (see, for example, Oda 1978: 42). It is suggested that the ethnonym did not reappear until 1921, when Soviet advisors at a conference in Tashkänt proposed that the name 'Uyghur' be used to designate the peoples hitherto known by names denoting oasis origin. This proposal was duly adopted in 1934 by the then Xinjiang provincial government (Gladney 1990: 4), while the ethnonym was officially reinstated in the 1950s as part of the nationalities classification project (Fei 1980).

Based on fieldwork conducted in 1989-90, Justin Rudelson (1997) emphasises the continued separateness of oasis identities in contemporary times, and highlights cross-cutting social group allegiances among Uyghurs (intellectual, peasant and merchant). He suggests that although shared enmity towards the Hui (otherwise known as Tungans or 'Chinese Muslims'), who were perceived as allies of the Han administration, helped Turkic Muslims envisage themselves as a single group prior to 1949, today's Uyghurs are the Uyghurs *primarily because the government gave them that name*. In this way, the study of Uyghur intra-group identity was until recently neglected in favour of theories of relative ethnicity. The latter approach conceptualises Uyghur identity largely as a product of enmity towards the 'Other', be that Hui Muslims or the Chinese state and, as such, tends to *produce* the hypothesis that Uyghurs maintained no coherent group identity prior to incorporation within the modern Chinese nation. As far as the ethnonym is concerned, there is now some doubt over whether the ethnonym ever completely disappeared. Ablet Kamalov (2007) has shown that it was in popular use in the 1920s among Soviet turcologists,

who established the academic field of Uyghur ethnological studies. Still more significantly, in a recent interview (2013), a Uyghur historian pointed to the existence of archival materials, hitherto suppressed by Han historians, which demonstrate the continuity of the 'Uyghur' ethnonym *among highly educated persons* from the fifteenth through the twentieth centuries. Whether or not this material is ever permitted to come to light, the 'no name, no identity' theory ignores the existence of a host of over-arching cultural commonalities between oases prior to 1949.

The co-edited volume *Situating the Uyghurs between China and Central Asia* (Bellér-Hann, Cesàro, Harris and Smith Finley eds. 2007) signalled a decisive shift in the field by undertaking a close empirical examination of the complexities of Uyghur social and cultural life, and seeking to re-frame Uyghur identity as a hybrid located somewhere at the nexus of the Chinese and Turkic Central Asian civilisations. The following years saw the publication of a historical anthropology documenting Uyghur collective identity from the late 18th to mid-20th centuries (Bellér-Hann 2008) and a contemporary ethnography of personal and community identities among Uyghur men in Ghulja (Dautcher 2009), each rich in social and cultural detail. The present work builds on this new corpus by expanding the dual framework adopted in Bellér-Hann et al. (2007) and including a broader range of external influences, including Middle Eastern, mainland European, and Anglo-American cultures. At the same time, it bridges the gap between the old corpus (focused on relational aspects of identity) and the new (focused on intra-group identity). While bringing the Han Chinese back into the picture, and acknowledging the central role they—not just the state but equally the people—play in stimulating Uyghur ethnic identity, I show how the resources Uyghurs use to build symbolic resistance are based on a preexisting and long-standing intra-group identity, based on shared social, cultural and religious practices. Rudelson, despite asserting that Uyghur identity exists only as a result of the group's formal classification by the Chinese state, himself acknowledges that this classification 'effectively drew *a map within which the Uighurs already saw themselves living*' (Ben-Adam 1999: 192, my emphasis).2 Following Newby (2007), I will suggest that factors other than an ethnonym had provided a strong sense of identity and difference to the dwellers of the disparate oases long before the arrival of the Chinese Communists. These factors include common cultural assumptions, patterns of social interaction, religious practices and moral

<sup>2</sup> Ben-Adam is a pseudonym.

values, a shared history, and an attachment to the land, all pre-dating not just the establishment of the PRC but also the arrival in the Western Regions of Qing troops.

In exploring the interplay between relational and intra-group aspects of identity, I employ at base Harrell's concept of the 'asymmetrical dialogues between the center and the periphery' (1994: 7, 27). At first glance, this appears to resemble the proposition that minorities are 'created' through the intervention of the nation-state. Yet in his exposition of the development of ethnic consciousness, Harrell nuances the term 'creation' to denote either the development of something anew where it did not previously exist, or the further development—sharpening, focusing or intensifying—of *something that already existed*. Sharing at least 500 years' history of common social, cultural and religious norms since the last Buddhist Uyghurs converted to Islam in the sixteenth century (150 years of which are now richly illustrated in Bellér-Hann 2008), and a common attachment to the land developed during the course of over a millennium,3 it might justifiably be argued that the Uyghurs belong to the latter category. This acknowledged, we must incorporate pre-existing shared practices into discussions of evolving Uyghur identities and contemporary Uyghur-Han relations.

My second theoretical proposition concerns minority representation. A. Tom Grunfeld, writing in the mid-eighties, claimed that China's minorities have 'no independent voice' save for the Tibetans in exile, and the Kazakhs and Uyghurs living in the Soviet Union' (1985: 55). When Grunfeld stated his position, no Western scholar had studied in Xinjiang since 1949, thus one reason for the apparent lack of minority voice could have been the absence of a medium to report it. This situation changed at the end of the eighties, and several scholars have since demonstrated that minority nationalities in China may exercise considerable agency in the articulation of independent perceptions, narratives and practices. Ralph Litzinger (2000) focused on how elite members of the Yao minority population represented their own culture, history, and identity to a range of Chinese and Western observers. Louisa Schein (2000), while examining how Miao ethnicity is constructed and reworked by the state and non-state elites, considered also how the Miao demonstrate 'ethnic subjectivity' by repre-

<sup>3</sup> Until 840, the Uyghurs lived on the steppes of today's Mongolian republic. Following the collapse of the Uyghur Empire (744-840), they began three separate migrations, one of which went west, crossed the Tianshan mountains, and occupied the Turpan Basin (see Mackerras 1968, Haneda 1978, Geng 1984 and Barfield 1989).

senting and consuming aspects of their own culture. Mary Rack (2005) built on this work to show how state-sponsored ethnic minority cultural events, and the elite ethnic classifications embedded within them, stimulate 'explorations of alternative identities' by Miao urban elites in the rural south. Maris Boyd Gillette's (2000) study examined how a community of urban Hui uses consumption to re-position its members within the official Chinese paradigm of development, and demonstrated that modernisation is possible without the assistance of the state. Finally, Nimrod Baranovitch (2003) demonstrated how rock and pop music enabled marginalized groups in China—including minority nationalities—to claim a new public voice that is often independent of the state. And yet the assertion that China's minority groups have no control over their representation continues to be repeated. For example, in Gladney's view, minorities have had little choice in their media representation, while the lack of a free press makes it *impossible* for them to represent themselves without state mediation. He further asserts that no indigenous 'subaltern scholarship' has yet begun to develop in China (Gladney 2004: 78; 265; 362). While minority peoples may be powerless to prevent the dissemination of essentialised representations by the Chinese state, or the circulation of pejorative stereotypes by Han people, indigenous voices may nonetheless employ creative means to make themselves heard, as ethnographers working in Xinjiang in the mid-nineties discovered.4 Furthermore, Uyghur scholars are increasingly demonstrating the possibilities for a limited degree of local advocacy (see Sulayman 2007 and Dawut 2007). In acknowledging only direct methods of opposition (demonstrations, riots) and ignoring the less obvious yet enduring dynamics of symbolic resistance, we risk understating the possibilities for minority self-representation and deny contemporary Uyghurs social agency. In this book, I will show that individuals in Xinjiang are not merely passive recipients of state policies and representations; they are also creative agents capable of finding subtle, symbolic means of representing alternative identities and expressing oppositions.

#### *Identity from a Minority Perspective*

As Harrell (1994: 4) has argued, an understanding of the effects of a civilising project can be gained only by looking at it from the perspectives of *both*

<sup>4</sup> See Bellér-Hann (2002), Bovingdon (2002), Cesàro (2000), Harris (2002), Smith (2002; 2007) and Smith Finley (2007a).

sets of actors, i.e. the centre *and* the peripheral peoples. Yet so far only one monograph has explicitly addressed the construction of alternative discourses by China's Uyghur nationality. Gardner Bovingdon (2010) explores how multiple nationalisms in Xinjiang—Han state nationalism and Uyghur ethnic nationalism—compete through a 'politics of representation'. Within this framework, he focuses one chapter (Chapter 3) on what he terms the Uyghurs' 'counterstrategies' (Bovingdon 2010: 6). In this volume, I go a step further by taking the subjectivist approach to ethnic identity, according to which identities are selected by group members themselves, and an ethnic group can only be defined and structured *from within* (Eriksen 1993: 37). As children become young adults, they begin to consciously choose markers with which to build personal and group identities, ethnic identity being just one of these. During this process of self-ascription, the features taken into account are not merely the sum of 'objective' cultural differences but those which *actors themselves* regard as significant (Barth 1969: 14). In other words, in choosing particular aspects of culture as symbols of identity, in-group members exercise social agency by defining the ways in which they differentiate themselves from other groups (De Vos 1975: 16). Or, to put this in more recent terminology, individuals put together 'cultural toolkits' made up of symbols selected from the cultural repertoire available to them (Lamont and Mizrachi 2012: 368). These identity symbols prepare in-group members for social interaction, and warn others of the accepted boundaries of that interaction. While ethnic group identification can afford a sense of pleasure in belonging (Dashefsky 1976: 8), out-group perceptions of identity symbols can also lead to the creation of negative stereotypes, leading to ethnic prejudice and ethnic conflict. External prejudice can impact on some individuals psychologically, causing them to absorb negatively perceived traits, and leading to low self-esteem ('internalised oppression', see Chapter 7, section on first-generation *minkaohan*). On the other hand, it may fuel ethnic pride and serve to reinforce ethnic identity (see Chapters 2-6).

## *Sartrean Notions of We-hood and Us-hood*

Early ethnicity theory—termed the 'primordialist' approach—posited that 'objective cultural characteristics' distinguished ethnic groups one from another, while group membership satisfied the psychological and emo-

tional need to 'belong' (Shils 1957). It soon became clear, however, that while in-group members did not always maintain the same cultural characteristics continuously through time and space, they often continued to identify themselves with the group (Barth 1969: 12), creating the need to modify the primordialist approach. Evidently, ethnicity could not arise simply from a collection of fixed cultural traits; it must also involve other factors, which allowed for the continuity of the group even after the demise or transformation of those traits. The factor that came to be viewed as crucial was relativity. In his extensive research on psychological boundary maintenance in ethnic relations, Fredrik Barth wrote: 'Categorical ethnic distinctions do not depend on an absence of mobility, contact and information, but do entail social processes of exclusion and incorporation' (1969: 9). Thus, ethnicity cannot occur in isolation, but develops when an ethnic group is in regular contact with another group or groups from whom it considers itself substantially different: '*I* identify myself with a collective *we* which is then contrasted with some *other*...What *we* are, or what the *other* is will depend upon context' (Edmund Leach, cited in Epstein 1978: 100). To explore this further, we can draw on Sartrean theory. Jean-Paul Sartre considered 'We-hood' to mean shared experiences *within* the group such as language, religion, position in the labour-force or notions of origin. He described 'Us-hood', on the other hand, as a sense of loyalty among group members created through enmity towards or competition with the 'others' (Sartre, 1943). As Eriksen has argued, both We-hood and Us-hood are essential for an ethnic category to come into existence; not only must group members have similar cultural experiences in common, they must also share a marked sense of feeling different from others in their environment. Ethnicity thus involves commonalities (complementarisation) *and* differences (dichotomisation), and group identities must always be defined in relation to that which they are not (Eriksen 1993: 35). It is only when cultural differences affect interaction on a regular basis, i.e. when they are *made to matter*, that the social relationship assumes an ethnic element (Eriksen 1993: 10-12).

The concepts of We-hood and Us-hood can be usefully applied to the impact of the Chinese 'civilising project'. Uradyn Bulag has characterised China as 'a mosaic of many territorial nationalities whose historic homelands have been incorporated into the modern Chinese state and whose positions have been transformed from being sovereign or semi-sovereign people on China's periphery to [being] minority nationalities' (2002: 8). The Dai of Xishuangbanna, for instance, were forced to shift from a

consciousness of themselves as an independent nation to one of an ethnic group within a larger political system (the Republic of China, then the PRC). Previously, they had enjoyed a 'fully formed ethnic identity' as a solidary group organized around their own lords, distinct from the subordinate upland peoples in their kingdom (Harrell 1994: 32-33). The imposition of the cultural values of the centralised Chinese state and its Han majority engendered in peripheral peoples a particular reaction, and in fact 're-emphasised the very difference between centre and periphery that the civilising project ostensibly sought to eliminate' (Harrell 1994: 17). This 'difference' is none other than the pre-existing, taken-for-granted social and cultural practices of the peripheral group (We-hood); the 'reaction', the group's heightened awareness of those practices in relation to the contrastive practices of the centre (Us-hood). My fieldwork data strongly suggest that, since the time that Rudelson was conducting research in Turpan, contemporary Uyghur identity has undergone significant changes, in response both to changing socio-economic circumstances within Xinjiang and to the changing international political context.

### *The Art of Symbolic Resistance: From Cultural to Religious Symbols*

Following the collapse of the USSR, and the formation of the Commonwealth of Independent States (CIS) in 1991, the Uyghurs became the only sizeable Central Asian population without an independent state to their name. The following years saw the rapid growth of hope and nationalist aspiration among young males, fuelled by a number of fanciful predictions, which I refer to elsewhere as 'the 1997 theories' (Smith 2000). These included the notion that China might attack Taiwan as a result of cross-strait tensions; that the United Kingdom might refuse to hand back its colony, Hong Kong, and that an Anglo-Chinese war would ensue (cf. Bovingdon 2010: 92); and that Chinese leader Deng Xiaoping's imminent death might lead to a power struggle within the CCP, resulting in large-scale civil unrest. Any one of these situations, it was hoped, might create the necessary conditions of domestic instability for Xinjiang to seize its independence.5

<sup>5</sup> Colin Mackerras (2001: 289; 300), while accepting that there were long-standing domestic reasons behind the growth in separatism in Xinjiang, argued that it was the collapse of the Soviet Union 'combined with the rise of ethno-nationalism in an era of globalisation' that provided the impetus for an ethno-nationalist movement in Xinjiang. In a later article (Mackerras 2005: 8), he lists a number of other catalysts, including Muslim demonstrations in China in 1989, the Baren rebellion in south Xinjiang in 1990, the rise to

These years saw urban Uyghurs deliberately separate themselves from the Han using a combination of strategies. Simultaneously and, I suggest, not coincidentally, there was a sharp increase in separatist violence, although this was confined to a small number of individuals prepared to take direct action against the state.6 In addition to the incidents documented by Millward (2004) and Dillon (1995; 2004), there were other cases which went unreported. Polat, a Chinese-educated taxi driver in his forties, described in 2002 how Uyghur activists in Atush had been involved in an armed conflict with the authorities in 1994, following which he claimed the perpetrators were 'taken into the hills and shot'. In a particularly dramatic incident in 1996, nine activists stormed the home of a local Uyghur leader in Kucha, slit his throat and those of his wife, younger brother and brother's wife, and cut out their tongues. Given that the leader was widely accused of 'letting the Uyghur people down and failing to represent them', this act may be interpreted as a silencing of perceived 'ethnic traitors'.7 Yet despite these incidents, it must be emphasised that Uyghur collective resistance in the 1990s was on the whole non-violent.

Scholar and human rights worker, Nicolas Becquelin, has observed that ethno-national movements 'have dynamics of their own, which have the potential to override obstacles set up by the dominating power' (2004a: 377). While a minority of pro-independence activists in the mid-1990s employed direct and violent methods of resistance against symbols of the Chinese state (including sabotage of government installations, and assassinations of police, military personnel and indigenous officials), the desire for separation was usually subtly expressed via symbolic means. Prior to the Ghulja disturbances in 1997, the main vehicles for this were:


power of the Taliban between 1996 and 2001, and the acceleration of modernisation in general.

<sup>6</sup> Although James Millward (2004: 16) suggests that the high numbers of arrests linked to separatism in 1996 could result simply from the introduction of the Strike Hard campaign itself, local testimony supports the idea of a rising trend leading up to the Ghulja disturbances of 1997.

<sup>7</sup> Personal communication with the daughter of a local Party official. See also Beckley (1997) and Bellér-Hann (2002: 78).

3. Dissemination of alternative representations of Uyghur-Han relations via the medium of popular song.

In contrast to Rudelson, who dismisses verbal attacks, jokes and curses against the government and Han people as 'passive' and of little import (Ben-Adam 1999: 209-10), I argue that daily repetition among Uyghurs of feelings of cultural antipathy, moral disapprobation, and anger served to underline differences, reinforce hostility, and heighten Uyghur-Han tensions. Moreover, I show how negative stereotypes conveyed via home/family education quickly reproduced in the young generation a multiplicity of negative feelings towards Han people. This social effect is neatly exemplified in an incident occurring in the mid-1990s where a small Uyghur child appearing on live television refused to speak to a Han person introduced to him as 'Uncle', and instead rejected him using the derogatory term *Khitay*.8 Needless to say, this event quickly became the subject of popular story-telling.

Following the Ghulja disturbances of 1997, their military suppression, and the subsequent series of 'Strike Hard' campaigns against 'splittism' (*fenlie zhuyi*) and 'illegal religious activities' (*feifa zongjiao huodong*), the situation radically changed. In the eyes of the state and many of its Han subjects, these disturbances seem to have represented a precocious attempt to break away from the state patron. As Blum once pointed out, the Han Chinese expectation that all within their universe will respect their superiority causes them to express astonishment at the 'ingratitude' of those minorities who 'inexplicably want the Han to leave' (2001: 126-7). In daring to pursue the 'separatist' dream (at least according to the state's version of events), disaffected Uyghurs had failed to fulfil their citizen's duties of safeguarding nationality unity and the unification of the motherland, and were thus labelled 'ingrates'.9 Heavy restrictions were placed on freedom of speech, so that any expression of Uyghur dissatisfaction became a punishable crime, labelled 'local nationalist' or 'splittist' (Becquelin 2004a: 375).10 In Ghulja, a 'Three-No' policy was established: no

<sup>8</sup> For the historical origins of the term *Khitay*, see Millward and Perdue (2004: 43).

<sup>9</sup> See my article on the Ürümchi riots 2009 (Smith Finley 2011a) for an expansion of this idea.

<sup>10</sup> There are echoes here of oppression by warlord governors in Republican Xinjiang. During Yang Zengxin's rule (1912-28), signs in restaurants forbade political discussion (Mackerras 1995: 20). In the late 1930s, when Sheng Shicai suppressed his Turkic and Tungan (Hui) opponents, the Soviet secret police network (*Bao'andui* or Security Preservation Corps) was reportedly so feared that conversation in Qumul was restricted to melon grow-

questioning [the riots], no telling [about the riots] and no visiting [those imprisoned following the riots].11

Following the events of 11th September 2001, China's leadership saw an opportunity for manoeuvre, and re-positioned itself alongside the United States in the global 'war against terror' (Millward 2004). From this point forwards, the Chinese government defined all expressions of discontent in Xinjiang—whether violent or non-violent—as 'separatist' and 'terrorist'. In so doing, it hoped to 'obtain *carte blanche* from the international community to take whatever action it sees fit in the region' (Dillon 2004: 156-57; see also Becquelin 2004b). The situation worsened in 2002, when the public recital of an allegedly 'nationalist' poem during the *Nawruz* celebrations in Ürümchi led to intensified censorship of the cultural sector.12 Cultural products 'in the area of the ideological front' were now deemed to have had 'an adverse impact on society', and the Chinese authorities began to claim that those now using arts and literature to 'distort historical *facts'* were the same people who had employed violent, terrorist tactics in the past (Becquelin 2004b: 43-44). As a result, two means of symbolic resistance previously in operation—negative oral stereotyping and alternative representations in popular song (and popular culture more broadly)—were no longer viable. Under these circumstances, Uyghurs increasingly found themselves facing a seemingly bi-polar world, in which they could either resist and face marginalisation, or accommodate to ensure personal survival.

Yet symbolic oppositions are necessarily fluid: they emerge, adapt and disappear in response to changing political conditions. This dynamism of symbolic boundaries is 'a necessary element in responding to the ongoing challenges of being marginalised' (Vasquez and Wetzel 2009: 1570). While the third means of symbolic resistance—ethnic segregation—continues to be enforced by most Uyghurs (an exception is a small number of deeply acculturated school-age *minkaohan*), and now includes a strengthened emphasis on endogamy (see Chapter 6), a new symbolic boundary has emerged: Islam itself. Since the turn of the millennium, new scholarship has documented the return to the mosque and the revitalisation of orthodox Islamic practice in Qäshqär (Waite 2003; 2007) and Ürümchi (Smith

ing, the weather and market prices. When a person 'disappeared', he/she was said in code to have 'gone to Chöchäk,' a remote place close to the Soviet border (Forbes 1986: 151; 155).

<sup>11</sup> Details are found in Classified document No. 175 (Dillon 2004: 95).

<sup>12</sup> *Nawruz* (Persian) is the first day of Spring and the start of the year in the Iranian calendar. It usually occurs on March 21 or the previous/following day, depending on where it is observed. Central Asian peoples and some Uyghurs in Xinjiang have recently also begun to celebrate this day. The young Uyghur poet was Tursunjan Emet.

Finley 2007b), as well as the economic, social and psychological (cathartic) role increasingly played by Sufi pilgrimages (Harris and Dawut 2002; Dawut 2007). The use of Islam as symbolic resistance is to be carefully distinguished from ethno-religious *culture* (everyday practices influenced by Islamic mores), which formed the basis for counter-stereotypes employed by Uyghurs in the mid-nineties to distinguish the in-group favourably against the Han. While many in-group practices are informed directly or indirectly by Islam, their ethnic (rather than their religious) quality was stressed in the mid-nineties. Since 1997, however, the emphasis has shifted from the ethnic quality of practices to the religious. For instance, intermarriage is currently deemed impossible on grounds of religion, whereas prior to 1997 it was more often said to be obstructed by different ethnic customs and habits.13

The 'pure' route to an orthodox form of Islam is not however being taken by all Uyghurs. As some predicted earlier, 'the dynamics of penetration and resistance between the centre and […] a periphery populated by indigenous peoples can be expected to generate at the same time increased sinicization and increased ethno-national unrest in the future' (Becquelin 2004a: 378). Thus, at one end of the spectrum, we find Islamic 'reverts': individuals who have embraced the ritual of *bäsh waq* (five prayers a day), have given up alcohol and cigarettes, and link the fate of the Uyghur nation directly to their personal, moral salvation. At the other extreme, we find (a small number of) deeply acculturated youths, largely unskilled in their mother tongue, who sometimes criticise the shortcomings of the in-group, and hanker after all things Han, including in some cases a Han romantic partner. Dotted between the two, we find an assortment of 'hybrid' individuals who more or less combine different aspects of religious and cultural affiliation with a Chinese-language education in a bid to survive in an increasingly competitive socio-economic landscape. The multiple facets and layers of their identities may be selectively mobilised and fore-grounded at different times and in different contexts, so that their position on the spectrum also fluctuates according to context (cf. Gladney 1996). Among them are a set of would-be urban 'cosmopolitans'; new world citizens, who via their preference for Central Asian, Turkish, Middle Eastern, Mediterranean and Anglo-American cultural products, reach for positive globalised

<sup>13</sup> Of relevance here is Colin Mackerras's musing: 'An interesting question […] is whether the revival of identity feelings among Muslims in China is primarily *ethnic* or *religious*. It could of course be both, but one of them would possibly still predominate …' (2005: 5).

identities independent of the state's representational machine and transcendent of national borders.

In this complex climate of re-Islamisation, acculturation and globalisation, poles of religious and secular have perhaps inevitably become more pronounced among both elites and the common people. In particular, the growth of reformist Muslim ideologies has created the potential for increased religious debate. Based on his ethnographic research in Qäshqär, Edmund Waite describes a range of popular reactions towards reformists. These include on one hand resentful accusation—'they keep their wives secluded—like in a cage'—and on the other hand positive appraisal—reformists are 'good Muslims who believe in reform (*islahat*) and progress (*täräqqiyat*)' (2007: 172). Yet the division between religious and secular in Xinjiang is in general far from clean-cut. The divide between nominal or partially observant Muslims and strictly observant Muslims cannot be neatly explained in terms of *minkaohan* (Chinese-educated = secular) and *minkaomin* (Uyghur-educated = observant) as one might expect. Instead, my interviews reveal that while some Uyghur-educatedindividuals now go to university and embrace secularism, some Chinese-educated individuals, albeit a minority, are today embracing religion. Uyghur intellectuals tended in the past to follow secular ideologies, and this continues to be the case for a considerable proportion of this group. Consider this excerpt from an interview in 2004 with Tashmämät, Uyghur-educated intellectual in his forties originally from Qäshqär:

Political resistance can and often does take the form of religion. But we, the intelligentsia, don't have much time for religion. We don't see it as the best path for Uyghurs to take. The *taliplar* (religious students) don't have a detailed idea of what they want to create; they haven't thought it through on a deep level. Most ordinary Uyghurs think very illogically […] Very few—only the intellectuals—are able to think about issues calmly and rationally.

His view reflects not only the widening gap between reformist and secular thinkers but also the traditional disdain of the intellectual class for the opinions of the layman (see also Chapter 4 on cultural elites). Other intellectuals, however, have begun to gravitate towards Islamic and Arab culture in their research, and may themselves be experiencing a parallel process of re-Islamisation to sections of the lay public (see for example Rewaidula and Rewaidula 2004). Tashmämät and his colleague, Shökhrät, also a Uyghur-educated intellectual originally from Qäshqär, complained in 2004 that there was currently far too much 'blind Arab worship' in contemporary Uyghur scholarship: 'Everything is interpreted in terms of Arab culture, Arab civilisation … but the truth is that Uyghur history and culture are much more nuanced, much more complex than that.' The trend is reflected among the public in a growing preference for Arabic over Uyghur names (Sulayman 2007; Waite 2007).

Thus, a new intra-ethnic dichotomy has emerged: one which separates 'those who pray' from 'those who don't'. A foreign manager at a 5-star Ürümchi hotel, also interviewed in 2004, categorised the city's inhabitants into two clear (albeit simplistic) types: the 'modern Uyghur' who has lost his sense of history, and the 'traditional Uyghur' who remains 'pure' through adherence to religion. I observed in 2002 and 2004 that, in order to cater for both types, many Uyghur restaurants now have separate floors, with those who smoke and drink seated on one floor and those who abstain seated on the other. Gülhärä, a 20 year-old Uyghur-educated female student from Döngköwrük, confirmed that segregation in restaurants was a new trend in Ürümchi: 'You see, people who don't drink don't want to sit with people who do.' Two male migrant workers from the south dated the change from around 1998, and suggested that restaurants now separate smokers and drinkers from abstainers because they can afford to, i.e. have made enough money to re-invest in larger premises. Nor is this new system limited to Xinjiang; I observed similar practices in a Uyghur dance restaurant in the eastern Chinese city of Shanghai in 2013.

The intra-ethnic dichotomy is also mirrored in other forms of interaction. While in the mid-nineties Uyghurs would introduce each other in terms of whether they were Uyghur-educated or Chinese-educated, in the 2000s I observed how individuals had begun to introduce one another rather in terms of *whether they prayed*. The implication was that if a person actively practised Islam he/she was a good and reliable person. Sometimes the distinction was based on a direct contrast between Ürümchi folk and Uyghurs from other towns, with the former described as 'no good', 'broken' (*buzuq*), or 'having no religion and no ties to their people.' It also appeared to inform choices about whom one could trust. One day in 2002, while discussing sensitive political topics in a cafe with Abdukerim, an observant university student from Aqsu, and his friend from Ghulja, I discovered that the two had, incredibly, met one another that same day at the mosque. When I asked them how it was that they could trust each other enough to discuss politics in one another's presence, Abdukerim replied: 'The ones who pray trust one another [*namaz oquydighanlar bir-birigä ishinidu*]. And the young people trust one another. The old people are scared of talk of independence. So are those who were born in the Cultural Revolution.

But those who were born after 1968 … .!' The view they expressed implied that only non-practising Muslims could potentially be *espion* (spies), and connected to a remark Abdukerim had made on a previous occasion: 'Uyghurs need religion to unite them.'

#### *A Brief Political History*

In order for the account of symbolic resistance and boundary maintenance which follows to make sense, it is necessary to set it explicitly in political context. Only by doing so can the reasons for that resistance and for the protection of boundaries be fully understood. As is widely known, the territory now called Xinjiang did not fall decisively under Chinese rule until its conquest by the Manchu Qing empire in the mid-eighteenth century. Following, there were several local rebellions, some of which resulted in a temporary state of independent rule, while others did not. Following an unsuccessful uprising launched by Manchu and Uyghur officials in Üch Turpan in 1765, the region fell quiet until 1828, when descendents of a Sufi clan known as the Khojas (former rulers of the Tarim Basin oases) launched a series of invasions of southwest Xinjiang. These attacks did not however overthrow Manchu citadels, and ended in the 1850s (Tsing 1961; Saguchi 1978; Millward 2004: 2-3). When rebellion broke out among Hui Muslims in Northwest China in the 1860s, it quickly spread to neighbouring Xinjiang, where Uyghurs joined Hui rebels and sometimes staged separate rebellions in the cities. Subsequently, Uyghurs dominated the south and west, while the Hui controlled the east and north. In the interim, Yaqub Beg, an adventurer from neighbouring Khokand, established an emirate extending from southern Xinjiang to Turpan, which lasted for twelve years (Chu 1966; Kim 1986, 2004; Millward 2007: 117-118). In the late 1870s, the Qing re-conquered most of Xinjiang, and in 1881 Russia was forced to return most of the Ili Valley territory it had previously annexed. Thousands of local Turkic Muslims (known then as 'Taranchis', today as Uyghurs) emigrated to the north at that time rather than fall again under Qing rule (Millward 2004: 2-3).

At the end of the nineteenth century, influential industrialists and merchants in Qäshqär, Turpan and Ghulja launched a new movement to modernise Uyghur education, similar to the 'jadidist' ('new method' education) movements developing elsewhere in Central Asia. These reformist schools instilled Turkic—and subsequently Uyghur—nationalist ideas in young Uyghurs during the 1920s, many of whom then got involved in the rebellions

of the 1930s (Millward 2004: 4; Bellér-Hann 2008: 58). The first, known as the Qumul Rebellion, occurred in 1931 when the predatory behaviour of a Chinese military commander towards a local Uyghur woman resulted in his assassination and a series of uprisings against the Chinese warlord administration in Ürümchi. A confused period of shifting political and ethnic loyalties followed, involving the entry of Hui warlord, Ma Zhongying—'Big Horse'—from Gansu (supported by the Chinese nationalist— Guomindang—government in Nanjing), mutiny among Han forces in north Xinjiang, mutual competition between Uyghur armies, and an alliance of Chinese warlord Sheng Shicai (based in Ürümchi) with the Soviets. Between November 1933 and February 1934, a 'Turkish-Islamic Republic of East Turkestan' (TIRET), which opposed both Chinese and Soviet influence, briefly flourished in Qäshqär (Hedin 1936; Forbes 1986). It would later be characterised as 'a milestone of Uyghur nationalist history', and is frequently cited by advocates of Uyghur independence today (Millward 2004: 5).

From the early 1940s, the Guomindang (GMD) government re-established control over Xinjiang, ousting Sheng Shicai, and cutting off Soviet influence. In 1944, the Rebellion of the Three Regions (*Üch Wilayät Inqilabi*) erupted among Kazakhs and Uyghurs in the north, allied with Kirghiz in the south, in response to warlord rule, economic hardship and Han chauvinism among GMD officials (Lattimore 1950; Forbes 1986). Initially characterized by interethnic competition and ambiguity towards the role of Islam, by the summer of 1945 it had evolved into a secular, socialist government backed by the Soviet Union: the 'East Turkestan Republic' (ETR), based in Ghulja, northern Xinjiang. Under Soviet pressure, the ETR entered into a coalition government with the GMD in Ürümchi in August 1945; however, tensions remained high, and the ETR is said to have 'continued virtually without GMD interference' in the north (Millward 2004: 5). Finally, in late 1949, Chinese Communist forces occupied Xinjiang, facing only minor military resistance, mainly from independent Kazakhs under the leadership of Osman Batur (Shih 1963; Millward 2004: 5).

This brief political history suggests several recurrent themes helpful for an understanding of modern Uyghur identities and of Uyghur 'separatism' in the 1990s. The 1864 Muslim Rebellion was the first instance in which the region's Muslims (Hui and Uyghurs) united against a common non-Muslim enemy (Manchus and Han), using influential Sufi leaders as figureheads. From then on, Islam became a common focus for the mobilisation of Muslim groups against external rule. However, a pattern can be discerned whereby alliances forged between ethnic groups along Islamic lines quick-

ly broke down once victories had been won. In addition, divisions in political ideology are perceptible, with different trends emerging in the north, south, and east of the region. Since at least the early 1930s, Uyghurs and other Muslim groups in south Xinjiang have tended to mobilise political opposition along Turkish-Islamic lines. TIRET, the independent administration set up in Qäshqär in 1936-1937, was staunchly anti-Soviet and anti-Han, but also anti-Hui, considering this non-Turkic group to be allies of the Han (cf. Bellér-Hann 2008: 59). It looked to reformist Islamic regimes like Turkey, Afghanistan, and the Middle East for inspiration and support. On the other hand, political action in the north, despite initially using Islam as a spiritual focus, has tended to advocate a secular secession. In the north, therefore, it has been possible to unite Muslim and non-Muslim groups against Han rule. The progressive northern secessionists of the 1940s maintained close links with the Soviet Union, had often been Soviet-educated, and are known to have used sophisticated Soviet-made arms. Still, there were some Turkish-Islamic partisans within rebel ranks who opposed the notion of a secular secession. In the eastern oases, the story is different altogether. Local rulers in Turpan and Qumul voluntarily submitted to Qing encroachment in the 1700s, and, with the exception of the Qumul Rebellion in 1931, there have been no anti-Chinese uprisings in these areas. The fact that the Qing dynasty established a Muslim banner system in the eastern oases while deeming it necessary to appoint indigenous *beg* officials in the southern oases further indicates the comparatively amicable relations the east has traditionally enjoyed with China.

Another characteristic of uprisings in Xinjiang before 1949 is that Uyghurs rarely took the initiative in opposition movements. Rather, they tended to follow the leads of others and voice their grievances on the back of others' movements. For example, the 1864 Muslim Rebellions were in each case begun by the Hui, with Uyghurs joining in later for their own reasons (a century of Manchu exploitation of resources and labour in Xinjiang). Similarly, it took an outsider in the form of Yaqub Beg from Khoqand (located in modern-day Uzbekistan) to unite Uyghurs and other groups against Qing rule for twelve years from 1864 until 1877. While Yaqub Beg is a popular folk hero and nationalist symbol for many Uyghurs today, histories of this period indicate that many locals considered life under the Qing to have been more comfortable: 'Because the rulers were mostly foreigners, people began to see them as the exploiters of their toil, not as the protectors from the infidels' threat' (Kim 1986: 189). The popularity of Yaqub Beg among Uyghurs today cannot emanate from his actual policies of that

time, since they were in many cases detrimental to local interests (see Kim 1986). Rather, it appears to stem from the Uyghurs' yearning for a leader with strength and charisma. Throughout their history, Uyghurs produced no leader who could effectively unite them against an outside threat. Finally, the Ili Revolt of 1944 was begun by Kazakhs, although Uyghur progressives later came to prominence in the ETR.

Looking at patterns of Uyghur protest in Xinjiang in the reform era (see Dillon 1995, 2004; Smith 1999, Chapter 3; Millward 2004), four main sources of discontent can be identified:


In the 1980s, Uyghur protesters alleged Han ethnic (cultural) discrimination on several occasions, manipulating Chinese state discourse to call for 'nationality solidarity *on the basis of* equality between nationalities'. Notably, these demonstrations were peaceful, and there was no call for secession. On the other hand, grievances over socio-economic inequalities, perceived to result from accelerated Han in-migration, led to violent protests and were often linked to the call for 'self-rule'. Reactions to nuclear tests conducted at Lop Nor by the Chinese government initially took the form of peaceful protest. However, following a string of experiments conducted between the eighties and nineties, including the detonation of a bomb in 1987 estimated by Nils Olof Bergkvist, a spokesman for the Swedish Defense Ministry's Hagfors observatory, to be at least six times as powerful as the bomb that levelled Hiroshima (*Los Angeles Times* 1987), opposition became violent. Kyrgyz and Uyghurs who rioted in south Xinjiang in 1990 identified the tests as one reason for their disaffection.

After 1990, earlier calls for more (or real) autonomy were replaced by overt demands for independence, and violent protest along ethno-political lines became more frequent. But, once again, the nature of protests differed by region. The Baren riots of 1990 in the south suggested the use of Islam to mobilise opposition to the Chinese state. However, during armed insurrections taking place in Dacheng and Bole on the Russian border in 1991, there was no mention of 'holy war' or of 'Eastern Turkestan', suggesting that strategy was different north of the Tianshan (Heavenly Mountains). Rather, some insurgents demanded independence from China, while others

requested the right to become USSR nationals (i.e. be repatriated to the Soviet Central Asian homelands). Some Uyghurs in the northern oasis of Ghulja, on the other hand, began in 1995 to call for the establishment of 'Uyghuristan,' a name that potentially reduces other nationalities in the region to the status of minorities under Uyghur control. Violent incidents took the form of attacks on Han cadres, police, and military, assassinations of perceived 'collaborators' among indigenous Uyghur cadres, and occasional bombings. The violence climaxed in 1997, when several hundred local people in Ghulja, enraged by increasingly acute restrictions on religion (including the banning of the all-male social gathering known as the *mäshräp*), came out in protest on the streets. Banners and slogans included calls for 'ethnic equality' and 'independence', as well as expressing religious sentiments. Protests continued for several days, resulting in the arrest of demonstrators, and eventually turned violent. While anti-riot police and troops reportedly used dogs, tear gas, fire hoses, beatings, and live ammunition on both demonstrators and bystanders, rioters are said to have torched vehicles and attacked Han police and civilians.

It is of course no coincidence that the new calls for independence followed certain influential domestic and international events, including the pro-democracy movement in Beijing's Tian'anmen Square (1989), the collapse of Marxist-Leninist parties in Eastern Europe (1989), and the disintegration of the USSR and formation of independent Central Asian states (1991). Moreover, Uyghurs in Xinjiang were further encouraged to aspire to political change by the series of humanitarian interventions made by highprofile multilateral organizations in the 1990s on behalf of certain peoples:

If the leaders of nationalist movements believed it likely that powerful actors would lend a hand, they were likely to think and act more boldly […] the message that national self-determination was again possible and that it might enjoy international support spread to other parts of the world […] Learning of NATO's intervention in Kosovo, many Uyghurs hoped that foreign assistance would make this [a sovereign Uyghur nation] possible, since they could not achieve independence on their own (Bovingdon 2010: 2-3).

Yet following the climax of violence in the 1997 Ghulja disturbances, the situation in Xinjiang quietened down, partly as a result of the Strike Hard punitive campaign (Millward 2004: 28; 32), but mainly because most Uyghurs now saw clearly that foreign assistance would not be forthcoming, a sinking realisation compounded in August 2002 when the US embassy in Beijing announced that it considered the East Turkistan Islamic Movement

Figure 2: Wood-turners at the handicrafts alley, Qäshqär Old Town (1996)

(ETIM) as 'a terrorist organization associated with al Qaeda' (Millward 2004: 13).

All this begs the question of how accurate a descriptor the word 'separatist' really is when referring to contemporary Uyghurs as a whole people. Throughout the relatively short period of increased violence during the 1990s, available data sources indicate that some social groups (e.g. young male intellectuals and young unemployed men) were more involved in separatist activities than others (e.g. women, middle-aged and elderly Uyghurs) (cf. Smith 2000). It is important to take a nuanced approach to this question, one that is facilitated by consideration of Anthony Smith's (1981) theorisation of the concept of 'separatism'. Smith distinguishes between 'territorial' and 'ethnic' separatism, where 'territorial separatism' is based merely on political boundaries, geography and a sense of *political* community (e.g. British immigrants to North America; Creoles in Latin America; Australians and New Zealanders in Oceania), while 'ethnic separatism' is based on a sense of being united by, and differentiated from others by, an alleged 'community of *culture*'. In this latter case, a group considers itself different from its rulers or neighbours in one or more significant cultural dimensions, so that 'separation' becomes a means of protecting the group's

cultural identity. In certain spatial / temporal contexts, the cultural uniqueness of an ethnic community may demand political separation, so that it can 'run its own affairs according to inner laws of the culture community, uncontaminated and unmolested by external influences' (Smith 1981: 12-13). In these cases, territory and geography feed the sense of unique culture, and are an integral part of that culture (Smith 1981: 14). However, the desire for political separation does not necessarily or always lead to 'political separatism': the wish to secede and form one's own sovereign state (Smith 1981: 16). While 'nationalism' normally seeks to extend the scope of ethnic community from purely cultural and social to economic and political spheres, this too does not automatically suggest full secession. It is also possible for a culturally distinctive ethnic group to seek 'autonomism', defined as full control by ethnic group members over every aspect of cultural life, and in some cases extended to all aspects of political, social and economic life (Smith 1981: 19). Such a configuration is potentially achievable through federation (Canadian federalism provides a useful example).

Applying these concepts to Xinjiang as a case study, I would tentatively suggest that while the number of Uyghur individuals seeking full secession (i.e. pursuing 'political separatism') undoubtedly increased during the midlate 1990s, the number of those participating in violent, direct action to achieve this was small. While many other individuals surely longed for independence in their hearts, they cannot strictly be called 'political separatists' because they did not take concrete action towards that goal. Still others perhaps never really believed that independence was either achievable or desirable, preferring rather to concentrate on the pursuit of full equality for Uyghurs within the Chinese political system (i.e. 'autonomism'). Since the Ghulja riots in 1997, this last view has arguably come to dominate the debate within Uyghur society.

#### Chapter One

inequalities: the socio-economic backdrop 23

# InequalItIes: the sOCIO-eCOnOmIC BaCkdrOp

In this first chapter, I set out the socio-economic background against which contemporary uyghur identities have developed and uyghur-han relations have deteriorated. I combine my own empirical data gathered in Xinjiang between 1995 and 2004 with published research findings focused on China's development policies—including the Great Western development (*Xibu dakaifa*) strategy—in order to demonstrate the impacts of han in-migration and a changed population composition. I ask to what extent resistance among the uyghurs may be linked to escalating social (read: ethnic) inequality and discrimination in the spheres of language, education, employment, wealth distribution, resource exploitation and political representation, and question whether it is sustainable to bring minority nationalities limited development without granting them equal empowerment.

#### 1.1. *Han In-migration*

prior to 1949, most han people immigrated to Xinjiang spontaneously. the majority maintained strong links with their hometowns at first, but later came to identify themselves as *Xinjiangren* (Xinjiang folk) and adopted elements of local material culture (ren and Yuan 2003: 97-98). state-sponsored migration to the region has taken place since the times of qing emperor qianlong (mid- to late eighteenth century); however, it has been most vigorously pursued as a state policy by the prC (ren and Yuan 2003: 89). particularly during the 1950s and 1960s, the Chinese state directed the mass transfer of hundreds of thousands of han 'pioneers' to Xinjiang, achieved through propaganda, 'encouragement' centred on state work units, and active recruitment of youths from urban schools. the aim of post-1949 population transfer was (and is) officially stated as the 'economic and cultural development' of areas dominated by 'less advanced' minority nationalities. Yet already by the 1970s, scholars had begun to suggest that the long-term goal of han migration was the sinicisation of frontier areas as a

#### 24 chapter one

means to strengthen China's territorial claims (Pye 1975: 497; Bush 1970: 278). Contemporary scholars tend to echo this view, representing the project as a desire to strengthen central control and ensure social stability in China's sensitive borderlands (Li 1990: 64) or as a 'relentless' effort to outnumber compact native populations along China's peripheries (Bulag 2002: 19). A second pragmatic aim of population transfer has been the reduction of population pressure in eastern China (Ren and Yuan 2003: 90).

Li (1990: 51; 62-64) identified five phases of population development in Xinjiang between 1949 and 1984:1

*Phase 1 (1949-53)*—arrival of People's Liberation Army (PLA) personnel;2 'volunteer' wives for demobbed soldiers; political cadres; college students, engineers, doctors, technical personnel and other skilled workers (cf. Schwarz 1963: 63).

*Phase 2 (1953-61)*—arrival of Xinjiang Production and Construction Corps (XPCC) personnel;3 'volunteers' to carry out land reclamation and industrial construction;4 'floating migrants' (*ziliu yimin*) escaping the famines

<sup>1</sup> I extend my thanks to the anonymous reviewer, who alerted me to the fact that the name of the scholar, Li Yuanqing, then of the Population Research Unit at Xinjiang University, was mistakenly rendered in this publication as 'Yuan Qingli'. I use the correct surname, Li, throughout this text.

<sup>2</sup> Initially, almost 200,000 PLA troops were dispatched to Xinjiang to quell minority resentment and revolt. In 1952, a large section was discharged from the army and reassigned *junken* duties that combined economic tasks (production and construction) with military tasks (safeguarding social stability). These demobbed soldiers later formed the backbone of the Xinjiang Production and Construction Corps (Li 2010: 49), along with former Nationalist troops stranded in Xinjiang (Esposito 1977: 169). Only in 1957 did the *Xinjiang Daily* report that local resistance to Chinese Communist rule had finally been rebuffed (McMillen, cited in Grunfeld 1985: 62).

<sup>3</sup> The XPCC had established 27 state-owned farms in Xinjiang by the end of 1952. These were intended to serve as the 'principal Han communities' for incoming migrants (Li 1990: 66). Some XPCC personnel were demobilised veterans of the Korean war, recruited from East China by the Xinjiang authorities (Schwarz 1963: 64). Also known as the 'militia corps' (*bingtuan*), the contemporary XPCC has been described as 'a Han population engaged in economic activities in times of peace, but which can be mobilised in case of military emergency.' Accounting for approximately one in three registered Han in Xinjiang and one in seven of the total regional population, it may be seen as a 'powerful colonising force' (Becquelin 2000: 77-80; 2004: 366-7; cf. Sines 2002: 12).

<sup>4</sup> Those earmarked for land reclamation included young peasants and high school graduates. To encourage voluntary migration, the government praised the Northwest to an unrealistic degree, and planted letters supposedly written by happy new settlers in the press. Those intended for industrial construction included oil engineers, statisticians, metal workers and construction company managers, and hailed from Beijing, Zhengzhou, Tianjin, Shandong and Shanghai (Schwarz 1963: 63-5).

of the Great Leap Forward;5 and political exiles of the Anti-Rightist campaign (1957-58).6

*Phase 3 (1961-62)*—return of migrants to hometowns following national recovery from famine and in response to the 1962 'Redundancy Movement' in urban Xinjiang.

*Phase 4 (1962-76)*—arrival of 'volunteers' at the start of the Cultural Revolution, some of whom helped to strengthen border defences (cf. Esposito 1977: 167); and 'floating' rural migrants attracted to Xinjiang's cities and state-owned farms.

*Phase 5 (1976-84)*—slowing of population growth following implementation of family planning; return of political exiles to China proper.7

According to Ren and Yuan (2003: 96), spontaneous migration accounted for less than one third of all immigration during the period 1950-1970. This may in part be because the first railway line reached the eastern oasis of Qumul (Ch. Hami) only in late 1959. Since most 'planned immigrants' (*guojia jihua yimin*) in the 1950s were male, this created a heavily imbalanced sex ratio by the end of the decade (1: 197.51, increasing to 1: 296.8 in the 24-29 age group) (Ren and Yuan 2003: 90; 95). To control the subsequent flow of wife-less immigrants back to their home provinces, the government embarked on a project of social engineering. On one hand, it encouraged young women from Tianjin, Shandong, Shanghai and Hunan to move to Xinjiang; on the other, it promised financial assistance to established immigrants who could bring a girlfriend to Xinjiang from the East. Incentives included help with marriage arrangements, housing and employment, and a transfer of *hukou* (household registration). Pressure was also put on earlier settlers to ask their dependents to join them in the northwest (Schwarz 1963: 71). After the railway line reached Ürümchi in 1962, some established immigrants began to bring families and relatives to Xinjiang voluntarily. Although the 1964 census continued to reflect a high concentration of males aged 20-30 years (Li 1990: 60; 65), after 1970 the sex ratio stabilised at 1: 101 (Ren and Yuan 2003: 96). Despite the improved marriage opportu-

<sup>5</sup> Per capita food production in Xinjiang was 288.5 and 245.7 kilograms respectively for the years 1960 and 1961, compared with national per capita outputs of 215 and 223 kilograms for the same years (Li 1990: 67).

<sup>6</sup> Many of these exiles came from Shanghai, a suspected centre for 'rightist' intellectuals, and may have been destined for 'remoulding' through physical labour (Schwarz 1963: 65).

<sup>7</sup> Around 10,000 'professional intellectuals' left Xinjiang to return to their home provinces following the end of the Cultural Revolution (McMillen, cited in Li 1990: 64).

#### 26 chapter one

nities, however, writer Vikram Seth reported extreme reluctance among Xinjiang migrants during his travels in 1981. Complaining of their banishment to 'New Zealand' (Ch. *Xinxilan*), that is, *Xin*jiang, *Xi*zang (Tibet) and *Lan*zhou, they remarked: 'We could just as well be on the moon' (1993: 39).

Conversely, in the 1990s many migrants moved to Xinjiang of their own volition, keen to raise their standard of living. Popular destinations included the Xinjiang Production and Construction Corps, towns bordering the second Europe-Asia railway (Shihezi, Kuytun, Bortala), and national stations of petroleum exploitation and cotton production.8 The net migration rate increased from 0.4 to 2.43 per cent across the nineties, constituting the fourth highest rate after Beijing, Shanghai and Guangdong (Ren and Yuan 2003: 97). There was an especially pronounced increase in the number of Han migrants to southern Qäshqär, facilitated by the completion in 1995 of the Täklimakan Highway, which runs south from northern Qorla. There the proportion of Han rose from 7.7 per cent in 1982 to 9 per cent in 1997 (Ma 2003: 112-3). Against this background of spontaneous migration, Han scholars called in 1996 for new resettlement plans that would attract a young rural workforce out of impoverished central provinces and into Xinjiang's rural desert regions. Those who took advantage of the plans would benefit from the conferral of property rights that permitted the division, transfer, rental and inheritance of land (Wang and Chen, cited in Becquelin 2000: 75-76).

It is difficult to pinpoint the extent to which the overall population composition has been altered, since post-1980 rural reforms allow surplus rural labour to migrate without a change of household registration or *hukou* (hence the term 'floating population', Ch. *liudong renkou*). According to the Xinjiang Statistical Bureau, Xinjiang's rate of population increase over the period 1953-1982 ranked second only to Heilongjiang (Li 1990: 50). The average annual rate of increase between 1949 and 1993 was 3 per cent, with immigrants comprising 23.13 per cent of the total (Ren and Yuan 2003: 93- 95). As a result, the proportion of Han Chinese in Xinjiang increased from 5.5 per cent in 1949 to 20.5 per cent in 1962 and 45 per cent in 1966 (Tal, cited in Pye 1975: 497). By 1970, it stood at 40 per cent (Dillon 1995: 31). While the 1990 census registered a dip to 37.6 per cent during the eighties, mainly owing to lower Han birth rates (Becquelin 2000: 69), this share had again increased to 40.6 per cent by the year 2000, representing a net gain of 3 per

<sup>8</sup> Wang (2004) notes that no statistics exist regarding the number of oil field workers contracted in Xinjiang.

cent over 1990 (Becquelin 2004a: 369). Across the decade, the Han population grew by 31.6 per cent (almost one third), while the minority population grew by just 15.9 per cent (Becquelin 2004a: 369; cf. Wang 2004).

In terms of the relative populations of individual groups, in 1990 there were 5,695,626 Han Chinese resident in Xinjiang compared with 7,194,675 Uyghurs (Hoppe 1992). In 1996, an official Chinese survey estimated the total of Han at 6,424,400 persons compared with 10,468,500 persons of minority nationality, including Uyghurs and others;9 and the 2000 census confirmed a Han population of 7.49 million. As noted by Becquelin, this figure was significantly higher than the figure of 6.87 million published for 2000 in the Xinjiang Statistical Yearbook 2001 (2004a: 369). Geographer Stanley Toops and Tibet historian Tsering Shakya both suggest that statistical data remain to some extent a propaganda instrument in Xinjiang and Tibet, where the political sensitivities of population issues impact directly on state security. Shakya, for instance, noted in 1999 that China had repeated identical population statistics for Tibet for the previous several years.10 Toops points out that while recent Chinese censuses accord with international standards, published population figures are not accurate since they do not count military personnel or 'unofficial' migrants in Xinjiang. At the same time, he asserts that most Han who moved there after 1964 did so illegally (unofficially): 'The government either looked the other way or channelled and facilitated a movement of people […] whose presence in Xinjiang […] promoted the policies of the Chinese Communist Party and the state' (2004: 241-2, 257-8).

Turning to my interview data gathered in Ürümchi in 1995-6, the perception among Uyghurs of a rapid increase in Han migrants was reflected on a daily basis, as in this comment: 'We used to say that one Han brought ten more with him. Now we say that one Han brings another hundred.' Respondents like Shökhrät, a young male intellectual, complained in 1995 of large numbers of Han people arriving at Ürümchi's rail station daily, and argued that the huge sacks they carried indicated their intention to stay.11 Some

<sup>9</sup> Xinjiang statistical communiqué for 1996, *Xinjiang ribao*, Ürümchi, 14 March 1997 in SWB (Asia Pacific), 7 May 1997, FEW/0485 WS2/8.

<sup>10</sup> Tsering Shakya. Talk on his book *The Dragon in the Lands of Snow*, Waterstones bookshop of Leeds, 3rd February 1999.

<sup>11</sup> Guomindang plans in 1946 to build the first railway line into Xinjiang led to it being dubbed the 'Trojan horse of Han colonisation' by the Ili group, and to the collapse of the short-lived coalition government (Dreyer 1976: 25). Similar feelings regarding the adverse effects of enhanced infrastructure have been documented for Tibet. Blum notes that while official Chinese articles often cite the material advantages bestowed on the Tibetan people

#### 28 chapter one

Uyghurs in southern Xinjiang reported in interview in 1996 a sharp increase in Han migrants to urban centres like Khotän during the first half of the nineties. While most agreed that Han were concentrated in urban areas, popular stories demonstrated an awareness that they had also begun to appear in rural areas, as in the tale about the horrified Uyghur who found a Han Chinese shoe-mender setting up business in a remote village.

In accordance with state discourses during the nineties which denied the use of incentives to encourage Han migration and blamed demographic change on the *'*floating population' (*liudong renkou*), many *minkaohan*  (Uyghurs educated in Chinese) also used the term when bemoaning increased Han in-migration. Adalät, a health professional speaking in Ürümchi in 1995, blamed 'floating' migrants for regional pollution, arguing that the air and water had become dirty and that vegetables and fruits had lost their flavour. The notion also surfaced in puns invented by Uyghur *minkaohan*, such as: *Hanliu lai le, manliu zou le!* [When the cold weather comes, the overflow leaves!]12 One young male carpet trader joked in 1996 that Han migrants should be dubbed not the *liudong renkou* [lit. flowing and moving population] but rather the *lailiu renkou* [the population that comes and stays].13 A small number of *minkaohan* respondents interviewed in the mid-nineties repeated the government line that immigration was 'necessary' to help Xinjiang achieve rapid economic development; however, it was not always clear whether this constituted a genuine personal belief or a politically safe response.

Since the turn of the millennium, the state's goal of encouraging Han migration has been explicitly articulated, most notably within the 'Great Western Development' (*Xibu da kaifa*) strategy approved in July-August 2000. According to Becquelin, the strategy consolidates what he calls 'domestication' policies pioneered in the 1990s. Han settlers are now conceived as part of a broad plan known as *ningjuhua* (process of improving cohesion), euphemistically labelled 'the peacock flying West', and colloquially

by the modernizing effort of the Han Chinese, improvements in communications may be read locally as an attempt to foist Han people and Han culture on the Tibetan homeland: 'Roads are not welcome when they bring conquerors with ever greater speed and efficiency' (2001: 129). In Inner Mongolia, the city of Chining on the newly built Beijing—Ulan Bator railway line doubled its population between 1951 and 1956 (Schwarz 1963: 70).

<sup>12</sup> This plays on two rhyming words, the second of which is invented: *hanliu* 寒流 and *manliu* 满流.

<sup>13</sup> *Lailiu* is another invented word constructed from the characters 来 (come) and 留 (stay); in this case, further comic effect is achieved from the fact that the verbs for 'flow' and 'stay' are homonyms in Chinese, each being pronounced *liu*.

referred to as *chan shazi* (mixing sand) (Becquelin 2004a: 373-74; 2000: 74). The strategy has been surrounded by academic controversy both within China and abroad over its feasibility and implementation (Li 2010: 59). While government discourses speak of the urgent need to raise the standard of living in China's West and reduce inequalities *between regions*, officials rarely refer to the need to address inequalities *between ethnic groups*. Recent immigrants have been attracted by the extension of the railway line to Qäshqär, completed in 1999, and in some cases were actively recruited via job fairs and employment hotlines aiming to 'bring talent west' (Sines 2002: 14). In this way, covert processes of territorial and ethnic assimilation have been replaced by overt ones: '… as the long sought-after objective of seamless integration of minority nationalities into the "Chinese nation" draws near, history has entered an end-game, and there is no necessity to proceed with caution and coded double-speak any more' (Becquelin 2004a: 374). This development is reflected too in the purposeful alteration of official nomenclature when referring to China's minority groups. The first indication came in 1995, when the English translation of the title of Chinese journal *Minzu tuanjie* was changed from 'Nationality Unity' to 'Ethnic Unity', altering the representation of the minority nationality from one of formerly self-governing entity to incorporated minority community (Bulag 2002: 21).14 This practice has been mainstreamed into the 2000s, so that official Chinese documents now routinely translate the word *minzu* as 'ethnic'; thus the PRC is a 'multi-ethnic' rather than a 'multi-national' country and the Uyghurs a firmly incorporated 'ethnic minority' rather than a potentially independent 'minority nationality' (Becquelin 2004a: 359).

#### 1.2. *Language Policy*

The most noticeable impact of increased Han in-migration, and one which affects all other spheres in a chain reaction, has been the policy of gradual institutionalisation of standard Chinese (*Putonghua*). This has occurred at the expense of Uyghur, formerly the regional *lingua franca* (Dwyer 2005: 12-13). As Han communities in Xinjiang were consolidated by consecutive waves of migrants, and the total of Han residents drew near to the total of Uyghurs, confidence in numbers gave rise to ever more sino-centric policies

<sup>14</sup> In his book tracing the development of China's Mongols, Bulag (2002) borrows Kymlicka's (1995) contrastive definitions of 'national minority' (a distinct society seeking autonomy) and 'ethnic minority' (immigrated into another state and desiring integration).

Figure 3: Ever decreasing Uyghur fonts, Hongshan, central Ürümchi (2004)

concerning language use and language domain, as well as to increasingly visible displays of Han chauvinism (*Da hanzu zhuyi*—the assumed superiority of Han language and culture).

PRC policy towards minority languages and scripts has fluctuated between the progressive and the assimilatory. Early 1950s policy was tolerant in that although the 'major minority languages' of autonomous regions were required to share space and resources with Chinese in the domains of state administration, law, education and media, their use was promoted and legally guaranteed (Dwyer 2005: 7). By the second half of that decade, however, CCP proximity to Soviet politics had resulted in the replacement of the Arabic-based Uyghur script with a modified Cyrillic script, usage of which peaked between 1955 and 1958 (Dwyer 2005: 19). This was swiftly followed in 1958 by another switch, this time to a Pinyin-based Latin alphabet, reflecting state fear of the development of cultural alliances along Soviet Central Asian borders in the aftermath of the Sino-Soviet split (Bellér-Hann 1991: 73-74; Wei Cuiyi 1993). From the early 1960s, local media organs tried to negotiate these so-called 'right' (1949-50) and 'left' (1956-62) policy deviations by once more broadcasting programmes in local languages, but adjusting their content to stress a common culture that leaned towards Han culture (Grunfeld 1985: 64). Yet another script innovation was enacted in 1971, this time in the form of a Latinised Uyghur script based on the phonetic equivalents of Chinese characters and introduced by the PLA (Esposito 1977: 168). The Latin script came into wider use in 1974 and was formally adopted in 1976. By 1978, it was employed by the mass media, by 70 per cent of intellectuals and by 50 per cent of laymen (Jarring, cited in Bellér-Hann 1991: 73-74, and in Dwyer 2005: 20).

The 1980s once again ushered in a period of tolerance, with the re-introduction of the Arabic-based Uyghur script and expanded support for minority spoken and written languages in the form of literature, news reporting, publishing, radio and television broadcasting, and film (Dwyer 2005: 11). The return of a modified Arabic script was welcomed by most (if not all) Uyghurs, who in the context of a growing non-Muslim population saw its protection—and the defence of the Islamic religion—as an important task (Bellér-Hann 1991: 76; 79-80). Two important Uyghur literary anthologies emerged during this period: the *Uyghur Klassik Ädäbiyatidin Namunilar* [*Specimens from Classical Uyghur Literature*, 1981] and the *Qädimi Uyghur Yazma Yadgarliqlardin Tallanma* [*A Collection of Ancient Uyghur Written Remains*, 1984] (Naby 1991: 20-21). The first of these introduced works dating from the Islamic period, presented in the contemporary vernacular, and gained broad appeal beyond literate circles because, as Eden Naby put it, 'the images and metaphors of the Islamic past form the main base of present culture' (1991: 21). The second brought together descriptions and translations of inscriptions and texts produced in the pre-Islamic Uyghur alphabet. Together, they constituted the earliest sources of information on Uyghur cultural heritage available to the literate layperson and, according to Naby, embodied the Uyghur need to 'justify their cultural autonomy and the value of the continuation of their language and literary traditions by pointing to a rich past' (1991: 22).

As observed by several scholars, the script changes of the first few decades of the PRC were almost certainly ideologically motivated: the Cyrillic-based script by Soviet communism and friendly Sino-Soviet relations;

#### 32 chapter one

the Latin-based one by Chinese communism and the Sino-Soviet split (Bellér-Hann 1991: 74; Benson 2004: 194). In addition to distancing Uyghurs educated between 1950 and 1976 from the Arabic-based script and the Muslim identity, these discontinuities made it difficult for that middle generation to engage in written communication with the older and younger generations educated in the Arabic-based script (Bellér-Hann 1991: 75, 80; Dwyer 2005: 21). This situation is vividly illustrated by Seth, who found trilingual signs—in Chinese, Latin and Arabic—in Turpan in 1981, and observed that while a Uyghur *aq saqal* ('white beard'—elder) could write to his grandson in Arabic, his son could communicate in writing with neither (1993: 1; 18). The old man meantime suggested to Seth that the script had been deliberately changed to Latin to prevent Xinjiang Uyghurs from reading Arabic script publications filtering in from Soviet Central Asia, and thus to prevent pan-Turkic identifications. This notion was accompanied by widespread suspicions that the frequent script changes represented the state's attempt to divide the generations (Benson 2004: 197).

Even in the conciliatory 1980s, reality did not always reflect policy. Though regional autonomy allowed individuals to use minority language textbooks in theory, in practice, Uyghur and Kazakh scholastic publishing was often limited to the humanities and social sciences, reflecting the Han assumption that minority languages are unsuited to modern science. Many books were simply translated from the Chinese rather than created in the local language (Dwyer 2005: 36). Uyghur-language editions of regional newspapers often substituted Chinese terms for Uyghur lexemes or were literally translated from the Chinese edition. Uyghur actors in dramas or films intended for minority consumption were required to say their lines in Chinese and have them dubbed in Uyghur, in order to ensure political control of script content. Needless to say, all of these practices accelerated the sinicisation of the minority-language lexicon (Dwyer 2005: 48-9). Grunfeld, visiting the region in the eighties, reported that language remained a 'sore point' since far more minority people learned Chinese than Han bothered to learn the local tongues (1985: 66).

From 1990 onwards, China's language policy became increasingly 'reactive' and tied to geopolitical considerations (Dwyer 2005: ix). In Xinjiang, the threat of transnational (Central Asian) solidarities prompted the state to attempt after 1996 to obscure the Turkological roots of the Uyghur, Kazakh etc. languages. Thus, higher education institutions now have *Zhongyuxi* (Departments of China languages) and *Zhongyu wenxue xi* (Departments of China languages and literatures) rather than Departments of Turkic-Altaic languages and literatures, as in the past (Dwyer 2005: 31-2).15 Even classes in Uyghur literature were now taught entirely in Chinese, and were subject to spot checks by *til saqchi* (language police) (Dwyer 2005: 51). Yet although media (broadcasting) and educational (textbook) material have been greatly affected by the shift to a monolingual language policy since the mid-nineties, a combination of literary works, scholarly journals and non-scholastic print publications continues to successfully maintain minority languages in Xinjiang (Dwyer 2005: 47-8).

When I conducted fieldwork in Xinjiang in 1995-6, educated Uyghurs in Ürümchi regularly complained that Han people did not respect their native language. I sometimes observed this phenomenon personally. On one occasion in 1995, as I chatted to Ghäyrät, a migrant worker from Aqsu, about the Uyghur language, the young Han manager of the restaurant outside which we sat exclaimed: 'They [the Uyghurs] don't *have* a language! They just speak one of our China's regional dialects [*women Zhongguo de yige fangyan*].' A remark frequently on the lips of Uyghur respondents during this period was: 'Do you know, some Hans have lived here for thirty years and never even thought of learning our language?'16 Other Han residents scoffed that Uyghur was a 'simple' language and 'easier to learn than Chinese', although few if any of them appeared to have studied it. Exceptions included some Han scholars affiliated to the Xinjiang Academy of Social Sciences (whose remit it was to study the cultures and societies of minority nationalities) and some long-term Han settlers residing in majority Uyghur communities in the south. The latter seemed to enjoy greatly improved relations with Uyghur neighbours as a result of their willingness to learn the local language and adapt to local culture. This phenomenon is also reflected in Seth's account. He speaks warmly of encountering a Han policeman who had lived in Turpan for twenty years and spoke fluent Uyghur (1987: 12). More recently, Chinese scholar Ma Rong documents that in one horticulture garden in Qaghiliq (Ch. Yecheng) in the Qäshqär region, managers pursued a policy of deliberately mixing Han and Uyghur workers in their everyday routines. As a result, 20 per cent of Han workers (usually

<sup>15</sup> One Uyghur respondent pointed out in 2002 that while every effort has been made to divide the common Turkic identity, the differences among the Han group have been deliberately obscured. On the same note, Pye pointed out that sub-groups among the Han—Central Han, Hokkien, Tiechiu, Hailam, Khed and Hakka—have 'all the social attributes usually associated with ethnic differences' (1975: 491).

<sup>16</sup> This echoes attitudes in the 1800s, when Chinese rulers in Xinjiang made no effort to learn the local language, literature or customs (Warikoo 1985: 99).

#### 34 chapter one

longer-term residents) spoke Uyghur while 30 per cent of Uyghur workers spoke Chinese, and the two maintained a closer relationship than was usual (2003: 119).

Uyghur academics maintained in the 1990s that Han residents were intolerably proud of the Chinese language. A young male scholar named Shökhrät spoke scornfully in 1995 of a Han colleague who had described Chinese as 'the most excellent language in the world [Ch. *shijie zui youxiu de yuyan*].' According to my observations, such attitudes frequently aggravated Uyghur-Han relations. In 1996, I witnessed the following debate between Shökhrät and a Han taxi driver in Ürümchi, sparked by my announcement that I was studying Uyghur:

Han driver: What's the point in learning Uyghur? We don't even use it!


However, not all respondents interviewed in the mid-nineties shared this level of awareness (or cynicism); some passively accepted the Han assumption that Chinese language and culture were superior to their Uyghur counterparts. For instance, Mobaräk, an economics student from Kucha, remarked in 1996: 'The Hans must be more advanced than the Uyghurs! All information about science and technology comes to Xinjiang in the form of Chinese language texts.' It did not occur to her that new developments in science and technology imported from Western societies might have been translated into Uyghur as easily as they had been translated into Chinese, given the political will and efficient language planning (cf. Dwyer 2005: 28-29).

Language discrimination also affected the publishing industry. Sinologist and political scientist Lucian Pye observed in 1972 that only 10 per cent of the 600,000 volumes housed at the Central Institute for Nationalities (*Zhongyang minzu xueyuan*) in Beijing were published in minority languages, while 90 per cent were published in Chinese (1975: 507). During my first field trip in 1995-6, educated, urban Uyghurs frequently complained

of limitations imposed on the print-run of Uyghur language publications. As Räwiä, a foreign languages graduate, remarked: 'When a new Uyghur book is going to be published, there is always a Han who says "What do you want to print that many copies for?"' On the other hand, according to Sadiq, a young male employee at an Ürümchi publishing house, in 1995 Han writers were at liberty to publish any amount of material on Uyghur customs in Chinese. It was also evident that books of a particular type were not being published at all. My own search for materials on Uyghur cultural traditions in Ürümchi's bookshops—in either Uyghur or Chinese—proved futile throughout 1995-6, although materials published years earlier could be found in Uyghur homes. I asked Uyghur publishers why this was so. Some were reluctant to offer an explanation, afraid of the consequences of doing so; others asserted that the *Xuanchuanbu* (Propaganda Department) controlled all publications closely, and that manuscripts produced by Uyghur authors were checked especially carefully. According to one respondent interviewed in 1995, this had produced a thriving black market in previously published materials on Uyghur history and culture: 'It's all the rage at the moment! You can only find good material in people's private collections … but you won't find it in the bookshops.' Another joked that nobody dared to publish at present, eliciting this humorous retort from a colleague: 'Is it that no-one dares? Or that no-one is allowed? (*Bu gan haishi bu rang?*)'17 Shökhrät, a young male intellectual, added that although some Uyghur writers were prepared to take risks, publishers often refused to publish their work out of fear of political retribution. Those books that were published in the Uyghur language tended to present information through the lens of the CCP's political interests. As Aliyä, a postgraduate student based in Ürümchi, explained in 1996: 'I don't trust the textbooks here because they are all vetted by the Public Security Bureau (*Gong'anting*).'18 The lack of availability of Uyghur language materials made it difficult for young Uyghurs growing up in a Han-dominated urban environment to learn about their cultural heritage and arguably may have been a deliberate effort to obstruct the development of ethnic pride.

While the institutionalisation of the Chinese language seemed to occur in a more or less *de facto* process across the 1990s as more and more minority parents recognised a pragmatic need for their children to master the

<sup>17</sup> This is a play on a partial rhyme in Chinese: *gan* (dare) and *rang* (allow).

<sup>18</sup> On the subject of politically motivated interpretations of Xinjiang's history, see Bovingdon (2001).

#### 36 chapter one

Chinese language, the blanket adoption of Chinese-medium education the so-called 'bilingual education'—since 2002 signalled that the process of 'linguistic convergence' was now openly endorsed by the state.19 As Schluessel notes, this micromanagement of language policy has made Chinese in schools 'not a choice, but an imposition, one […] which unintentionally triggers a reaction of resistance and resentment', and has led locally to charges of 'linguistic genocide' (2007: 259, 270; cf. Uyghur Human Rights Project 2007). Meanwhile, although Article 49 of the Law on Regional Autonomy (2001) states that Han cadres 'should learn the spoken and written languages of the local minority nationalities' (Dwyer 2005: 33), the reality is very different. In a survey conducted in Ürümchi in 2000, Herbert Yee found that while 91.5 per cent of Uyghur cadres spoke excellent, good or fair Chinese, 80.5 per cent of Han cadres spoke poor Uyghur or no Uyghur at all. Not surprisingly, the author concludes that 'the Han cadres are clearly handicapped by their language incompetence' (2003: 446-7). Among laypersons, Yee's study found that 79.2 per cent of Uyghurs spoke excellent, good or fair Chinese compared with only 17.3 per cent of Han Chinese who possessed the equivalent skills in Uyghur (2003: 436). When asked whether the Uyghur language had been successfully promoted in Xinjiang, 60.8 per cent of Uyghur respondents stated either that policy implementation had not solved regional language problems (for example, the relative status of languages) or that the policy had been poorly implemented, or not implemented at all (Yee 2005: 47).

### 1.3. *Education*

The issue of relative value of minority languages and Chinese (*Putonghua*) has long been central in the evolution of education policy in Xinjiang. Perhaps the first Chinese official to perceive a need to introduce 'modern Chinese education' to the region was General Zuo Zongtang (1812-1885), who stated: 'When the young Moslems become more and more interested in studying Confucian teaching […], new foundations for *bringing the* 

<sup>19</sup> The term 'bilingual education' has been aptly defined by Schluessel as 'a euphemism for the mandatory increase in the use of Mandarin in minority-language-speaking children's school environments'. The concept of 'bilingual education' dates back to 1987, when a working group formed to discuss its possible implementation as a means to help students 'transition' from their native language to the dominant one (Dwyer 2005: 34-5; Schluessel 2007: 256). The goal of 'linguistic convergence' echoes Stalinist policy, which viewed language as an essential political tool to ensure national unity (Schluessel 2007: 251; 254).

*Moslems to terms* will have been established' (cited in Warikoo 1985: 99, my emphasis). But his proposal to introduce free public schools which would teach Chinese language, literature and philosophy was not adopted by his successors after 1884. We know that during the 1880s, primary schools were attached to mosques and attended by both sexes, while pupils learned to read and recite verses from the Qur'an. However, the proportion of children attending school was low overall, and girls discontinued study after the age of 12 'for fear of getting molested in the towns' (Warikoo 1985: 98-99). By the 1930s, the most popular schools were those established by ethnic and religious educational associations featuring native language instruction; such schools outnumbered those of Chinese warlord Sheng Shicai's government by five to one (Bellér-Hann 2008: 336-337; Han, cited in Schluessel 2007: 253). Later, under the East Turkestan Republic (1944-49), schools were set up for separate ethno-national groups, so that Uyghurs, Kazakhs, Kyrgyz, Uzbeks, Tajiks, Tatars and Taranchis were each taught in their mother tongue.20 At the same time, however, pupils studied Chinese for 6-8 hours per week in addition to a 'foreign language' (Han, cited in Schluessel 2007: 254).

After the inception of the PRC in 1949, education policy towards minority nationalities oscillated alongside the national politics. While minority students wishing to enter institutes of higher education had the national foreign-language requirement waived, their numbers remained severely limited by the fact that most teaching was delivered in Chinese rather than Uyghur; a government report published in 1973 stated that a total of just 3700 students from a minority background had successfully graduated from Ürümchi's Xinjiang University (Benson 2004: 196). For over a decade during the Cultural Revolution, minority languages were not taught at all, while the leading institution of higher education for minorities, the Central Institute for Nationalities, remained closed for 6 years between 1966 and 1972 (Grunfeld 1985: 65). The most 'progressive' policy periods fell in the early years of the PRC (1950-55—see previous section on Language) and between 1980 and 1995. From the beginning of the 1980s, the state pursued a 'separate but equal' education system involving parallel Chinese-medium and minority-language schools (where 'minority-language' usually meant Uyghur) (Schluessel 2007: 255-6). This liberalisation created a significant space for

<sup>20</sup> Taranchi (which means 'farmer' in Chagatay) is the name given to a Turkic agriculturalist people of mixed Iranian origin, who were transferred from the Tarim basin to live in the Ili valley by Qing military governors following their conquest of the Zunghar Khanate.

#### 38 chapter one

minority linguistic and cultural expression; however, there were also problems. Corresponding to the increase in Han residents, funding for minority-language schools fell steadily across the nineties while funding for 'regular education' in Han schools grew (Becquelin 2000: 85). Dwyer observed that the minority-language classroom in the Xinjiang University pre-school received only half the volume of materials and toys compared with its Chinese-medium counterpart. At the same time, the state gradually expanded tuition of Chinese as a second language. In 1980, relatively prosperous urban minority students began to study Chinese between fourth grade and the end of high school, and from 1984 tuition in Chinese as a second language began to be introduced from the earlier stage of third grade primary school (Dwyer 2005: 36; cf. Schluessel 2007: 256).

Urban parents confronted with the choice between sending their children to a Chinese-medium school or classroom (*Hanzu xuexiao; Hanzuban*) or a minority-language school or classroom (*minzu xuexiao; minzuban*) often felt forced to choose the former in order to maximise their children's life chances in a Han-dominated urban society (Dwyer 2005: 36-8). Nathan Light estimates that up to 50 per cent of Uyghur families in Ürümchi may have chosen to school their children in Chinese over this period, although the proportion will have been far lower in cities with a large Uyghur population (cited in Dwyer 2005: 38). The logic of this parental decision to sacrifice a child's training in the mother tongue and culture in favour of socio-economic mobility is brought into sharp relief by a demographic study on the socio-economic composition of China's 15 largest minority groups (Poston and Jing 1987). This study found that groups with educational distributions (percentages of people educated to different levels) similar to the Han also tended to be similar in terms of occupation and industry. Furthermore, it concluded that those groups similar to the Han in educational distributions were similar to them on *all* characteristics of composition, while those who differed from the Han differed from them on all characteristics. In other words, the more assimilated a minority is in terms of degree of population distribution, degree of intermarriage with the Han, etc., the greater the proportion of its population in college and professional occupations, and the lower its illiteracy rate and percentage engaged in agriculture.21 Poston and Jing's study showed that the state's

<sup>21</sup> In terms of education distributions, the Mongolians and the Manchu were the least differentiated from the Han in 1982, reflecting their wider geographic distribution and high levels of intermarriage with the Han; while the Tibetans and the Hani were the most dif-

goal of socio-economic advancement for the minority nationalities had been reached among only a few groups in the 1980s, and that the Uyghurs, as a more geographically differentiated and less integrated group, were not as well off in socio-economic terms as the Manchus, the Mongolians, the Hui or the Bai (1987: 718-9).

From the mid-eighties, preferential policies were introduced into the university admission process in an effort to close the 'ethnic gap' in living standards. Minority applicants were now given extra points, based on their home region, ethnicity, course of study and medium of tuition. Since *minkaohan* (Chinese-educated) applicants received more points than *minkaomin* (minorities educated in their mother tongue) (Sautman 1998: 92), one ramification of the policy is that it further progressed the institutionalisation of the Chinese language. Meanwhile, though the aim has been 'real' or 'actual' equality (*shishishang de pingdeng*), related discourses which refer to the 'special needs' of minorities—are mired in what social scientist Barry Sautman calls the 'social evolutionism' of elite Han ideology (1998: 88), and tend to reinforce majority Han stereotypes of 'stupid' or 'backward' minorities (Dwyer 2005: 11). In relation to this point, some minority intellectuals fear that affirmative action policies will indeed render Uyghurs less competitive in the long run, by lowering standards and expectations (Yee 2005: 48). Others regard them as 'half-measures' since they ultimately fail to address the core problem of erosion of Uyghur language and culture, symbolised by the dominant position of Chinese in the university sphere (Sautman 1998: 101).

At the same time, while the policy has increased minority access to universities, progress is threatened by the trend towards marketisation of education, which has in turn impacted on levels of enrolment of minority students. Sautman's study found that because Han are mainly resident in prosperous urban areas and enjoy a higher income, they can better afford the 2,000 yuan annual fee, whereas 77 per cent of Xinjiang's minority tertiary students are from peasant families. Although many students receive state subsidies and need only pay 900 yuan per year, even this fee is beyond the reach of some (Sautman 1998: 92). The additional year that minority individuals must spend studying Chinese in order to reach a level in that

ferentiated, and the Uyghurs fell somewhere in the middle. In terms of occupation and industry distributions, the Mongolians and the Hui were the least differentiated from the Han, reflecting their geographic distribution and, in the case of the Hui, linguistic proximity to the Han, while the Uyghurs ranked 4th and 5th respectively (Poston and Jing 1987).

#### 40 chapter one

language adequate for university-level study is also viewed by some as imposing an unjust financial burden (Benson 2004: 208; Taynen 2006: 47). As a result, while the total number of minority students in PRC universities and colleges rose from 9.7 million in 1980 to more than 16 million in 1995, the overall percentage of minority students, on the increase until 1992, fell thereafter to below the minority share of the national population. In Xinjiang, minority enrolment in regional institutes of higher education varied from 56 to 60 per cent in the late 1980s and early 1990s, but decreased to 52 per cent in 1994. In the same year, 50 per cent of students at Xinjiang University in Ürümchi were from minority nationalities compared with 60 per cent in 1986, while in Turpan only 48 percent of college students were Uyghurs compared with their 78 per cent share of the local population. Meanwhile, most students studying at the Xinjiang Medical College (*Yixueyuan*) and Xinjiang Industrial College in the mid-nineties were Han (Sautman 1998: 90-92; cf. Benson 2004: 208). The percentage of minority nationality lecturers also lagged far behind that of minority students, despite government attempts to increase their numbers, and most taught at the less prestigious minority area colleges or institutes (Sautman 1998: 93).

The interviews I conducted in 1995-6 amply support the above findings. First, with regard to the dominance of the Chinese language in higher education, student respondents explained how those *minkaohan* who attended a Chinese-medium school from an early age (and whose first language was consequently Chinese) could enter university far more easily than *minkaomin*. For a *minkaomin* student to get into university was considered an enormous achievement. Aliyä, a *minkaomin* postgraduate in Ürümchi, explained in 1996: 'Uyghurs always have to take one step further than Hans in everything. You see, we have to master Chinese [as a second language] before we can make any real progress.' Others interviewed in the same year confirmed that the linguistic difficulties attendant to university entrance resulted in fewer minority students in Xinjiang's universities than Han, despite the state's preferential policies.

Second, urban parents in Ürümchi confirmed that after careful consideration of the social and economic climate, Chinese-medium schools were effectively the *only* choice if they hoped to increase the life chances of their children. These were forward-thinking people, who recognised the necessity for a strong command of Chinese in order to succeed in a highly discriminatory labour market, and who saw that the youth needed the education in science and technology offered at Han schools (and taught less well using inferior materials at under-funded minority-language

schools) if they were to keep pace with economic development. In 1995-6, I observed many bilingual Uyghur children playing beside Ürümchi's *loufang* (residential blocks): they spoke Chinese with a perfect accent, and announced brightly that they attended Han schools. This trend was less observable in the Uyghur-dominated southern oases, where I spent three months in the summer of 1996, since parents remained unconvinced of the merits of a Chinese-medium education. At that time, there were proportionally fewer Han (and less need to speak Chinese) in the south, with the Täkilimakan Highway only recently finished (1995) and the Qorla-Qäshqär railway still under construction. Most development had taken place north of the Tianshan (Heavenly Mountains), so that most southerners had not experienced the employment competition that characterised urban Ürümchi. In the absence of compelling socio-economic reasons to do otherwise, southern parents naturally prioritised the study of the native language and culture. This inclination was summed up by a middle-aged male teacher in Khotän, who stressed in 1996: 'Of course, Uyghur schools are best. After all, we are Uyghurs.' There had also been a move in the south (where religious observance was historically stronger) to re-strengthen Islamic education in reaction to urban secularising trends. In 1994-5, some parents in Khotän and Qäshqär had removed their children from Chinesemedium schools and placed them in *mädris* (Islamic schools teaching Qur'anic studies) or in private religious training with *mollas* (trained mullahs, who have studied Islamic traditions and Islamic law).22 In rural areas Chinese-medium schools were rare at this time, and I only encountered one case of non-Han parents seeking a Chinese education for their offspring: this was a mixed Uyghur-Kyrgyz couple I interviewed in 1996, who planned to school their children in Ürümchi. Most rural people were farmers or engaged in traditional professions, and received a Qur'anic education, if they received one at all. During an extended stay in rural Aqsu, I noted that few of the adult farmers I encountered were literate, while

<sup>22</sup> This counter-movement was prematurely halted in the wake of a mass protest on 7th July 1995. The protest arose when the imam at Khotän's largest mosque was arrested for directing a gathering of five thousand people at Friday prayer not to transfer their children into *mädris*. He advised them: 'Send your children first to [secular] state schools and encourage them to study well modern science. There will be time enough to master their religion later on.' Following the imam's arrest, protestors marched to government buildings, where they were met by police who deployed tear gas. Two died in the ensuing violence. This event is mentioned obliquely in a special report in *Lien ho pao*, Hong Kong, 11 October 1995 in SWB (Asia Pacific), 28 November 1995, FE/2472 G/12. In an ironic response, the Chinese government banned private Qur'anic tuition and began to close *mädris*.

school attendance was erratic owing to the requirement for children to help out in the fields at harvest times.

Third, with regard to the adverse impacts of marketisation, many Uyghur families said in 1995-6 that they faced difficulties financing their children's studies, as school, university and college fees became increasingly expensive. This was especially true for rural families, whose annual incomes remained at roughly half those of urban residents in the 1990s (cf. Benson 2004: 197-198). Some respondents also complained about the bribes required to 'buy' university entrance if their children failed the entrance examination. In a situation where the standard of living among Uyghurs was lower than among Han, and unemployment levels higher, some respondents described 'sacrificing everything' in order to raise the money for their children's education. One peasant in Aqsu interviewed in 1996 described how she worked from dawn until dusk, sowing and harvesting crops by hand, in order to pay for her teenage daughter to study in Ürümchi. Disillusioned urban parents in Ürümchi, Kucha and Qäshqär stated simply that they were unable to shoulder the expense, lamenting that their children had worked hard and passed university entrance examinations, only to find they lacked the funds needed to pay for tuition. According to a young woman from Kucha speaking in 1996, 'poor children with intelligent minds' were frequently denied opportunities in this way. The situation even led some older respondents to reminisce about the progressive education systems operating in Xinjiang during the Republican era.23 Some educated, urban dwellers suggested that the combination of the Chinese language requirement and high tuition fees was designed not merely to favour Han students but also to 'keep Uyghurs stupid.' When a Khotänese imam was arrested in 1995 for advocating a calculated combination of state education in science and technology plus private training in religion (see footnote 21), Uyghur intellectuals interpreted this as a reflection of government fears of increased knowledge and increased empowerment. As Shökhrät, a young male scholar in Ürümchi, remarked in 1996: 'They [the authorities] want to keep us in the dark. They would rather we know nothing. That way we pose no problems to the state.' The implication is perhaps not so farfetched, given that there is a pre-history of this type of strategy. Warikoo

<sup>23</sup> Bellér-Hann (2008: 329) describes a differentiated fee-paying system in Yäkän [Yarkand] in 1930 which allowed children from poor families to pay only half the fees paid by those from rich families. This system of subsidisation in fact closely resembles the current system as detailed above (Sautman 1998); however, the cost of tuition in relation to standard of living may in real terms be higher today.

observes that Chinese rulers in the 1800s 'did not allow the light of modern education to fall upon their Muslim subjects' (1985: 99), while the first ruler of Republican Xinjiang, Yang Zengxin, similarly tried to 'keep Uyghurs uninformed' (Mackerras 1995: 20). Other respondents suggested that the policy aimed to prevent Uyghurs from competing with Han in education and employment. One Ürümchi mother, who had originally planned to send her daughter to a Chinese-medium school, told me in 1995 how she had been obstructed each time she tried to enrol her. Han teachers repeatedly said that they were 'busy', and asked her to 'come back later'.

By the mid-nineties, Chinese as a medium of tuition dominated at university level, while minority-language tuition was also being slowly removed in a non-systematic fashion at primary and secondary level. The Great Western Development strategy and the 'bilingual education' policy, introduced from 2000, systematised and sped up this process on the grounds of supposed communicative inferiority, as demonstrated in the below argument by then Regional Party Secretary, Wang Lequan:

The languages of the minority nationalities […] do not contain many of the expressions in modern science and technology which makes education in these concepts impossible […] This is why the Chinese language is now used […] This way, the quality of the Uyghur youth will not be poorer than that of their Han peers when they grow up (cited by Becquelin 2004a: 376).

Linguistic anthropologist Arienne Dwyer has rejected this claim as 'simply untrue'. She counters: 'All languages, given proper language planning, are capable of complex expression' (2005: 37). In fact, efforts had been made since the late 1950s to modernise the Uyghur lexicon, but 'major minority languages' such as Uyghur, Mongolian and Tibetan ended up falling behind Chinese in terms of capability of expression owing to the lack of a strong official mandate. The fact that scientific and technological terms are currently introduced from the Chinese and not from English entails considerable irony. Dwyer notes that Uyghur and English are stress-based languages (unlike Chinese, which is tonal), and that Uyghur has a larger phonetic inventory than Chinese, making English sounds potentially easier to learn for Uyghur speakers than for Chinese speakers. Moreover, the Uyghur language already contains many loan words from Russian and Persian. When Uyghur students learn the English word 'computer', they are taught the Chinese *diannao*. If they had learned this lexeme directly from English, they would have found it almost identical to the Russian loan in Uyghur, *kompyuter* (Dwyer 2005: 27-8). These irritations led many of my

#### 44 chapter one

Uyghur respondents to argue in 1995-6 that it made far better sense to learn English directly.

In September 2002, Xinjiang University (formerly the flagship of truly bilingual education) ceased to offer Uyghur-medium instruction completely amid public disquiet. In defence of the move, the Dean spoke of 'language deficits' in Uyghur students which caused them to fall behind Han students and prevented them from finding work. In 2004, *Xinjiang Daily* announced that all ethnic minority schools and Chinese schools were to be 'consolidated' by 2011, and that minority students would be mixed with Han students. In that year alone, 50 minority and Chinese schools in Xinjiang were merged. Concomitant with these developments is the replacement of minority-language educators with Chinese-speaking (either Han or *minkaohan*) teachers, so that educators who teach in the Uyghur language have been forced to retire or else take extra classes in Chinese in order to pass the HSK (*Hanyu shuiping kaoshi*) language proficiency test (Dwyer 2005: 38-9; Schluessel 2007: 257). Yet progress has been slower than the state would like. Because of a lack of qualified minority teachers (where 'qualified' means proficient in Chinese), Ma Wenhua of the XUAR Education Department estimated in 2006 that only 5 per cent of minority primary schools had so far begun to teach in Chinese. In a bid to further strengthen the 'bilingual education' policy, seven rural prefectures began to teach Chinese at nursery school level that same year, offering financial incentives to pre-schoolers and their teachers who joined the programme (US Congressional-Executive Commission on China 2006). The 'bilingual' policy direction has been met with deep ambivalence by the local population, with many viewing school mergers as 'linguicide' (forced extinction of the Uyghur language and, by extension, the Uyghur culture). Dwyer (2005: 38-40) suggests that such perceptions are now fostering the very identity polarisation (Uyghur versus Han) that Beijing seeks to neutralise. Based on my interviews and field observations held between 1995 and 2004, I strongly support that proposition.

#### 1.4. *Employment*

The growth of the Han population in the region, combined with the growing dominance of the Chinese language in education, has impacted dramatically on the labour market. Post-1949 waves of migration produced a new distribution of population (Han Chinese as opposed to local Turkic

peoples) along a central axis north of the Tängritagh (Tianshan) range, taking in Ürümchi, Zhangqi, Shihezi, Qaramay and Bortala. These cities in northern Xinjiang have become the industrial and commercial bases and transportation centres for the region (Li 1990: 66). They have also benefited exclusively from the selective introduction of market mechanisms (Becquelin 2004a: 373). Originally earmarked as the headquarters of the XPCC, Shihezi has benefited most of all, from the construction of roads, offices, schools, hospitals, factories and Xinjiang's first high-rise building in 1951-2 to the ongoing 'preferential treatment' it receives, such as annexation of land, central government subsidies, the city's upgrade from county-level to prefecture-level status in 1984,24 and the inauguration of its Economic and Technological Development Zone (ETDZ) in 1994 (Li 2010: 50-52; 54-56; 61).

The beneficiaries of accelerated development in the northern 'New Towns' have almost exclusively been Han. The provincial government, meeting in October 1954, decided that while suitable jobs would be found in the cities for incoming Han migrants, local peasants who came to urban areas would be reminded of the importance of agricultural production and 'persuaded' to return to their villages. As a result, Han people were disproportionately settled in urban areas favoured with new job creation, such as Ürümchi (fast-developing regional hub), Qaramay (regional centre for the oil industry) and Shihezi (main base for agricultural and militia corps— XPCC). According to two reports published in Beijing in 1956, the state covered the moving expenses plus 50 per cent of the rail fares for Han who resettled in Xinjiang at this time (Schwarz 1963: 68-9). Many of the recognised 'big enterprises' in Xinjiang in the 1990s originally developed from enterprises that moved in from other parts of China in the 1950s-60s (Shanghai, Changchun, Tianjin, Lanzhou and Xi'an). Such relocations often involved the migration of thousands of Han households; the First Construction Company in Changchun, for example, transferred in 2500 households (Ren and Yuan 2003: 100-101).

From the mid-1980s, permanent settlement—rather than seasonal 'labour migration'—became the norm (Ren and Yuan 2003: 99-100), and organised migration seems to have been supported by a number of covert government initiatives. For instance, some areas were granted extraordi-

<sup>24</sup> Only 3 of the 16 prefecture-level regions in Xinjiang—Ürümchi, Qaramay and Shihezi—had been awarded prefecture level city status by 2005. All have majority Han populations (Li 2010: 56).

#### 46 chapter one

nary permission to accept 'professional people' (those educated to above college level) *without requiring the agreement of the original work unit* (Li 1990: 70). Many of the permanent migrants (including organised and spontaneous migrants) moved into XPCC areas; such movements accounted for 33.12 per cent of all immigration to Xinjiang between 1991 and 1993. Others took jobs in the rapidly developing industries (especially oil exploration), new urban enterprises (often small-scale private enterprises) or new infrastructure projects such as the second Europe-Asia Continent Bridge, which connects Eastern China with Kazakhstan (Ren and Yuan 2003: 101; 104). Thus, while industrial and commercial development *per se* ought to have increased job opportunities and standards of living among local people, economic growth tended to disproportionately benefit new Han settlers.

In 1997, the percentages of Han people within the populations of the developed cities of Ürümchi, Qaramay and Shihezi were 72.7 per cent, 76.4 per cent and 95.1 per cent respectively (Ma 2003: 112). In Ürümchi, the influx of Han from China proper has been countered by a steady in-migration of rural minorities from southern Xinjiang.25 However, Uyghur migrant workers have not enjoyed equal access to employment in the city, while Uyghur university graduates are required to return to their hometowns following the culmination of their studies. This requirement stands in stark contrast against the right to free mobility enjoyed by their Han counterparts (Sautman 1998: 101). Within the XPCC areas (both north and south), populations are composed predominantly of Han employees. Ma (2003: 121) gives the example of one regiment farm in Shule County, part of an XPCC agricultural division in the Qäshqär region, where 91.7 per cent of workers are Han. Even within state-run agricultural and pastoral farms, breeding stations and horticultural gardens (which are not responsible for safeguarding national security), Han workers can make up anything between 10 and 40 per cent of the workforce, on the questionable basis that the 'technological requirements' of this type of work can be met only by Han personnel (Ma 2003: 117). At the same time, the import of companies from China proper has continued, resulting inevitably in increased competition for contracts. Local Uyghur contractors in the 1990s complained of being unable to compete with construction companies that had moved in from the east (Sautman 1998: 102).

<sup>25</sup> The percentage of 72.7 per cent calculated for the Han in 1997 represented a fall from 74 per cent in 1995 (Wang 2004).

At the start of the new millennium, there were no signs that these patterns were going to abate. Within the broad thrust of the Great Western Development strategy, Wang Lequan planned in 2000 to further speed development in 'areas with favourable conditions' (that is, those populated by Han), specifically, Ürümchi, Changji, Shihezi and Qaramay (Sines 2002: 11). To this end, new measures were introduced to allow Han sojourners in urban areas equal status to permanent residents in terms of access to public goods and services. These included the transfer of household registration and the possibility of dual registration. The same privileges were not however afforded to Uyghur migrant workers in the cities (Becquelin 2004a: 370). Once again, Shihezi was earmarked by the XPCC to receive 'disproportionately more attention' in terms of policies, investment, technological development and human resources. As a result, by 2006 around 75 per cent of share-holding industries, over 50 per cent of private industries and nearly 80 per cent of foreign-supported enterprises within the XPCC were concentrated in that city (Li 2010: 60).

Just as Han settlers have colonised areas targeted for development in Xinjiang, so too have they dominated white-collar, government and hightech occupations within those areas. Many newcomers arriving in Xinjiang in 1955-6 had experience of petroleum engineering and were sent directly to the oilfields in Qaramay. The wage-grade system was reformed so as to differentiate between Han and non-Han workers: 'A more sophisticated wage scale would tend to segregate the Chinese and non-Chinese workers […] Thus, although doing essentially the same type of work, the Chinese workers could be given higher wages under the new system' (Schwarz 1963: 63; 66-67).26 Schwarz suggests that Han migrants were probably also paid cost-of-living subsidies (based on the costliness of imported commodities) not made available to local farmers and herdsmen (Schwarz 1963: 68-9). According to Pye, sources of ethnic inequality after 1969 included Han monopoly of better farming lands, government offices, and better paid jobs on the railroads, in the industries and in the servicing of the military (1975: 503).27 By the 1970s, 'de facto control of Xinjiang lay squarely in the hands

<sup>26</sup> This situation was mirrored in Qinghai, where in 1958 a high-level Party official told the *People's Daily* that newcomers occupied higher positions, while the Qinghai Daily reported on the case of a Han settler who benefited from manipulation of the labour points system (Schwarz 1963: 72).

<sup>27</sup> Again, there are echoes of population transfer policies in the Republican era. In 1942, Jiang Jieshi announced the transfer of 10,000 officials and their families to Xinjiang, following which outsiders soon began to replace locals in positions of influence (Dreyer 1967: 24).

#### 48 chapter one

of Han Chinese males', while the tendency for Han incomers to move straight into elite managerial, political, and technical positions had set the stage for ethnic tensions (Hannum and Yu 1995: 6). In 1982, labour distributions in Xinjiang clearly reflected the tendency to deploy Han rather than minority workers into high-tech posts within the secondary industries. 84.3 per cent of minority nationalities were employed in the primary industries (mining, farming and fishing), a figure 11 per cent higher than the national average. On the other hand, only 6.06 per cent were employed in the secondary industries (manufacturing, production of consumables, and telecommunications), a figure 10 per cent lower than the national average. Only the figure for minority nationality employees working in the tertiary industries (provision of services to businesses and individuals) fell within 1 per cent of the national average at 9.64 per cent (Wang 2004: 9). In their discussion of elite occupations, Hannum and Yu (1995: 11) noted that despite constituting 55 per cent of the total regional labour force in 1990, minorities comprised only 36.3 per cent of those holding elite posts. Minority nationality employees were particularly under-represented in the government occupations and in professional and technical jobs; in the former case, the authors attribute this trend to the fact that increased educational attainment is no substitute for political loyalty. An unexpected finding of their study was that while the chances of attaining an elite job among minority individuals with primary level education were just 39 per cent of those of Han peers with a similar level of education, minority persons with senior high school attainment appeared to enjoy better chances than Han counterparts (Hannum and Yu 1995: 14; 16). On the other hand, Ma Rong's research confirmed that the percentage of Han people working in administrative offices and manufacturing or mining industries in Xinjiang was far higher than for the Han nationally. In 1990, 12.4 per cent of Han residents were technicians compared with just 5.3 per cent of the Hui and 4.1 per cent of the Uyghurs. Conversely, only 38.1 per cent of Han worked in farming and forestry compared with 66.9 per cent of the Hui and 84.1 per cent of the Uyghurs (Ma 2003: 110-11). Calla Wiemer's statistical research showed that in the year 2000 non-Han employees occupied just 30.1 per cent of *zhigong* positions [white-collar occupations], compared with their 59.4 per cent share of the regional population (2004: 179). Owing to the status quo of filling white-collar jobs with Han incomers, a social survey conducted at the turn of the millennium showed that while 65.2 per cent of Han respondents welcomed new migrants from China proper, a total of 59.6 per cent of Uyghur respondents either said that such migrants were unwelcome or declined to comment ('don't know'), this latter representing a guarded response to a sensitive question (Yee 2003: 442).

While permanent settlement by Han migrants has been the main reason for the changed labour structure in Xinjiang, the annual in-flow of seasonal workers has exerted additional pressure on the labour market. A report produced by the Intelligence Cadre Department in Xinjiang estimated at the end of the eighties that the level of temporary migration had reached approximately 200,000 individuals each year (Li 1990: 64). This figure rose to 400,000 in 1992 and had tripled to 1.4 million by 1997 (*Beijing Review*, cited by Wang 2004). Meanwhile, in rural Qäshqär the number of permanent Han settlers is said to be on the decline, replaced by a floating population of contract workers, who come each spring to work on construction projects or in the vegetable fields and leave again in the autumn. Because they do not intend to stay, their contact with local people is minimal, and mutual misunderstandings have reportedly led to cultural and religious conflict (Ma 2003: 118-9).

Since the second half of the nineties, employment discrimination has become a very tangible problem in Xinjiang (as it has across the whole of China). The difference in Xinjiang is that discrimination occurs directly along ethnic and linguistic lines. As Dwyer notes, while all languages in China are theoretically equal, *de facto* language policies promote an unequal division of power and resources, both linguistic and material (2005: 50). Furthermore, in order to be hired into an elite job, applicants need not only superior (Chinese-medium) education but also superior contacts (Gladney 1997: 290; cf. Sautman 1998: 102). This means that skilled and high-tech jobs in the industrial and energy sectors often go to Han candidates by default, because they tend to be better connected with (invariably Han) company managers. Colin Mackerras, writing in 2001, observed: 'Many Uyghurs have told me that they feel it is generally the Han Chinese who call the tune in Xinjiang. They believe that it is Han who generally get the best jobs and that Han employers tend to look down on Uyghur applicants as less "civilized" and less well educated' (2001: 299). While the Regulations on Work with the Urban National Minorities (1993) state that enterprises should be encouraged to recruit minority workers, requiring large state enterprises in Xinjiang to fulfil a 25 per cent minority quota, this falls far short of the 60 per cent minority share of the regional population. Meanwhile, oil exploration is exempted from the quota on the grounds that only graduates of petroleum technical school can be employed in developments like the initial Täkilimakan desert exploration of 1995. Chinese dissident

#### 50 chapter one

author, Wang Lixiong, in his controversial book *Wo de Xiyu, ni de Dongtu* [My Western Regions, Your East Turkestan], reports with first-hand detail on deepening nepotism and corruption in urban Xinjiang, giving the highprofile example of the monopoly on the sale of mineral water held by the son-in-law of former Party Secretary, Wang Lequan (Veg 2009a). Preferential policies, while helping more minority applicants into positions as state cadres, are less helpful when minorities seek posts in private enterprise. China's law on enterprises currently has no preferential policy provision for minority nationalities, so that enterprises contracted to Han managers normally employ only Han. Some enterprises contracted to minority managers employ minority workers, and ethnic affairs cadres try to get more minority workers hired in their municipalities, but with increased marketisation, cadres admit that new quotas need to be extended to the private sector (Sautman 1998: 97).

Ethnically based employment discrimination has greatly worsened since the 1997 disturbances in Ghulja. Following this event, Chinese state discourses began to routinely refer to 'Muslim separatists' when discussing anything relating to Xinjiang, and this was joined by the label of 'terrorists' following the attack on the World Trade Centre in New York and the subsequent 'war on terror'. The insidious effects of mass-mediated discourses on terror have been far-reaching, as Dwyer reminds us (2005: x). I would suggest that it is the prevailing sense among Han residents that Uyghur locals are 'ungrateful', disloyal and dangerous that has made Han employers so loath to hire Uyghur workers. Even independent minority businesspeople are not immune; many have encountered ethnic prejudice when applying for bank loans or administrative authorisations (Becquelin 2000: 85). Literature on the Western Development strategy does nothing to dispel the fear that this trend will continue. The stated purpose of the campaign is to remedy the problem of unequal economic development via privatisation of state-owned enterprises and responsiveness to market forces (Sines 2002: 5; 8). This, unless accompanied by stringent legislation guarding against ethnic discrimination, can only consolidate the prevailing pattern of preferential treatment of Han applicants by private firms. As Sines observes: 'Since local agents are [considered] either incapable or untrustworthy, it seems development will come mainly by the activity of outside agents […] Western Development is still entrenched in a colonial ideology. National minorities still seem to lack agency for their own development and are talked about as if they have little or nothing to bring to the table' (2002: 16). In the summer of 2000, Han students in Ürümchi were excited

about the employment prospects offered by the increased investment in private enterprise; young urban Uyghurs, however, greeted the programme with scepticism.

As a result of this and other factors discussed above, under-employment and unemployment among the Uyghurs has reached serious levels. According to the 1996 Statistical Communiqué for Xinjiang, there were 98,000 unemployed persons in urban Xinjiang, or 3.8 per cent of the total population. Tellingly, the report gave no ethnic breakdown.28 While Han employers frequently cite 'an insufficient grasp of the Chinese language' as the reason for making Uyghurs redundant or for failing to hire Uyghurs, respondents commonly claim that labour discrimination is in fact ethnically based (cf. Baranovitch 2007a: 464, 489). In the second half of the nineties, minority businesspeople reported the eviction of minority managers in small-scale, state-owned enterprises and their replacement with Han employees (Becquelin 2000: 85). According to a study by one Han scholar, by 1998 one third of the Uyghur population in Xinjiang was 'underemployed', while in some districts as many as 45 per cent of the population had been forced to migrate elsewhere to find work (Jiang, cited in Becquelin 2004a: 372). Uyghur migrants to urban centres additionally face disadvantaged status—and as a result unequal access to resources—since they retain their rural *nopus* [Ch. *hukou*, household registration]. Moreover, female migrants pass this status on to their children (Bellér-Hann and Hann 1999: 26). In 1996, a temporary urban residence card cost 360 yuan, while a permanent residence permit cost an astounding 15,000 yuan (Benson 2004: 209), the equivalent of 3-4 years of income for a rural resident dependent on subsistence farming. As a result, Yee's study conducted at the turn of the millennium concluded that unemployment is a major cause of tension in urban Uyghur-Han relations, with minorities the first to be laid off as the market economy takes hold (Yee 2003).

These problems have only rarely been acknowledged by the state. A high-level report from the Xinjiang CCP committee in 2001 was exceptional in its honest acknowledgement that the policy of 'choosing from both sides' (i.e. hiring workers from among both Han and minority nationalities) had become impracticable, and that the government was increasingly failing to intervene in discriminatory employment practices under the market economic system (Becquelin 2004a: 375). In India, this problem has been one factor behind widespread inter-communal riots. There,

<sup>28</sup> *Xinjiang ribao*, Ürümchi, 14th March 1997 in SWB (Asia Pacific), 7 May 1997, FEW/0485 WS2/6.

#### 52 chapter one

Muslims are widely discriminated against in both public (government jobs) and private (private sector jobs; bank loans), despite an official government ideology that commits itself to non-discrimination on the basis of race, caste or religion (Marsot 1992: 165). In Xinjiang, where the problem is demonstrably more acute in the private sector (Han managers are at liberty to hire and fire whom they please), the problem has equal potential to generate social unrest and by extension national instability. Improving the educational level of minority persons will be pointless unless the state can ensure that there are legal safeguards to protect equal opportunities: 'A system that produces educated ethnic youths only to prevent them from obtaining good jobs appears bound to alienate a growing proportion of them' (Becquelin 2000: 86).29 This view is echoed by Herbert Yee, who points out that while thousands of minority graduates are without jobs, Han graduates from China proper can easily land a job in the Xinjiang regional government, leaving educated young people from minority groups 'unhappy and restless' (2003: 449).

The findings summarised above are fully corroborated by my own field research. Already in summer 1995, my respondents in Ürümchi noted that only a small percentage of Uyghurs—those who mastered the Chinese language well enough to go to university—were suitably qualified to seek work in the Han-created and Han-dominated urban job market, where all transactions were carried out in Chinese. Those not fluent in Chinese were automatically deprived of opportunities available to the Han population. Shökhrät, a young male intellectual speaking in interview in 1995, described how the 1993 Forum for Talent Exchange (*Rencai jiaoliuhui*) in Ürümchi had been closed early after Han company representatives were accused of intending to employ only Han people or *minkaohan* (Chinese-educated Uyghurs), and students from Xinjiang University threatened to march in protest. The state's response to this potential for conflict was not to introduce equal opportunities legislation but rather to issue a limited number of entry passes for the same event in 1995 and to send these to selected companies and work units only (i.e., those which employed Han). Shökhrät nonetheless obtained entry to the 1995 Forum, and observed that only one company stall had its Chinese-language details reproduced in the Uyghur

<sup>29</sup> This reflects the prevailing situation during GMD rule in Xinjiang. As Dreyer observed, the imposition of Chinese-language tuition within the GMD's educational programme might have been more successful had it been combined with economic opportunities for those who had completed the course of instruction (1976: 40).

script. He decided to test employer attitudes by approaching a computing firm and offering them his services, but was told by representatives: 'We hadn't considered Uyghurs' (*Women meiyou kaolü Weizu*) (cf. Gilley 2001; Taynen 2006: 62). Nor was this phenomenon limited to northern or urban Xinjiang. In 1996, southern respondents also complained of Han privileges in the labour market. One case described by Nijat and Mälikä, two secondary-school teachers in Khotän, involved plans in 1995 to install an all-Han workforce in a local cotton factory.30 They reported that, when the local Uyghur leader objected, protesting that the workforce should at the very least be half Uyghur and half Han, he was removed to Beijing for 're-education.' According to my own observations, this imbalance was clearly reflected in the ratio of Han to Uyghur employees in some state work units. For example, the History department at the Xinjiang Academy of Social Sciences employed only three native Uyghur scholars in 1996 compared with forty-four Han scholars.

Respondents asserted that, often, the only posts available to them were blue-collar ones, and that they were expected to do the dirty and tiring jobs while Han people occupied the white-collar jobs. This only deepened local resentment. Qadir, a taxi driver in his thirties from Khotän, complained in 1996: 'If more Han migrants come, they will take our jobs! It is like that already. Uyghurs do all the exhausting work, for instance, farming, and Hans do all the nice jobs.' These claims are amply supported by government statistics, which show that while, in 1989, 148,000 minority individuals held 'professional and technical posts', by 1995 this figure had seen only a modest increase, rising to 181,000 individuals (Benson 2004: 214). 'Ethnic nepotism' was frequently the focus of popular story-telling. Mälikä, a secondary school teacher in her fifties, complained in 1996 that Han in eastern China were being offered incentives to move to Xinjiang: 'A Han will come to Xinjiang, do some petty business for maybe a year, and then all of a sudden become a high-ranking cadre! You don't see Uyghurs becoming cadres overnight.'31 One evening, as I watched television in Khotän in July 1996, a programme was aired which told the story of two Han migrants who had

<sup>30</sup> The Xinhua news agency reported in 1997 that hundreds of thousands of "surplus labourers" from China proper come to Xinjiang during the cotton harvest. Of these, some return home when the season is over, while others stay on in Xinjiang. Xinhua news agency, Beijing, 19 May 1997 in SWB (Asia Pacific), 22 May 1997, FE/2925 G/8.

<sup>31</sup> As early as 1989-90, Justin Rudelson (1997: 68) had noted a similar anecdote circulating in Turpan about a Han shoemaker who arrived from China proper, became mayor of the oasis within a week, and governor by the end of the year.

#### 54 chapter one

come to Khotän two years before to work on cotton production, and were recently made leaders of their units. Such publicity, surely aimed at building confidence among the (then) less established Han communities in the south, not surprisingly gave rise to local perceptions of inequality. Like the Han settlers of the 1950s-60s, many of whom were reluctant to relocate to an inhospitable climate and alien cultural environment, Han migrants who arrived in south Xinjiang in the early nineties seemed uneasy about moving to a politically unstable region. While some declined to confirm that they had been attracted by government incentives, a female respondent in her twenties admitted in 1996 that she and her companions had been offered better jobs, the chance to become high-ranking cadres, and the opportunity to build bigger houses on their own land, if they relocated to south Xinjiang. In one hotel in Khotän, I personally overheard two Han migrants telephoning their relatives in China proper in 1996 to report their arrival. Both men were to take up teaching posts immediately, an arrangement that had evidently been made before they came. Other evidence suggested that new migrants in the mid-nineties had needed no persuading. A Han bank clerk in his thirties, who had relocated to Aqsu from Jiangsu in the 1980s, explained in 1996 that whereas the government used to 'send people here to build up the region,' Han settlers were now coming voluntarily. This move was viewed as a happy relief from housing shortages and unemployment in the cramped east.

The preference of some work units and many private companies for ethnic Han employees had already produced a high rate of unemployment among *minkaomin* (Uyghurs educated in their mother tongue) by 1995. In Ürümchi, every day I observed crowds of Uyghur men of employable age standing around idly at the Döngköwrük bazaar. Respondents told me that these men had no jobs to go to, and asserted that Han migrants were 'stealing their jobs.' In the south, Nijat, a secondary school teacher in his fifties from Khotän, observed in 1996: 'The more Hans flock to Xinjiang, the higher the unemployment level among Uyghurs.' The fact that these migrants occupy posts which were created by Han investment and development does little to lessen local resentment. Most respondents argued that while the government had promised to modernise Xinjiang and improve living conditions, those improvements were intended exclusively to benefit Han people, while Uyghurs were forced into an increasingly marginalised position. *Minkaohan* respondents spoke with visible shame about unemployed Uyghurs who consented to do work that they themselves considered an embarrassing and undignified compromise. Räwiä, a *minkaohan* fortunate

enough to hold a post in an Ürümchi work unit, remarked helplessly in 1995:

I look at Uyghurs playing *sunay* and *naghra* at opening ceremonies for [Hanmanaged] department stores and shops and I feel ashamed for them. They are really pitiful. What good will that shop do for them? Will they be able to work there? Will they be able to buy the goods inside?

According to popular opinion, most of those individuals liable to anti-state, separatist activities in the mid-1990s were unemployed persons. As one Ürümchi publisher suggested in 1995, jobless, unmarried Uyghur men had nothing to lose in pursuing independence.

# 1.5. *Wealth Distribution*

The selective development of new Han communities north of the Tianshan coupled with the dominance of the Chinese language in education and ethnic discrimination in employment (all described in the previous sections) has resulted in an uneven distribution of wealth—or income gap in Xinjiang. This is clearly discernible along north-south, urban-rural and Han-minority lines. According to David Wang, in 1990 the rich urban centres of north Xinjiang—Ürümchi, Changji, Qaramay, Qumul and Shihezi were settled by majority Han populations of between 66 and 95 per cent. By contrast, more than 90 per cent of those living in the poorest areas of south-west Xinjiang—Qäshqär and Khotän—were Uyghurs. In some counties, this figure exceeded 95 per cent (Wang 2004; cf. Wiemer 2004: 177, 188 and Li Yuhui 2009: 12). At the same time, Han people constituted a disproportionate slice of the population in urban centres of the south. In 1990, the proportions of Han resident in the county towns of Qäshqär, Shufu, Shule and Yängi Shähär [Yengisar] in the Qäshqär region were 23.3 per cent, 36.8 per cent, 17 per cent and 30.9 per cent respectively. Compare these figures with the far lower proportions of Han within each county's population as a whole (i.e. incorporating the rural areas): 22 per cent, 1.9 per cent, 6.3 per cent and 1.7 per cent (Ma 2003: 114-7). In 1996, 13 of the 20 counties in which Uyghurs accounted for over 90 per cent of the population were designated 'key poverty alleviation areas' by the state (Zhang and Huang, cited in Becquelin 2000: 68).

The quality of medical care oscillates along a similar axis. It is true that the transfer of skilled Han medical staff into the region brought about a decrease in the crude death rate from 20.81 to 13 per thousand between

Figure 4: Supplicant at Kokhmarim mazar (Snake Mountain shrine), Khotän (1996)

1949 and 1958, and to less than 10 per thousand by 1966. However, it is also the case that better medical care has been available in newly opened areas populated by Han settlers than in areas traditionally settled by Uyghurs. For example, in 1984 the death rate in Uyghur-dominated Qäshqär in southern Xinjiang was 12.06 per thousand compared with just 2.17 per thousand in Qaramay, a city in northern Xinjiang populated by Han oil field workers (Li 1990: 55-6). While part of this discrepancy arises from the higher proportion of young people (migrant workers) living in Qaramay, the contrast remains dramatic. Differences in quality of life can also be observed in the average life span of a Xinjiang resident, which is higher in the Han-dominated cities than in the Uyghur-dominated countryside (Li 1990: 56).

My interview data reflect these statistics. Respondents claimed that at the national level economic development since 1978 had done little to benefit minority areas, in particular those situated in the west. Ghalip, a history graduate who ran a restaurant in Ürümchi in 1995, asserted that per capita standard of living in Xinjiang had halved since 1949, and that both

Xinjiang and Tibet were 'among the poorest regions in China'. These claims are questionable, since in reality economic development in Xinjiang has far exceeded that in Tibet and is more buoyant than that of a good many other provinces. However, it is also known that the ratio of total value of agricultural and industrial outputs in minority areas (relative to the value of outputs across China as a whole) was just 7.85 per cent in 1949 and 7.2 per cent in 1990 (Sautman 1998: 99). In other words, while local standards of living may have risen overall, the gap between western and coastal regions is greater now than ever, a fact also acknowledged by the state since the turn of the century. At the regional level, respondents in the urban capital pointed to ethnic differentiation in standards of living, noting that many urban Uyghur homes had no telephone, and that almost no Uyghur family owned its own car. By contrast, they pointed to the brand new Mercedes and BMWs being driven by Han residents. While such vehicles were in fact often owned by state work units and shared by multiple employees during the 1990s, the daily sight of Han faces behind the steering wheel nonetheless served to stoke local resentment. According to a Han clerk working in the Bank of China in Aqsu in 1996, his branch had registered only six Uyghur customers, compared with the Bank of China's Qäshqär branch, where a Uyghur clerk confirmed in the same year that 50 per cent of customers were Uyghur merchants.32 The small number of Uyghurs using the Bank of China in Aqsu New Town likely reflected the higher proportion of Han in the urban population of Aqsu as well as their monopoly of higher-status positions. In the same way, the higher percentage of Uyghurs banking at the Qäshqär branch probably reflected the smaller proportion of Han settlers in the Qäshqär region (thus higher numbers of Uyghurs in state posts) as well as Uyghur dominance of traditional modes of cross-border trade. Interviewed in rural Aqsu in 1996, Uyghur peasants acknowledged the rise in their standard of living since the introduction of economic reforms (claiming an average annual income of between 5000 and 6000 yuan), but bemoaned the return of social inequality alongside those reforms. Some even looked back with nostalgia at the social egalitarianism of the Maoist period.33

<sup>32</sup> We can assume that both figures are slightly exaggerated as the Han clerk may have intended to portray Uyghurs as 'too stupid' to hold down a job or have a bank account, while the Uyghur clerk may have wished to emphasise that many Uyghurs were rich and successful.

<sup>33</sup> Uyghurs of a certain generation remember that during the Great Leap Forward years wheat and maize were distributed in egalitarian fashion (cf. Bellér-Hann and Hann 1999: 5).

#### 58 chapter one

My personal observations supported the existence of an ethnic income gap. 'New Towns' in Xinjiang were largely composed of brick-built *loufang* (multi-storey) or *pingfang* (single-storey) residential blocks. These were custom-built for new Han settlers and included the considerable benefits of running water and electricity. Traditional Uyghur homes in the 'Old Towns' and rural areas often lacked such facilities. Since the allocation of a *loufang* home then depended upon holding a post in a state work unit, only a small number of *minkaohan* lived in the New Towns, and this had created perceptions of housing discrimination. One day in 1995, Räwiä, a *minkaohan* employee in an Ürümchi work unit, contrived a social visit to my home to communicate her views surrounding social injustices (ethnic inequalities) with regard to housing. She stressed: 'If you look at the buildings they [Han people] build, you will rarely find a Uyghur living in one of them! I don't care how much propaganda they put out [about ethnic equality]. Everyone can see; everyone can use their eyes.' Similarly, Uyghur locals were rarely seen travelling on aeroplanes or comfortable sleeper buses between oases in the mid-1990s. The more expensive modes of inter-oasis transport were mainly used by Han passengers, while Uyghurs travelled on rusty old vehicles furnished with hard seats.

Preferential policies, while certainly improving the lot of some minority individuals, have been unable to halt the escalating income gap between ethnic groups. Minority respondents observed during the mid-1990s that preferential policies remained 'too weak' to overcome the gap in education and living standards between most minorities and the Han (Sautman 1998: 101). As a result, the average annual income of an urban resident of Xinjiang (mainly Han) at that time was 4,252 yuan compared with an annual income of just 935.5 yuan among rural residents (mainly Uyghur) (Benson 2004: 214). A half-decade later in 2000, Yee conducted a social survey on ethnic relations in which 38 per cent of urban Uyghur respondents stated that their standard of living had risen more slowly than that of Han residents, and 46.9 per cent deemed the rate of development 'about the same' for both groups, this latter possibly reflecting a guarded response to a sensitive question. On the other hand, some 15.1 per cent claimed that Uyghurs were benefiting faster than Han (2003: 443). This view may reflect the emergence of a small Uyghur middle class composed of successful businessmen (*chong sodigär*) who have thrived on renewed opportunities for cross-border trade (Benson 2004: 213; Mackerras 2005: 18).34 When questioned on income

<sup>34</sup> See Erkin (2009) for an analysis of how the Uyghur middle class in Ürümchi expresses a contrastive ethnic identity through the selective consumption of foodstuffs, music and real estate.

disparity, however, feelings were clearer, with 46.6 per cent of Uyghur respondents acknowledging a large income gap between the two groups (Yee 2003: 443-4). This pattern prompted Colin Mackerras to muse at the turn of the millennium:

There seems little doubt that the Han are generally somewhat or much better off than most of the minority population. Most of the modern sector in a very advanced city such as Ürümchi appears to be dominated by Han, not minorities […] there are no doubt Han people who infringe official policy and do discriminate against Uygurs (2001: 299).

Government statistics released in 2002 showed that the annual per capita GDP in Xinjiang had increased by 14.3 per cent in (Han-dominated) urban areas, compared with just 8.8 per cent in (Uyghur-dominated) rural areas; and that southern Xinjiang—with its 95 per cent Uyghur population—had an average per capita income only half the regional average (Becquelin 2004a: 371-72). The gap was particularly visible when comparing the richest city in the north with the poorest town in the south: in 2002, per capita GDP in the regional capital Ürümchi was calculated at 16,493 yuan, almost ten times the per capita GDP of Khotän (Yee 2005: 45).

A quantitative survey conducted by sociologist Zang Xiaowei in Ürümchi in 2005 identified an additional reason for the aggregate ethnic income gap in Xinjiang: the impact of traditional gender roles on the division of labour in Uyghur households. Zang argues that, in order to 'compensate for their human capital deficit' (i.e. inferior education), Uyghur men are supported by their families to compete in the urban labour market. In other words, to allow their husbands to devote all their time and energy to work, Uyghur wives take care of all household roles, including housewifery and childcare. This 'mother penalty' leads to Uyghur women taking career breaks, and doing low-status and part-time jobs, which in turn enlarges the income gap between Uyghur and Han women, thus partially explaining the overall Uyghur-Han earnings differential (Zang 2012: 245). While this study provides an important modification to previous findings that identify employment discrimination and educational differences as the source of the Uyghur-Han income gap, it is important to remember that this gendered division of labour among contemporary urban Uyghurs is in fact reproduced by pre-existing disparities in human capital. Whether they wish to or not, Uyghur families are obliged to maintain this traditional structure for two reasons: first, in order to best enhance opportunities for (Uyghur) male earnings in a highly discriminatory urban labour market;

#### 60 chapter one

and, second, to ensure that Uyghur mothers can safely transmit Uyghur language and culture to the next generation via the vehicle of home education, increasingly important in an era of Chinese-medium state education.

# 1.6. *Resource Exploitation*

The issues that have drawn perhaps the bitterest opposition among Uyghurs are those relating to resource exploitation. Within the politically contested territory of Xinjiang, it is not surprising that local conflicts often focus on the land, its appropriation and its use, as demonstrated in the following confrontation observed by a female health professional named Adalät on a crowded Ürümchi bus in 1995:

A young Uyghur girl clutching a tiny baby got on the bus. No-one gave up their seat for her. In the past, Uyghurs always stood up and offered their seats to the old, and to women and children. But now the buses are full of Hans, and they never give up their seats. The girl leant against a seat where a young Han male was sat opposite his girlfriend, and tried to steady herself and the child. The man glared up at her and said 'Uggh, how dirty! How stinking!' A Uyghur policewoman standing nearby was incensed. She walked over and said to him 'Let's get things the right way round. Who's dirty? And who stinks? We've given you the whole of our land Xinjiang, yet you can't even give up your seat for a tiny baby!' The man was silent. When he still didn't move, the policewoman grabbed him by the collar, hauled him out of the seat, and bade the girl sit down.

Another female respondent, Räwiä, who was employed by an Ürümchi work unit, observed in 1995 that her Han colleagues often betrayed what she described as a 'guilty conscience' with regard to their presence in the region. According to her, this guilt manifested itself in the form of paranoia and over-sensitivity in the workplace: 'Every time we speak Uyghur in front of Hans, they think we are saying something nasty about them! If we sing a song, they think it is a mockery of them! In Chinese, they would call it *zuozeixinxu* [Lit: Those who thieve must fear discovery].' As a *minkaohan*, she used the Chinese language rather than Uyghur to question Han hegemony.

State and popular discourses regularly justify the Chinese presence in Xinjiang by pointing to the efforts made by the Han militia corps in land reclamation as well as the centre's investment in the regional economy. That the Chinese state has developed the land is indisputable. From the 1950s, the Xinjiang Production and Construction Corps (XPCC, *Xinjiang* 

*shengchan jianshe bingtuan*), described by Donald H. McMillen as an 'important symbol and vehicle of Communist Party of China (CPC) efforts to control and integrate the strategic and traditionally non-Han borderlands', harnessed rivers to build reservoirs and hydroelectric power stations, reclaimed desert land for the development of farms and enterprises, and built whole new cities in formerly remote areas (Shihezi being a key example) (McMillen 1981; Esposito 1977: 169-70; Ren and Yuan 2003: 100). To quote from a recent study focusing on the contemporary incarnation of the *bingtuan*, the XPCC 'reclamation warriors' of the Maoist period 'held a gun in one hand and a hoe in the other' (Cliff 2009: 87). Then, through the 1980s and 1990s, the state built a vast new infrastructure throughout the region, consisting of highways, railway networks, and aviation, communications and pipeline routes (Zhao 2001: 211-12). However, the contentious issues are: who was the land reclaimed for? Who profits from local resource exploitation? And, to what extent has regional development benefited all ethnic groups equally? According to Warikoo, in nineteenth-century Xinjiang (a time when the Han presence remained negligible) there was no need to reclaim additional land because farmed territories were sufficiently fertile to ensure cheap and plentiful food for the whole of the (sparse) local populace (1985: 80). This situation changed dramatically in the context of accelerated Han in-migration, and issues of land use now sit squarely at the centre of local conflicts.

One issue that has loomed particularly large in rural areas is cotton cultivation. Since the early 1990s, financial incentives have drawn increasing numbers of Han workers to Xinjiang to grow cotton. In addition to increasing regional population pressure and heightening labour competition (40 per cent of cotton production is controlled by the Han-dominated XPCC), this has led to large-scale reclamation of additional land. While two thirds of the reclaimed territory could be described as 'arid land' previously considered unfit for use, Xinjiang local media reported in 1998 that the remaining third consisted of planted areas, fertile forests and prairielands, and had been commandeered for cotton production by project managers who did not obtain official authorisation. Irresponsible land reclamation of this sort has substantially contributed to environmental degradation in Xinjiang, for example, desertification in areas surrounding the Tarim river (Becquelin 2000: 83-84). Meanwhile, in pursuit of its claim that cotton cultivation promotes 'socio-economic differentiation'—and thus improved livelihoods—among local peasants, the state has imposed cotton-growing quotas on rural Uyghur households. These have proved

#### 62 chapter one

extremely unpopular, with locals complaining that the operation is labourintensive yet unprofitable (Bellér-Hann 1997: 95; Becquelin 2000: 81-2). Their concerns were reflected in official statistics, which showed that in 1999 it was 15 per cent cheaper to import cotton than to produce it locally, and that this disparity had led to the formation of a regional stockpile. Indeed, it was reported that even cotton factories in the Han-dominated XPCC city of Shihezi had run into debt in consecutive years (Li 2010: 58). Bellér-Hann and Hann (1999:19; 23) firmly reject the notion that cotton cultivation in Xinjiang represents 'a commitment to social justice' on the grounds that the distribution of small cotton quotas across all villages ensures that no one peasant family can become a specialist producer. At the same time, the system prevents those peasants living close to cities from entering into commercial competition with urban Han. In sum, it suggests the state's unwillingness to allow private accumulation in the south, and its fear of the empowerment this might bring.

Another source of discontent has been the perpetuation of subsistence farming by a state seemingly reluctant to allow rural Uyghurs to transition to a differentiated or commercially oriented agriculture. Essentially, peasants have been forced to continue to focus on subsistence as a result of the state's produce procurement policy. A similar policy had been practised back in the 1800s, when Uyghur peasants were obliged by local begs, tax collectors and Manchu mandarins to sell corn and other produce to the

Figure 5: Threshing ground, Bay, Aqsu (1996)

authorities at prices far lower than the prevailing market rates (Warikoo 1985: 87). In 1996, Bellér-Hann and Hann conducted a study in three villages near Qäshqär, where they observed a continuing 'struggle over the harvest'. They describe how some households chose to retain their produce and pay the fine for failing to deliver rather than forfeit their self-sufficiency and be forced to purchase grain on the market (1999: 7-8; 11). They further illustrate how rural officials engaged in 'subversive acts' designed to help their fellow villagers explain shortfalls: 'Some seemed to hold a deep conviction that the peasant was right and the state was wrong' (1999: 20; 23). Finally, they cite respondents who were critical of policies which prevented the sale of land, or which required them to grow new crops or varieties purely because the yield was larger (Bellér-Hann and Hann 1999: 20). This latter complaint I too encountered while conducting fieldwork among peasants in rural Aqsu in 1996. Instructions (or rather, orders) to grow new varieties of maize made no sense to my respondents, who considered that their taste was vastly inferior to those they had grown for generations.

Across the second half of the 1990s, the XPCC—whose domains were originally situated mainly in the north—expanded into the rural south, where it 'opened up' new lands in Qäshqär county and other areas. Based on their rural field study undertaken in 1996, Bellér-Hann and Hann warned that if these reclaimed lands were operated by Han migrants from other parts of the country rather than by Uyghur natives, 'this can only be a recipe for conflict at some future date' (1999: 25). Yet there seems to have been little consideration of the tensions that such a policy might create. In 1998, *Xinjiang Daily* advertised household responsibility contracts to potential Han settlers, together with a promise of a 2-year fee waiver. Aside from the dual injustices of settling newly opened lands with Han migrant rather than Uyghur communities, and furnishing incomers with preferential land rental agreements not available to locals, the land itself cannot support such a high population density. Consequently, struggles have arisen over water, electricity and confiscation of land for the construction of roads. One 1996 study undertaken by a Han scholar found that the allocation of water resources often catalysed communal conflicts in a township south of Turpan, and that Uyghur farmers felt aggrieved that their limited resources were being taken from them by Han settlers (Li, cited in Becquelin 2000: 84). Becquelin, citing his own field research conducted between 1997 and 1999, documents a similar situation in Qiemo, a remote township in south-east Xinjiang. There, less than 2.5 *mu* of tillable land was available per person by the year 2000, and the resulting pressure on land use and

#### 64 chapter one

water supplies had led to fierce clashes between Uyghur and Han farmers (Becquelin 2000: 76; 84).

The issue most vehemently debated during the 1990s and arguably the principle factor underlying escalating regional unrest during that decade was state (read, majority Han) exploitation of Xinjiang's natural resources. During the 1800s, residents of Qäshqär had controlled the export of raw materials to Soviet Central Asia, including wool, cotton, silk, hides, hemp, gold dust and silver, and had received Russian roubles, gold and other currencies in exchange. With these earnings, they imported manufactured goods from India, Russia and Britain, apparently preferring Russian-made cotton cloth and chintzes to Indian textiles. These exchanges meant that raw materials eventually became prohibitively expensive on the domestic market and beyond the reach of the poor; they also caused the local silk and cotton weaving industries to languish (Warikoo 1985: 85-6; 106-7). With regard to mineral extraction, historical records suggest that local peoples were mining quantities of coal and iron sufficient to meet their limited demands in the nineteenth century. In 1868-9, iron furnaces were observed in Qizil, where that metal was used to fashion household implements. Coal was also burned in domestic hearths. However, according to Warikoo, generally speaking, local people knew little of the mining techniques they would need to benefit from deposits of copper, lead, gold and jade, and it was Chinese interest which first led to the working of these mines (1985: 79).

Today, Xinjiang has proven reserves of 2.5 billion tons of petroleum and 700 billion cubic metres of natural gas (Becquelin 2004a: 364-5). In order to extract these resources, China has established oil wells in the oasis of Qaramay (Uy: 'black oil') in northern Xinjiang, and in the Turpan-Qumul basin in the south. In 1992, Xinjiang produced just 6 per cent of China's oil (8.8 million tonnes), mostly from the Qaramay fields. From this time, ambitious plans emerged to extract oil from the Tarim basin in the Täklimakan desert. In October 1992, a Chinese drilling team struck oil at the Tazhong field, thought to be three times the size of the Daqing field (which then supplied around 40 per cent of China's annual oil output or 142 million tonnes). Subsequently, China launched a crash programme in 1993 to develop the Tarim's Lunnan oilfield. Foreign oil companies reportedly paid up to 1.4 million US dollars each for Chinese data relating to the foreign bidding block, and it is likely that this money went straight to the stateowned China National Petroleum Corps (CNPC) (Goldstein 1993: 57). Between 1996 and 2005, crude oil output in Xinjiang increased from 14 million

tons per year (most of it processed and used within the region),35 to 24 million tons per year, and President of the PetroChina Tarim oilfield, Sun Longde, estimated that the field would produce 100 million tons of oil annually by the year 2010 (Pocha 2006). Up until 2005, the oil industry was controlled by the China National Petroleum Corporation (CNPC) and SINOPEC; however, from 2006 local government introduced tax exemptions, tax reductions, low land rent, low land transfer fees, and other new regulations in a bid to encourage increased domestic and foreign private investment (Center for Energy and Global Development 2006).

The fact that Xinjiang's oilfields fall under the management of central rather than regional government further deepens local suspicions. As Gladney observed with regard to pro-independence sentiments in the 1990s: 'The initial rumblings of discontent centred on a desire to see the benefits of the north-western region's oil and mineral wealth begin to flow back into the region' (1997: 287). One controversy stemming from centralised control concerned the fiscal reforms of 1994. Their purported aim was to tighten fiscal control from the centre as a means to allow for redistribution of tax revenue among coastal and interior provinces, thus easing the east-west wealth gap. Unfortunately, within the reforms, the policy of reducing taxes on manufactured products and increasing taxes on raw materials unwittingly penalised Xinjiang, which is a key producer of the latter (Becquelin 2000: 72-74). In interviews held in the second half of the nineties, regional officials stated that central and foreign companies benefited more from local oil than did local people, pointing out that only 30 per cent of tax income from Xinjiang oil and only 50 per cent of tax income from petrochemical products returned to local government (Sautman 1998: 102). Local Uyghurs made similar comments to the *New York Times* in 2006 that the region received little benefit from its own energy reserves because energy production was 'controlled by China's state-owned companies, consumed by China's coastal cities, and taxed in a way that central government gets most of the revenues' (Pocha 2006). In response, Wang Lequan, then Communist Party secretary for Xinjiang, acknowledged that approximately 75 per cent of oil taxes did go directly to Beijing, but protested that much came back in the form of transfers or federal funding. He also pointed out that free natural gas was being supplied to people who lived near the Tarim

<sup>35</sup> Xinhua news agency, Beijing, 26 March 1997 in SWB (Asia Pacific), 28 March 1997, FE/2879 G/8.

#### 66 chapter one

Basin, although the real aim of this policy was environmental: to prevent protected poplar trees from being felled for fuel (Pocha 2006).

Far from addressing resource-related reasons for regional disaffect, state policies on resource extraction have increasingly focused on the need to cut China's energy imports, control its own energy supplies, and thus ensure energy security. This aim is especially clear in the documentation surrounding the Great Western Development strategy, which describes the West-East natural gas pipeline, stretching from Lunnan in the Tarim basin to Shanghai, as a 'critical project' (Becquelin 2004a: 364-5). Initially, oil had been carried to the rail head at Qorla via a small-bore pipeline; but the state quickly realised that even an upgraded rail route would have insufficient capacity to cope with increased extraction. It therefore began planning for a controversial and expensive 4,200 kilometre pipeline. This route would transport the oil east to Luoyang (Henan province), the home of a major refining centre, and from there to Shanghai and perhaps even Beijing (Goldstein 1993: 54-55; Pocha 2006). According to a 1997 report published by the UN Centre for Human Rights, Xinjiang was required to deliver 80 million tons of petroleum to central government during the period of the Ninth Five-Year Plan (1996-2000);36 a volume equivalent to 16 million tons per year. Thus, national dependency on Xinjiang's natural resources makes political secession quite inconceivable for the foreseeable future, leading one Uyghur activist to remark cynically in 2006: 'Our oil has become our curse' (Pocha 2006).37

Xinjiang's abundant natural resources attracted an overwhelming influx of Han settlers to resource-rich areas between 1990 and 1995, as documented by Ren and Yuan (2003: 96). The lucrative oil and mining industries have generally taken the form of independent economic organisations, and tend to be staffed by Han workers. To give some examples, in 1990, 66.4 per cent of residents in the Poskam County oilfields in southern Xinjiang were Han (Ma 2003: 116-7; 121). In 1993, 90 per cent of adults living in Qaramay—an oil city heavily dominated by Han—worked for the Xinjiang Petroleum Bureau, while some 30,000 Chinese workers were deployed on the new oil exploration in the Tarim basin. These migrants were reportedly paid

<sup>36</sup> Xinhua news agency, Beijing, 26 March 1997 in SWB (Asia Pacific), 28 March 1997, FE/2879 G/8.

<sup>37</sup> Alim Seytoff of the Uyghur American Association (UAA), based in Washington (cited in Pocha 2006). Resource exploitation had also occurred under the Chinese nationalists. In 1942, Jiang Jieshi set an artificial exchange rate in order to allow Chinese merchants to ship local products out of Xinjiang at great profit (Dreyer 1976: 24).

'princely wages' of up to 2000 yuan per month (compare this with the 5000 yuan per year earned by the average Uyghur peasant), plus large bonuses for early completion. Although they enjoyed little job security and were obliged to work 12-hour shifts, they were accommodated in prefabricated houses and trailers, equipped with dining hall, satellite TV and air conditioning. They were also permitted to take long breaks in their hometowns (Goldstein 1993: 55-56). In 1996, none of the hundreds of workers residing at the Lunnan oil station was Uyghur. Its Han manager, Sun Tairong, attributed this to the 'poor education level' of Uyghurs, but also to the fact that he spoke no Uyghur while most locals spoke no Chinese (Pocha 2006). Han enthusiasm to come to Xinjiang to exploit oil fields and mines has done much to exacerbate local ethnic tensions (Gladney 2003: 462). Han oil workers tend to have less contact with—and thus less cultural understanding of—local Uyghur farmers than Han farmers who reside long-term in the rural community. For these reasons, disputes over the use of natural resources are often 'intensified by nationalistic sentiments', so that managers are obliged to educate workers in nationalities policies in order to avoid conflicts (Ma 2003: 121).

Han in-migration has impacted on other types of resources too; for example, the rising population density in Xinjiang has increased the burden on food production. In 1966, the region produced 400 kg of grain per capita, but already in 1980 this figure had shrunk to 301 kg per capita, so that an additional 100,000 tons of grain had to be imported (Heberer 1989: 94). Lastly, there is the controversy surrounding commercial development and use of local cultural resources, including historical and religious sites. Rahilä Dawut writes that in some places in Turpan and Khotän, profits made from shrine (pilgrimage) tourism are going directly into the pockets of Han-owned tourism companies and not to Uyghur townships and villages, with the result that local people feel both cheated and marginalised. To make matters worse, in some locations, locals have even been charged entrance fees to enter their own shrines to carry out Sufi prayers and other rituals (Dawut 2007: 155-57).

In my interviews conducted in 1995-6, a prevalent local opinion was that, despite promises to help Xinjiang develop and modernise, China had failed to re-invest the profits gained from resource exploitation back into the region. A female merchant based in Ürümchi, who engaged in cross-border trade with Russia, spoke with passion, anger and resentment in 1996 about how Uyghurs 'didn't have their own country', adding that Han had 'spirited away Xinjiang's rich natural deposits.'Her husband, a teacher, called the

#### 68 chapter one

Han a 'hateful race of people' and pointed to their 'dirty habits' (see Chapter 2 on stereotypes) before going on to assert: 'Every day they take oil off to China proper.' Dilbär, the young wife of a research fellow in Ürümchi, drew attention in 1995 to enduring poverty in the region, and linked this to the state's failure to plough revenues from resource extraction back into local Uyghur communities: 'There are many poor people in south Xinjiang. We have oil, gold, so many raw deposits. But we don't see the benefit of them. The Hans control them.' Pro-independence sympathies during this period indeed appeared to be strongly linked to Uyghurs' desire to gain control of the region's natural resources. Turghun, an Ürümchi chef in his fifties, argued in 1996 that if only Uyghurs could gain political control of the region and its plentiful minerals (he listed gold, oil, coal, silver and iron ore), every Uyghur house would have a car and a telephone. Speaking at an Ürümchi outdoor market in the same year, a 20 year-old petty entrepreneur from Aqsu told me excitedly: 'Do you know that if just one day's worth of [oil] cargo to China proper were stopped, it would be sufficient to supply Uyghurs in Xinjiang for three years?'38 Several others had gathered round to listen, and all nodded in agreement. The young man then continued with his speech, insisting that Uyghurs 'had to win independence within the next few years, or else in twenty years there would be no natural resources left.' For Ghäyrät, another market trader, the future of Han people in Xinjiang depended on the evolution of their economic policies, particularly those concerning natural resources. He told me in 1996: 'They can stay if they help us to develop; but the ones who just squeeze money out of Xinjiang and steal our raw materials can leave!'

Respondents also claimed that the growth of the Han migrant population had resulted in price rises of food and other commodities.39 Two female respondents based in Ürümchi, Räwiä, a linguist employed by a state work unit, and Tajigül, the young wife of a restaurant owner, cited the example of meat. Räwiä explained in 1995 how Han people in Xinjiang had gradually become used to eating mutton, and as a result purchased it in increasing volume, causing the market price to rise dramatically. Consequently, she remarked ironically, people on a low income could no longer afford to buy mutton: 'Uyghurs sell kebabs … but the Hans eat them!'

<sup>38</sup> The respondent claimed that this information had been learned from economics students based at Xinjiang University.

<sup>39</sup> Local media reported in 1997 that there had been complaints about the soaring price of vegetables. *Xinjiang ribao*, Ürümchi, 1 March 1997 in SWB (Asia Pacific), 18 March 1997, FE/2870 G/9.

A final adverse effect of regional development and increased population density has been resource degradation. In the past, Chinese policymakers rarely considered the environmental effects of an increased population, increased farming, artificial irrigation, mining or industrial activity, either in Xinjiang or in China proper. In Xinjiang, increased population density drove local people to fell desert vegetation for fuel, leading to severe deforestation.40 According to Han scholar Xia Ri, writing in 1983, 73.2 per cent of peasants in southern Xinjiang lacked firewood for between three and six months of the year, meaning that many households could cook only one meal per day in the summer (cited in Ren and Yuan 2003: 103). 'New' oasis agricultural systems were constructed, involving large irrigation projects which took control of Xinjiang's rivers in order to provide reservoirs, irrigation and drainage to Han-populated 'New Towns' like Shihezi (population density = 491 persons per square kilometre, compared with 7 persons per square kilometre in other arid zones). The same attention was not however paid to the 'old' natural oasis areas inhabited by minority nationalities. As a result, whenever Han migrants moved into fragile border areas of these Old Towns, water supplies were depleted, accelerating the process of desertification (Ren and Yuan 2003: 102). Further problems were created by mismanagement of scarce water resources. For example, water from the upper reaches of rivers was used without being monitored, while reservoirs retained water from the middle reaches. Consequently, water levels dropped in the lower reaches, water quality deteriorated, and in some cases rivers dried up completely (Ren and Yuan 2003: 102-103). The regional capital Ürümchi has itself encountered serious problems of water supply, heating supply and environmental pollution in the wake of rapid urbanisation (Li 1990: 70). When I lived in the city in 1995-6, the water supply and sometimes also the electricity supply would be turned off for two to three hours per day in order to save water and energy resources. More recently, concerns have emerged over the impacts of global climate change on the region, namely, that glaciers in the Tian, Kunlun and Altai mountains are melting, raising the spectre that downstream city residents may eventually run out of drinking supplies. The 3,800-metre Ürümchi No.1 glacier has lost more than 20 per cent of its volume since 1962. Meanwhile, over-

<sup>40</sup> Similar effects have been noted in Inner Mongolia, where demographic change resulting from Han in-migration has resulted in reclamation of grassland for agricultural purposes, leading to deforestation and desertification. As noted by Naran Bilik: 'This ignorance of the special environmental features of minority regions and the consequent upset of the ecological balance are another cause of ethnic tension' (1996: 142-143).

#### 70 chapter one

exploitation of river systems and oases has exacerbated the problem, with the volume of water in the region's once vast Aibi lake decreasing by twothirds over the past 50 years (Watts 2009). Finally, the development of domestic tourism threatens to exert severe environmental impacts on sites of natural beauty. In 2000, tourist facilities at the Heavenly Lake, formerly inhabited by nomadic herdsmen, included a chair lift, a shopping centre, a food court, speed boat rides and several Chinese-style temples and pavilions, while the lake itself was suffering from siltification (Sines 2002: 15).

# 1.7. *Political Representation*

Finally, though it is not my intention to tackle the subject of regional autonomy and political leadership in Xinjiang (for this, see Bovingdon 2004a), I want to briefly reflect on the effects of excluding Uyghurs from meaningful political participation. The Uyghurs' right to decide their own political future was in fact taken from them long before their incorporation within the PRC in 1949. In 1928, a young Chinese Communist Party had reaffirmed Sun Zhongshan's vision of a 'free federation', composed of China proper, Mongolia, Xinjiang and Tibet, and had extended this in 1931 to grant minority nationalities *the right to self-determination or separation* (Dreyer 1968: 97; Pye 1975: 494; Grunfeld 1985: 57). Yet these rights disappeared from CCP discourses after 1936. From then on, Mao spoke of how the Mongolian, Mohammedan and Tibetan peoples would 'automatically become part of the Chinese federation *at their own will*' (Grunfeld 1985: 57, my emphasis). This somewhat contradictory position was repeated in the Common Programme of 1949 (Dreyer 1968: 97), but by 1952 the message published in the government mouthpiece *People's Daily* was clear:

… any national movement which seeks separation from the Chinese People's Republic for independence will be reactionary […] it would undermine the interests of the various races and particularly *the foremost majority of the race considered* (Grunfeld 1985: 58, my emphasis).

In other words, minority secession was no longer considered to be in the interests of the Han majority. This message was codified in the PRC Constitution of 1954, which declared that the national autonomous areas were 'inalienable parts of China' (Grunfeld 1985: 58). In Grunfeld's view, the political U-turn reflected state anxiety, deriving from external factors including Soviet designs in Xinjiang, Japanese alliances with the Manchus and Mongols during the Sino-Japanese war, US support for the Guomindang,

the establishment of US military bases in Korea, Japan and Taiwan, and American-sponsored clandestine activities along China's southern border (1985: 60).41 Pye similarly pointed to concerns regarding external threat, but looked further back into Chinese history to suggest an almost primordial attachment to the land, developed in response to foreign aggression:

As the Communists began governing, the issue of the 'national minorities' had to yield to the spirit of Chinese nationalism, first during the Japanese war and second during the confrontation with the Soviet Union. For reasons which spring from deep within the Chinese spirit and which were reinforced during the era of Western encroachment [...] the Han Chinese have developed a powerful sense of territoriality (1975: 489).

In his view, this visceral fear of losing ground had been exacerbated by the Nationalists' failure to uphold Chinese claims to Outer Mongolia, and by negative encounters had in minority areas during the Long March (Pye 1975: 494-5). For whatever reasons, in the early 1950s the CCP felt it necessary to introduce a 'steel framework of administration' made up of Han cadres into the newly established Inner Mongolian Autonomous Region (1947), Xinjiang Uyghur Autonomous Region (1955) and Tibet Autonomous Region (1962) (Pye 1975: 497). At the same time, it trained a new generation of 'revolutionary minority cadres' at the Central Institute for Nationalities, established in Beijing in 1951. There, students on average trained for five years before returning to their people to provide 'appropriate leadership' (Pye 1975: 503-4).42

The pattern of mutual mistrust was established when minority individuals first attempted to voice objections to government policy during the Hundred Flowers movement of May 1956. This movement, inspired by the relaxation of control in the Soviet Union that accompanied Khrushchev's denunciation of Stalin,43 invited criticism of the Chinese Communist

<sup>41</sup> This anxiety endured as late as the 1970s. See Esposito (1977: 164) on the military doctrine taught to Ordinary, Basic and Armed militia units in Xinjiang at that time, involving a preparedness for war against US imperialism and 'social imperialism' (i.e. 'Soviet revisionism').

<sup>42</sup> Before the Cultural Revolution (which shut the Institute down for 6 years), average enrolment was around 2800 students; in the mid-seventies, 1200 students were enrolled, of whom 30 per cent were CCP members and 70 per cent Communist Youth League members (Pye 1975: 503-4).

<sup>43</sup> In a 'secret speech' to the 20th Congress of the Communist Party on 25th February 1956, Nikita Khrushchev had denounced the regime of 'suspicion, fear, and terror' built up under former dictator Joseph Stalin, and stated his wish to break the 'Stalin cult' of the previous thirty years. This speech was not made public until 18th March 1956, and then only in Belgrade and Washington. In Eastern Europe, it led to 'de-Stalinisation' and raised

#### 72 chapter one

Party's policies through the famous slogan 'Let a hundred flowers bloom, and a hundred schools of thought contend.' When minority intellectuals took the opportunity to speak up against the imposition of the Han language and the policy of population transfer, their opinions were not taken on board. Rather, they were met with a 'rectification campaign' designed to 'correct tendencies towards local nationalism' (Dreyer 1968: 98). In the early years of the Cultural Revolution—when minority languages and cultures were denounced as belonging to the 'Four Olds' (old customs, old culture, old habits, old ideas)—central mistrust of the peripheral peoples swelled once again, this time reflected in a reduction of minority cadres in Xinjiang from 60 to 40 per cent of the total (Grunfeld 1985: 64).

As this pattern suggests, central to the state's reluctance to grant minority peoples political empowerment are the twin issues of political reliability (in relation to the CCP) and national loyalty (to the PRC). Sautman observed the 'truism' that in China party heads are more powerful than state officials (1998: 94). At the political centre, Beijing, there are no minority nationality members within the CCP Politburo, while in minority areas Party Secretaries are only rarely chosen from among the minority nationalities. While all three Party Secretaries in Turpan were minority individuals during the relatively conciliatory period between the end of the Cultural Revolution and 1991, since that time—and probably in reaction to the 1990 Baren riots in south Xinjiang—the Party Secretary, county secretaries and staff in party organisation departments have all been Han. This has led to local perceptions of an autonomy that is purely lip service: 'all talk no action' (Sautman 1998: 95). In other multi-ethnic societies such a superficial form of 'autonomous status' has led to ethnic unrest. For example, in Aceh, most Acehnese feel betrayed by the integration of their territory into the Indonesian Republic, negotiated by break-away forces within PUSA (the All-Aceh Religious Scholars' Association), and consider their special status a 'farce' (Siapno 2002: 163). In Xinjiang, the political imbalance is visible not only in leadership positions; minority individuals are underrepresented in the party ranks as a whole. In Qäshqär in the midnineties, over 50 per cent of Party committee members were Han or Hui, although these groups made up only 19 per cent of the local population (Sautman 1998: 95). At the turn of the century, Uyghurs remained considerably less well represented in the Party than Han, comprising just 37.37 per

expectations of change, particularly in Poland and Hungary. The full text of the speech was published in Russia only in 1988.

cent of CCP cadres in Xinjiang, compared with their 61.58 per cent share of the regional population (Mackerras 2001: 290).

Minority nationalities also lack representation within certain sectors of the state administration. Here, it is not simply a question of political and national loyalty but also one of perceived competence. Sautman cites a scholarly discourse which appeared in a 1992 issue of the Chinese journal *Lingdao kexue* [Leadership Science]. In it, minority cadres were negatively stereotyped as retaining the economic, cultural and religious 'backwardness' of their cultures, and therefore being slow to 'mature' (Sautman 1998: 95). The same author observed in Turpan during the mid-nineties that financial and economic departments tended to be composed of mainly Han, while ethnic and religious affairs departments were made up of mainly minority persons. This pattern was acknowledged obliquely by one Han official, who described it as a matter of 'natural form'; the assumption was that Han cadres are better suited to handling modern institutions such as banking, enterprise and economic planning, while minority cadres are better suited to managing traditional institutions such as customs, habits and religion (Sautman 1998: 96). Similar to their limited representation in the CCP, at the end of the nineties minorities comprised a far smaller proportion (48.83%) of state administrative cadres than warranted by their 61.58 per cent share of the regional population (Mackerras 2001: 290).

Grunfeld earlier predicted that ethnic equality would remain 'an elusive goal' as long as minority cadres were unable to acquire positions of power within either the CCP or the government. He further observed that ethnic equality would not be possible until the economic conditions of the minorities approximated those of the Han (Grunfeld 1985: 67). The two issues are of course connected: in order for Uyghurs to have an influence upon regional policies affecting their socio-economic status (i.e. policies governing language use, medium of education, employment legislation, wealth distribution, economic development and resource exploitation), they must be permitted—and trusted—to occupy key positions in the regional Party organisation and administrative structures. They must also be free to air their views about the policies that affect them without being accused of acting in an 'unpatriotic' manner. Yet the response to one of Yee's survey questions in 2000 suggests that the reality is otherwise: asked whether Xinjiang was a part of China, 47 per cent of Uyghur respondents strongly agreed, while 16.7 per cent said that they were not sure. Only 2.1 per cent of respondents dared to openly disagree with the statement (2003: 439). Given that a far higher proportion of Uyghurs could be expected, on a

#### 74 chapter one

level playing ground, to qualify that statement—whether this involved advocating outright Xinjiang independence or a gradual move towards a multi-nation, federal system—the result reflects deep political apprehensions: an understanding among individuals that they must conceal their true opinions if they are to save their skins. If Uyghur political life continues to be stifled in this way, the Chinese state may eventually face a crisis of resentment of the type that periodically arises in the Middle East and North Africa. Were that to occur, it would be the direct result of 'the unwillingness of regimes to discuss the small-scale discontents and everyday problems of their people' (Anderson 1997: 21).

# *Conclusion: Development – Equality = Conflict*

There is considerable evidence that first-generation Han settlers who came to Xinjiang in the fifties and sixties made a genuine effort to learn the Uyghur language and to adapt to local Uyghur customs. Particularly in south Xinjiang, early Han settlers evidently constituted such a tiny minority of the local population that they were obliged to respect local linguistic and cultural norms (including the avoidance of pork), and taught their children to do the same. As a result, some children born to original Han settlers grew up speaking the Uyghur language fluently. In 1996, I met a Han male in Aqsu who had been raised in the south, spoke perfect Uyghur, and appeared to maintain comparatively good relationships with the Uyghurs around him. That same year in Khotän New Town, I came across several Han post office employees in their thirties and forties, who spoke Chinese with a heavy sing-song lilt, adding *ya* onto the end of each sentence in the style of the Khotänese dialect. In the local hospital, a Uyghur doctor spoke warmly of the Han nurse with whom he worked, explaining that she 'had been born and grew up in Qorla.' In 1996, I also came to know a young woman of mixed parentage (one Uyghur and one Han parent) named Liu Lan. Despite taking her father's Chinese name and speaking Chinese as her first language, she told me that she (and indeed her whole family) adhered to Muslim dietary prescriptions: 'No-one eats pork in Khotän, really. Everyone eats *qingzhen* (halal). Only Hans who have just arrived eat pork, when they're not used to eating mutton.'

This image of mutual respect and accommodation contrasted sharply with respondents' descriptions of ethnic relations in the 1990s. It was clear that with the steady increase in Han settlers, relations between locals and migrants had greatly deteriorated. Räwiä, a foreign language specialist in her thirties interviewed in 1996, compared her experience of growing up in Ürümchi in the sixties and seventies with the situation in the mid-nineties:

When I was little, there was none of this 'minority nationalities are inseparable from the Han nationality' propaganda. It wasn't necessary because everyone *was* united then. We used to play with Han children. Now, you rarely see Han and Uyghur children playing together.

Her description closely corresponded to the picture painted by Salamät, a health professional in her thirties interviewed in the same year:


In this way, both women pointed to the changed population composition and related effects as the root of worsening Uyghur-Han relations over the past two decades. It is instructive that fresh waves of Han migrants in the 1990s have also angered some of Xinjiang's first-generation Han residents, who consider themselves *lao Xinjiangren* or 'old Xinjiang folk'. For instance, elderly Han farmers have claimed that the new arrivals have destroyed the *modus vivendi* they had developed with local Uyghurs over the years, while at the same time increasing competition for water, fertilisers and seeds (Becquelin 2000: 85).

Some scholars writing in the first half of the 2000s suggested that the relative quiet in the region since the 1997 Ghulja disturbances signalled the success of the state's development policies. Dru Gladney, while acknowledging that increased Han participation in Xinjiang's oil and mining industries continued to exacerbate ethnic tensions, concluded that the government's bid to buy support through stimulating the local economy 'seemed to be working' (2003: 462). Colin Mackerras offered a similarly upbeat analysis in proposing that the Western development strategy was 'broadly succeeding in creating wealth' (2005: 16). I would question both

#### 76 chapter one

viewpoints, for they ignore the fact of ethnic differentiation. For whom is economic development working? And for whom is it creating wealth? While the standard of living in Xinjiang may have been raised *in general*, this will be insufficient to satisfy local groups who are keenly aware that Han settlers in the urban north have benefited from development *to a far greater degree*. In South Xinjiang (where Uyghurs are most populous), poverty remains severe, and this surely accounts for the radicalism that Mackerras and others have connected to this part of the region. Although a small middle class of independent Uyghur businessmen has emerged in the era of the free market economy, the majority of Uyghurs are certainly not enjoying a 'more-or-less prosperous life' (Mackerras 2005: 18). I would rather support Mackerras's second proposition: that the growth of the regional economy may have acted to heighten local resentment 'because the Uyghurs may have considered that the Han Chinese were battening on what should have been their land' (2001: 301). This being the case, the state would be wise to recognise that the bestowal of development *without equal empowerment* may well lead to revolt as it did in other, overtly colonial contexts during the twentieth century (cf. Harrell 1994: 35). This argument is also supported by some intellectuals interviewed by Yee's research team, who argued that the roots of Uyghur separatism were 'complicated' and suggested that rather than nationalism *per se*, 'policy mistakes' and 'cadre incompetence' had provoked separatism in Xinjiang (Yee 2003: 444). Yee for his part identifies 'bad', 'dated' and 'ill-advised' minority policies as a key factor straining Uyghur-Han relations (2003: 450; 452).

At the core of this collection of ill-advised minority policies, we find a constant pattern whereby Han settlers to Xinjiang have continuously reaped privileges derived from their 'close(r) connection to the main organs of political and economic power' (Côté 2011: 1856). This has included the compliance of the centre in allowing settlers to impose the Chinese language and displace the dominant language of the receiving community (Uyghur); the opening of the most prosperous economic sectors (e.g. oil) to Han settlers at the expense of local Uyghurs; the fostering of a state discourse that justifies the presence of Han Chinese in peripheral areas; the keen responsiveness of the centre to the periodic demands made by Han settlers; and the centre's reluctance to use repressive force to deal with Han mobilisation (the most recent example being the Ürümchi riots of 2009) (Côté 2011: 1856-57; 1860-61; 1866). As Isabelle Côté adroitly observes, in all these aspects, Han settlers to Xinjiang may be easily compared with Russians who migrated to the outlying Central Asian republics, with *pieds* 

*noirs* who migrated to the French colony of Algeria,44 or with Jewish settlers living on the West Bank (2011: 1957).

The Uyghur people will not come to identify with the Chinese state until—and unless—they are given a true stake in the nation's fate. In a 'multi-nation state' (Kymlicka 2011) such as China, the ruling elite will not gain the cooperation of its peripheral citizens through threat of force (cf. Blum 2001: 7); it should rather encourage them to belong by offering them linguistic, cultural and socio-economic equality alongside the majority nationality. Indeed, it might consider adopting the strategy it has pursued towards Taiwan since 2001, i.e. promoting mutually beneficial economic transactions and collaborations between Han and Uyghurs (rather than ones that serve only the majority group). A dependence on the stick at the expense of the carrot will only result in unrest as popular violence reacts to state violence. At the same time, the carrot can only gain mass allegiance if it is offered equally to all citizens. As Kramer suggested in 1996 in the context of the 'new Middle East', success depends upon resolute pragmatism, a solution based on peace, development and democratisation, on a promise to feed, house, and employ masses of people. Such an economic pragmatism will only work if it brings swift gratification, 'because it speaks to interests, not identity' (Kramer 1996: 285). In the case of Xinjiang, gratification must be not only swift but also fairly distributed; if the majority Han continue to be seen as the preferred benefactors, then there can be no peace and understanding in the region. On the other hand, should Han settlers in Xinjiang perceive the introduction of policies intended to equalise opportunities among the region's diverse ethnic groups as leading to a potential loss of Han dominance and privilege, this could equally lead to unrest from a different quarter (cf. Côté 2011: 1864).

<sup>44</sup> The term *pieds noirs* denotes people of French descent born or living in the French colony of Algeria, and may refer to the black boots or shoes worn by French settlers in contrast to the typically bare-foot Algerian peasants. Forced to flee Algeria following France's retreat from the war of Algerian independence in 1962, many *pieds noirs* felt abandoned by French policy and homesick for Algeria.

chapter one

PART TWO

inequalities: the socio-economic backdrop 79

SOUNDS OF DISCONTENT

chapter one

#### Chapter two

stereotypes: talk breeds conflict 81

# StereotypeS: talk breedS ConfliCt

the period from the disintegration of the USSr in 1991 to the Ghulja disturbances of 1997 may be conceptualised as a run-up to an 'imagined independence': a period of sharp growth in ethno-nationalist sentiment, fed by a heartfelt belief among many young Uyghur men that domestic and international politics would soon converge to bring Xinjiang a chance for independence. Some said that tensions between China and america in the taiwan straits (1995-96) would result in Sino-US military conflict. others argued that the United kingdom had no intention of allowing the peaceful return of hong kong to China (1997) and that an anglo-Chinese war would surely ensue. Still others believed that the imminent death of CCp helmsman, deng Xiaoping, and the transition to a new generation of party leaders would throw the Chinese nation into political chaos, creating a chance for restless peripheries to break away. this was a period punctuated by the sounds of discontent. in Chapters two, three and four, i consider the three main forms of symbolic resistance deployed by the Uyghurs during this period: stereotypes and counter-stereotypes; ethnic boundaries; and alternative representations in popular song. the first form involved the daily oral repetition of negative ethnic stereotypes of han people, even as Uyghurs simultaneously circulated positive counter-stereotypes of the ingroup. writing in 1976, June t. dreyer observed of turkic Muslim groups in Xinjiang: 'their capacity for effective cooperation […] has generally been shown less in a positive sense than in opposition to a common enemy, such as han colonization' (1976: 22). the process of unpacking the stereotypes which coloured this period confirms this assertion in that the positive values Uyghurs assigned to their own group were frequently conceived in relation to the negative values they associated with the han (cf. Vasquez and wetzel 2009: 1569). at the same time, however, i will show how those positive values were not created or imagined, but based firmly on centuries of shared daily practices, derived mainly—although not exclusively—from islam. i will also demonstrate how negative stereotypes directed at the dominant out-group served to alleviate feelings of powerlessness and resignation, providing a powerful sense of agency, and allowing Uyghurs 'the

#### 82 chapter two

symbolic revenge of the downtrodden' (Eriksen 1993: 24). I conclude with the proposition that 'talk'—or the sounds of discontent—did much to breed inter-group (Uyghur-Han) hostility, segregation and conflict during this period; furthermore, it allowed ordinary Uyghurs to rise above the 'inferiority complex' that had characterised earlier generations and begin the process of building a positive and dignified self-image. Counter-stereotypes in this sense became 'the rhetorical and strategic tools deployed […] in reaction to perceived stigmatisation, racism and discrimination', or a means through which to 'stigmatise the stigmatisers' (Lamont and Mizrachi 2012: 366, 373).

# 2.1. *Social Stereotypes: Theoretical Frame*

As defined by Cinnirella, stereotypes are belief systems which associate attitudes, behaviours and personality characteristics with members of a particular social category or group (1997: 37). They may include 'active stereotypes' (circulating in society and endorsed by particular groups) and 'dormant stereotypes' (indirectly disseminated via the mass media). Since some groups may have greater access to mass communication technology than others and are better placed to diffuse stereotypical images, the latter type invokes the question of power (Cinnirella 1997: 40-41). Stereotypes can exist in isolated form—'individual stereotypes'—and help the individual to make sense of the social world (furnish him or her with 'predictive power'). Alternatively, they may be communicated, shared and endorsed by members of the in-group, thus becoming 'shared' or 'social' stereotypes with the potential to create prejudice and inter-group conflict (Cinnirella 1997: 38). Where individuals have not 'shared' stereotypes with their peers, they often rely exclusively on dormant stereotypes, internalising massmediated images that later come to influence face-to-face interactions (Cinnirella 1997: 41-42). In extreme cases, social stereotypes can 'numb our consciences and tempt us to succumb to violent and destructive impulses' (Barfoot 1997: 20).

Henri Tajfel viewed social stereotypes as involved in a 'cyclical model of causality' in which they both cause and reflect inter-group relations (Tajfel 1981: 157-61). Advocating the study of their *contents* as a means of understanding those relations, he argued that stereotypes serve three main social functions:


While the first two functions immediately suggest the situational variability of social stereotypes across time, it is important to acknowledge that these variations take place in the context of a pre-existing core set of beliefs—what Schültz (1972) called the *Relevanzstruktur* or 'structure of relevance'. To a significant extent, members of a particular social group will behave in similar ways and espouse similar attitudes in a variety of contexts. Following Cinnirella (1997: 44-47), I argue below that Tajfel's theory of social stereotypes can be directly applied to instances of ethnic and national stereotyping, and illustrate the point with reference to the Uyghur case.

Studies have shown that ethnic and national stereotypes impact on the individual from an early age, with the family unit mediating a child's perceptions and providing him or her with one of many available (and often contested) sets of beliefs related to other ethnic and national groups (Cinnirella 1997: 50). Stereotypes are also inculcated in the young through teaching materials employed in primary and secondary schools as, for example, when standard representations of 'national types' are disseminated in foreign language textbooks (Barfoot 1997: 11-12). In Xinjiang, contemporary history textbooks construct and reproduce the official state narrative on Xinjiang: that Uyghur people have been an inseparable part of a harmonious Chinese nation since antiquity, and will always be so (Feyel, 2011). Yet at the same time as endorsing and internalising stereotypes of the outgroup, the in-group may reject the stereotypes imposed upon it, instead building alternative versions of their social identity in order to redefine the basis upon which they are judged (Cinnirella 1997: 41; 49). This case study will demonstrate how negative stereotypes of a monolithic Han 'Other', articulated by and among Uyghur adults in the family setting and more broadly in the Uyghur-language domain, exerted a high level of 'community pressure' on in-group members not to interact with Han people unless

#### 84 chapter two

absolutely necessary. This dynamic created a particular social identity among young Uyghurs that would prove to be highly resistant to change, and led to strained inter-ethnic relations and an increased level of segregation in Xinjiang in the 1990s (see Chapter 3). At the same time, however, I will illustrate how the deployment of positive counter-stereotypes, founded on a set of superior values derived from Islamic morality, functioned to build in-group esteem and cement a positive Uyghur ethnic identity (cf. Vasquez and Wetzel 2009: 1557-1558). This in turn served to fuel ethnonationalist aspirations during a post-Soviet period of change in which almost anything seemed possible.

# 2.2. *Theorising Han Chauvinism: The 'Civilising Project'*

As acknowledged above, the counter-stereotypes deployed by Uyghurs during this period were constructed at least partly in relation to negative stereotypes imposed on them by the Han. We cannot and should not overlook the fact that 'the attitudes of the dominant group towards the dominated are extremely significant for understanding relationships between groups' (Blum 2001: 12). In other parts of the world, racial issues have featured prominently in majority-minority discourses—notably in the Americas; yet these have been less in evidence in China. One suggested explanation is that if we accept that different degrees of civilisation are truly genetic and 'inborn', then China's civilising project ceases to make sense, since it becomes impossible to 'civilise' peripheral peoples over time (Harrell 1994: 15). Stereotypes typically employed by Han people to describe Uyghurs can be usefully framed with reference to the concept of *sino-centrism*. The ancient Confucians sought to classify peoples as closer to or more distant from the 'centre' (the Chinese imperial court) based on how much culture (*wenhua*) they were thought to command, where 'culture' is defined in Confucian terms (Harrell 1994: 4). The centre saw its relationship with the rest of the world as a process of transforming it from raw and untutored to fully civilised:

The inequality between the civilizing center and the peripheral peoples has its ideological basis in the center's claim to a superior degree of civilization […] the process of domination is one of helping the dominated to attain or at least approach the superior cultural, religious, and moral qualities characteristic of the center itself (Harrell 1994: 9).

This 'culturalist' way of thinking permeated not only the imperial court but also the consciousness of the common people (Harrell 1994: 20). It is documented that all strata of Tang society constructed the 'barbarian' as 'an antipode to accepted norms', frequently dehumanising foreigners as animals: stupid, simple and cruel (Abramson 2003: 146). Qing rulers made attempts to 'civilise' barbarians by pressuring Muslim subjects to accept the Confucian principle of *xiao* (filial piety) and to participate in the court ceremonies of the ancestral shrines (Israeli, 1984: 276). In everyday life, Muslims were described using the disrespectful term *Huizi* rather than the respectful alternative *Lao hui*;1 and Islam was called *xiao jiao* (the Small Religion) in contrast to Confucianism, which was known as *da jiao* (the Big Religion) (Israeli 1984: 284). Needless to say, the assumption through history of Han cultural superiority has not necessarily been shared by non-Han peoples; there are many examples of non-Han groups forming cultural and political alliances which excluded the Han, as in the case of Manchu-Mongol collaboration during the Qing dynasty (Grunfeld 1985: 57).

Following the fall of imperial China, the sino-centric mindset described above was supposed to dissipate. The revolutionary ideals of the Chinese Republic (1912- ) and later the Chinese Communist Party (1921- ) advocated a 'progressive' approach, which valued the concept of national (ethnic) equality. Sadly, these projects were stymied—'co-opted', to use Harrell's (1994: 25) term—by residual Confucian beliefs. The Five-Coloured Flag (*wuseqi*) initially adopted as the national flag of the Republic of China in 1912 featured five horizontal stripes representing five races under one union. However, the visual design seemed to embody precisely the social hierarchy the revolution strove to replace, and for this reason was not favoured by Sun Yat-sen. The flag consisted of a red stripe (representing the Han) at the top, followed by yellow (Manchu), blue (Mongol) and white (Hui) stripes, and finally a black stripe (representing the Tibetans) at the bottom. Meanwhile, the various linguistic terms denoting 'barbarian' survived until the 1930s, when the Guomindang (GMD, Chinese Nationalist Party) was forced to flee to rural areas dominated by minority nationalities

<sup>1</sup> *Hui* or *Huihui* was the generic term for a Muslim in China during the Ming and Qing Dynasties. It is thought to originate from the earlier *Huihe* (回纥) or *Huihu* (回纥), the name given to the steppe Uyghur kingdom (744-840) (Lattimore 1950; Gladney 1996; Lipman 1997).

and thus found it prudent to introduce more neutral terms like *zhongzu* (ethnic group) and *renzhong* (race) (Dreyer 1976: 18).

In the PRC period (1949- ), the notion of national (or nationality) equality has been fundamentally undermined by the theory of the 'five stages of history', in which the Han rank highest on the scale. This theory continues to be taught as scientific fact in the national school curriculum, suggesting that Han prejudice against minority peoples remains ingrained among policy makers:

As long as such an innate, almost visceral Han sense of superiority remains, the actual program of the Communist project will be based on the unconscious assumption that Han ways are better, more modern ways. Peripheral peoples who act like Han—who are educated, Hanophone, *cultured*—will be treated equally with their Han compatriots (Harrell 1994: 25).

In this context, minority individuals must accept sinicisation if they seek to win Han respect and attain social equality. These attitudes of the dominant are of course reminiscent of the chauvinism that prevailed among Western colonisers of Asia and Africa in the 1800s. Marsot documents how 'high-handed and superior attitudes' among the British administration in the Indian sub-continent were partially responsible for the Sepoy mutiny of 1857 (1992: 160-61). Similarly, Anderson remarks how Western interest in the Middle East and North Africa 'was more a function of [that territory's] physical than its human or cultural resources' (1997: 26). In the latter context, lack of interest in, and respect for, the peoples and cultures of the Islamic world produced local resentments against Westerners and allowed Islamists to successfully mobilise popular resistance.

For their part, Chinese Communist politicians have periodically acknowledged the dangers of Great Han chauvinism (*Da hanzu zhuyi*)—attitudes of Han cultural superiority. Yet even as they did so, their written and spoken discourses were awash with evaluative statements which differentiated between Han and minority nationality groups. Speaking on 23rd October, 1951, Premier Zhou Enlai said: 'Appropriate reforms within a nationality are a necessary stage through which it must pass *in order to develop and progress and reach the level of the more advanced nationalities*  …' (Chinese Communist Party 1953: 25) [my emphasis]. To give another example, in the first 'Policy towards Nationalities' of 1953, Chairman of the Commission for Nationality Affairs, Li Weihan, emphasised the need to keep manifestations of Great Han chauvinism and 'local nationalism' (minority chauvinism) in check. Yet in the same breath he described the

'significant assistance' that the Han nationality would render in the 'future advancement of the fraternal nationalities' in the military, political, economic, and cultural fields (Chinese Communist Party 1953: 27). Han chauvinism played out both in a tendency towards 'commandism' and in the deeply embedded assumptions of superiority among cadres dispatched to the hinterland. Richard Bush records how soldiers of the Communist First Field Trip Army were cautioned to respect Muslim customs as they proceeded to the Northwest in 1949. The military order showed clear recognition of important boundaries:

The soldiers were ordered not to speak to Muslim women […] They were to refrain from eating pork, drinking wine, or smoking in the homes of Muslims […] Further, they were not to take baths at Muslim bathing places and were to wash their hands before drawing water at a Muslim well (Bush 1970: 268-9).

Despite these instructions, *Renmin ribao* [People's Daily] reported in 1953 that resistance in the Northwest was due to cadres' 'lack of respect for the religious belief of the national minorities, or for their customs, habits, languages, or writings' (Bush 1970: 271). It is easy to see how such attitudes could provoke continued local resistance against CCP activists in minority areas in the first years following the establishment of the PRC.

By 1957-8 (the first period of so-called 'left deviation'), the term 'Han chauvinism' had almost disappeared from state discourse, while 'local nationalism' was increasingly viewed as the truly dangerous tendency (Moseley, cited in Grunfeld 1985: 62-64). This shift would repeat each time there was a 'left deviation' in modern Chinese history. June T. Dreyer documents how the 'Han man's burden' of helping the 'backward fraternal nationalities' was fraught with problems, not least the need for Beijing to regularly admonish cadres for their arrogance (Dreyer 1968: 98). Rusticated (*xiafang*) Han youth exiled to Xinjiang during the Cultural Revolution—and described as unmarried and frustrated individuals—are said to have 'victimised Muslims, hurt their feelings, ridiculed their beliefs and customs and upset their institutions' (Shichor 1994: 77).

In the reform era (1978- ), reports suggest that Han chauvinism has continued and may even have worsened, in spite of an initial round of conciliatory state policies towards minority nationalities. In *From Heaven Lake*, an account of a journey through Xinjiang and Tibet in the 1980s, Chinesespeaking author Vikram Seth expresses amazement when he meets a Han named Sui who does *not* expound chauvinist views: 'Unlike almost all Han

#### 88 chapter two

Chinese I have met […] Sui […] has not once indicated any dislike of Lhasa or of Tibet or of Tibetans' (1993: 74). Uyghurs whom I interviewed in Xinjiang in the mid-nineties stated that Han chauvinist attitudes were increasingly widespread and had burgeoned in direct proportion to the growing Han population. Some respondents distinguished between firstgeneration Han settlers who came to Xinjiang in the fifties and sixties and more recent immigrants arriving in the eighties and nineties. They explained that the former had learned Uyghur (then still the regional lingua franca) and made cultural—including dietary—concessions in an environment where Han language and culture were marginalised. By contrast, the latter joined a large community of Han Chinese, whose members increasingly perceived that incomers need not make concessions to local cultural practices. In this context, respondents maintained that new Han migrants were frequently hostile towards them. Räwiä, a *minkaohan* language specialist in Ürümchi, told me in 1995: 'The original Han settlers were alright. We grew up with them side by side. They wouldn't say anything to hurt you. It's the ones who came later. They brought bad habits with them.' This phenomenon was reflected also in the regional media; *Xinjiang ribao* [Xinjiang Daily] reported in 1996 that there had been problems between new Han immigrants and local people because the former had not received a 'systematic education in the nationalities policy and the etiquette of nationalities.'2 In south Xinjiang, Qadir, a taxi driver in his forties from Khotän, observed in 1996 that the new settlers 'know no gratitude.'3

Recent field studies conducted by both Western and indigenous scholars suggest the continuation of this trend. Colin Mackerras, writing on the causes of separatism in Xinjiang at the turn of the century, recalls 'a Han friend, normally polite and well-mannered, who suddenly assumed the manner of an overlord when she came across a Uyghur crowd' (2001: 300). Jennifer Taynen describes an incident where three Han female students refused to eat lunch in a Uyghur restaurant on the grounds that mutton was 'unclean' (2006: 53); their refusal to compromise seems to have been a sarcastic imitation of the reaction that a Uyghur would have if asked to eat

<sup>2</sup> *Xinjiang ribao*, Ürümchi, 22 May 1996 in SWB (Asia Pacific), 13 June 1996, FE/2637 G/10.

<sup>3</sup> His comment is ironic in the context of Han reactions to the Ghulja disturbances of 1997. In the eyes of the state and many of its Han subjects, that event represented a precocious attempt to break away from the state patron. In its wake, a growing consensus formed among Han people that *all* Uyghurs (not just those who dreamed of independence) were 'ingrates' (Smith Finley 2011a: 81).

in a Han restaurant, but in the latter case this would have been based on legitimate religious proscriptions. Rahilä Dawut describes problems of Han tourists' insensitivity when visiting Uyghur *mazar* (shrines), citing the example of Qirmish Ata Mazar in Aqsu prefecture. Despite being asked to respect Islamic teachings while inside the shrine, such as prohibitions on exposed arms and legs, consumption of alcohol, speaking in a loud voice, spitting, dancing and playing music, she writes (diplomatically) that it has been 'difficult to ensure that every visitor obeys these rules' (2007: 159-60). Another example of contemporary chauvinism is found in the state's insistence that Uyghur names are reproduced on passports and other international documents using the Chinese transliteration system, and in reverse order (surname-name) in conformity with the Chinese naming tradition. The resulting discrepancy between the name on the document and that used by the individual in everyday life creates all manner of problems for Uyghurs who go to foreign countries to study, do business or visit relatives (Sulayman 2007: 118-120).

While the language of Great Han chauvinism may not be continuously in evidence, it quickly comes to the fore given a particular set of social or political circumstances. In a study of the ambiguous attitudes of Han people towards 'primitives' in the 1990s, Susan Blum writes that Chinese people do not all share a pure sense of Han nationalism but that some do, sometimes. She contends that they can be simultaneously open- and closed-minded, affectionate toward and dismissive of ethnic minorities, and that these contradictions spring from an inherited group consciousness: 'People talk for themselves and think for themselves, but they and their language are not created anew with each individual birth; they use *words and thought patterns that precede their own existence*' (2001: 11) [my emphasis]. In the wake of the Ghulja disturbances of February 1997, during which a number of Han civilians were attacked and killed on the streets after security forces opened fire on a peaceful Uyghur demonstration, negative stereotypes of minority groups were widely deployed. Following this event, the most serious instance of ethnic violence in Xinjiang since the Baren riots of 1990, a government official in Beijing was quoted as saying: '… ethnic minorities are like untamed mustangs. No matter how well you feed them, they may run off at any time.'4 The government subsequently embarked on its campaign to 'develop the West', apparently hold-

<sup>4</sup> *Ping kuo jih pao*, Hong Kong, 17th February 1997. In Summary of World Broadcasts (Asia Pacific), 18th February 1997, FE/2846 G/2.

#### 90 chapter two

ing firm to the belief that the roots of inter-ethnic conflict lay in regional economic inequalities, in particular, the gap in standard of living between East and West China. In the academic literature on the Great Western Development campaign, Han scholars discussed the thorny problem of some minority villages continuing to exhibit what they coined 'countryside disease' despite poverty alleviation measures. In conclusion, they warned that material advancement provided no guarantee that such people would not slip back into their 'innate backwardness' (Sines 2002: 9). While this statement evidently reflected urban-rural prejudice as much as majority-minority discrimination, it was noteworthy in appearing to subvert the long-standing belief of the centre that civilisation can be 'learned'. A similar reaction followed the Shaoguan factory fight and subsequent Ürümchi riots of July 2009. Han locals interviewed in Guangdong referred to Uyghurs' 'low level of civilisation' and their predatory behaviour towards young Han women, while Han residents in Ürümchi characterised Uyghurs as 'savages' and Muslim extremists (Smith Finley 2011a: 82-83).

# 2.3. *Different Ways of Seeing: Counter-stereotypes*

As suggested above, worsening inter-ethnic relations between Uyghurs and Han in the 1990s resulted at least in part from a growing reluctance among Han settlers to 'do in Rome as the Romans do'. Finding themselves the numerical majority in some urban centres (notably, Ürümchi and Aqsu), many assumed a sense of safety in numbers and began to expect Uyghurs and other local peoples to adapt to imported Han culture, the majority culture in the national context. The fact that in 1996 Han respect for Uyghur religio-cultural sensibilities remained relatively intact in Qäshqär, where Han people were still a minority, supports this theory. In the context of rapidly consolidating Han communities, identity symbols ascribed to, and presented positively by, Uyghurs have come to be viewed increasingly negatively by incoming 'Others'. Robert B. Ekvall noted an early example of this process in his 1939 study of cultural relations on the Gansu-Tibetan border. There, Han Chinese who came into increased contact with Tibetans considered the latter's mode of dress—long coats with no trousers—indecent (cited in Shibutani and Kwan 1976: 102). This type of ethnic prejudice can at length cause in-group members to develop an inferiority complex, so that ethnic stereotypes become 'self-fulfilling prophecies' (Eriksen 1993: 24). I have discussed the dynamic of 'internalised oppression' in evidence

in Xinjiang during the Cultural Revolution period elsewhere (Smith Finley 2007a).

Yet ethnic prejudice imposed from without may also have the opposite effect of fuelling pride among group members, leading to a strengthened ethnic identity and group incorporation. It is important to remember in this respect that notions about the virtues and vices of others can be held by dominated groups as well as dominating ones (Eriksen 1993: 23). Vasquez and Wetzel provide rich examples of the ways in which two marginalised groups in the US—Mexican Americans and Native Americans—establish their social worth and collective dignity by making discursive comparisons with, and drawing distinctions from, the American mainstream: 'Emphasizing an alternative set of values that they believe are absent from and indeed superior to white society enables group members to demonstrate their merit and validate their ways of being' (2009: 1557-1558, 1569). It is precisely this dynamic of positive representation via counter-stereotyping which I observed while conducting fieldwork in Ürümchi in 1995-6. While Uyghur resentment of Han chauvinism did occasionally explode in direct action or violence, it far more frequently took a verbal form. When I informed respondents that I was researching Uyghur culture, they generally asked two questions. The first was: 'Do you like the Uyghurs?' The second was: 'Do you like the Hans?' or 'What do you think of the Hans?' They thus began by testing my loyalties, in order to know what they could and could not say in my company. If I replied that I liked Han people and had Han friends in Beijing, respondents would nod doubtfully and become less friendly and open. If I said I *dis*liked Han people, respondents would immediately pour forth a long list of negative stereotypes of that group. Especially among Uyghur intellectuals, a sense of cultural competition permeated all their discourses on Han language and culture, which they consistently compared negatively with Uyghur language and culture. For example, Han failure to learn the Uyghur language was explained in terms of Han inability to grasp a more complex language, and contrasted negatively against the superior Uyghur intellect (cf. Bellér-Hann 2001a: 56-57). These counter-stereotypes were deployed by respondents to positively distinguish the in-group and to underline Uyghur cultural superiority, and had their foundations in a pre-existing common heritage.

#### 92 chapter two

### 2.4. *Common Heritage: Foundations of Counter-stereotypes*

In *Pivot of Asia*, the eminent Inner Asia scholar, Owen Lattimore, observed that the oasis peoples of Xinjiang had called themselves by localised hometown appellations (e.g. Turpanliq) for several centuries until the political call was made in 1921 for the name 'Uyghur' to be revived. That the name was re-adopted in 1934 and came back into use, he argued, indicated an increasing national consciousness (1950: 125-6). Yet this consciousness was only possible because the oasis dwellers *already possessed* a high level of social, cultural and religious homogeneity—one, moreover, that pre-dated the return of the ethnonym by many centuries. As Lattimore notes:

… the Uighurs—that is, the people who speak one of the many mutually intelligible dialects of the Uighur language—are all Moslems, all share the same literary and artistic heritage, live chiefly by agriculture, have a common history, form 95 per cent of the population in the Tarim Basin (where 90 per cent of them are concentrated), and in other respects show a high level of uniformity (1950: 119).

Lattimore even goes so far as to suggest that Uyghurs in the 1940s *already constituted* a 'viable nationality' in terms of national consciousness combined with a territorial base for existence (1950: 122). In a later socio-economic study, Warikoo contends that nineteenth-century Uyghurs were socially and spatially segregated by intervening deserts between oases, and that this left them 'unable to stand united under a common leader against alien rule' (1985: 92). Nonetheless, he too acknowledges that the oasis peoples shared a common religious, linguistic and historic heritage; in other words, the foundations on which to build a broad national identity were already in place. In fact, the oasis-dwelling peoples formed the main core of a settled indigenous population which had long contrasted itself against indigenous nomadic groups (Kazakh; Kyrgyz), non-Turkic groups (Mongols, Manchus), immigrant groups (Han; Hui) and even its culturally close Turkic neighbours (Andijanis, Khoqandis) (Newby 2007: 16-20). In the context of territorial contestation (revived by the collapse of the USSR), growing inter-cultural conflicts and sharpened socio-economic competition, I argue that Uyghurs in 1990s Xinjiang drew heavily upon this shared heritage in order to reposition the Uyghur ethnic group as culturally superior to its oppressors. In so doing, they evoked a widely recognised dynamic of national or group mobilisation wherein cultural values and norms are presented as 'unchanging indicators of the "essence" of the group' and couched in terms of an 'implicit moralising dichotomy' (Breger and Hill 1998: 9). In

such contexts, stereotypes used to draw distinctions between groups frequently simplify and ignore social diversity within the group, so that group characteristics are presented as if they are universal to all, inherent, 'natural', and therefore unchangeable (Breger and Hill 1998: 11). The notion of ethnic culture is thus 'homogenised' so that group members assume that cultural conventions such as religion, language, norms and expectations are the same for all, and completely accepted and practised by all, and that there is one 'pure' set of traditions with little internal variation between regions or social classes (Breger and Hill 1998: 9).

Certainly, Uyghurs like to distinguish between the unique features of people hailing from different oases, and these differences can sometimes evoke a certain sense of local rivalry. All Uyghurs prefer their particular hometown, and insist that it is better than the others. However, such assertions rarely take the form of an attack, and do not stem from some powerful ethnic sentiment. Uyghur attitudes towards Han migrants are by contrast characterised by disgust, anger, bitterness, passion, and a strong sense of injustice. In the mid-nineties, the vast majority of Uyghur respondents deployed three discrete sets of negative stereotypes in order to reject Han people as a monolithic, culturally inferior 'Other'. The first (comparatively mild) set related to Han failure to subscribe to Uyghur cultural norms that are not necessarily or directly connected to Islam. For instance, Uyghurs ridiculed the slight Han physique, characterised Hans as 'killjoys', and criticised the *Khänzu mijäz* [Han disposition]. The second set was based upon Han failure to conform to Islamic social laws deeply ingrained in Uyghur social life. Respondents expressed universal disgust at the Hans' indiscriminate diet, their modes of washing, their 'bad habits' and their 'inhumanity'. The third set—which evoked the most anger and bitterness—characterised Han people as invaders and exploiters. While this kind of stereotype was sometimes expressed in humorous form, it was more often delivered in vitriolic tones.

#### 2.5. *Stereotypes relating to Uyghur Cultural Norms*

#### *Effeminate Men*

Respondents often ridiculed Han people for their disinclination—relative to Uyghurs—to eat meat. For a Uyghur, meat is an essential and coveted part of the everyday diet, and a meal is considered incomplete, and the taste of food inferior, without it. Moreover, given that meat consumption

#### 94 chapter two

varies according to income in Xinjiang (Cesàro 2000: 226), consuming and *being seen to consume*—meat is an important class statement. This is also one reason why Uyghur hosts hope to entertain guests with a spread of 'prestigious foods', especially meat-based dishes (Bellér-Hann 2008: 205). Some respondents laughed behind the backs of Han colleagues, claiming that they rarely cooked meat at home, feasting on it only at the state's expense (i.e., at work-related banquets). They considered this behaviour degrading and a negative marker of social status. Räwiä, a language specialist in Ürümchi, and her Uyghur neighbour scoffed in 1995:

Hans don't eat meat at home. That's why they stuff themselves every time the government is paying. That's the only time they ever eat meat. They eat plenty then!

Behind the stereotype lay the implication that the slight and 'effeminate' physique of Han men, compared with the strong, broad build of Uyghur men, was the result of consuming less meat. In Han China, what Kam Louie terms the 'cerebral male tradition' in the form of the *caizi* or 'fragile scholar' figure (i.e. *wen* masculine ideals) came to dominate over *wu* masculine ideals of martiality in Chinese discourses (2002: 8); the literati elite thus hoped through the mastery of literacy and ritual (*li*) to assert their political, social and cultural dominance (Abramson 2003: 133). Conversely, a highly muscled physique, though implying strength and martiality, was associated with the demonic creatures depicted in Buddhist iconography; it offended the prevailing Han belief that the body must balance the universal opposing forces of *yin* (the feminine principle, represented by fat) and *yang* (the masculine principle, represented by bone) (Abramson 2003: 135, 137). In contemporary times, the *wen* version of manliness is being challenged by alternative discourses circulating among minority group males, and serves to subvert existing power relations among men/ ethnic groups and construct an alternative social hierarchy (Hillman and Henfry 2006: 255-56; 268). For instance, Tibetan males are positively represented (by Tibetan men and women, and also by Han women) as tall, strong, broad-shouldered, sturdy, confident, generous, manly, handsome, decisive and brave. As the authors point out, this description corresponds closely to *wu* masculine ideals (martiality) in China (Hillman and Henfry 2006: 259-260).

Uyghur respondents asserted that an inferior physique rendered Han males less successful than Uyghur counterparts when participating in sports. A key example of this belief emerged when in the summer of 1995 young male Uyghurs in Ghulja organised a multi-ethnic football match as part of a series of activities intended to address social problems among the youth (in particular, alcohol use and intravenous drug use) and steer them towards healthier—and more moral—pursuits (Roberts 1998b; Dautcher 2004: 281-92). Amid state paranoia surrounding the possibility of separatist activities at that time, the football match was cancelled, tanks were driven on to the football pitch, and goals were removed from the pitch to prevent the gathering (cf. Millward 2004: 17). This led to widespread local protest, and the story was quickly circulated around the region via the 'rumour grapevine', so that Sadiq, a young publishing house employee in Ürümchi, complained to me soon after: 'The Hans disallowed the contest because they don't want us Uyghurs to be better than them at anything! They're no good at football. They haven't got the physique for it.'

To a significant extent, this stereotype was constructed upon local knowledge, passed down across centuries, of the contrastive lifestyles referred to by China historian John K. Fairbank as 'the steppe and the sown'. While the Han were moulded by intensive agriculture involving a sophisticated rice culture, the Inner Asian nomads depended on a sheep and horse economy for their livelihoods: 'From his flocks, the nomad secured food, sheepskins for clothing, shelter in the form of felt for his yurt, and fuel in the form of sheep dung' (Fairbank 1992: 23-4). China's 'culturalism'—the sense of 'cultural and aesthetic superiority' among Han people, founded on sedentary social institutions—was thus developed in relation to Inner Asian hunters, horsemen and warriors (Fairbank 1992: 25). This consciousness, described by one scholar as the product of a 'unique ethnoscape' constituted by civilisation versus barbarism (Bulag 2002: 3-4), endures to this day. Historically, meat was far less central in the Han diet than in that of the Turkic peoples. On the steppes, the nomad's diet consisted solely of meat and dairy products; and even after Uyghurs migrated West to become sedentary oasis dwellers, they reputedly continued to be 'voracious meat-eaters', with the common people consuming mutton, beef, camel flesh, yak, wild sheep, stag and antelope, while the ruling elites feasted also on horseflesh and wild fowl (Warikoo 1985: 97).

By contrast, the main staples in China were cereals, first cultivated as food crops during the Zhou Dynasty (ca. 1046-256 BCE). Although members of the imperial court began to eat domesticated chicken (originating from Thailand) by 5500 BCE, sheep, cattle and goats (from West Asia) by 4000 BCE, and pork (native to China) by 4000 or 3000 BCE, little meat or poultry

#### 96 chapter two

found its way into the diet of the common people.5 The poor ate almost nothing but grains (millet, wheat, and sorghum in the north; rice in the south), and became skilled in the arts of smoking, salting, sugaring, pickling and drying foods so as to be ready in times of drought and famine (Chang 1977: 9-10; 13). Whenever possible, they supplemented their rice with native soybean or cucumber, but meat was prohibitively expensive and eaten in small pieces on special occasions only. Regular consumption of large amounts of meat was 'traditionally the preserve of the wealthy', a distinction reflected also in the Chinese language, with the term *roushizhe*— 'meat-eaters'—applied to the ruling class (Chang 1977: 9, 15; Kieschnick 2005: 193). As Buddhist vegetarianism was gradually popularised in China from the sixth century, what was previously a sign of status and privilege became instead a symbol of decadence and gluttony. Vegetarianism, on the other hand, was associated with the virtues of self-sacrifice and simple living, hence the term that most closely approximates 'vegetarianism' in Chinese, *sushi* or 'plain eating' (Kieschnick 2005: 193-194). Now, commoners began to add *doufu* (bean curd) to their stir-fried dishes as a source of protein. Prior to 1949, poor locals in Hunan, Hubei and Sichuan could not afford to buy meat, instead eating the parts of animal carcass that the rich discarded (brain, intestine, fish heads, and chicken feet). Meanwhile, hot chilli peppers were dubbed 'meat for the poor' because they were a good accompaniment to rice when substituted for meat or salt,6 and dishes were submerged in oil in an attempt to make up for the lack of meat. The first Sichuanese hotpots have been described as 'famine food', eaten for nutrition and warmth rather than for taste.7 Against this historical background, the tendency of some Han colleagues to eat meatless dishes at home while consuming meat at the expense of their superiors allows con-

<sup>5</sup> The *Book of Rites* (one of the Five Classics of the Confucian canon) indicates that foods served to the emperor included beef, mutton, pork, fish, cured meat and intestine. Meanwhile, the *Book of Songs* (*Shijing—*the earliest existing collection of Chinese poems and songs) states that there were in excess of 130 edible plants in China during the Western Zhou period (ca. 1046-771 BCE), including more than 30 kinds of vegetable. China Internet Information Center. Retrieved from: http://www.china.org.cn/english/imperial/25995.htm (accessed 4 February 2012).

<sup>6</sup> China Internet Information Center. Retrieved from: http://www.china.org.cn/english/ imperial/25995.htm (accessed 4 February 2012).

<sup>7</sup> Wu Dan, manager of The Happy Old Family, a popular hotpot restaurant in Chengdu. Retrieved from: http://www.4panda.com/sichuan/about%20sichuan/cuisine.htm (accessed 10 March 2012). See also Daniel Allen, 'Fire and Rice: Chowing down in Chengdu'. Retrieved from: http://www.beijing-visitor.com/index.php?cID=440&pID=1793 (accessed 10 March 2012).

temporary Uyghur intellectuals to pour scorn upon them in an articulation of reverse class prejudice.

Objective cultural differences aside, I believe it is not stretching the point to suggest that contemporary Uyghur criticisms of Han people for eating *insufficient* amounts of meat can also be seen as a counter-stereotype, constructed in relation to the pejorative Han view which considers the meat-intensive diet of Turkic-Islamic peoples 'savage'. When relating my experiences in Xinjiang to Han friends in Beijing in 1996 and 1997, they would frequently make (half-serious) jokes about Uyghurs 'slipping raw meat inside their shirts to snack on when they get hungry'. Similar perceptions were also regularly expressed by Han residents in Ürümchi, stirring different reactions among Uyghur respondents. Some turned the stereotype into a source of humour, as in the following altercation, observed in the first days of my initial field trip in 1995-6:

A Han Chinese at the local market made apologies to me for the Uyghurs' diet of kebabs: 'They like their meat. It's their custom. They're a bit of a savage race...Didn't used to have fixed homes, you know.' I questioned whether a nomadic lifestyle could be equated with savagery, and Ghäyrät [a Uyghur male migrant worker] laughed. Later, another Han market trader arrived, his jacket sleeve split to the elbow. Apparently, he sold fish and had to reach into a fish tank all day. I joked about him trying to start a new fashion, and Ghäyrät laughed. Then he said: 'That Han must be very savage, walking around in a sleeve like that', and everyone roared with laughter.

However, a far larger number of respondents were infuriated by the stereotype, which they considered offensive and patronising. In light of this, it is entirely feasible that, in mocking the meatless diet and poor physique of the Han, contemporary Uyghurs are constructing a reactive counter-stereotype, based on social rank, which represents the majority as originating from an 'economy of scarcity' still in evidence as recently as the Republican era (Fairbank 1992: 298-9).

#### *Androgynous Women*

Han women too came under fire for possessing the wrong type of physique. In the eyes of female Uyghur respondents, Han females were androgynous and possessed none of the curvaceous attributes of Uyghur women (i.e. had slim hips and no bust). This perceived bodily deficiency, added to their 'ugly' facial features—they lacked the high-bridged nose, large round eyes, and luxurious black eyebrows of Uyghur counterparts—and their 'dull' dress sense, meant that Uyghur respondents located them at a considerable

#### 98 chapter two

cultural distance. While Islamic discourses more usually centre on the need for women to restrain their sexuality by hiding the contours of the body, a sense of the ideal female figure can nevertheless be gleaned from scholarly writings on Muslim women and from Muslim feminist literature. For the first generation of Arab female novelists (late 1950s-60s), for instance, the female body was characterised as the 'incarnation of womanhood', and their protagonists complained that male peers saw them simply as a 'collection of female physical attributes'—hair, lips, breasts and legs (Zeidan 1995: 139-40). In a study of young women in contemporary Indonesia, Bennett describes how her respondents defined femininity in terms of 'moderation' and 'containment'. 'Moderation' indicates the need to adhere to normative ideas of femininity, thus, bodies that are 'too slim', 'too tall', 'too dark', etc., are deemed 'unfeminine'. Moreover, the fact that women in the study sought (albeit cautiously) to 'maximise their physical attributes' and 'accentuate their figures' even within the parameters of Islamic feminine modesty (2005: 49), suggests that the ideal female figure is typically womanly, that is, it has curves in the correct places. It is for this reason that the patriarchy perceives the need to 'contain' it lest it distract men from their moral pursuits (cf. Mernissi 1975: 30-34).

# *'Killjoys'*

In Vasquez and Wetzel's study, a Mexican American respondent characterises mainstream white Americans as 'bland' and 'lacking culture, colour, feeling and passion' (2009: 1562). My Uyghur respondents evoked a similar stereotype in 1995-6 in referring to Han people as 'killjoys', and asserting that they 'don't know how to enjoy themselves'. To some extent, this stereotype results from objective cultural differences between the two groups. It is generally accepted that Uyghur people tend to be more extrovert than Han. Uyghur social activities, life cycle rituals (notably, weddings and circumcisions), Islamic festivals and pilgrimages are all based around musical performance and dance (Harris 2005: 629). Mujat Tokhti describes the lively and 'uninhibited' (Ch. *haofang*) nature of Uyghur wedding celebrations, where guests sing and dance for long hours (1995: 148). Song and dance traditionally played an important role in flirtation and courtship, whether at *mäshräp* gatherings, wedding parties or other chaperoned social occasions. Vikram Seth, travelling through Xinjiang in the 1980s, describes local songs in Turpan as 'based on repartee between lovers—usually a tubby man with a wicked moustache and a woman with flashing eyes'

(1987: 10). Han people, on the other hand, often characterise themselves as 'reserved' (Ch. *hanxu*) (Fang and Faure 2011: 322), and, for most, singing, dancing, and being the centre of attention does not come so easily. This tendency is also routinely stereotyped in PRC state media. Chris Hann provides an example from the Third Ethnic Games held in Xinjiang in 1986. An event which involved only minority nationality participants but no Han, the emphasis was on folklore more than competitive sports, and television footage showed scenes such as minority people playing traditional musical instruments and singing in idyllic mountain regions while Chinese officials 'looked on passively' (Hann 1991: 226-227). During the course of my fieldwork in 1995-6, this cultural disjuncture often produced the comic spectacle of a handful of Han guests trying (and failing) to simulate the flamboyant style of Uyghur dance at a Uyghur wedding, as Uyghur guests struggled to contain their mirth. Meanwhile, Uyghurs living in work unit complexes in Ürümchi and other 'New Towns' regularly received complaints from Han neighbours after staying up until the early hours singing, playing musical instruments and dancing. The relative disinclination of Han people to 'perform', together with the internalised Confucian notion of conflict avoidance, prevents them from taking part in such revelry themselves. A similar binary has been expressed by Tibetan men, for whom manliness (Tib: *phokhyokha*) is embodied in the act of singing and dancing in public, and directly reflects male confidence (Hillman and Henfry 2006: 263).

In addition to these objective differences (which are based on 'ideal types', and do not account for the many exceptions which inevitably populate any group), there may be another basis for the 'killjoy' label. Like stereotypes concerning consumption of meat (above), it may be formulated—consciously or otherwise—in response to a common stereotype circulated among Han that all ethnic minorities are 'good at song and dance' (*neng ge shan wu* 能歌善舞) (Blum 2001: 68-9; 75). While this is certainly a more positive stereotype than some others, it is nonetheless an essentialising simplification; moreover, in the view of some scholars, it suggests the state's deliberate intention to contrast the 'lingering primitiveness' of minority cultures negatively against Han 'modernity' (Gladney 1994, 2004; Blum 2001). It also fails to acknowledge the rich diversity of positive cultural characteristics associated with Uyghur society and Islamic societies in general. In this context, I would suggest that the counter-stereotype of Han people as 'bad at song and dance' functions to reclaim

#### 100 chapter two

musical ability as a positive cultural attribute of the Uyghur people in order to challenge its co-optation and 'de-civilisation' by the state.

# *The 'Han Disposition'*

Many Uyghur respondents criticised Han people—as a group—for their temperament, labelled pejoratively the *Khänzu mijäz* [Han disposition]. This term encompasses a wide spectrum of ills, mostly related to the indirect, *hanxu* tendency in Han communication style. Traditionally, Han Chinese have valued a person who does not indulge in excessive verbal communication, considering them wise and trustworthy: 'Speaking is silver, silence is gold' (Fang and Faure 2011: 322; 327). Uyghurs, however, view this style of communication negatively, and consequently see Han people as lacking in social skills. Some respondents speaking in 1995-6 characterised the 'Han disposition' as a curt manner, introverted nature or tendency to sullen moods, and linked it to the 'killjoy' stereotype above. For others, the label denoted an indirect way of speaking, indicative of a 'slippery' or dishonest character. This type is mentioned also in Susan Blum's study, where it proves problematic for one individual of mixed Han-Bai ethnicity. Yongyong, whose mother is Han and his father Bai, finds himself caught between the two cultures when he hears someone say that Han aren't *laoshi* [honest]. He feels compelled to protest: 'But my mum is Han, and *she's*  honest!' (Blum 2001: 6). The negative perception seems also to have been shared by Christian missionaries operating in China in the nineteenth and twentieth centuries, for whom peripheral peoples were honest, simple and hard-working, while Han were clever but sly, sneaky and untrustworthy (Harrell 1994: 22). On the other hand, the term 'Han disposition' was alternatively employed to describe a person whose words and acts were loud, tactless or abrupt. While this second definition seems to contradict the first, it in fact reflects the observable duality of implicit versus explicit communication in the contemporary behaviour of urban Han. Whereas Han Chinese—like other East Asian peoples—are traditionally averse to the explicit, direct and visible, they also value *renao* [noise and good cheer]; the type of communication chosen depends upon context and counterpart. Though a person may be reserved and implicit in formal settings, that same person can be expressive and explicit in informal settings, where these are important for business relationships (Fang and Faure 2011: 324; 326). Behaviour also varies according to the nature of the relationship; the deeper the trust, the more directly Han Chinese tend to communicate (Fang and Faure 2011: 326).

In all of the above contexts, the 'Han disposition' describes a person unversed in social etiquette and ill-equipped for social interaction, where these terms are defined *according to Uyghur cultural norms*. In a fascinating discussion, Ildikó Bellér-Hann describes how the *mijäz*—a central component of Uyghur identity—has a 'strong bodily aspect', so that each individual is believed to be born with either a 'hot' or a 'cold' temperament. Since a person's specific disposition may be modified by the foods he or she consumes, it is important to maintain a 'moderate' temperature. It is in this context, then, that the Han *mijäz* is seen as an 'uncontrolled disposition' which results from Han disregard for the rules governing diet (2002: 72-74).8

# 2.6. *Stereotypes relating to Islamic Social Practices*

#### *'They Eat Anything'*

As noted by Uyghur historian Mujat Tokhti, every aspect of Uyghur family, social and economic life has been, and continues to be, heavily influenced by Islam (Tuoheti 1995: 147; 149; 124). As a result, Islamic social laws featured prominently in the counter-stereotypes that Uyghurs deployed to demonstrate their cultural superiority relative to Han. One of the most frequent complaints among respondents concerned the 'unclean' Han diet, in particular, the consumption of pork. Stereotypes of Han food culture derive directly from negative comparison with Uyghur food culture, in which Islamic dietary requirements proscribe certain *haram* [unclean, unlawful] foods. In Xinjiang, the proscription on *haram* foods has been more or less strict since the Buddhist inhabitants of Turpan became the last Uyghurs to convert to Islam in the thirteenth century, and has long been exaggerated in situations of inter-ethnic contact (cf. Bellér-Hann 2008: 71). Moreover, during periods of orthodox Islamic rule, such as the reign of Khoqandi adventurer Yaqub Beg (1864-77),9 the prohibitions on 'unlawful meats' were extended to include not only pigs but also rats, cats, dogs and

<sup>8</sup> Interestingly, research conducted among successive waves of Uyghur migrants to Kazakhstan reveals a cleavage between 'Chinese Uyghurs' and 'local Uyghurs', with the former accusing the latter of having adapted to Russian ways and developed a 'Russian temperament' (*mijäzi orus*) (Bellér-Hann 2001: 60).

<sup>9</sup> The city of Khoqand lies in contemporary Uzbekistan, at the southwestern edge of the Ferghana Valley. The crossroads of two ancient trade routes and the major religious centre in the region, it was the capital of an Uzbek kingdom known as the Khanate of Khoqand (1740-1876) (see Newby 2005).

#### 102 chapter two

donkeys (Warikoo 1985: 97). This is corroborated by Bellér-Hann, who cites one Hindu observer's remark that while horse meat was halal in Turkestan, donkey meat was considered unclean or haram and thus a pollution offence (2008: 388). My respondents in 1995-6 criticised Han people for exercising no discrimination in the slaughter and consumption of animals, often appropriating the Hans' own popular sayings in order to indict them. The saying *Neng chi shenme jiu suan shenme* [Only edible things have value] was cited as evidence of an unhealthy preoccupation with food consumption (over other activities). Other respondents adapted a famous northern Chinese saying as follows: 'Hans eat anything! Of things with four limbs, the only things they don't eat are tables; of things that fly, the only things they don't eat are aeroplanes!'10 In respondents' minds, this attitude pointed to a fundamental and shameful lack of discrimination.

These differences in eating habits, used to distinguish 'pure' Uyghur culture from 'unclean' Han culture, also figured in the notorious bus and market stories circulating during the 1990s. Such tales focused on the colourful description of ethnic conflicts occurring as a result of Uyghur and Han residents coming into too close contact in a crowded place. The following story was related by Shökhrät in 1995:

There was a Uyghur on a crowded bus. He did something that upset a Han, who turned around and said: 'Why don't you go and get on your donkey cart?' The Uyghur replied: 'I can't. You Hans have eaten all our donkeys and horses!'

Historically, both Turks and Persians have eaten horsemeat. In the Kazakh culture, for instance, traditional foods containing horsemeat and mutton reflect the nomadic lifestyle;11 and it is well-known that Uyghurs in the northern oasis of Ghulja eat horsemeat as a delicacy. However, an important distinction lies in the respective classifications of horsemeat and donkey meat in Islamic law. The majority of Islamic jurists rule that it is permissible to eat horsemeat, citing *Sahīh al-Bukhāri* (5191) and *Sahīh*

<sup>10</sup> The original saying circulated among northern Chinese goes: 'The Cantonese eat anything! Of things with four limbs, the only things they don't eat are tables; of things that fly, the only things they don't eat are aeroplanes!'

<sup>11</sup> Typical Kazakh dishes include *shuzhuk* (a type of sausage made from smoked horse meat) and *kuyrdak*, which is prepared from a freshly slaughtered horse, sheep, or cow, and involves boiling pieces of the animal's heart, liver, kidneys, etc. in oil, and serving them with onions and pepper. See the dedicated web portal of the Turkish Cultural Foundation: http:// www.turkish-cuisine.org/english/pages.php?ParentID=3&FirstLevel=19 (accessed 13 February 2012).

Figure 6: Travel by donkey cart, Kucha Old Town (1996)

*Muslim* (1942), where it is narrated that: 'At the time of the Prophet, we slaughtered a horse and ate it.' On the other hand, Muslim scholars are fully agreed that the meat of domestic donkeys is *haram*, citing *Sahīh al-Bukhāri* (5520) where it is narrated that Jabir ibn 'Abd-Allah said: 'On the day of Khaybar,12 the Prophet (peace and blessings of Allah be upon him) forbade the meat of domesticated donkeys, but He granted us a concession allowing us to eat the meat of horses'.13

Uyghur respondents often expressed revulsion at the sight and smell of pork. On one occasion in 1995, Shökhrät, a male intellectual in his thirties,

<sup>12</sup> The Battle of Khaybar was fought in the year 629 between Muhammad and his followers and the Jews living in the oasis of Khaybar, located 150 kilometers from Medina in the north-western part of the Arabian peninsula (modern-day Saudi Arabia).

<sup>13</sup> Retrieved from: http://islamqa.info/en/ref/70320 and http://islamqa.info/en/ ref/85534 (accessed 12 March 2012).

described his reaction on suddenly coming across a pig's carcass in an Ürümchi outdoor market: 'I almost threw up. It was the first time I felt really nauseous from looking at pork. It had only just been slaughtered, and still had its head and everything. Uggh!' On another occasion that same year, I observed how he pointedly avoided using a public telephone kiosk, because he caught sight of the Han owner carving up pork inside. Respondents complained of the 'offensive' smell of pork, both when raw and when cooking. Mobaräk, an economics student originally from Kucha, stated in 1996 that she and her friends would circumvent market stalls where pork was on sale in order to avoid passing near the raw meat. Uyghur respondents living in residential buildings in Ürümchi's work units during this period complained of the smell of frying pork that emanated daily from Han neighbours' homes.

Still others complained in 1995-6 that Han people 'show no respect for the animal once dead', and pointed out that the Han banqueting culture (*yanhui wenhua*) results in large amounts of food waste. In deliberate contrast to these deplorable practices, they observed that Uyghurs slaughter only sheep and cows, and use every scrap of the carcass, to show respect for the animal's sacrifice. The disdain for food waste derives from an Islamic teaching which exhorts believers to live frugally and use their wealth thoughtfully: 'Allah says: "Eat and drink, but be not excessive. Indeed, He does not like those who commit excess" (*Sūrah al-A`rāf*: 31). Believers are reminded not to gorge themselves on food and drink beyond necessity, but to conduct themselves appropriately, eating only the few morsels they need, and spending the remaining money on charity.14 Over and above the issue of food waste, it is made clear in the Islamic scriptures that waste of all kinds must be avoided, since its root is ingratitude. The correct response to the beauty and bounty of Allah is gratitude and contentment with a little; a person who continually consumes and is never satisfied no matter how much he has, is ungrateful and the 'brother of Satan'.15

<sup>14</sup> Sheikh Muhammad al-Hamad, "Eat and drink, but be not excessive". Retrieved from: http://en.islamtoday.net/artshow-267-3155.htm (accessed 12 March 2012). It is interesting to note that Tibetan male respondents also criticise Han men for their excessive concern with food, agreeing that 'real men just ate their food and did not worry about how good it tasted'. This view is reflected in a popular Tibetan saying, which states that when men of a low level come together they speak of eating and drinking; men of a high level rather speak of the great warrior and statesman, Gesar (Hillman and Henfry 2006: 262).

<sup>15 &#</sup>x27;Ecology and Islamic values I'. Retrieved from: http://www.islamweb.net/emainpage/ index.php?id=101557&page=articles (accessed 12 March 2012).

Stereotypes of the Han diet as 'unclean' occasionally led my Uyghur respondents in 1995-6 to liken Han people to *haram* animals, constituting a serious insult in Uyghur culture. They compared their 'ugly' facial features to those of pigs, or joked that Hans had 'ended up looking like the pigs they eat'. By contrast, respondents compared themselves—or rather their personalities—to sheep, implying that the Uyghur people have a docile nature, and are easily led. Politically aware respondents saw this is an obstacle to ethnic incorporation, and joked that it was the consequence of eating too much mutton. Bovingdon cites one instance of black humour in which an older Uyghur male sitting on a crowded minibus observed that the fundamental problem in Xinjiang was that 'sheep and pigs are forced to live together in one pen', provoking an uproar of laughter among the other Uyghur passengers (2010: 89).

On one occasion in 1996, I observed a Uyghur respondent indirectly compare a Han resident to a dog (another *haram* animal).16 We were walking down a narrow passageway amid a maze of houses in a Uyghur 'Old Town' in southern Xinjiang when a gate suddenly opened ahead of us, and a dog was heard barking. My companions, Aliyä, a female postgraduate student, and her mother were afraid that the dog was about to emerge, and took a few steps back. When a Han male stepped out instead, Aliyä burst out laughing and remarked (in Uyghur): 'Ah well, there's not much difference!' Besides this, the only other times I heard the word 'dog' used as an insult was with reference to policemen; in that context, it was levelled equally at Han and Uyghur police officers, each viewed as 'unclean' owing to their apparent allegiance to the Chinese state.

# *'Dirty'*

Uyghur respondents in 1995-6 frequently stereotyped Han people as 'dirty', with reference to the latter's washing customs. In their view, Hans, in washing their hands in a bowl of water, wash in their own dirt. Many claimed privately that Han people use the same bowl with which they wash their feet to later prepare vegetables for their dinner. They pointed out that Uyghurs, by contrast, *pour* water over themselves, ensuring that the water, once dirtied, does not come into further contact with the skin. It was customary in the nineteenth century to offer guests a precious-metal pitcher

<sup>16</sup> Dogs are considered impure in many traditional Islamic societies, and a hadith mandates ritual ablution for anyone who has come into contact with dog saliva (Esposito 2003: 70).

#### 106 chapter two

and basin (or a simple ewer for persons of lower social rank) with which to wash hands before a meal (Bellér-Hann 2008: 206-7), and this practice is still maintained by elites in Ürümchi today. On the same basis, respondents berated the Han habit of shaking water off their hands after washing. Uyghurs, they pointed out, never shake water off their hands, but gently squeeze them together and hold them aloft until handed a towel, or by their sides until their hands have dried naturally. This practice ensures that the dirtied water does not accidentally splash someone else. Over the course of a year, I witnessed only one young girl, aged 19, shake water off her hands, and she did so in a dark corner of her home when she believed no-one was watching.

A principal source of cultural misunderstanding in this area is the contrastive concepts of cleanliness that govern respectively the Islamic ritual environment and the 'scientific', secular world. Uyghur and Han people practise different methods of washing. As in many Islamic desert societies where water is in short supply, Uyghurs have traditionally taken the *kichik tärät* ['small wash'—refers to the ritual purification of the hands and lower arms, feet and ankles, and facial orifices] more regularly than the *chong tärät* ['big wash'—refers to the whole body].17 For a Uyghur, the 'small wash' was more important since it was this that was required for entry into the mosque. While the 'big wash' has become a more regular habit for urban Uyghurs living in residential buildings with running water, it remains an infrequent practice in many rural areas. Moreover, the Islamic concept of waste avoidance also extends to water usage. The Prophet Muhammad is known for his attention to conservation, and is said to have urged believers to avoid using more water than necessary when performing ritual ablutions.18 Evidently, since Han people are non-Muslims, they do not share the religious notion of taking a 'ritual wash', which renders a person 'pure' before Allah and ready to perform prayers. Furthermore, most Han migrants to Xinjiang settle in cities or 'New Towns', where they live in modern residential apartments replete with running tap water and shower. They can therefore easily access water for the purpose of bathing. The lack of sanitation among peripheral peoples is a point frequently raised by Han ethnologists when telling their tales from the field, with minorities seen as

<sup>17</sup> The Arabic terms denoting the same practices are *wudu* (ritual ablutions)and *ghusl*  (a full bath).

<sup>18 &#</sup>x27;Ecology and Islamic values I'. Retrieved from: http://www.islamweb.net/emainpage/ index.php?id=101557&page=articles (accessed 12 March 2012).

polluting: at once dirty and dangerous (Harrell 1994: 13). In Xinjiang, while some Han social scientists are aware of the minutiae of Uyghur customs relating to washing practices, few ordinary Han could say the same. This is because the two groups scarcely interact outside the work unit, and do not form relationships or friendships where a deep level of cross-cultural exchange might be possible. Uyghur scholars, for their part, give detailed and positive accounts of Uyghur customs and habits, some of which are published in Chinese (see for example Tuoheti 1995).19 It is unclear however to what extent Han residents form part of the readership of scholarly publications produced from within the Uyghur community.

# *'Ill-mannered'*

Many respondents characterised Han people as 'ill-mannered', complaining regularly about them breaking wind, spitting, or blowing the contents of their nose onto the ground in public places. Tajigül, the young wife of an Ürümchi restaurant owner, exploded with disgust one evening in 1996, exclaiming:

I hate the Hans! I really do hate them. They're so dirty. The way they sit at a table eating, and then turn and spit on the floor ... even in their own homes! They may wear fine clothes, but their manners are filthy! They don't know what courtesy is.

In this way, she drew a distinction between wealth and morality, pointing out that the superior living standards of many Han residents did nothing to assuage their lack of social etiquette. By implication, she characterised Uyghurs as culturally and morally superior, on the basis that no Uyghur would ever commit the above acts in public. There was evidence that other Turkic-Islamic groups shared Uyghurs' revulsion at these habits. On another occasion in 1996, I was eating *polo* (pilau rice cooked in sheep fat with carrots, raisins and mutton) in an Ürümchi street restaurant. Opposite me sat an elderly Han male, on my right sat a Kazakh girl, and on my left a young Uyghur couple. Suddenly, the Han male stopped between mouthfuls, turned his head, and spewed a long train of spittle onto the floor. The Uyghur girl clicked her tongue loudly, muttered something under her breath, and threw the Kazakh girl a disgusted glance. In response, the

<sup>19</sup> Mujat Tokhti is a Uyghur historian who graduated from the Central Institute for Nationalities in 1982. He spent a year in Japan as a Visiting Research Fellow in 1992, before commencing his PhD research in East Asian history / Uyghur studies at Tokyo University in 1994.

#### 108 chapter two

Kazakh girl pulled a face and said she 'felt sick'. The Uyghur customers later complained vehemently to the chef: 'Can't you do something about those Hans? Spitting in front of us while we eat!'

The question of breaking wind is a subject of some interest on internet forums devoted to discussions of Islamic teachings, with many followers writing in to ask for advice.20 It is generally agreed among Islamic scholars that to break wind in front of others is dishonourable and humiliating. Beyond this, the rules surrounding the breaking of wind before or during prayer are quite intricate. Many argue that the mere imagining of gas in the stomach or passing of wind does not invalidate ritual ablutions, citing the following *hadith*: 'The Prophet was asked about a man who felt something during his prayer—should he stop praying? He said, "No, not unless you hear a sound or detect an odour"' (*Sahīh Al-Bukhāri*, 1915; *Sahīh Muslim*, 540).21 Thus, if after passing wind there is no offensive smell, a person may enter the mosque and pray. If, however, the passing of wind produces an unpleasant smell, it is not permissible to enter the mosque, because this will cause annoyance to other worshippers and the angels. The emission of sputum or mucus, meanwhile, runs against the central Islamic notion of spiritual and physical hygiene. Cleanliness is an integral part of *iman*  (faith), and inner purity can be attained only by maintaining outer cleanliness. According to Islamic scholars, cleanliness comprises three types: purification (ritual ablutions); keeping the body, clothing, and environment clean; and removing the dirt that collects in various parts of the body. With regard to the second type, a Muslim must pay attention to personal hygiene (since the body is a gift from Allah) while taking care also to maintain hygienic surroundings. Filthy conditions cause bodily disease, and Islam exhorts followers to keep body and mind free from disease (Stacey 2009). It is especially important to ensure that food preparation areas and hands that touch food are clean, and a Muslim is expected to refrain from spitting or blowing his nose whilst eating.

<sup>20</sup> See for example: http://www.ummah.com/forum/showthread.php?122946-Breakingwind-and-Wudu (accessed 22 March 2012).

<sup>21</sup> A *hadith* is a report of the sayings or actions of Muhammad or his companions, and the tradition of its chain of transmission. There are six major collections of *hadith* in Sunni Islam, among which *Sahīh Al-Bukhāri* and *Sahīh Muslim* (sometimes called 'the two *Sahīhs*') are considered the most authentic. They have been described as 'an indigenous product of Muslims' understanding of their own scriptural tradition', as a 'standard of truth', and as possessing the status of 'undeniable historical reality' (Brown 2007: 36; 38).

On occasion in the mid-nineties, I would observe a Uyghur spit on the floor in public. When I remarked on this apparent contradiction, Uyghur respondents invariably retorted that the bad habit had been 'learned from the Hans'. This was a strategy frequently adopted by Uyghur acquaintances in the field to deal with anomalous behaviour that contravened standard in-group representations of 'pure' Uyghur culture; it will be seen again in the example which follows.

# *'Disrespectful to Elders'*

Uyghur respondents in 1995-6 criticised Han people for 'showing no respect to their elders,' often citing the example of cigarette and alcohol consumption. Young Hans, they asserted, think nothing of smoking and drinking even getting drunk—in front of older relatives. They contrasted Uyghur youth favourably against Han, pointing out that some (i.e. practising Muslims) abstain from cigarette and alcohol consumption altogether, while others take care to abstain in front of parents and grandparents, in recognition of avoidance rules between males. In Ürümchi during the mid-nineties, religiosity differed greatly between generations, with most elderly people attending the mosque each Friday, while many young people went only during Islamic festival periods. The act of abstinence before grandparents thus constituted a considered mark of respect for those members of the community actively engaged in Islamic practice. More broadly, Islamic scholars argue that smoking is *kabīth* [repulsive], and that its 'repugnant odour' offends those not accustomed to smoking. They cite the following *hadith* from the two *Sahīhs* in support: 'Whoever has eaten onion or garlic is to keep away from us and our *masājid* [mosque], and is to keep to his house.'22 The smell of cigarette smoke is deemed 'no less repulsive' than the odour of onion or garlic on a person's breath.23 On the other hand, alcohol consumption in the presence of elders is problematic because revellers run the risk of losing their social inhibitions and failing to treat elders with the required respect. Worse, they might unwittingly disclose secrets or expose faults, causing shame, regret and social disgrace.

<sup>22</sup> Sheikh Muhammad bin Ibrāhīm bin Abdul-Latīf Āl-ish-Sheikh. Retrieved from: http:// www.islamawareness.net/Smoking/sfatwa004.html (accessed 13 February 2012).

<sup>23</sup> Garlic is also problematic in certain other religions. In Mahayana texts imported to China from India in the early fifth century, the consumption of the 'five pungent plants' (*wuxin*), including onions, chives, garlic, and so on, was condemned, with the Buddha noting: 'people despise men who smell of garlic' (Kieschnick 2005: 190-192).

#### 110 chapter two

This counter-stereotype posits that abstinence from cigarettes and alcohol before elders reflects a heightened sense of respect and social morality among young Uyghurs as compared with young Han. Yet at the same time it raises the spectre of hypocrisy: after all, is a Muslim permitted to smoke and drink in the first place? This is a question not restricted to Xinjiang but debated across the Islamic world. The basis for the controversy is that smoking did not exist in the time of the Prophet; thus, religious scholars may draw only on general principles of Islamic jurisprudence to support their arguments for prohibition. Those who rule that smoking of cigarettes is impermissible take their evidence from two sources: the Qur'anic scriptures and the *hadith*; and scientific expert witness (physicians and researchers of medicine). The core argument drawn from the scriptures revolves around the tripartite issue of *harm*: harm to one's religion, to one's health, and to one's financial condition.24 Since smoking impedes prayer, causes physical disease, and is considered wasteful and excessive, a Muslim who smokes is guilty of all three sins. To the scriptural evidence, scholars add the copious studies by medical experts which find that smoking causes fatal illnesses. Interestingly, the issue of a fatwa against smoking was a central feature in the popular Iranian protest of 1891-2, which saw the ulama and merchants join together to resist the shah's granting to the British of a monopoly on tobacco trade (Esposito 2003: 321).

The issue of alcohol prohibition is easier to support through the scriptures, since the intoxicating nature of fermented drinks is referred to di-

<sup>24</sup> Regarding 'harm to one's religion', smoking is said to obscure the mind, impeding the performance of acts of worship and/or fasting. A *hadith* in *Sahīh al-Bukhāri* is cited in support, wherein the Prophet is said to have declared: 'Every intoxicant is *haram*'. While the term 'intoxicant' is more often applied to alcohol, its broad definition denotes a substance producing a condition of diminished mental and physical ability, hyper-excitability, or stupefaction, or which poisons or produces symptoms of poisoning in the human body. It can thus be argued that smoking is *haram* because nicotine is a stimulant and poisons the body. Regarding harm to one's health, religious scholars point out that Islam teaches its followers to take care of their physical health, citing the following *hadiths*: 'Be not cast by your own hands to ruin' (*Al-Baqarah*: 195); and 'Do not kill yourselves' (*An-Nisa'*: 29). Regarding harm to one's financial condition, it is noted that Allah forbids the 'squandering of money'. By extension, smoking is deemed a waste of resources which might have been used to benevolent (charitable) ends. See the rulings of Sheikh Ahmad Kutty, Senior Lecturer at the Islamic Institute of Toronto and the Saudi scholar, jurist, exegete, and Arabic grammarian, Sheikh Abdur-Rahmān an-Nāsir as-Sa'di (1889–1956). Retrieved from: http://www. islamawareness.net/Smoking/sfatwa004.html (accessed 13 February 2012). See also the approbation of Sheikh Abdul-Muhaymin Abu Samh (Imam of Masjid al-Harām, Mecca) of the ruling on the impermissibility of cigarette smoking. Retrieved from: http://www. islamawareness.net/Smoking/sfatwa004.html (accessed 13 February 2012).

rectly. For instance, Chapter 4 Verse 43 of the Qur'an exhorts Muslims not to conduct prayer while in a state of inebriation: 'O ye who believe! Approach not prayers with a mind befogged, until ye can understand all that ye say'. Given that a Muslim is required to pray at five regular intervals throughout the day, this command makes it difficult for him to become drunk in the first place. In case that he does, and then fails to appear at the mosque, his friends will take note of his absence. A Muslim who falls away from the religious community in this way is said to be 'like a lost sheep among wolves', and the religious brotherhood will normally work to 'save' him (i.e. encourage his abstinence) through peer pressure. This particular dynamic is currently in full swing in Xinjiang as more and more Uyghurs return to orthodox Islam (see Chapter 5). A later verse in the Qur'an declares total prohibition: 'The devil wants only to cast among you enmity and hatred by means of strong drink and games of chance [gambling] and to turn you from remembering Allah and from prayer. Will you then desist' (Chapter 5, Verses 90-91).25 Like smoking, alcohol consumption is believed to cloud the mind (distracting the believer from Allah), cause disease and waste money, all of which are prohibited (see above). Moreover, it is considered responsible for a wide range of social ills, including the creation of enmity between people, participation in unlawful sexual relationships, domestic and other forms of violence, suicide, gambling, impaired judgement and risk taking: in short, the destruction of individuals, families and communities. As a result, conservative religious scholars list alcohol consumption among the major sins, warning that all Muslims will be held accountable on the Day of Judgement for how they used their youth and strength, how they consumed their time, and how they spent their wealth.26

# *'Heartless'*

Uyghur respondents often characterised Han people as 'heartless' or 'inhuman'. For example, on one occasion in 1996, a middle-aged female teacher in Ürümchi expressed regret for the death of a Uyghur acquaintance's mother, and observed that 'he must be feeling very sad'. In response, her teenage daughter, Sänäm, retorted: 'If he were a *Han*, he wouldn't be upset!'

<sup>25</sup> Esposito (2003: 140). See also the following article published in the *Islamic Journal* (December 1995)*.* Retrieved from: http://www.islamawareness.net/Alcohol/alcohol\_about. html (accessed 13 February 2012).

<sup>26</sup> Retrieved from: http://www.islamreligion.com/articles/2229/; http://islam.about. com/od/health/f/alcohol.htm; http://www.inter-islam.org/Prohibitions/khmrmk.htm (accessed 13 February 2012).

#### 112 chapter two

Her mother admonished her gently, saying: 'People of *every* race must be upset when their mum dies.' Yet despite the common-sense truth of this statement, Sänäm persisted in her view. Her negative perception can be partly understood if we consider the different ways in which the respective groups deal with death. While I was staying in a small village in Aqsu prefecture in the summer of 1996, an elderly neighbour passed away, and the village was engulfed in a vocal outpouring of grief. As I discussed the villagers' reaction with a local middle-aged teacher, she observed: 'The Hans don't know how to cry*.*' Further prompted, she explained that Han people do not know how to cry *in the correct way*. When a Uyghur person dies, their relatives will 'cry' (*yighilimaq*) in a way peculiar to Islamic custom. This 'crying' or wailing takes the form of communal prayer and the recitation of formulaic lines (Bellér-Hann 2008: 289-90). It first takes place at the moment of death as a way of saying farewell (Tuoheti 1995: 148). It is then repeated at *näzir* mourning ceremonies held on the third, seventh and fortieth days after the death, and again on the one-year anniversary of the death. The 'crying' takes a very public form, and demonstrates to relatives, friends, and neighbours the love that one felt for the deceased. As noted by Bellér-Hann, female mourning is particularly expressive, with women expressing their lamentations while simultaneously pulling their hair, scratching their faces and beating their chests (2008: 290). If family members do not cry in this way at these times, others assume that they cared little for the deceased and criticise them. Since Han people do not practise this particular form of grieving, many Uyghurs conclude that they must be devoid of emotion. There are also key differences in funerary practices. While urban Han have tended to cremate the dead since the inception of the CCP (in Uyghur eyes causing 'harm' to the body), Uyghurs adhere to the Islamic practice of burial. Although strict Islamic practice requires burial to take place within one day of the death, in Xinjiang it should happen within three days in order to allow distant relatives to attend the funeral. Prior to burial, the body is ritually purified by an *ahong*, cleansed three times by female relatives, wrapped in white cloth and sprinkled with perfume (Tuoheti 1995: 148-9; Bellér-Hann 2008: 287). Another Han funerary tradition which jarred with respondents was the playing of the *sunay* [shawm],27 since, as 22 year-old respondent Kamil explained in 1996, Uyghurs play *sunay—*an instrument with a loud, jubilant and celebratory sound—at their weddings. By contrast, Uyghur funerals are a sombre and solemn affair (Tuoheti 1995: 149; Bellér-Hann 2008: 287-92).

<sup>27</sup> The Uyghur term *sunay* is transliterated as *suona* (唢呐) in Chinese.

An incident in 1996, observed as I travelled by road from Khotän to Ürümchi two days after a fatal flash flood had occurred in Toqsun (Turpan prefecture), neatly illustrates the issue. As we came across a scene of total devastation, Han passengers crowded off the bus and began to speculate about the corpses strewn beside the road. They laughed among themselves, pretended to place bets on whether the bodies were male or female, and poked at them with their feet. Uyghur passengers, all of whom had remained on the bus, clicked their tongues loudly in disapproval, and shook their heads. For them, such behaviour was irreverent; as Muslims, they had been taught that the human body deserves respect in death just as in life. When I related the story to Uyghur respondents in Ürümchi, each had this identical reaction: 'You see? The Hans have no feelings.' A more recent example concerns the reaction of Han youths to the terrorist attacks on the World Trade Centre on September 11, 2001. Young Uyghurs studying in universities in China proper were shocked to witness jubilant celebrations among their Han classmates, who were apparently delighted that 'the US cultural imperialists had got their just deserts'. Uyghur students swiftly transmitted what they had witnessed back to relatives and friends in Xinjiang, so that respondents in 2002 pointed to 'the coldness of heart' of these Han students and labelled them *sarang* [madmen].

One further example can illuminate the stereotype. While conducting fieldwork in the late eighties*,* Justin Rudelson observed a cultural difference between Uyghur and Han communities in Turpan related to the disposal of the placenta (*hämra*) following childbirth. Among Uyghurs, the placenta is recognised as a constant companion through life, and, as such, is buried in the wall of the domestic residence at eye level (Ben-Adam 1999: 204). A similar tradition is also mentioned in a study of late nineteenth- and early twentieth-century Uyghur community practices, with the placenta being buried in the mud floor of the mother's natal household and symbolising her attachment to it (Bellér-Hann 2008: 282). It is said that if this is not done, the baby will be unlucky, may become sick, or may even die (Bellér-Hann 1999: 129-130). Han people, conversely, find this practice surprising, since in Chinese culture the placenta has long been believed to have health benefits and is made into medicine or eaten (Rudelson 1999: 204).28 This practice has been a source of disquiet for Uyghur women, par-

<sup>28</sup> Dried human placenta (*placenta hominis*, or *ziheche* 紫河车) continues to be used in traditional Chinese medicine today to treat a variety of ills, including insufficiency of the liver and kidney, impotence, infertility, deficiency of the lung and kidney, asthma of defi-

#### 114 chapter two

ticularly since following delivery in modern state hospitals only the male child's placenta is preserved for the manufacture of medicine, giving rise to the belief that a female child's placenta is simply discarded as rubbish (Bellér-Hann 1999: 130). While the human placenta is revered in many cultures around the world, the practice of 'placentophagia' (the act of eating the placenta after birth) is quite rare. Far more common—and closer to the ancient Uyghur practice—is the symbolic act of burying the placenta in the garden, as in the Hawaiian tradition where it is planted together with a tree which then grows alongside the child.29

While Rudelson did not elaborate on Uyghur views on the consumption of human placenta, one can imagine that it might be similar to the horror that greeted a controversial piece of Chinese performance art in 2000. The installation *Eating People*, part of a series by Beijing-based artist, Zhu Yu, which explored cannibalism and human hypocrisy, was presented at the infamous *Fuck Off* exhibition in Shanghai.30 It consisted of a series of photographs of the artist in his home, apparently preparing and consuming a human foetus. In a domestic media report, Zhu claimed to have cooked and eaten the corpses of babies stolen from a local medical school. He stated that he had vomited several times during the course of eating it, but insisted that the act was necessary 'for art's sake'.31 The piece was later included in *Beijing Swings*, a documentary film about the Chinese art scene broadcast in the UK on Channel 4. The documentary's presenter, *Sunday Times* art critic, Waldemar Januszczak, described the work of Zhu Yu and his peers as 'the most outrageous, the darkest art, of anywhere in the world' (BBC Online, 2003). The decision to broadcast the programme in the UK led to a raft of public complaints, provoking the artist to defend his actions by pointing out that no religion and no law forbids cannibalism: 'I took

ciency type, deficiency of Qi and/or blood, wasting diseases, shallow complexion and insomnia. See the Hunan-based TCM website: http://w ww.tcmtreatment.com/herbs/0 ziheche.htm (accessed 18 February 2012).

<sup>29 &#</sup>x27;Why eat a placenta?', BBC Online, 18 April 2006. Retrieved from: http://news.bbc. co.uk/1/hi/magazine/4918290.stm (accessed 16/2/2012).

<sup>30</sup> *Fuck Off* was curated by avant-garde artists Ai Weiwei and Feng Boyi as a reaction against the 'over-complaisant' Shanghai Biennale: the city's first attempt at an international survey of contemporary art. The exhibition's title was an interpretation of its Chinese name *Bu hezuo fangshi*(不合作方式, strictly translated as 'An Uncooperative Style'), but the blunter English version was preferred by its participants. http://artasiapacific.com/Maga zine/53/TheFreedomOfIrreverenceAiWeiwei (accessed 18 February 2012).

<sup>31</sup> http://www.artspeakchina.org/mediawiki/index.php/Zhu\_Yu\_%E6%9C%B1%E6% 98%B1 (accessed 16 February 2012).

advantage of the space between morality and the law and based my work on it' (BBC Online, 2003).

In Xinjiang, Uyghur respondents interviewed in 2002 were aghast at the act, which was seen as incontrovertible proof that their stereotypical view of Han people as 'inhuman' was accurate. Subsequently, Muslim netizens across the world would question Zhu's assertion that no religion forbids cannibalism, and some sought the advice of respected Islamic luminaries online. To cite some of their responses, an Egyptian scholar and *da`iyah*32 opined:

Hurting a human being is prohibited in Islam. Moreover, the sacredness of a human being and of his body after his death is the same as during his lifetime. Not only has Islam protected man's dignity during his life but also it has guaranteed the protection of his honour after his death. Therefore, it commands washing, shrouding, and burying him, and forbids breaking his bones and all forms of aggression against his corpse, thus differing with the nations that cremate their people when they die. This fact is stressed in the hadith of the Prophet which states, 'Breaking the bones of a deceased person is like breaking his bones alive (in sin).' Man is not created to be eaten but to live and worship Allah. After his death, his corpse is to be buried, which is a sign of the honour and dignity that Islam grants to man. It is not permissible by any means to eat human flesh. Cannibalism has no place in Islam. It is an act against humanity and contradictory to human nature. It is no more than an indication of cruelty, barbarism, brutality that is an indication of abnormality and harsh-heartedness.33

A second prominent Islamic scholar and author wrote in a similar vein: 'Cannibalism is against civilization and is no more than a call for backwardness'.34

In the years since the shock art scandal, the suggestion that Han Chinese eat babies has developed into an urban myth of mammoth proportions. In 2004, the theme was continued in Hong Kong-made fantasy horror film, *Dumplings* (Dir. Fruit Chan Gor), which, according to one Japanese critic, provides a 'haunting and even realistic glimpse at human horrors'.35 In

<sup>32</sup> A *da`iyah* is defined as a person who invites people to understand Islam through dialogue.

<sup>33</sup> Sheikh `Abdel Khaliq Hasan Ash-Shareef. Retrieved from: http://www.onislam.net/ english/ask-the-scholar/crimes-and-penalties/mischief/169146.html (16 February 2012).

<sup>34</sup> Sheikh `Abdul-Majeed Subh, a graduate of Al-Azhar University in Cairo, the chief centre for Arabic literature and Islamic learning in the world. Retrieved from: http://www. onislam.net/english/ask-the-scholar/crimes-and-penalties/mischief/169146.html (16 February 2012).

<sup>35</sup> Film reviewed by Kozo (2004) at: http://www.lovehkfilm.com/reviews/three\_ extremes\_dumplings.htm (accessed 17 February 2012).

#### 116 chapter two

choosing to eat dumplings made from minced foetus as a means to restore her lost youth, the film's heroine, Ching, 'reveals the ugly vanity inherent in the human desire to remain young and beautiful.'36 Since that time, a sequence of disturbing images taken from Zhu Yu's installation and Gor's *Dumplings* has circulated on the internet and been emailed around the world, resulting in furious debate.37 New York-based *The Epoch Times* stoked the mood on the web by reproducing a sensationalist report originally published in a Hong Kong magazine. The report's anonymous author claimed that infant corpses and aborted foetuses had become the newest health and beauty supplements in Guangdong province, where 'gourmet body parts' were allegedly sought-after delicacies. He/she suggested that customers with the right connections could obtain the 'highest quality' human parts (more mature foetuses) and the 'prime' human part (a male foetus). In conclusion, the author notes that the many 'forced abortions' resulting from the PRC's one child policy creates opportunities for such 'atrocities', and asserts that there has been a 'complete lack of morality and respect for human life' in China since the Cultural Revolution (1966-76) (*The Epoch Times* 2007). This story cannot be corroborated in any other source; and, given that *The Epoch Times* is run by Falungong practitioners seeking to counter what they see as disinformation disseminated by Beijing, may have been politically motivated.38 The piece was later re-posted online by US pro-life (anti-choice) campaigner, Jill Stanek, who wrote in her blog: 'The following, if true, is what abortion and the dehumanization of preborn babies has wrought. It is the most despicable outcome of abortion I have ever seen or reported.'39 Her post was greeted as the perpetuation of a racist myth by bloggers on the prominent US feminist website, www.feministe.us. Moreover, one American Chinese contributor threw doubt on the story by pointing out that the original article had appeared in a Hong Kong tabloid. This tabloid, owned by a former Red Guard who emigrated after the Cultural Revolution, 'causes shrugs and chuckles and head shakes among Hong Kongers,' and is known for 'running outrageous

<sup>36</sup> Ibid.

<sup>37</sup> See the following 'urban legend' websites, where netizens discuss the truth (or otherwise) and morality (or otherwise) of the act: http://urbanlegends.about.com/od/horrors/a/ eating\_babies.htm; http://40ana.blogspot.com/2006/09/eating-aborted-foetus-in-chinatruth.html (accessed 18 February 2012).

<sup>38</sup> See Chen (2007) on media groups founded by Falungong followers.

<sup>39 &#</sup>x27;Sweet and Sour Fetus: Chinese cannibalism', 9 April 2007. Retrieved from: http:// www.jillstanek.com/archives/2007/04/sweet\_and\_sour.html (accessed 18 February 2012).

stories about alien babies being raised by communist coyotes'.40 As pointed out by Snopes, a well-known online reference source for urban legends, myths and misinformation, the tale sounds plausible when set in China, since it is common knowledge that the PRC government considers organ harvesting from executed prisoners to be a reasonable use of discarded assets.41

Whatever the truth, this controversy provided ample opportunity for Uyghur respondents in 2002 to consolidate their view of Han people as 'heartless'. By grounding their collective identity and dignity in a particular moral order (Lamont 2000), this subordinated group was able to reposition itself above the majority Han. Furthermore, the controversy later provided ammunition for Uyghur political activists seeking to publicise the treatment of group members by the Chinese state and mobilise hatred against the Han people. A video, uploaded onto myspace.com in 2009 by a netizen calling himself simply 'Uyghur', was designed to create a visual and conceptual link between the act of foetus consumption, Han on Han violence, and Han on Uyghur violence. The film opens with an image of a human foetus placed on a chopping board, before cutting to footage of a Han Chinese man violently beating two fellow Hans as they lie screaming on the ground. The scene is accompanied by the captions: 'Do they still have [a] human heart?' and 'Other Chinese stand there and watch like it's a show'. Next, the camera returns to images of foetuses being prepared and stewed, while the caption reads: 'Is this how they [Han people] lost their human heart?' In the next frame, the narrative continues: 'This is what the Han Chinese do to each other … imagine what they do to others? Like Uyghurs and Tibetans?' At this point, the film cuts to footage of the factory riot which took place in Guangdong in June 2009, and lingers for some time on the image of dead (or dying) Uyghur victims. The following caption demands: 'Is there any human being who can beat a person like that and eat a baby child like this?' Next, the film cuts to footage of a Han mob attacking a Uyghur male on the streets of Ürümchi on 7th July 2009. For the first time, the accompanying caption raises the question of territorial ownership, reading: 'Han Chinese hunting for Uyghur[s] *in the Uyghur hometown*.

<sup>40</sup> http://www.feministe.us/blog/archives/2008/01/04/and-here-i-thought-it-was-cats/ (accessed 18 February 2012).

<sup>41</sup> Retrieved from http://www.snopes.com/horrors/cannibal/fetus.asp (accessed 18 February 2012). The authors dub this type of recycled misinformation a 'peacetime atrocity rumour'. A very similar story was recently published as a 'Reader's Letter' in *The Seoul Times*, which said it had received the letter in the form of an 'email report' (2012).

#### 118 chapter two

They're desperate to kill more … [a] few hundred wasn't enough' [my emphasis]. In closing, the video returns to images of human foetuses arranged on dinner plates, as the final caption reads: 'God forgive us, we couldn't stop them'.42 In this way, the Uyghur film-maker deliberately sets out to establish a stark and shocking contrast between moral and immoral, humane and inhumane, right and wrong, Muslim and infidel.

# 2.7. *Stereotypes relating to Colonisation and Exploitation*

# *Pest and Plague*

The third set of stereotypes circulating in 1995-6 comprised representations of pest and plague, and was founded on local perceptions of Han colonisation, including both state-sanctioned population transfer and voluntary economic migration. As detailed in Chapter 1, in the 1950s the CCP quickly got underway a large-scale resettlement of Han people to the West. Already in 1958, *Xinjiang Daily* reported that new Han settlers in Qäshqär had been dubbed 'locusts' by local people, who blamed the incomers for rising unemployment. The ensuing social tensions caused many Han to try to flee the region and return to their home provinces (Schwarz 1963: 73). The choice of label—locusts—naturally evokes the spectre of devastation. In their so-called 'gregarious behavioral phase', locusts congregate into thick, mobile, ravenous swarms, causing major crop damage and human starvation. That a desert people should use such a metaphor to describe aggressive immigration and resource appropriation is unsurprising, for the desert locust found in Africa, the Middle East, and Asia is among the most notorious. A desert locust swarm may cover one-fifth of the Earth's surface and threaten the economic livelihood of one-tenth of the world's humans, with each locust eating its own weight in plants each day.43

During the course of my first field trip, Uyghur respondents frequently complained that they 'could go nowhere in Xinjiang without bumping into a Han'. Ghalip, a history graduate who ran a restaurant in Ürümchi, related in 1995 how he had learned in his college textbooks (in 1990) that Han people comprised 77 per cent of the population in the regional capital. By

<sup>42</sup> Retrieved from: http://www.myspace.com/video/uyghur/chinese-cruelty-eatinghuman-uyghur-riot-china-xinjiang/60436862 (accessed 16/2/2012). The video was released on 13th July 2009, just over a week after the riots in Ürümchi began.

<sup>43</sup> See the National Geographic factsheet on *Locustidae* at: http://animals.nationalgeo graphic.com/animals/bugs/locust/ (accessed 24 March 2011).

1995, he—like many other Ürümchi Uyghurs—had the strong impression that this proportion had dramatically increased, observing: 'It's totally changed now. The Hans are everywhere. *Just like ants*' [my emphasis]. He connected rising social injustice and inequality directly to this perceived influx (see Chapter 1). Other respondents interviewed in 1995-6 lamented that Han had begun to move into the countryside, whereas once they had settled only in towns. Like the locust metaphor above, the comparison with ants conjures the image of invasion and infestation. Ants are known for their abundance. Scientists estimate that they form between 15 and 25 per cent of the terrestrial animal biomass (Schultz 2000: 14028). They are especially notable among insects for their 'ecological dominance,' composing at least one-third of the biomass of global insect fauna (Hölldobler and Wilson 2005: 7411). The total biomass of the global ant population is approximately equal to the total biomass of the human race (Hölldobler and Wilson 2009: 5). Ants have a tenacious capacity for survival, attributed to a high level of social organisation and their ability to modify habitats, tap resources, and defend themselves. As a result, they have successfully colonised almost every landmass on Earth, with the sole exceptions of Antarctica and a few large islands (Hölldobler and Wilson 1990: 471). The notion of omnipresence was reflected in the questions Uyghur respondents asked upon meeting me for the first time, for example: 'Are there Hans in your country?' or 'Are there Hans in Spain / Cyprus / Germany / France / Japan / Kenya?'44 If I replied in the affirmative, they would exclaim: 'The Hans are in *every* country... all over the world!' At the same time, Han residents themselves did much to reinforce this idea by frequently referring to Han 'resilience' and the group's enhanced capacity for survival.

Inescapably, the ant comparison also highlights the concept of colonisation. While ants are tiny, apparently insignificant insects, they form colonies that range in size from a few dozen predatory individuals living in small natural cavities to highly organised colonies consisting of millions of individuals and occupying large territories (Oster and Wilson 1978: 21-22). They live in territory that humans consider to be theirs, and seek to consume resources that humans need (for example, crop plants). Some species are known to 'enslave' other ants and force them to work for the colony,45 while others are capable of forming 'supercolonies' that can threaten,

<sup>44</sup> These were all other countries I had worked, studied, or travelled in.

<sup>45</sup> Retrieved from http://insected.arizona.edu/antinfo.htm (accessed 25 February 2012).

#### 120 chapter two

displace or even kill endemic populations.46 Ants are skilled at communication, and slave-making species are known to employ 'propaganda pheromones' to confuse enemy ants and make them fight among themselves (D'Ettorre and Heinze 2001). As one UK pest control firm put it: 'Ants are empire builders, and their colonies are designed to grow at any expense.'47

A third stereotype—the description of Han people as 'imported bedbugs'—suggests both infestation (invasion) and a parasitic relationship between exploiter and exploited. While Han Chinese have been known to refer to Uyghurs as *chouse* (the Chinese loan for 'bedbug', derived from the Uyghur *chusa*), implying that Uyghurs are filthy and an irritant, the term has been employed much more powerfully in reverse. Uyghur respondents speaking in 1995-6 frequently levelled charges of Han 'occupation' of their territory (seen as the Uyghur homeland); 'theft' of commodities needed by locals for subsistence (food and water); 'stealing' of employment opportunities which might have been enjoyed by locals; and the exploitation of Xinjiang's natural resources. Educated respondents pointed most often to the appropriation of oil, connecting this directly to the demand for energy in China's developed coastal belt. To cite an example, a middle-aged male teacher in Ürümchi asserted in 1996: 'The Hans are exploiting our natural resources. They have always done so. You can see for yourself: empty trains arrive in Xinjiang. Then they leave, filled with oil and coal, bound for China proper.' Such criticisms were already being circulated at the end of the eighties, as reflected in the popular joke documented by Rudelson: 'The train [from Beijing to Xinjiang] is said to enter from China proper speedily into Ürümchi making the sound "ach, ach, ach" (I'm hungry), and to leave slowly making the sound "toq, toq, toq" (I'm full)' (1997: 75). Young, unemployed males and self-employed stallholders were especially vehement when this topic was broached. A migrant worker from Aqsu told me excitedly in 1996: 'Xinjiang is teeming with oil! But every day it leaves Qaramay in freight carriages for the east.' This young man had previously spoken of his aspirations to a future 'Eastern Turkestan', suggesting a clear link between local perceptions of exploitation and the growth of ethno-political sensibilities during this period.

<sup>46</sup> See the National Geographic factsheet on *Formicidae* at: http://animals.national geographic.com/animals/bugs/ant/ (accessed 25 February 2012).

<sup>47</sup> Retrieved from: http://www.rentokil.co.uk/pest-guides/insects-and-spiders/ants/ index.html (accessed 25 February 2012).

The characterisation of Han migrants as 'thieves' provides an interesting counter to the prevailing belief among Han that all Uyghurs are petty thieves. A social survey conducted in Ürümchi in August 2000 showed that 19.5 per cent of Han respondents thought Uyghurs were dishonest and unreliable compared with only 8.4 per cent of Uyghur respondents who thought the same of Han (Yee 2003: 440). The stereotype is also regularly disseminated through the mass media. The character Dark Cloud in the film *Crouching Tiger, Hidden Dragon* (2000) provides a recent example on a national and global scale. Dark Cloud stands as the representative of all Xinjiang minorities: he is a bandit who lives in a desert cave, rides a horse, is dirty and unkempt, and lives by stealing. As Sines points out, these features function to clearly distinguish him from the Han 'civilisers' (2002: 16). In reality, of course, while petty thievery exists in all societies (including Han society) the majority of individuals tend to abstain from such activity. The same is true of Xinjiang, where nineteenth century travel writers observed that both nomads and cultivators were 'generally averse' to crimes such as theft, robbery, suicide and murder (Warikoo 1985: 100-101).

Meanwhile, the bedbug stereotype provides a complex and detailed metaphor for the process of Han in-migration and resource exploitation. Bedbugs, which are spread more rapidly in this age of increased immigration and international travel, are extremely resilient; their young can survive for months without feeding, while adults can survive for more than a year. In the right conditions, they can develop fully in just one month, producing three or more generations each year. Feeding mostly on the blood of mammals, bedbugs are commonly found where humans have established cities and homes. While they tend to congregate in particular areas, it is also common to find a single bug or eggs scattered more widely. Their mode of exploitation is insidious; they are nocturnal creatures, biting at night while humans are asleep, with hosts rarely aware of their feeding (Potter 2008). The bites cause physical discomfort (allergic skin reactions, inflammation and intense itching) as well as psychological stress (sleeplessness, anxiety and embarrassment linked to social stigma) (Harlan 2003; Potter 2008). The available comparison is clear: the capacity for migration and rapid mobility of the bugs is equated with the flood of Han migrants into Xinjiang during the 1990s and 2000s. Their resilience, capacity for survival, and ability to multiply at speed is mirrored by the large Han population (one quarter of the global population) which in turn reflects the capacity of Han people to endure hardship and famine. Han in-migrants have built their New Towns directly next to the Uyghurs' Old Towns (and

#### 122 chapter two

in one case—Ghulja—practically on top of the Old Town). However, they are not confined to the towns; individuals and communities can also be found from time to time in remote areas of the countryside. Han people 'batten' on mutton produced by Uyghurs and other local peoples, while competing with them for limited water supplies (see Chapter 1). Meanwhile, the dynamic of invisible feeding on blood finds a sinister parallel in the hidden extraction of oil via the underground Lunnan-Shanghai pipeline (see Pocha 2006). This parasitism in turn causes Uyghurs practical and emotional discomfort. On one hand, their financial capacity to buy mutton is reduced, as the result of rising market prices owing to increased demand. On the other, it creates anxiety among local peoples both for their future livelihoods and for the fragile natural environment upon which they depend. Many Uyghurs experience in addition a strong sense of shame or regret linked to the fact that they could not/did not prevent the infestation in the first place.

#### *'Laiguazi'* (来瓜子)

One further label attached to new Han settlers can be connected directly to the notion of invasion: the term *laiguazi*. This is a variation on the Chinese word *shagua* [idiot, literally 'stupid melon'], and might perhaps be translated as 'idiot newcomer'. The neologism itself is quite comical, and reflects the tendency of many Uyghurs to tackle the issue of Han in-migration with black humour. Consider the following joke told by Änwär, a *minkaohan* interpreter living in Ürümchi in 1996:

There were four men in a train: an American, a Japanese, a Uyghur, and a Han. The American pulled out a pack of Marlboro cigarettes, smoked one, and then threw the rest out of the window. The others were amazed. 'Why did you do that?' they asked. 'Oh, you find Marlboro everywhere in the States,' he replied. Then, the Japanese switched on his tape recorder, listened to half a song, and chucked it out of the window. Again, the others were amazed. 'What did you do that for?' they asked. 'Oh, we've got tape recorders coming out of our ears in Japan,' he replied. Then, the Uyghur picked up the Han, and threw him out of the window. 'Don't worry,' he smiled. 'There are Hans all over the place in Xinjiang!'

The above bears a strong resemblance to the popular joke form in the UK and Ireland, which begins: 'An Englishman, an Irishman and a Scotsman …'. This 'three nationalities' joke format involves placing the characters in a scenario, and describing how each reacts to it, while playing on the stereotypes of the respective nationalities.48 It is also found in other regions, where it follows a similar formula, that is, the 'foreigners' are usually portrayed as arrogant, stupid and naïve, while the 'insider' is smart, practical and, ultimately, victorious.49 Such jokes function not only to define the 'Other' but also recursively in the definition of the 'Self' (Eriksen 1997).

In the Uyghur version, Han people are portrayed as so numerous as to be dispensable. Extending the implications beyond this general notion, the subtext suggests that Han are to all intents and purposes useless (the Uyghur does not make use of his Han passenger before dispensing with him), and something that local people would like to be rid of (the Uyghur flings the Han from the window). This in turn can be linked to the image of Han settlers as 'barren chickens'—parasites who eat all the grain but give nothing back—popularised in new folk song during this period (see Chapter 4). My story-teller grinned when he had finished relating this joke, and remarked: 'I shouldn't be telling you things like this. I'll be arrested as a counter-revolutionary!' While uttered in good humour, a grim reality underlay his words. Local resentment of Han migration to Xinjiang was in 1995-6 at its fiercest since the founding of the PRC. In voicing their dissatisfaction, Änwär and his peers every day risked indictment as 'local nationalists' (pro-secession and anti-Party). Interestingly, the term *laiguazi* was employed not only by local Uyghurs but also by first-generation Han settlers, distinguishable from recent migrants by their willingness to learn the Uyghur language and to adapt to local customs, including the prohibition on consumption of pork.

### *Conclusion: Talk Breeds Conflict (or the Power of Oral Complaint)*

As Lazzerini has noted, complaint can take many forms, depending upon circumstance (time and place) and individual or group personality:

From culture to culture […] and from period to period, the volume of complaint as well as its appropriateness (function) within the dominant discourse has varied, thereby illustrating again the marvellous diversity of human response to life' (1992: 154).

<sup>48</sup> In Ireland, the punchline is usually based around the Scotsman's meanness (or frugality) or the Englishman's snobbery. In the U.K., it plays upon the Irishman's stupidity.

<sup>49</sup> In Scandinavia, the joke begins 'A Swede, a Dane and a Norwegian …'. The Swede is always depicted as a rich and arrogant child of the Enlightenment, the Dane as a slightly decadent hedonist, and the Norwegian as an uneducated country bumpkin (Eriksen 1997).

#### 124 chapter two

Recalling Franz Rosenthal (1983), he argues that the role of complaint is 'characteristically Islamic': one of two poles of man's approach to life, balanced by the pole of hope. While this generally takes the form of 'routine protest', Islamic history has also witnessed complaint that 'episodically reached such a degree of emotional and/or intellectual intensity, as well as popular support, as to engender actions resulting, on occasion, in dramatic and enormously consequential change' (Lazzerini 1992: 154). In Xinjiang during the 1990s, the level of oral complaint among the Uyghurs attained a height unprecedented since the CCP's accession to power in 1949. Daily verbalisation of negative stereotypes of the out-group (essentialised as an immoral 'Other'), contrasted against positive counter-stereotypes of the in-group, resulted in a harsh 'Us and Them'-type ethnic dichotomisation (Eriksen 1993: 27). This in turn contributed to the rapid consolidation of ethno-national identity and the resurgence of aspirations to Uyghur independence, this time not just at the elite level. Where these had previously been restricted to urban, intellectual discourses, they were now circulated among different classes as well as among rural-urban migrants, who subsequently took them home to the countryside.

In their bid to exclude Han people from their social, cultural and moral circle, Uyghurs drew upon a situation they had shared for more than five hundred years (and, in the case of the southern oases, longer still): their *Relevanzstruktur* (Schültz 1972) or *habitus* (Bourdieu 1977). This 'structure of relevance', composed of taken-for-granted cultural norms, social practices and religious rituals, gives meaning to individual and collective 'knowledge' and interpretations; inhabiting past and present actions, it also structures future ones. In emphasising the failure of Han people to adhere to the Uyghurs' self-ascribed structure of relevance, respondents drew on a set of broadly Turkic-Islamic (Central Asian) values, rituals and traditions—what Anthony Smith might refer to as a 'symbolic repertoire' of cultural elements such as language, religion, customs and institutions (2009: 25)—to stress the impossibility of co-existence, and, for some, the urgent need to take action. In characterising Han settlers as pest and plague, Uyghurs defied state and media discourses of Xinjiang as an 'inalienable' part of China since ancient times, as well as notions of 'nationality unity'. They challenged Han assumptions of territorial ownership, highlighted socio-cultural disjuncture between the two groups, and redefined themselves as an oppressed party in socio-economic competition with a privileged one. Furthermore, in characterising Han as insects, respondents de-individuated and dehumanised an entire people, rendering

them faceless and, for a small minority of activists, deserving of extermination. As social psychologist Philip Zimbardo explains, in this way, dehumanisation, enacted through words and images, can form a 'cortical cataract' that clouds an individual's thinking:

To modify an old adage: Sticks and stones may break your bones, but names can sometimes kill you. The process begins with stereotyped conceptions of the other, dehumanized perceptions of the other, the other as worthless, the other as all-powerful, the other as demonic, the other as an abstract monster, the other as a fundamental threat to our cherished values and beliefs. With public fear notched up and enemy threat imminent, reasonable people act irrationally, independent people act in mindless conformity, and peaceful people act as warriors (Zimbardo 2007: 11).

This creation of a 'hostile imagination' is most observable in the propaganda of nations headed for war (Keen 2004). An extreme example is available in the Nazi comic books of Julius Streicher (news editor of *Das Sturmer*, or *The Storm Troop*). In drawing a comparison between handsome, strong Aryan men and ugly, weak Jewish men, and characterising Jews as vermin or insects riding on the back of the devil, these books sought to spread anti-Semitic propaganda among the general public. They successfully created a perception of Jews as sub-human and inferior, yet simultaneously as a threat to the nation (Zimbardo 2007: 297-323).

That stereotypes in 1990s Xinjiang could be so powerful, influential and damaging to ethnic relations is in part possible because the two groups hold very different sets of cultural and social values, based on contrastive belief systems: Islam versus (residual and resurgent) Confucianism fused with Daoism and Buddhism. It should in theory be possible to manage differences in a multi-cultural society; however, for mutual understanding to be achieved, a certain level of interaction between groups is necessary, not to mention truly supportive government policies, properly enforced. The 1990s saw levels of ethnic chauvinism among Han in-migrants escalate out of control, while the state imposed an increasingly draconian set of local minority policies that belied the guarantees of linguistic and cultural equality enshrined in the PRC constitution. As a result, Uyghurs have chosen to steadily restrict interaction; in the 1990s, they enforced segregation to a far greater extent than even most Han, as will be seen in Chapter 3. Consequently, there have been few opportunities for either group to gain insights into the other's cultural practices. Many of the negative stereotypes held derive from cross-cultural misunderstandings. For Han people, 'culture' is inextricably linked to education, where 'education' is understood as: a) mastery of the Chinese language; and b) knowledge of Confucian codes of behaviour (*li*). By extension, an education in a minority language and/or the Qur'anic teachings is automatically equated with a 'lack' of culture, hence the stereotype of Uyghurs as 'uncultured' (*meiyou wenhua* 没有文化). Han people interpret the Uyghur male proclivity to settle interethnic disputes through physical combat as 'uncivilised' (*meiyou wenming* 没有文明) because they are raised according to Confucian mores, which advocate the maintenance of social harmony. Most are unaware of the requirement in Islam to defend the honour of a Muslim brother (or sister) when the latter is slandered or done an injustice.50 Similarly, it is because Han people lack detailed cultural knowledge of Uyghur religious practices that they fail to understand that the 'small wash' is important because it renders a Uyghur 'spiritually clean'; they observe simply that Uyghurs attach less importance to the 'big wash' and thus conclude that they are 'dirty' in a bodily sense.

Viewed in reverse, Uyghurs find it ridiculous that a slender male figure might be considered desirable among the Han because most are unaware of the central binary of civil and martial virtues (*wen-wu*) in East Asia, and of the primacy traditionally given to civil virtues (*wen*) in China. While the martial aspect (*wu*) has long been viewed by Han as a feature of the 'barbarian' (Benesch 2011), notions of Islamic masculinity often include requirements to fight or martyr oneself for the Islamic (or religious nationalist) cause or to protect the family honour (Ouzgane 2006: 4-5). Against this contradictory backdrop, even the minority of Uyghurs familiar with pre-modern Chinese literary discourse struggle to understand why the physical frailty, delicate and feminine features, and vulnerability of the *caizi* ('fragile scholar', 才子) figure in the popular genre of 'scholar-beauty romance' embodied a set of desirable masculine qualities (Geng 2004: viii). In the same way, Uyghur women cannot identify with the Chinese concept of *nüse* (female beauty), because it differs so dramatically from the (albeit unspoken) Islamic female aesthetic of womanly curves. As outlined in Confucian and Daoist literary and philosophical texts, women should be young, small and slim, have drooping shoulders and smooth white skin, and be elegant in gesture and behaviour (Man 2000: 170). Uyghur percep-

<sup>50</sup> See the Sunni Qur'anic exegesis *Tafsir al-Baghawi* (ca. 1122), written by Husain bin Masood al-Baghawi as an abridgement of *Tafsir al-Tha'labi* (ca. 1035). As recorded in *Sahih Muslim*, the Prophet Muhammad taught that all Muslims should act as one body in their love and concern for each other, since 'when any part of the body suffers, the whole body feels the pain'.

tions of Han people as 'killjoys' are not based on a primordial Han disinclination to have fun, but rather on the Han disinclination—and inability—to perform (sing, dance, play) *in the ways sanctioned by Uyghur tradition*. One has only to consider the popularity in China of karaoke or ballroom dancing, the practice of *taijiquan* (Tai Qi) in public parks, and the conduct of *yuanji* dance exercise on street corners,51 to know that Han people can also be willing performers in the right cultural context. Finally, Uyghurs raised to follow Islamic social practices, including dietary prescriptions, ritual ablution, life cycle rituals, etc., cannot fail but find many Han customs disrespectful and filthy (that is, 'impure' in a spiritual sense). To a Uyghur, Han consumption of pork and other *haram* foods, the washing of hands in water that is already dirtied, the display of bad habits (and their ill effect on the social environment), and the apparently flippant treatment of the human body after death, all represent the spiritual *impurity* of Han people as well as their lack of respect for living creatures either in life or in death.

However, it was not social and cultural disjuncture that most deeply provoked anger and bitterness in Xinjiang in the 1990s, but Uyghur perceptions of Han invasion and exploitation. Respondents connected the steady influx of Han migrants to the region directly to increased competition for land, jobs, commodities, and natural resources. Of these, the exploitation of oil and coal for the benefit of Han people living on the eastern seaboard, and the privileges accorded Han migrants in Xinjiang's fast-developing urban labour market were cited most frequently. Each of the dynamics described in this chapter can be framed with reference to Tajfel's (1981) typology of the functions of social sterotypes, as follows:

1. *Social causality*: a wide range of stereotypes were employed in 1990s Xinjiang to help *explain* a troublesome problem or event. That is, Han in-migrants were blamed for employment discrimination and rising unemployment; increased competition for land and water; rising meat prices; spiritual pollution of the environment (i.e. through the raising of pigs in ritually 'pure' territory); environmental pollution and resource degradation; depletion of local natural resources through extraction and export, and so on.

<sup>51</sup> *Yuanji* dance is an exercise that integrates martial art forms, fitness, music, dance, and entertainment. With its simple movements and its elegant form, it is widely considered a suitable daily exercise for both old and young.

#### 128 chapter two


In sum, these dynamics recall the theory of anthropologist, Abner Cohen, who argued that while ethnic identity is instrumentally deployed in the competition over scarce resources, it is nevertheless *circumscribed by ideologies of shared culture* (1969; 1981).

In his discussion of Han-Hui relations during the Qing period, Raphael Israeli observed that 'in the eyes of the thinking Chinese literati […] the gap between the two communities might have looked so hopelessly unbridgeable that they could easily have come under the sway of popular stereotypes, which *tended to make the gap look wider*. These stereotypes, which grew to grossly exaggerated misjudgements of the Muslims by the Chinese, were constantly fed by the alienation between the two communities, as reflected in contemptuous name-calling, vicious story-telling and eventually in pogrom-style onslaughts when the opportunity presented itself' (1984: 280-81) [my emphasis]. He went on to note that where stereotypes are concerned, 'it is popular belief, not genuine fact, that matters' (Israeli 1984: 282). In the same way, exaggerated stereotypes of Han people in 1990s Xinjiang, conceived partly in response to Han chauvinist representations of Uyghurs and partly in response to misguided state policies, were self-perpetuating and supported the growth of a deep-seated popular hatred towards Han people. This hatred would twice build to the point of explosion over the next decade and a half, triggered first by religious repression, leading to the Ghulja disturbances of 1997, and then by discrimination and injustice, leading to the Ürümchi riots of 2009. Susan Blum in her 2001 study cites a Han respondent as conceding: '...although the government said, we are the great family of nationalities, we should have national unity […] I think, we can't understand each other completely, we unite to-

gether, but have a distance' (2001: 131). In the next chapter, I will illustrate how hatred, established through universal stereotyping, led to the deliberate construction and maintenance of ethnic boundaries by Uyghur communities and ultimately to ethnic segregation in 1990s Xinjiang.

#### Chapter three

130 chapter three

# Making Culture Matter: syMboliC, spatial, and soCial boundaries1

author: do you have any han friends? shökhrät: of course not! author: Why not? shökhrät: i don't want to make friends with hans. i hate them. even if i did make friends with a han, he would still act in a superior way towards me.

(interview, 1995).

as described in Chapter 1, early han migrants tended to settle in newly opened areas in the north—new towns—separate from the old towns inhabited by uyghurs and other indigenous groups. Contacts between han newcomers and local peoples were therefore initially limited. those han settlers who did venture into the south of the region or into rural areas generally opted to follow the local ways of life, learning the uyghur language, adjusting their dietary habits and getting along well with their neighbours as a result. later migrants, however, settled not only in the northern and urban areas but increasingly in the southern and rural areas as well. at the same time, southern and rural uyghurs migrated to the regional capital Ürümchi and other urban centres, lured by economic development and new opportunities in education, employment and trade. as a result, contacts between the two groups increased dramatically, and there was a higher instance of boundary crossing. My interview data suggest that the increasingly chauvinist attitudes and behaviours of han migrants towards minority peoples (and local apprehension of these attitudes), along with increased competition for resources, lie at the heart of deteriorating uyghur-han relations. this can be clearly seen from the conversation presented at the start of this chapter: the respondent's indignant reaction to the question of whether he had any han friends emphatically underlined that hostilities derive not alone from religio-cultural differences and socioeconomic rivalries but also from the unbearable sense of superiority expressed by recent newcomers.

<sup>1</sup> an earlier draft of this chapter was published as an article in the refereed journal, *Asian Ethnicity* (smith 2002).

#### making culture matter: symbolic, spatial & social boundaries 131

Sino-centric attitudes sometimes provoked serious altercations in the 1990s, as young, resentful Uyghur males resorted to violence to settle disputes. While conducting fieldwork in Aqsu city in 1996, I heard from Uyghur respondents that a local Uyghur boy aged just 15 had been imprisoned for stabbing three Han residents who ridiculed him in a general store. Yet by far the more common response was a strong disinclination among Uyghurs to interact with Han people in any way except of necessity. If ethnic boundary-setting is the embodiment of nationalist (and ethno-nationalist) ideology and practice (Duara 1996), then the construction and maintenance of Uyghur-Han boundaries in 1990s Xinjiang can be viewed as an act of ethnic incorporation, reflecting the strengthened Uyghur (ethno) national identity as well as Uyghurs' outright rejection of the 'pragmatic nationalism' (Zhao 2004) pursued by the Chinese state. At the same time as circulating negative stereotypes of Han people and using these discourses of hate to distance the in-group from the out-group (see Chapter 2), Uyghurs simultaneously sought to construct and/or reinforce an intricate system of ethnic boundaries, designed to strictly delimit the conditions under which Uyghur and Han people could interact and to enforce segregation where they should not. My focus in this chapter is therefore how boundaries constructed by the dominated unfolded in everyday interactions (or non-interactions), and how these were mobilised by individuals to make claims concerning their moral worth and superiority in relation to majority group members (Lamont and Mizrachi 2012: 366-367).

# 3.1. *Barthian Theory: Ethnic Boundaries as Resistance*

Criteria for cultural difference may potentially include: racial uniqueness (some sense of genetically inherited difference), place of origin, economic independence (secured by community organisation within the plural society), religious beliefs and practices, aesthetic cultural forms (e.g. food, dress, dance), and language. In choosing some or all of these criteria as symbols of ethnic identity, group members demonstrate social agency by defining the way in which they differentiate themselves from other groups (De Vos 1975: 16). Anthropological studies have shown that a given set of cultural criteria is rarely maintained in its entirety through time and space (Barth 1969: 12). Most ethnic groups will have included cultural forms in the past that are clearly excluded in the present; in the case of the Uyghurs, one example would be Buddhist practice. Similarly, while a group spread

#### 132 chapter three

over ecologically varying territory will sometimes display regional diversities of cultural practice, self-identification as a group member may nonetheless continue; in Xinjiang, this manifests itself in oasis-based cultural specificities which are nonetheless subsumed under a broader regional and religious identity.

Ethnicity, defined as the will to express a distinct ethnic identity, is not an isolated phenomenon but a relative one. Ethnic distinctions cannot exist in a vacuum of contact and information; rather, they entail social processes of exclusion and incorporation embodied in the construction and maintenance of ethnic boundaries (Barth 1969: 9; Eriksen 1993: 10–12, 35). There can only be 'insiders' where there are also perceived 'outsiders'. As a result, for ethnicity to develop, an ethnic group must be in regular contact with another group or groups from which it considers itself substantially different, or against whom it has reason to seek to differentiate itself. Ethnic boundaries—like ethnic identities—are fluid and negotiable. They appear, change shape, and vanish again in relation—and reaction to changing social, political, and economic contexts. Accordingly, group members may employ different cultural criteria at different times, and in different places, in order to define themselves against different groups (cf. Gladney 1996).2 Because of its focus on fluidity and negotiation, ethnic boundary theory is extremely useful when considering the impacts on Uyghur identity and Uyghur-Han relations of the diverse socio-economic changes unfolding in Xinjiang over the past twenty years.

In the context of economic and political argumentation over rights and interests, identity is sometimes construed as being 'more about boundaries than about contents' (Blum 2001: 13). Yet, as the previous chapter demonstrates, the contents of stereotypes are vital to their successful deployment, and the same is also true of boundaries. When peripheral peoples are confronted by a hegemonic centre that seeks to define and educate them according to its own cultural standards, they are forced to come to terms with who they are and how they are different from the civilisers. In doing so, they draw upon their 'pre-project consciousness' (Harrell 1994: 28-9): group awareness of shared practices and norms predating the advent of the ci-

<sup>2</sup> In his 'relational alterity' model, designed to explain shifting, transnational Hui, Uyghur, and Kazakh identities, Gladney argues that 'people subscribe to certain identities under certain highly contextualised moments of social relation'. Locating the emergence of national identities in Central Asia within this field of contemporary and historical social relations, he shows how different identities are constructed in different social contexts vis-à-vis a number of different 'opposites'.

# making culture matter: symbolic, spatial & social boundaries 133

vilising project. It is here that practice theory becomes central in the study of ethnicity (Bentley 1987). As Central Asia anthropologist Jo-Ann Gross has observed, 'it is not enough to look merely for the primordial cultural symbols or the situational manipulation of identity. Rather, pre-existing shared values and commonalities, as personalized individual and group experiences, need to be brought back into the picture' (1992: 10). Common practices, learned in childhood, internalised, and repeated to the point where they become unconscious, both produce and reinforce a group's history of common experience, constituting a body of knowledge that bonds individuals to their ethnic group. This body of knowledge has been dubbed variously the *Relevanzstruktur* (Schültz 1972), 'habitus' (Bourdieu 1977) and 'cultural hegemony' (Comaroff and Comaroff 1992). Individuals who have grown up in a common environment and share common experiences know implicitly why other group members behave in certain ways, so that ethnicity becomes *implicit in practice*:

One of the fundamental effects of the orchestration of habitus is the production of a commonsense world endowed with the objectivity secured by consensus on the meaning (sens) of practices and the world, in other words the harmonization of agents' experiences and the continuous reinforcement that each of them receives from the expression, individual or collective (in festivals, for example), improvized or programmed (commonplaces, sayings), of similar or identical experiences. The homogeneity of habitus is what [...] causes practices and works to be immediately intelligible and foreseeable, and hence taken for granted (Bourdieu 1977: 80).

Or, to borrow the words of psychoanalyst Theodor Reik, individuals identify with in-group members because they share taken for granted meanings derived from an 'intimacy of experience' and a common 'rhythm of living' (Reik, cited in Epstein 1978: 140).

Central to the process of boundary construction in 1990s Xinjiang was the activation and exaggeration of religio-cultural (or ethno-religious) differences between the two groups—primarily, the consumption or avoidance of pork. Yet these differences did not prevent interaction to the same degree in the past; we saw in Chapter 1 how Uyghur children had once happily entered Han homes to play, provided that their dietary habits were respected. Also of significance is the fact that these same ethno-religious differences do not prevent Uyghurs from making pragmatic concessions when it serves their interests to do so. A prime example concerns *haram* and *halal* foods. Although Uyghur respondents were adamant that they would never eat in a *Hancan* restaurant (one serving Han cuisine, including

pork), Uyghur restaurant owners were perfectly willing to serve Han customers in their own establishments, apparently turning a blind eye to the 'contaminating' effect of Han use of their facilities (crockery, cutlery). Two scholars provide similar examples of this type of negotiation: while the Uyghur intellectuals Cristina Cesàro knew in Ürümchi in the mid-nineties rejected out of hand the possibility of entering a *Hancan* restaurant on the grounds that they were not *qingzhen* (halal), they happily frequented Western-style restaurants (which also listed pork dishes on their menus) since this meant that they could make strategic connections with English-speaking Caucasians (2000: 232). Jennifer Taynen, meanwhile, observed that none of her Uyghur acquaintances ever quizzed her about her cooking practices during her residence in Ürümchi, while all consumed refreshments prepared in her kitchen (2006: 53). In other words, cultural concessions can be made for a politically sympathetic interlocutor but are impossible for one perceived to be hostile.

### 3.2. *Symbolic Boundaries*

Ethnic boundaries between Uyghurs and Han in the 1990s were of three types:


Symbolic boundaries have been defined as 'conceptual distinctions made by social actors to categorise objects, people, practices, and even time and space […] Symbolic boundaries […] separate people into groups and generate feelings of similarity and group membership' (Lamont and Molnar 2002: 168). In Xinjiang, symbolic boundaries were constructed around language; time; endogamy; and the state-imposed birth control policy. With respect to language use, most urban Uyghurs distinguished clearly between the home and the outside environment, corresponding respectively to spaces in which they spoke Uyghur and spaces in which they consented to speak Chinese. Time was the vehicle through which Uyghurs and other local peoples underlined their relative belongingness to the land, ordering

their day according to the local time congruent with the region's topological position. The taboo on intermarriage underlined Uyghurs' religious and cultural identity, and embodied local resistance to central attempts to assimilate them bodily (genetically). Opposition to the one child policy similarly stressed the Islamic view that every child is sacred, while at the same time resisting changes in the regional population composition.

Perhaps the most powerful symbolic boundary during this period concerned *language use and language domain*. As noted by Dwyer: 'Language policy affects the domains, status and use of language varieties and the rights of their speakers. It […] provides a rallying point for or against ethnic identity' (Dwyer 2005: 6). As such, marginalised Native Americans in the US place a rhetorical emphasis on language when describing the cultural dimension of symbolic group boundaries (Vasquez and Wetzel 2009: 1568). Uyghurs in the mid-nineties demonstrated their preference for their mother tongue by making clear distinctions between the domestic and external environments, corresponding to times when they spoke Uyghur and times when they spoke Chinese (cf. Yee 2005: 39). These rules were followed regardless of an individual's level of fluency in Chinese, the sole exception being a small number of *minkaohan* (Chinese-educated) young people, who tended to have a stronger foundation in Chinese than in their native language and code-switched at home. Although many urban Uyghurs, particularly Ürümchiliks, are fluent or semi-fluent in Chinese, most emphasise that Chinese is a 'language of practical convenience'. Take, for instance, Tömür, a first-generation *minkaohan* in his forties employed by a state work unit in Ürümchi. He became very excited when I asked him in 1995 which language he generally used, and outlined language boundaries as follows: 'Oh, we may speak Chinese outside... but we all speak Uyghur as soon as we get home!' Here, 'outside' denotes environments in which Uyghurs are obliged to interact with Han people who do not speak Uyghur, usually state work units or Han-managed private companies.

I observed that in university dormitories, *minkaomin* (Uyghur-educated) students spoke Uyghur to one another in the mid-nineties. They also spoke Uyghur to *minkaohan* roommates until the latter showed signs of discomfort. Then, they would navigate the conversation in a mixture of Uyghur and Chinese, or entirely in Chinese (depending on the language ability of their interlocutors), until the point had been grasped. In this way, the rules were negotiated slightly in order to accommodate the *minkaohan*, who sometimes struggled to express themselves in Uyghur, especially in academic contexts.

#### 136 chapter three

On the street, it was taken for granted that conversations with other Uyghurs were carried out in Uyghur, regardless of whether the other person was an acquaintance or a stranger. In the Uyghur-dominated Döngköwrük district of Ürümchi, Uyghur people always spoke to me in Uyghur first, breaking into Chinese only when they perceived that I had not understood their point. The best examples of this type of boundary maintenance came from first-hand observations of interactions between Uyghur individuals. For example, on one occasion in 1995 I witnessed a Uyghur man approach the counter in Ürümchi's Hongshan department store and, without looking up, begin talking to the store assistant (whom he assumed to be Han) in Chinese. Halfway through his sentence, he glanced up, saw that she was Uyghur, and instantly switched to speaking in Uyghur. They both laughed, and he immediately apologised for his mistake. On another occasion in the same year, I boarded a bus in Ürümchi with Räwiä, a language specialist, and she addressed the Uyghur bus conductress (whose facial features were closer than usual to those of Han) in Chinese. Seeing the woman's expression change, Räwiä quickly realised her mistake, changed to Uyghur halfway through her sentence, and apologised profusely. Although Räwiä is herself *minkaohan* and therefore more comfortable with the Chinese language than with Uyghur, she was quick to correct her mistake, suggesting that this type of boundary is not porous but solid. Her behaviour represented a public acknowledgement of mutual origins as well as shared differences vis-à-vis the Han. On another occasion in 1996, I was eating in a Uyghur street restaurant in Turpan in the company of a young male intellectual from Ürümchi named Shökhrät. It was deep winter, and I was dressed warmly in clothing typical of an Ürümchilik woman. Since I had only just begun learning Uyghur at that time, we conversed in Chinese for the sake of convenience. As we left, we overheard two old Uyghur men in the corner mutter to one another: 'But she's a *Uyghur* girl! Why on earth is she speaking in Chinese?' To them, it was inconceivable—and incorrect that two Uyghurs should speak Chinese to one another while eating lunch in a non-Han domain.

Whether they wanted to or not, many Uyghurs in 1995-6 felt forced to master Chinese in order to better compete with Han migrants. However, respondents unanimously described the decision to learn Chinese as a strategic career choice, emphasising that the only way to succeed in the Han-created urban society was through the acquisition of Chinese. In this way, study of the Chinese language was viewed as a means of socio-economic survival. Regarding the growing trend of placing Uyghur children in

#### making culture matter: symbolic, spatial & social boundaries 137

Han-medium schools, Tömür said in 1995: 'I'll tell you, there's only one reason why we learn to speak Chinese, and that's just to get a better job!' This was a view echoed by Räwiä and many other Uyghur parents in Ürümchi throughout that first field trip. Aliyä, a female postgraduate studying in Ürümchi in 1996, further explained: 'We speak Chinese because of its dominant position in this society. There is no way around it. Uyghur is not [considered] as important as Chinese now'. Yet despite her recognition of this fact, she spoke in Uyghur at all times except when speaking to Han people in public places or when intermingling Uyghur and Chinese for the sake of *minkaohan* dorm-mates.

Most respondents in 1995-6 (again, with the exception of some *minkaohan*) preferred to read and write in the Uyghur script, whether that meant the modified Arabic script (often dubbed the Old Script) or the Latinised script (the New Script).3 Periodicals and journals containing minoritynationality literature had mushroomed in Xinjiang since the re-introduction of relaxed minority policies in the early eighties (Mackerras 1985: 77), and in 1986 at least twelve journals were being published in the Uyghur language (Naby 1986: 243). By the time of my arrival in 1995, over seventy different publications appeared in Uyghur, indicating a large audience of literate, urban Uyghurs who increasingly preferred to read in their native language. There were also a large number of Uyghur-language newspapers in addition to several Uyghur-language television channels. Most southern Uyghurs watched the Uyghur-language channels in preference to the Chinese ones, although Ürümchiliks (who tended to be at least semi-fluent in Chinese) also liked to watch soap operas or dramas on the Chinese channels. Owing to discontinuities in state policy towards minority scripts in Xinjiang (see Chapter 1), Uyghur scholars held conflicting opinions on whether Uyghurs should use the modified Arabic script in use until 1956 and reinstated with modifications in 1981, or the Latin script introduced in 1958 and popularised during the Cultural Revolution. Some academics I interviewed in 1995-6 said they wished that the Latin script had been retained (cf. Bellér-Hann 1991: 75). For example, Sultan, a specialist in literary

<sup>3</sup> An Arabic-based script had gradually become dominant in the region between the tenth and fifteenth centuries as Islam spread eastwards. This script, which had earlier been adapted for Persian, was modified for Chagatay, the *lingua franca* of the medieval Central Asian Turks, and has predominated up to the present day. The script was first revised for Uyghur in 1925, and after further revisions in 1982 and 1987 was formally adopted (Dwyer 2005: 19).

studies in his fifties, lamented the consequences of script discontinuity over the past forty years:

We should not have reverted back to the Old Script in 1981. Everyone had just got used to the Latin script, and it was easy to use. Languages and scripts are very precious. You can't go chopping and changing them on a mere whim. Too much damage has been done already.

Younger scholars complained that the Old Script was 'inconvenient'. For example, Shökhrät pointed out in 1996 that if young Uyghurs wished to learn a foreign language, they had first to learn the Latin alphabet. He also argued that international communication (by which he meant political alignments with Anglo-European countries) would be easier if Uyghurs still used the Latin script rather than the modified Arabic script. Other respondents speaking in the same year reasoned that the Old Script was extremely difficult to put into a computer, whereas the New Script was 'cleaner' and more immediate when using information technology.4 The religious renewal of the 2000s, however, has brought about a shift in script attitude, so that 'nearly all' Uyghurs now favour the Arabic script, owing to its fundamental association with Islam (Dwyer 2005: 19).

Urban Uyghurs, and particularly intellectuals, liked to turn the tables on Han chauvinists in 1995-6 by making the Chinese language the object of ridicule. One example concerned the idiomatic expression *Manman zou!* (lit. 'Walk slowly'), uttered to a person leaving a particular location. This expression is understood to mean something like 'Mind how you go' or 'Take it easy' in English. However, Uyghur respondents deliberately interpreted the phrase literally, as when Sultan, a specialist in literary studies in his fifties, joked in 1995: 'Why would anyone want to walk slowly? They should walk quickly…otherwise they will never get to where they are going before night falls!' Others suggested that, as a phonetic language, Uyghur is more modern than Chinese, as when Aliyä, a postgraduate student in Ürümchi, argued in 1996:

Languages based on phonetics have always been among the most advanced... Their script [Chinese characters] is actually really backward! All the earliest scripts were made up of pictographs. All they've done is taken the pictures, squared off the curved edges, put them into neat squares, and continued using them!

<sup>4</sup> See Dwyer (2005: 23-4) on how the advent of email has stimulated adaptations of Latin-based scripts—*Uyghur kompyuter yeziqi / Uyghur Internet yeziqi*—purely for internet use.

#### making culture matter: symbolic, spatial & social boundaries 139

That the same idea was on a separate occasion echoed by Aliyä's aunt suggested that this student's opinions had been strongly influenced by her family elders. Räwiä, a *minkaohan* language specialist speaking in 1995, ridiculed the fact that Han people did not even attempt to imitate Uyghur phonetics, instead substituting them with the sounds in Chinese that approached the Uyghur word most closely. Other respondents mocked Han people for adopting the same strategy when learning English. For them, this behaviour proved that Han people were not only less intelligent but also less flexible and less creative than Uyghurs. Sometimes, this type of counter-chauvinism could also be found on the lips of urban migrant workers, as when one petty entrepreneur in Ürümchi suggested in 1995 that the Chinese vocabulary was deficient because it often used combinations of words to express a concept for which there existed a single word in Uyghur: 'They've got all those characters, and yet they still can't make a word for 'puppy' or 'kitten.' They just say 'little dog' and 'little cat'!' He also implied that Han people were stupid because they apparently could not find simple phonetic loans for the names of foreign countries. Using the example of America, he contrasted the semantic translation in Chinese *Meiguo* [The beautiful country] negatively with the Uyghur phonetic loan *Amerika*. The fact that Uyghur respondents in the mid-nineties strove to compare the Chinese language unfavourably with Uyghur indicates both an awareness of Han chauvinist attitudes towards minority languages and a strengthening of Uyghur linguistic identity in response.

Yet in rural areas language boundaries were more blurred. Uyghur peasants living in relatively Han-free areas and untouched by increased competition and discrimination in education and work attached no stigma to the Chinese language. Many had received little education in either Uyghur or Chinese, and appeared glad to show off their limited knowledge of the Chinese language. In 1996, Ömärjan, a peasant in his sixties in Aqsu, translated all the terms for the crops he farmed into Chinese for me, with evident delight. For him, the ability to speak a few words of Chinese carried both novelty value and prestige, and was a source of pleasure and pride.

*Time* was the vehicle through which Uyghurs and other Central Asians in Xinjiang underlined their belongingness to the land vis-à-vis Han and Hui 'newcomers' in the mid-nineties. I was amazed to observe upon my arrival in 1995 that while the timepieces of one half of the population read 10 a.m., those belonging to the other half read 8 a.m. Further investigation into this strange circumstance revealed that Han and Hui people used central 'Beijing time' while Uyghurs, Kazakhs, Uzbeks, Tajiks, and others

used 'Xinjiang time', the local time congruent with the region's topological position and two hours behind (cf. Bellér-Hann 2001a: 57; Bellér-Hann 2002: 60). As Änwär, a *minkaohan* interpreter in his thirties from Ürümchi, explained in 1995, this deliberate distinction was made to underline indigeneity to the territory: 'You see, all the original inhabitants (Ch. *tumin*) of Xinjiang use local time. It's what we're used to.' By 'original inhabitants', he indicated people of Central Asian origin who had lived in the region for centuries. I therefore set my watch to local time after just a few days in the regional capital, prompting the American ethnographer, Jay Dautcher, to smile and remark: 'Well...what other time is there?' The gesture was also met positively by Uyghur respondents. When Räwiä noticed this act, she smiled, nodded, and said: 'When in Rome, do as the Romans do' (Ch. *Ru xiang sui su* or 'When you enter a village, follow the customs').5

In 1980, when conciliatory minority policies once more gained precedence over assimilatory policies, the Xinjiang People's Congress had taken the decision to switch to 'Ürümchi time' (another term denoting local time, two hours behind Beijing time); however, this policy change had to be abandoned in the face of resistance by Han settlers (Veg 2009a). During the initial weeks of my first fieldwork period in Xinjiang in 1995, it became clear that while state work units—including companies, shops, railway stations, and so on—used Beijing time, peoples of Central Asian origin were unanimous in their continued use of local time and their rejection of the official time designated by the hegemonic centre. They persisted in using local time despite the obvious inconveniences, for example, the constant need to check understanding with the question: 'Do you mean Xinjiang time or Beijing time?' The time issue was further complicated by the fact that Hui Muslims did not bother to differentiate between the two terms 'Beijing time' and 'Xinjiang time'. When asked which system they used, they usually replied 'Xinjiang time' or even 'Ürümchi time', yet close questioning revealed that they actually meant Beijing time. When I queried who did use local time, Hui respondents either declined to reply or stated that Uyghurs and Kazakhs preferred to use local time, describing them as 'the same kind of people' [i.e. indigenous Turkic Muslims]. The fact that Hui Muslims, like the majority Han, consider Beijing time to be the standard time for all

<sup>5</sup> Blum recorded a similar discrepancy in the mountainous areas of Western Yunnan, commenting: 'How could decisions in such a far-off place [Beijing] affect the texture of life here? The region did not even switch to daylight saving time in the summer; people called it "Beijing time"' (2001: 4).

regions of China, including Xinjiang, doubtless exacerbates the mutual mistrust between Uyghurs and Hui.

Some Uyghur respondents did however use Beijing time in 1995-6, in most cases because their job required them to, and so they had become accustomed to it. These individuals tended to work in state work units where time was a key feature of everyday routine (such as a train station or travel agency), or else their spouses did. Some were *minkaohan* (Chineseeducated) Uyghurs, who had attended Han schools throughout childhood and thus been exposed to Beijing time on a daily basis. That said, most Uyghurs working in state work units stubbornly set their watches to local time and calculated the time difference, a circumstance that obliged Han co-workers to compromise by coming to work at 10am Beijing time (rather than 8am as they would have done in east China) (cf. Bellér-Hann 2002: 60).

The time situation continues to stun visitors to the region. In the early 2000s, a contemporary adventurer arrived at a hotel in Tashqurghan to find fixed to the wall four large clocks, which gave the respective times for Islamabad, Paris, Qäshqär and Beijing. Noting that the time on the Beijing clock was two hours ahead of that on the Qäshqär clock, he comments: 'I wasn't surprised. Here we were as close to Europe as we were to the Chinese capital' (Bealby 2004: 75). As recently as 2009, a UK reporter remarked:

Even the time is out of joint—while the clocks are required to display official Beijing time, for practical purposes Xinjiang is two hours' behind, so that the streets are dark and deserted at eight in the morning and bright and alive with people at midnight (Parry, 2009).

*Birth control* was at the heart of a particularly stringent religious boundary between southern/rural Uyghurs and Han in 1995-6; indeed, it constituted perhaps the main focus of southern opposition to Han rule.6 The extreme sensitivity of the issue is clearly observable in the caution with which the state applied family planning policy in the region. According to Sautman, while Han residents of Xinjiang were obliged to practise birth control from 1975 onwards, restrictions were not imposed on non-Han groups until 1983.7 Inter-group differences can be seen in the markedly different birth rates

<sup>6</sup> It is generally conceded that Islamic beliefs are stronger in the south than in the north of the region. Historically, the southern oasis of Qäshqär was greatly influenced by the Islamic centres of Bukhara and Samarkand (Rudelson 1997: 49).

<sup>7</sup> At that time, the state had a projected population target of 1.2 billion by the year 2000. Within this, the quota for Xinjiang was 17 million (Li 1990: 71).

recorded for 1984. In Shihezi (with a 96 per cent Han population), the birth rate was just 5.99 per thousand, compared with much higher birth rates in Uyghur-dominated cities: 29.28 per thousand in Khotän, 27.5 per thousand in Turpan, 25.31 per thousand in Qäshqär, 22.33 per thousand in Aqsu and 21.21 per thousand in Ghulja (Li 1990: 53). Wang gave statistics for 1989 that similarly showed far higher birth rates and fertility rates in southwest Xinjiang—Qizilsu, Qäshqär, Khotän—than in the north, with the highest rates among the Uyghurs (2004). When birth control was introduced to non-Han groups, initial policy permitted urban minorities a maximum of two children (unless both were girls or one disabled) and rural minorities a maximum of three (or four in remote areas). This was greeted in 1985 by a riot in Ürümchi and demonstrations by Uyghurs resident in Beijing and Shanghai. In the regional capital, approximately 1000 Uyghur students demonstrated before the Regional People's Government Committee Building on 10th December to demand that birth control policy be applied only to Han migrants in Xinjiang, and not to the far less populous indigenous groups (Li 1990: 71).

Some compromise was subsequently introduced into the regulations in 1991 and 1992, so that minorities living in border or mountainous areas of Qäshqär prefecture were allowed four—and occasionally five—children. However, local people continued to complain that family planning was 'too lax' among the floating Han population in that prefecture (Sautman 1998: 100). Meanwhile, the fact that approximately 40 per cent of minority children in Xinjiang were born outside of the birth control plan in the early 1990s signalled continued resistance (Sautman 1998: 89-90). As Uyghur males attending a Ghulja mosque told Colin Mackerras in the first half of the nineties, the birth control policy had from the start been regarded as 'a plot to reduce the Islamic population' (Mackerras 1995: 120). In 1997, the birth rate for minority nationalities in Xinjiang remained high at 2.31 per cent, compared with just 1.4 per cent among the Han (Ma 2003: 108). Meanwhile, random interviews conducted in Khotän in 1999 revealed families of eight or nine children, suggesting that family planning rules continued to be flouted in that oasis too (Mackerras 2001: 294).

When I conducted my first field trip in 1995-6, educated Uyghurs in northern Xinjiang (especially urban Ürümchi) had already begun to embrace the concept of smaller families (cf. Clark 2001: 232-235). This may have partially resulted from internalised state discourses representing family planning as a 'modern' activity; for example, the authorities were keen to point out to Muslim citizens that modernising Islamic countries (such as Iran) practised birth control. For some, it may also have constituted a

#### making culture matter: symbolic, spatial & social boundaries 143

desire to emulate mass-mediated television images of Western career women and neat nuclear families. It was clear that the expense of raising a child and the adverse effects of multiple child-bearing on the female body were two important factors. Many of the urban women I knew chose to waive their right to bear two children. Räwiä, a language specialist in Ürümchi, said in 1995 that she felt 'one was enough', in an echo of policy slogans; like many others, she worried about the material resources needed to bring up a child in Ürümchi, where many parents hoped to send their children to costly Chinese-medium schools. Other urban women were concerned that pregnancy would ruin their figures. There was some evidence that a small minority of southern women shared these attitudes. One young woman dressed in smart modern clothes, who worked as a chambermaid in the New Town of Aqsu in 1996, said she had opted to have only one child because 'one was enough to handle.' I even encountered some rural women during that year who voiced fears that childbirth would 'make them old before their time'.

On the whole, however, southern and rural Uyghurs remained strongly opposed to the birth control policy in 1995-6. Southern families were far larger than northern families, and my Uyghur respondents in their twenties from Bay, Aqsu, and Qäshqär all hailed from families of between eight and twelve children. Southerners were extremely proud of the size of their families, and conversation often centred on how many sons and daughters someone had. To choose not to bear many children meant to become the subject of local gossip, as in the case of one respondent's mother who bravely chose to bear only four children in the Old Town of Kucha. Declaring children to be 'a blessing from Allah', southern respondents often enquired whether there was mandatory birth control in the UK, and wanted to know why Westerners usually had small families, that is, *whether they were constrained to by law*. Tursun, a peasant in his thirties from a small village in Aqsu, lamented in 1996 that Uyghurs could do nothing about birth control policy in Xinjiang 'because our king (*padishah*) is the Han'. Another recently married young Aqsuliq quipped during that same year that he would have twenty or thirty children if birth control was not enforced by the state. This figure was evidently exaggerated for comic value; yet his use of the fatalistic phrase 'It can't be helped' suggested that he considered the miracle of birth to be a question for Allah rather than the state.

Discontent revolved also around the perceived ill effects of using birth control methods recommended by the local authorities. Many women in the Xinjiang countryside had never seen a condom in 1995-6, and relied on

Figure 7: Rural community around a 500 year-old walnut tree, Khotän (1996)

the IUD coil and birth control pills, the latter of strengths rarely seen outside of the PRC. When I visited a family planning clinic in the New Town of Aqsu in 1996, local authorities were heavily promoting a new pill, which apparently acted as a 'morning-after pill' or in varying doses as a means to induce miscarriage. Glossy adverts on the clinic wall proclaimed it safe, painless and 100 per cent effective. None of the three women working there—one Hui and two Uyghurs—seemed uncomfortable about promoting this pill; they simply replied that the authorities 'keep a very tight grip on birth control in this area', and said that there was no way around it. They confirmed that local women took the pill on a frequent basis: 'Some take it once every two months. They take it one time... then, a few weeks later, the condom splits again and back they come.' One respondent from Khotän named Arzigül narrated in 1996 how she had been told to have a coil fitted following the birth of her third child. When this led to headaches, incessant bleeding and physical weakness, she had returned to the doctor, who removed the coil immediately and observed that she might have died had it remained in place. Indeed, many local women were reported to have died from using this type of coil, which seems to have been routinely fitted. Subsequently, Arzigül had switched to a birth control pill prescribed by her local hospital, but the headaches had continued.

#### making culture matter: symbolic, spatial & social boundaries 145

Those families who exceeded the legal quota of children in the midnineties faced fines of between 8,000 and 10, 000 yuan (compared with a rural household's average annual income of 5,000 yuan). As a result, many were forced to resort to unusual strategies to avoid punishment. Some families sent the fourth or fifth child to live with relatives elsewhere in the region; alternatively, a mother gave her child over to a childless woman to bring up. This strategy was recorded also by Bellér-Hann and Hann, who note that those peasant families which produced lots of sons in the midnineties often found their land allocation insufficient and coped with this by transferring their children elsewhere and having them adopted (1999: 6). My respondents in 1996 also confirmed that pregnant women in the south were periodically rounded up, loaded onto trucks, and taken away for forced abortions and sterilisations. Hard evidence of such practices (which are illegal even according to Chinese law) surfaces only rarely, but Amnesty International have documented two cases among Han citizens in China proper. The first involved city authorities in Linyi, Shandong province, who according to local residents began requiring parents with two children to be sterilised and forcing women pregnant with a third child to have abortions in March 2005. Officials are said to have detained family members of those who went on the run, beating them and holding them hostage until the exiles returned and submitted to the operation (Amnesty International 2005).8 The second case involved local officials in Puning, Guangdong province, who in 2010 launched a campaign to sterilize 9559 people in a drive to meet birth control quotas. According to Chinese media reports, 50 per cent of this target had already been achieved in the five days between 7th and 11th April. One local doctor described performing surgeries from 8am until 4am the next morning, while local reports suggested that at least some of those affected had not consented to the operation. Like the case reported in 2005, the Puning city authorities are said to have detained 1377 family members as a mechanism to pressure individuals targeted for sterilisation into undergoing the operation. Many of these relatives were elderly and held under house arrest in cramped conditions (Amnesty International 2010). While the national birth control policy is applied to all ethnic groups in China, including the majority Han, the above-mentioned enforcement measures tend to be carried out more often in rural and minority nationality-dominated areas, where resistance

<sup>8</sup> This case was championed by China's celebrated and controversial blind lawyer, Chen Guangcheng, now in exile in the United States.

tends to be higher for cultural and/or religious reasons. The policy is therefore experienced as 'an ethnic issue affecting ethnic relations' in the autonomous regions (Yee 2003: 451) and as the articulation of Chinese hegemony over Uyghur land (Bellér-Hann 2002: 59).

The final symbolic boundary upheld in 1995-6—and the one that has come to dominate at the present time (see Chapter 6)—was Uyghur insistence on *endogamy*. This was a qualified, politicised endogamy, in that respondents expressed a hierarchy of intermarriage preferences within which only the Han Chinese were excluded. Although Uyghur and Han people had clearly intermarried in the past, by the mid-nineties this idea was unacceptable and the source of deep social stigma. Invariably, the justification proffered for avoidance of marriage to a Han was differences in religio-cultural practices. As one young Ürümchilik woman explained in 1996:

We believe in different religions. Before, there was intermarriage, but nowadays there is none. Once a couple gets older, they start to realise their customs and practices are different. And their religions are different. One person says one thing and the other says something else. They can't agree. It almost always ends in divorce.

For Uyghurs and other Central Asian Muslims, the inseparability of national customs from Islamic values is considered an axiom (Atkin 1991). Many older respondents noted that past experimentation with Uyghur-Han partnerships had proved for many that marriage to a non-Muslim can be fraught with difficulties. One *minkaohan* respondent in her thirties concluded in 1996: 'Two separate races of people still have areas that are very different at the end of the day. These differences cannot be resolved with love.' She was typical in that, despite the fact that *minkaohan* are generally considered linguistically and culturally closer to Han Chinese, most nevertheless reject the option of intermarriage with a Han person. Liu Lan, a 20 year-old woman of mixed Uyghur-Han parentage from Khotän, explained in 1996 that 'national sentiment had proved too strong' on the side of her Uyghur mother, leading to her parents' divorce twelve years earlier.

According to personal observation, young Uyghurs rarely had romantic relationships with Han people in the mid-nineties, and those who did came under immediate attack from Uyghur elders and peers. In 1995, I interviewed one nineteen-year-old girl from Ürümchi who was then dating a Han boy. She told me:

#### making culture matter: symbolic, spatial & social boundaries 147

It's really hard for us to even go out anywhere. If other Uyghurs see us together in public, they give us trouble. If Uyghur men see us together on a public bus, they swear at us and hit us. Uyghur women aren't so bad, but they still make comments.

She confided that her older sister, who worked as a model in Beijing, also planned to marry a Han person. Fearful of the weight of in-group disapproval, their mother had refused to allow the younger daughter's marriage to take place in Xinjiang, despite the fact that her Han partner had given up pork and begun to learn the Uyghur language (thus removing what barriers he could that might obstruct the match). She agreed, however, to the elder daughter's marriage, since the wedding would take place far away from the socially and politically strained environment of Xinjiang, and her married life would be spent in Beijing. This mother's strategy at least suggested that it was not necessarily religio-cultural differences *per se* that ruled out intermarriage with a Han at that time, but rather the threat of censure and stigmatisation from within the Uyghur community.

# 3.3. *Spatial Boundaries*

On his visit to China in 1342, the Berber Moroccan explorer, Ibn Battuta, observed a 'dissonance' between the Hui Muslim and Chinese sections of the cities that was so great that the markets within the Muslim sections resembled those in Muslim lands (Israeli 1984: 282). According to Blum, at the turn of the millennium, there remained whole Hui towns in southeast Yunnan where no non-Hui willingly went as a result of incidents of provocation and violence against outsiders (Blum 2001: 134). In Xinjiang, a number of spatial boundaries were carefully maintained during the mid-nineties, the most obvious being ethnically based patterns of *settlement and residence*. Early Han settlers arriving in the region in the 1800s followed a deliberate 'policy of isolation' from the local population: 'Chinese traders, artisans, soldiers and officials did not like to have direct intercourse with the local people (Warikoo 1985: 108). Similarly, although Han migrants were initially sent to live in minority communities during the early PRC years, the policy was abandoned in the 'consolidation' phase starting in 1957. Subsequent migrants were settled in 'dispersed form' either in agricultural cooperatives or within entirely new villages built exclusively for them. This abrupt policy change was attributed to the problem of 'age-old discrimination' (Schwarz 1963: 70) (see also Chapter 1). A similar

strategy had earlier been pursued in Russian Turkestan, where 'there were a few officials and military garrisons, creating "white" colonies in quarters specially built for them […] Russians and Muslims formed distinct communities, coexisting but not mixing with each other' (Bennigsen and Bryan 1989: 250). It has been suggested that local Uyghurs may have perceived this pattern of Han settlement as 'encirclement', leading to a new focus on Uyghur-Han rivalry (Rudelson 1997: 38). Probably, the state hoped to manage religio-cultural differences (in particular, the raising of pigs by Han as against Uyghur avoidance of pork) and in this way avoid inter-ethnic conflicts. The policy will also have made the prospect of in-migration more appealing to Han settlers, many of whom were reluctant to move to Xinjiang, which they perceived as a distant and hostile territory in physical and cultural terms.

During all three of my field trips to Xinjiang conducted in 1995-6, 2002 and 2004, the pattern of Uyghur-Han spatial segregation persisted. Each oasis had an Old Town (Uy: *kona shähär*) and a New Town (Uy: *yengi shähär*). While the populations of the Old Towns were composed almost entirely of Uyghurs, those of the New Towns were composed mainly of Han migrants. The exceptions to this general rule were the regional capital and the northern oasis town of Ghulja. In the former Chinese military outpost of Ürümchi, there were originally very few Uyghur residents.9 This historical fact continues to reflect in the population and residence patterns of the city, which is a New Town in which Han migrants occupy all districts except one: the Uyghur-dominated south-eastern district of Döngköwrük. In Döngköwrük, there were very few Han residents in the mid-nineties, while comparatively small numbers of Uyghur residents were scattered throughout the Han districts. In Ghulja, it is difficult to discern a separate Old Town; rather, Han migrants have superimposed their New Town on the pre-existing Old Town, gradually infiltrating and diversifying its traditional mosaic of Uyghur back alleys and courtyards. As a Han taxi driver explained: 'There *is* only the Old Town here.' In 2002, one district in Ghulja was predominantly inhabited by Uyghurs: the area around Shenglilu (the aptly named Victory Road). Revolving around a traditional Uyghur bazaar

<sup>9</sup> Historically, southern Xinjiang was the centre of economic activity, accounting for 70.3 per cent of the regional population before 1949 compared with just 25.5 per cent living in the north. Although the population of Ürümchi increased from 132,000 in 1949 to 571,000 in 1961, the increase in urban and non-agricultural populations outstripped the increase in rural and agricultural populations. By 1982, Han migrants were dominant within a population of slightly less than one million (Li 1990: 68-70).

# making culture matter: symbolic, spatial & social boundaries 149

Figure 8: Wardrobes for sale, Kucha Old Town (1996)

Figure 9: Buskers at Victory Road bazaar, Ghulja (2002)

#### 150 chapter three

where popular cultural items were on sale, it was visibly poorer than the Han-dominated part of town. It was also much livelier, periodically hosting travelling Uyghur musicians who were very popular with passers-by. During my field trip, I was spectator to three elderly bearded gentlemen from Qäshqär, one playing the *rawap*, the second playing the *dap*, while the third sang and danced in the street to great applause.

A similar pattern of ethnic segregation has been observed in south Xinjiang by the Beijing University sociologist, Ma Rong. In a survey conducted at the turn of the century, he found that ethnically segregated population distributions restricted contact in the form of language and cultural exchanges between Uyghur and Han residents, and that this exerted a negative impact on Uyghur-Han relations (2003: 122). According to this important study, ethnic groups live in separate communities, in terms of both residence and social life, at county, township and village levels in the southern oases of Qäshqär and Khotän. To underline the extent of segregation, in the environs of Qäshqär, Han migrants are concentrated in just three out of eleven counties. In rural areas of Qäshqär prefecture, different ethnic groups also form separate administrative units in terms of agricultural production, including production brigades (*dadui*) and squads (*xiaodui*) (Ma 2003: 114-5; 118).

It would however be untrue to suggest that spatial boundaries of residence were non-porous. New Towns did include a proportion of Uyghur residents, often those employed by state work units, while impoverished Han migrants and petty entrepreneurs could sometimes be found in the Old Towns (cf. Bellér-Hann 2002: 68). Many Uyghurs were willing to make compromises and move into the New Towns if a good job and a modern home beckoned. *Loufang* homes (Soviet-style six-storey residential blocks popularised in the PRC after 1949) had the benefit of water supplied on tap, removing the tricky problem of the shared water source. In addition, it was clear that some *minkaohan* respondents had during the course of their Chinese-medium education (or from Western films and dramas broadcast on Xinjiang television) internalised supposedly 'modern' notions of hygiene and beauty. As a result, some thought the new *loufang* homes cleaner and more attractive than the dusty, earthen-built one-storey Uyghur homes of the Old Towns. I sometimes overheard the children of *minkaohan*  parents complain in 1996 that the Old Towns were dirty and muddy, and remark upon how 'beautiful' Beijing was, as represented through the mass media. In some cases, their parents had been to China proper to study or visit, bringing home reports of glittering mirrored skyscrapers.

#### making culture matter: symbolic, spatial & social boundaries 151

Despite this concession, Uyghurs in the mid-nineties sought nonetheless to maintain spatial segregation between the two groups as far as possible, and they did so by placing a deliberate emphasis on *Islamic dietary prescriptions*. The boundary was in the first instance symbolic, as illustrated by the *zhuancha* (brick tea) scandal (cf. Cesàro 2000: 231). My respondents in Ürümchi narrated the story as follows: some years earlier, a Uyghur reporter had allegedly circulated photographs, some of which showed Han workers slaughtering pigs within a brick tea factory complex in China proper, others showing Han workers trampling the tealeaves while sweating profusely. Although there was a lack of clarity surrounding the source of this evidence, the ensuing scandal united Uyghurs in a boycott of the brick tea product for five months, inspired both by the apparent presence of pigs in the factory grounds but also by the image of Han sweat soaking into the tealeaves. The Chinese government had subsequently been forced to launch a campaign to persuade locals to resume drinking the tea (its sales account for a large proportion of trade profits from the three Chinese provinces of Hubei, Hunan and Sichuan). At the time of interview (1996), many respondents were not sure what to believe, and some brick tea lovers had begun to drink it once more. In another example, Uyghurs allegedly refused to drink water from the Tianshan mountains after the ashes of the notorious Chinese political figure, Wang Zhen, were scattered there following his death in 1993 (Wang Lixiong, cited in Veg 2009a).10 In a third example, symbolic avoidance was taken to comic extremes when a Uyghur computer programmer on a study trip to Beijing in 1995 was given a plate decorated with Chinese politician Li Peng's signature. He refused it, saying: 'No, thank you. It's not *qingzhen*!' In this way, Uyghur actors employed religious differences in order to assume the moral high ground in relation to Han people and Han products. It has been suggested that the earlier adoption of the term *qingzhen* by the Hui (Chinese Muslims) equated to a 'turning of the tables' on Confucian society by implying that Islam (rather than Confucianism) was the true faith, and deploying this as a means to maintain the 'purity' of one's relations (Ben-Adam 1999: 194-5).

In interviews, Uyghur respondents consistently associated Han Chinese with forbidden (*haram*) foods, and presented this as the main barrier between the two, in cultural, spatial and social terms. As a middle-aged Uyghur doctor in Khotän observed in 1996:

<sup>10</sup> Wang Zhen, one of the Eight Elders of the Chinese Communist Party, led one of the seven columns of the Northwest Field Army under the command of General Peng Dehuai between 1945 and 1949, and enacted the 'peaceful liberation' of Xinjiang in October 1949.

# 152 chapter three

There are still huge differences in culture between the two nationalities, for instance, food. A Uyghur will not eat pork, although the Han do. If a Uyghur ate pork, he would no longer be a Uyghur. It is as simple as that.

Han migrants too were keenly aware of the complications of Islamic dietary prescriptions. One young Han woman living in Ürümchi in 1996 explained with evident distaste that Uyghur rules governing diet were so strict that in the (rare) case of a Han person marrying a Uyghur, the Han was obliged to drink a bitter concoction to 'sterilise his/her intestines' before converting to the Muslim diet. In Ürümchi and other urban centres, most restaurant signs stated whether the food served inside was *hancan* ('Han cuisine', including pork and therefore 'forbidden') or *qingzhen* ('pure and true', specialising in *halal* dishes of mutton and beef). Within this basic distinction, there were three types of restaurant:


Uyghur respondents avoided the *hancan* restaurant without exception, as a taken for granted social fact. However, this was not simply about the practical need to avoid forbidden foods. The refusal to dine in Han establishments or with Han peers represented also a political statement and a counter-assertion of Uyghur cultural and moral superiority. This is underlined by the fact that respondents in 1995-6 also attempted to dissuade me from entering *hancan* restaurants, entering into arguments with Han proprietors who invited me to eat in their establishments: 'She doesn't eat Han food! She eats Muslim food, don't you?' In this way, they tried to ensure that I ate only *qingzhen* foods while in Xinjiang, thereby aligning myself culturally, politically and symbolically with the Uyghur in-group. As Blum reminds us:

Commensality (eating together) in China is the single most important act of social cohesion. Banqueting is the one form in which people forge obligations of reciprocity with one another […] food is a point of intersection among many aspects of Chinese culture, including […] a sense of harmonizing of diverse elements […] (2001: 137).

In rejecting the possibility of commensality, Uyghurs not only challenged the state discourse of ethnic harmony (*minzu hexie*) but also sidestepped

the expectations of Uyghur-Han reciprocity that might have been generated in that context. This same boundary marker has, in the absence of other linguistic and cultural distinctions, been used by some urban-dwelling Hui (Chinese Muslims) to distinguish themselves from Han Chinese. This is because, otherwise: 'Urban homes of the Hui have little to distinguish them from Han homes except perhaps for an Islamic calendar' (Ben-Adam 1999: 197). On the other hand, some ambitious Hui cadres have chosen to renounce tradition and enter the food-sharing world of the Han so as to avoid exclusion from the important social and political interactions it entails (Blum 2001: 137).

Although I never observed—or heard of—a Uyghur individual who compromised and consumed *Hancan* foods, nor were the rules surrounding avoidance of pork completely hard-and-fast in Xinjiang. Uyghurs constantly negotiated this boundary, which was permeated with contradictions and concessions according to context and circumstance. For instance, their definitions of forbidden (*haram*) and permitted (*halal*) foods did not always or necessarily accord to those items explicitly forbidden in the Qur'an; rather, Uyghurs defined categories of *haram* and *halal* in accordance with the people or ethnic group the foods were associated with. In this way, articulation of the 'food boundary' became a form of Uyghur resistance against the Han people (cf. Cesàro 2000: 230, 234). If the food boundary were truly solid, then one might have expected Uyghur respondents to avoid *any* establishment where food came into contact with Han people. Rather, I found that respondents ate in Han-managed *qingzhen* restaurants under certain circumstances. Uyghurs employed by the Xinjiang Social Sciences Academy (to which I was affiliated) normally ate lunch in nearby Han-managed *qingzhen* restaurants in 1995-6. It made no difference whether employees were *minkaohan* or *minkaomin*; they did so in order to make a concession to Han colleagues, thus easing interactions in the workplace. Han colleagues made a similar concession in patronising restaurants that did not serve pork, as a mark of respect for Uyghur colleagues' cultural sensibilities. Groups of young Uyghurs, who normally ate in Uyghur *qingzhen* restaurants, sometimes held birthday parties or other special events in Han-managed *qingzhen* establishments. They did so because they appreciated the opportunity to combine local Muslim dishes with cold vegetable dishes (*liangcai*) popular among Han, and thus enrich the spread of food on the table. Yet it was noteworthy that they were rarely if ever accompanied by Han peers.

#### 154 chapter three

The willingness to patronise restaurants whose Han managers did not conform to the Muslim diet and chose to serve all-*halal* dishes simply to increase profits (i.e. to attract both Han and Uyghur customers) inevitably raises the question of ritual purity. I asked one of my closest respondents in 1996 why these Uyghur customers consented to eat in a restaurant whose staff doubtless came into regular contact with pork at home and probably also during their lunch break. Shökhrät, a young male intellectual from Ürümchi, protested:

But have you noticed that it is a certain type of Uyghur who goes there? Those who went to Han schools (*minkaohan*). They're not so fussy about things like that; also, those who didn't receive a higher education. They haven't thought of all the implications, you see.

In saying this, Shökhrät unwittingly suggested that the only type of Uyghur to refuse to eat in Han-managed *qingzhen* restaurants was the educated *minkaomin* like himself. For one thing, there were relatively few *minkaomin*  who had graduated from university and remained in Ürümchi at that time, suggesting a small demographic. For another, one wonders how he could tell *minkaohan* and *minkaomin* apart just by looking at them. Thirdly, I myself knew young *minkaomin* who accompanied *minkaohan* peers to these establishments, when it was a case of celebrating a birthday. To give another example, although respondents attempted to prevent me from entering *hancan* restaurants, it seemed to make no difference to them that I might eat pork when in my home country, the UK. Indeed, Uyghur respondents apparently chose to believe that I did not; they certainly never questioned me on this point. In this way, the food boundary was selectively deployed as a means to reject Han people, while the same rules were not applied to Anglo- and mainland Europeans (cf. Cesàro 2000: 232).

Given the choice, Uyghur respondents normally elected to dine in Uyghur rather than Hui *qingzhen* establishments, with Uyghur restaurants easily distinguished by the sheep carcasses hung outside. Some complained that Hui people did not know how to cook *läghmän* (thick round noodles in spicy tomato, red pepper and mutton sauce) properly, and put in too many chilli peppers. Yet Uyghur employees of the Xinjiang Social Sciences Academy did sometimes eat in Hui restaurants, because there were no Uyghur restaurants in the immediate vicinity of the work unit (located in north-western Ürümchi). Cesàro has argued that Uyghur reluctance to eat in Hui restaurants can be attributed to a general lack of trust between the two groups, since Uyghurs inevitably associate Hui with the Han Chinese

(2000: 230). This corresponds with my own findings on Uyghur attitudes towards the Hui, for example, their disapprobation of the Hui preference for 'Beijing time' over 'local time'.

On university campuses, kitchens and dining halls were segregated. This system had been reintroduced after the excesses of China's Great Leap Forward, during which 'peoples […] with widely differing diets were initially lumped together on communes and served one menu in communal mess halls' (Dreyer 1968: 99). It had subsequently been reintroduced for a second time following the conclusion of the extreme phase of the Cultural Revolution. When the Central Nationalities Institute re-opened in Beijing in 1972, it included a special kitchen for Muslim students (Pye 1975: 504). Lunchtime visits in 1995-6 to Ürümchi's Universities of Medicine and Education revealed separate queues before the *hancan* and *qingzhen* serving hatches, the former composed entirely of Han students, the latter of mainly Uyghur students. In a separate policy to manage dietary differences, universities usually provided separate dormitories for Uyghur and Han students. Similar arrangements of spatial segregation were observed by Yee, who reported that student hostels and staff quarters on university campuses, as well as living quarters in cities and townships, were divided along ethnic lines, ostensibly in order to 'preserve and respect minority cultures' (2003: 450). However, like Ma Rong above, Yee concluded that rather than preserving ethnic peace, the strategy had simply created 'ethnocentricity'. Interestingly, these criticisms echo those of Western social scientists who have challenged multi-culturalist policies on the grounds that they are likely to create perverse effects such as heightening ethnic prejudice or reproducing ethnic segregation (Kymlicka 2011: 7).

Like settlement patterns above, on-campus boundaries in Xinjiang were not completely hard-and-fast. Some Uyghur students explained in 1996 that they sometimes chose voluntarily to live in mixed dormitories together with Han students, so that they might practise their Chinese. In this way, they were prepared to make cultural concessions where they stood to gain in socio-economic terms: they knew that Chinese language ability was crucial for success both at university and in the job market. Such students emphasised however that Han students took care not to eat 'forbidden' foods in the dormitories. Evidently, I was not able to participate in dormitory life for long enough to observe whether this was really the case.

Perhaps most crucially, Uyghur respondents in 1995-6 cited the problem of 'unclean' utensils, crockery, and cutlery as a key factor preventing them from visiting Han homes, thereby ensuring spatial segregation in the

#### 156 chapter three

domestic sphere. Whenever I was invited to dine at a Han home, my closest respondents would do their utmost to dissuade me from accepting their invitation, protesting: 'You don't want to go there! The Han eat pork, and the food will be cooked and eaten out of the same pans!' Yet here too there were many inconsistencies in practice. As described above, Uyghur proprietors served food to Han customers in their restaurants every day, meaning that Uyghur customers used the same plates, bowls, chopsticks, and spoons used by customers who ate 'forbidden' foods; yet Uyghur customers nonetheless continued to patronise these establishments. When I raised this issue with my respondents, most acknowledged the discrepancy, but pointed out that there was no easy solution to the problem since they needed Han custom to keep their businesses ticking over. Shökhrät, a male intellectual in his thirties originally from the south, differentiated between this public, commercial environment and the private, domestic environment, explaining: 'On the rare occasions when a Han colleague comes to a Uyghur home, the bowl, dish, and chopsticks he uses are thrown away afterwards'. I later asked another respondent in his thirties in rural Aqsu whether he followed the same guidelines. He shrugged, said that for his part he would simply wash the crockery in boiling water, and joked that such an extreme attitude was to be expected from a Qäshqärliq.11

Finally, Uyghur respondents in 1995-6 frequently expressed disgust at the sight of live pigs or at the smell of pork cooking. Aware of this religiocultural specificity and in the context of smashing the 'Four Olds', Red Guards during the Cultural Revolution had forced Muslims to raise pigs in many parts of China (Heberer 1989: 109). In Xinjiang, mosques had been torn down and Uyghurs and Kazakhs forced to eat pork and cremate their dead (Bush 1970: 295); Muslim clergy had even been forced to walk through the streets with pigs' heads tied around their necks (Grunfeld 1985: 64). Conversely, in the mid-nineties, Han migrants living in rural areas were obliged to raise their pigs on all-Han settlements well removed from Uyghur communities owing to the problems presented by the shared water supply. When I asked a Uyghur peasant living in Aqsu prefecture in 1996 whether Han people who lived nearby raised pigs, he replied decisively: 'No, they can't. They don't dare. We all use water from the same rivers and streams. If they raised pigs, the pigs would drink from the streams, the meat would

<sup>11</sup> The inference here is that the Islamic faith is particularly strong in that southern oasis, which is 'widely regarded as both the place of origin of all Uyghur and the stronghold of Uyghur-Islamic civilisation' (Bellér-Hann 2001: 62).

#### making culture matter: symbolic, spatial & social boundaries 157

be washed in the streams, and then the water would flow downstream to us! That would cause big problems.'12 Conducting her fieldwork at the same time in another rural southern oasis, Bellér-Hann reported a similar state of affairs, noting that the pork taboo was so much internalised that only open breach led to its verbalisation: 'Certain prescriptions […] were evidently so consistently observed and taken for granted by actors and observers alike that they were not considered worth mentioning' (2008: 311). Perhaps the best example of Uyghur avoidance of *haram* animals concerns a direct observation I made when travelling between oases in 1996. At one stage, our passage was delayed by a flood across the road. After a time, a truck loaded with live pigs made it through the water from the other side, and drove past. A Han passenger, glancing at the driver of the truck, remarked with evident surprise to his neighbour (also a Han) that the driver had been a Uyghur. His companion's response was scornful: 'Impossible! When did you ever see a *Uyghur* transport pigs?'

In the southern oasis towns, where the Han population remained limited in 1996, there were still relatively few *hancan* restaurants. In Khotän, I found just a handful of *hancan* breakfast stalls and street restaurants opened in the New Town by newly arrived Han settlers. According to a Han bartender at the Qinibagh hotel in Qäshqär, there were still only two main markets for 'products for Han Chinese' (her euphemism for pork), and pork was not sold outside of these locations. In Aqsu, a Uyghur male chef insisted (with some bravado) that Han entrepreneurs in the New Town had not been able to open many *hancan* restaurants, since 'pork stinks when it's cooking'. In Ürümchi, on the other hand, the high percentage of Han residents meant that *hancan* establishments had quickly spread across the whole city with the exception of the Uyghur district, Döngköwrük. As a result, Uyghur respondents living in the regional capital in 1995-6 complained of changed policy regarding the sale of pork. Shökhrät, a male intellectual in his thirties, maintained that in the past there had been just two designated locations in Ürümchi where pork could be sold, in recognition of the local culture. However, following the influx of Han migrants, pork was no longer outlawed, and could be easily found in markets all across the city. My personal observations supported these claims. In the winter of 1995, I was amazed to find a whole frozen pig's head placed like

<sup>12</sup> Ben-Adam reports similar findings with regard to the Hui, observing (a little indelicately) that 'in rural areas Huis have created for themselves isolated ethnic enclaves making what could be called "qingzhen islands in a pork-infested sea"' (1999: 198).

a trophy at the centre of a Han meat stall at the local outdoor market in Ürümchi. Although the Han stallholders set their stalls up at one end of the market while Uyghur and Hui stallholders sold *halal* meats at the other, this act seemed unnecessarily provocative, given local religious and cultural sensibilities. Similarly, in a predominantly Han homestead just outside Aqsu New Town, a pigpen had been placed in front of the public toilets in 1996. Any Uyghur inhabitant wishing to use the facility would therefore have been faced with a pig's snout (as I was), appearing at the air-vent in the toilet wall and accompanied by a cacophony of snorting and snuffling.

Interviews conducted with Han migrants in Ürümchi in 1995-6 confirmed that attitudes among the incomers were changing fast, as when a Han restaurant manager argued: 'We [Han] eat pork, and they [Uyghurs] don't. But that doesn't mean that we don't respect their culture. What about respecting each other's culture?' This finding corresponds with Dawut's study of the commercialisation of the Qirmish Ata Mazar (shrine) in Aqsu prefecture, in which she described Han tourists' transgressions of cultural and religious boundaries in 1998, such as the setting up of food and drink courts directly on the premises, and extensive littering including liquor bottles (2007: 159-160)

While there was no question that Uyghur respondents disliked the sight of pigs and the smell of pork, I frequently had the impression in 1995-6 that urban Uyghurs deliberately exaggerated their disgust as a symbolic means of articulating dislike of the Han people. For instance, a young female named Mobaräk studying in Ürümchi in 1996 assured me that she and her friends always took a long detour at the local market in order to circumvent and avoid a Han butcher's stall. Yet I never observed her do so when she and I strolled around the market, suggesting that community pressure from the in-group was at least partially responsible for this practice. Similarly, Uyghur petty entrepreneurs at the local market were perfectly willing in 1995-6 to set up their shish kebab stalls directly in front of *hancan* establishments, knowing that this could bring mutual economic benefits. From the kebab seller's point of view, Han diners liked to eat lamb kebabs, and could afford to buy them in substantial quantities; meanwhile, Han restaurant proprietors knew that Han customers were more likely to patronise a *hancan* restaurant if they could enjoy a combination of traditional Han dishes and local specialities. Hence the common occurrence of a regular partnership between a particular Han restaurateur and a particular Uyghur kebab seller, and the frequent sight of the latter running in and out of the former's *hancan* restaurant with a steaming plate of shish kebabs.

### 3.4. *Social Boundaries*

With the exception of pragmatic relationships formed within the workplace, Uyghur respondents in 1995-6 did not willingly socialise with Han people; rather, social segregation was maintained between the two groups in the home, on the street, and in all social situations, with few exceptions. If I asked respondents whether they spent time socially with Han people, the enquiry invariably brought a negative click of the tongue or decisive shake of the head. Again, the principal justification given for segregation was contrastive dietary habits. Writing on Uyghur-Han interactions in the eastern oasis of Turpan at the end of the eighties, Rudelson identified pork as the main factor influencing interaction between Uyghur and Han people in Turpan: 'These social borders may appear invisible… but they become salient in structuring interethnic social, religious and commercial interactions' (1997: 63). On a purely practical level, the inclusion of pork in the Han diet made interaction in many social situations difficult, since food is invariably a central part of this type of interaction. Take, for example, the following conversation held with a Uyghur woman in her forties in Ürümchi in 1996:


Not only did Uyghur respondents avoid visiting Han homes, creating a 'distinct asymmetry' in inter-ethnic hospitality (Bellér-Hann 2001a: 58), but they also stated that they felt extremely uncomfortable when Han Chinese came into their homes (usually this would be a senior colleague from the work unit). A Uyghur peasant in his forties in Aqsu told me in 1996 that on the rare occasions when Han colleagues visited his home or the homes of neighbours, they always asked the Han guest: 'Would you like something to eat?' It is well known in Uyghur culture that asking a guest *whether* they would like to eat indicates the reverse inclination, since, conventionally, a Uyghur host will simply produce tea and refreshments and enjoin the guest to tuck in: '*Yäng! Iching*!' (Eat! Drink!) Ethnic segregation was plainly visible on the streets of Ürümchi and other New Towns in 1995-6, where Han

#### 160 chapter three

friends walked hand in hand, and Uyghurs arm in arm, but mixed groups were never seen. Han people only rarely attended Uyghur social gatherings, and then it was normally in an official capacity. For example, a Uyghur employee might feel obliged to invite his Han boss from the work unit to his wedding party; or else a Han superior might drop in on a Uyghur wedding party to pay his respects. Alternatively, a Uyghur businessman might consider it advantageous to take a Han businessman to a high-class Uyghur dance restaurant in order to clinch a deal.

There were additionally a great many objective reasons why Uyghur and Han people found it difficult to socialise together. First and foremost, the vast majority of Han people cannot speak any Uyghur, and would have felt lost in an environment where Uyghur is spoken almost exclusively. Second, Islamic dietary requirements aside, Uyghurs serve and consume food in a different way to the Han. At house parties, they sit cross-legged on the floor and feast from a traditional cloth used for that purpose (*dastikhan*), whereas Han Chinese sit at table. Third, Han and Uyghur people have very different traditions of music and dance (see Chapter 2). Music and dance are the key foci of Uyghur social gatherings, which serve as an arena in which to make music, sing, dance, pursue courtships, and generally take centre stage.13 For instance, at a birthday party in an urban Uyghur restaurant, the owner is expected to produce Uyghur music cassettes once the guests have finished eating. The male guests then approach the all-female table to invite the women to dance in the traditional form, or to waltz or fourstep, depending on the rhythm and tempo of the song. Where a party is held at home, space is more limited and thus furniture will be pulled back to clear an area for dancing. Guests take turns performing Uyghur songs to the accompaniment of *dutar* or guitar, and everyone joins in. Special guests are honoured by being invited to dance or sing for the others. At the highclass Uyghur dance restaurants in Ürümchi, the evening begins with music played on traditional stringed instruments such as the *dutar*, *tämbur*, and

<sup>13</sup> Traditional Uyghur social events include Islamic ritual ceremonies (weddings, funerals, circumcisions), house parties (*olturush* or 'sittings'), orchard gatherings, and the *mäshräp:* an out-door mass gathering held during the summer months, at which men and women feast, play *dutar*, sing, dance, tell stories and jokes, and play games. A popular spot for the *mäshräp* in 1996 was Khotän's biggest water-control point, on whose banks young men and women converged at weekends. For a description of the historical origins and various forms of the *mäshräp*, see Zheng and Luo (1989: 134–7). Contemporary social events include birthday parties held in urban street restaurants, trips to high-class urban dance restaurants, all-Uyghur dances held in universities, and even visits to the work unit *wuting* (dance hall).

#### making culture matter: symbolic, spatial & social boundaries 161

*rawap*, as guests dine on Uyghur dishes. Later, well-known artists appear to sing modern Uyghur and Central Asian pop songs, accompanied on electric guitar and keyboards. At this point, one Uyghur guest will move onto the dance floor and start off the dancing. The following extract from my 1996 diaries describes the scene:

Now one woman got up and began to dance in the traditional form... confident, smiling. Like all Uyghurs, the perfect performer. Her eyes flashed as she looked proudly upon her table of friends. She was dancing for them. Whoops and cheers came from many areas of the room, and she was joined on the dance floor by pairs of men and women, smiling, winking, clicking their fingers, and swivelling their wrists.

From then on, dancing will alternate between the traditional form and ballroom dance, until the finale: ten minutes of Eurobeat disco. During the course of the evening, there may also be solo performances by professional dancers or fashion shows. A comparable love of flamboyant, public performance does not figure in the social lives of most Han, who do not consider themselves natural show-offs. While many Uyghurs love to be in the limelight, Han are usually loath to be 'the bird that sticks its head out of the nest' (*chutou niao* 出头鸟).

Fourth, Uyghur men and women conventionally conform to traditional patterns of gender segregation at such events, in the form of the 'modesty code' (*namahram*).14 For example, at birthday parties in Ürümchi restaurants Uyghur adults usually sit at sexually segregated tables. During house visits, women often retire to one room while the men sit in another. At wedding parties, men sit together at tables arranged down one side of the hall, while women sit at tables placed along the other. At funerals, also, men and women separately pursue their given roles. It is not only that they are adhering to Islamic social laws but also that they feel more comfortable in the company of their own sex. This type of sexual segregation is culturally alien to Han people, who were taught throughout the Maoist period that women 'hold up half the sky', and were encouraged to play, study and work together with women.

<sup>14</sup> This code governs women's everyday behaviour in relation to men, who are separated into two sub-sets: those covered by the 'incest taboo', who are not available for marriage (for example, fathers, sons, brothers); and those not covered by the taboo, who might constitute potential marriage partners. In the presence of the latter, women must at all times maintain spatial distance and observe the rules of female modesty (Bellér-Hann 1999: 122).

#### 162 chapter three

Fifth, participation in Islamic ritual ceremonies (weddings, funerals, etc.) requires 'inside' cultural knowledge. There is a strong feeling among urban Uyghurs towards *mähällä* (neighbourhood), seen as a community which determines social relations between individuals, and provides unity and solidarity. Within the *mähällä*, obligations and responsibilities are placed upon individuals in return for support and services, so that members celebrate weddings and festivals together, organise funerals and rituals together, and help one another in need. Uyghurs residing in urban *loufang* often complain that *mähällä* life is disappearing in the big cities, particularly Ürümchi, and consider this the result of the large Han presence. Räwiä, a language specialist in Ürümchi, explained in 1996:

When Uyghurs have a death in the family, that's when they most wish they had other Uyghurs as neighbours. Especially during the first week after the death, friends and relatives come to the house with gifts of food, since the family is too beside itself with grief to cook. With so many guests coming every day, the family will really wish it had Uyghur neighbours, who would simply throw open their doors and allow the guests to overflow into their houses.

Similar difficulties have long existed among the Hui Muslim communities to be found in every province of China proper. In his historical study, Israeli observes that Chinese Muslims were simply out of place in the Confucian setting:

They went their own ways, in prayers and ceremonies, in their calendar and festivals, in their weddings and burial of the dead, in their socializing and eating habits, in their travelling and settled living … [And yet daily life] necessitated contacts and their resulting conflicts (1984: 286).

The popular religious activities of the Han (worship of the spirits of the ancestors) were equally kept separate from the rituals of the Muslim community, so that villagers would seek to drive away the spirits of dead Muslims by shouting: 'If despicable Muslims come, give them a segment of the pig's intestine!' (Hsu, cited in Israeli 1984: 286).

Still, it should be clear that these differences alone would not absolutely prohibit Uyghur-Han social interaction; in theory, ways might be found to manage the differences. In fact, Uyghur disinclination to socialise with Han people in the mid-nineties was based upon much more than the sum of the above differences. Some urban Uyghurs liked to frequent the *wuting* (dance hall)—a Han innovation—within their work unit, and many

# making culture matter: symbolic, spatial & social boundaries 163

had mastered the respective arts of waltz and four-step. Surely this venue ought potentially to have formed a space in which Uyghurs and Han could share a love of dance. Yet after a brief greeting, Uyghur and Han individuals from a common work unit chose to sit in separate groups; Uyghurs invited fellow Uyghurs to dance, while Han approached other Han. Occasionally, Uyghur men asked Han women to dance (often their nextdoor neighbours), but all returned to their separate seating arrangements afterwards. Furthermore, my direct observations suggested that Han people were not only unlikely to join Uyghur social gatherings but patently unwelcome to cross that boundary. At a private New Year's party held in a Uyghur friend's street restaurant in Ürümchi in 1996, groups of Han twice stumbled through the door and demanded to be served food, or stood grinning at the scene. The owners said nothing. Instead, they threw them an initial glance, looked quietly angry, and then averted their gaze as they waited for them to leave. When this had happened for a second time, my host rose with a face like thunder and bolted the door after them. Evidently, anyone might be expected to feel annoyed if their private gathering were gate-crashed by strangers; yet I could not help feeling that had the intruders been a group of Uyghurs, the reaction of the proprietors would have been quite different. This observation is supported by Yee's 2000 survey, which found that both Uyghur and Han respondents were inclined to befriend people from the in-group (2003: 437).

# 3.5. *The Consequences of Boundary-crossing: Bus Stories and Street Fights*

Situations where Uyghur and Han people found themselves involuntarily crammed together in a limited physical space, such as a crowded bus or outdoor market, could breed conflict, and was a favourite topic of Uyghur storytelling in the mid-nineties. Fights broke out on buses over incidents as simple as a Han passenger stepping on a Uyghur's foot. I often observed Uyghur and Han individuals go out of their way to avoid sitting next to one another. One respondent narrated in 1995 how two Uyghurs spread themselves across three seats simply to prevent a Han policeman from sitting down beside them. Street fights were also a common occurrence during this period, and guaranteed to take place if a Han male had the audacity to approach a Uyghur female (see Chapter 6 on the female embodiment of national honour). In the south, respondents reported in 1996 that

#### 164 chapter three

conflicts often occurred in public shower-houses, these being one of the few places where Han and Uyghurs were obliged to interact. In one small village in Aqsu, the shower-houses were not only sexually but also ethnically segregated; according to locals, this was 'to prevent fights breaking out.'

#### 3.6. *Managed Interaction*

The sole arena in which Uyghur-Han interaction seemed to be on the whole successfully managed in 1995-6 was the workplace. Mutual pursuit of a good job, a regular salary, and a better livelihood in a competitive urban society meant that working Uyghurs had learned to live with Han peers within strictly defined limits. Observations of Uyghur and Han employees in state work units and Han-managed companies in the regional capital Ürümchi suggested that on the surface at least Uyghur-Han relations were relatively amicable. At the Xinjiang Social Sciences Academy, Uyghur and Han colleagues courteously (if archaically) referred to one another as 'Han comrades' and 'minority comrades'. While colleagues frequently gossiped about one another—a phenomenon to be found in any workforce in the world—this penchant for gossip was an example of Uyghur-Han commonality more than difference. Uyghur and Han colleagues usually asked one another about their lunch plans; and if it was decided to lunch together, Han colleagues compromised by accompanying Uyghur peers to Hanmanaged *qingzhen* restaurants or Hui Muslim restaurants.

Uyghur and Han colleagues also compromised in limited social interactions at official celebratory events organised by the work unit (e.g. the fortieth anniversary of the Xinjiang Uyghur Autonomous Region in 1995). It must be remembered, however, that these events take place in working hours, and thus all employees are obliged to attend. On such occasions, some Uyghur colleagues would join Han peers in drinking large volumes of *baijiu* (a spirit distilled from sorghum or maize) through a series of celebratory toasts. This symbolic 'honouring' of comrades went some way to smooth ethnic interactions, if on a temporary and artificial basis. The only occasions on which Han colleagues paid a visit to the homes of Uyghur peers was during Islamic festivals, e.g. the Qorban or Rozi festivals, as a public gesture of courtesy and cultural respect. These interactions, too, were often managed with the help of a generous amount of alcohol.

A certain sense prevailed in the Xinjiang Social Sciences Academy (a work unit closely connected to the Party organisation) in 1995-6 that

Figure 10: "Han and minorities breathe the same air, share the same destiny, and their hearts are joined as one", Ürümchi (1996)

harmony should be preserved, and appearances kept up, in the workplace. Some Uyghur employees chose to maintain this through ironic humour, as in one conversation I observed between a Uyghur scholar and his Han peer in 1995. When the latter enquired of the Uyghur scholar whether they would go to a certain meeting together or whether he should go alone, the Uyghur colleague replied: 'Together, together! Of course we must go together ... after all, we *are* inextricably linked (Ch. *libukai* 离不开), aren't we?'In this way, he used a Chinese-language term frequently used in CCP state discourses to describe state-enforced ethnic harmony in comic terms.15 Whether he hoped in this way to lighten the atmosphere or was being sarcastic is debatable; on this occasion, the Han colleague was not sure whether to laugh or take offence, and left the room in confusion. Sometimes, relations in the workplace broke down along ethnic lines, particularly where minority nationality employees felt that their native languages and/or cultures were being marginalised or belittled by Han peers. In extreme cases, a Uyghur colleague who privately cherished ethno-nationalist

<sup>15</sup> Uradyn Bulag has explored the power relation inherent in *minzu tuanjie ('nationality unity'),* concluding: 'insofar as it is meant to uphold the welfare of a wider community at the expense of a smaller one, *tuanjie* implies a coercive unity.' At the same time, he notes that it plays to a basic human desire for stability and harmony (2002: 13-14).

#### 166 chapter three

aspirations might try to create a symbolic mutiny in relation to Han colleagues.

Another workplace in which the two groups managed their interactions was the local outdoor market. There, Uyghur and Han market traders had established mutually advantageous relationships and patterns of transaction, in an example of what Barth refers to as 'ethnic symbiosis' (1969: 20). I regularly observed scenes in Ürümchi in 1995-6 where Uyghur and Han entrepreneurs collaborated as a means of stimulating one another's business growth. I have described above how Uyghur migrant workers rented kebab stands in front of *hancan* restaurants and pursued mutually beneficial arrangements alongside Han proprietors. One respondent named Ghäyrät would cat-call to Han passers-by, with a broad grin across his face, and invite them to eat in the 'partner' *hancan* restaurant before his stand. In return, the Han owner would recommend Ghäyrät's mouth-watering kebabs to all the diners sitting in his restaurant. Since Uyghur, Han and Hui market traders were self-employed (thus reliant on their own efforts to subsist), they belonged to the same status group and had much in common to talk about when business was slack. Often in 1995-6, they could be seen sat around chatting about how much money they had made that day, or how much they had paid for a particular article. Where the talk was limited to business or money matters in this way, the atmosphere remained perfectly amicable. However, when culture was touched upon, a sense of ethnic rivalry entered the equation, so that Uyghur and Han traders teased one another about differences. Consider the following altercation, observed in 1996:


Yet in this environment such exchanges rarely caused long-term offence. Everyone was happy to continue warming their hands over the kebabsellers' charcoal embers, or devouring hot chestnuts. Here I found a good working atmosphere despite the religio-cultural differences mentioned in this and the previous chapter. Inter-personal relationships remained basically untouched by perceptions of economic inequalities because these

### making culture matter: symbolic, spatial & social boundaries 167

Uyghurs at least could see that Han market traders were no better off than themselves (cf. Bellér-Hann 2002: 67-68). Still, as with Uyghur-Han relationships in the state work unit, social interactions rarely extended beyond the work environment. Ghäyrät and others often played pool with Han peers at the market pool tables when taking a break from their stalls since, as Ghäyrät remarked, he did not mind whom he played so long as his opponent was a good player. However, the game was usually conducted in nearsilence, and was characterised by a tangible element of ethnic competition. When the game was over, the players thanked one another and went their separate ways.

# *Conclusion: The Roots of Constructed Segregation*

The second symbolic means through which Uyghurs in Xinjiang resisted the Chinese state in the mid-nineties was boundary construction. Building an intricate system of symbolic, spatial and social boundaries, respondents consciously articulated their separate identity in relation to Han settlers in an 'Us and Them' type ethnic dichotomy. These boundaries were not based upon empty rhetoric; in constructing them, respondents drew upon specific aspects of a previously taken for granted, shared history of group values and norms. In deploying boundaries as a means to segregate themselves from Han people in linguistic, religio-cultural and territorial terms, they exercised their social agency and marked themselves as culturally and morally superior to the coloniser.

Different boundaries were emphasised in different parts of the region, owing to contrastive local conditions. For educated, urban Uyghurs in the north, who competed with a majority Han population on a daily basis in the spheres of higher education and employment, language use, time and endogamy were foregrounded as the key means through which to express resistance. For southern and rural Uyghurs, who had not yet seen a comparable influx of Han settlers and thus had not struggled with the socioeconomic competition that results from a changed population composition, language had not yet become an issue. Indeed, those peasants who had had the opportunity to learn a few words of Chinese considered the new skill rather novel. Here, resistance was more commonly expressed in the form of local opposition to the state family planning policy. This boundary clearly marked rural Uyghurs as Muslim subjects and reflected the

#### 168 chapter three

historically stronger Islamic identity in the south (particularly in Qäshqär and Khotän). In treating childbirth as a matter dependent on Allah's will, these respondents rejected the authority of the Chinese state in favour of Islamic fatalism. Oral reports circulating on the Uyghur grapevine of sickness and death caused by inappropriate birth control methods, and of mandatory abortions and sterilisations, not only strengthened southern opposition to family planning but also reflected that opposition.

However, for Uyghur respondents across the region, the 'food boundary' was the primary means of drawing ethnic distinctions vis-à-vis the Han, and by extension of resisting Chinese hegemony. This particular boundary served to emphasise the failure of Han people to adhere to Islamic social norms and was deployed by respondents as a means to maintain symbolic, spatial and social segregation on a daily basis. Yet Uyghur respondents often compromised this boundary, most often in situations where they stood to gain socially or financially. For instance, they found jobs in state work units and moved to New Towns in order to secure economic stability: a clean, modern home; a steady income; and a better quality of life. Uyghur restaurant owners permitted Han customers to use the same chopsticks, cutlery, and crockery as their Uyghur customers because they needed (or wanted) the extra custom. Uyghur university students opted voluntarily to share dorms with Han students because they understood that they could improve their studies and career opportunities if they created the conditions in which to practise and improve their spoken Chinese. Uyghur petty entrepreneurs co-operated with Han counterparts at the local market because they knew this to be a winning strategy in terms of maximising Han custom in the catering business. In all of these contexts, priorities related to study, work, business and/or livelihood overrode supposedly non-porous religious requirements. However, the compromises made in these situations were not extended to other situations. Presumably, it should have been possible to invite Han people to Uyghur social events provided that they conformed to Uyghur dietary habits; yet for the most part Uyghur respondents were evidently unwilling even to consider this.

Yet we know that Uyghur and Han residents interacted more readily in the past. Urban Uyghur children in the 1960s and 1970s had played in Han homes on the mutual understanding that differences of diet could be 'managed'. It had been common for Uyghur colleagues to visit Han colleagues *in their homes* during the Spring Festival in the 1980s (Aysha Eli, communication). Previously, a Han person might even marry a Uyghur, provided that he or she agreed to convert to the Muslim diet. In the 1990s, however,

#### making culture matter: symbolic, spatial & social boundaries 169

there were almost no instances of Uyghur children or adults socialising with Han neighbours, and the force of public disapproval had made intermarriage practically impossible. That Uyghur children no longer had Han friends, Uyghur adults no longer stepped into Han homes at festival times, and intermarriage between adults was no longer an option suggests that, in addition to long-standing religio-cultural differences, there were other, new factors making the estrangement between Uyghur and Han adults more pronounced, and leading Uyghur parents to keep themselves and their children firmly apart. A similar reversal of behaviour had been seen in India in the 1930s, when renewed tensions between the Muslim and Hindu communities led to the birth of a Muslim nationalist movement and the 'Two Nation Theory'. In the changed political context, Muhammad Ali Jinnah argued that despite centuries of former coexistence, there had never been a single nation in India but rather many nationalities and peoples; and that while Hindus and Muslims lived side by side, they neither intermarried nor dined together because they belonged to two different civilisations based on conflicting ideas (Esposito 1991: 93-94).16

So, what new factors had brought about this deepened estrangement? To assert that increased interaction between Uyghur and Han peoples had by itself led to heightened ethnic tensions would be simplistic. In fact, the proportion of Han settlers in the total regional population had already reached 40 per cent by 1970, and has stayed relatively stable ever since according to the national censuses taken in 1990 and 2000.17 On the other hand, instances of ethnic conflict did not accelerate until the end of the 1980s (Dillon 1995; Smith 1999, 2000; Millward 2004). Firstly, we must acknowledge the 'contamination effect' created by the image of collapsing Marxist-Leninist politics in Eastern Europe in 1989 and a disintegrating Soviet Union in 1991. Following the latter event, five independent Central Asian republics were established adjacent to Xinjiang (Kazakhstan, Kyrgyzstan, Tajikistan, Uzbekistan and Turkmenistan), with the result that the Uyghurs, the Tatars and the Salars became the sole remaining Central

<sup>16</sup> Jinnah was leader of the All-India Muslim League from 1913 until Pakistan gained independence in August 1947. He rose to prominence in the Indian National Congress, initially expounding ideas of Hindu-Muslim unity but later embracing the goal of creating a separate Muslim state. He was Pakistan's first Governor-General from 15 August 1947 until his death on 11 September 1948.

<sup>17</sup> The pattern remains the same in the 2010 census, with the Han population static at 40 per cent owing to a more rapid rise in the local Uyghur population (see *The Economist* 2011).

#### 170 chapter three

Asians in Xinjiang without an independent country to their name.18 Moreover, the proliferation of international news dissemination via the national and global mass media across the eighties enabled Uyghurs to sit and watch as their Muslim cousins took control of their own economic, political, social and cultural structures and policies, as well as of their abundant natural resources.

Yet these external conditions would not have triggered mass resistance to Chinese rule in Xinjiang had state policies been working and had local Uyghurs felt fairly treated. Clearly, there were also many internal factors at work, above and beyond religio-cultural differences, for while those differences had existed between the two groups all along, Uyghur attitudes towards the Han had changed significantly over time. It is often said that Uyghurs initially 'welcomed the Han Chinese with open arms' (see Chapter 4 on Ömärjan Alim's song 'Mehman Bashlidim'/'I Brought Home A Guest'). Yet Uyghurs in the 1990s stated passionately that they wanted the influx of Han migrants to stop. The fact that many respondents distinguished between the amicable relationships they enjoyed with first-generation migrants and the conflicts they have experienced with more recent arrivals further confirms the change of heart. The good relationships they maintained with original settlers were characterised by a willingness on the part of Han residents to learn the Uyghur language and to adapt to, and participate in, Uyghur social and cultural practices. Looking back to Republican times, it is known that Uyghur *mazar* (shrines) were frequented by Han as well as Uyghur pilgrims (Bellér-Hann 2008: 333), and that the two groups occasionally joined forces in communal prayers for rain: 'Interethnic cooperation was employed to ensure the efficacy of the rituals; Islamic prayers were accompanied by the recital of Chinese prayers' (Bellér-Hann 2008: 351). Worsening relations in the 1990s, on the other hand, resulted from a growing unwillingness on the part of new Han migrants to embrace or even try to adapt to Uyghur culture. As a result, deep-rooted prejudices on both sides resulted in the limiting of Uyghur-Han interactions to working or formal relations (cf. Yee 2003: 449).

Perhaps most salient among the internal factors was a growing sense among Uyghurs that they were the victims of widespread ethnic prejudice

<sup>18</sup> The Salars are said to have originated from a Turkmen tribe (Schwarz 1984: 39–40), and therefore might be said to have their own country in Turkmenistan. The Tatars and the Salars in Xinjiang numbered only 4,821 and 3,660 persons respectively in 1990, compared with 7,194,675 Uyghurs (Hoppe 1992: 360).

#### making culture matter: symbolic, spatial & social boundaries 171

and socio-economic inequalities. In that context, the use of religio-cultural differences as symbols demarcating Uyghur-Han boundaries may be seen on another level as the articulation of demands for inter-ethnic equality. Faced with the daily reality of socio-economic marginalisation at school, university, and in the labour market, many Uyghurs felt deep resentment at perceived Han privilege. Many were excluded from job opportunities in state work units and Han-managed companies by default because they did not possess the necessary level of fluency in their second language, Chinese, and were not assimilated to the Han culture to an acceptable degree. Unqualified for white-collar jobs, most ended up doing blue-collar jobs (in the cities) or remaining in subsistence farming (in rural areas). The daily sight of urban Han residents living in clean new housing and driving (and apparently owning) brand new cars, or televised 'success stories' of new Han migrants to southern Xinjiang, only served to exacerbate feelings of social injustice. While the state had clearly developed an exciting new labour market in Xinjiang, including projects which sought to exploit the region's natural resources, few urban Uyghurs believed that they had profited. Although most Uyghurs had seen an improvement in their standard of living since 1949, urban dissatisfaction stemmed from the fact that Uyghurs now compared their situations and living conditions with those of their Han neighbours. Thus, while religio-cultural differences alone might have been managed in such a way that Uyghurs and Han could interact in situations where those differences were not felt to matter, the added complications of ethnically based discrimination and socio-economic inequalities led Uyghurs in the 1990s rather to exaggerate those religio-cultural differences as an expression of symbolic resistance against the Chinese state and Han residents. Colin Mackerras suggested at the turn of the century that a tendency had developed among some Uyghurs to *want* the Han to look down on them and the economic gaps to grow 'so that they can feel greater resentments and find grounds for rebellion against the Han and against China'(2001: 300). While I agree that Han chauvinism and interethnic inequalities make good weapons for young activists seeking to mobilise ethnic peers into seeking separation from the Chinese nation, this strategy could not possibly have succeeded had majority-minority discrimination and inequalities not existed in the first place. Similarly, the pattern of increased segregation in the late nineties noted by Nicolas Becquelin, especially in the south of the region, resulted not from instrumentalist strategies towards independence (which had in any case waned after the

# 172 chapter three

1997 Ghulja disturbances) but from renewed state repression of Uyghurs' religion and customs and of local economic elites (2000: 90).

#### Chapter Four

illuminists: popular song & the waking of the uyghur nation 173

# IllumInIsts: popular song and the wakIng oF the uyghur natIon

anthropologist stevan harrell once stated: 'the answer to whether the subaltern can speak is that the subaltern can speak on the sufferance of the civilizer' (1994: 34). according to this thesis, peripheral peoples have a voice only when granted one by the centre, and the centre grants that voice only on condition that peripheral peoples speak in its favour. yet in some contexts it has been possible to promote an alternative politics through the vehicle of elite or popular cultural forms. In 1930s algeria, for example, Islamic reformers of the algerian association of ulama (aau)1 found fertile ground in popular culture—specifically, poems, especially those recited orally—as they sought to mobilise Islam for nationalist purposes: 'the major themes found in these works are a celebration of the past glories of Islam; the portrayal of the enemy as a powerful demon, an unbeliever (Christian France) who violates the law of Islam; the belief that the current humiliation in Islam at the hands of the unbeliever will be avenged' (esposito 1984: 82-83). such a strategy is not without precedent in Xinjiang. as Bellér-hann points out, legends about folk heroes, muslim saints, historical events, and so on, have long been communicated in oral form via a technique of memorising which formed the core of formal and informal learning: 'orally transmitted tales and songs were potentially the property of all, they could be repeated at home and handed down to the next generation' (2000: 32-33, 39). uyghur intellectual elites of the 1920s employed a 'poetry of resistance' and oral recitation as a way to disseminate political ideas and encourage uyghur nationhood against a backdrop of heavy taxation by Chinese warlord governors, yang Zengxin and Jin shuren (rudelson, 1997: 146-53). moreover, conditions of accelerated modernisation since the early eighties have paradoxically favoured the continuation of oral transmission rather than its disappearance. literate uyghurs regard the official

<sup>1</sup> the aau, founded in 1931 and formally organized on may 5th, 1935, was a body of muslim religious scholars who, under French colonial rule, advocated the restoration of an algerian nation rooted in Islamic and arabic traditions. the association was heavily influenced by the views of muslim jurist and reformer muḥammad ʿabduh (1849–1905).

#### 174 chapter four

media as unreliable owing to high levels of state censorship, and find orally circulated information far more credible than written sources. Indeed, oral transmission has been the only way to spread news concerning violent incidents provoked by the state family planning policy, such as attacks on hospital staff or local officials for the part they played in the performance of forced abortions on Uyghur women (Bellér-Hann 2000: 73, 84). From the late 1980s through the 1990s, orally transmitted popular culture played an important role in the mobilisation of Uyghur national identity in the context of dramatic political change at domestic and international levels. Political dissidence in Xinjiang, unable to find voice in the form of official associations, found alternative expression in the historical novel, folk song, humour, stories and oral histories (cf. Harris 2002, Smith 2007, Dautcher 2009 and Bovingdon 2010). It was in this context that the late Uyghur novelist, poet, and historian, Abdurehim Ötkür, became a symbol of Uyghur aspirations to national independence. In September 1996, an estimated ten thousand Uyghurs attended Ötkür's funeral in Ürümchi's Döngköwrük district; they carried his casket high above their heads, and stopped traffic for four hours. The works of certain folk singers from this period similarly struck an emotional resonance with the Uyghur populace, and in some cases transcended underlying social divides of oasis origin, occupation, generation, political orientation, level of education and degree of religiosity. By focusing on the inter-ethnic boundary between the Uyghurs and a monolithic Han 'Other', these singers blurred to a significant degree intraethnic divides within the Uyghur group itself. At the same time, the hierarchy of older/younger sibling inherent in state rhetoric surrounding Han-minority relations (Harrell 1994; Bovingdon 2002) was challenged and re-cast as a relationship of coloniser to colonised. This chapter deals with the third—and arguably most effective—means of symbolic resistance operating during the mid-1990s: the dissemination of ethno-nationalist ideologies via Uyghur popular song. It opens with two examples of the state's preferred type of minority artist—whom I term the 'musician-comprador'—before going on to discuss the advent of 'new folk' and the possibilities for mass mediation of ideas furnished by the pop music cassette. Through an interpretation of key songs chosen from among the repertoires of two rival 'Voices of the Uyghurs', Ömärjan Alim and Abdurehim Heyit, I show how lyrical metaphor served to:

a. re-cast the fraternal relationship (as represented by the state) between Han Chinese and Uyghurs as one of coloniser and colonised;

illuminists: popular song & the waking of the uyghur nation 175


At the same time, new folk paradoxically doubled as a locus for Uyghur disunity, as cross-cutting identifications of social class, religiosity and locality created divides between its two main protagonists. In the final section of the chapter, I draw on interviews conducted with the two musicians and/ or their close friends and family members to demonstrate the social, religious and regional bases of the discord between them.

# 4.1. *The Musician-comprador*

As Harrell (1994: 34) outlines, from the state's point of view, the ideal form of the subaltern voice is the 'compradore elite', a term which denotes members of peripheral or colonised peoples who willingly participate in the civilizing or colonising project. China's compradors tend to be linked to professional song-and-dance troupes [Ch. *gewutuan* 歌舞团], state-sponsored organisations known for the co-optation and control of traditional art forms and minority representation. Minority nationality artists who have trained in these troupes since 1949 learn how to present their ethnic identities in ways acceptable to the Chinese state (Baranovitch, 2001: 362- 66). When they perform at official live or television shows, the choice of works is normally vetted by troupe leaders and by the local Cultural Bureau (*Wenhuaju* 文化局), the state organ responsible for monitoring the arts (Mackerras, 1985: 71-2). Lyrics contain messages of 'nationality unity' (Ch. *minzu tuanjie* 民族团结), that is, unity among China's fifty-six nationalities, and often promote the government's minority policies (Harris, 2004: 9). Below, I consider two prominent examples of the musician-comprador in contemporary Xinjiang: Baha'erguli and Kelimu.

#### *Baha'erguli* (巴哈尔古丽)2

Bahargül Imin was born in Khotän in south Xinjiang. There, she joined the regional culture troupe in 1977 as a dancer, solo singer and presenter before becoming a 'soldier-musician' within the Xinjiang Military District culture troupe in 1980. In 1981, she entered the Shanghai Music College to learn singing under the direction of a Han music professor. Seven years later, in 1988, she entered the PLA Arts College under the tutelage of a Uyghur music professor. Across her career, Bahargül has won a number of prestigious awards. In 1984, she was named 'Outstanding Vocalist' at the first ever National Young Singers Television Awards (*Quanguo qingnian geshou dianshi dajiangsai* 全国青年歌手电视大奖赛), while in 1986 she came second in the Minority Song category at the same event. In 1995 she won the Bronze award at the China Central Television (CCTV) Chinese Music Contest, and she has twice won the highest accolade in the Chinese arts sphere (in 1996 for one of four songs she recorded for MTV, and again in 1997): the 'Best Works Award' (*Wu ge yi gongcheng jiang*五个一工程奖) conferred by the Central Ministry of Propaganda.3 Bahargül has been officially categorised as a Grade 1 performer (*yiji yanyuan* 一级演员) in the PRC.4

Bahargül's best known songs include 'Our Xinjiang is the Most Beautiful Place' (*Zui mei de hai shi women Xinjiang* 最美的还是我们新疆), 'Girls and Boys of the Tianshan' (*Tianshan ernü* 天山儿女), 'Hello Ürümchi' (*Wulumuqi, ni hao* 乌鲁木齐,你好), 'Please Taste a Piece of Hami Melon' (*Qing ni chang kuai Hami gua* 请你尝快哈密瓜), 'Anarkhan, My Black Eyes' (*Ana'erhan, wo de hei yanjing* 阿娜尔汗我的黑眼睛) and 'Those who Visit the Borderlands Don't Want to Leave' (*Ren dao bianjiang bu xiang zou* 人到边疆不想走).A glance at just two of these provides immediate insights into her popularity with state authorities. While she appears in Uyghur traditional dress in the music video for 'Our Xinjiang is the Most

<sup>2</sup> Information about Bahargül's career is taken from her online profile on Chinese website hudong.com: http://www.hudong.com/wiki/%E5%B7%B4%E5%93%88%E5%B 0%94%E5%8F%A4%E4%B8%BD (accessed 5th January 2012).

<sup>3</sup> The *Wu ge yi gongcheng jiang* (五个一工程) is a project initiated in 1992, and originally awarded a prize for the best works in the following 5 categories: drama (including traditional Chinese opera such as Peking opera, dance and musicals); film; TV series; books (social sciences only); and theoretical articles (social sciences only). From 1995, two more categories were added: songs and radio series. However, the award is still called *Wu ge yi gongcheng jiang*.

<sup>4</sup> The title *yiji yanyuan* is awarded to performers considered to have made an outstanding contribution to the sphere of arts and culture. Performers with this status are issued with a special subsidy by the PRC State Council.

# illuminists: popular song & the waking of the uyghur nation 177

Beautiful Place', and is surrounded by professional dancers—male and female—performing traditional Uyghur dance, she sings the song in Chinese (not Uyghur). This in itself is not necessarily problematic, given the multilingual quality of contemporary urban Xinjiang. However, she delivers the song in a typically Chinese vocal style, and the lyrics appear to suggest a Han settler's perspective on the region rather than a local one. Lines such as 'I've roamed many places, but our Xinjiang is the most beautiful place' evoke the notions of internal (Han) migration from East to West as well as Han 'ownership' of the Western regions; later on, 'Lights twinkle in the New Oil Town in the Tarim' offers hope to migrants for a more prosperous life; while one of the closing lines 'A piece of precious [valuable] land has attracted the attention of the world' conjures up rivalries over Xinjiang's natural resources, both past (the 'Great Game' of the early twentieth century) and present (foreign bids for oil exploration). The central and oft-repeated phrase *women Xinjiang* ('Our Xinjiang') is itself problematic, blurring as it does the important distinction between 'local' peoples and newcomers.5 While Bahargül's songs have been phenomenally popular with the Han Chinese audience, a post on a Uyghur exile web forum in 2005 illustrates the song's somewhat different impact on those who consider themselves indigenous to the territory:

I have recently travelled by express train between Ürümchi and Kashgar. I'm shocked to discover that all of its radio announcements on the train are in Chinese ONLY over the course of the whole trip. It even broadcast a song entitled as 'Xinjiang is the most beautiful place' by Bahargül both upon departure and arrival and even that song was also in Chinese! I felt like a stranger travelling in my hometown.6

The Han settler's perspective is conveyed still more clearly in a second song titled 'Those who Visit the Borderlands Don't Want to Leave', penned by Han composers Cui Zong and Tian Ge. In the first verse, Bahargül sings: 'I can't forget the walled New Town', 'I can't forget the desert oases', and even 'I can't forget the Qaramay oil', before going on to laud the beautiful fields and gardens of Qäshqär and the pastures of the Heavenly Mountains. The verse concludes with the words 'The apple tree branch has caught my sleeve and won't let go', further suggesting an attachment to the 'fruits' of the region. In the second verse, Bahargül proclaims: 'I'm in love with the

<sup>5</sup> See also Dwyer (2005: 30) on the term *Xinjiangren*.

<sup>6</sup> Post dated 27th April 2005 on the website of the Uyghur American Association (UAA) based in Washington. http://forum.uyghuramerican.org/forum/showthread.php?940- Chinese-ONLY&s=b59fabb2993dae07ccd8a940ff74ac61 (accessed 5/1/12).

#### 178 chapter four

perfumed snow lotus;7 I'm in love with the magnificent tamarisk;8 I'm in love with the grapes of Turpan; I'm in love with the fine wines of the Ili valley', again underlining a focus on the region's superior natural environment and its produce. Ultimately, the song wraps up with the stereotyped image of local Turkic hospitality:


The whole is accompanied by the inviting (or ominous) refrain *Lai, lai, lai, lai, lai …* (Come, come, come, come, come …). Uradyn Bulag provides a vivid description of a similar representation in Inner Mongolia, in the form of a gigantic statue of a Mongol 'hostess', standing before the Hohhot municipality building with a ritual scarf and wine cup in her hands, and intended to symbolise the warm Mongolian welcome to the Chinese 'guests'. In this case, the host-guest roles are in reality reversed given that the Han population in the city now stands at 90 per cent of the total (2002: 15). Later in this chapter, I will contrast these ideas with what Ömärjan Alim has to say in his song 'Mehman Bashlidim' [I Brought Home a Guest].

<sup>7</sup> *Saussurea*: a genus of perennial herbs in Eurasia and North America, having mostly pink or purple tubular flowers gathered in heads that form a common inflorescence. Most species are found in the subalpine and alpine zones of mountains, growing in meadows and mountain tundras and on stony slopes, screes, and cliffs.

<sup>8</sup> *Tamarix chinensis*: a shrub or small tree of Eurasia, having small, scale-like leaves and clusters of white, pink or red flowers.

<sup>9</sup> Lyrics and sheet music retrieved from www.tom163.net (accessed 5/1/12). The translation is this author's.

#### illuminists: popular song & the waking of the uyghur nation 179

Popular reaction to Bahargül in the regional capital Ürümchi was mixed. Interviewed in 2002, Zunun, a service industry employee in his thirties, described her as an 'army musician', and her role within the PLA as one of 'singing songs to raise the troops' spirits'. Perhaps this explains why Bahargül sings mainly about the beauty (and booty) of the region, in an effort to convince local troops that the 'new dominion' is worth protecting and hardships worth enduring. In this sense, the songs can be seen as directed at both Han troops and Han settlers, all the more so in the aftermath of the 2009 Ürümchi riots and subsequent paranoia surrounding alleged syringe attacks.10 While this informant conceded that Bahargül's tendency to sing in Chinese rather than Uyghur rendered her less appealing to some sectors of Uyghur society, he also highlighted the attraction for many Uyghurs of her 'sonorous' voice. A second respondent, Jelil, an observant graduate in his late twenties, claimed in 2002 that Bahargül had lost popularity within the Uyghur community following her divorce from her Uyghur husband and subsequent remarriage to a Han Chinese: 'No-one likes her now. Before, she was married to a Uyghur, a good man with a good job. They had two beautiful children. She was really blessed [*bäkhtlik*], you know? But then she became friends with this Han who produced her show […] So I don't want to listen to her songs.' In Jelil's view, Bahargül's fall from grace was precipitated by the social stigma surrounding her divorce and exogamous marriage (or, more precisely, her remarriage to a Han).11 He refuted the suggestion that her delivery in Chinese was off-putting for the Uyghur audience, pointing out that she also sang songs in Uyghur, and that even had she sung entirely in Chinese, some groups of Uyghurs would have continued to like her, for example, the Chinese-educated (*minkaohan*) youth.12 I have explored the increasing irrelevance of language mode in the sphere of popular culture elsewhere (Smith Finley, 2011b). Certainly, I was able to find online footage of Bahargül performing Uyghur folk songs in the Uyghur

<sup>10</sup> See Smith Finley (2011a: 83-84) on the breakdown of confidence among Han residents following the Ürümchi riots in 2009.

<sup>11</sup> I discovered later in an interview with a close family friend that Bahargül's situation had in fact been quite complex. According to this friend, Bahargül had been very unhappy with her first (Uyghur) husband because the combination of her showbiz lifestyle with his male pride and jealousy had caused the two to quarrel endlessly. Interestingly, this respondent's own mother was a performing artist, whose marriage had similarly ended in divorce as a result of subverted gender roles: 'My mother was always out and about socialising, and my father couldn't bear it. Even though he's a *minkaohan*, he's still a Uyghur man, right?'

<sup>12</sup> See Baranovitch (2007b) on the popularity among Uyghur youth of pop idol, Arken Abdulla, who sings apolitical songs in both the Chinese language and the Uyghur language.

#### 180 chapter four

language and, judging by the related discussion threads, this seemed well received by at least some Uyghur fans. One post tagged to the 'Mudänkhan' video on YouTube in 2007 read in Uyghur: 'Sizge kop rehmet!!! [Thank you very much!] This is my favorite Uighur classic. I can't get myself off this piece since I found it yesterday. Bahargul sings it very well!!!'13

In addition to promoting the notion of Han belonging within the Western regions through the performance of songs that uphold Chinese national integrity, Bahargül has otherwise proved useful to the authorities by acting as ambassador for China's Islamic population on significant political occasions. In 1979, she went to Pakistan as part of a China youth delegation, where she performed 'Long Live Pakistan' in Urdu to great local acclaim. In 1985, she performed the ironically titled 'Sweet Song to Greet our Honoured Guests' at the 30th anniversary celebrations of the establishment of the Xinjiang Uyghur Autonomous Region. This choice of song involved a double meaning since, in one sense, the 'honoured guests' referred to were the Chinese politicians who had travelled from Beijing to join the regional celebrations; however, the underlying reference to 'guests' in a broader sense—Han settlers in Xinjiang—is unmistakeable. While the artist may or may not have chosen this song herself, she nonetheless proliferates the idea of 'peaceful liberation' of Xinjiang by acquiescing in the song's delivery (again, contrast this with Ömärjan Alim's *Mehman*, discussed later in this chapter). In 1986, Bahargül joined a Xinjiang artists' delegation to the UAE and Saudi Arabia, promoting the state's preferred image of a China in which Muslims flourish; and in 1992 she attended the World Festival of the Chinese People (*Shijie huaren jie* 世界华人节) in Hong Kong, representing the Uyghur group as a 'Chinese people', or at least as a willing Muslim minority in the majority Chinese nation.

#### *Kelimu* (克里木)14

Kelimu was born into a musical family in 1940 in Turpan, east Xinjiang. His father, at one time Artistic Director for the local song and dance troupe,

<sup>13</sup> Posted by amina0311. http://www.youtu.be/watch?v=e3GN6muuCh0&feature=relat ed (accessed 5/1/12).

<sup>14</sup> Information about Kelimu's career is taken from a number of online profiles posted on Chinese websites, including: http://baike.baidu.com/view/1306300.htm; http://www. xuje.com/yxz/left28/index.html; and http://www.rwabc.com/diqurenwu/rw\_detail. asp?people\_id=4606&id=7652(all accessed between 6/1/12 and 8/1/12). See also the article posted at: http://ent.sina.com.cn/v/m/2005-03-22/0935682674.html.

#### illuminists: popular song & the waking of the uyghur nation 181

was famed for his mastery of the *suona* [shawm],15 while his mother was an acclaimed dancer. The young Kelimu is described in Chinese sources using the stereotype perennially applied to minority peoples of the Northwest and beyond: *nenggeshanwu* ('having a flair for song and dance'). At the tender age of eleven, he joined the army as a 'literary soldier' (*wenyibing*文艺兵) and began to study dance, before moving to Beijing at 18. He is credited with being the first singer to perform Uyghur folk songs live in Chinese (in 1959), an act which purportedly 'helped Han people to learn about the Uyghurs', and which gained the praise of Chairman Mao himself. Kelimu was the first person from Xinjiang to join the Communist Youth League (1957), and would later become a member of both the Chinese Communist Party and the Chinese People's Political Consultative Conference (CPPCC).

Often seen on stage in full PLA uniform, Kelimu enjoys quasi military status and describes himself as 'a soldier in the service of the people'.16 His willingness to give concerts in remote rural areas of China proper (famously, in a mountain village in Hunan), together with his aim of 'encouraging a deeper understanding of and positive outlook on life', have earned him the nickname 'the people's artist'. He also regularly performs for PLA border defences stationed in remote areas such as Tibet. In assuming the all-embracing identity of a proud *Xinjiangren* [Xinjiangese]—'Come, friends, come and see our Xinjiang! [*Lai ba, pengyoumen, kankan women de Xinjiang*!]'—it could be argued that he reproduces the image of the Uyghurs as 'benign hosts'. In this way, Kelimu, like Bahargül, fulfils the function of attracting new Han settlers to the West as outlined in the post-2000 Great Western Development campaign. For his artistic services—or

<sup>15</sup> The *suona* is thought to have been developed from the Central Asian instrument known variously as *sorna*, *surnay*, or *zurna*, of which the Chinese name is likely a transliteration. It has a conical wooden body like that of the European oboe, but incorporates a tubular brass or copper mouthpiece to which a small double reed is affixed, and a detachable metal bell at its end.

<sup>16</sup> Over the course of his career, Kelimu earned the nickname 'Soldier Afanti'. Afanti (阿凡提) is the Chinese transliteration of Äpändi, a 13th century Sufi mystic, who appears in thousands of popular folk tales across the Turkic and Islamic world. For examples of these tales in the original Uyghur and English translation, see the website of the London Uyghur Ensemble at http://www.uyghurensemble.co.uk/en-html/en-ependi-letipe.html. Known as Nasreddin in Persian and Arabic, and Nasreddin Hoca in Turkish, this populist philosopher is sometimes witty, sometimes wise, but very often the butt of a joke. Nasreddin stories usually have a pedagogic nature. Although most are set in an early small-village setting, the tales, like Aesop's fables, deal with timeless concepts and purvey a pithy folk wisdom that invariably triumphs over all.

#### 182 chapter four

perhaps more accurately for his contribution to the safeguarding of a unified China—he was among the first cohort of 'literary soldiers' to be accorded the status of Civilian General (*wenzhi jiangjun* 文职将军).17

Although principally a singer (treble voice) and dancer, Kelimu is also an acclaimed songwriter, and has won national awards for such dubious compositions as 'Uncle Qurban, Where Are You Going?' (*Ku'erban dashu ni shang nar?*), 'Lamb Kebabs Oh So Fragrant!' (*Yangrouchuan xiang you xiang*), Afanti Song (*Afanti zhi ge*), 'Tarim River, River of My Hometown' (*Talimu he, guxiang de he*), and 'Ode to the Beloved Party [CCP]' (*Songge xian gei qin'ai de dang*). According to the artist's profile published on baidu. com,18 this song repertoire 'brims with ethnic characteristics and humour', and Kelimu 'wears a smile wherever he goes.' Needless to say, this representation does much to perpetuate the stock image of singing, dancing, smiling, joking ethnic person found in state and popular Han discourses (Blum 2001: 69, 74-5). Kelimu has twice performed in Chinese Central Television's Spring Festival Variety Show (1998, 2008), and is said to be the 'darling' of his (Han) audience. His original songs have been performed and covered far and wide. Yet to what extent do his creations truly reflect the 'authentic flavour of Xinjiang' (*dididaodao de Xinjiang wei*), as claimed by his Han fan base? Consider the refrain in 'Lamb Kebabs Oh So Fragrant!':


<sup>17</sup> The honorary title *wenzhijiangjun* (文职将军) was introduced in 2001. It makes provision for civilian army personnel at skills level 3 (on a scale of ten) to be accorded the status of Major-General. Civilian cadres at or above skills level 3 wear an upgraded uniform, including a cap adorned with gold cord and epaulets decorated with pine branches and the General's badge, and are addressed as 'General' (将军). Cadres at skills level 1 are accorded the status of Lieutenant General (中将), those at skills level 2 the status of Major-General (少将), and those at skills level 3 the status of Colonel or Major-General (大校、少将).

<sup>18</sup> The online Chinese-language search engine Baidu was founded in 2000 by internet pioneer Robin Li. The name Baidu (literally, 'hundreds of times') is inspired by a Song Dynasty poem, which compared the search for a retreating beauty amid chaotic glamour with the search for one's dream while confronted by life's many obstacles: '… hundreds and thousands of times, for her I searched in chaos, suddenly, I turned by chance, to where the lights were waning, and there she stood.' http://ir.baidu.com/phoenix.zhtml?c=188488& p=irol-homeprofile (accessed 11/2/12).

#### illuminists: popular song & the waking of the uyghur nation 183

### 更加爱新疆. She loves Xinjiang still more!19

Far from conveying the complexities of Uyghur culture, religion and society to the Chinese audience, this song seems to serve rather as an advertisement to would-be Han settlers. The focus is firmly on food culture (produce)—particularly lamb kebabs—and the image conveyed is of a well-nourished lifestyle in a new land with which newcomers quickly fall in love. Certainly, the song contains underlying ironies (although it is impossible to discern whether Kelimu intends these). For example, the lines about putting on weight and gaining in strength might be interpreted as a jibe at Han men, whose slight and effeminate build is often ridiculed by their heftier Uyghur male counterparts (recall Chapter 2 on counter-stereotypes). Furthermore, the song as a whole evokes the controversial issue of Han settlers rapidly developing a taste for lamb and causing meat prices to rocket, much to the horror of locals who find themselves priced out of the meat market. However, the enduring effect is of a simplistic, comic and above all *harmless* representation of Uyghur culture, delivered in Chinese and punctuated by transliterated forms (Chinese phonemes) of friendly Uyghur words that Han Chinese are likely to know, for instance, *ya-ke-xi* (*yakhshi*, good) or *sa-la-mu* (*Salam*, greetings). As I observed while travelling between Ürümchi and Ghulja with a Han taxi driver in 2002, it is these 'domesticated' forms of Uyghur song that have proved most acceptable to the Han audience. My companion enthusiastically introduced the artist to me thus: 'This is Kelimu. We Hans living in Xinjiang really like his songs; they're brimming with ethnic flavour!'

Kelimu's most recent project was a film broadcast during the Spring Festival 2010, entitled: *The Hometown Sonata: A New Year's Comedy*. Describing this as the pinnacle of his career, Kelimu said he hoped the film would 'make Xinjiang better known in China proper and across the world'. The project was, I would argue, deliberately constructed to reflect the state's ideal political configuration of a (re-)unified Greater China, involving a mixed-ethnicity cast consisting of Kelimu, two Taiwanese artists Wu Qilong and Wu Chenjun, mainland Chinese stars Chen Lina and Wang Dongfang, and two child stars from Xinjiang, A'erfa and Zilalan. The interwoven ethnic fabric seemingly lacked only a Tibetan artist. Filmed on location in Turpan in 2009 and broadcast on 11th February 2010 by 35 provincial and city-level television stations across the PRC, the film entailed a comic look at contras-

<sup>19</sup> Lyrics retrieved from the PRC tourism website *Wulumuqi lüyou* (Ürümchi Travel). http://www.wlmqxj.com/gl/wh/1391.html (accessed 8/1/12).

#### 184 chapter four

tive national (ethnic) customs. In the style of a comedy of errors, a music teacher runs into an arrogant female company director in 'traditional and mystical Turpan', where they encounter a group of 'kind and generous children' and an 'honest ethnic uncle' (played by Kelimu). According to the plot synopsis, the ensuing story is characterised by a 'relaxed, happy, humorous and rich ethnic style', creates 'endless laughter', and 'plays an important role in conserving ethnic culture'.20

# 4.2. *Popular Singers as Illuminists: The Advent of 'New Folk'*

So much for the musician-comprador; however, this co-opted version of minority artist constituted only a single thread within the musical fabric of Xinjiang in the 1990s. Along with negative oral stereotyping (Chapter 2) and processes of inclusion and exclusion via ethnic boundary construction (Chapter 3), an alternative form of popular song—'new folk'21—proved more effective in uniting Uyghurs against the Han than the earlier 'poetry of resistance', than competing attempts by regional intellectual elites to project a national consciousness in the late eighties (Rudelson 1997), and even than Islam. For the purposes of the following discussion, I adopt the third orientation of the term 'popular' as deployed in popular music and cultural studies: popular-as-populist (Kassabian 1999: 116-17). This approach focuses on ways in which artists and audiences express social and political positions through music production and consumption. Over time, discourses on the term 'popular' have gradually embraced the notion that audiences consciously engage with their popular culture, rejecting the Frankfurt School position that they are passive victims of a 'false consciousness' imposed by music producers as part of the dominant culture. In other words, the ways in which musical texts are used and interpreted are the 'ongoing product of people's attempt to represent their own experiences, and to speak in their own voices instead of hegemonic codes' (Manuel, 1993: 8).

In the Xinjiang context, the popular-as-populist model requires modification to account for local political conditions: artists in democratic societies can be openly subversive and, indeed, deliberately seek sub-cultural capital through open rebellion. However, artists representing

<sup>20</sup> http://baike.baidu.com/view/2973859.htm (accessed 8/1/12).

<sup>21</sup> This term was coined by Harris in her survey of the popular music industry in Xinjiang (2002: 272).

#### illuminists: popular song & the waking of the uyghur nation 185

dominated ethnic groups in the PRC cannot so readily challenge state hegemony. The Chinese empire is no stranger to the cynical use of music in nation-building; it has used popular music and song variously to assess the attitudes of the empire's subjects, to mobilise the nation against foreign invaders (most famously, the Japanese), and to represent the nation in a way that stresses ethnic inclusiveness (Tuohy 2001: 110-118; see also Perris 1985: 93-122). It has even employed musical repetition—daily broadcasts on television, radio and public loudspeakers—in its bid to create a 'voice of the people' (Tuohy 2001: 117; 123). Mindful of this and of the potential political repercussions of 'subversion' (in any form), 'new folk' singers remained purposely enigmatic when questioned about meanings and motives behind their texts. When in 2002 I met with family members of the artist Ömärjan Alim (discussed later in this chapter), they declined to discuss those of his songs which address the issue of indigenous collaboration with the Chinese state (beyond a discreet chuckle at the mention of the song title 'Barren Chickens'). In fact, Alim's wife seemed reluctant to discuss song content with me at all, observing only: 'People may understand the words in different ways … it won't be the same for everyone.' They carefully explained that the artist did not write the song words himself. His songs, they said, were based on texts produced by writers and poets, and *brought to him by other people*. Presenting Alim as a willing minstrel who would perform anything so long as someone wished it, they denied any deliberate intent on the part of the artist: 'Ömärjan isn't responsible for what he sings; he doesn't know what it means. He's uneducated, isn't he, without knowledge [Uy: *bilimsiz; bilim yoq*].' It did not seem to matter that while it was common knowledge that any Uyghur—including uneducated ones—could grasp his song words, she claimed that Alim himself could not. I later learned that this line had been successfully used to defend Alim against charges of 'ethnic splittism' in court.22

The possibility of multiple interpretations meant that new folk singers could often avoid detection by the government censor. Even without the use of metaphor, the Chinese state has sometimes struggled to interpret song lyrics. State leaders in the 1980s and 1990s, for instance, were unable to reach agreement on whether rock singer Cui Jian's lyrics about the Chinese Communist Party's revolutionary heritage were patriotic or ironic (Tuohy 2001: 118). But, while song words remain on the surface ambiguous,

<sup>22</sup> Personal communication (November 2002) with Rachel Harris, School of Oriental and African Studies, University of London.

#### 186 chapter four

the people can—and do—read what they like into them, and their readings are invariably tempered by the social, political, and economic climate in which they live. This fact was readily acknowledged by a Han scholar, speaking in 2002, who observed: 'They always write indirectly, so that if you confront them, they can deny that they meant what you think they did. What they write is nonsense, but people believe it. So it must be corrected' (cited by Bovingdon 2010: 81). Regardless of whether artists were willing (or able) to claim agency, the act of consumption of their material by the Uyghur people was political: 'a process of making meaning from, and contributing meaning to, popular culture' (Kassabian, 1999: 115).

'New folk' music differs radically from that performed by state-sanctioned song-and-dance troupes, consisting mainly of solo recordings of contemporary compositions. In terms of performance style, it enjoys an expressive freedom afforded by the octave-based modal structure (usually heptatonic or pentatonic). Its unique rhythms, associated with the traditional folk singing style and traditional stringed instruments such as the *dutar* (two-stringed lute), are characterised by extensive use of asymmetry and syncopation. Moreover, its song lyrics address prevalent social and political issues, often borrowing or adapting lines penned by well-known and less well-known—poets and writers. New folk was successful in conveying nationalist ideas in the nineties for a great many reasons. Firstly, it carried messages to the Uyghurs in their preferred cultural form. De Vos lists aesthetic cultural forms such as food, dress, music, and song as one group of 'markers of cultural difference' that might be selected by group members in the subjective definition of their group identity (1975: 9). Song and dance have been integral to Uyghur culture since ancient times and continue to be the mainstay and focal point of Uyghur social life, predominating at *olturush* ('sittings' or gatherings) in the home, birthday parties, Uyghur dance restaurants, weddings, *mäshräp* [gatherings for feasting, story-telling, and music-making],23 university dance halls, and the bazaar

<sup>23</sup> These days, women can take part equally with men in a non-religious form of *mäshräp*, intended to allow people to rejoice through the medium of the performing arts (Mackerras, 1985: 68). Up until the 1950s, however, participants were all male. The *mäshräp* functioned as a rite of passage into manhood, a vehicle for regulating moral, religious, and social etiquette, and a means of forming male peer groups. In the mid-late 1990s, the 'traditional' form of *mäshräp* was resurrected in the city of Ghulja, North Xinjiang, and among Uyghur communities in Almaty, Kazakhstan, with the modern aim of creating and maintaining Uyghur national culture (Roberts, 1998b: 675; Millward 2004: 16-17; Dautcher 2009: 272-282). Such *mäshräp* quickly came under government suspicion in the People's Republic as popular arenas for the dissemination of separatist ideologies and literatures.

#### illuminists: popular song & the waking of the uyghur nation 187

with its myriad cassette kiosks (Mackerras, 1985: 62-75; Smith, 1999: 231-33; Harris, 2002: 267-70). The saying, 'When a Uyghur child can walk, he can dance. When a Uyghur child can speak, he can sing' continues to be widely cited by Uyghur youth today; see for example a *New York Times* reporter's conversation with Abdul Gheni, a young Uyghur in Qäshqär 'who had discarded the traditional skullcap in favor of a baseball cap worn backwards' (Strauss 1999). Significantly, a love of, and flair for, musical performance is also one criterion by which Uyghurs differentiate themselves from the Han whom they often dismiss as dull and introverted (see Chapter 2 on counterstereotypes).24

The orality of popular song also made it accessible to Uyghurs from all backgrounds, rural and urban. Although the late Uyghur novelist, poet and historian, Abdurehim Ötkür, was an undisputed symbol of Uyghur nationalist aspiration in the 1980s and 1990s, his media—poetry and the historical novel—were circulated predominantly in urban, often intellectual, spaces.25 The extent to which his works reached the rural population—with its higher instance of illiteracy—must surely have been limited. Furthermore, the fact that the Uyghur language has undergone repeated script changes during the period of Chinese Communist rule means that Ötkür's work published in the Arabic script ('Old Script')—is inaccessible to a generation of Uyghurs educated in the Latin script ('New Script') between 1960 and 1982 (see Chapter 1). In contrast, song words, communicated orally, can be received and understood by all. In this way, new folk singers borrowed the poetry of Ötkür and other nationalist writers, adapted it for use in their song lyrics, and were able to transmit nationalist sentiments—originally expressed in written or print form—far beyond urban, intellectual circles and straight into the heart of the countryside.26 There is an interesting comparison here with the Albanian nationalist movement during the period from the start of the Albanian War of Independence in 1898 through the declaration of an independent Albania in 1912 to its recognition by the Great Powers in 1918. Sugarman writes that villagers gained a 'deep-seated

<sup>24</sup> Mackerras remarks that the spoken play [*huaju*] popular among Han Chinese has singularly failed to make an impression among the Uyghur because it lacks music, the ingredient so essential to their cultural identity (1985: 66).

<sup>25</sup> Ötkür's novels were fictional accounts of events that occurred before the Chinese Communists came to power in 1949. In them, he expressed themes and ideas relevant to the present, escaping the censor by dressing these in analogous historical situations (Rudelson, 1997: 163-65).

<sup>26</sup> Harris gives examples in Ömärjan Alim's work (2002: 278).

#### 188 chapter four

sense of themselves as being Albanian' only after nationalist poems were transformed into men's narrative songs (1999: 441-45). In this process, figures such as *çetë* [armed guerrilla unit] commander Sali Butka composed revolutionary poems that fused features of village song texts with nationalist themes in a form of folk poetry. Sugarman hypothesizes that *çetë* members, who were encouraged to become literate, began to sing the poems as songs, and that these were then taken up by fellow villagers when members returned home.

The third source of new folk's appeal was the capacity of its musical (sonic) texts to evoke emotion during a time of growing national awareness: the unique qualities of the vocal (predominantly a head voice but also articulated by nasal inflections) and the sounds of the *dutar* rendered it enormously affective. The traditional Uyghur singing style features subtle tone shifts of the melodic line, free melismatic ornamentation, and, as with other Turkic musical traditions, a tendency to employ ululations (an inflection that draws on howls and/or cries to give the tone a 'lift'). With Ömärjan Alim's and Abdurehim Heyit's songs being sung almost exclusively within a minor (and modally heptatonic) tonal structure, this voice sometimes evokes fragility and grief, sometimes rage and frustration. When combined with affective lyrical content, it can give rise to powerful emotions. On this subject, Harris cites a Uyghur song composer as follows:

A people who have suffered long oppression have soft hearts, they are easily shattered. There is much in their hearts that is unsaid. There is a special tragic note to their music. (2002: 273)

In a similar vein, *Time* journalist Matthew Forney characterised Heyit as the 'man of constant sorrow', in a reference to American political folk singer, Bob Dylan (2002).27 The *dutar*, meanwhile, assumes an atavistic role as, like the Serbian one-stringed *gusle*, its sound 'travels across the dark centuries' (Thomas 1999: 172-3), linking Uyghurs with their ancestors and sustaining cultural continuity through the retention of traditional strumming patterns.

Unlike co-opted Uyghur artists Kelimu and Bahargül, who sing for the most part in Chinese and (arguably) become mouthpieces for China's allinclusive national design, new folk singers sing in the Uyghur language.

<sup>27</sup> The Dylan comparison was more recently evoked in relation to Uzbek political singer/songwriter Dadaxon Hasanov; or, rather, to borrow the author's words, it would have been applicable 'if Bob Dylan had been stalked by the KGB and played the dutar' (Kendzior 2007: 322).

#### illuminists: popular song & the waking of the uyghur nation 189

This ensures that messages reach rural Uyghurs who may know little or no Chinese, and carries significant symbolic value. The Chinese state has frequently used vocal music as a vehicle for the dissemination of national standard speech (Mandarin) in the hope that local, regional, and ethnic loyalties will transform into national (Chinese) ones (Tuohy, 2001: 117). Where minority languages are retained in folk songs, the aim has been to represent ethnic diversity while lyrics stress inter-ethnic harmony. By favouring the Uyghur language, new folk singers in the 1990s not only challenged the assumed superiority of the Han language but also constructed an alternative (Uyghur) national voice. For these reasons, new folk became the ideal vehicle through which to reflect prevalent social and political concerns in Xinjiang and to construct and reproduce an alternative social, political and national consciousness. In presenting the phenomenon as a two-way process, I follow Susan Tuohy, who models musical nationalism in China as a 'mutually transformative process of making music national and of realizing the nation musically' (2001: 108-9).28

# 4.3. *Cassette Culture and Mass Mediation of Political Messages*

Disaffection with Chinese rule—and with the Han migrants who represent it—had until now been strongest in Xinjiang's urban centres. In the earlymid nineties it began to spread to rural areas via an affordable means of mass communication: the low-budget popular music cassette. As Peter Manuel showed in his study of mass-mediated cassette culture in North India, cassettes as a form of 'new media' are affordable, easy to produce, duplicate and pass on, and largely evasive of central control. This makes them the ideal vehicle for socio-political mobilization, even when banned by the censor (1993: 2-4; 238). A potent example of this process was observed in Uzbekistan following the Andijon massacre in 2005. A political protest song, composed by Dadaxon Hasanov and entitled '*Andijonda qatli om bo'ldi*' [There was a Massacre in Andijon], was distributed throughout Uzbekistan on audio cassette almost immediately after it was recorded. Serving as a mode of reportage through which illicit information may be spread, it quickly became a 'poem of witness' (Kendzior 2007: 318). Hasanov, who had sung provocative songs about the *vatan* (homeland) since

<sup>28</sup> Again, parallels can be drawn with processes observed in Albania. Sugarman sees Albanian literary figures as simultaneously *producing* the nationalist discourses of the Albanian Rilindja ('rebirth') and *being produced* as national subjects by them (1999: 421).

#### 190 chapter four

the 1970s, and whose works had been banned in Uzbekistan for two decades, was voted one of the most popular and respected figures in the country in 1991, a fact that attests to the prevalence and influence of underground music and media (Kendzior 2007: 324).

In Xinjiang, cassettes are transported from their place of production to large towns north and south of the Tianshan. Uyghur merchants then buy the cassettes from urban distributors and sell them in small rural bazaars. At each stage of the process, further 'pirated' copies pass from hand to hand. Moreover, as in China proper, growing numbers of young Uyghur men are leaving rural areas to look for work in urban centres (cf. Bovingdon 2010: 14 on political effects of educational and employment 'pilgrimages' to Ürümchi). These migrant workers become quickly familiarised with 'Us and Them'– Uyghur vs. Han—discourses (stronger in the city due to more intensive penetration by Han settlers) and with 'nationalist' songs heard at the bazaar. When they return to the countryside, a copy of the cassette may accompany them, or else the memorized song lyrics, in a process of oral transmission. The cassette is then played (or the song performed) in the rural home before relatives, neighbours and friends, and nationalist images and ideas are reproduced in rural settings. This 'performative' dimension renders popular song 'an active means by which to experience the nation' (Tuohy 2001: 109).

During this period, two key players emerged: Ömärjan Alim and Abdurehim Heyit. Like Inner Mongolian musician, Teng Ge'er (see Baranovitch 2001),29 they began to try to subvert Beijing's control over the arts in the pursuit of an alternative ethnic and national agenda. As each competed with the other in his bid to become the primary 'Voice of the Uyghurs', they attracted varying degrees of attention from the state censor, being prohibited from public performance and having their works confiscated on an intermittent basis. While both succeeded to perhaps an unprecedented degree in uniting Uyghurs from a variety of backgrounds, an interrogation of their relationship to one another reveals religious and social tensions that acted to undermine their efforts.

<sup>29</sup> Baranovitch writes that Teng Ge'er's 1989 release 'The Mongol' was still frequently sung in rural areas in the mid-1990s, suggesting that 'his alternative [ethnic] representation is accepted by a large portion of the people he claims to represent'(2001: 371).

#### illuminists: popular song & the waking of the uyghur nation 191

# 4.4. *Ömärjan Alim: The People's Choice*30

During my visit to Ömärjan Alim's home in Ghulja in 2002, family members told me how Ömärjan had learned to play *dutar* the old way: not by formally taking lessons or studying music at college, but from his own father via the traditional method of oral transmission.31 Asked why the artist enjoyed such popularity, his wife (who clearly sought to protect him) shrugged, gave an embarrassed laugh, and then replied: 'Who knows? It's impossible to say! I guess they like his voice, the way he plays.' Alim's 17 year-old daughter, however, was less discreet. Her response was: 'It's the content, the words.' Perhaps encouraged by her boldness, the mother added: 'Yes. His words are straight; direct [Uy: *ochuq*]; anyone can understand them.' This was a sentiment echoed by respondents from many different social backgrounds. Patimä, a *minkaomin* music seller in her thirties interviewed in Ürümchi in the same year, explained that Uyghurs (whom she referred to as 'We') liked Alim because the song words were 'very deep, very meaningful' [Uy: *mänisi chong*]. Polat, a *minkaohan* taxi driver in his forties interviewed in Ghulja in 2002, also identified content as the key to Alim's success: 'The meanings are pretty big, you know, heavy … they reflect the problems between the Uyghurs and the Hans. The way all the power lies in the hands of the Hans.' Aliyä, a postgraduate student living in Ürümchi, had listened to Ömärjan's songs ever since entering high school in the late eighties. She proclaimed him 'well-known to everyone' in her southern hometown and, although unable to recite a whole song, claimed that if she saw the singer on television, she could immediately recognise the song and explain the content. Tashmämmät, a young male on vacation in Ürümchi in 2002, confirmed that Alim was 'very popular' in his hometown of Khotän, and moreover that Khotänese peasants were all familiar with his songs. Anarkhan, a middle-aged woman from Ürümchi working in the service industry, owned many of his cassettes released in the 1990s, and explained in 2002 that she and her husband loved Alim's ability to *affect*: 'When we hear his songs, it does something to us … It gets us right in the chest, makes us feel really agonized [Ch. *nanshou*]'.32 Pressing her hands to her heart,

<sup>30</sup> An extended paper relating to this section on popular singer Ömärjan Alim, and presenting additional song material, was previously published in Biddle and Knights eds. (Smith 2007).

<sup>31</sup> See Bellér-Hann (2000) on the tradition of oral transmission of knowledge in Xinjiang.

<sup>32</sup> This respondent was a *minkaohan* who had attended a Chinese-medium school during the Cultural Revolution; as such, she spoke in a mixture of Uyghur and Chinese during interview.

#### 192 chapter four

she described this as 'a good sort of pain'. Her description suggested a clear identification with song content as well as a strong element of catharsis. Emotional release was also suggested by Häsän, a *minkaomin* teacher in his thirties from Kucha, interviewed in 2002, who had begun to listen to Ömärjan after starting university in Ürümchi: 'His words say what is in the people's hearts. When they listen to the music, they become broken-hearted.' In this way, Alim's clear and emotive vocals successfully solicited identifications among Uyghurs from disparate oases in the north and south, and from urban and rural, educated and working class backgrounds.

A 2002 study by Harris and Dawut demonstrated the important role played by catharsis in a ritual context. Focusing on pilgrimage events at the Ordam Mazar [shrine] festival, the authors describe a crowd of Uyghur pilgrims gathered to listen to a group of musicians performing 'Tashway', a mournful piece attributed to a nineteenth-century religious mendicant. According to their description, the *rawap* player is 'on the edge' and has a 'raw voice full of emotion'; the audience gives a long cry when the piece comes to an end; and, later, a tranced dancer 'weeps and shakes, speaking of his troubles as others weep at his words and make the movement of ritual cleansing' (2002: 105-6). At a second gathering at the Imam Hasim Mazar, the authors find Uyghur women performing *monajat* songs ('prayers of supplication' associated with grief and mourning), causing other women at the ceremony to weep openly. On the basis of their observations, Harris and Dawut find that *mazar* festivals 'serve as a focus of prayer and for the outpouring of personal grief' (2002: 108; 112). Building upon this, I would argue that the audience's experience of Alim's popular songs may be characterised as an outpouring of grief at the level of the group, or indeed the nation.33

Another aspect which may have attributed to Alim's popularity was his willingness to defy the state censor, at least as perceived by fans. Polat, a taxi driver interviewed in Ghulja in 2002, seemed almost proud as he talked about the singer's political difficulties: 'Alim's been in trouble, I think. He hasn't been allowed to sing or perform. They said his lyrics encouraged "ethnic splittism". I heard that he was in America, in fact.' Learning that Alim had not been at home when I visited his Ghulja residence, he questioned whether the artist had really been touring the south at that time, as

<sup>33</sup> Forney draws a comparison between Uyghurs and other 'oppressed peoples' in history, such as the Irish, the Gypsies and the Jews, suggesting that Uyghurs 'draw strength from music' (2002).

#### illuminists: popular song & the waking of the uyghur nation 193

his relatives had claimed, implying that he may in fact have been in prison.34 By romanticising and making a martyr of the 'people's voice' in this way, ordinary Uyghurs may attain, through them, a feeling of agency and control, a sense of satisfaction that they are resisting their plight (I return to this notion below in the discussion about Abdurehim Heyit's followers in Qäshqär).

Perhaps the most famous of Alim's songs—and one that was banned is 'Mehman Bashlidim' [I Brought Home a Guest], released on the cassette *Pärwayim Päläk* [Why Should I Care?] in the early 1990s. 'Mehman', which is apparently modelled on a poem included in a cassette compilation titled *Pighan* [Rooster's Cry] and published during the 1980s (mentioned in Bovingdon 2010: 95), employs an allegory in which the coloniser (Chinese state/Han settler) is depicted as 'the guest who never left' and the colonised (local Uyghurs) as slaves forced into the hostile desert:


It is not difficult to see this song's appeal in the context of a growing Han migrant presence, increased ethnic discrimination and accelerated exploi-

<sup>34</sup> Given that a relative at Alim's father's house had earlier in the day offered to ask his wife to ring him in Aqsu so that I could speak with him directly, I consider that the family's explanation on this occasion was the truth.

<sup>35</sup> This refers to the back half of the traditional *supa* platform, the best seat a Uyghur host can offer his guest.

<sup>36</sup> Translated from the Uyghur by the author (Smith 1999).

tation of local resources. Indeed, some have argued that the topic of mass Han migration has been fruitfully employed by 'separatists' as propaganda against the Chinese polity: 'This situation is very useful for them to inspire anti-Han sentiment' (Wang 2004). 'Mehman', like 'Those who Visit the Borderlands Don't Want to Leave' (discussed above), depicts the Uyghurs as hosts; yet here they are represented as not 'happy' but disenchanted hosts: deceived, robbed and ultimately excluded in their own homeland. Uradyn Bulag uses the term 'sheepishness' to describe the peacefulness of the modern Mongols in comparison with their historical ferocity (2002: 3). In contemporary Xinjiang, too, this metaphor is often deployed. Uyghurs remark that they are too much like the sheep that grace their dinner table—naïve, docile, and apt to follow others' lead—and joke that they should eat less mutton. As Tashmämmät, an intellectual in his forties from Qäshqär, proposed in 2004, this idea (arguably) reflects the tendency of Uyghurs throughout history to accept foreign hegemony at face value:

The Hans are a crafty race; their way of thinking is extremely sly. We [Uyghurs] say they show us their facade but not their rear … What they say and what they are really thinking are two completely different things! The Uyghurs, on the other hand, are a simple and straightforward [*addi*] people; they are gullible, easily fooled. This is how the Hans have been able to control us.

Shökhrät, a male intellectual in his thirties interviewed in 2004, employed the same notion to explain events on the eve of the Communist takeover, asserting: 'Uyghurs believed everything the Red Army said when they entered Xinjiang in 1949. Five years later, the political campaigns and purges had already begun; but by then it was too late.'

The song 'Mehman Bashlidim' proved hugely popular among Uyghurs from all social groups in 1995-6, when respondents would forcefully communicate Uyghur attachment to, and rightful ownership of, the land. Both Shökhrät, a young male intellectual, and Räwiä, a language specialist, accused Han Chinese of creating a 'myth of origin' regarding their presence in the region, and claimed that they frequently staged archaeological tricks in order to 'prove' it.37

Nor was this view restricted to the educated classes; migrant workers in Ürümchi were equally scornful of Chinese historiography, as when Batur, a baker, exclaimed indignantly in 1996: 'They [the Hans] say they were here first. They even say it in the books they write. They say they have *always*

<sup>37</sup> See also Bovingdon's account of a lively exchange between two women—one Uyghur and one Han—on the question of indigeneity to the region (2002: 49-52).

#### illuminists: popular song & the waking of the uyghur nation 195

been here!' A more recent example of this type of controversy concerns the origin of the mummies excavated at the Small River Cemetery No. 5 in the Tarim basin between 2003 and 2005, including the now-famous Beauty of Loulan. Li Jin, a Han geneticist at Fudan University, stated in 2007 that the mummies' DNA contained markers indicating an East Asian or even South Asian origin. In 2010, however, a team of Chinese scholars led by Hui Zhou of Jilin University published their conclusion that the people were of mixed ancestry, having both European and some Siberian genetic markers, and probably came from outside China. Li Jin was a co-author (Wade 2010). Hundreds of Western-looking Mummies had in fact been discovered in the 1970s by construction workers labouring on housing developments, roads and industrial sites in Xinjiang, but had been 'swiftly reburied' in the back rooms of a museum: a 'politically convenient place to stow the evidence and stonewall all the inconvenient questions they posed' (Pringle 2010).38

Harris has noted that Xinjiang's bazaars function as an 'unofficial pop chart', where the number of shops and restaurants playing a certain cassette provide a reliable guide to the latest hit (2002: 270). In the summer of 1996, one cassette by Ömärjan Alim blared ceaselessly from kiosks and eateries across Xinjiang: it was called *Qaldi Iz* [Traces]. In Ürümchi, Turpan and Kucha, in Qäshqär, Khotän and Aqsu, Alim's voice rang out in private and public, rural and urban spaces, temporarily turning this cassette into the Uyghur national soundtrack. A sound engineer involved in its production attributed its extraordinary success to Alim's ability to communicate with the peasants (Harris 2002: 278). But 'plain language' was clearly not the only source of the cassette's overwhelming popularity; this derived also from its lyrical content. As outlined above, when asked what they liked about Ömärjan Alim's songs, the vast majority of respondents indicated the song-words, which evidently articulated their situation as they saw it. Including the gentle title track, 'Qaldi Iz' [Traces], which draws on a poem and novel by the late nationalist writer Abdurehim Ötkür,39 almost every song invites reflection on common grievances and is conceived within an implicitly or explicitly assumed framework of political opposition to the Han.40 Yet the

<sup>38</sup> Currently, the mummies can be viewed at the Xinjiang Regional Museum in Ürümchi, opened in 2005, where they are displayed alongside much more recent Han mummies, a decision viewed by one visitor as a political strategy to confuse chronology (Coonan 2006).

<sup>39</sup> For an English translation of the poem 'Iz' [Traces], see Allworth and Pahta (1988).

<sup>40 &#</sup>x27;Wäsiyät'[Testament], which tackles the social problem of drug abuse and addiction among young Uyghurs, might be deemed an exception, although many respondents related

#### 196 chapter four

album as a whole was not banned. Only the song 'Qaldi Iz' was restricted, receiving no airplay on regional radio, while Alim was instructed not to perform the song in public. Given that 'Qaldi Iz' was hardly the most controversial song in the collection, the authorities' reaction begs the question: why were the other songs not censored? There are several possible explanations. Perhaps the authorities had simply been unaware of the content. If this were true, it might indicate a strengthened sense of in-group loyalty during this period, with the result that no member translated the song words for the censor.41 The title track 'Qaldi Iz', on the other hand, had come under immediate suspicion because it was based upon a celebrated poem by a known nationalist author. Alternatively, in-group informants may have considered the songs not worth reporting to the authorities given that the songs involved criticism of Uyghurs by Uyghurs, rather than of the Chinese state or the Han people.

Above, I analysed 'Mehman Bashlidim' as a metaphorical representation of the Han 'coloniser'. Following, I examine two songs taken from *Qaldi Iz*, the first dealing with the Uyghur 'collaborator' (those accused of placing personal ambition above ethnic loyalty), and the second reflecting the troubled quest for Uyghur national unity. Highlighting the singer's agency in group identity construction, I show how these representations reflected popular perceptions held among disparate groups of Uyghurs across the region, forging a common sense of 'emotional unity'. At the same time, it reproduced relational configurations of Uyghur identity in the cities, and went a considerable distance in creating a broad-based Uyghur national identity that might straddle both urban and rural communities.

# *The Uyghur 'Collaborator'*42

Three songs on *Qaldi Iz* dealt with the thorny subject of Uyghur indigenous officials, widely considered 'collaborators' in the politically charged atmosphere of the mid-nineties.43 'Tughmas Tokhu'[Barren Chickens] conveyed

even this problem to a widespread sense of political impotence and to the absence of equal opportunities under the Han regime.

<sup>41</sup> This helpful observation was made by James Millward, following my presentation of a preliminary paper on *Qaldi Iz* to the 2003 Annual Meeting of the Association for Asian Studies in New York.

<sup>42</sup> I employ the term 'collaborator' in this chapter, although the closest equivalent to the original Uyghur term *maqulchi* ['person who says "OK"'] would be 'Yes-man'.

<sup>43</sup> For an in-depth treatment of *Qaldi Iz* as a complete work and its role in reproducing the Uyghur nation, see Smith (2007).

#### illuminists: popular song & the waking of the uyghur nation 197

symbolic meaning on multiple levels, constituting at once an indictment of Han colonisation of the land, a blackly humorous critique of enforced birth control policy, and a dig at ineffectual and self-serving Uyghur cadres. The image of the parasite chicken—which devours the grain and gives nothing back—was evoked again in 2001 as Colin Mackerras observed that an improved economy in Xinjiang had heightened local resentment since Uyghurs 'may have considered that the Han Chinese were battening on what should have been their land' (2001: 301). 'Bäzilär' [Some People] articulated popular Uyghur perceptions of the weakness and passivity of indigenous officials (and would-be officials), characterising them as 'addicted to power', 'without conscience' and as 'stray dogs licking at the dog bowl'. The dog metaphor is traceable to the Qing colonial period, when indigenous officials were dubbed 'dogs with human faces' (Kim 2004: 34). It surfaced once more in the early seventies, when minority graduates of the Central Institute for Nationalities in Beijing were required to return to their communities and spread the word of Mao. Despite the attempt of the centre to 'mask the issue of assimilation' by training cadres in the traditional folkways, the latter were inevitably viewed as 'agents of Han domination' and as 'running dogs' of the Hans (Pye 1975: 506-7).44 In the mid-1990s, the metaphor again appeared in the modified form of the 'stray dog', evoking one of Alim's older themes: the displacement of Uyghurs from their homes and their ejection into the hostile desert. Although dogs in the wild exist in hierarchically ordered packs, living and hunting together as a team, the stray dog, once separated from the pack, becomes an opportunist who takes his chances. The domestic dog, meanwhile, may be characterised as parasitic, betraying his own kind to accept an easy life under a dominant race. Interestingly, the dog metaphor has also been employed by Uzbek artists. In the song he wrote about the Andijon massacre of 2005, political singer Dadaxon Hasanov depicts a brutal scene of civilians 'torn apart like ribbons' and 'shot down on the command of dogs' (Kendzior 2007: 324); meanwhile, in a poem about the same event, Uzbek poet Haydarali Komilov demands of the Uzbek military forces: 'Do you have parents or are you a dog? You have bitten your own people, you wretches!' (Kendzior 2007: 327).

<sup>44</sup> The Chinese term *zougou* ['running dog' or stooge] was earlier used to describe Chinese nationals who collaborated with Western or Japanese imperialists. *Zougou* translates into Uyghur as *yalaqchi* ['one who licks'].

#### 198 chapter four

In 'Äwliya Dostum' [My Immortal Friend], Alim explored the phenomenon of 'social climbers', that is, those perceived to collaborate with the Chinese state in order to advance their social status:


The first verse of 'My Immortal Friend' expresses the visceral inability (of a Uyghur loyal to his ethnic group) to kowtow to 'social climbers' for favours ('my heart burns'; 'my tongue burns'), and brims with scorn. The second verse condemns the lack of integrity of low and middle-level indigenous officials. We hear how they parade their superior status and act as though invincible before the 'chick' (the powerless Uyghur people), and yet become servile before the 'lion' (the Chinese state or higher-ranking Uyghur leaders). If a 'fly' (an ordinary Uyghur of average status) seeks their help, they 'whine' and make excuses; yet if struck by a 'stick' (variously interpreted by respondents as Han or higher-ranking Uyghur officials, or as a thick wad of cash), they are silent and complicit. In other words, such officials do favours only for those from whom they stand to gain. One respondent explained the ironic use of the word 'immortal' in this song with

#### illuminists: popular song & the waking of the uyghur nation 199

reference to the saying '*Asman egiz, yär qattiq*' [The sky is high, the ground is hard]. In other words, those who over-reach themselves have further to fall. This notion, too, can be traced back to the early seventies, when the humble origins of minority cadres trained at the Central Institute for Nationalities (then chosen for their 'revolutionary class backgrounds') made them 'appear in the eyes of just about everyone as suspiciously opportunistic individuals […] rude social climbers' (Pye 1975: 505-6).

All three songs ultimately evoke the social ideal of the 'man of the people,' summed up in the Uyghur proverb: '*Yakhshi är älgä ortaq, yaman är malgha* [A good man is in touch with the people; a bad man only with property].45 In addition to voicing keenly felt popular resentment, these songs appear to have been intended to prick the consciences of the officials themselves. Again, a parallel may be drawn with Albanian nationalism, with Sugarman hypothesising that the patriotic slogan 'Albania hates a traitor' was probably added to folk songs included in nationalist intellectual Mitko's 1878 collection long after the songs' creation, with the purpose of awakening and uniting the Albanian people (1999: 425). The notion of 'selling out' the nation has long been a central concept in Uyghur social and political life, where a large number of words are deployed to express national betrayal, for instance, *milliy kha'in* [ethnic or national traitor]; *milliy munapiq* [scum of the nation]; *wätän satquch* or *satqin* [person who 'sells' the nation]; and *maqulchi* [collaborator or 'Yes-man'].

Looking back at the period of Manchu rule, it is apparent that the Qing court was able to successfully co-opt existing local officials and bend them to its own agenda. Warikoo (1985) writes extensively on this topic, noting that native *begs* and Islamic clergy inherited their positions, frequently surviving changes in central government, and collaborated with one another in oppression of the common people. According to him, the Qing furnished native *beg* officials with government land grants and even allocated them bondsmen to cultivate the land. For their part, *begs* often exacted forced labour from the peasants, working alongside alien Chinese officials to line their own pockets, while ignoring the interests of the population. In this way, they became 'handy instruments of extortion' for the benefit of Chinese officials, who meantime appeared blameless. Their image cannot have been much improved in local eyes when they were

<sup>45</sup> According to Ämät's annotation, the proverb implies that a man who thinks of his people and native place is pure while one who thinks only of the road to riches is selfish (2001: 399).

#### 200 chapter four

required after 1877 to wear official Chinese dress and sport pigtails (Warikoo 1985: 81; 96; 105-7).

In the early years of the PRC, the Chinese Communists often found themselves similarly allied with local minority elites and feudal landlords in Xinjiang and Tibet as they worked to convince local peasants that support for the Party was in their interests (Grunfeld 1985: 62). Once the new China had been consolidated, the Chinese government commenced to handpick trainee minority cadres according to the appropriateness of their political attitudes. Suitable candidates were sent to the Central Institute for Nationalities (now renamed the Central University for Nationalities) in Beijing, where they normally completed five years of political education before returning to their home regions to take up post. In this way, non-Chinese officials in Xinjiang are appointed only after having been indoctrinated and proven loyal to the regime. They are to all intents and purposes 'powerless' and do not necessarily represent the views of local people (Shichor 1994: 76). Those who do voice local dissatisfactions are swiftly dismissed from office. Colin Mackerras once observed that 'no Uyghur could rise to real power who espoused local nationalism' (1985: 77), and this remains the case today.

In the mid-nineties, amid an atmosphere of fast-growing desire for selfdetermination, the social (community) consequences of this 'powerlessness' sharpened. As later summed up by Nicolas Becquelin, the selective introduction of market mechanisms in the local economy (designed to encourage Han migration) institutionalised ethnic discrimination, and the erosion of minorities' preferential rights in education and state sector employment combined to leave indigenous officials—the 'mediating echelon'—unable to counter the local sense of being colonised (2004a: 373). With Han people far better placed than minorities to take advantage of local business opportunities, minority leaders and their deputies who 'stood silent' rapidly lost credibility (Sautman 1998: 99). When I conducted fieldwork in Xinjiang in 1995-6, the Uyghur general public had become extremely disillusioned with its indigenous representatives. Räwiä was one urban dweller who despaired that Uyghur leaders did not stand up to central government, declaring in 1995: 'Our leaders say nothing. I don't know if they don't care or if they are just too scared. *I* would speak up about things if I were a cadre' [respondent's emphasis]. On another occasion, she suggested that Uyghur leaders were 'in no danger of execution', and demanded to know why they supported adverse policies proposed by Han officials. Similar views were recorded among rural Uyghurs, although these were

# illuminists: popular song & the waking of the uyghur nation 201

generally aimed at local, low-level officials. During my stay in rural Aqsu in 1996, Tursun, a peasant in his thirties, remarked ironically that much of the tax he paid made up wages for Uyghur cadres and labelled them 'parasites'. A wealth of information on the rural situation at this time can be found in Bellér-Hann and Hann's study (1999) of subsistence and subversion in southern Xinjiang. They found that large-brigade (*dadui*) and small-brigade (*xiaodui*) village officials, who are always Uyghur, were paid from taxes collected from local peasants, while township officials (*zibyoluq*) did not engage in subsistence activities, but rather lived in state-owned accommodation, ate in restaurants, and even dressed differently from the mass of peasants (Bellér-Hann and Hann 1999: 13). Local peasants claimed that opportunities for corruption among township officials had increased across the decade, and frequently 'turned Marxist rhetoric about living off the surplus labour of others against the very people whose main qualification was to be well versed in these communist doctrines' (Bellér-Hann and Hann 1999: 23). Particularly during times of ethnic unrest, these individuals found themselves caught between two poles of mistrust:

The position of these [indigenous] officials became more precarious whenever the political situation became tense […] there was the ever-present possibility that an official could be accused [by the state] of contributing to political resistance, 'splittism'. This factor, combined with the unpopularity of key policies and the unpleasantness of having to enforce them […] meant that people did not actively seek these jobs (Bellér-Hann and Hann 1999:15-16).

During the 1990s, Uyghurs frequently criticized in-group individuals thought to co-operate too closely with Han authorities. Starting from the summer of 1993, some of these criticisms developed into direct action. A small number of militant Uyghur nationalists began to make assassination attempts on Uyghur cadres, police, soldiers and religious personnel, whereas previously they had targeted only Han cadres, public security officials and military.46 According to Dilshat, a respondent who worked in the tourism industry in Qäshqär in 1996, a violent attack on a 'much hated' imam from the Heytgah mosque that took place in 1996 (one of his ears was cut off and his arms slashed with meat cleavers) was intended as a

<sup>46</sup> Compare shifting boundaries in South Africa with regard to 'internal' and 'external' Others. In the 1970s, Inkatha encouraged their members to defend the Zulu kingdom against 'outsiders' (at that time the government, whites, Indians, and Xhosas). Yet in the 1980s, Zulus opposed to the Zulu kingdom were targeted as 'the enemy within' (Morris Szeftel, African Studies Seminar, Department of Politics, University of Leeds, 7/12/94).

#### 202 chapter four

warning to others not to 'listen to' (collude with) the Han hegemony.47 Karim, a health professional from southern Khotän, supported this interpretation, stating in 1996: 'That imam's links with the authorities were rather too close'. Violence did not always stop with a warning either, so that in 1995 a Hong Kong newspaper reported (slightly sensationally) that the Uyghur separatist movement was 'an organisation like the "Palestinian National Liberation Front", often conducting activities such as assassination with their spearhead pointed chiefly at local government officials, especially those in the public security departments'.48 Certainly, the number of assassinations of indigenous cadres rose significantly during the years 1995-1997 (the same was true for attacks on Han police and soldiers). In one particularly horrific incident in 1996, not reported in the Chinese press, a Uyghur cadre and three of his relatives were attacked and killed in his home in Kucha. Again, the nature of the attack was significant: assailants cut out victims' tongues before slashing their throats and then blowing themselves up with explosives (Beckley, 1997; Bellér-Hann 2000: 84). Incidents of this type provide the context for the closing lines of Alim's song 'My Immortal Friend'. While one reading of the parting shot 'Farewell my friend' suggests that Uyghurs must say goodbye to a bad character trait (the inclination to prostitute oneself for power and material advantage), a darker interpretation warns that it is Uyghur social climbers—the 'immortal friends'—who must bid farewell to their lives. As one respondent put it, the song's closing line then assumes the meaning: 'You die!' [*Ölisän!*]

Dru Gladney (1997: 288) suggested that in-group attacks and assassinations in the mid-1990s reflected long-standing 'divided loyalties' among the Uyghurs. While I agree that Uyghurs sub-divide themselves into many different groups, some of which may mistrust the others in certain situations, I would contend that the period prior to the 1997 Ghulja disturbances in fact constituted a time of unprecedented unity; a time when more Uyghurs than ever before shared a common desire to eject the colonial power. Seen in this light, the attacks reflect not intra-ethnic but inter-ethnic rivalry, with indigenous officials seen as quasi-Hans—willing representatives of the state—and traitors to the common ethno-national cause.

<sup>47</sup> See Smith (1999: Appendix I) for a catalogue of ethnic disturbances in Xinjiang between 1949 and 1997.

<sup>48</sup> *Lien ho pao*, Hong Kong, 27 February 1995 in Summary of World Broadcasts (Asia Pacific), 1 March 1995, FE/2240 G/4.

### illuminists: popular song & the waking of the uyghur nation 203

At the turn of the millennium, the outlook for popular trust in officialdom was bleak. In Yee's 2003 study, 52.4 per cent of Uyghur respondents felt that cadre-people relations in Xinjiang were only 'fair' or 'poor', although distrust of Han cadres (39.5 per cent) surpassed distrust of Uyghur cadres at 23.5 per cent. Given the high sensitivity of the questions, we can assume that the true figures were considerably higher. Yee also stated that those who adhered more closely to Islam tended to view indigenous cadres—who are also CCP members—as 'traitors to their fellow Uygurs' (Yee 2003: 449). Observing that cadres play a key role in maintaining harmonious ethnic relations between groups, he concluded that the lack of popular trust toward both Han and Uyghur cadres signalled a 'red light for ethnic relations' (Yee 2003: 441). Following the ethnic riots in Ürümchi in July 2009, the Uyghur-language website Uyghurlar.biz published a report which expressed 'deep sorrow for our Uighur brothers and sisters' and declared seven days of mourning for the dead. In the subsequent online discussion, one forum participant accused the Uyghur head of the regional government, Nur Bäkri, of taking the Chinese government's side in the unrest, comparing him to a 'sword in the hands of a butcher' (BBC Monitoring, 2009). Earlier that same year, Ilham Tokhti, Uyghur professor of economics at the Central University for Nationalities, had also levelled sharp criticism at Bäkri during the annual session of the National People's Congress after the latter spoke of 'a more fierce struggle against separatist unrest' in Xinjiang. Describing the leader as 'incompetent', Tokhti subsequently posted the following appeal on his personal blog, Uyghur Online:

My message to the Xinjiang government is, 'You should know that there is no peace without equal development between Han immigrants and native Uyghurs in Xinjiang. Similarly, there is no stability in the Uyghur region without freedom of speech.' My message to the central government is, 'Don't listen only to what the local government officials in Xinjiang say—listen to the people. Don't just make decisions based on government research—also look at independent research. This will be very helpful for protecting the unity of the nation, and the long-term prosperity of the country.'

In so saying, Tokhti drew two clear distinctions: firstly between the ideals of 'nationality equality' and 'freedom of expression' (as enshrined in the PRC Constitution) and local realities in Xinjiang; and secondly between what indigenous Uyghur officials *say* about the situation in Xinjiang and the truth. In a speech delivered on the first day of the riots, 5th July 2009, Tokhti's blog was targeted by Bäkri, who labelled its author an 'instigator of the clashes' along with exiled Uyghur leader, Räbiyä Qadir. Tokhti's final

#### 204 chapter four

entry before his disappearance (temporary house arrest) was published through a US server on the morning of 7th July and read simply:

As the editor of Uyghur Online, I want only to tell Nur Bäkri, 'You are right, everything you say is right, because you will decide everything. I have already offended too many powerful people, including yourself and others whom I don't want to and don't dare to offend. But right or wrong, there will be justice (Radio Free Asia, 8th July 2009).

Despite the local perceptions of indigenous leaders discussed above, it is of course impossible to categorise all Uyghur officials in black and white terms as either a 'man of the people' or a 'collaborator'. Interviews suggested rather that individuals occupy different points on a scale between altruism and self-interest, and that these positions are far from static.49 An excellent example of this concerns a young male respondent in Ürümchi who joined the state police force between my first and second field trips in 1996 and 2002. He explained in 2002 that, in the wake of the Ghulja disturbances and the crackdown that followed, Uyghurs faced a 'fork in the road', at least until such time as China's domestic situation radically changed. They could choose either to accommodate to the state, with its associated risks of linguistic and cultural assimilation; or to continue to resist and ultimately face exclusion. While native Uyghur police are routinely mistrusted and even labelled 'dogs' by other Uyghurs, my interview with this individual suggested that his position and his motivations were complex. In particular, his descriptions of his everyday practices at work involved preferential treatment of Uyghur prisoners wherever possible and suggested his retention of a high level of ethnic loyalty.

# *The Bad Character Trait*

Two songs on *Qaldi Iz* dealt with the flawed national character, as perceived by group members: the tendency of Uyghurs towards political indifference and mutual betrayal. They were 'Ränjimäyli Özgidin'[Let's Not Blame Others] and 'Häsrät' [Sadness], the second of which I reproduce below:


<sup>49</sup> See also Stan Toops' illuminating comparison of Uyghur leaders Tömür Dawamat and Ismail Amat, in which the former is characterized as a 'cipher' for state-generated rhetoric, while the latter is shown to pursue a viable political agenda that reflects the realities of the region (1992).

#### illuminists: popular song & the waking of the uyghur nation 205


In lamenting Uyghur national disunity in this way, Ömärjan Alim may be said (intentionally or not) to have assumed the role of 'illuminist' in the 1990s, providing a timely incentive for national enlightenment.50 Respondents declared 'Häsrät' to be one of Alim's most emotive songs, its sadness deriving from a preoccupation with the Uyghurs' failure to unite (tendency to walk multiple 'narrow lanes' rather than one 'wide road') and the group's penchant for infighting. Two university graduates living in Ürümchi in 1996, Aliyä and Äziz, framed this problem in relation to another regional minority, explaining: 'The Kazakhs are a much better people than us. They are not jealous of one another. They fight big wars, we fight small ones'. Compare this idea with lines 7-8 of 'Häsrät', where we find another modification of the dog metaphor. This time, petty jealousies cause the dogs to fight over the spoils ('snap and gnaw at each other's success'). The two also quoted the Uyghur saying 'Paltining sepi yaghach tur' [The axehandle is ever made of wood]. According to this image (which is somewhat reminiscent of the 'sword in the hands of a butcher' metaphor mentioned in the previous section), the axe blade symbolises the Chinese state as it oppresses the Uyghur people (the wood) with the help of Uyghur 'traitors' (the axe-handle). The perception of disunity was also reflected in the claims of some respondents that some Uyghurs worked for the state as political spies. Dilshat, who worked in tourism in Qäshqär, asserted in 1996 that certain local people were supplying information to the Chinese authorities 'in the hope of bettering their personal circumstances', creating fear and mistrust within the Uyghur community.

As discussed above, despite the fact that nearly every song on *Qaldi Iz* [Traces] was framed in the context of escalating Uyghur-Han estrangement

<sup>50</sup> I borrow this term from the Albanian *iluminist*, connoting a poet who gives his life to spreading a message of self-elevation to his fellow countrymen (Sugarman, 1999: 447).

#### 206 chapter four

and a call to Uyghurs to mobilise (respondents agreed that the collection included 'lots of dangerous songs'), the album was not banned. One final and rather sinister explanation (one that I speculate on elsewhere, Smith 2007) might be that the state *was* aware of the song content, but allowed the cassette to continue circulating in the hope that its focus on political passivity, self-interest and opportunism might paradoxically work to foster negative self-identity and low self-esteem among Uyghurs. After all, state officials had by that time learned that banning any but the most inflammatory texts and audio-visual media could work to increase their appeal and their popularity, turning the cassette's creator into a political martyr.

# 4.5. *Abdurehim Heyit: The Intellectuals' Choice*

Abdurehim Heyit was born in Qäshqär in southern Xinjiang, where he began to study the *dutar* when he was six years old. At the age of seventeen he enrolled in a state-run music school, where he was encouraged to sing in an Italian operatic style, but found the training 'foreign' and far removed from the authentic Uyghur sounds of his childhood. In 1980, the school sent him to join the central song-and-dance troupe for minorities in Beijing, where he remained for two decades. In an article in *Time* magazine, Heyit described his life to date as a 'constant artistic struggle' (Forney 2002).51 With his troupe tightly controlled by the State Propaganda Department, he was for many years prevented even from playing the instrument of his choice. When I met the artist in Qäshqär in August 1996, he was having difficulties releasing his latest album through Xinjiang-based producers, who are subject to greater scrutiny by the censor than their counterparts elsewhere, and was not permitted to tour. Two years earlier in 1994, he had managed to release his trilogy, *Mung-zar* [Sorrow], only by approaching a recording studio in Xi'an, a city located at a suitable distance to the east of the troubled region. The more political of his original compositions, for example, 'Stubborn Guest', a song about an old man's plea to a lodger who overstayed his welcome (note the resemblance to Alim's 'I Brought Home a Guest', discussed above), have been disseminated on privately manufac-

<sup>51</sup> This article has now been removed from the *Time* website and archived, but is reproduced both in English and Uyghur (Latin script) on the website of a Uyghur exile organisation based in Washington, the Uyghur American Association (UAA). http://forum. uyghuramerican.org/forum/showthread.php?27152-Mengguluk-Hesrette-Qalghan-Adem-TIME-2002-Yil (accessed 21st January 2012).

#### illuminists: popular song & the waking of the uyghur nation 207

tured cassettes, a process that involves paying extra money to the producers concerned in order to compensate their risk (Strauss 1999). Always careful to absolve himself of direct agency, Heyit told me in 1996, smiling: 'They don't like my song words. *They say* I'm a nationalist!' [my emphasis]. As a professional musician, attached to a song-and-dance troupe, Heyit is theoretically required to appear in official, state-controlled variety performances on a regular basis. Yet he told me that he often declined to appear in these televised performances, remarking: 'I don't want to play the songs the producers ask me to play... and they won't let me play the songs I want to play!' In 2000, Heyit transferred back to the regional song-and-dance troupe in Ürümchi.

Despite the political constraints, Heyit's formal training has enabled him to draw parallels between Uyghur and other musical traditions around the world. His complex compositions for *dutar* are influenced by Persian, Arabic and Turkish styles, 'soaring and dipping along the Arabic scale[s]' (Strauss 1999). The Arabic scales are not even-tempered; the 5th notes are tuned based on the 3rd harmonic, and the tuning of the remaining notes depends on the *maqam* being played.52 This means that the same note by name may have a slightly different pitch in different pieces of music, depending on the melodic flow. The use of quarter tones is common, although these rarely fall exactly halfway between two semitones. My first encounter with Abdurehim Heyit in 1996 occurred in the Emin musical instrument store, located down a dusty alley in the Old Town of Qäshqär. As I pondered which *dutar* to buy, the singer happened to stroll past, and was immediately hailed by the store's owner. In the eye-opening half hour that followed, I watched Heyit test the *dutar* I had selected (for its modest, unadorned appearance), and glue in an additional fret to create that all-important quarter tone. The rest of the afternoon was spent in a Uyghur restaurant with Heyit and a group of his friends, where we were honoured with a private performance. When I met up with him again in 2002, he spent some time explaining the similarities between traditional Uyghur melodies and certain famous Turkish and Japanese melodies. He also illustrated how his songs draw heavily on the Twelve Muqam (*On ikki muqam*), a set of musical suites found in varying forms across Central Asia and the Arab world

<sup>52</sup> In Arabic music, a *maqam* is a set of notes with traditions that define relationships between them, their habitual patterns, and their melodic development. Each *maqam* has a different character which conveys a mood. Since classical Arabic music is mostly melodic (that is, it excludes harmony), the choice of *maqam* greatly affects the mood of the piece. http://www.amukhtar.com/en/a/a3.html (accessed 21 Jan 2012).

(see Harris 2007). Asked whether there were any similarities between Uyghur and (Han) Chinese melodies, his implicitly political response was: 'Some cultures are more alike than others'.

For Heyit, music constitutes one of the most important ways to transmit Uyghur culture and tradition across generations, and is an art form that enjoys enormous staying power: '[Young] people here can wear American T-shirts, but their hearts stay the same. The music, because it's such a symbol of the heart, changes slowly' (Strauss 1999). In 1999, he completed the first book ever written on classical Uyghur music. In interview in 2002, he stressed the importance of 'high culture' and was dismissive of some aspects of Uyghur popular culture; for example, when asked his opinion on the 'new flamenco' fusion popularised by Arken Abdulla (see Smith Finley 2011b), he replied that while fusion in itself is an acceptable art form, 'all musicians need a good grounding in classical and/or traditional music'. He later admitted that he could not tolerate electrified music at all, an attitude reminiscent of the uproar created within the folk music community when American folk singer Bob Dylan switched to electric guitar in 1965 (Yaffe 2011).53 This attitude has been observed also among Uyghurs based in Kazakhstan, and reflects a defensive response to cultural threat, whereby 'any kind of musical innovation is heard as deviation from authenticity and evidence of the encroaching influence of the other culture' (Harris 2005: 642).

Abdurehim's formal musical training was frequently acknowledged by his audience. Jelil, an observant university graduate interviewed in 2002, described him as a 'first-class singer and musician'. Tahirjan, a migrant worker in his thirties from Aqsu, declared him in 2004 the best *dutar* player in Xinjiang, remarking that his fingers 'move like lightning' on the strings. Both he and Zunun, who worked in the service industry in Ürümchi, praised the artist's fine singing voice, describing it as resonant, sonorous and clear, and observing that he 'sings from deep inside and not from the throat.'

Those of his fans who observed orthodox Islamic practice drew attention to his 'high moral standing', that is, the fact that he practises *bäsh waq* (prayer five times daily) and abstains from alcohol and cigarettes. Abduke-

<sup>53</sup> It would be interesting to know what sort of relationship Heyit has with his nephew, Marat; on the day this individual led a reporter for *New York Times* to Heyit's parents' house in Qäshqär, he wore a T-shirt emblazoned with the logo of the Western heavy-metal band, Kreator (Strauss 1999).

#### illuminists: popular song & the waking of the uyghur nation 209

rim and his friend, both university students who grew up in rural Aqsu, asserted in 2002: 'Heyit is a good man! He writes good songs, and he is a good Muslim!' Heyit's religiosity, while reassuring to his fan base, was however threatening to the authorities. According to Heyit's most ardent fans in Qäshqär, those songs with religious (Islamic) overtones could be sung only in his living room; the state censor prevented their inclusion on cassette recordings, and he was forbidden to perform them in public. In 2000, he was photographed for the sleeve of *On Meeting* [*Uchrashqanda*] wearing a traditional Uyghur robe and boots, and sporting a full Islamic-style beard. Subsequently, Party leaders at his work unit (Ürümchi's song-and-dance troupe) made him remove this because it was 'too religious-looking'. Two years later, a slogan chalked onto the blackboard outside his office read: 'Resolutely re-educate in the battle against splittism' (Forney 2002).

Respondents highlighted the 'subtlety' of Heyit's work. Jelil, an observant graduate, compared him favourably against Ömärjan Alim, asserting in 2002: 'Abdurehim's lyrics are less direct, more nuanced, better phrased'. This subtle lyrical approach, however, may have contributed to the failure of Heyit's audience to grow beyond comparatively narrow parameters. Möminjan, a young man from Turpan in his twenties and studying in Beijing, appreciated Heyit's work very much, explaining in 2002 that his songs were 'extremely deep' [*nahayiti chongqur*]. He noted however that many people could not understand his song words, and his songs therefore may have held less appeal. In his opinion, it was necessary to have quite a high level of education or intelligence (*säwiyä keräk*]) in order to grasp the underlying meanings. Whenever I met educated young people who admired Heyit, they generally qualified their comments in this fashion. This view was also reiterated in interviews I conducted with music stallholders in Döngköwrük, Ürümchi's Uyghur district, during August 2004. Respondents included both men and women, and their ages ranged from early twenties to late forties. Among them, several suggested that older people [*chonglar*] like Heyit's music the most because they appreciate his voice, but felt that he was not especially admired by the youth, who tended to listen to popular artists like Änwär Hakim, Abdullah Abdurehim (Xinjiang's so-called *nakhsha cholpan* or 'song star'), Möminjan or Tashmämmät Batur.

I also observed that fewer respondents were aware of Heyit compared with Alim, who seemed to be more of a household name. One exception to this was perhaps Heyit's hometown of Qäshqär, where he was well known to everyone in the community. There, songs like 'Stubborn Guest' and 'Silk' (described as a paean to Uyghur culture) were 'anthems' and

#### 210 chapter four

sources of local pride, and his latest cassettes were put on prominent display at the famous Sunday market (Strauss 1999). His celebrity was not this pronounced in other oasis towns, that is, it had not spread so far beyond intellectual circles. Zunun, a *minkaomin* in his thirties working in the service industry in Ürümchi, confessed in 2002 that he had not heard of Heyit before I brought the artist to his attention in 1996. On the other hand, he was familiar with Alim's songs, and claimed even to have recognised him in person in the street when the singer came to Ürümchi to give a concert in 1995. Anarkhan, a *minkaohan* in her late thirties working in the service industry in Ürümchi in 2002, also seemed unfamiliar with Heyit's work, although I gave her the background to his career. Nor had she heard of Osmanjan Sawut, considered 'the greatest poet since [Abdurehim] Ötkür' by Heyit, and a close friend of the artist. In the case of Sawut, the main barrier for Anarkhan was probably script: Sawut's poems were printed in the Uyghur Old Script (a modified form of Arabic), whereas Anarkhan had been educated in the New (Latin) Script during the Cultural Revolution. Those lesser educated respondents who had heard of Heyit were often familiar only with one or two of his most famous songs, for instance, 'Chillang Khorizim' [Crow My Rooster!], and could not reproduce Heyit's song words in full (though they could recite Ömärjan Alim's or Arken Abdulla's lyrics with ease). In a reflection of the comparatively incomplete dissemination of Heyit's work, Zunun remarked in 2002: 'Abdurehim Heyit is a great man, but not in the same way as Abdurehim Ötkür. The people who know Ötkür are many, but the people who know Heyit are few.'

All this said, Heyit was evidently much admired in intellectual circles in Xinjiang. When I visited him at home in 2002, he showed me a *ghäzäl* in classical Uyghur,54 published in the scholarly journal *Xinjiang wenhua* [Xinjiang Culture]. The *ghäzäl*, entitled 'Abdurehim Heyitqa Bir Ghäzäl' [A *Ghäzäl* for Abdurehim Heyit], compared him to the great Uyghur poets and artists through history, underlining his role as a transmitter of culture across time, and stating that he had 'touched the hearts of the people'. During that visit, several of Heyit's friends and one relative came to pay

<sup>54</sup> The *ghazal*, also spelled *ghazel*, *gasal*, or *gazel*, is a genre of lyric poem in Islamic literature. The ghazal developed in Arabia in the late 7th century from the *nasib,* which itself was the often amorous prelude to the *qasida* (ode). These enormously popular works were imitated not only in Arabic but also in Persian, Turkish, and Urdu poetry up until the 18th century. http://www.britannica.com/EBchecked/topic/232525/ghazal (accessed 21 Jan 2012). Among the Uyghur, this form of classic love poetry acquired the more general meaning of love song, and was reproduced as both a written and an oral genre (Bellér-Hann 2000: 40).

#### illuminists: popular song & the waking of the uyghur nation 211

their respects, closely followed by one of Heyit's students and the poet Osmanjan Sawut. Two friends read the *ghäzäl* aloud with great reverence, before presenting the musician with the only existing double VCD of a concert he had given in Qäshqär the year before. Heyit subsequently treated us to a spine-chilling, impromptu performance on *dutar* in his front room. He first played *Ömüt* [Hope], the title track of his then current release, a powerful instrumental whose progression seemed to me to represent the feverish build-up to the Ghulja disturbances in 1997 and the gradual anti-climax and simmering down that followed them. Next, he announced that he would play 'Chillang Khorizim' [Crow My Rooster!], from the triple album *Mungzar* [Regret] (1994). The room was abuzz with excitement at this, and the dark and seering song with its rousing refrain of '*Oyghansun!—*Let her [the Uyghur nation] awaken!—was followed by loud applause and a grateful chorus of thanks.

Furthermore, it was Abdurehim Heyit—not Ömärjan Alim—who caught the eye of the international press. Prior to the *Time* article published in 2002, Heyit had already been profiled as 'a master of folk tradition with a fire for protest and poetry' and as 'one of the most important contemporary Uighur composers and musicians' by the *New York Times* (Strauss 1999).55 The article had been brought to the attention of the artist, who kept a copy, preserved in plastic, in the original language. Most recently, in 2007, Abdurehim Heyit and Uyghur ghijäk master, Äkräm Ömär, were involved in a transnational collaboration with the Mike Del Ferro trio through the Jazz Bridges Western China project, an initiative of the American Voices collective which produced a joint performance and live recording of Duke Ellington's 'Caravan'.56

Partly because of the huge popularity of 'Chillang Khorizim', and partly because of his strong and sonorous voice, Heyit is sometimes dubbed the 'Rooster of Xinjiang'. The song is evidently a personal favourite of the artist, who became very excited when he learned that this song had been given special mention in the online version of *Time* magazine (Forney 2002). The

<sup>55</sup> Reprinted as Neil Strauss, 'A Folk Star in China Blends Song and Protest*,'International Herald Tribune*, 2nd September.

<sup>56</sup> American Voices describes itself as 'engaged in cultural diplomacy through jazz, hip hop, country, Broadway, classical and other musical programs with over 80 countries around the world'. Their aim is to bring American musicians together with local performers of traditional musics around the world, and make video recordings of the ensuing gala concerts available on the internet. Their projects have involved musicians in Western China, Afghanistan, Iraq, Kazakhstan and beyond. http://www.youtube.com/watch?v=I0akWqlxv2A&fe ature=related; http://www.youtube.com/user/americanvoices (accessed 21/1/2012).

Figure 11: Abdurehim Heyit, dutar virtuoso, Döngköwrük, Ürümchi (2002)

song words are adapted from a poem of the same name written by the great Uyghur poet and novelist, Abdurehim Ötkür, while the music is composed by Heyit:

#### *Chillang, khorizim! Crow, My Rooster!*

Nimä degän uzun kechä What a long night this is Häjäp tang atmay qaldi Daybreak never comes Chillang khorizim, chillang! Crow my rooster, crow! Güliräykhanim oyghansun! Let her awaken! Chillang khorizim Crow my rooster, Yarni chillap oyghiting! Wake my sweetheart!

Tanggha qarap közüm taldi My eyes are weary, yearning for the dawn Ukhlaq jahanim oyghansun! Let my sleeping world awaken! Meni tashlap ukhlighan My sleeping sweetheart who forsook me Yüzigä qanat yeyip Spread your wings, stroke her face

#### illuminists: popular song & the waking of the uyghur nation 213

Asta sipap oyghiting! And gently wake her up! Achsa yarim kirpigini When she opens her lashes Yänä uyqu basmisun Let her sleep no more Uyqu birlä män gheripni Let her not return to sleep Mängü tashlap kätmisun Forsaking lonely me for good Chillang khorizim, chillang! Crow my rooster, crow! Tang aldida yaringiz Crow to tell that Chillang khorizim, chillang! Crow my rooster, crow! Chillang, bu gülistangha Crow, to this flower garden Oyghansa yarim khäwär bering When my sweetheart awakes Män baghri qangha Tell suffering me Oyghansa yarim khäwär bering When my sweetheart awakes Män baghri qangha Tell suffering me

Tang atidu däp, chillang! Crow to tell that daybreak comes! Sizni kütidu däp, chillang! Your sweetheart awaits you before sunrise!

A signpost regarding the underlying message of the song is provided on the recorded cassette by a narrator, who introduces the song with this sentence in Uyghur: 'The morning rooster wakes the people from their ignorant slumber'. In adding this, a clear link is provided between the 'slumbering people' and the corresponding metaphor in the song: the 'sleeping sweetheart'. As a result, educated and less well educated respondents alike could understand immediately that the song concerned itself with the concept of national enlightenment—'waking up the Uyghur nation'. An equally clear link connected this song/poem to the work of an earlier, influential Uyghur nationalist from Turpan: Abdukhaliq Uyghur (1901-1933) (Rudelson 1997: 145-154). As a child, Abdukhaliq had travelled with his father throughout China, Russia and Finland. Observing the political systems operating in other nations, he later returned to Turpan, where he established the Turpan Revolutionary Central Committee and attempted to bring enlightenment to his people through the vehicle of poetry. If we look at Verse 2 of one of his best-known poems, *Oyghan* [Awaken!], the ideas are similar to those expressed in Ötkür's more recent poem/Heyit's song:


#### 214 chapter four

The second song I want to discuss is 'Bilim Ishqidä' [On Knowledge], and is collected in Volume 2 of the privately released cassette trilogy titled *Mungzar* [Regret] (1994). Interviewed in 2002, the artist singled the song out as one of the best he had released to date (Forney 2002). The lyrics are an abridged version of an original poem by Uyghur poet Nimshehit,57 written during the period of the Three Districts Revolution against Guomindang rule (November 1944):


<sup>57</sup> A collection of poems in Uyghur by Nimshehit entitled *Yüräk sözi sha'irlar* [Words from the Hearts of the Poets] was published in 1980 by the Millätlär näshriyati [Nationalities Publishing House] in Beijing. See entry online at the Open Library (a project of the non-profit Internet Archive, funded in part by the California State Library and the Kahle/ Austin Foundation): http://openlibrary.org/search?language=uig (accessed 21 Jan 2012). A collection of Nimshehit's literary works was apparently also published by the Shinjang khälq näshriyati [Xinjiang People's Press] in 1995, but no further details are available at time of writing.

<sup>58</sup> *Nomus* [shame].

illuminists: popular song & the waking of the uyghur nation 215


Once again, the song is introduced by the narrator, who sets the tone by observing: 'A boy who has studied is greater than his father'. The song should first be understood within the context in which the original poem was produced: during the height of the Jadidist movement to modernise Uyghur education, first imported by wealthy Uyghur industrialists and merchants who travelled in Central Asia, Turkey and Europe. The movement flourished in Xinjiang from the 1920s onwards during which time new teachers' colleges were established, new journals and newsletters published, and Turkic (and latterly Uyghur) nationalist ideas instilled in school pupils and readers. Many of those inspired by this 'Uyghur enlightenment' later became involved in the rebellions of the early 1930s (Millward 2004: 4).

Particularly since the 1997 Ghulja disturbances, the idea of empowerment through knowledge is again circulating in intellectual circles in Xinjiang. According to this ideology, if increasing numbers of Uyghurs can achieve a higher level of education, political awareness will grow, and Uyghurs across the region will develop a strong sense of ethnic identity and nation. As Jelil, an observant university graduate, explained in 2002:

The most important thing of all is knowledge. This is the foundation of everything … our people need knowledge, education. Over time, the number of educated will grow and grow, until every Uyghur is educated! Through this transitional period, the educated ones will ensure that their own children are educated, and bring their influence to bear on uneducated people of the same generation.

This generational model for political mobilisation evokes an earlier model that had been touted during the mid-nineties: the belief (hope?) that the passing of the cautious older generations and the coming of age of the post-1980 youth would result in a united resistance against the Chinese state (cf. Smith 2000). The more recent model incorporates two new ingredients, in recognition that the passing of time alone will be insufficient: the first is knowledge; the second is peer influence or 'community pressure' (see Chapter 6 on selective endogamy for an illustration of this powerful mechanism). As Jelil pointed out in 2002, according to this way of thinking, one should not blame those Uyghurs who become informants or 'sell the nation' in order to save their own skins, for they are without knowledge [*bilimsiz*], unenlightened, and do not know any better: 'They do not see

#### 216 chapter four

things in the long term. We must forgive them and concentrate on educating them, our people.' This elitist attitude suggested that uneducated individuals were less likely to have a strong sense of ethnic identity or loyalty; yet this was not supported by my interview data, which showed that people from all walks of life could receive and appreciate the political messages conveyed in Ömärjan Alim's songs.

'On Knowledge' exhorts the listener to make knowledge his constant companion, warning against less enduring ('fickle') attractions ('loves'), such as women and material wealth. It urges the Uyghur people to seek knowledge when they are in need of comfort, rather than the alternatives. Here, one might read an undertone that pleads with young people not to seek solace in alcohol and drugs (see Dautcher 2004; 2009 on social pathologies) but to build their self-esteem ('safeguard your worth') through education instead. Male listeners are warned not to become distracted by women ('beauties resembling the full moon') who are attentive only in pursuit of personal gain ('when your wealth declines, they'll desert you without shame'). There are puritanical undertones here, deriving from Islamic morality, and possibly also a criticism of the artist Ömärjan Alim, who was at the centre of a scandal involving a nightclub in Ghulja during this period. The final verse underlines the transience of material wealth in comparison with knowledge ('you are rich today, but tomorrow none will have greater debts than you'). With reference to Islamic fatalism, it is emphasised that a person who prioritises his own physical or financial satisfaction over the well-being of his people will eventually see his luck run out (cf. earlier discussion of how 'social climbers' have further to fall).

The third song I will analyse is 'Män Aq Bayraq Ämäs' [I'm No White Flag], and is a very slight abridgement of another influential poem by Abdurehim Ötkür (the part in italics is omitted in the song version). Like the previous two songs, it was included in the trilogy, *Mung-zar* [Regret]. The first part of the song runs as follows:


<sup>59</sup> *Tianshan* is the colonial name in Chinese for the Heavenly Mountains dividing north and south Xinjiang. In the poem form, the original Uyghur language name is used: *Tängri-*

#### illuminists: popular song & the waking of the uyghur nation 217


Part One celebrates the magnificent beauty of the Tianshan (Heavenly Mountains) and its tallest peak, the King of Heaven, which may be read as a metaphor for the local Turkic people (if not necessarily for the Uyghur people alone), who, like this mountain, are anchored to the land. The worth of the Turkic people is represented by glittering silver and gleaming pearls; their proud demeanour by the mountain peak's upright stature. The rising and setting of the sun stresses continuity of the land and its people, with each blessed by the respect of the seven stars;61 while the description of the hardy snow lotus may be seen to reflect the determined presence and courage of the people. The peak represents by turn the free spirit of the

*tagh.* The artist has changed the name in the song form, possibly in order to disguise the underlying nationalist content and thus avoid the censor.

<sup>60</sup> *Hörliqa*: a legendary bird bringing freedom.

<sup>61</sup> The mystical number seven has played an important role in the myths and rituals of the Central Asian shamans. Here, it likely refers to the notion of traversing the seven celestial regions (the seven planetary heavens), which originated in Mesopotamia. The Altaic shaman is thought to mount and descend through the seven heavens, just as he passes successively through the seven underworld 'obstacles' (Eliade 1964: 274-279).

people (freebird), the flourishing of the people (flower in full bloom), and the tranquillity of the people (swan lake).

In Part Two, the melody changes direction and rhythm, becoming more agitated, and with the vocal frequently rising to the top of the octave. In this section a conversation ensues between the King of Heaven and a passing observer:


At the beginning of this conversation, the observer mistakes the tall, white peak for a white flag of surrender, raised by a retreating army. The King of Heaven's response is amused but also indignant: he retorts that if he is all white, that is because of his advanced years (an implication that he has been here for a very long time), and is certainly not a sign of surrender. He adds that his disposition is 'far from being a white flag', in other words not at all meek or vanquished. Meanwhile, the poet's choice of *beard*, *hair* and *face* appears to be deliberate: the Turkic people tend to grow beards, while the Han do not; Turkic people are more likely than Han to see their hair turn white; and Turkic people are at least partly Caucasoid, unlike the Han, hence the white face. The implication is that the people who have existed in this region for so long are Caucasoid, unlike the Han.

The next section of the poem uses many examples and illustrations to stress the presence of the mountain peak (Turkic people) in this region since ancient times as well as the part they played in the dawn of civilisation:

Adäm ata wä Hawa When Adam and Eve Jännättin chiqqan chaghda Left Paradise Tunji tughulghan oghli Their first born son Män idim bu zeminda On this land was me Boz echip uruq chachqan I tilled the fields Yeshil wada terimdä Sowed grain in the spring Nohning kemisi topanda And when Noah's Ark Yol tapalmay azghanda Lost its way in the Flood Aghzimda zeytun chishläp I was the pigeon Hayatliqning nishani I was the bearer of news Ashu khäwär män idim Bringing signs of life Qosh münggüzlük Iskändär When Alexander the Great Bashlap kälgändä khätär Brought danger62 Uning ötär yolini It was I Tosup qoyghan män idim Who blocked his way Shärqliqning küchini I who showed him Tonutup qoyghan män idim The strength of the East Yipäk yükligän karwan I who hosted Qonghuraq jaranlatqan The silk-laden caravans Ashu qädim yollarda Who rang the bells Häm qoruqchi pasiban On those ancient roads Häm bolghantim sahipkhan To the caravans Äy janabiy muhtiräm Hey, respected sir Kheti mändä pütülgän Were written about me Ni-ni chegish tügünlär Many mysteries Mening bilän yeshilgän Solved through me Dora degän möjizä Gunpowder—these miracles

Barghan käptär män idim Holding an olive branch in its beak Ash-tuz berip karwangha I who gave food and shelter Bilip qoyung shuni häm Take note of this, what's more Tarikh degän kitabning The words in the book of history Qäghäz, kompas, mätpä'ä, Paper, compass, printing press

<sup>62</sup> Alexander III of Macedon, better known as Alexander the Great, was born in Pella, the ancient capital of Macedonia in 356 BC. Educated by the philosopher Aristotle, Alexander inherited a powerful but volatile kingdom after his father, Philip, was assassinated in 336 BC. He quickly dealt with his enemies at home, reasserted Macedonian power within Greece, and then set out to conquer the Persian Empire. Against overwhelming odds, he led his army to victories across Asia Minor, Syria and Egypt without suffering a single defeat. His greatest victory was at the Battle of Gaugamela, in what is now northern Iraq, in 331 BC, following which he became 'great king' of Persia at the age of 25. Over the next eight years, Alexander led his army a further 11,000 miles, founding over 70 cities and creating an empire that stretched across three continents, from Greece in the west, north to the Danube, south into Egypt and as far to the east as the Indian Punjab. The area was linked together in a vast international network of trade and commerce, and united by a common Greek language and culture. http://www.bbc.co.uk/history/historic\_figures/alexander\_the\_great.shtml (accessed 22 Jan 2012)


Öz qolumda täwrätkän I rocked with my own hands Demäk ta älmisaqtin You see, since the dawn of history Shunga beshim aqarghan That's why my hair turned white

This version of history, while true for the mountain peak, is selective when it comes to the Turkic people. Archaeologists have concluded that 4000 year old mummies discovered in the Tarim basin are of mixed ancestry, having both European and Siberian genetic markers. Meanwhile, historians tell us that the people today known as the Uyghurs originated on the steppes of what is now the Mongolian Republic, migrating west to the region now called Xinjiang only in 845 AD (see for example Barfield 1989). This of course raises the question of relative—and contested—indigeneity, and throws doubt upon the claims made in the poem with regard to Adam and Eve, Noah, Alexander the Great, and so forth. Probably, the poet's intention is to omit the earlier history of the steppe Uyghurs in favour of emphasising the ancient roots of his ancestors within the territory in which they now reside. In doing so, he imaginatively reinterprets the past in the present, creating a subjective authenticity which legitimises the contemporary territorial claims of the Uyghur nation (cf. Hann 1991: 231). Yet it is certainly true that the Turkic people have been attached to the land *for far longer* than the Han, at least if we are talking about populations in large numbers. And it is also possible to argue that even if the steppe Uyghurs were not residing in Xinjiang before 845 AD, they subsequently intermarried and merged with the indigenous Indo-Europeans they found there to form one fused group. Also of note is the fact that the poet does not employ the term 'Uyghur' at all; this could be either because he hoped to avoid the censor, or because he rather espoused a Pan-Turkic ideology like so many other intellectuals born in the republican period.

The poem/song ends on a proud and triumphant note as follows:

Bu hör zemin wätinim This free land is my homeland Anga baghliq jan-tenim Body and soul to it tied fast Chach-saqalning aqlighi The whiteness of my hair and beard Tarikh qäsirining uli The foundations of history's mansion Kökkä taqashqan beshim My head stretched high into the sky

#### illuminists: popular song & the waking of the uyghur nation 221


Thus, we are assured that the Turkic people will not shy away from troubled times, but will doggedly continue to hold fast to the land; the 'tall mountain peak' has not surrendered, nor will it. As noted earlier, in the song version performed by Heyit, four lines are omitted from the original poem (highlighted in italics above). These appear to frame the poem firmly within the context of socialist revolutionary resistance against the Guomindang (Chinese Nationalists) in the 1940s, thus the mountain peak is said to have a 'heart red as embers' and to 'hold the red flag', representing the allegiance of the local peoples to the Chinese communist saviours. This was very likely a ploy by Ötkür to avoid the political censor in the present by setting his epic in the past: a strategy he used repeatedly and to great effect in his historical novels (cf. Rudelson 1997: 163-164). The act of removing the lines re-positions this poem within contemporary times, indicting the current Communist aggressor in place of the past Nationalist one.

# 4.6. *Religious and Class Cleavages*

As I hope I have demonstrated above, during the 1990s Xinjiang enjoyed two 'Voices of the Uyghurs', each of whom made an important contribution to raising political awareness and/or safeguarding the continuity of Uyghur tradition and culture. Yet as Breger and Hill observe: '… even within the ethnic community identities can be plural, cross-cutting, conflicting and contested, with different ideologies supporting them' (1998: 10). Given that the two artists appeared to be pursuing similar goals and that some though not all—of their songs addressed similar themes (the host-guest

#### 222 chapter four

dynamic, collaboration versus non-collaboration), one might easily have assumed that they supported one another's efforts in the realisation of a common goal. This however was not the case, as interviews with the artists and close relatives and friends quickly proved.

During several meetings with Abdurehim Heyit in Ürümchi in 2002, I learned that he and his entourage of friends in fact held rather negative opinions of Ömärjan Alim. When first asked his thoughts on the other 'Voice of the Uyghurs', Heyit did not describe him as a 'friend' but stated simply that he knew him. He agreed—a little reluctantly—that the tracks on Alim's album, *Qaldi Iz*, were 'good songs', but lowered his gaze and did not elaborate. Seeing this, I enquired playfully whether there was an element of competition between the two men. Heyit shook his head vehemently at this, an action I initially interpreted as a sign of political caution (a desire not to be connected with Alim, who had by then run into political difficulties). Later, in a conversation with two members of Heyit's circle, I realised that the issue was more complex. Asked what he thought of Alim's work, Jelil, an observant graduate, commented in 2002:

You only have to look at the level of artistic merit … you simply cannot discuss Alim and Heyit in the same breath. Abdurehim's music has many faces! It changes all the time. It moves on, develops …! Alim's music always sounds the same, it always begins from the same basic structure; it is simply variations on a theme.

Uyghur poet, Osmanjan Sawut, added to this (in Chinese, for my benefit): 'Alim's material is really coarse [*cu de hen*]! It's like a slap in the face, too direct. The poetry is simple, elementary; the meanings too obvious.' Hearing this, it occurred to me that Heyit may have felt partly insulted by the suggestion that he could be compared to Alim. Jelil then remarked, almost as an afterthought, that Alim's methods were a bit 'dangerous' and so Abdurehim would certainly not wish to be associated with him. However, this was evidently far less salient than the question of inferior artistry.

In a subsequent meeting in 2002, I suggested that Alim's lyrics were perhaps *deliberately* simplistic in order to reach a wider, rural audience. This theory was however quickly rejected. Heyit and Sawut's indignant answer was:

He doesn't even write his own words! Someone else writes them for him! Some second-class poets in Ghulja … Ghulja has never produced any brilliant poets. And Alim's *dutar* playing and his voice are totally inferior; they just don't make the grade!

#### illuminists: popular song & the waking of the uyghur nation 223

To this, Jelil added that Alim had not received a college education, and therefore has no knowledge (*bilimsiz*) and is unable to read newspapers and scholarly journals. Heyit, on the other hand, had gone to college to study the arts, and reads newspapers and journals every day. Thus, in his view, the two could not be placed in the same category. The above responses indicated a sense of social superiority, linked to occupation/profession, based on Alim's lack of formal musical training and poor education; it was a type of class snobbery found both in traditional Uyghur society and in traditional Chinese society. Furthermore, the generalisation concerning Ghulja poets betrayed a strong element of 'hometown chauvinism', to use a term coined by Freeman (2011). The initial charge levelled at Alim—that he does not write his own song lyrics—might just as well be applied to Heyit's songs, many of which draw heavily upon the poetry of Abdurehim Ötkür. It seems that it was not the borrowing of poetry in itself that was anathema, but rather the borrowing of poetry *from Ghulja*. It is important at this point to note that none of those present were natives of Ghulja, but rather of Qäshqär, Qumul and so forth. When I ventured to suggest that Alim's end goal—to enlighten Uyghurs and encourage ethnic unity—at least was a worthy one, the two agreed, but again qualified the concession by questioning Alim's artistry: 'Yes, there is some good lyrical content. What he's saying, what he's trying to do, is good. But the vehicle through which he's doing it is sadly lacking.'

From here, the discussion developed into an intense debate on high versus popular culture, with our host, Saqim, continually trying to intervene in order to offer us food (and diffuse the contention). Abdurehim, while acknowledging that a person could not truly know a people or nation without speaking to individuals from all walks of life (including kebab traders) insisted that only high culture can sustain that people's cultural continuity through time: 'Look at the Twelve Muqam, for instance. It is the artists, poets, historians, musicians who will go down in history, who will be remembered; and, with them, the Uyghur people.' Jelil agreed strongly with this statement, adding that while high culture can represent a people, popular culture has no longevity and is quickly forgotten. Having in the back of my mind the comments made by ordinary Uyghurs regarding the inaccessible nature of Heyit's song words and vocal style, I asked how high culture could sustain a people if the majority of that people could not understand it. The two friends replied: 'Our 'high culture' *can* reach the peasants these days … ten years ago, they could not have understood these song-words. But now they can.'

#### 224 chapter four

To what extent their brand of high culture truly made an impression on people outside of intellectual and artistic circles remains unclear. It is possible that some peasants might have possessed an educational level which allowed them to appreciate oblique meanings deeply embedded in metaphor, beautiful but impenetrable. We know for example that Uyghur teachers in rural villages and small towns were able to 'read between the lines' of local reportage published in literary magazines during the 1990s, considering this an important source of information about implementation and repercussions of new policies, cadre corruption, inequality and poverty (Bellér-Hann 2000: 83). It is perhaps more likely that Alim's lyrics reached the peasants first and established a precedent, heightening political awareness in the countryside, and encouraging ordinary folk to listen to music and poetry in a more analytical fashion. However, on testing this hypothesis against the experiences of respondents, it did seem that on the whole Heyit's work was less able to permeate people 'from all walks of life'. To cite some examples, Zunun, a *minkaomin* in his thirties working in the service industry in Ürümchi in 2002, expressed a clear preference for Alim's songs, and attributed the negative comments made by his competitor to class snobbery: 'I still like Alim's songs, no matter if Heyit and his friends say they are crude … all my relatives in Bay [rural Aqsu] know Alim and like his songs. Heyit's crowd *would* say negative things about him, because they look down on him!' Aliyä, a postgraduate student interviewed in 2002, was similarly displeased at the criticism levelled at her favourite artist, and urged me not to convey it to Alim himself. In her view, Abdurehim Heyit was simply jealous of Alim's stardom; she disagreed that the latter's work was 'crude', pointing out that although he does not write his own songwords, he nonetheless selects which poetry to set to music: 'He chooses lines that reflect what the people are feeling. That takes an intelligent mind, intellect.' She also stated that Alim played *dutar* very well, which suggests that many Uyghurs may in fact appreciate the home-grown sound of a self-taught instrument.

Another seam of criticism centred on Alim's perceived lack of morality. Saqim, a television director in his thirties who had previously produced Alim's music videos, claimed in 2002 that although Alim was 'great' in the early years, he later 'lost sight of his original aims' and 'began to walk a broken road [*buzuq yol*]'. According to him, the artist had begun to sleep around, despite being a married man with a family. Connected to this charge of philandering, Alim came under fire from several respondents for

# illuminists: popular song & the waking of the uyghur nation 225

not practising what he preached. Abdukerim, an observant university student from Aqsu, protested in 2002:

Alim's a hypocrite! You know how he did that song 'Aldanma Singlim' [Don't be Taken In, Little Sister]?63 Well, a while after, he opened a disco-restaurant! And Uyghur girls went there, and drank and smoked and took heroin! What do you think of that! And some people went to see him and said 'Ömärjan, why do you sing about protecting our children, but then open this disco?' They beat him up. And he said he was very sorry and a month later had to close it down.

If female visitors to this establishment had indeed taken heroin, this made the artist double the hypocrite in the public's eyes; one of his best loved songs exhorts young people not to fall prey to drug addiction. Asked about the disco rumours, a male restaurateur who had migrated to Ürümchi from Aqsu to find work, replied in 2002: 'Yes, they say that Uyghur girls go in there and go wild. I think it probably is true; the story wouldn't be circulating otherwise. Rumours don't appear out of nowhere, do they?' Yet other respondents reacted angrily to the suggestion, as when Aliyä, a postgraduate student, told me in the same year:

I don't believe at all what others said about Alim running around with other women. I watched him with his wife every day here at the *Yixueyuan* [Medical College] hospital […] and he was attentive from dawn to dusk. I watched him go back and forth from her bedside.

The close details of the scandal were hard to verify; however, it was clear that something of this type had occurred. Restaurateurs living in Alim's hometown of Ghulja stated in interview in 2002 that Alim had previously bought a *ristoran* (high-class dance restaurant—somewhat different from a disco or nightclub) in the town, but that he no longer owned it. The artist's wife, interviewed in 2002, corroborated this story, but was irritated by and scornful of the criticisms:

You heard he took all sorts of girls in his disco, right? That was before. Just for one month. Actually, it wasn't even his disco. One of his friends opened it and he sang there every night for a month. And then he stopped. That's

<sup>63 &#</sup>x27;Aldanma, Singlim' was a didactic song warning young Uyghur women not to mar their youth by frequenting discos and consorting with (often Han) men. It alludes to a small minority of young women in Ürümchi who are said to have adopted 'bad modern ways', where modernity is perceived to have been introduced by the Han and as running counter to Islamic social mores. Some girls, known in Chinese as *sanpei xiaojie*, have assumed a role comparable to that of the Japanese 'hostess', drinking, dancing, and, in rare cases, sleeping with Han businessmen for money.

#### 226 chapter four

it. It's all a pack of lies [*yalghan*]. There are always all kinds of stories flying around about famous people—singers, dancers … all rubbish.

When I asked whether she thought this behaviour might stem from jealousy, she responded with an emphatic affirmative in Uyghur: 'Hä---!' Thus, another irony was suggested: a section of Alim's audience, which had once identified closely with his songs condemning in-group envy, was now itself succumbing to the 'bad character trait'. Abdukerim, an observant university student, also questioned the artist's morality in 2002, citing an occasion when fans at a concert in Aqsu waited an eternity for Alim to come on stage, only to be told that the singer was 'ill' and would not be playing that night. In fact, Abdukerim claimed, the artist had been inebriated:

We could see him, backstage, swaying around, blind drunk! It was only a small place, you could see everything. The people were furious! 10 yuan a ticket is a lot, but I paid it because I wanted to see him so much. Everyone smashed the windows in Alim's car that night!

While this last claim was likely an embellishment, it is easy to imagine how this experience could create disillusionment, especially among fans who observed orthodox Islamic practice. The artist's wife confirmed in 2002 that Ömärjan did not practise *bäsh waq*, exclaiming: 'Goodness me, no! What, with him drinking and all? You can't drink and still pray! Oh yes, he drinks, smokes, has a good time [*oynaydu*]. You see, he never learned how to pray as a child. He wasn't taught.'

In 2002, in a changed environment where possibilities for combining music and politics had dwindled, Alim stood accused by some of 'selling out'. His double release *Häsät Qilma* [Don't Fall Prey to Envy, 2002] contained no explicit reference to the Uyghur/Han conflict, focusing instead on in-group social problems. As the artist's wife explained: 'The words reflect problems in society, people's everyday lives. For instance, this song [then playing on the family's video recorder] exhorts people not to waste food when the peasants are starving.' Asked whether the lyrics were aimed at the Han banqueting culture (*yanhui wenhua*), she and Alim's sister retorted in sarcastic tones: 'Hey, Uyghurs do it now too!' This indictment of in-group failings, while appreciated by some as a necessary stage of awareness-raising or vehicle for catharsis, was easily interpreted by others as a shameful expression of low self-esteem, an act of 'hanging out the Uyghurs' dirty washing'. One might speculate that, following the political trouble generated by earlier songs, Alim was dismissed from his work unit as a warning, then instructed to continue in a castrated form. In this way, the

#### illuminists: popular song & the waking of the uyghur nation 227

state was partly successful in orchestrating the hero's public fall and weakening his status and power as an ethnic martyr. A similar political strategy was employed by the Uzbek government in 2005-6, when it released rebel poet Haydarali Komilov from prison, ordering him to publicly retract his poem on the Andijon massacre, '*Nima qilib qo'yding oqpadarlar?!*' [What have you done, you wretches?!], and to discredit the broadcaster who had dared to air it. As Kendzior observes, this manoeuvre fully demonstrated how threatening the Uzbek government considered poetry to be (2007: 328).

Alim was also condemned for becoming too attached to money and the 'high life'.64 Saqim, a television producer, claimed in 2002 that the price of Alim's CDs had gradually risen until ordinary people could no longer afford them. He also asserted that the artist had latterly become very 'arrogant', and would do nothing for anyone unless they paid him.65 These accusations were ironic indeed, given Alim's earlier song about the relationship between indigenous cadres and 'thick wads of cash' ('Äwliya dostum', above). The public perception that Alim cared only about self-elevation was not lessened by remarks made by his wife, for example, her admission that Alim often drank alcohol when in the company of top cadres [*chong bashliqlar*]; and her (unlikely) claim that her husband 'had never been in trouble with the state'. Jelil, an observant university graduate interviewed in 2002, blamed Alim's ethical decline on his poor education: 'We can forgive him and all those who "sell out". After all, they are without knowledge [Uy. *bilimsiz*]'. His comment reflects the common belief that uneducated Uyghurs are less likely to maintain a strong sense of ethnic identity and loyalty (cf. the poem/song 'Bilim Ishqidä', discussed above), but also a new sympathy for the rural class and a desire to support and inform (rather than patronise) Uyghur peasants. Ironically, this is exactly why Alim's 'plain words' played such an important role in 1996.

Other respondents did not accept that Alim had sold out; rather, they characterised him as a political martyr, and circulated rumours of his arrest

<sup>64</sup> Another Ötkür poem adapted by Abdurehim Heyit and included in the privately released trilogy, *Mung-zar* [Regret], was called 'Khäjlä, kha'inlar, khäjlä!' [Spend, Traitors, Spend!]

<sup>65</sup> The family was apparently quite wealthy. Ömärjan's *loufang* home was located near the No.16 Primary School in Ghulja, and was lavishly furnished with shiny pouring vessels, leather sofas and intricate Islamic wall covings. When we discussed the family's desire that Alim's daughter go abroad to study, and I warned them of the high tuition fees required of international students, her mother replied confidently: 'Oh, we have money! [Uy. *Pul bar!*] Money for this girl's study, we have.'

#### 228 chapter four

and persecution. They claimed (with evident pride) that he had been discovered by the censor and consequently warned by the government to cease singing songs that 'threaten the unity of the nationalities'. Several reported that, despite being previously attached to the local song-anddance troupe, Alim was no longer attached to a work unit (a circumstance which may have indicated state punishment in kind). A group of restaurateurs from Ghulja, who ran a fast food outlet in Ürümchi in 2004, confirmed this, adding that Alim was now a 'businessman'. They rejected the idea that Alim had been arrested or imprisoned after the release of *Qaldi Iz* [Traces] in 1995, stating that he had incurred a stiff fine, but subsequently continued to release new compilations. They did not accept that the recent albums were more costly than earlier ones, or that the artist had in any sense sold out. Given the Uyghur proclivity to 'hometown chauvinism', it seems possible that fellow Ghuljiliqs might generally have been more sympathetic towards their oasis star, while criticisms tended to emanate from the oases of the religious south. Yet some of Alim's southern-born fans, now situated in the regional capital, stood by him. Aliyä, for instance, claimed in 2002 that she could sing 'any one of a thousand songs' Alim had released since the mid-nineties. When asked to name her favourite, however, she chose not a new song but an old one, 'Oghlum' ['My Son']: a song about 'teaching one's son how to be a man' (a man of the people).

Criticisms of Heyit were comparatively few, and mostly centred on the notion (discussed earlier) that he is too aloof, and his art too lofty, to reach the common man on the street. During a discussion held in 2002, Aliyä, a postgraduate student, and Häsän, a teacher, concluded that Heyit's work was far less well known than Alim's. They estimated that it reached the local populace in his hometown of Qäshqär,66 plus the educated elites in urban centres, but enjoyed less influence beyond these parameters. Presented with the proposition that 'high culture' may in fact carry less impact than popular culture if it fails to resonate with people's hearts across social classes, they were in full agreement.

Tahirjan, a migrant worker in his thirties from Aqsu, complained that Heyit was more prone to compromise than Alim, observing in 2004: 'Heyit has a bit of a problem … Some of his songs 'look angrily' at the government [*hökümätni güläydu*], but others seem to praise it. Ömärjan Alim, now *all* of his songs are uncompromising!' He did however concede that

<sup>66</sup> When Abdurehim Heyit performed in Qäshqär with the Xinjiang song-and-dance troupe in 2001, he followed his 45-minute act with an encore of equal length (Forney 2002).

### illuminists: popular song & the waking of the uyghur nation 229

artists in Xinjiang were obliged to try to maintain a fine balance between indicting the authorities and pacifying them. He also acknowledged the intensified monitoring of the arts that characterised the period since 2001: 'Heyit can't sing right now … he can't say what he wants to say [*eytalmaydu*]'.

Having heard much from Heyit and his companions on Ömärjan Alim, I was curious to learn what Alim and his circle thought of Abdurehim Heyit. Disappointingly, during my 2002 trip to Ghulja, Alim was out of town, and I was able to meet only with his relatives. When I mentioned Heyit's name to Alim's wife, she was initially unsure to whom I referred. Her mother jogged her memory by singing a stretch of one of Heyit's songs, imitating his deep, operatic voice in a somewhat facetious manner. Alim's wife listened, clicked her tongue, then said dismissively: 'Ah—the one with the funny moustache? I think I know who you mean. No, I don't really like him.' She subsequently began to make excuses for her patchy knowledge, explaining that she herself worked in the People's Courts and so 'had little to do with music or the arts'. It was difficult to discern whether she was ignorant of all music and art, or whether it was rather the case that she was more familiar with popular than with high culture. It was evident, though, that neither she nor her mother was impressed by Heyit's religious piety, as symbolised by his facial hair. The gap between the two 'voices' derived not only from social (class) difference, related to educational level, but also from another tension in Uyghur society: the discrepancy between an observant and a secular lifestyle.

# *Conclusion: New Folk as a Locus of Solidarity and Disunity*

The Uyghurs were arguably more united as a people and potential nation in the mid-1990s than at any other time previously: the result of common opposition to the Chinese state (and the Han migrants who are its symbol), constructed upon on a solid foundation of norms and values long shared by the in-group. However, there remained significant obstacles to a successful ethno-political mobilisation, the military power of the state aside. Only three social groups (and I present these as 'ideal types') fully subscribed to a separatist ideology in this period: young male intellectuals, young male petty entrepreneurs (often migrant workers), and unemployed males, all located in urban settings. Older people, women, and peasants, on the other hand, did not necessarily share these aspirations. Views on

#### 230 chapter four

Uyghur independence—and how to achieve it—differed significantly across generations, between urban and rural areas, and between men and women.67 As Naby had predicted a decade earlier (1986: 245), the attitudes of the youth had indeed become more radicalised. Yet the old were not given to impetuous acts, and both they and the middle-aged bore the scars of their persecution during the Cultural Revolution period (1966-76). Many women believed that the Chinese state 'would never let go of Xinjiang', and felt it made better sense to focus on achievable goals, such as the attainment of equal rights for Uyghurs. Many of those living in the countryside had not yet been touched by competition from Han migrants for education, employment, land and resources, so that their grievances were less than those of urban Uyghurs (with the important exception of rural opposition to birth control). Few had begun to look beyond their immediate, everyday circumstances (agricultural subsistence) to consider the political future of the Uyghur nation.

That said, I believe that the dissemination of alternative representations of Uyghur identity via the medium of popular song, together with the oral circulation of counter-stereotypes against the Han and the maintenance of distinct ethnic boundaries, went a considerable distance towards the construction of a viable nationalist discourse. Owing to the new possibilities of the cassette tape and the increased movement of Uyghurs between city and country, that discourse, for perhaps the first time since the founding of the PRC, began to spread slowly to the countryside and beyond the educated elite. Confucius once said: 'In altering customs and changing habits, nothing is better than music'.68 Between 1991 and 1997, two Uyghur *dutar* players constructed and disseminated representations of the Uyghur nation that contrasted sharply with the carefully packaged images of colourful, 'happy', *passive* ethnic minorities produced by the Chinese state (Gladney, 1994; 2004).69 These alternative representations, while not necessarily painting a nuanced picture of complex social conditions in the region, proved powerful reflections of popular grievance. Each artist produced strong identifications among his audience, and one of them succeeded in

<sup>67</sup> See Smith (2000) for an analysis of generational differences in political attitude in Xinjiang in 1995-6.

<sup>68</sup> Confucius, quoted by Kong Yingda, *Liji zhengyi* (cited in Tuohy 2001: 107).

<sup>69</sup> I was fascinated during a stay in Kucha in 1996 to find a microcosm of competing musical nationalisms, with state-sanctioned, co-opted minority music blaring out of public loudspeakers in the predominantly Han-settled New Town, as independently-produced Uyghur pop musics blared from cassette kiosks in the predominantly Uyghur-settled Old Town.

#### illuminists: popular song & the waking of the uyghur nation 231

creating an emotional resonance that reverberated among Uyghurs from all walks of life. Evidently, when taken too far by persons of a particular disposition, nationalist discourses can—and do—lead to violence both within and between ethnic groups, so that songs of nationalism may—as in the former Yugoslavia—'reaffirm hatreds between different peoples' (Hudson, 2000: 168). Thus, it could be argued that one extreme consequence of Alim's songs against 'collaborators' (and one perhaps unforeseen by the singer himself) was an increase in violent attacks and assassination attempts on Uyghur cadres and religious personnel.

The steady growth of nationalist sentiment culminated in the (then) worst instance of unrest in Xinjiang since 1949. On 5th-6th February 1997, a peaceful protest took place in Ghulja against the execution of thirty young Uyghur separatists, the arrest of a hundred Muslim students with allegedly pro-independence views, and state suppression of perceived 'illegal religious activities' such as the male social gathering known as the *mäshräp*. The protest was sparked into a full-scale riot after security forces opened fire on the first line of demonstrators, killing a young child. During the course of the riot, some young Uyghur males attacked Han civilians on the streets and set their bodies alight. These disturbances were met by an intensified government crackdown in the political and religious spheres, and eventually the cultural sphere (Amnesty International 1999, 2007; Human Rights Watch 2001; Kellner 2002: 16-27; Millward 2004: 16-17). Those Uyghurs later detained on suspicion of separatism [*fenliezhuyi* or 'ethnic splittism'] came from diverse backgrounds, and included secondary school teachers, peasants, merchants, a surgeon, a factory worker, a local Uyghur cadre, and the 'millionaire businesswoman' Räbiyä Qadir, this latter in a high-profile attempt to make a public example (Ch. *sha ji gei hou kan* or 'kill the chicken to show the monkey'). Some said that Qadir's arrest and incarceration (in 1998) had been made on the basis of information leaked to the Chinese authorities by internal 'spies', and attributed this behaviour to an in-group reluctance to see others achieve. Märyäm, divorcee and music trader in her thirties from Ghulja, told me in 2002: 'We can't unite, Uyghurs. We're always envious of others who do better [Uy. *bashqilarni körälmäy*], others of higher social status, and have to pull them down. Like Räbiyä Qadir. Even a woman of conscience like that!'

In 2002, a poet from Ürümchi named Tursunjan Emet recited a poem in Uyghur at the end of a concert at the Xinjiang People's Hall in the regional capital. The Party committee subsequently ruled that the poem contained an 'anti-government' message and labelled the mere act of recital an 'eth-

#### 232 chapter four

nic separatist crime'. Following this incident, the Chairman of the Xinjiang provincial government opened an investigation, vowing to purge all those who 'openly advocate separatism in the name of art'. In a move reminiscent of the Cultural Revolution at its height, cadres were instructed to use politics as the *only* standard when judging artistic and literary work. Emet went into hiding, and was later detained and subsequently released some weeks or months later (Human Rights Watch 2005). In the wake of this renewed cultural crackdown, Xinjiang's new folk singers were forced to search for an alternative means of expression. This period was so politically tense that, according to Abdurehim Heyit's friends, it was 'simply too dangerous to use nationalist lyrics'. The answer for Heyit was to resort to the power of music itself, or 'sonic nationalism'. The title track of his 2002 release, *Ömüt* [Hope], was a powerful instrumental, conveying by turns passion, desperation and determination. In the absence of lyrical metaphor, the *dutar* became a 'political instrument … giving voice to the collective feelings of the group' (Cooke and Doornbas, 1982: 53). Asked in 2002 whether there was still hope for the troubled region, the singer responded as always in guarded fashion, yet his message was clear: '*Hayat bar bolsa, ömüt bar* [Where there's life, there's hope'].

PART THREE

illuminists: popular song & the waking of the uyghur nation 233

'SILENCE': AFTER THE 1997 GHULJA DISTURBANCES

chapter four

#### Chapter Five

reverts: sources and dynamics of islamic renewal 235

# reverts: sourCes and dynamiCs oF islamiC renewal

# *Introduction*

as i have argued in part 2 of this volume, non-violent symbolic resistance among the uyghurs was expressed through three main vehicles during the 1990s: negative stereotyping of the han (and positive counter-stereotyping of the in-group), symbolic, spatial and social segregation from the han, and alternative representations presented through the medium of 'new folk'. however, the Ghulja disturbances of February 1997, their military suppression and the ensuing state crackdowns represented a turning point in uyghur national politics: simultaneously an ethno-political climax and anti-climax, which ushered in an era of silence. in the wake of this event, uyghurs found themselves increasingly confronted with a bi-polar world in which they could resist and face marginalisation, or accommodate to ensure survival. as dilbär, an 18 year old *minkaomin* high school student, explained in 2004: 'there are basically two roads open to uyghurs, although they are not always strictly demarcated. in my family, for instance, i'm studying at university, but my little brother is taking religious education; so the two roads can co-exist within one family.' similarly, the elder son in a family i knew in Ghulja in 2002 had chosen to study in a secular state school and try to succeed from within the system, while the younger son refused outright to attend a Chinese-medium school, instead taking private religious education and planning to go into business independently. in the wake of tightened controls over verbal utterances considered to be 'local nationalist' and intensified censorship of cultural products from 2002, the situation worsened further. now, two forms of symbolic resistance previously in operation—oral stereotyping and alternative representations in popular song—were no longer viable. while ethnic segregation continued to be enforced by the vast majority of uyghurs vis-à-vis the han (a small minority of deeply acculturated, school-age *minkaohan* are an exception), a new symbolic boundary emerged among some sectors of the population. that new boundary was islam itself, a renewal of orthodox religious practice, a return to the mosque, as distinct from the social practices informed by Islam which had featured in the negative stereotypes—and positive counter-stereotypes—deployed in the mid-nineties.

In Part Three of the book, I will suggest reasons why resistance during the post-1997 period has taken the form of Islamic renewal alongside a strengthened emphasis on (selective) endogamy, and illustrate how Uyghur urban youths have learned to negotiate multiple and hybrid identities, some of which transcend national borders. In this chapter, I examine the new form of symbolic resistance which emerged during this period of crackdown and reduced civil rights: the return to orthodox Islamic practice. Beginning with a discussion of the ways in which globalising forces aided the flow of traditionalist Islamic ideologies into Xinjiang from the start of the eighties, I go on to identify and analyse six sources of religious renewal, including Islam as:


In conclusion, I underline the diversity of reasons for the current Islamic renewal, and stress that this process has been symbolic, peaceful and cathartic in Xinjiang rather than violent or extremist, as the state suggests.

# 5.1. *Islamic Renewal in Ürümchi since the Late 1990s 1*

By 1989, the number of mosques in Xinjiang had increased by 5.8 times compared with a decade earlier to some 20,000; but, of that figure, 17,540 were said to be in the south (Wang 2004), where Uyghurs greatly outnum-

<sup>1</sup> It is hard to gauge the nature and complex sources of the Islamic renewal underway in Xinjiang other than through qualitative ethnographic means. Statistics on numbers of mosques in the region are useful initial indicators, but do not tell us how many people are visiting the mosques or with what frequency, nor who is going in and why. And they tell us nothing about observant Muslims in education or attached to state work units, who are forced to conduct prayer in private. My empirical data are based on interviews and observations carried out mainly in Ürümchi but also in Ghulja during the summers of 2002 and 2004. Interviews were conducted with a combination of long-term respondents (those I had known for eight years or more) as well as new respondents.

ber Han Chinese and have historically been more observant than those in the north. By the end of the 1990s, the number had reached 23,000, and was continuing to grow (Mackerras 2001: 292). There are indicators that religiosity had begun to revive in the south by the mid-late nineties. While conducting fieldwork among southern peasants, Ildikó Bellér-Hann noted that observance of public worship on Friday, praying once a day and fasting during Ramadan were usually cited as important requirements for Muslim males (2008: 311). In 1996, peasants in rural Aqsu showed me printed leaflets (origin unknown) containing information about the Day of Judgment, which they said had 'scared them into praying', whereas their attitude towards prayer had originally been lax. A revival of traditional Islamic almsgiving was noted, whereby peasants who produced a surplus sometimes donated it to the poor; this charity was often coordinated by indigenous *dadui* [brigade] officials (Bellér-Hann and Hann 1999: 10). The picture of local officials' participation in spontaneous religious renewal is further suggested by the story of a *xiaodui zhang* [production team leader] who led his community in building a new mosque, an act characterised as 'an initiative which came from his neighbours and had nothing to do with his position in a formal political hierarchy' (Bellér-Hann and Hann 1999: 13). Although appointment to an official post precluded religious participation, many rural officials ignored this until the rule was reinforced in 1996 in response to increased separatist unrest. Even then, they often returned to the mosque following retirement as this was 'a precondition for acceptance in the local moral community' (Bellér-Hann and Hann 1999: 16-17). In urban Qäshqär, there was a strong reliance in 1993-4 on the orally transmitted knowledge of religious elders. However, this local knowledge has latterly been challenged by reformist Muslims claiming to uphold a more 'orthodox' form of Islam, a development attributable to increased transnational links and greater access to the scripturalist traditions (Waite 2007: 173-177).

In the north, the process of religious renewal proceeded more slowly. Mackerras observed in 1990, 1992 and 1994 that older men formed the main core of mosque-goers in Xinjiang, while adherents from the younger generations were conspicuously absent (1995: 116). Nor was the trend in evidence when I conducted my doctoral fieldwork in the regional capital Ürümchi in 1995-6. Significantly, it gained momentum only towards the end of the decade, indeed, only after the 1997 Ghulja riots and subsequent crackdown on 'illegal religion'. While the 1996 Strike Hard campaign had targeted 'political activists' and 'pro-independence activists', the 1997 post-Ghulja campaign focused on 'splittists' (political separatists), 'counter-

#### 238 chapter five

revolutionaries' and 'religious fundamentalists'. After the events of 9.11 and China's decision to jump aboard the US bandwagon known as the 'war on terror', repression of religious activities in Xinjiang deepened further as state campaigns began to speak of 'international terrorists' and 'jihadists' (Dwyer 2005: 53-57).

The current return to prayer has necessitated the replacement of the city's original mosques with new ones many times their size and more attractive and splendid. While larger mosques have been built on existing sites with the aid of state funding, there is also evidence of new sites being opened up for small-scale mosque construction, funded by private donations. By the time of my field trip in 2002, the former *Aq Meschit* [White Mosque] at Shanxihang had been replaced by a larger, new mosque, with beautiful turrets and domes and intricate carved wooden doors (the completion date on the façade read 2001-8-21). Where the doors of the former mosque had opened straight onto the road, the new mosque was set back, allowing for a large forecourt to accommodate greater numbers of worshippers. The replacement Yan'anlu mosque remained under construction in 2002 but was completed by the time of my field trip in 2004.2 It was a huge and magnificent building, again set back off the road, with a large square forecourt in front. I asked two petty traders what had become of the tiny old Yan'anlu mosque; they responded enthusiastically: 'That's gone! Pulled down! […] With numbers of worshippers increasing so rapidly, they couldn't fit them all in [*patmaydu*]! The new mosque has two floors!' Another large, ornate, multi-storeyed mosque being built in the Döngköwrük district in 2002 was also complete by 2004, making a total of four huge mosques in the relatively small, Uyghur-dominated area between Nanmen and Yan'anlu.

Clearly, the government is complicit in the construction of these large mosques, which have benefited greatly from state funding in addition to private donation, and has its own reasons for supporting the venture. As Dilbär, eighteen, female high school student, remarked in 2004: 'Of course, one should not forget the element of tourism: the government knows that if it builds mosques and other buildings with "ethnic" characteristics, foreigners will come to see them and will be interested in the local "culture" … then it can say "Look! We built all these mosques for our Muslims; aren't

<sup>2</sup> In 2002, a large impressionist painting of the planned new mosque was mounted before the construction site. Though the mosque was not yet in use, people milled excitedly in the forecourt.

Figure 12: Friday prayer at the new Yan'an Road mosque, Ürümchi (2004)

#### 240 chapter five

we good to them?"' Many respondents interviewed during the same year highlighted the relatively lenient policies towards religion in Ürümchi (where Uyghurs are outnumbered by the Han and considered comparatively loyal to the state), and suggested that the government hoped to buy the loyalty of its more 'civilised' (read, Chinese-educated or Chinese-speaking) Uyghur population by granting it freedom of religion.3 Such policies contrast starkly with the strict religious policies in operation in the south, where Uyghurs are dominant in number, considered less 'civilised', and historically resisted Chinese rule. As Rudelson has pointed out, both strategies of control are doomed to fail since the suppression of Islam invites opposition, while its tolerance leads to an intensified sense of separateness from (Han) Chinese society (1997: 48). A second factor behind the Islamic face-lift in the regional capital may be the state's desire to maintain good diplomatic and economic relations with other Islamic nations.4

However, as observed with reference to the spread of Christian church meetings, a politics of permission is at the same time a politics of response; religious resistance in China often takes the form of a 'negotiated compromise with state agencies' (Feuchtwang 2000: 167). Tourists and Islamic foreign dignitaries are not alone in taking advantage of these new Islamic spaces. The volume and composition of mosque-goers has changed radically. On each occasion in 1995 and 1996 when I visited Ürümchi's principal mosques on a Friday (traditionally the most important day of prayer), I observed only a handful of elderly men going in. The sole exception was on the occasion of a funeral, when the former Yan'anlu mosque was attended by a larger crowd of elderly and middle-aged men, one of whom commented wryly: 'The only time we go to the mosque these days is when someone dies.' Compare this with an observation made at the *Aq Meschit* on a Monday in summer 2002:

Six years earlier, you would barely have seen a soul going into the mosque on a non-Friday, just the odd bearded elder. What a difference now! Some men were already inside when I got there […] Men drifted in continuously until about 12.45; unlike in the mid-90s, these were not only men in their sixties in long coats, hats and *ötük* [tall, leather boots]; there was an equally large number of young men, ranging from teenagers to those in their thirties,

<sup>3</sup> *Sovetlik* (Kazakhstan-based) Uyghur nationalists believe that the Chinese government has deliberately promoted religion—the 'opium of the masses'—in Xinjiang as a means of keeping its Muslim population from becoming revolutionary (Roberts 1998a: 521).

<sup>4</sup> Note that for this reason mosques in Peking and Shanghai—two cities which received a high number of foreign visitors—were kept open during the Cultural Revolution (Bush 1970: 296).

as well as some younger boys taken in by fathers and grandfathers. The middle-aged men were conspicuous in their relative absence. A few were there, certainly, but they were greatly outnumbered by the upper and lower generations. Of course, this generation had grown up during the Cultural Revolution [and so had not been educated in the Arabic script, or in prayer rituals].

Some respondents interviewed in Ürümchi in 2002 suggested that these worshippers may have been southern traders and businessmen passing through; others noted that there are simply more Uyghurs (migrants) in the regional capital now. Certainly, there is evidence that reformist Muslims from Qäshqär may be congregating in Ürümchi because of the city's greater religious freedom (Waite 2007: 172). Yet the suggestion that 'only southerners go in the mosque' probably reflected a southern bias against Uyghurs in the capital, who are often assumed to be religiously 'empty' due to their relative proximity to Han settlers. Ürümchiliks (Uyghurs born and raised in Ürümchi) confirmed that many Ürümchi Uyghurs are now themselves returning to religion. Two male traders selling fruits before the new Yan'anlu mosque in 2004 explained: 'Now the youth are going into the mosques … this is a very good thing! It demonstrates the strength, the power [Uy. *küchlük*] of Islam!' The young mosque-goers I observed seemed to be mainly businessmen and petty traders, although this clearly reflected the mosques' location in the Uyghur-dominated southeast, where few Uyghurs are employed by work units or Han companies and are therefore free to go to prayer. Zunun, a state employee in his thirties, pointed out in 2004 that one reason why mosques are smaller and fewer in the northwest of the city is that most Uyghurs there are attached to work units and cannot get time off to pray, nor are they officially permitted to pray. He was keen to stress that this does not necessarily mean that these men are not observant (I will return to this theme later).

An increase in religiosity is also visible among some young women in Ürümchi. In 2002, I observed that the number of women wearing full veils or tying headscarves beneath the chin (rather than at the nape of the neck) had greatly increased in Döngköwrük compared with the mid-nineties, and included very small girls. Textile traders, young and old, prayed in designated rooms in the bazaars. Respondents confirmed these observations to be an accurate reflection of the trend. Gülhärä, a 20 year-old Uyghur-educated student, explained in 2002: 'The middle-aged women don't wear headscarves, but a lot of young women are wearing them now around Döngköwrük, passed under the chin […] some wear the full veil with just 242 chapter five

Figure 13: Female modesty as national honour, Yan'an Road, Ürümchi (2004)

their eyes showing. Even the tiny girls wear scarves.' According to Yultuz, another 20-year-old Uyghur-educated student speaking in 2002, some women are covering up on their own initiative, while others are persuaded to do so by their husbands following marriage. However, it would be simply untrue to claim a strong element of male coercion; persuasion, where it occurs, is usually gentle and based on explanation of Islamic teachings. In some cases, wives are more observant than their husbands. Ghäyrät, migrant worker from Aqsu in his thirties, told me in the same year: 'A friend's wife in my hometown conducts five daily prayers, but he drinks, smokes and has a good time. He often says to her "You don't need to wear a veil," but she does so because she wants to. No-one is forcing her.'

Perhaps more indicative of a general shift in levels of religiosity in the capital is the emerging evidence surrounding prayer beyond the public eye, that is, among students and state employees. Despite their exposure to the secular state curriculum and state prohibitions on their participation in religious activities, I detected a strong desire among Uyghur-educated high school and university students to pray, and a willingness to expend considerable energy in circumventing the law that forbade this. Interestingly, some *minkaohan* (Chinese-educated Uyghurs) are also taking part in prayer, especially those exposed to a strong 'counter-education' in the family home. Gülhärä explained in 2002:

I pray at home. We're not allowed to pray at college, but we do anyway, in the dorms where no-one can see us … some *minkaohan* [Chinese-educated Uyghurs] pray too, those who've been taught well in their family, whose mum and dad showed them how. We can't wear headscarves at school. I would like to, perhaps when I graduate. But, if I get a job in a state work unit, I still won't be allowed to wear one! I'll wear it at home then.

As Fuller and Lipman observe, the surveillance and suppression of religion has thus had the paradoxical effect of strengthening the central role of Islam in Uyghur life (2004: 344).5 For Dilbär, eighteen, a Uyghur-educated student living in Han-dominated northwest Ürümchi in 2004, family education played a key role in maintaining religious identity. Her grandparents scolded her if she didn't wear a scarf, while her parents were less strict on head coverings but forbade her from wearing what she termed 'wild clothes' (short sleeves, etc.). In her view, the Chinese authorities were reluctant to allow young girls to wear headscarves at school for fear that it would encourage or pressure other girls to veil up and thus 'exert a bad influence' (i.e. strengthen the common, religious identity). Despite the considerable time and space constraints imposed by the daily school routine and the law, she confirmed that she and her classmates tried to pray as often as they could:

I like to pray, but it's hard. Because we're so busy rushing to and fro from school. I did pray regularly a while back; I put a long black veil over my head […] but then I missed a few days, being so busy. And it's no good stopping and starting. I want to start praying again, though. It's the same for the boys, they barely have time to rush out for lunch when they're at school, so it's hard to go to the mosque. But many go on Fridays.

<sup>5</sup> Prohibition of religious practices has often had the reverse effect in China; during the Cultural Revolution, the number of underground 'house churches' grew, with bonds among believers strengthening in response to the common risk of professing belief (Feuchtwang 2001: 164).

# 244 chapter five

Bahar, eighteen, Uyghur-educated daughter of a first-generation *minkaohan*, told me in 2004 that while she found the secular climate of the state education system stifling, she nonetheless made sure to pray at least once a day:

I know how to pray. But I only know the words, I don't know the actions [Uy. *härikätlär*] […] I pray every day at night-time, just mouthing the words, you know. I really like to pray, but I'm very busy with school, so I don't have much time. Most of the boys at my school go to the mosque on Fridays even if it means missing their lunch [laughs].

I later interviewed some of her male classmates, and discovered a strong tendency at their schools to apply peer pressure on those who did not enter the mosque. Ömär, eighteen, a Uyghur-educated student interviewed in 2004, described the process as follows:

The strength of religion has grown, yes … it's a powerful force, an abstract thing, it's hard to explain. But I know one reason is that we put pressure on one another to go to the mosque … in my class [high school], the male pupils ragged me for not going on Fridays, and so I began to go. And another lad has started to go now because I railed at him and said 'You didn't go to the mosque …' And so it goes on. And we don't trust people if they don't believe in Allah.

He described it as 'the force of propaganda' and explained that classmates 'feel good about themselves and each other' if they go to the mosque. He was visibly glad as he described his feeling on seeing the errant classmate at the mosque the previous Friday, 'wearing a snowy white *doppa* and looking really happy'.

The process of peer pressure is also in operation in the broader community (where it may be dubbed 'community pressure'), and was mentioned at the other end of the spectrum by Ibrahim, an elderly religious man in his seventies originally from Khotän. He explained in 2004:

Numbers of religious devotees have increased in Ürümchi partly because people tend to imitate one another [Uy. *bir-birsini doraydu*]. They watch the VCDs, listen to the cassettes, read the books and become afraid; afraid of Allah. Some give up alcohol and cigarettes, and when others see this, they feel ashamed and embarrassed and do the same.

However, he questioned the depth of religious knowledge among some new devotees, suggesting that many don't understand the meanings behind *namaz*, but 'read them in Arabic as if they are blind men [Uy. *qarghularchä*]'. Patigül, a music trader in her early thirties interviewed in the same

year, similarly suggested that some new mosque-goers were not profoundly religious individuals: 'Everyone goes in the mosque now, not just those who recite *namaz*, but also those who don't … the thieves [Uy. *oghri*] go in the mosque now too!' This strongly suggests that, for some people at least, the renewed popularity of the mosque masks a 'ritual of defiance' rather than the simple rediscovery of religiosity.6

# 5.2. *Globalising Forces: Border Trade, Pilgrimage, Study Abroad and News Media*

As a border region, Xinjiang has been characterised variously as the critical nexus between China and Inner Asia, the Middle East and Europe (Zhao 2001: 220), and as holding a pivotal position at the crossroads of six cultural and geographic regions (Becquelin 2000: 65). It is a territory that has been affected at different times—and over time—by differing cultures and shifting politics on all sides. Most recently, it is the opening of China (including Xinjiang) to the West since the 1980s and the emergence of the independent Central Asian states in 1991 which have once more 'dramatically altered the geocultural and geopolitical dynamics of the Eurasian landmass' (Zhao 2001: 220). For decades, the 'closed' nature of Chinese communist politics had restricted access to religious ideologies in other Muslim societies (see Waite 2007: 165-166; 168-169). By contrast, over the course of the nineties, forces of globalisation have vastly increased the flow of ideas and feelings from outside Xinjiang into the region (cf. Mackerras 2001: 301). Like contemporary individuals everywhere, today's Uyghurs are subject to 'an extraordinary diversity of information […] images, concepts and lifestyles from well beyond their immediate locales' (Kim 2005: 446). My data show that a) imported Islamic materials; b) pilgrimage to Mecca; c) Islamic theological studies abroad; and d) increased access to media coverage of Muslim issues worldwide are all playing a part in dynamics of re-Islamisation in Xinjiang.

Border trade between China and Central Asia and other neighbouring countries has been developed rapidly over the past two decades, as has domestic and cross-border infrastructure. The volume of cross-border trade reached \$94 million in 1991 and \$1.02 billion by 1999 (58 per cent of the region's total trade volume); Kazakhstan is Xinjiang's largest trading part-

<sup>6</sup> Feuchtwang gives examples of similar 'rituals of defiance' in Tibet (2001: 173).

ner, with border trade being conducted at 27 open ports (Zhao 2001: 207). According to Xinjiang's Outline for the Ninth Five-Year Plan, border trade was to be further expanded in the north-west and south-west of the region during this decade (ibid.: 216-217), implying expanded trading relations with Kazakhstan, Tajikistan, Afghanistan and Pakistan. By 1995, 96 per cent of Xinjiang's villages and towns had been linked by highways, and there were five international flight routes linking Xinjiang with Central Asia, Iran, Europe and other countries; meanwhile, China's Xinhua news agency reported in 2000 that a new railway would link Xinjiang with Kyrgyzstan and Uzbekistan. A network of three vertical and three horizontal highways is planned for completion by 2020, which will link north and south Xinjiang via the Tianshan mountains and join the region both to China's eastern provinces and to Central Asia and other countries to the west (ibid.: 211-2). In addition, China has begun to develop international, cross-border and domestic tourism in Xinjiang (ibid.: 214). In this way, the state has allowed the Xinjiang regional government substantial freedom in approving foreign trade and establishing border stations, as well as in setting up companies aimed at exploiting business opportunities in the rich Islamic countries (Shichor 1994: 82).

Yet increases in border trade and improvements in communications may also backfire, leading to increased religious and cultural exchanges between Uyghurs and Central Asians and other foreign nationals to the west. For instance, while the Chinese government would have liked to use foreign Muslim investment to build factories, in reality several large donations have gone to building mosques and madrasahs in Xinjiang, Gansu, Qinghai and Ningxia (Gladney 2004: 317).7 The isolation of Muslims in Xinjiang has therefore come to an end, and closer religious connections with the outside world have reinforced religious identity and the 'sense of separateness from the Chinese' (Dillon 2004: 167).8 In particular, the import

<sup>7</sup> Compare the situation in Central Asia: while the new republics were initially enthusiastic about the prospect of economic support from the Middle East, they have latterly been cautious in playing their 'Islamic card' due to fears of imported extremism. However, their economic situation has at times compelled them to look in this direction for help. A reluctant Nazarbaev was forced to accept assistance from Oman, Saudi Arabia and Iran despite the fact that their donations were tied to funds for the construction of new mosques in Kazakhstan (Tazmini 2001: 76).

<sup>8</sup> Dillon particularly mentions religious influences from Central Asia; however, given the fact that many Central Asians rather look to their cousins in Xinjiang as the true keepers of religious customs, influences from Saudi Arabia and the Middle East may be more critical. With regard to Central Asia, the reverse trend has been happening, i.e. Uyghur

of Islamic materials is contributing to the current trend of renewing faith or, in some cases, learning about religious rituals for the first time. There was formerly an extreme scarcity of religious publications, and Uyghurs had little access to alternative forms of knowledge in the print media; the Qur'an itself was not translated into Uyghur until 1986 in the contemporary era (Waite 2007: 165).9 By contrast, imported Islamic materials—some of which are translated, produced and copied locally—now flood the region in the form of books, pamphlets, audio cassettes, CDs and VCDs, some sanctioned by the state, others 'illegal'. It is known that imported Islamic literature has been confiscated by customs officials at Xinjiang's borders since at least 1993 (Shichor 1994: 71). Such materials have nonetheless been finding their way into the region. In July 1996, *Xinjiang Daily* called for a crackdown on what it called 'illegal publications and audio and video products that promote religious fanaticism' (Dillon 2004: 88). The problem remained at the top of the state agenda in 1999, when on January 1 a special 100-day 'Strike Hard' campaign was launched in Ürümchi against pornography and 'illegal publications with political problems', a code for separatist and unofficial Islamic materials (ibid.: 127).

In an interview conducted in 2002, Jelil, an observant graduate in his thirties, confirmed that works about Islam are now being translated into Uyghur and circulated, and that many people are learning about Islam from these previously unavailable educational materials. Two male music sellers in their thirties speaking in 2004 confirmed the trend, explaining:

Uyghurs are more developed now [*täräqqi qilghan*] … they know more about Islam, their minds are broadened. This knowledge has come from books. We didn't have books before … not about Islam. They sell them on the street; you can find them right outside this shop.

Dilbär, eighteen, female Uyghur-educated high school student, drew attention in the same year to the local production and proliferation of such materials:

We think about the next world, you know, what will happen when we die, will we go to Heaven or Hell? And we know more about Islam now; before there was very little information or instruction about Islam. Now, lots of

traders from Xinjiang have been spreading Islamic practices in Kazakhstan with the goal of 'educating' their Russified *sovetlik* relatives. See Roberts (1998a: 522-3).

<sup>9</sup> China's first translation of the Qur'an in the vernacular was published in Shanghai in 1952, presumably in Chinese and intended for consumption by Hui Muslims. Its preface noted that it had previously been 'impossible for ordinary Muslims in China with no knowledge of Arabic to fully understand the teachings of Islam' (Bush 1970: 271-2).

#### 248 chapter five

materials have been translated into Uyghur. Much of it is written out by hand or printed in small pamphlets.10

Mihray, a middle-aged woman in full headscarf (i.e. passed under the chin) who worked on a religious book stall near the Yan'anlu mosque, spoke in 2004 of Uyghurs previously unversed in prayer rituals learning these from newly available publications and audio-visual media. Other interviews conducted in 2004 supported the emergence of this trend toward religious self-education, with one woman in her sixties showing me a book from which she had recently taught herself how to conduct *namaz*. She had grown up in a religiously lax southern rural community and was now learning for the first time how to conduct the five daily prayers. Edmund Waite has observed a similar process of religious self-education taking place in Qäshqär (2007: 173).

Second, the interactive experience—globally and locally—of pilgrimage has both contributed to the development of a broad, supra-national Islamic identity and increased knowledge of Muslim issues in the outside world. Pilgrimage resumed in the first decade following China's opening, with 6500 Xinjiang pilgrims visiting Mecca between 1980 and 1987 (Shichor 1994: 80). Following the collapse of the Soviet Union in 1991, the Chinese authorities were quick to accuse outside separatist groups of 'subverting pilgrims' en route, and in 1996 they announced a limit on numbers allowed to go on the *hajj*, with officials complaining that local imams had returned to China 'dressed like Arabs' (Dillon 2004: 66, 90). Notwithstanding, official *hajj* numbers for the late nineties regularly surpassed 6000 each year, though this figure is for the whole of China (Gladney 2003: 463); and it is certain that many more travelled in a private capacity. In 2001, more than 6000 pilgrims visited Mecca, mostly privately financed (Gladney 2004: 235). Micallef and Svanberg argue in their writings on contemporary Central Asians that the pilgrimage tradition has provided 'a space for popular discussion that is difficult to monitor' (1999: 158-60). Whether or not some Uyghur pilgrims are being subverted by 'outside separatists', the *hajj* experience is at least serving to increase religiosity among pilgrims and their neighbourhoods back home, and to enhance their sense of being part of a broader Islamic community. We know that during the Qing period, Hui

<sup>10</sup> Compare the Islamic revival in Central Asia following *glasnost* and *perestroika*, where Muslims 'printed previously banned Islamic literature and simple pamphlets that described how to pray' (Micallef and Svanberg 1999: 157).

Muslim notables who visited Mecca 'told their fellow Muslims of the marvels of the Islamic world and of their brethren there' upon their return to China (Israeli 1984: 297). Similarly, the rituals attached to Uyghur pilgrimage in the pre-socialist period are said to have 'had a bonding effect within the mosque community, which was further strengthened by the presents distributed by the returning pilgrims. The stories pilgrims told about their journey also became communal property, and they served the integration of the pilgrim's community into the wider community of Islam' (Bellér-Hann 2008: 349). My contemporary data point to a similar process occurring today. Interviewed in 2004, one respondent described it as follows:

Now, people's knowledge of Islam is deeper. Not only have they been able to read more about it, but a lot more people are also going to Mecca. They come back and tell us what they saw, what they heard; they tell us about the Kaba. They disseminate information [*täswiqat*], encourage us with propaganda. They tell us all they've heard about what will happen to us in the Afterlife if we are not good Muslims … we're afraid of those things.

In this way, religious information is passed on by pilgrims and 'community pressure' is exerted on others to be better Muslims.11 Mecca (and the Middle East) may thus gradually come to embody an alternative centre of identification.12 A similar phenomenon was observed in Daghestan in the former Soviet Union, where from the early nineties young people took religious training in the Middle East, with teachers from the Middle East, or with graduates of Middle Eastern teachers. These individuals grew beards (in place of moustaches), wore Arab-style trousers, and abstained from drugs and alcohol (Gammer 2005b: 836).

Some arguments have been put forward against the notion of pilgrimage leading to the development of a supra-national Islamic identity. It has been reported, for example, that many *hajj* returnees mention an increased sense of affinity with the Uyghur ethnic group rather than with members of the international community (Gladney 2004: 224). To be sure, pilgrims notice

<sup>11</sup> Community supervision was used in the late 1800s to enforce Islamic norms in Uyghur communities. See Bellér-Hann (2008: 344) for details of the series of sanctions imposed on men who failed to attend the all-important morning prayer, including: verbal reminder; threat of beating; being undressed and having cold water poured over the head; whipping; expulsion from one's home; oral examination by an *akhun*; and, finally, excommunication from Muslim burial rites.

<sup>12</sup> See Becquelin, who writes: 'The concept of alternate centres highlights why […] assimilation (and thus political loyalty to Beijing) will never be either complete or irreversible' (2004a: 377).

#### 250 chapter five

national differences in terms of the contrastive customs of other groups with whom they come into contact; yet I would suggest that this does not preclude the broader development of a common religious identity or an appreciation of universalist Islam. Rather, contrast with the 'other' occurs simultaneously with comparison. There has also been a suggestion (in contradiction to the above) that pilgrimage, rather than improving social cohesion within the community back home, actually creates social divisions (Gladney 2003: 461). Gladney suggests that increased travel to the Middle East and exposure to new and radical ideas has prompted criticism of Muslim practices at home and has thus contributed to 'factional struggles' among China's Muslims (presumably between Sufi sects or subsects).13 This notion is perhaps better applied to the Hui Muslims of the Northwest than to the Uyghurs, especially when one examines Hui history. However, there are indications that the dissemination of reformist ideologies by some *hajj* returnees has created limited discord between observant and nominal Muslims in Xinjiang, as the former put pressure on the latter to abandon local, syncretic religious practices (see Waite 2007: 172). Even so, it appears that for each person who resents the proselytising activities of the returnees, another person is persuaded to enact a lifestyle change.

Another conduit for the re-Islamisation process has been study abroad. At the start of the reform era, several Muslim countries expressed a desire to strengthen religious exchanges with China's Muslims. In the mid-1980s, King Riyadh of Saudi Arabia recommended that Sino-Saudi religious exchanges be expanded after observing the 'overall adverse condition of Islam and Muslims in Communist China'; access to China's (and Xinjiang's) Muslims was subsequently facilitated by the establishment of diplomatic relations between Saudi Arabia and Beijing in July 1990 (Shichor 1994: 80). Also during the eighties, the Islamic Development Bank gave China over 4 million US dollars to spend on four projects in the Northwest, including the enhancement of the Xinjiang Islamic Academy (Dillon 2004: 44). In 1989,

<sup>13</sup> Clearly, pilgrimage *can* bring ideological influence to bear on local communities; consider for instance the three Sumatran pilgrims who returned from Mecca in 1804 to launch the puritanical, revivalist Padri movement, which attacked lax local practices and aimed to protect locals against European robbery and slavery (Choueiri 1997: 9). To give an example from the Chinese context, during the eighteenth century, several Muslim clergymen and scholars from Hezhou District in Gansu went to the Middle East on pilgrimage or to study, and were responsible for the import of Sufi activism. In 1750, Ma Laichi returned to convert the Salar to Khafiya teachings and practices; in 1761, Ma Mingxin brought back the less tolerant ideas of Jahriya revivalism, which was to lead to communal violence (Goodman 2005: 330-1).

Ayatullah Ahmad Jannati, president of Iran's Islamic Propagation Organization, stated his preparedness to strengthen Iran's relations with China's Muslim community through the exchange of cultural and scientific delegations and (significantly) theology students (Shichor 1994: 80). Three years later, in September 1992, Iranian president Rafsanjani visited Qäshqär, led prayers in the Heytgah mosque, and was cheered by local crowds as he walked briefly round the square afterwards (Dillon 2004: 136). Educational exchanges have subsequently been set up with a number of Islamic nations, exposing both Hui (cf. Gladney 2004: 280) and Uyghurs to international aspects of their religious heritage. Of thirty four students from China enrolled in Islamic and Arabic studies courses in 1993 at Egypt's Al-Azhar University, twenty eight were Uyghur. These students were studying on private scholarships provided by relatives (some of whom lived abroad), the Muslim World League in Saudi Arabia and the university itself (Gladney 2004: 234, 237).14 Since 1996, China and Egypt have officially exchanged twenty students a year. Yet in 2003 there were around 350 Chinese nationals studying in that country, suggesting that a significant proportion of these students were self-funded (Gladney 2004: 237).

Abdullah, an observant Muslim I interviewed in Ürümchi in 2002, had spent a protracted period between 1994 and 1999 in Pakistan, initially studying theology at a religious school. The experience had rendered him antagonistic to controlled Islam in Xinjiang, and he had been arrested in 1999 following an argument with an individual he labelled the 'government imam' at the Nanmen mosque. The latter had berated him for saying 'Amin' at the end of prayer, a practice outlawed at that particular mosque.15 Others at the mosque (presumably government informants) had told the police, who had subsequently arrested him and questioned him about his activities in Pakistan.16 Abdullah's view on the episode was as follows:

<sup>14</sup> In the late nineteenth and early twentieth centuries, most of the Muslims who were to exert 'substantial influence' on Islam in China studied at Al-Azhar University, the foremost Sunni Islamic training institute. Gladney (2004: 236-8) suggests that for this reason, only those students who are less religiously inclined are considered for state funding.

<sup>15</sup> Declaration of the word 'Amin' aloud at the end of the Quranic verse derives from Hanbali prayer rites, and is said to be an indicator of those who follow the reformist path (Waite 2007: 169-170).

<sup>16</sup> The Chinese authorities have arrested some young Uyghurs who returned to Xinjiang after studying Islamic law in Pakistan, claiming that they are Taliban members of Pakistani or Afghan citizenship. See also Becquelin (2004b).

#### 252 chapter five

I don't like this imam. He has sold his conscience to the Chinese, like an infidel. Look at his long white beard! He has sold it to the infidels! He has sold his heart. They give him money, you know, to be a traitor … by the back door route [*arqi ishik*] … They wanted me to be an imam! They offered me this money too! But I said 'No, thank you very much.' Never!

I interviewed several other observant Uyghurs in 2002 who claimed that Chinese officials had attempted to buy their political loyalty in this way. Meanwhile, the government has been reluctant to issue passports to Uyghurs since 1997, owing to concern that foreign travel will infect Uyghurs with extreme nationalist and Islamist ideas (Fuller and Lipman 2004: 348). During my trips to the region in 2002 and 2004, respondents confirmed that visa applications for study abroad were being carefully screened and the political and religious background of all applicants checked.

In view of state attempts to restrict and monitor pilgrimage and study abroad, the primary vehicle for the development of a global Islamic identity in Xinjiang may be the audio-visual mass media. As Dilbär, high school student, eighteen, explained in 2004, Uyghur-language materials printed in the Arabic script, including books, pamphlets and newspapers, remain less accessible to *minkaohan*, whose first language is Chinese, unless the ideas and beliefs they contain are explained orally by family or friends: 'It depends on their family situation; they can't read the script as fast or as readily, and so it will depend on what their parents talk about at home.' On the other hand, audio tapes, VCDs, and radio and television broadcasts national and international—can reach all Uyghurs regardless of language background. Kim lists a number of theoretical conceptualisations of the globalisation phenomenon, among which the notions of 'accelerating interconnectedness' (the intensification of global social relations and consciousness of world society) and 'action at a distance' (when actions of social agents in one part of the world come to have significant consequences for distant others) are most relevant to our context (2005: 446). One example of these dynamics is radio broadcasting, which has brought many local Muslim populations into touch with what is going on in the rest of the world:

People are now […] anxious to formulate Islamic positions on current political, economic, and social issues—or, in other words, to think of the world's problems in Islamic terms […] Muslims come to know how dissatisfaction and protest against injustices can be and have been, in other places, framed by reference to Islam (Piscatori 1986: 30).17

<sup>17</sup> Perhaps equally aware of the uses of mass media, Uyghur exiles in the Central Asian states called in 1992 for a resumption of Uyghur language radio broadcasts to China, the

In this way, processes of Islamic revival—like ethnicity—may entail *reflexive* as well as socio-psychological and political dimensions in the global age.18 Gladney (2004: 280) has attributed the current depth of Islamic knowledge among urban Hui (compared with rural Hui) to increased educational levels and greater media exposure. Uyghurs too have been increasingly exposed to mass media—national, international and illegal during the reform era. To give an example of the obvious subversive potential of illegal media, in the mid-nineties, Radio Free Asia regularly broadcast news reports to Xinjiang via its Uyghur language programmes, which are rumoured to have encouraged Uyghurs to believe that they could achieve independence.19 According to Uyghur university students interviewed in the regional capital in 2002, the authorities had responded in the late nineties by going door to door and confiscating radios from rural homes in the south. They had apparently been unable to successfully block the airwaves.20

However, media reports need not be directly subversive to exert an influence; the emancipatory effects of exposure to international news broadcast through official channels is clearly illustrated in the following interview with Zunun, service industry worker in his thirties, conducted in 2004:

Before, people were less educated, less informed, and they feared the government. But now their educational and cultural level has been raised. They read what America has been saying in the newspapers and see it on the TV […] America attacks China's human rights record; China declares its rejection of America's stance. And so we understand about human rights and no longer feel frightened. We say 'We are Uyghurs and we are Muslims.'

In this way, discussion of concepts such as freedom of speech and freedom of religion in the news media is providing a reference point against which Uyghurs can compare their own situation.21 As Lull (cited in Kim 2005: 447)

last having been broadcast in 1979. Meanwhile, in 1999 the Chinese authorities despatched truckloads of television and radio equipment to Xinjiang in a bid to ensure that party propaganda reached even the remotest outposts of the region. See Dillon (2004: 43, 67).

<sup>18</sup> On reflexive aspects of ethnicity, see Roosens (cited in Eriksen 1996: 7).

<sup>19</sup> Radio Free Asia describes itself as 'a private, nonprofit corporation that broadcasts news and information in nine native Asian languages to listeners who do not have access to full and free news media'. Its stated purpose is to provide a forum for a variety of opinions and voices from within these Asian countries.

<sup>20</sup> Deputy regional CCP secretary Keyum Bawudun had called for a crackdown in May 1998 on 'illegal radio and television stations and networks' (Dillon 2004: 121).

<sup>21</sup> Ömär, high school student, eighteen, described in 2004 a similar dynamic whereby people were buying and reading a state publication entitled '100 Questions and Answers on Human Rights' (available from the Xinhua bookstore), designed to refute Western

#### 254 chapter five

has suggested, globalised diffusion of television programming may thus spark critical reflexivity in television audiences. Or, put another way, knowledge gained through global television may provide viewers with the 'tools for self-understanding' (Kim 2005: 449). This type of critical reflexivity is increasingly understood to be a central tendency among contemporary subjects under the conditions of global modernity, with the media serving as nucleus (ibid.: 448).22 Mass media may be still more influential when combined with powerful visual images; it is especially through the daily consumption of international news reports and often shocking accompanying images that horizons have been broadened for the contemporary Uyghurs, leading to a series of wider Islamic identifications.

At present, Islam is increasingly assuming an important function in Uyghur social and cultural life and individual psychology, as well as in the political consciousness of some individuals. It should be stressed, however, that the return to orthodox Islamic practice and the more or less conscious employment of Islam as a means of alternative ethnic representation or symbolic protest is taking a non-violent form. While re-introducing Islam into the picture, I eschew the automatic confusion of Islam with Islamism (understood as militant fundamentalism), a tendency perpetuated in Chinese state discourse, the press, and sometimes also in the academic literature. To give some brief examples: Justin Rudelson, writing on the Beijing bus bomb of February 1997, states: '*Muslim terror* had struck China's leadership in their own nest' (1999: 190); Yueyao Zhao, in his otherwise insightful analysis of regional development policies, characterises Xinjiang as an ethnic frontier region 'whose ethno-religious linkage with the outside world acts like a double–edged sword', holding the potential to bring economic prosperity but also *Islamic militancy* (2001: 220-1)23; Michael Dillon

allegations of human rights abuses in China. From this, they were able to glean all the basic theories underpinning Western notions of freedom of speech, association and religion. In this way, the state itself provided Uyghurs with the means of reflecting negatively on their daily experience. Rudelson provides another amusing example: in 1990, the Chinese government distributed a pamphlet entitled 'The One Hundred Mistakes of Turghun Almas' *Uyghurlar*', which aimed to discredit this banned, alternative history of the Uyghurs. The result was of course the opposite to that intended; many more people were exposed to Almas' arguments and most found the account very compelling (Ben-Adam 1999: 208).

<sup>22</sup> Kim (2005: 448) suggests that critical reflexivity induced by increased 'trans-cultural interaction' through the medium of imported movies has led young Korean women to pursue feminine emancipation, with divorce rates in South Korea rocketing from 5.8 per cent in 1980 to an astonishing 33 per cent in 2000.

<sup>23</sup> In seeming to link ethnic separatism unambiguously to religion, Zhao gives the impression that pro-independence activities have been motivated by political Islam—

(2004: 5-6) claims that 'Islamist movements' have flourished in south Xinjiang, but fails even to define 'Islamism';24 and Colin Mackerras, while rejecting Raphael Israeli's view that Islamic renewal in China may constitute 'the first omens of an Islamic fundamentalist wave', appears to assume a direct connection between Islamic identity and radicalism, stating: '… there appears very little doubt that Islamic radicalism, *and consequently Islamic identity*, increased in strength during the period between the 1990s and 2001 […] and that it was one of the factors responsible for a series of violent incidents' (Mackerras 2005: 7-8, my emphasis). While extreme Islamist views may have pushed a small number of individuals to violence during the 1990s, I contend that it is problematic to equate Islamic identity or Islam *per se* with radicalism or to suggest that one necessarily leads to the other. Whether owing to the repressive 'Strike Hard' campaign, the PLA military presence in the region, a historical tendency towards non-radical, traditional Islam or a generally docile, 'sheep-like' national character, Islam among the Uyghurs—as an alternative form of representation—is currently taking the form of peaceful, symbolic opposition.25

My interviews, conducted in 2002 and 2004, uncovered next to no evidence of militant Islamist activism among the Uyghurs. Only one respondent, Adil, a 45 year-old male, suggested that what he termed the 'small Islamic wars' then going on in Tajikistan, Uzbekistan and other parts of the Muslim world were influencing the process of Islamic renewal in Xinjiang. There existed some romanticism—often among young women—of the notion of armed resistance or of Uyghur fugitives on the run from the state. Patimä, a service industry worker in her thirties from Ghulja, proudly pointed out in 2002 a house where she claimed *taliplar* lived (the word simply means 'religious students' but she used it in a romantic sense to denote radical separatists). Yet almost all Uyghurs—including those who romanticised separatism—vehemently condemned the use of violence against civilians—a method that neo-fundamentalists are known to

including at the international level—throughout the region's history. In fact, secular ideologies have played an equally important role, while militant Islamist ideologies have so far barely penetrated the region. He also implies that religio-cultural exchange will automatically result in the rise of militant Islamic fundamentalism, whereas it may equally inspire an innocuous trend of Islamic renaissance.

<sup>24</sup> Dillon's information derives entirely from Chinese official and scholarly sources, whose impartiality must be questioned.

<sup>25</sup> As Piscatori points out, 'the wretched of the earth' rarely have enough energy to mount revolutions, although it is true that they may often turn to religion even if they do not take to arms (1986: 34-5).

#### 256 chapter five

espouse. Dilbär, eighteen, Uyghur-educated high school student, told me in 2004 that she had been proud to see Uyghur fighters being trained in Afghanistan on national television and remarked defiantly that 'where one Islamic fighter is suppressed, two more will spring up'. At the same time, however, she stressed that while she believed in the aim of setting up an independent East Turkestan (which she referred to using the blanket term 'Islamic state', in stark contrast with Uyghur exile rhetoric, which distances itself from political Islam),26 she could not condone acts such as the bus bombs of 1997. Indeed, she insisted that most Uyghurs believed those bombs had been planted by Han or Hui 'rebels'.

Meanwhile, current Uyghur understandings of the term 'Islamic state' are varied and vague, often conflicting with one another, and ranging from secular Islamic government (as in Turkey) to theocratic state (as in Iran). Views differ greatly as to whether such a state is even desirable. Consider the following excerpt from a 2004 interview with Ömär, eighteen, Uyghureducated high school student:

Some people now treat religion [Islam] and becoming independent as inseparable [Uy: *din häm mustäqil bolush birni qildi*]. By that, I don't mean that Islam is seen as an opposition force, as a method of [direct] resistance; how can we resist the Chinese state? Our power is very small. What I mean is that some people link religion and an independent state in the sense that, in the new state, we would be able to practise our religion freely […] We could pray when we want, wear beards and veils if we want. But others say 'Don't go there!' They're afraid of what an independent Islamic state would mean. They think they would be forced to wear veils and not allowed to work, that they would sit at home all day. But we persuade them using the example of Turkey. We don't know yet whether the state and Islam would be one, or whether they would operate separately as in modern, secular Turkey, but we persuade the women by citing the Turkish example.

Ömär firmly rejected the notion of employing Islamist violence to overthrow the state. However, he indicated that increasing numbers of people are becoming attracted to the *abstract* notion of an 'Islamic state', in the sense that they believe it to be a place where they could practise religion freely, in contrast to current experience within the Chinese polity. Mean-

<sup>26</sup> Prudence has led most exiled Uyghurs to mute the Islamic component of their identity so as not to excite Euro-American fears. This is particularly the case since the events of 9/11. Interestingly, however, Fuller and Lipman (2004: 336; 340) interviewed Uyghur nationalists in the diaspora who praised Islamist activists as 'allies in the struggle against Chinese rule'. This suggests that some sympathy may now be developing among secular intellectuals for moderate Islamist methods.

while, his admission that some people have reservations about linking independence directly to Islam highlights the conflicting feelings among different groups of Uyghurs towards the creation of a theocratic state. Certainly, the implications of setting up such a state—with its concomitant attitudes towards women's rights and the status of non-Muslims—make it harder to rally some women and some nominal or partially observant Muslims around the aim of independence.27

Only four respondents mentioned violent methods in interview, three of these educated, observant Muslims and one a petty trader. All were from south Xinjiang, suggesting a possible link between extreme poverty, religious repression and a growth in religious militancy.28 But at the time of interview none had been involved in any kind of violent resistance; their beliefs remained at the theoretical level. Abdukerim, graduate in his twenties, and his friend from rural Aqsu had concluded by 2002 that 'independence is not going to be achieved through negotiation, only through war.' Abdukerim in particular talked wildly about teaching peasants how to use guns. However, their remarks seemed to me to amount to little more than bravado. Like telling stories of resistance, their boasting served rather to rescue injured national pride. Abdullah, a graduate in theological studies speaking in 2004, claimed to have spent six months in Qabul in 1992 training with the Taliban, learning how to use guns and become a *komandir* [military commander]. At one point during our conversation, he actually produced a brass bullet from his pocket, and announced: 'I have a gun now. The Chinese don't know about it!' Yet given that this episode occurred in broad daylight before Ürümchi's Nanmen mosque, one has to ask whether a person truly involved in clandestine political resistance would risk discovery in this way. When I related the episode to Uyghur friends, many suggested that this person was a government plant, someone strategically placed before the Nanmen mosque with instructions to give a certain impression of Uyghurs to nosey foreigners who asked questions. That this person was staying in a hotel that commanded a fee of 300 yuan per night (according to him, 'in hiding from the authorities') and was in possession

<sup>27</sup> The establishment of Islamic republics in Pakistan and Iran has had a negative effect on certain aspects of public life, in particular the freedom of the press, the status and rights of women and non-Muslim minorities, and the exercise of political opposition (Esposito 1987: 19).

<sup>28</sup> Dillon (2004: 168) may be correct when he suggests that separatism is especially attractive to the poorest sections of the rural population, and that Islam is becoming increasingly politicised and a potential conduit for separatist sentiment. Evidently, however, this course of action will be attractive only to individuals of a certain disposition.

### 258 chapter five

of a room key for that hotel, rendered such an explanation quite plausible. Only Tahirjan, a trader in his thirties from Aqsu interviewed in 2004, expressed a view on terrorist methods that resembled neo-fundamentalist discourse in its attitude towards civilian casualties. He told me:

With all terrorist acts, there is a necessary sacrifice to be made; there's no way around that. It's like this: imagine I have planted a field of melons. I check them every day to see if they are ripe. I eat the ripe ones and leave the unripe ones to ripen. That's impossible; what actually happens is I harvest them all at once and the unripe ones are lost. A small loss for a bigger gain. It's the same if I say I'm going to strap a bomb to myself. OK, the bomb goes off and I kill a handful of people close by, but I also must die myself. It's the same logic.

I asked him whether he considered the Muslims in the Twin Towers in New York to be a 'necessary sacrifice' in the same way, and he answered in the affirmative. He refuted the suggestion, however, that child victims in the Iraq war were also an unavoidable sacrifice for a greater gain, arguing that the Afghanistan and Iraq wars were not just wars. Tahirjan had clearly thought carefully about the controversial issues surrounding terrorist methods. Yet the fact remains that he was a hard-working, honest migrant worker who had not to date been involved in violence himself.

In the following sections, I will show that Islam has in recent years assumed a number of important roles in Xinjiang, and that some—though by no means all—of these may be seen to constitute Uyghur opposition to the Chinese state and/or its policies at the symbolic level. The current return to orthodox Islamic practice embodies (in no particular order of importance):


In the first case, through their alternative allegiance to Islam (rather than to the state) individuals are demonstrating symbolic resistance against colonial oppression at home; for some, this is further extended to include oppression of Muslims worldwide. In the second case, individuals are expressing their growing discontent at widening socio-economic inequalities

between Uyghurs and Han migrants in the wake of successive regional development campaigns (see the Introduction to this volume). In the third, they are choosing Islam as a 'pure' alternative to the cultural impurity imported by secular education in particular and Han modernity in general. In the fourth, they are seeking psychological consolation and catharsis amid the peace and social cohesion of the mosque community, with religion defined as a retreat and 'an area of certainty' separate from the world of changing politics (cf. Feuchtwang 2000: 172); at the same time, their patronage of this exclusively Islamic space underlines a continuing desire for spatial segregation from Han Chinese, who increasingly encroach on formerly exclusive Uyghur spaces (e.g. the Döngköwrük district). The fifth case embodies a search for personal and, by extension, group (national) salvation through the vehicle of Islam. It includes both those individuals who believe religious laxity to be at the root of the Uyghurs' national demise and those whose primary motivation may be fear or superstition.

For some re-Islamising Uyghurs, all the above factors are playing a part; for others, some factors are important while others are less so; for still others, different factors may be fore-grounded at different times. Individuals also display differing levels of awareness towards the process, with some respondents lucid and analytical about the catalysts, while others remain more or less unconscious of the motivations for their behaviour. This latter point was highlighted by Tashmämät in 2004, an intellectual in his forties originally from Qäshqär:

For many people going to the mosque now, it is a superficial manifestation. These middle school kids you mention who are going to the mosque on Fridays—for them, it's not a deep commitment or understanding. They are in fact expressing their discontent [*naraziliq bildürmäk*] with the current state of society. Perhaps 'opposition' [*qarishi qilmaq*] is too strong a word and certainly many won't realise that what they're doing is a kind of symbolic opposition. But it *is* a kind of resistance insofar as it's an expression of discontent with government policy and with the Hans. Such expressions of discontent can take many forms … at present they are taking a religious form.

His description suggests 'a certain artlessness, almost ingenuousness' in local experiences of Islamic renewal, as also documented for other Muslim countries which underwent revival (Piscatori 1986: 37).

#### 260 chapter five

# 5.3. *Islam as Symbolic Resistance against Perceived Oppression of Muslims*

I will first consider Islam as a vehicle for symbolic opposition to oppression of Muslims, at the national level (with the Chinese polity viewed as coloniser), but also at the global level (with individuals expressing empathy with similarly 'oppressed' Muslims in the Middle East and beyond). The active role played by Islam in many of the anti-colonial independence movements of the twentieth century is well known (see Esposito 1991: 60- 95). Among examples of such movements, the case of North Africa seems to be quite relevant to contemporary Xinjiang. French colonial policies in North Africa often exemplified the most extreme form of European imperialism in the Islamic world: 'Through a concerted and sustained program, total political and cultural assimilation was attempted and promoted under the French policy of "naturalization" […] French was imposed as the official language and the language of instruction […] Arabic was reduced to a foreign status' (Esposito 1991: 75-76). In schools in Algeria, children studied the French cultural heritage. This policy of total assimilation to all things French caused young, educated Muslims to reassert their own, indigenous identity, beginning with the Arab-Islamic heritage (ibid.: 76). Thus, the principal aim of Algerian Salafist and patriotic reformer, Ibn Badis, was 'to counter French persistent endeavours to deny the mere existence of Algeria as a political and national entity prior to its occupation' (Choueiri 1997: 28). In Egypt, resentment towards continued European hegemony fuelled nationalist fervour and similarly provided a common goal for opposition movements otherwise divided by class, region, tribe, or degree of religious (versus secular) commitment (Esposito 1991: 61). In Tunisia, nationalists 'used the fear of loss of identity to assert their role as the defenders of Islamic identity and culture' (Esposito 1991: 81).

For some individuals in contemporary Xinjiang, Islam may similarly be conceived in one sense as a reaction to increased political and cultural oppression. A clear illustration of this stance would be the demonstration prior to the disturbances in Ghulja in February 1997, during which protesters carried placards proclaiming 'We have one God, not two'.29 In post-1997 Xinjiang, the Chinese state moved quickly to consolidate its programme of internal colonisation. The acceleration via the Western Development Campaign of China's policy of 'domestication' pursued in the 1990s has led to further Han in-migration as the state attempts to stabilise the region by

<sup>29</sup> Interview in 2002 with Qurban, unemployed male in his twenties from Ghulja.

altering the population composition.30 This ill-advised policy is being executed in an already explosive climate of socio-economic competition between Han Chinese and minorities, and environmental degradation. Meanwhile, Uyghur-medium tuition, outlawed since 2002 in Xinjiang's universities (the exceptions are Uyghur language and literature departments), is now increasingly rare in secondary and primary schools as well (see Dwyer 2005), despite constitutional guarantees of linguistic freedom for minority nationalities.31 The trend towards Mandarin-only tuition, combined with the *de facto* institutionalisation of the Chinese language in education and employment since the mid-nineties means that Uyghur parents are increasingly forced to place their children in Chinese-medium classes in order to ensure their competitiveness and indeed their survival in the urban milieu (Chapter 5 deals with this emerging group of Chineseeducated Uyghurs, the *minkaohan*). There they are taught in the Chinese (rather than the Uyghur) language, learn about the Chinese (rather than the Uyghur) cultural heritage, and absorb the Chinese (rather than the Uyghur) version of history. A majority of Uyghurs believes that these policies surrounding Han migration and Uyghur language and religion are intended to result in cultural genocide.32

The recent acceleration of sinicising ('domesticating') economic and cultural policies aside, it is significant that religion did not initially resurge—or at least not to this degree—under these conditions alone; nor did it feature prominently in rising Uyghur nationalism in the mid-late nineties.33 Rather, I would suggest that the heightened repression of religion since the mid-nineties has particularly invited this religiously oriented response (although the increased censorship of other forms of expression—see earlier—may also have played a part). This would appear to fulfil Voll's earlier (1987) prediction that repression of religion and cultural practices could, against all odds, result in a politicised form of Islam

<sup>30</sup> The term 'domestication' was coined by Becquelin (2004c).

<sup>31</sup> Article 4 of the Constitution of the PRC reads: 'All nationalities have the freedom to use and develop their own spoken and written languages and to preserve or reform their own folkways and customs' (Legislative Affairs Commission of the Standing Committee of the National People's Congress 1994: 13).

<sup>32</sup> cf. Fuller and Lipman (2004: 324): '… this demographic shift appeared to almost every Uyghur interviewed […] to be one of the two greatest threats to their culture and society, the other being state control over religion'.

<sup>33</sup> Fuller and Lipman's study, conducted in the late nineties, concluded that the phenomenon of Islam reinforcing nationalist movements by 'investing secular nationalism with religious and emotional content of a universal nature' had so far scarcely been manifested in Xinjiang (2004: 340).

#### 262 chapter five

in China. Exactly this dynamic has already been witnessed in Chechnya and Daghestan in the former Soviet Union. Although Soviet policies such as creating an official Islamic administration (viewed as 'stooges of the godless authorities') and deploying secret police to control believers were designed to destroy the hold of Islam over society, the end result was conversely to ensure the endurance of Islam and turn it into a political tool. As cogently argued by Moshe Gammer, in a natural backlash to the Soviet socialist state and code of law, the Islamic alternatives of *imama* and *shari'a* were 'instinctive antitheses' (2005b: 833-834, 842).

Many Uyghur respondents scorned the government's attempts to limit religious activity in Xinjiang, and proclaimed its efforts doomed to failure. Ömär, a high school student speaking in 2004, confirmed the renewed centrality of the mosque in Uyghur public life and firmly rejected the notion that the state could control Islam: 'The mosques are a public place, right? They can't get rid of them […] They may try to control religion, but they can't, not even with their [state-trained] imams … at the end of the day, the imams are still Uyghurs.' Compare this with the positive attitude of the Jadids towards the state Islamic establishment as mediator in Soviet Central Asia: 'Soviet Moslem sermons pick up [the] general theme of Islamicsocialist compatibility by finding predictions of the Soviet state in the Koran, by referring to the building of communism as "the great earthly ideal of the Prophet Mohammed"' (Voll 1987: 149). By contrast, respondents in Xinjiang displayed a keen awareness of state manipulation of the attitudes of imams and the content of their sermons. While more radicalised individuals rejected these state imams as 'sell-outs',34 others suggested that the state should not assume the loyalty of state imams one hundred per cent, implying that individuals may be placed at different intervals on a scale stretching from full loyalty to the ethnic group to full loyalty to the state. The remark may also have referred to imams of some smaller mosques in the south, who have been selected not by the state but by the local community (cf. Fuller and Lipman 2004: 334). Qurban, eighteen, an unemployed Uyghur-educated male in Ghulja, insisted in 2002 that state repression would never succeed in cowing Uyghurs for in the end they would always have religion: 'The Hans think that if they hit [repress] us,

<sup>34</sup> Such criticisms are not new among China's Muslims; 'progressive' imams were labelled 'men without a religion' by fellow Hui during the Hundred Flowers period, while the Party's co-optation of religious personnel was condemned as an attempt to 'rule the Hui people by the Hui people' and to 'destroy the foundation of Islam' (Bush 1970: 283; 286).

then this will make us afraid! But it doesn't work. We are not afraid of them. We are only afraid of Allah!' It is perhaps precisely this fearlessness that the state so dreads and expects to translate into militancy and violence.

While early interviews strongly supported my hypothesis that Islamic renewal in Xinjiang was to a large extent a reactive phenomenon, it was noticeable that not all respondents used the words 'resistance' or 'opposition' to describe it. Piscatori, in asking people throughout the Islamic world why they had become more devout, was invariably told that 'it had happened almost imperceptibly' (1986: 36-7); respondents cited none of the social, economic or political dislocations that he later came to identify as central factors underpinning the phenomenon. In Xinjiang, too, I found differing levels of awareness among respondents. Dilbär, eighteen, female Uyghur-educated high school student, mused in 2004:

Is religious renewal an act of resistance against China? Probably [smiles] … insofar as 16 year-old Uyghurs go in the mosque even though the law states that we must be 18, I think you can call it an act of resistance. The Chinese say that children should first learn science and technology at school, then decide what they want to do about religion when they are adults. They say under-18s are not mature enough to decide. We don't agree, we think that's wrong. If we want to go to the mosque, we go [Uy. *kiriwerimiz*]! That much may be characterised as resistance.

In this way, Dilbär could only conceive of resistance in terms of breaking the law, and did not consider the wider phenomenon of Islamic renewal to be oppositional. She pointed out that if Uyghurs wanted to resist the Chinese, there were other ways, giving the example of the 'East Turkestan advocates' [Uy. *Shärqiy Türkistan qormaqchilar*, Ch. *Dongtu*].35 She admitted that her heavily romantic image of armed resistance had been absorbed solely from national and regional newspapers and recent Chinese publications on *Dongtu* such as Liu and Du (2003), available from the Xinhua bookstore. Thus, the Chinese government's exaggeration of *Dongtu*, far from demonising the notion of armed resistance, seems instead to have romanticised it in the public imagination. The point I wish to make, however, is that Dilbär was only able to conceptualise 'resistance' as direct, illegal and possibly violent action. This explained why respondents at first found it hard to conceive of Islam as a non-violent, symbolic form of resistance. Ömär, eighteen, male Uyghur-educated high school student speaking in

<sup>35</sup> In recent Chinese scholarly and official state literature, the term *Dongtu* has invariably been conflated with the notion of 'Islamic terrorist'.

#### 264 chapter five

the same year, was at first similarly wary of describing the Islamic renewal in such terms:

Of course, we have to resist/oppose in order to achieve independence, that much is self-evident. But, no, I don't think going to the mosque is equal to resistance. Mosques are public places, not private; we're free to go to them if we want. A symbol, you say? Oh, I see, not *actual* resistance, but a symbol of opposition? Yes, then I suppose it could be described in that way.

In this way, it was frequently necessary to distinguish between direct and symbolic resistance, with the Uyghur verb *qarishi qilmaq* [to resist, oppose] evoking notions of actual struggle, war, demonstration and secret, illegal activity for most lay respondents.

Perhaps not surprisingly, some respondents from the older generation flatly denied that the strengthening of Islam had anything to do with increased repression. Very elderly Uyghurs had probably never *stopped* being religious; some may also have been reluctant to make such a statement publicly for fear of political reprisal.36 Other respondents found it hard to link religious renewal to repression, given the state's recent policy of comparative leniency in the regional capital (I will return to this later). For example, Patigül, a female music trader in her twenties, protested in 2004 that on the contrary the government actually encouraged religion in Ürümchi. Still other respondents initially suggested innocuous reasons for the phenomenon, but later conceded that it involved a reactive dynamic. Tömür, a teacher in his fifties from Turpan, explained in the same year:

We'll all be underground one day, we'll all turn to dust. And then we will all be judged by Allah. So we go to the mosque, try to earn merit by doing good deeds [*sawab elish*]. People didn't worry about that years ago, you say? Well, Islam, religion, is a major component of Uyghur customs and habits, and if you try to squash or eliminate a people's cultural practices, those practices will just rebound and mushroom all the more … .

When I subsequently suggested that while the current Islamic renewal may have been triggered by state repression of religion, it was also intimately related to worsening political, cultural and socio-economic conditions in Xinjiang over the past decade, he and his friend (who had been listening intently) nodded emphatically and cried: 'Exactly right! [*Äppällä!*]' On the other hand, some respondents immediately made the connection to state

<sup>36</sup> Older persons in Xinjiang tend generally to be far more cautious than the younger generations, and will often avoid talking about politics or other 'dangerous' topics (see Smith 2000).

religious repression., Tashmämät, intellectual in his forties from Qäshqär interviewed in 2004, situated the current trend squarely within a repetitive cycle:

There is a historical pattern, a cycle of religious repression being followed by religious resurgence in Xinjiang. In the sixties and seventies—the Cultural Revolution period—religion was totally repressed, but it didn't disappear. After 1978—and Deng's conciliatory policies towards minorities—religion resurged once more. The current situation is a repetition of that cycle.37

A male restaurateur in his forties, originally from Qäshqär, explained in the same year: 'I suppose Islam is stronger now because the government is trying to block religious activity. If the government tries to block or limit Islam [*hökumät tosisa* …], then people's faith only becomes stronger.' Ömär, Uyghur-educated high school student, eighteen, elaborated on this idea:

The reason why 1997 [the Ghulja disturbances] happened and why Islam has grown further since then is because the government decided somewhere around 1995-6 that all things related to Islam—religious activities in the mosque, Islamic schools—were related to separatist politics, and that Uyghurs were meeting in those places to plot with one another. And so they clamped down [Uy. *ching tutush*]. The laws forbidding under-18s to go into mosques and so on were previously in place; the difference is that they weren't enforced. Now they began to enforce them.38

In this way, 'what might not be resistance on its own, becomes potential resistance when it is categorized as illegal' (Feuchtwang 2000: 172). His female cousin Aynur, a Uyghur-educated graduate in her early twenties, further pointed out that numbers going into the mosque had particularly increased following the 1997 Ghulja disturbances, and again following the events of 9/11 and their aftermath. This suggests that the current Islamic response is at least in part a reaction to increased religious repression, initially on a national scale, and more recently on a global scale.

# 5.4. *Islam as a Response to Failed Development*

On one level, then, Islamic renewal in Xinjiang can be understood as a form of symbolic resistance against perceived Chinese oppression on a national

<sup>37</sup> For a general analysis of the cyclical nature of Islamic revivals and causal patterns of religious resurgence throughout the religion's history, see Dekmejian (1995: 8-22).

<sup>38</sup> A sign in Uyghur posted over the door of every large or small mosque forbids entrance to minors (Fuller and Lipman 2004: 335).

#### 266 chapter five

scale and perceived anti-Muslim oppression on a global scale. On another level, it may be conceived particularly as a response to failed development, or the failure of China's economic policies to benefit all ethnic groups in the XUAR equally. It is generally agreed among scholars of the Middle East that the basis of the Islamic response in most Arab countries has been a combination of socio-economic and cultural/psychological factors. Ayubi (1991: 118-119) identifies a general pattern such that when development faltered and promised rewards were not achieved, segments of civil society rebelled against the state using religion as a catalyst for mobilisation and resistance. Since an alien cultural system (Western liberalism) had usually been imposed at the same time, this resistance often took the form of retraditionalisation or a return to Islam. Writing on the transition to Islamic militancy, Esposito (1991: 176) describes the specific conditions needed to propel people into becoming activists in terms of frustrated expectations. He notes that most Arab countries went through development processes in education, urbanisation and industrialisation, which raised the hopes and aspirations of many sectors of the population. When governments failed to deliver, it was the youth who could not get jobs or buy houses, and their 'heightened, then shattered, expectations' that caused them to form the mainstay of militant Islamic movements. Ayubi (1991: 176) further colours this frustration as a sense of 'relative deprivation' resulting from an incomplete process of modernisation, arguing that it is not so much modernity itself which lies at the heart of disaffection as unequal access to that modernity. Or, as he puts it: 'The Islamists are not angry because the aeroplane has replaced the camel; they are angry because they could not get on to the aeroplane' (Ayubi 1991: 177).

The notion of 'relative deprivation' is easily applied to the Chinese context. As in the Arab world, where the impact of global market forces widened the economic gap between social groups leading to class polarisation and conflict (Dekmejian 1995: 28), China's economic policies of the nineties and the more recent Western development campaign have so far singularly failed to deliver desired socio-economic improvements equally across social and ethnic groups. Stefan Feuchtwang has consequently characterised revitalised religions and the spaces occupied by their adherents as 'alternative arenas for telling or performing stories which reflect on the corruption, inequality and lack of security and support seen and experienced by the subjects of the Chinese People's Republic'(2000: 161).

In Xinjiang, the most visible gap is between ethnic groups, namely, the Han Chinese and the minorities, especially Uyghurs. My interview data suggest strongly that many have lost faith in the ability—and indeed the intention—of the Chinese polity to deliver economic development to Uyghurs, especially to *minkaomin* or those who have learned Uyghur (rather than Chinese) as their first language.39 At the same time, a majority believes that 'sinicising' policies towards Uyghur culture (especially language and religion) are intended to result in cultural genocide. In this context of political anti-climax and impotence, economic disappointment and cultural alienation, some respondents characterised Islam as the only thing left for them to believe in. In 2004, Möminjan, a young man from Turpan studying in Beijing, attributed the return to Islam to a steady deterioration in political, cultural and socio-economic conditions for Uyghurs. He claimed that ten or more years ago, things had actually been really good for the Uyghurs [Uy. *häqiqätän yakhshi idi*], despite the vocal disaffection that had characterised the mid-1990s. In comparison, he proclaimed the current situation 'very bad indeed'. Two male music sellers in their forties interviewed in 2004 in Ürümchi linked the return to the mosque and orthodox Islamic practice to the failure of the Chinese state to furnish promised opportunities: 'They [re-Islamising Uyghurs] think "Nothing good, no goodwill has ever come to us from the Hans" and so they're praying to Allah instead.' In this way, the disenfranchised, seeing a prosperous future recede on the horizon, 'naturally cling to the only comfort of the present, the pillar of traditional faith, and, in doing so, would give political expression to the frustration and indignities of living on the margin …' (Piscatori 1986: 34-35). Another male music seller in his thirties explained in the same year:

In the Qur'an, it says that we Muslims should fight back if persecuted […] There are jobs, opportunities for Hans, but none for Uyghurs. If you happen to be the son or daughter of a rich Uyghur, you know, a politician or manager of some kind, then money can buy you a job. Otherwise, you get nothing.

Asked why Uyghurs were turning back to Islam, Adil, a 45 year-old male from Ürümchi, replied in 2004 that Uyghurs were 'believers', that they 'had faith' like believers of other monotheistic religions. Yet when pressed as to

<sup>39</sup> These also tend to be the more observant Uyghurs. Fuller and Lipman have suggested that religious Muslims in particular 'do not see themselves as benefiting from the external investment, enhanced domestic spending, and resource development programmes that the central government has initiated …' (2004: 325).

#### 268 chapter five

why Islam's popularity should have suddenly grown in recent years, he observed: 'Well, we Uyghurs are all alone, aren't we?' During a follow-up interview, he expanded on this to place the blame directly on socio-economic deprivation: 'Why are they [re-Islamising Uyghurs] going to the mosque? Because even if they study really hard, there's nothing for them when they come out of school or university … they have no seat in society [*orun yoq*]. All they have left is Allah, and so they place all their hopes in him.'40

Yet Adil's friend (a young male student) condemned this course of action, arguing that while those who turn to religion may have Allah and their prayers, without education or a job they cannot eat. Ghäyrät and Tahirjan, *minkaomin* traders in their thirties originally from Aqsu, expressed similarly dichotomous views toward education in 2004. Ghäyrät enthused: 'With education, you can do anything, build anything. Look at the Jewish communities scattered around the world—they're all rich and well regarded! […] You can do business well if you're educated.' Yet Tahirjan disagreed, arguing that even if Uyghurs educate themselves, it will get them nowhere in the current environment of ethnic discrimination. In this way, while one sector of the Uyghur population is abandoning the system for religion, another sector continues to believe that the only way to advance indeed to survive—is through the system, that is, through study.

Two other interesting factors link the phenomenon of Islamic renewal directly to socio-economic deprivation. Gülhärä, twenty, Uyghur-educated student and daughter of a trader, explained in 2004 that education in Xinjiang was now so expensive that many people on low incomes or without employment could not afford to send their children to state schools, and so had them study religion privately. The inability of growing numbers of Uyghur families to pay escalating education fees may thus be helping to bring about re-Islamisation by default.41 Secondly, the scarcity of employment for Uyghurs, created by a high level of redundancies in state-run enterprises combined with uncontrolled ethnic discrimination across the

<sup>40</sup> Compare Anderson, who cites a young Algerian Islamist as follows: 'You can remain unemployed and celibate because there are no jobs and no apartments to live in; you can work in the black market and risk being arrested; you can try to emigrate to France to sweep the streets of Paris or Marseilles; or you can join the FIS [Front Islamique du Salut] and vote for Islam' (1997: 24-5).

<sup>41</sup> Ben-Adam (1999: 202-03) similarly writes that the cost of secular education, together with the scrapping of the policy of guaranteed placement for university graduates, means that Uyghur peasants in Turpan no longer consider secular education a viable option for their children; this has in turn raised the prestige of Islamic education.

job market, has led to a return to independent business. This is true particularly among Uyghur-educated or uneducated Uyghurs, whose Chinese language abilities are often more limited, but also increasingly among university graduates, who leave college only to find that, despite their acquired Chinese language skills, their ethnicity obstructs employment in state work units or private Han-run companies. This 'return to business' has in turn brought about a return to Islam, firstly because businessmen [*sodigärlär; tijarätchilär*] are free to go to the mosque while work unit and company employees are forbidden to by law;42 and secondly because businessmen traditionally make a point of going to the mosque to pray for good business. Patigül, a female music trader in her late twenties, explained in 2004:

When there are redundancies, it's always the Uyghur men who get the push because they don't know Chinese. They [Han employers] say 'You didn't learn it well.' So more men are doing business and the *sodigärlär* tend to go in the mosque because they can. They pray that their businesses will go well, that their trade is healthy … they ask for Allah's help.

Zunun, service industry worker in his thirties speaking in the same year, further explained: 'We Uyghurs say that, to do business, we must be pure [*pak*], otherwise the business will not go well. And so the *sodigärlär* go to the mosque and pray to Allah for business success. And the women traders do the same, only not in the mosque.'43 Some respondents went so far as to suggest that only the businessmen went in the mosque (not other social groups), but this was not supported by other interview data or observation, and probably derived more from the sense of distrust that exists among some independent entrepreneurs towards Uyghurs attached to state work units or Han companies.

<sup>42</sup> The Chinese authorities issued a statement in May 1996 threatening to punish indigenous party cadres who persisted in taking part in religious activities (Dillon 2004: 86). In the post-1997 period, this rule has been extended to anyone on the state payroll (except for state imams) and to school pupils and students (cf. Fuller and Lipman 2004: 324).

<sup>43</sup> There is a long-standing connection between trade/craft guilds and the Sufi brotherhoods in the region. Following its foundation in Central Asia during the 1400s, the Naqshbandiya brotherhood became 'a major factor on the popular level among the craftsmen of the cities' (Bennigsen and Bryan 1989: 248). There seems to have been a similar connection between Uyghur craft guilds and the Sufi brotherhoods in pre-socialist Xinjiang, with terminologies overlapping while each trade had a patron saint and a *risalä* (text of the profession), which explained the sacred origin of the trade in Islam (Bellér-Hann 2008: 325). Some tombs of craftsmen are today sites of pilgrimage (Dawut 2007: 151). In the context of the wider Islamic world, Choueiri describes the intertwining of Sufi orders and artisan guilds during the first phase of modern Islamic history (1500-1770) as 'entrenched means of spiritual and material endeavours' (1997: 2).

#### 270 chapter five

In a minority of cases, the decision to do business rather than seek employment within a work unit or Han company is deliberate and political. Abdukerim, graduate in his twenties from Aqsu, explained in 2002:

I don't want to work in a work unit [Uy. *idarä*] now. They wanted to send me to Xinjiang University to be a teacher, but I refused. If you work in an *idarä*, they control you, they have leverage to ask you to spy for them, translate things for them … they will use your language skills against your own people! If you do business, you're free; they have no hold on you.

He admitted that most Uyghurs working in business do so only because they failed to get a job in a work unit, but confirmed that he knew perhaps twenty others whose thinking was as politicised as his own. Äsäd and Ablät, *minkaomin* graduates in their twenties, expressed the same motives in 2004 for going into the computer trade: 'We're not going into a work unit. It's better to be in business, there's more freedom. In business, you're your own boss, no-one tells you what to do.' Michael Dillon also confirms this trend, writing that one banner held during the peaceful protest that led to the 1997 Ghulja disturbances read 'We want nothing from the government', a feeling he believes reflects the attitude of many Uyghurs who 'prefer to run their businesses, avoid working for the state, and have as little as possible to do with what they regard as an alien and oppressive government' (2004: 97).44

A final factor leading some Uyghurs to choose to return to business is the poor wages offered by state work units. Bahargül, *minkaomin* in her twenties and manager of a street restaurant, told me in 2004:

I worked in an *idarä* for just a year, but then I came out. Because the salary was so low, the wages in work units are terrible. So I left and started my own business. I'm starting small with this dumpling shop and hope to build it up from there.

She confirmed that she knew other Uyghurs who had abandoned jobs in work units for the same reason. This case represents a deliberate decision by some Uyghurs to pursue better, money-generating opportunities elsewhere.45

<sup>44</sup> Goodman (2005: 338) reports a parallel phenomenon among new Salar industrialists in the north-west, who are taking a localist approach to economic activism, providing jobs and opportunities for the local community and donating funds to communal causes. Reasons in this case seem however to be purely entrepreneurial rather than political.

<sup>45</sup> One example of a successful, Uyghur-run business venture is the *Istanbol* supermarket in Döngköwrük, one of a number of halal grocery stores that have sprung up in Ürümchi, Ghulja and other towns in the past few years. When I visited it in 2002, it was clean,

An emerging small group of 'big businessmen' [*chong sodigärlär*]46 is meanwhile re-investing money generated by big business in the construction and improvement of religious buildings, contributing in a material sense to the Islamic renaissance in the region. Ömär, eighteen, Uyghureducated male high school student, explained in 2004:

Another reason why Islam is more prominent now is that [some] Uyghurs have more money … traders and businessmen give money to the mosques. There are two types of mosque construction going on, basically. There are the big mosques, mainly funded by the government, then there are the smaller ones springing up all over the place—those are often funded by donations. The [Ürümchi] government doesn't care as long as the new buildings conform to safety regulations and don't have too many storeys. At the end of the day, they control the show; they encourage traders to fund the construction, offer to chip in a bit of money, put their imams in place, and there you have it!

This suggests that the current re-Islamisation process in Ürümchi is supported by the re-establishment of significant amounts of private *waqf* endowments donated by the business community, and that city government is allowing this, provided that religious buildings and activities remain within legal bounds.47 These factors lead inevitably to an accelerated return to the mosque, promoted by the return to the business world, while the state's acquiescence indicates the partially negotiated nature of religion as symbolic resistance in the regional capital.

### 5.5. *Islam as an Egalitarian Ethos*

One reason why Islam is increasingly attractive to Uyghurs suffering from socio-economic inequality and discrimination is its egalitarian ethos. Marsot suggests that the only way for many Muslims to 'tame the beast' of modernisation (read capitalism) is to have 'an anchor […] an ethical foun-

bright and entirely staffed by Uyghurs. It sold many household goods and foodstuffs imported from Turkey, and products often carried labels in Turkish and English, with no Chinese translation. Those Chinese goods it did stock, were those made by Uyghurs; these carried labels in Uyghur and Chinese. The clientele, too, was almost entirely Uyghur, so that the store constituted a rare example of a Han-free space.

<sup>46</sup> This group has been described as 'a growing Uyghur middle class' by Colin Mackerras (2001: 299). At present however, few Uyghurs may be described as 'big businessmen', while the vast majority of Uyghurs involved in business are petty traders.

<sup>47</sup> Gladney (2004: 318) notes that Muslims throughout China are now building mosques on their own initiative and sending increasing numbers on the *hajj*, using private, community-based funding.

#### 272 chapter five

dation, which Islam seems to offer because it is a Weltanschauung [world view] with a promise of a just and integrated society' (1992: 169). Although in the mainstream Sunni tradition the *ulama* often taught that anyone in effective possession of political power had to be obeyed and that an unjust ruler was better than civil strife, there is at the same time a long history of Islam providing the rationale for opposition to injustice. Shiism, because of the nature of its emergence as a sect, hosts a particularly rich literature on justification for revolt and martyrdom against unjust rule,48 while within the Sunni tradition too militants have concerned themselves with the cause of equality, often invoking the ideas of Islamic figures sympathetic to the poor or oppressed (Ayubi 1991: 61-63).49 The Egyptian independence movement, described as the culminating phase of Salafism, similarly grew out of an unacceptable situation of socio-economic inequality. Proposals for economic and political reforms put forward by Hasan al-Banna and the Muslim Brethren were couched in terms of 'the achievement of social equity' and 'a guarantee of adequate opportunity for all' (Choueiri 1997: 39- 40). Mazrui (summarised in Kramer 1996: 282) sees such movements in very positive terms, suggesting that future Muslims will make history through a reliance on Islam's ethos. In his view, there are lacunae in capitalism whose worst effects can be mitigated by Islam, understood not as a religion but as an egalitarian value system.

The case of contemporary Xinjiang demonstrates that Islam, as a response to inequality, can express itself peacefully and symbolically. In Qäshqär, reformist clerics have dismissed age-based hierarchies central to local practices, and are emphasising instead the egalitarian ethic that underpins the Islamic orthodoxy: equality of all believers before God (Waite 2007: 170). My ethnographic data gathered in Ürümchi suggest that people are increasingly attracted to this ethos. Jelil, an observant graduate in his late twenties, explained in 2002:

Islam has gained strength because people know more about it now; they have more knowledge at their disposal. They understand that Islam is a real and true religion [Uy. *riyalist; häqqiqi*]. By that, I mean that everyone in

<sup>48</sup> In Shiite Islam, the Twelfth imam is believed to be in a state of Occultation, awaiting the opportune moment to return and fill the earth with justice, in place of oppression (Choueiri 1997: 14).

<sup>49</sup> Ayubi (1991: 65-66) gives the example of Hasan Hanafi, the contemporary Egyptian leftist reformist, who argues that the purpose of the Islamic sermon should be to 'call for struggle and for retrieving the rights of the poor from the rich' and 'to strengthen the wretched and to confront Zionism and imperialism' (for a brief biography, see Esposito 2003: 107).

Islam is equal before the eyes of God. People understand that Islam is a beautiful religion, that Muslims are a beautiful people.50

In particular, respondents compared Islamic egalitarianism favourably against China's socialist modernisation, whose dynamics have vastly increased the gap between rich and poor, while creating more room for ethnic and gender discrimination in the labour market. Increasing numbers of Uyghurs have begun to wax nostalgic about the Maoist era. Their nostalgia draws on the undeniable fact that Uyghurs enjoyed more socio-economic equality (if not material wealth) during much of that era than in the contemporary reform era, compared both with other Uyghurs and with other ethnic groups, including Han Chinese.51

In a further suggestion of the reflexive relationship with global media and subsequent forging of 'global solidarity' along Islamic lines, many Uyghurs have now broadened notions of socio-economic inequalities between individuals and ethnic groups to include socio-economic and political inequalities between nations. Respondents demonstrated a keen awareness of unequal power relations between different countries, particularly where they perceived a Muslim country to be oppressed by a non-Muslim one. In 2004, Tahirjan, migrant trader in his thirties from Aqsu, linked the situation of other oppressed Muslims directly to the Uyghur experience in Xinjiang, lambasting Israel's treatment of the 'defenceless Palestinians' before moving immediately to the issue of local inequalities between Uyghurs and Han Chinese, which he evidently equated with inequalities between nations: 'Doing this kind of work, Ghäyrät will earn maybe 100 *yuan* a day; if he were a manager, he would be earning at least 500 *yuan* a day. But he'll never be a manager; because Uyghurs have no equality and no freedom. Even if you study hard, graduate from university, it's impossible to rise through the ranks.' It should be noted that, by modern Chinese standards, 100 *yuan* a day is not a bad earning for a petty entrepreneur, especially when compared with the average monthly wage in the

<sup>50</sup> While generally refuting the possibility of self-representation, Gladney (2004: 266) allows that some Muslims in China are beginning to communicate pride in their tradition of Islamic learning. *Häqiqät*, a main principle of Islam, was explained by Uyghur *molla*  Abdul Qadir (circa 1930) as 'the truth which purifies the heart'. Islamic mysticism considers *häqiqät* to be the profound reality, the way to which is opened only by union with God (Bellér-Hann 2008: 341).

<sup>51</sup> Similar nostalgia for this 'era of certainty' has been recorded among Han Chinese in China proper, with respondents characterising those years as 'a time of income security, straight dealing, and simplicity' and contrasting them favourably against the self-seeking cadres of the reform era (Feuchtwang 2000: 166).

#### 274 chapter five

regional capital Ürümchi (Han: 1,141 yuan/month; Uyghurs: 892 yuan/ month) (Zang 2011: 152). The problem here is the perception of relative deprivation: the belief that a higher percentage of Han Chinese are in a position to earn better salaries than Uyghurs.

# 5.6. *Islam as a Response to Frustrated Ethno-political Aspirations*

In tandem with failed economic expectations, a second 'frustration factor' contributing to re-Islamisation in Xinjiang has been the failure of ethnopolitical aspirations—the Uyghurs' failure to gain independence from China in 1997. In this conceptualisation, I follow earlier psycho-analytical explanations of Islamic revival in the Middle East and Indonesia.52 Following the 1997 Ghulja disturbances and the ensuing state crackdown on proindependence activities, the various '1997 independence theories' circulating since the mid-1990s collapsed in what I have termed an 'ethnopolitical anti-climax'. The subsequent process of religious renewal may for some individuals be characterised as a shift away from faith in Uyghur nationalist ideologies and towards faith in Islam. Ghäyrät, Uyghur-educated migrant trader from Aqsu, explained in 2004:

About 10 years ago [1992], no-one thought about independence. But then around 1995, 1996, 1997, we began to believe it could happen […] We were convinced that China would fall out with another country; that the UK wouldn't give Hong Kong back [and that international conflict would create the necessary domestic conditions for secession]; that the UK, the US or some other country would help us secede […] Now few people still believe we can achieve independence. I don't believe it any more. Many have lost faith like me, they realise that no-one is going to help us.

<sup>52</sup> Several scholars of the Middle East argue that the Islamic resurgence since the 1970s followed closely behind the routing of the Arabs from Israel in 1967 (see Ayubi 1991: 59); Esposito argues that this led to an Arab identity crisis, a greater concern for authenticity, and a 'return to the mosque' based on 'a desire to link the present with historical and traditional values' (1987: 17-18); for Dekmejian, the defeat 'precipitated a sharp decline in Pan-Arabist sentiments and a new quest for an ideology of victory and success' (1995: 26); Piscatori (1986: 26) notes that the moment, termed *al-nakba* ('the disaster'), was the culmination of a long series of setbacks and humiliations for the Arabs. On ethnic/ regional rebellion in Aceh, the indigenous Acehnese historian Sjamsuddin argues that where the region had once operated as a single nation that extended its power into other parts of Sumatra, under Sukarno it was incorporated into East Sumatra and dominated by that region from 1950, leading to a sense of betrayal and 'mass frustration' among Acehnese Muslims (Siapno 2002: 162).

Abdukerim, observant graduate from Aqsu in his twenties interviewed in 2002, confirmed this changed outlook: 'Now we only believe in Allah! Not in anyone else, not in America or the UK.'53 Yet he said other things that suggested that he secretly continued to hope that the US would intervene in the Uyghur problem. By 2004, increased ambiguity in the US stance toward Xinjiang separatism had caused him to lose faith in any notion of outside assistance, and he had grown more militant in his attitude. He boasted about Uyghurs who had trained in Afghanistan, and compared the Uyghurs' situation of colonial oppression with that of the Chechens, a comparison frequently made in Xinjiang along with the Palestine example.54 This militancy, however, remained at the level of talk:

Before, many Uyghurs in the south believed what the World Uyghur Youth Congress said,55 but now they think it is all just empty words tumbling out of their mouths. It's the same with Radio Free Asia, that's funded by the US government, right? We really believed their broadcasts, but then the Amer-

<sup>53</sup> In 2002, I asked Jelil, a strictly observant Uyghur-educated graduate in his thirties, why Uyghurs should turn back to Islam when Allah had so far failed to deliver the social justice and the homeland they so desired. He smiled, and explained patiently: 'You ask why we believe when Allah has not given us Xinjiang: this is because Allah neither punishes nor rewards in a hurry [*aldirimaydu*]. This is written in the Qur'an.' The negative component of this notion—punishment—implies that God watches all, and that while retribution may take time, the final 'balancing of the books' or Day of Judgement will surely come. This concept is also present in many other traditions, as demonstrated in the Oxford Dictionary of Proverbs, which quotes several variations on the following saying: 'The mills of the gods are late to grind, but they grind small' (physician and philosopher Sextus Empiricus, c. 160-210 AD, *Against Professors* I.xiii.287); 'God's Mill grinds slow, but sure' (English poet and orator, George Herbert, first published 1640; see Herbert 2001, No.747); 'Though the mills of God grind slowly, Yet they grind exceeding small; Though with patience he stands waiting, With exactness grinds he all' (translation by American poet, Henry Wadsworth Longfellow, of the 17th-century German-language poem 'Retribution' by German epigrammatist, Friedrich von Logau; see Longfellow 2000: 697). The positive component of this notion reward—implies that God will be good to those who show patience, as in the proverb 'All good things come to those who wait', likely an adaptation from the Bible: 'The Lord is Good unto them that wait for him, to the soul that seeketh him' (Lamentations 3:25).

<sup>54</sup> Sjamsuddin similarly characterises the situation of the Acehnese as a struggle against colonisation, one fought not just by Acehnese but by different communities around the world fighting against colonialism and economic exploitation. He firmly rejects the characterisation of Muslim Acehnese by Indonesians as 'fanatical and treacherous' (Siapno 2002: 153).

<sup>55</sup> The two Uyghur exile organisations, East Turkestan National Congress and World Uyghur Youth Congress, merged at a convention in Munich on April 16-18 2004 to become the World Uyghur Congress (WUC), with Erkin Alptekin (son of Isa Yusuf Alptekin) elected the first President. The main objective of WUC is to promote the right of the Uyghur people to use peaceful, non-violent and democratic means to determine the political future of East Turkestan, and to promote democracy, human rights and religious freedom for Uyghurs.

#### 276 chapter five

icans did nothing to help us. Instead, Bush is calling us terrorists!56 […] So we decided the only way is to fight, get weapons, guns, and train to fight the Chinese.

Ultimately, however, he saw the path to Uyghur independence resulting from a combination of study, wealth and religious faith. He insisted that education (either Islamic or modern) and money would not alone unite the Uyghurs in opposition to the state: religion would form the crucial component, the glue to bring the disparate groups together.57 A combined sense of frustration and betrayal (by potential foreign allies) is thus leading some Uyghurs rather to rely on their own efforts, with religion forming the foundation of a desired national unity.

# 5.7. *Modernity and the Return to Cultural 'Purity'*

I will turn finally to cultural and psychological aspects of religious renewal in Xinjiang. Etienne (in Ayubi 1991: 61) has termed Islam 'cultural and symbolic capital' that may be invoked both by rulers and ruled. Yet the dynamics of re-Islamisation may not be entirely instrumental, if they are instrumental at all; they are also tied up with complicated psychological processes. The transition to modernity has been a significant contributing factor to Islamic revival all over the world, straining the social and political fabric and leading people to 'turn to traditional symbols and rites as a way of comforting and orienting themselves' (Piscatori 1986: 27). Because in the past religion answered questions on life and death, it now becomes 'the means by which individuals hope to answer the new question of what it is to be modern, and, in so doing, to gain perhaps a reassuring, common world-view' (ibid.: 31).

On Islam-based reactions against 'imitative' developmental activities (that is, the wholesale adoption of Western models of modernisation), Ayubi writes: 'If the objectives of "effectiveness" [man's better mastery of his physical and social environment] are not defined within one's own cultural frame of reference […] feelings of "alienation" may intensify and

<sup>56</sup> The reference is to the US designation in 2002 of the East Turkestan Islamic Movement as an international terrorist organisation.

<sup>57</sup> Fuller and Lipman (2004: 344) predict that Islam 'will serve to reinforce nationalist feeling and community identity and solidarity, compelling even secular Uyghur intellectuals to recognise its power, though probably not to become Islamists themselves.'

the quest for "authenticity" may acquire a special urgency' (1991: 48-49). In Xinjiang, as previously in the Middle East, modernity has been introduced from without: China's programme to 'domesticate' the Uyghurs, accelerated since the late nineties via the Western development campaign, has entailed the imposition of Chinese models of regional development and increasingly the Chinese language and culture. This gives rise to two related issues: the impact of rapid modernisation (affecting developing societies across the globe), and the concept of competing cultural civilisations. As Mackerras observes: 'Modernisation affects issues relating to identity. It impacts on the feelings people have towards their own ethnic group and religion, sometimes weakening them but, *if in a context where modernisation threatens ethnic or religious identity*, sometimes strengthening them' (2005: 10, my emphasis).58 This reactive phenomenon has also been observed among Muslim communities in other parts of China. Gillette (cited in Mackerras 2005: 10) found that Hui in Xi'an in the 1990s considered Muslim practices to be more scientific and 'purer' than those introduced by the Han. In Xinjiang, too, recent research by Tsui (cited in Mackerras 2005: 10) into rural-urban migration shows that Uyghur migrant workers have retained a strong sense of ethnic difference in the city, with Islam among the most important markers of that difference.59 Meanwhile, the strong emphasis on Chinese cultural civilisation in the unified curriculum has created resentment among Uyghurs and led to calls for the revival of the old Islamic education system (Mackerras 2001: 300).60 The practice of

<sup>58</sup> See also Han scholar Wang Lixiong (summarised in Dreyer 2003: 424), who argues in a book published in Hong Kong that modernisation has failed in Tibet because it was forced in from without with no regard for Tibetan culture.

<sup>59</sup> Similar reactions have been observed elsewhere in Asia, e.g. among Malays who moved to cities, where they were exposed to modernisation in an environment dominated by communities of Chinese and Indian immigrants. The ensuing confrontation with different cultures and religions led to a re-assertion of Malay self-identity centred on Islam (Marsot 1992: 167); similarly, Acehnese women living in cities, where there is more ethnic diversity and a greater presence of foreigners, are more conscious than rural Acehnese of defending their Islamic identity against 'outside threats': 'In reaction to what ulama perceive to be the corruptive and demoralised influences of "Westernization" (e.g. "free sex", women walking alone in public places or coming home late at night), female roles in cities have become much more restricted …' (Siapno 2002: 172).

<sup>60</sup> Formal education prior to 1949 was almost exclusively under the control of the religious authorities; primary schools (*mäktäp*) were maintained by mosque communities and children were taught the basic rules of religious morality and social skills, the Arabic alphabet, classical Persian and the rudiments of religious praxis. Further education could be obtained at an Islamic college, where the curriculum comprised three branches of science: Islamic law, syntax and logic (Bellér-Hann 2008: 326-338).

#### 278 chapter five

Islam may thus be seen as a symbolic means of confronting the state: 'By embracing Islam, Uyghurs reject the atheism of Chinese communism as well as its goals of modernisation and social liberation' (Rudelson 1999: 207).

My interview data suggest that the re-Islamising sector of the Uyghur population is creating a new means of differentiation, a new boundary between itself and Han Chinese, based less on national customs and habits as it was in the nineties (though many of these evidently derived from Islam) and more on orthodox religious symbolism: prayer, abstention from alcohol and tobacco, and a return to modest forms of dress, especially among females. The process is perhaps best summed up by Polat, a Chinese-educated taxi driver in his forties from Ghulja, speaking in 2002: 'Uyghurs are returning to the mosque now, the youth too. They are searching for a kind of purity [*pak*], it's a way of defining their status as Uyghurs, their ethnic status … you know, separate from the Hans.'61 This purity is conceived in direct opposition to the 'impure' forms of modernity imported by the Chinese state. In another interview conducted in 2002, Abdullah, a graduate in theological studies, criticised the Chinese education system, where male and female pupils are permitted to study in mixed classes, and advocated segregation of women in the educational milieu: 'In Pakistan the women are good, they go to school and study in separate classrooms, away from the men. Not like here!'62 Later that day, he pointed out two Ürümchi women covered from head to foot, and observed: 'Look, aren't they beautiful?' I had noted the same reaction in Shökhrät, a secular intellectual in his thirties, that same year as we watched Turkey's football team play in the televised World Cup. He had indicated Turkish female supporters in the crowd wearing headscarves passed beneath the chin, and expressed strong approval. In this way, the enforcement of Islamic ethics 'may appeal to the young males as a way of ensuring "coverage" (*satr*) for, and control over, their female folk. The enforcement of such ethics may also appeal to the students or working female who may *voluntarily* choose to veil herself …'(Ayubi 1991: 47). Purity in Islamic practice simultaneously

<sup>61</sup> According to Fuller and Lipman, respondents in their study were conscious of mosque attendance as 'a means of reinforcing the distinctiveness of the Uyghur community from the dominant Han population and the Chinese state' (2004: 339).

<sup>62</sup> The preference for segregated tuition and a religiously oriented education has recently also been reported among other Muslim communities of the northwest, such as the Salar (see Goodman 2005: 332).

becomes a manifestation of social protest (especially in its extreme, ultra-Islamic forms) and a means by which the female may 'avoid the intrusions and molestations of males' in a modern society where the sexes come into more and more frequent contact (ibid.).

While the return to Islam embodies a search for religious and cultural purity for many men and women, it at the same time provides a form of psychological and spatial escapism for Uyghur men (and their wounded national honour), with the mosque characterised as one of the last remaining Han-free environments. Ömär, eighteen, Uyghur-educated male high school student, explained in 2004:

I think it's because they [Han Chinese] control us in all other spheres of our lives. You know, we're not allowed to have religion, to pray, in our working lives or at school. We never have time to go to the mosque … so, whenever we can, we do. It's the place where we breathe a sigh of relief, we feel overjoyed to see each other, we interact freely with one another; it's the one place where they cannot control us […] Everywhere else there are Hans, we have to deal with them, but at the mosque we are all Uyghurs!

In a follow-up interview conducted later that month, he continued:

The biggest reason why we go into the mosques now is trust, so that we can trust one another; to create unity among the Uyghurs.63 And to feel relieved, tranquil [Uy. *khatirjäm bolush*] … we feel happy when we're there, we can hear big ideas, big words taken from the Qur'an, we can get away from all the pressures of society, of the outside world … from the strains of interacting with Hans, from the strains of work. We feel disappointed and angry outside, but in the mosque we feel calm.64

Ablät, twenty, Uyghur-educated graduate doing business in Ürümchi, echoed Ömär's analysis in 2004, casting the Han-Uyghur relationship in terms of master and servant: 'Why are so many going in the mosque now? Because we don't have any freedom […] when we go to the mosque, we're free. It's the *only* time we're free. When there isn't someone [Han] saying "Do this, do that!"' Several other respondents similarly described the mosque as a space where Uyghurs could avoid daily humiliation suffered

<sup>63</sup> The point about consolidating Uyghur unity is complex since, while religion is certainly bringing re-Islamising Uyghurs closer together, it is simultaneously widening the chasm between strictly observant and nominal Uyghur Muslims, with partially observant individuals falling somewhere in between.

<sup>64</sup> An interesting comparison might be made here with the role of *mazar* festivals in the rural south, said to serve as 'a focus of prayer and for the outpouring of personal grief' (Harris and Dawut 2002: 112). Both might be characterised as a much-needed vehicle for personal (and national) catharsis.

#### 280 chapter five

at the hands of the Chinese. Consider this response from Jelil, Uyghureducated observant Muslim in his late twenties speaking in 2002: 'As the contradictions between the Hans and the Uyghurs get worse, the Uyghurs are clinging harder to Islam. They pray to Allah and ask for his help. Because they are being humiliated [Uy. *khorlimaq*] by the Chinese, tyrannised'. Shökhrät, intellectual in his thirties originally from Qäshqär, suggested in the same year that the desire to 'find peace' was one of two principal motives for going to the mosque, the other being the manifestation of symbolic resistance. Möminjan, a Turpanliq in his early twenties studying in Beijing, while acknowledging that re-Islamisation in Xinjiang involved elements of resistance against perceived oppression of Muslims—national and global—felt that the phenomenon was 'more about finding peace, regaining faith through religion'. Together with the mosque and Uyghurowned grocery stores like the *Istanbol*, a third remaining exclusively Uyghur space is the halal family restaurant. Märyäm, music trader in her thirties from Ghulja, laughed in 2002 as I observed the absence of Hans in a family restaurant in Ürümchi's Döngköwrük district, and retorted with audible pleasure: 'They don't come in here!'

Thus, while some Uyghurs attribute the return to the mosque simply to increased availability of translated educational materials about Islam, the dynamics involved are complex and multi-layered. Individuals respond differently to different stimuli, and diverse causal factors underlie the phenomenon. In contemporary Xinjiang, some Uyghurs are welcoming the salvation brought by Islamic materials and the mosque community because they have a psychological need to believe in Islam, catalysed perhaps by a brewing identity crisis of the type often referred to when describing the Islamic resurgence in the Middle East, one characterised by alienation, insecurity and powerlessness (Dekmejian 1995: 27). As in the Arab world of the 1970s, these individuals increasingly see Islam as 'the only remaining medium of identity and authenticity' in a situation of hopelessness and pessimism (ibid.: 31). Of possible further relevance is Ayubi's analysis of the historical tendency among Muslims to 'emphasise ijtihad [independent interpretation and judgement] during times of advancement when innovation and creativity were required, and to stress taqlid [imitation; traditionalism] in times of decay and stagnation, *when the rulers did not tolerate difference and freedom*, *and when the populace looked for elements of certainty in the midst of confusion or for consolation in the face of oppres-*

*sion*'(1991: 56, my emphasis). For such individuals, Islam is the sole remaining certainty in a society which allows them no space.

# 5.8. *Islam as a Vehicle for Personal and National Reform*

Some interviewees indicated strongly that Islam is providing a means of reform or salvation, both on a personal, individual level (saving the self) and at a national level (saving the Uyghurs). To return briefly to the North African example: the neo-Salafiyya reform movement in that region was a response partly to French colonisation, but also partly to internal decay within the Islamic community, in this case attributed largely to Sufism, whose popular religious practices had undermined the purity of the Islamic orthodoxy. The movement signalled the rediscovery of a 'Protestant work ethic' in personal practice, whereby hard work and social responsibility, abstention from alcohol, and fasting were stressed. At the same time, the reform embraced the modern sciences, so that modern Quranic schools were established, which combined study of Islam with that of modern disciplines (Esposito 1991: 77-78). Significantly, a similar 'work ethic' is also mentioned by Dekmejian in his discussion of the attributes of mainstream, moderate Islamists (fundamentalists), described as follows: 'a firm knowledge [of] and belief in Islam, willingness to face hardship, a sense of duty, a lifestyle guided by Islamic tenets and obedience to legitimate authority' (1995: 35).

Similar dynamics of reform in personal practice can be seen in contemporary Xinjiang, occurring in tandem with the return to prayer and inspired jointly by newly available materials on Islam and a desire to halt escalating social crises. A 31 year-old male restaurateur originally from Qäshqär touched in 2002 on the central notion of reform as he joked about the different strains of re-Islamisation occurring in the region: 'Well, you know, some people change for the worse, but some change for the better [*bäzilär buzuläydu, bäzilär tuzuläydu*]! All five digits on one hand are not alike.' The implication was that while some may turn to militancy and misuse Islam (the respondent's disapproval of this course is indicated in his use of the verb *buzulush—*'break'—to characterise 'broken' individuals), others saw Islam as a vehicle for reform. Patimä, a middle-aged clothing trader in her forties interviewed in 2004, identified the deaths from heroin abuse of many young Uyghurs over the past decade as a direct catalyst in the drive for personal reform. She noted that, as a result, many young people had

#### 282 chapter five

now begun to go to the mosque, and to try to do good deeds [Ch. *shanshi*] and live good lives.65 A male in his early twenties from Qäshqär explained in 2004 why he had given up alcohol and tobacco:

I used to drink and smoke and I didn't used to recite *bäsh waq*. But now I've given up drinking and I'm slowly giving up smoking. Because I recite *bäsh waq* and you can't drink or smoke if you do that […]66 I understand about Islam now. Lots of material has been translated into Uyghur, you see.

Speaking in the same year, Dilbär, eighteen, Uyghur-educated high school student, confirmed this trend among the youth, noting that it was occurring largely because alcohol did not mix with religion (thus re-Islamising Uyghurs must necessarily decide to abstain) and partly under parental duress. While some young people abstained altogether, others thought it acceptable to drink a little with a meal or at an event, so long as this was done with propriety. She referred to the saying *mäslik päslik* ['alcohol is degrading'], adding that people often made excuses for bad behaviour by blaming it on the effects of alcohol.

Other scholarship confirms this general trend. In Turpan, local *mollas* are voicing fervent opposition to popular social practices such as the drinking of alcohol and mixed-couple dancing, and Islam has gained in popularity among young Uyghurs, increasing numbers of whom are now fasting during Ramadan (Rudelson 1999: 202). Meanwhile, Fuller and Lipman's (2004: 341) respondents cited Islam and its associated practices as the Uyghurs' best defence against problems such as drug and alcohol addiction and domestic violence. A revived version of the traditional *mäshräp*—male peer gatherings designed to regulate moral, religious, and social etiquette had (until they were banned in 1997) successfully addressed the issues of alcohol and drug abuse, problems that many in the Uyghur community felt the Chinese authorities had ignored (Roberts 1998b; Dautcher 2004; Uyghur Human Rights Project 2006). More recently, a discursive form of proselytising based on the use of advice has become an important channel for the

<sup>65</sup> The problem of heroin-related drug abuse and death peaked in the nineties and has since begun to improve, perhaps partly as a result of the circulation of popular songs with didactic messages warning Uyghur youth away from heroin. See, for example, Yasin Mukhpul's 'Sirliq Tuman' [Strange Smoke] (Harris 2002: 274) and the song by Abdukhahar broadcast on Uyghur-language radio in Ghulja in 1996 (Dautcher 2004: 288). On the soar in heroin use in the 1990s, see Dautcher 2004 (288-93).

<sup>66</sup> Generally speaking, intoxicating drinks were considered *haram* in the pre-socialist period. A source from the late 1800s reads: 'Wicked people prefer brandy to water. All these people will be damned by God and his Prophet and will not receive blessings' (Bellér-Hann 2008: 387).

dissemination of orthodox ideas in Qäshqär by self-proclaimed *sünnätchi*  (those who follow the practice of the prophet Muhammad) (Waite 2007, 171). Thus, abstention—like prayer—is being perpetuated in contemporary Xinjiang through a powerful mixture of family and community pressure, as well as pressure exerted on the self.

It is noteworthy that some respondents treated the Uyghurs as a group so that the notion of personal reform was, by extension, conceived in terms of group (or national) reform. Möminjan, Turpanliq in his early twenties studying in Beijing, explained in 2002:

Previously, Uyghurs had lost their religion … they were smoking, drinking, playing around, even injecting heroin. But now they're realising that they have taken the wrong path [Uy. *khata yol*], and so they have given up alcohol and cigarettes and have gone back to prayer. They are finding their way back to the correct path.

A male music seller in his forties interviewed in the same year explained that many more people could go into the mosque now because they had given up alcohol, which was 'harmful'. His colleague rejoined: 'Nothing good, no goodwill has ever come to us from the Hans, and so they're praying to Allah instead', immediately suggesting a perceived link between secularisation/sinicisation (and the corresponding decay in orthodox practices) and Uyghur national fate; or the belief that 'harm' being done on a personal level is intimately entwined with harm being done on a group or national level. Again, this process is reminiscent of the reaction of some Arabs to the 1967 defeat by Israel, who concluded that they were 'inadequate believers who had not lived up to the ideals of Islam and therefore deserved their fate […] Muslims needed to be better Muslims […] if God was to spare them further calamity' (Piscatori 1986: 26).

For some, reform appears to be partly inspired by the renewed dread of what will happen after death, probably instilled by readings on Islam and information disseminated by *hajj* returnees. In 2002, Päridä, a waitress in her twenties, identified fear and superstition as one catalyst driving Uyghurs back into Ürümchi's mosques and away from alcohol: 'They're worried about the Afterlife! They say "Things will go badly for you" [Uy. *Yaman bolidu däp*].' The process of reform was often characterised with reference to the notion of *sawab elish* (earning merit through good deeds). Dilbär, eighteen, Uyghur-educated female high school student, explained in 2004:

Faith in Islam is based on *sawab elish* –– if we go to the mosque and pray, we are good Muslims. All will be well at the Day of Judgement. We are all

#### 284 chapter five

afraid of that day, when all the dead will rise and go to meet Allah. And they will be judged on the good and bad things they have done.

Ömär, eighteen, Uyghur-educated high school student, described similar attitudes among some young people at his school. But while some respondents in 2004 insisted that fear and superstition were significant factors inspiring Islamic renewal for some people, others firmly rejected the idea, shaking their heads and even laughing at the suggestion. In their view, the trend required a political explanation.

# 5.9. *Internal Divisions within the Uyghur Muslim Population*

As Gladney (2003: 457) has argued, another factor which calls into question the popular, monolithic view of the Uyghurs as united around separatist or Islamist causes is the existence of internal religious tensions (in addition to contradictions based on regionalism, class, etc.). Several of the patterns of internal division found in Islamic communities worldwide are also present in contemporary Xinjiang: for instance, the contradiction between reformist orthodox practices and popular Sufi practices or local, syncretic religious traditions; and the clash between the holistic view of Islam and the secular vision.

There is some evidence that the renewal of orthodox practice in Xinjiang is currently creating a doctrinal contradiction between the new reformists and those who practise long-standing, popular Sufi practices. As Voll notes, Sufi orders 'have a special local character and often continue what strict Muslim teachers think of as non-Islamic or pre-Islamic activities in an Islamic garb' (1987: 135). This basic tension has in the past produced internal counter-movements in both Chinese and Russian Turkestan,67 but the two forms of practice have at the same time integrated to a significant degree. From the eighteenth century on, Sufism is said to have 'absorbed the orthodox emphasis on the Qur'an, hadith reports, and the person of the Prophet' in Russian Turkestan (Lazzerini 1992: 155). Similarly, in pre-Republican Xinjiang, membership in Sufi brotherhoods did not formerly ex-

<sup>67 &#</sup>x27;Neo-Sufism' provided the framework in eighteenth-century Russian Turkestan for movements of Islamic revival and purification (cf. the Naqshbandiya tariqah), directed at the perceived compromises of the regular Sufi orders (Voll 1987: 136); in 1915 in south Xinjiang, a Tatar Islamic reformist denounced the annual fete at the Appaq Khoja mausoleum and the devotional activities practised there throughout the year as 'contrary to the spirit of Islam' (Zarcone 1999: 232).

clude participation in mainstream religious life; *mollas* were often members of such organisations, and mainstream/Sufi distinctions would have meant little to locals (Bellér-Hann 2008: 308-9).68 During the Republican period, however, there appears to have been an ideological split between the two streams, with an indigenous record stating that some of the *ulama* did not tolerate dancing, while the recital of mystic poetry and appearing with uncovered head in their presence were prohibited (ibid.). Today, many practices normally considered immoral—such as hashish smoking and gambling—may be practised at *mazar* festivals, and orthodox reformists in Qäshqär have accordingly begun to denounce popular traditions such as *mazar* pilgrimage (Harris and Dawut 2002: 112).69 The fact that orthodox Islam advocates abstention from alcohol and tobacco and prohibits musical performance at large tombs cannot but throw doubt on Chinese official theories that link 'Wahhabism' and the *mazar* (Harris and Dawut 2002: 115).70

Reformist clerics and their adherents are also increasingly questioning local, syncretic religious traditions. Waite writes that while in the mid-90s most Uyghurs continued to view *näzir* (feasts to commemorate the dead) as an essential aspect of Islamic practice, by 2004 its prevalence had substantially declined. As he points out, this is one example of an area where reformist and state interests may coincide, since the state has an interest in discouraging ritual activities outside the regulated space of the mosque as well as excessive expenditure on feasts, which can only aggravate poverty among southern Uyghur communities (Waite 2007: 175-176; 178).

The third contradiction and one that is becoming increasingly salient is that between the holistic notion of Islam and secularism, another tension prevalent across the wider Muslim world:

Many Muslims […] believe in the holistic nature of Islam in the sense of its being a way of life and not simply a religion (*dunya wa din*). This fusion of

<sup>68</sup> According to Arabov, orthodox Islam and Sufism are very much intermingled in contemporary Tajikistan. Some *mollas* serving holy places are also Sufis, and the majority of believers are 'not able to make out the difference' (2004: 346-7). On the other hand, he distinguishes between the *ishans*, which preach a conservative, moderate form of religion, and 'Wahhabi' extremists calling for the creation of an Islamic state.

<sup>69</sup> Despite the general connection between shrine visiting and Sufism, it should be stressed that pilgrims themselves do not necessarily belong to organised Sufi brotherhoods (Bellér-Hann: personal communication).

<sup>70</sup> See Harris and Dawut's (2002: 105) description of a performance of devotional music in the courtyard of the mosque attached to the Ordam *mazar*, where the large all-male audience was 'quiet and calm, seated, smoking cigarettes.'

#### 286 chapter five

matters of belief with matters of conduct in Islam makes it difficult to separate religion from politics […] "secularism" is therefore considered by many to be a concept not only alien to, but also incompatible with Islam' (Ayubi 1991: 50-51).

As Ayubi (1991: 51) notes, the tendency for some Arabs to view secularism as something alien and imposed from without also derives from the fact that in the (Arab) Islamic memory, the concept can be related only to periods of colonial hegemony, or, alternatively, to experiments by Arab nationalists with Western developmental formulas. In Xinjiang, secularism and secular education were introduced following the region's incorporation into the People's Republic. While religious elites composed of strictly observant Muslims have been maintained, these have been co-opted and controlled by the state since 1949 and increasingly so since the flurry of separatist activities in the mid-late nineties; state-appointed imams teach orthodox Islam strictly within authorised limits. Meanwhile, an ageing secular elite, educated in the Soviet Central Asian republics prior to the Sino-Soviet split in the sixties and influenced by Jadidist modernism, has gradually been replaced by a new PRC-educated elite, most of whom speak Chinese, but some of whom speak Uyghur as their first language. Members of this elite are required to pursue an ostensibly secular orientation and are prohibited from practising religion either in state educational institutions or in state organs or Han-run companies.71

As evidenced in history, the internal divisions that invariably characterise Islamic populations can be temporarily overcome to ensure success of a political Islamic movement, especially given certain environmental conditions. The example of the Arab world shows that diverse groups experiencing a sense of emotional dislocation brought about by modernisation/ industrialisation, new forms of education and rural-urban migration 'may consequently find in the bond of religion not only a sense of psychological consolation but also a new focus of identity and an alternative network of relationships' (Ayubi 1991: 175). Where the development process is of an 'imitative', dependent type, those social and cultural dislocations may render disparate groups susceptible to feelings of resentment of foreign domination and to calls for religio-cultural authenticity. It is precisely in this broader environment, Ayubi argues, that potential sympathy to an Islamic call may exist and that an 'Islamic revival' may emerge. The difference of

<sup>71</sup> See Naby (1986) for a survey of four generations of religious and secular elites in Xinjiang.

the Islamic thesis in itself 'seems to fulfil a certain function vis-à-vis an order [the State] that is deemed to be at most evil and corrupt, and at least a failure or a non-fulfilled promise' (Ayubi 1991: 175).72 Yet Islam's role as a unifier of diverse groups is often transient. Ayubi observes that while the Islamic revival has brought together different Arab groups (radical students, merchants) in their sense of alienation and quest for authenticity, these groups 'cannot go hand in hand … except in transitional situations, because the social "function" of Islamism is different for each group' (1991: 174-75).73 In most cases, religion has been a means rather than an end for the younger generations of modern, educated nationalists, so that 'once political independence had been achieved, Islam tended to recede from public life as political elites set about the process of nation building' (Esposito 1991: 95). As I will demonstrate in this book, in the case of Xinjiang, diverse groups of Uyghurs are at some distance from uniting under the symbolic banner of Islam. Moreover, were they to do so at some point in the future, it is unlikely that this unity would last beyond the revolution.

### *Conclusion: 'All Five Digits on one Hand are not Alike'*

As Esposito (1987: 24) observed in the late eighties, the renewed emphasis on Islamic identity and solidarity following the Islamic resurgence of the 1970s and the international contact made possible by modern communications and technology ought to have led to the forming of international linkages between Muslim governments. In reality, however, governments have appealed to Islam only where this might further their

<sup>72</sup> In colonial Algeria, nationalists of differing political orientations were able to build a mass movement which mobilised popular support by employing Islam as a common denominator: 'Islamic symbols and language offered an alternative rallying point and ideology capable of reaching out to the masses of Muslims and their religious leaders to unite them with elites in an opposition movement' (Esposito 1991: 78). Similarly, Morocco's political party Istiqlal (Independence), initially a small group of young, educated urban reformists, was able to multiply its membership to become a mass political movement by adopting the organisation and nomenclature of the Sufi orders (Esposito 1991: 79); and, in the Philippines, a combination of political and economic factors and a heightened Muslim consciousness led to the formation of a Muslim protest movement that sought to transcend linguistic and clan differences (Esposito 1987: 21).

<sup>73</sup> This can be clearly seen in the example of the Iranian Constitutional Revolution (1905-11), where once the constitution movement passed from one of simple opposition to the shah to one of actually drafting the constitution, the differing visions of religious traditionalists and Western-oriented reformers quickly came to the surface (Esposito 1991: 85).

#### 288 chapter five

national interests, while 'the logic of global Capital […] draws a map that transcends religious and national boundaries (Siapno 2002: 168). Dillon (2004: 142-55) and Bovingdon (2004b) have shown how political and economic alliances struck between China and the independent Central Asian states under the auspices of the Shanghai Co-operation Organisation (SCO) have consistently taken precedence over Islamic solidarity with the Muslim cousins in Xinjiang.

At the public level, however, new contacts enabled by pilgrimage to Mecca and the availability of translated writings by leading Islamic ideologues have served to 'increase the sense of an international, or transnational, dimension of Muslim identity …' in all corners of the world (Esposito 1987: 25). This symbolic dimension of global Islam can be clearly observed in Xinjiang today. The sense of symbolic community—local and global—experienced by re-Islamising Uyghurs can be situated within the broader frame of the internationalisation of Islam in an age of modern technologies. Through international trade, pilgrimage, study abroad and, above all, the mass media, Muslim communities around the world come to recognise mutual commonalities; through these identifications, they consolidate their sense of belonging to the virtual *ummah*. Raphael Israeli (1984: 289) has suggested that, unable to express themselves politically within the system, Hui Muslims sublimated through the universal *ummah* their political frustrations within China as well as their political aspirations outside of it.74 A similar dynamic may be occurring in Xinjiang today, where a feeling of victim-hood created by domestic Chinese oppression (and repression), combined with an empathy for similarly 'oppressed' peoples in the Middle East, is finding expression in an unprecedented sense of global Islamic solidarity and the visible demonstration of opposition through the symbolic vehicle of Islamic practice.

<sup>74</sup> China's Muslims, while recognising the local religious community as the focal point of their identities, have nonetheless always had an appreciation of the universal *ummah*; this is especially true of the Hui, whose origin legends state that their ancestors were of Arab/Persian origin, but true to a lesser degree for the Uyghur also. Israeli writes that Hui Muslims in the Qing period consciously cultivated a sense of membership in the global Islamic community, and characterises their daily prayers in the mosque as 'not only a communal worship, but also a way of identifying with all Muslims who faced the same *qibla* (direction of prayer)' (1984: 297). Posters displayed by Hui Muslims on walls and telegraph poles in northwest China during Ramadan in 1950 (shortly after the Communist takeover) proclaimed 'Let all the Muslims of the whole world unite and fight against tyrants' (Bush 1970: 267-8), again underlining the sense among Hui of belonging to the global *ummah*.

Clearly, the Uyghurs do have a choice in their representation. For instance, while it may be difficult for minority nationalities to express themselves directly in the official media, it has nonetheless been possible at times to do so via a roundabout route; the alternative representations presented in metaphor and allegory in the 'new folk' songs and historical novels of the nineties are an obvious case in point. Since the turn of the century, we are also seeing the emergence of an indigenous Uyghur scholarship, which takes the forms of cautious advocacy.75 But representation need not take the obvious, direct forms of argument and debate in newspapers and journals, or street demonstration; there are also subtle, symbolic means of exercising agency and making oneself heard. While Uyghur discontent had been expressed in the mid-1990s through negative stereotypes of the Han (in conjunction with positive counter-stereotypes of the in-group), in ethnic segregation, and in literary and musical metaphor, in the absence of those vehicles (since deemed 'local nationalist' and 'ethnic splittist' activities) some are now venting discontent in the form of a return to religion, although this is only one possible response. As Tashmämät and Shökhrät, male intellectuals originally from Qäshqär, emphasised in 2004: 'In an Islamic country, religion and the state are one, religions and daily customs and practices are one. This is why political resistance can take the form of religion and why it often does.' When Uyghurs express their discontent through vehicles such as stereotypes and ethnic segregation (both of which draw on differences in religio-cultural practices) or through visible participation in Islamic practices (veiling, growing a beard, praying and participating in the mosque community), they are drawing on a sense of 'we-hood' enjoyed for at least the last 500 years. In creating and maintaining these stereotypes and boundaries, or re-emphasising the centrality of religious ritual in everyday life in symbolic opposition to the state, they draw on a repository of social, cultural and religious practices and norms basically common to all Uyghurs since the last of the Uyghur Buddhists converted to Islam. Thus, religious culture 'links personal to collective truths which are large in potential scale, beyond the limits of a generational or a group identity or a particular period of a history, a government, regime or state', and may challenge the practices of that state (Feuchtwang 2000: 163).

But will the Islamic renaissance currently observed in Xinjiang develop into a militant form of Islamism? This is by no means a foregone conclu-

<sup>75</sup> See the contributions by Rahilä Dawut and Äsäd Sulayman in Bellér-Hann et al. (2007).

#### 290 chapter five

sion; that transition requires certain conditions. The 'Islamic alternative' has invariably become more radical within regimes that brook no dissent and regard Muslim criticism of any kind as a threat to their survival (Piscatori 1986: 33). Anderson (1997) and Hafez (2003) have convincingly illustrated how radical Islam emerges particularly in reaction to political repression and exclusion, in addition to economic deprivation. The Central Asian example is instructive. While Tazmini (2001) argued forcefully for the innocuous nature of most strains of Islam in Central Asia, even she conceded that Islam could potentially be further radicalised if: a) regimes pursue oppressive policies; or b) they fail to meet economic and social needs. Haghayeghi (1994: 196) had earlier made similar predictions. By 1997, the cycle of state repression and religious response in Uzbekistan was clear. Three years after a major crackdown in the Ferghana valley in late 1994 involving arbitrary arrests of independent Muslims, 'disappearances' of Islamic leaders, detention of bearded men and harassment of veiled women, police officers were murdered in Namangan—allegedly by Uzbek 'Wahhabis'—leading to a second, harsher crackdown, during which hundreds of men were arbitrarily detained (Tazmini 2001: 73). A more recent example is the Andijan incident of May 2005 (see earlier), which involved an armed crowd springing from jail 23 members of the moderate Islamist group Akramiya, who stood falsely accused of 'Islamic extremism'. That protest resulted in over 500 civilian deaths after Uzbek authorities reacted brutally to the situation, which they later characterised as a coup aimed at forming an Islamic caliphate. Yet the Islamic elements in the Andijan crowds were said to be more concerned with government corruption and poverty than with Islamic extremism. The quick resort to violence for which Islamists are often noted may thus be at least partly attributable to the violence visited on them by their own governments (Anderson 1997: 28-29). In Daghestan, as in most parts of the former Soviet Union, the term 'Wahhabi' was routinely applied to young people in the 1990s, simply on the basis that they had completed religious studies either in the Middle East, or with teachers from the Middle East. Yet when actual Wahhabi squads came from the Arab states to support the Chechen struggle for independence in 1999, Daghestani volunteers flocked to the militia in order to ward them off, afraid of being dragged into the Chechen war (Gammer 2005a: 104). There are surely lessons to be learned from patterns of state repression and resistance in other parts of the Muslim world; the future direction of the moderate, traditionalist movement currently observed in Xinjiang will depend largely upon the course of state policy, external stim-

uli, and the emergence (or otherwise) of an effective leadership (cf. Dekmejian 1995: 6).

Israeli, applying the concept of *aman* (pledge of security: here, the pledge to respect the religious beliefs and practices of Muslims living within a non-Muslim territory) to the Chinese context, writes:

In "normal" times, when law and order prevailed, and the level of cultural oppression was tolerable, Muslims accepted their fate without much ado […] In times of disorder […] Muslims might come to the conclusion that they had better take measures to protect themselves since nobody else would […] if Islam in China was no longer a workable way of life […] a Muslim was no longer bound to respect its laws and regulations, and he would be justified in rising against it (1984: 294-5).

This conceptualisation may be neatly applied to the current context. In an online article, Andrew Forbes, author of the seminal republican history *Warlords and Muslims in Chinese Central Asia*, rightly reminds us that there is no fundamentalist tradition in Xinjiang (1996). Yet the intensification of religious repression in Xinjiang may in time cause previously moderate Uyghurs to become radicalised, though this is only one of many possible responses. Mackerras notes that China's reaction to the events of 9/11 and subsequent designation of all separatist aspirations in Xinjiang as 'terrorist' has stimulated, not dampened, religious fervour: '… quite a few people who may not have any terrorist inclinations have seen themselves branded as terrorists and treated as such. This kind of state reaction has, in turn, sharpened feelings of Islamic identity' (2005: 11). Thus, the Chinese state runs the risk of fulfilling its own prophecy. If all oppositions are treated as dangerous and subject to repression, opposition that would be loyal if it were tolerated becomes disloyal because it is not (cf. Anderson 1997: 29). My empirical data, gathered in Xinjiang in 2002 and 2004, strongly suggest that for many Uyghurs, Islam is not the root of disaffection (in an ideological sense) but rather a vehicle for alternative ethnic representation. Islam, as a symbol of resistance against perceived oppression of Muslims (at both domestic and global levels), an egalitarian ethos, a religio-cultural system and a psychological outlet is providing respite from a host of well documented, pre-existing political, cultural and socio-economic disaffections.

Ayubi (1991: 59) has argued that, following the 1967 Arab defeat by Israel, there were three strands of re-Islamisation in the Middle East: a marked growth in general religiosity; an increase in the membership of the mystic orders; and the emergence of militant, neo-fundamentalist Islamic groups. Militant groups were largely influenced by a 'religious interpreta-

#### 292 chapter five

tion' of Israel's victory such that Israel was perceived as a 'religious formula' that had conquered and flourished where nationalist, socialist and pan-Arab formulae had failed. Thus, 'the hopes of people in the Arab world in the fifties and sixties hinged upon socialism and pan-Arabism; when these failed (and Israel succeeded) there was a strong "return to Islam", but this time with a fairly right-wing orientation' (ibid.: 65). Some might argue that the conditions for a militant response exist in Xinjiang today,76 where the accelerated return to Islam following the suppression of the 1997 Ghulja disturbances can be characterised on one level as a response to the various, failed ethno-nationalist theories that projected Xinjiang independence for that year. Yet the revival does not at present bear the features of 'religio-political fundamentalism' as outlined by Dekmejian (1995: 6), at least not at the activist end of the spectrum: for one thing, the movement lacks charismatic leadership committed to revolution, partly (but not wholly) because of the state's close control of religious personnel;77 for another, it cannot be described as a movement in the sense of an organised network since, as in Central Asia prior to independence, religious figures lack 'a broad-based organization which could increase the cohesion of the Islamic community' (Atkin 1992: 63); third, while an embryonic ideology of salvational beliefs and practices may have begun to form, for most this has not led to the development of what Dekmejian terms a 'fundamentalist personality'.78 For now, the vast majority of re-Islamising Uyghurs are following non-violent paths of religious renewal, ranging somewhere between Ayubi's category of 'growth in general religiosity' and Dekmejian's

<sup>76</sup> See Becquelin (2004a: 378), who warns that radical Islam, with its nearby centres in Afghanistan, Pakistan and the Ferghana valley, may prove a magnet for ethno-national aspirations in the near future, notwithstanding the historical characteristics of non-radical, traditional Islam in the region.

<sup>77</sup> There is anecdotal evidence that in the southern town of Khotän, local Muslim clerics claiming to represent Allah and supporting the notion of 'holy war' are currently competing with state imams for influence (Mackerras 2001: 297). Fuller and Lipman (2004: 334) also mention religious leaders at small, neighbourhood mosques who are not on the state payroll but are chosen by the local community on the basis of their moral character and Islamic learning; however, they point out that, as with the larger mosques, state informers attend their sermons and monitor their activities. Indeed, reformist cleric Ablimät Damolla was dismissed from his post at the Toqquz Tash mosque in Qäshqär in November 1997 following a sermon in which he advocated the combination of religion with education and accused past rulers of keeping Uyghurs in ignorance and superstition (Waite 2007: 170-171).

<sup>78</sup> See Dekmejian's (1995: 32-35) archetypal configuration of the 'modal personality' of the radical Islamic fundamentalist.

manifestation of 'spiritual fervour' but not yet translating into 'revolutionary activism'.

Finally, it would be inaccurate and disproportionate to suggest that all re-Islamising Uyghurs are doing so for the same reasons, or even that all Uyghurs are returning to the mosque. In fact, the causes and sources of the Islamic renewal are far more complex, as I have outlined, while the return to Islam among some individuals has simultaneously created a greater polarisation between observant and nominal Muslims. In the next chapter, I deal with the *minkaohan* (or Chinese-educated Uyghurs), who rather than pursuing the 'pure' path of re-Islamisation that separates the self from the state, seek rather to survive in the new society through accommodation, negotiation and hybridisation.

#### Chapter Six

294 chapter six

# endogamiStS: gender, nation and a SeleCtive taboo

intermarriage has long been viewed by ethnicity theorists in america and europe as an indicator of ethnic assimilation and integration, perceived in terms of acculturation—the extent to which mixed couples adapt cultural practices in the context of intermarriage—and of the degree to which a minority partner maintains social networks (or not) following intermarriage (Spickard 1989: 10-13). the key assumption was that the minority group sought to assimilate into the dominant society, where that society was resistant. Yet we know too that the context for intermarriage has not always or necessarily been dictated by the dominant group; it can equally be controlled by the minority group. research in the US illustrated that while the 'melting pot' thesis had been superceded by the assimilationist school of thought by the turn of the twentieth century (i.e. the expectation that non-anglos give up their distinctive cultural traits), Japanese americans and Jews 'had agendas of their own […] shaping the settings of interethnic relationships in which intermarriages took place' (Spickard 1989: 5). the rationale for endogamy has been described as 'the maintenance of group boundaries by forbidding the introduction of outsiders […] the reinforcement of intra-group ties and sense of identity'; moreover, in societies where boundary maintenance is considered paramount, endogamy may become prescriptive and be deployed alongside segregation (Cerroni-long 1984: 26-27). in this way, intermarriage—and its avoidance—becomes a key variable in the examination of ethnic relations and boundaries.

before considering intermarriage trends in China, it is useful to look at the experience of russia. there, the tsarist regime actively encouraged mixed marriages in the baltic and Western regions of the empire during the second half of the nineteenth century in an attempt to integrate the imperial peripheries. this policy anticipated the active promotion of intermarriage which would characterise the Soviet period (Werth, in edgar 2007: 584). From the mid-1930s, mixed marriages were encouraged as evidence of the 'friendship of [Soviet] nations', a means of accelerating the emergence of a pan-Soviet people, and a force for 'modernising' the historically 'backward' region of Central asia, in terms, for instance, of gender equality (Edgar 2007: 581-583). This 'enlightened' policy was compared positively against eugenics theory and the practices of racial segregation then prevalent in North America, Nazi Germany and South Africa. Yet, as Lynne Edgar points out, it had much in common with the agendas of racial blending pursued in Latin America (*mestizaje*) and Australia or New Zealand (biological Europeanisation) in that it ultimately favoured the dominance of one language and culture over others (Edgar 2007: 584; 593). Edgar notes a strongly Russo-centric element in the notion of hybridity celebrated by the Soviet Communists, with mixed families more likely to converse in the Russian language in the home and to adopt Russian-style home furnishings and foods. In Turkmenistan, the policy led to abandonment of the Islamic circumcision ritual for sons; in Kazakhstan to mixed couples using Russian as the standard channel of communication; and in Tatarstan to the rejection of traditional gender norms and the equal division of housework (Edgar 2007: 590). Yet, ultimately, some in the non-Russian republics rejected the policy—and its emphasis on the use of Russian as a regional *lingua franca*—as a form of 'Russification'. By the 1970s, the rate of intermarriage had slowed and in some republics had been reversed (Edgar 2007: 586). In all the above cases—the Soviet Union, Latin America and Australia / New Zealand—state promotion of intermarriage served not to weaken the hold of indigenous cultures as originally anticipated but rather to increase the number of individuals who identified with indigenous groups (Edgar 2007: 595).

In China, state policies towards ethnic intermarriage oscillated over the centuries. During the Tang dynasty (618-907), widely considered to be an era of unprecedented Chinese cosmopolitanism, intermarriage with 'barbarians' was so common that assimilation of non-Han cultural traits 'often confounded efforts to clearly distinguish Chinese from barbarian lineages […] and, most of the time, Tang people were not interested in making that effort' (Abramson 2003: 123). The phenomenon was not however without its critics. A typical story from the Tang tells of a young Han member of the gentry, who is tricked into marrying a fox spirit posing as the daughter of a rich barbarian merchant; this, according to Abramson, 'reflects conservative anxieties over the regular phenomenon of Han bluebloods contracting marriage alliances with social climbers, often of non-Han ethnicity, for financial or political reasons' (2003: 140). By the time of the Song dynasty (960-1279), the imperial court prohibited intermarriage between Hans and others, although the 'Northern barbarians' encouraged it. Conversely, the Ming (1368-1644) adopted a policy of forcible assimilation of 'peoples with

#### 296 chapter six

coloured eyes' (*semuren* 色目人), prohibiting Mongolians and others from seeking endogamous marriages and actually *requiring* them to marry with Chinese. During the early Qing period (1644-1911), Manchu-Han intermarriage was prohibited, but as the northern Manchus increasingly moved south into Han territory, Manchu rulers were obliged to accept the phenomenon as a social reality. They however continued to ban Miao-Han, Mongol-Han and Hui-Han unions. Following the fall of imperial China (1911), the Republican government sought to encourage intermarriage as a means to speed the assimilation of peripheral groups (Li Xiaoxia 2006: 82- 83). Thus, Li Xiaoxia succinctly sums up the history of Chinese policy towards intermarriage as follows: 'It can be seen that, throughout history, where the ruling classes supported intermarriage it was with the goal of softening ethnic contradictions; where they restricted intermarriage, it was with the goal of divide and rule' (2006: 83). Meanwhile, the attitude of PRC leaders today remains similar to that maintained in the early period of CCP rule (1950s). While the state does not promote intermarriage, nor does it prohibit it: 'It is not so much that the state supports intermarriage as that it supports the free marriage rights of individuals' (Li Xiaoxia 2006: 85). This view is however not shared by certain dominated minority groups, who see the state rather as attempting to forcibly assimilate them.

In this chapter, I examine the second prominent form of symbolic resistance operating during the post-1997 period: selective endogamy. This ethnic boundary was already in force in the mid-1990s (see Chapter 3), but has since 1997 become paramount in the struggle, as perceived and represented by respondents, to resist the colonial state and ensure cultural survival. I illustrate how, following the demise of earlier boundaries which are no longer politically or legally viable, a mixture of parental prohibition, community supervision, and self-censorship is employed to guard against inter-ethnic courtship and marriage with Han people. This process is applied particularly to Uyghur daughters in what Bellér-Hann refers to as a 'one-sided form of ethnic boundary-crossing which is regulated by the dominated group' (2001a: 59). I consider in turn the racial, religious, cultural, historical and political barriers to Uyghur-Han intermarriage, as articulated by respondents. I then analyse the ways in which the taboo could occasionally be managed, such as through a mixed couple's physical removal to a 'neutral' context, or on condition of the religious conversion of the Han partner. Finally, I present a hierarchy of intermarriage preferences constructed by Uyghurs in contemporary Xinjiang, and propose that this system of selective endogamy is deployed to maintain ethnic honour (and

avoid ethnic shame) in a context of Uyghur national and political impotence.

#### 6.1. *The Statistics on Intermarriage*

Contemporary Chinese scholars regard intermarriage as one of the most important measures of ethnic integration, with an intermarriage rate of 10 per cent or higher between groups thought to indicate 'relatively good ethnic relations' (Li Xiaoxia 2004: 20; cf. Zhang Suqi 2005: 24). Of the total number of minority individuals in mixed unions in 2000, 81.58 per cent were married to a Han spouse (Li Xiaoxia 2004: 20). However, the occurrence of intermarriage between minority and Han persons varies between regions and between groups. A study conducted in Kunming, Southwest China, found that ethnic intermarriage was prevalent, and that 'status homogamy' was the norm for 70 per cent of mixed marriages in that city (Xing 2007: 170; 175). In other words, most spouses valued equal social status more highly than common ethnicity. Furthermore, where one spouse enjoyed higher status than the other, the data suggested that gender was more significant than ethnicity, that is, husbands of *any* ethnic status (be they Han, Hui, Yi or Bai) in any combination of mixed marriage tended to enjoy higher socio-economic status than their wives (Xing 2007: 178). At the same time, Xing's findings suggested that, as the size of urban minority elites increased after 1990 as a result of preferential minority policies, minority individuals increasingly sought to match both status *and* ethnicity when choosing a mate.

In Xinjiang, the situation is quite different. As Ren and Yuan point out, while intermarriage between Han migrants and minority females is in general on the increase, the practice is increasing with *all groups but the Uyghurs* (2003: 99). Ma Rong's research team, conducting a survey in 1997 on ethnic relations in Xiahepu township, Qaghiliq (Ch. Yecheng) county in Qäshqär prefecture, failed to find a single instance of Uyghur-Han intermarriage. While they did find three examples in a neighbouring horticultural garden, it is noteworthy that all occurred within the same family. A Sichuanese farmer (first-generation Han settler) who moved to Xinjiang in 1961 had married a local Uyghur woman; later, two of the couple's three mixed-heritage daughters had entered mixed marriages themselves (Ma Rong 2003: 120). This suggests that ethnically mixed marriages commonly repeat down generations (cf. Li Xiaoxia 2005: 63-64). In Herbert Yee's study

#### 298 chapter six

of Uyghur-Han relations in urban Xinjiang, conducted in 2000, 78 per cent of Han respondents gave the green light to intermarriage, but less than one third of Uyghur respondents approved. Indeed, the actual level of Uyghur approval was probably lower if we account for the likelihood of respondents giving politically correct answers.1 It is instructive that only seven out of Yee's 378 respondents (0.018 per cent) were in a mixed Uyghur-Han marriage (2003: 436-7). Zhang Suqi's study, conducted in Turpan in the early 2000s, showed an increased number of Uyghur women intermarrying with Han males, but the concrete numbers involved remained small: just four instances in 2001, rising to nine instances in 2003 (2005: 25). Despite finding that Uyghurs are more likely to intermarry with Han than with any other group (Kazakhs and Hui sit in second and third place), Li Xiaoxia placed the intermarriage rate among Uyghurs at just 0.62 per cent (2004: 21). In their comparative study of ethnic intermarriage in Beijing and Xinjiang, Mamet, Jacobson and Heaton found that, of all minority groups in Xinjiang, the odds of exogamy were lowest among Uyghur males (just 0.2 percent) and Uyghur females (0.46 per cent) (2005: 198, 200). According to another study conducted in rural areas of Qäshqär prefecture, strict endogamy remains the rule, with locality, ethnicity and religion cited as the main factors in spouse selection. 99.39 per cent of couples in rural Qäshqär were in all-Uyghur marriages at the time of this survey (Hu Xinxia 2006: 15-16). Hu observes that this ratio reflects the high population concentration of Uyghurs in southern Xinjiang as well as what he terms the 'limitations imposed by Islam'. The data shows, however, that those 'limitations' did not necessarily imply parental intervention, since a total of 69.5 per cent of the marriages resulted from 'free love' (57.1 per cent) or introduction by a friend (12.4 per cent). Here, as in Turpan, the number of intermarried couples had doubled between 2000 and 2004, but the proportion of such unions within the total population remained negligible (2006: 16).

<sup>1</sup> For further examples of what can happen when formal survey methods are used to research sensitive topics in Xinjiang, see Herbert Yee's study on Uyghur-Han relations in Xinjiang and his discussion of the political difficulties encountered during data collection and analysis (2005). 40 per cent of those invited to participate in the survey declined. Of those respondents who agreed to participate, non-credible responses included the following: 82.5 per cent of Han respondents said that they felt the Uyghur language was 'useful' (p38); 41.8 per cent of Uyghur respondents said that Han people were 'totally trustworthy' (p43); 87 per cent of Uyghur respondents claimed to be proud of their Chinese citizenship (p44); and 46.8 per cent of Uyghur respondents agreed that 'Xinjiang has been part of China since ancient times' (p47); Each of these findings is entirely at odds with data collected in unofficial interviews.

In interviews conducted in 2002 and 2004, my respondents in Ürümchi observed that intermarriage had 'very slightly increased', estimating the percentage of parents who might consent to a Uyghur-Han match conservatively at 1 per cent and generously at 10 per cent. Most believed that the majority of intermarriages occur in the regional capital, Ürümchi, a perception supported by one study, which found that the odds of exogamy in Xinjiang increased within the following four categories: young people, highly educated people, urban residents, and residents of northern Xinjiang (Mamet, Jacobson and Heaton 2005: 199). All respondents noted that Uyghur women were more likely to take Han husbands than the reverse. This gendered pattern is also confirmed in domestic and international scholarship (Zhang Suqi 2005: 24; Mamet, Jacobson and Heaton 2005: 196, 201). Yet, if anything, the increase in intermarriage seems to have *heightened*  the degree of communal resistance as family members and local communities respond to the perceived cultural threat. This process has been observed elsewhere in China. In Inner Mongolia, where traditional cultural markers such as language have been gradually eroded, endogamy has become the most important marker of difference between Mongols and Hans (Borchigud 1994). Similarly, Ben-Adam notes that Hui Muslim families living outside Beijing, who cannot trace their ancestral origins beyond four or five generations, are trying to rebuild an uninterrupted descent from Hui ancestry by not marrying their daughters to non-Hui (1999: 198). In Xinjiang, the response may be motivated especially by the introduction of mandatory Chinese-medium education since 2002 and group fears of 'linguistic genocide' (see Chapter 1). The social uproar which has greeted even a slight increase in Uyghur-Han mixed unions was neatly reflected in the TV soap opera *Xinjiang guniang* [Xinjiang Girls], broadcast in Chinese and Uyghur in China proper and Xinjiang during the summer of 2004. This drama series followed the romantic lives of four daughters of a Uyghur scholar based in Beijing, one of whom was involved in a romantic relationship with a Han, and depicts the attempts of the Han suitor to win over the male elders in the Uyghur family, portrayed as tradition-bound, inflexible and obstructive to the match. It is indicative of the current climate that, by the end of the series, the marriage did not go ahead; rather, viewers' outrage was left suspended as the acculturated, Beijing-based father urged his daughter's suitor to bide his time until her extended family abandoned its 'old-fashioned' ways (Smith Finley 2013).

In most inter-ethnic contexts, patterns of endogamy and exogamy are deeply influenced by the images that people construct of each other and

#### 300 chapter six

of themselves, that is, by stereotypes rather than actual experience (Spickard 1989: 19; Shibata 1998: 83). Xinjiang is no exception. Below, I discuss the many ways in which family and community currently bring prejudice to bear on the marriage choices of young Uyghurs. Downs, in her work on interracial unions in the US, once asked how, in a time of deeply troubled race relations, does one 'reconcile group loyalty with friendship—even love—across the color line?' (1975: 160). While concepts like race and colour have been less significant in Xinjiang (and in China in general), markers such as religion, culture, historical memory and politics ensure that a small minority of young urban Uyghurs are currently wrestling with the same conundrum.

# 6.2. *Parental Prohibition and Excommunication*

Studies show that of all the problems an intermarried couple may have, the attitudes of opposing members of the families are often the most enduring (Evers 1975: 157). Familial disapproval and/or rejection can exert significant external stress on an intermarried couple (Cerroni-Long 1984: 41-42). This is especially true of cultures in which there is a strong tradition of filial piety and respect for parents, grandparents and elder siblings, such as in the Middle East. In these cases, parental and familial influence can be decisive in making or breaking a mixed union. One study concluded that, owing to the importance of the family, intercultural marriage between Middle Eastern people and others may be impossible 'unless the man or woman refuses to bow under parental pressure' (Brown and Farahyar 1994: 180). An Arab female graduate who married an American professor was disgraced by her mother, who refused to speak to her following the wedding (Brown and Farahyar 1994: 180). Similarly, Middle Eastern men who fell in love with American women while studying in America in the 1990s were almost always faced with parental pressure to marry a girl from their home country. Relatives even resorted to the lengths of visiting the man in the US in order to dissuade him from entering into the mixed union, or to pressure him into an arranged marriage instead (Brown and Farahyar 1994: 175).

Other situations where parents may obstruct an exogamous match include the scenario where one ethnic group has a history of political and/ or socio-economic conflict with the other group, or has a strong religious identity and corresponding desire to maintain religious endogamy. Here, too, methods of obstruction are various and often extreme, and familial

sanctions can remain in place beyond the marriage in the form of a 'complete and apparently lasting severance of the family ties' (Beigel 1975: 72). As a result of US-Japanese hostilities during the Second World War, *Nisei* (second-generation) Japanese Americans intermarrying with white Americans were often greeted with parental boycott of the wedding ceremony, opposition by grandparents, and even the breaking off of family relations following the marriage until—and sometimes beyond—the arrival of grandchildren (Spickard 1989: 63; 113). The mother of one Japanese American woman threw her out of the house shortly before her wedding to a white male, while another young *Nisei* bride was struck by her younger brother for proposing to marry a white (Spickard 1989: 63). A third *Nisei* Japanese American respondent considered eloping with her black boyfriend after her parents forbade her to see him, but ultimately could not stand to break with her family. Fifteen years later, she remained single and bitter towards them (Spickard 1989: 67). The reactions of first-generation Jewish immigrant parents in the US to what constituted 'inter-faith marriage' were equally dramatic. Confronted with the desire of their Americanborn youngsters to marry Gentiles, some Jewish mothers threatened suicide, while Jewish fathers often disowned their child, even sitting *shivah* (i.e. conducting a funeral ceremony) and pretending he/she was dead. One mixed couple saw each other in secret for over twenty years rather than tell the male partner's first-generation Jewish parents that they wished to marry (Spickard 1989: 188-189). A Jewish respondent to Beigel's study rejected the idea of marriage with a non-Jewish black girl outright because, as he put it, his orthodox Jewish parents 'would die' if they heard of such an intention (1975: 70). Where such unions did go ahead, Jewish families often continued to shut out the miscreants and their Gentile spouses long after the marriages had been consecrated: 'And for years afterwards, gossip circles rang with tales of their misdeeds' (Spickard 1989: 188-189). The arrival of a baby did not necessarily lead to reconciliation in these cases. Although some couples and parents resumed relations following a birth, others remained hopelessly estranged (Spickard 1989: 113; 190).

Yet the belief in love itself has led young couples throughout history to defy family and community sanctions by opting to elope, a 'last resort' course of action still employed today in the most difficult situations of inter-ethnic intimacy (Ellman, cited in Spickard 1989: 194). In a context where an ideology of love and marriage as based on *affect*—'companionate marriage'—has become the global norm (Hirsch and Wardlow 2006: 14-26), individuals increasingly invest their faith in the power of love and its abil-

#### 302 chapter six

ity to overcome adversity. Elopement is most likely to occur where parental or community reactions to a match are extreme. Thus, where one or more sets of parents break with a couple over marriage, the greatest loss may be to the families (Evers 1975: 157). In contemporary Guyana, where ethnic and religious endogamy remains the norm and boundary-crossing is taboo, Indian parents often interfere when a child chooses an African partner, sometimes throwing the child out of the house and severing home and family ties. In these circumstances, some young mixed couples have chosen to elope (Shibata 1998: 87).

In contemporary Xinjiang, parental and familial opposition to intermarriage—and Uyghur-Han intermarriage in particular—has been extreme. Writing at the turn of the century, Rudelson noted that Uyghur students who undertook long-term study in China proper and ended up marrying local Han were invariably disowned by their parents or told not to return to Xinjiang (Ben-Adam 1999: 204). In the early 2000s, cases were reported from the Ghulja [Ch. Yining] marriage registry office involving mixed couples who had managed to get hold of the necessary documents and had registered their marriage in secret. On discovering this, family members are said to have erupted in protest:

When their parents found out, they sped to the registry office, where there was a great hullabaloo. In some cases, fathers and close family friends even drew knives and threatened murder. In order to avoid this sort of thing occurring, the Yining municipal marriage registry office subsequently ensured that in all cases of intermarriage the officer first held a conversation with the couple, then required them to write a statement confirming their desire to register their marriage and outlining the attitudes of their parents towards the match. Only where neither set of parents expressed explicit opposition was the couple permitted to complete formal registration procedures (Li Xiaoxia 2006: 82).

Indeed, familial intervention has been so determined in Xinjiang that the authorities were forced to issue a new set of regulations on the 1st October, 2003, entitled 'Ordinance Governing the Registration of Marriages'. These regulations simplified marriage registration procedures in Xinjiang with the aim of weakening couples' former reliance on their families (Li Xiaoxia 2006: 82). Even in the eastern oasis of Turpan, where locals are generally perceived to be more receptive to interactions with Han people, parents often react negatively to a mixed Uyghur-Han union. The Uyghur spouse in one such mixed marriage related in interview how his father had opposed the match from the beginning, finally agreeing to fund the wedding

but declining to attend the ceremony. The father remained cold towards the Han wife until the couple's baby was born (Zhang Suqi 2005: 26).

My respondents in Xinjiang were on the whole unanimous that Uyghur-Han intermarriage is not an option at the current time, and confirmed that the vast majority of Uyghur parents simply will not allow it, regardless of whether they have opted to educate their child in a Uyghur- or a Chinesemedium school. Gülshäm, a twenty year-old *minkaohan* (Chinese-educated) student from central Ürümchi, was incredulous when I enquired in 2004 whether her father would consent to her marrying a Han male, exclaiming: 'No! My Dad would never let me marry a Han!' Although Gülshäm's (divorced) father occasionally teased his daughter with the suggestion that he might re-marry with a Han woman, the dividing line between humorous banter and real life was clear.

Uyghur parents adopt a variety of strategies in contemporary times to ensure endogamy—or at least, selective endogamy (marriage with anyone but a Han). Gülhärä, a twenty year-old *minkaomin* (Uyghur-educated) student from the Uyghur district of Döngköwrük, stated in 2004 that her father would 'go mad' if she dated a Han peer. One way in which he acted to prevent this possibility was to forbid her to date anyone—Han, Uyghur or otherwise—until she had completed her higher education and entered the working world. In doing so, he outlawed romantic relationships during precisely that time (teenage years) when she was most impressionable and most likely to come into contact with Han peers. Like many other Uyghur parents in the city, he imposed a curfew on his daughter, requiring her to return home before dark each evening. Like parents in other countries and contexts of intermarriage, Uyghur parents regularly used the threat of excommunication in the case that their offspring brought home a Han partner. Aynur, a *minkaomin* university graduate in her early twenties, reacted with horror when I playfully suggested in 2004 that she might marry a Han peer, and exclaimed that her parents would disown her if she did such a thing: 'My parents would kick me out of the house!' Hearing this, her male cousin, Ömär, a *minkaomin* high school student, grinned and added: 'They would say "She's no daughter of mine!"'

The fact that sons and daughters are fully aware of the potentially adverse reactions of their mothers and fathers acts to further enhance endogamy through self-censorship. The majority of young people have no desire to enter into confrontation with their parents, and so the desire to avoid conflict often leads couples to separate rather than pursue their mixed union. Rabiyä, a twenty year-old *minkaomin* student, explained in

#### 304 chapter six

2004: 'Even if the couple love each other, the parents won't usually agree to it. Then they break up. There's no way around it (Uy. *Amal yoq*).' When I suggested that the couple might decide to elope, she and her *minkaomin* classmate, Aygül, firmly ruled the possibility out: 'It won't do for them to run away! That's no good! (Uy. *Bolmaydu!*)' In another example from 2004, Mälikä, a twenty-year old *minkaomin* student who secretly worked as a hostess, was pursued for some time by her Han manager. This young man apparently fell in love with Mälikä and repeatedly asked for her hand in marriage. Although she considered him a gentle and kind partner, she eventually informed him that marriage would be impossible since her parents 'would never agree to it'. She subsequently ceased working in the karaoke venue, and the two lost touch.

While some young Uyghurs involved in mixed relationships buckle under parental and community pressure, others opt for desperate measures, choosing to defy their parents and run away. Abdukerim, a *minkaomin* graduate in his twenties from Aqsu, conceded in 2002 that one common response to parental pressure is for children to elope with their partner. As seen in other contexts, eloping Uyghur youth typically emphasise the sanctity of love, as described by two young male *minkaomin* migrant workers in 2004:

The children say 'This is the person I have fallen in love with'. And, if the parents don't agree, then the son or daughter leaves and goes some place far away, only coming back when they have children. Even then, they may ignore the parents and look the other way in the street; then the parents feel humiliated.

In refusing to recognise the inviolable nature of love, Uyghur parents may thus be rejected by their children in a process of 'counter-ostracisation'. A significant level of rebellion has inevitably contributed to an increase in the percentage of Uyghur parents who would consider intermarriage. As Märyäm, a music trader in her thirties from Ghulja, explained in 2002: 'If they don't, their children will go and get married anyway [Uy. *qiliweridu*]!' For the protagonists themselves, or those potentially attracted to Uyghur-Han mixed unions, rebellion may prove irresistibly attractive within a climate of prohibition. When I asked what she would do if her father forbade her to marry a Han, Gülshäm, a twenty year-old *minkaohan* student, replied in 2004: 'Well, I suppose we would have to elope (Ch. *siben*)!' As a *minkaohan* towards the more acculturated end of the accommodation spectrum, she used the Chinese term, and was plainly excited by the prospect.

Gülhärä, twenty, a *minkaomin* student, confirmed in the same year that parental prohibition in some cases had the reverse effect of encouraging youth rebellion: 'What can they [individuals in Uyghur-Han mixed relationships] do? Would *you* listen to your parents if they told you that you couldn't marry someone? There's nothing for it [but to elope] (Uy. *Amal yoq*).' Once a couple has eloped, the action is irreversible owing to cultural perceptions of female lost chastity. Tahirjan and Ghäyrät, migrant workers in their thirties originally from Aqsu, cited a humorous Uyghur saying in 2002, as follows: 'If the parents won't at first agree to the marriage, they will after the couple has eloped for fifteen days!' The implication, evidently, is that after spending fifteen days alone with her male partner, the girl will no longer be 'pure' (*pak*) and thus no other Uyghur man will have her.

Where parental excommunication takes place in Xinjiang, it is extremely distressing for both parties. Interviewed in 2004, Patigül, a music trader in her late twenties from Döngköwrük, narrated the story of her female friend who lives in Tieluju (a Han-dominated district in northwest Ürümchi) with her Han husband. The root of this situation lay at least partly in the fact that this husband had not converted to Islam, despite opting to abstain from consumption of *haram* meats: 'She and her parents don't talk at all. Her parents won't have her in the house. So, every time at Rozi or Qurban festivals she's in tears because that's the time when families normally get together, when relatives visit one another …'

One circumstance in which parental opposition may be lessened and the path of intermarriage eased is where one parent is deceased (Khatib-Chahidi et al 1998: 55) or otherwise absent. Osman is a *minkaomin* market trader who married a Han female trader who worked alongside him at the local outdoor market in the early 2000s. Both his parents had died when he was just fifteen, and he had been raised by his elder sister, who cooked and cared for him. In the absence of parental prohibition, Osman had been relatively free to marry his Han partner; yet I learned from his best friend that none of his remaining relatives (including his sister) had attended the wedding, suggesting that significant stigma remained.

#### 6.3. *Community Supervision*

As an institution, marriage may be interpreted as ensuring the perpetuation of a given ethnic group and the stability of a given community; as such, it becomes a social rather than an individual value (Cerroni-Long 1984: 33).

#### 306 chapter six

Put differently, intermarriage is community—not private—business. Empirical studies on intermarriage return repeatedly to this theme. Grearson and Smith note that people in intercultural relationships 'must be prepared for extra attention, some of it subtle and some of it not so subtle' (1995: xi). Breger and Hill hypothesise that mixed marriages are often treated with suspicion because they call into question important boundaries between 'self' and 'other' (1998: 9). In an attempt to protect those boundaries, the in-group may initiate a process of social stigmatisation, reinforcing negative discourses and stereotypes of the out-group. These discourses in turn influence the reactions of extended kin and the local community, and ultimately affect the social acceptability of an exogamous marriage (Breger and Hill 1998: 4; 12). The force of community supervision has impacted heavily on—often preventing—mixed unions in many different parts of the world, particularly in societies where endogamy was previously the norm. Societal hostility or ostracism may thus constitute a significant external stress factor vis-à-vis a mixed marriage.

The strength of community supervision will depend upon the particular circumstances, but one scenario in which the community is likely to act to prevent intermarriage is where there is racial conflict between groups. Typical examples include historical memory of racial (black-white) segregation in the US, and Aryan-Jewish discrimination in Germany during the Second World War. In the case of black-white unions in the US, censure could hail from either community involved. In terms of acceptance within the white community, the process of 'national homogenising' sometimes drew the minority partner into a generalised American identity, but at other times reception remained hostile. A black respondent to a study conducted in the 1970s confessed to being beset by anxieties, convinced that white men would attack him with fists and knives, make remarks, or spit at him if he were seen with a white girl on the street (Beigel 1975: 75). A black mother expressed the fear that her children, growing up in a white neighbourhood and therefore likely to date white partners, 'may be hurt, either by white parents who reject them or by the insults or violence of whites who see them together in public' (Evers 1975: 156). Meanwhile, negative reactions within minority communities could also work to make them push inter-married couples and mixed people away (Spickard 1989: 370). This was the case for a black musician in the 1970s, who would carefully avoid being seen in public with his white girlfriend through fear that black society and community organisations would sever all relations with him (Beigel 1975: 76). A female, white German respondent to the same study,

who married a black American soldier, narrated how her husband's black relatives and friends rarely talked to her, while she had nothing to talk about with them. As a consequence, she felt keenly disliked and ended by divorcing him because she felt rejected by his community (Beigel 1975: 72-3). Often, censure hailed from both communities at once. Also writing during the 1970s, Poppy cites a couple (black male, white female) who discovered that even total strangers would not allow them to 'bypass' colour:

The real world started crowding Bob and Katie. Her parents stonily refused to talk to him […] He lived in the uneasiness of things felt but not said—a look in the street from a white man, a black woman; an apartment door slammed in his face with the 'For Rent' sign still on it (1975: 158).

Memories of interracial conflict can take an extremely long time to heal. As recently as the mid-90s, one black American in a mixed union commented:

If there are problems, it is not the couple as such that causes them. Problems are caused for them by other people. Interracial marriages do not cause problems. Small-minded, bigoted individuals with their attitudes cause the problems (Davenport and Davenport-Zanolli 1994: 260).

In the case of Second World War Germany, just weeks after Nazi flags went up across Berlin, German friends and colleagues suddenly turned cold towards Jewish peers they had previously played cards and drank with, and German-Jewish intermarriage became the target of social stigmatisation (Stoltzfus 1996: 38).

Community reluctance towards intermarriage may be further accentuated where spouses derive from countries engaged—or previously engaged—in wars or occupation of one by the other. This was the case in the US following the incarceration of Japanese Americans during the Second World War. While some *Issei* (first-generation) Japanese Americans 'were not shy about harassing others who did not share their prejudice', others ostracized those who married non-Japanese on racial, hereditary or physical grounds (Spickard 1989: 61). Such difficulties also feature in contemporary Greek-Turkish unions, where enduring memories of Ottoman expansionism (1458-1830) and the Turkish invasion and division of Cyprus in 1974 mean that attitudes are often marked by contradictions and ambiguities (Petronoti and Papagaroufali 2006: 557).

Religious differences are often a key motivation for community censure of marriage choices, particularly in the case of Judaism and Islam. As expounded by Rabbi Jonathan Sacks, Chief Rabbi of Great Britain, 'Judaism

#### 308 chapter six

is not a private faith of individuals, it is embodied in communities, above all the family' (cited in Zemmel 1999: 63). According to this view, the very survival of the Jewish religion depends upon the decision of Jews to marry other Jews. As a result, first-generation Jewish immigrants to the U.S who chose to inter-marry were often shunned by their communities, and some lost their places in Jewish burial societies and synagogues. In general, Orthodox Jews were more likely to condemn than Reform Jews, underlining the extra difficulty attendant where one or both partners hail from a strictly observant community (Spickard 1989: 186-7). Islamic societies similarly seek to organise the minute practices of family and social life through the collective enforcement of public morals, thus the power of community supervision can be equally great (Ayubi 1991: 35). The Arab-Islamic culture traditionally laid more emphasis on 'external' than on 'internal' moral enforcement, favouring precautionary safeguards over internalised prohibitions (Ayubi 1991: 37). For this reason, most Islamic manifestos include a detailed account of the moral precepts that the public is to observe and oversee collectively, especially in the area of sex, women and the family. This concept of public and collective (rather than private and individual) morals is partly related to the importance of shame-related values in Arab society over guilt-related ones. Each person is expected to take an active stand in 'enforcing good and prohibiting evil' (Ayubi 1991: 42-44).

Finally, socio-economic competition related to the distribution of resources among groups may invite potential for community opposition to intermarriage. Indeed, under 'extreme conditions of intolerance', it can make intermarried individuals and their offspring vulnerable to violence and even elimination, as in the case of ethnic genocide in the former Yugoslavia and Rwanda (Hartley 2010: 239). Individuals of mixed Hutu-Tutsi background continue to be vulnerable to violence from militias and neighbours in contemporary Rwanda, and inter-ethnic marriages have become less common (Hartley 2010: 233).

In many cases, community opposition serves to precipitate the demise of a mixed relationship; however, like parental excommunication, it can also strengthen the resolve of those involved. For example, where only one community rejects the match, this may engender resentment in the embattled individual, leading them to draw closer to the 'outside' community. In Spickard's study, one *Nisei* (second-generation) Japanese American respondent who dated white boys reflected:

The *Nisei* […] made fun of me because I still went around with *hakujin* [White] and *kurombo* [Black] kids. None of the other Nisei did that at all. They just stuck in their own group with their noses up in the air as if they were so good. They made me sick' (1989: 65).

Similarly, in contemporary Guyana, Indian women's experience of expulsion from their Indian communities following their exogamous marriage has alienated them from Indian cultural traditions and values (Shibata 1998: 97). Alternatively, some lovers may begin to conduct their relationship with increased discretion in reaction to unpleasant comments, and this may act to consolidate their relationship (Barbara 1989: 16). Indeed, feelings of fascination and attraction to difference may be magnified where social pressure is marked (Barbara 1989: 21).

In Xinjiang, there is a fairly long history of community supervision of marriage choices in general, and opposition to intermarriage with Han in particular. Documenting the evolution of Chinese policies towards ethnic intermarriage in Xinjiang during the last century, Li Xiaoxia provides many historical examples where the attempt by a Han male to take a Uyghur wife resulted in rapid and grave local conflict. The most famous of these is of course the case which led to the Qumul rebellion in 1931 and the demise of the warlord Jin Shuren's regime (2006: 76). Wary of the frequently adverse reactions among Uyghur communities to Uyghur-Han unions—including verbal humiliation, beatings and threats of murder—in 1946, General Zhang Zhizhong of the National Revolutionary Army of the Republic of China issued a legal prohibition against Muslim-Han intermarriage on the grounds that it led to 'ethnic friction and disunity' (Li 2006: 76).

Notwithstanding the above, other evidence suggests that the frequency of intermarriage and corresponding levels of community supervision in Xinjiang have fluctuated over time. For example, following the end of the Yaqub Beg regime (1864-1877) in Qäshqär, which strictly enforced Islamic morals, although the higher echelons of Uyghur society continued to adhere strictly to Islamic values, the masses are said to have lapsed once again into 'laxity'. Among women, this laxity was reflected in the tendency to go about unveiled and in the preparedness of some to marry a Chinese (Bellér-Hann 2008: 312-313). Local opinion in contemporary times has it that Uyghurs and Han Chinese intermarried intermittently in the past, particularly in the eastern oases, this apparently explaining why Uyghurs from Turpan are more likely to physically resemble Han Chinese. One respondent, a middle-aged male restaurateur named Adil, maintained that some Uyghur girls had been 'quite up front about going with Han soldiers' during the Guomindang period (1911-1949).

In the 1950s, while Muslim-Han intermarriage was not prohibited, it was nonetheless treated with kid gloves, and marriage permissions were not given in cases where families or communities opposed it. Only in the 1960s did the policy begin to relax slightly, with registry offices in some districts adopting a positive view on intermarriage as a reflection of 'true equality and unity'. Still, many couples were forced to separate by family and friends, or else had to elope to get married (Li 2006: 77-80). By the early 1980s, growing numbers of Uyghur and Kazakh young people had begun to seek marriages outside of the in-group, but local opposition remained fierce and policy cautious. In 1982, official procedure included two steps: first, the couple would be asked to reflect carefully on their decision; if they persisted, the state sought to protect their individual rights and prevent 'thirdparty interference'. Where a couple demonstrated true love and perseverance, the state tried to facilitate the match by 'persuading' the parents.2 However, families frequently resisted by withholding documents the couple needed to register their marriage. Owing to the high level of community opposition, the authorities were obliged to pass a law in 1986 allowing those who could not produce the relevant documents to marry following a series of alternative official checks (Li 2006: 80-81).

From the 1990s onwards, the Uyghur community has if anything become increasingly hostile to Uyghur-Han intermarriage. A Uyghur research assistant involved in Yee's turn-of-the-century study of ethnic relations in Ürümchi insisted that he would not allow his daughter to marry a Han Chinese lest he and his family were 'looked down upon by his fellow people' (Yee 2003: 450). Similarly, a contemporary Han scholar records how the relatives of a Uyghur woman in Turpan who took a Han husband would regularly hunt down the pair and beat them in an attempt to force separation, with the result that the couple finally left Turpan altogether (Zhang Suqi 2005: 25). My own interviews suggest that, currently, few Uyghurs dare to pursue inter-ethnic relationships with Han people openly lest they encounter public disapproval. As Adil, a middle-aged male restaurateur, put it in 2002: 'If we see a Uyghur marry a Han and see them together in the

<sup>2</sup> Methods of 'persuasion' had been used with the opposite aim by Nazi Germany in 1935, when the Gestapo (internal secret police) intervened in the marriage registration process and requested registrars to inform them of proposed German-Jewish marriages so that they might 'enlighten' the Germans involved (Stoltzfus 1996: 47).

street, we feel anger in our hearts; we feel it is shameful [Uy. *sät körünidu*].' Uyghur-Han couples who appear together in public risk the verbal and sometimes physical (violent) censure of the Uyghur community. In 1995, I interviewed a 19 year-old girl from Ürümchi who had a Han boyfriend. She told me: 'It's really hard for us to even go out anywhere. If other Uyghurs see us together in public, they give us trouble. If Uyghur men see us together on a public bus, they swear at us and hit us. Uyghur women aren't so bad, but they still make comments.' Mixed couples are especially likely to encounter public disapproval in the Uyghur-dominated areas south of the Tianshan (Heavenly Mountains), and in certain urban districts where Uyghurs are concentrated (e.g. Döngköwrük in Ürümchi). In 2004, Patigül, a music trader in her late twenties based in Döngköwrük, was astonished when I ventured to say that I had occasionally seen mixed couples walking around in the regional capital, exclaiming:

What?! Were they married or courting? Well, you certainly don't see them here; they can't walk along together here. They would get sworn at. People would hurl jibes at them and ask the girl 'What's wrong? Couldn't you get a Uyghur boyfriend? [Uy. *Sanga chiqmidi ma?*]'

Nurmämät, a *minkaomin* kebab trader in his early forties speaking in the same year, supported this assertion, distinguishing between the centralnorthern Han-dominated parts of the city and the Uyghur-dominated south-east: 'If a mixed couple walks along here [north-west], nobody says anything, because there are lots of Hans around here. But if they walk along in Döngköwrük [Döngköwrük: south-east], people will swear at them, even hit them.' The distinction was further qualified by Gülshäm, twenty, a *minkaohan* student interviewed in 2004, who drew the physical line at Nanmen (South Gate). According to her, mixed couples who appeared together south of this area were likely to be openly cursed; at the very least, onlookers would curse them under their breath. However, she added that the problem was not so acute in the Beimen (North Gate) area of central Ürümchi.

Perhaps worse than the disapprobation of strangers is the dismay of friends and colleagues in the Uyghur community, who will employ all means necessary to 'persuade' Uyghur peers to abandon a mixed union. In 2002, Zunun, a *minkaomin* service industry employee in his thirties, narrated the story of an inter-ethnic courtship between two co-workers, a Uyghur female (*minkaomin*) and a Han male. He described how all the Uyghur staff members, who worked on different floors of the building, had

#### 312 chapter six

joined forces in urging their Uyghur friend not to marry this Han colleague, even though she insisted that she loved him. By the time I returned to Ürümchi in 2004, their relationship had broken up. In this way, the most effective form of community supervision may come not from strangers but from one's colleagues and friends.

The pressure of selective endogamy impacts with equal force upon *minkaohan*, who speak Chinese as their first language and interact mainly with Han at school, thus in theory are more likely candidates for a mixed union. Despite being visibly acculturated in dress, hairstyle and manner, and speaking poor Uyghur but fluent Chinese, Burkhan, a fifteen year-old *minkaohan* boy, insisted in 2004 that he would never marry a Han peer. His explanation for this position centred on the potential for social ostracisation that the act entailed: 'If I did [marry a Han girl], no Uyghur would come to visit me or be friends with me; no-one would have anything to do with me. That's what my parents say.' In a third case, Gülshäm, twenty, a *minkaohan* student, described in 2004 what happened when she broached the subject of a Han boyfriend with a *minkaomin* friend:

I asked my best friend—we met after I started university—what she would do if I got myself a Han boyfriend. She was furious, and said she would kick me in the face! … that the friendship wouldn't be worth having if I did a thing like that … that she would break off our friendship and have nothing to do with me. She said: 'What on earth are you thinking of, going with *that*  kind of person? I would really look down on you if you did.'

When I asked why her friend should have such an extreme reaction, she referred negatively to the insularity of her in-group, as she perceived it: 'Uyghurs think that Hans eat pork and therefore that they're dirty, you know … Uyghurs' 'ethnic thinking' [Ch. *minzu guannian*] is very strong. They set themselves apart, they're stand-offish; they exclude the Hans from their lives.' This exchange between two young women clearly demonstrates the force of community pressure as boundary maintenance, employed in this instance to prevent inter-ethnic courtship and indeed any kind of ethnic interaction with Han Chinese.

In the second of the above two cases, Gülshäm's irritation with the perceived narrow-mindedness (as she perceived it) of her in-group led her to applaud the bravery of other young Uyghurs who openly defied the taboo. I observed this directly one day as we sat on a restaurant terrace in central Ürümchi. Espying a mixed teenage couple who happened to pass by with their arms draped around one another, she was all smiles and effusively praised their courage. On another occasion, she related her surprise and

pleasure upon sighting a young mixed couple 'holding hands in Nanmen (South Gate) of all places!'

Anticipation of negative community reaction led Osman, a market trader in his thirties from Aqsu betrothed to a female Han co-worker, to opt for a quiet wedding away from friends and relatives in the early 2000s. His close friend, Zunun, confirmed in 2002 that the wedding had not been the usual, huge social event continuing for many hours, conceivably because Osman had felt too embarrassed to cement the match—and transgress boundaries—in front of his Uyghur peers. His wife, too, had foregone the banquet that traditionally surrounds a Han wedding ceremony. Moreover, the pair continued to be affected by community sanctions in their married life. In 2004, a friend of Osman, a trader originally from Khotän who also worked at the market, voiced strong disapproval of Osman's choice of mate:

He's a good man, he has a good heart. But we don't approve of him marrying that Han woman. She is not a Muslim, and she won't agree to take the religion. There's nothing to be done. Why are they together? They just play with this [points to mouth] and with this [points to genitals]. If you put a man and a woman in a room together, what do they do? That's all it's about.

He thus defiled the couple's relationship by describing it in purely sexual terms.

Petronoti and Papagaroufali (2006) reported the difficulties of investigating the private lives of Greek-Turkish mixed couples, who were hyperconscious of prevailing public suspicion towards 'mixture'. They describe how respondents sometimes adopted 'evasive modes of speech', becoming defensive, and accusing researchers of seeing only 'problems' with their relationships, while at other times they welcomed the opportunity to publicise their views (2006: 561). In Xinjiang, it proved difficult even to obtain direct interviews with intermarried couples. Perhaps owing to his anticipation of negative community reaction, the market trader Osman normally avoided discussing his mixed marriage, both with Uyghur friends and with me, despite having known me for some ten years. As a result, I often had to rely upon the accounts of family members and friends. I was able to interview just one middle-aged man in a mixed union directly, in the year 2002. This man was of mixed Central Asian and Han Chinese heritage, and had in 1983 married a first-generation *minkaohan* Uyghur woman, who had been educated in Chinese during the Cultural Revolution period. The couple had been together for twenty years, and had a son of eighteen. The

#### 314 chapter six

husband denied that the couple had encountered problems in the community when they got married, but conceded that marriages such as theirs were extremely rare. When I enquired whether other people ever made comments about their marriage, his response was defiant and defensive: 'Well, what can *they* say? It's got nothing to do with anyone else, only with us two.' He thus stressed the individual, independent character of love and romance over Uyghur (Central Asian) traditions of marriage. This couple may have been less targeted for criticism by peers owing to the fact that the husband was only half Han; there was therefore potential to rationalise the match by placing emphasis on his Central Asian rather than his Han Chinese heritage.

# 6.4. *Perceived Barriers: Racial, Cultural, Religious, Historical, Political*

Intermarriage typically involves factors such as demography, class, race, culture, religion, history and politics. Whereas many earlier studies of intermarriage focused on structural factors like demography and class, later studies expanded this basic framework to incorporate cultural difference, specifically, the mutual images held among and between communities (Spickard 1989: 6). In some contexts, barriers to mixed marriage were explicitly characterised as racial. In the US, parents of *Nisei* Japanese Americans framed their opposition to interracial unions in terms of Japanese racial superiority and conflicting cultures, including modes of living and philosophy (Spickard 1989: 62). Then, around 1970, a wave of pan-Asian activism on West Coast college campuses discouraged social contact between Asian and non-Asian students, leading some of those dating non-Asians to break up with their partners through an obligation they felt to the Asian race. This represented a response to a century of perceived oppression, ranging from wartime incarceration of Japanese to negative cultural stereotypes and limits on occupational advancement (Spickard 1989: 68).

Meanwhile, environmental resistance to black-white unions was variously attributed to class, racial prejudice, fear, culture (perceived lack of acculturation), feelings of sexual inferiority (especially in relation to black men) and biological differences (Beigel 1975: 69-70; cf. Hernton 1975: 151-2); but, of these, the racial barrier or 'colour line' loomed largest. In earlier times, the white community had outlawed unions with black people and even legislated against them; later, the Black Pride movement led many

within the black community to strive for endogamy in what Spickard described as a form of 'compensatory race hatred' (1989: 300). An aggressive pride in blackness that grew in the years following Reconstruction (1865- 1877) led many to reject interracial marriage out of a desire to build a strong, confident, internally integrated, economically sound, and politically protected Black people (Spickard 1989: 298).3 The activist William Du Bois, writing at the turn of the nineteenth century, opposed 'the wholesale intermarriage of races during the present generations' as a potential social calamity based upon broad cultural, ethical and traditional differences (Spickard 1989: 298). This theme was continued in Marcus Garvey's 1924 manifesto 'What We Believe':

The Universal Negro Improvement Association […] is against miscegenation and race suicide. It believes that the Negro race is as good as any other, and therefore should be as proud of itself as others are … It believes in the purity of the Negro race and the purity of the white race …' (Spickard 1989: 300).

With the rise of the Black Power movement in the 1960s, integration into the White world ceased to be important altogether as the Black Panther party exerted pressure on politically oriented young Blacks not to date or marry non-Blacks (Spickard 1989: 303):

Let the white man keep his women, and let us keep ours … . Our women are the most beautiful … .what would we do married to a white woman? Her people don't want you in the neighbourhood around them, and our fast awakening people don't want you to bring her back into our neighbourhood any more to live with us (Malcolm X, cited in Spickard 1989: 303).

Although the movement stopped short of banning interracial romance altogether, a black man dating a white woman was no longer awed as a taboo-defier; rather, he was criticised for shaming his own race (Downs 1975: 161).

In extreme cases of racial division, individuals who are themselves free from racial prejudices and seek to intermarry with a member of a stigmatised group may be rejected by their families. Such was the experience of altruistic Germans who sought to marry Jews after anti-Semitism became the official state language of Nazi Germany (Stoltzfus 1996: 18). One German woman was greeted with a furious response in 1935 after informing her mother that she intended to marry a Jewish man. The latter responded:

<sup>3</sup> When civil rights campaigner Frederick Douglass married a white woman in 1884, reactions from the black community were mixed. Some defended him; others were shocked at his 'betrayal of their race' (Spickard 1989: 297).

#### 316 chapter six

'Incredible! You can't marry a Jew at this point!', before ordering her daughter never to show her face in the family home again (1996: 50-51). Subsequently, Charlotte's elder sister's husband—an SS officer—threatened to have her Jewish partner arrested as the couple made their way to the marriage license bureau (Stoltzfus 1996: 52). Historical memory of racial and religious persecution later led Jewish Americans to state the 'desire to maintain ethnic solidarity' as a principal reason for opposition to intermarriage (Spickard 1989: 186). A similar example is provided by Khatib-Chahidi et al (1998: 55), who cite a case in which Jewish parents disliked a future sonin-law because he came from a country with a history of imperialist and anti-Semitic politics. In such contexts, mixed couples and their children may be abused, harassed or even molested by groups on both sides for 'betraying' their culture, race or religion (Breger and Hill 1998: 23).

Religious differences can prove an especially strong barrier, with the potential to provoke strong opposition from communities as well as conflicts within couples. Observant persons who practise their religion regularly will normally experience more difficulties than those who do so only occasionally (Barbara 1989: 209). Yet even for individuals who are not especially observant, religion can nonetheless affect attitudes, expressions and the way in which people adopt positions (Barbara 1989: 172). Of the world religions, Judaism and Islam maintain especially strict norms with regard to marriage outside the faith. In Judaism, halachic law dictates that the child of a Jewish mother is Jewish, regardless of the father's faith (Zemmel 1999: 51-52). This means that Jewish women often face less opposition to intermarriage than Jewish men. While familial opposition to intermarriage can in some situations lead to trauma and the termination of the engagement, the conflict may be mitigated if the Jewish female promises to raise the couple's children as Jewish (Zemmel 1999: 52). In Islam the reverse is true in that, according to the Quran, Muslim males may intermarry, provided that spouses follow 'religions of the book' (that is, Christianity or Judaism). Muslim females, on the other hand, may marry only within the Islamic religion. This norm derives from the system of patrilineal descent, which rules that the offspring of a marriage take the name and religion of the father (Yamani 1998: 154; Li Xiaoxia 2006: 75-76; Edgar 2007: 590). For this reason, in the Soviet context it was usually Russian women who married Central Asian men and 'introduced lace curtains and cabbages to Central Asian villages' (Edgar 2007: 592). This norm is often perceived as a double standard by Muslim women, especially the highly educated. For example, Brown and Farahyar document the anger of Middle

Eastern females studying in America regarding the male expectation that Muslim women conform to traditional patterns of courtship behaviour (1994: 174). In Greece, we find an example of a case where two religions may simultaneously resist interfaith unions. Here, because Orthodox Christianity has been closely linked to the formation of the Greek nation, endogamy is applied energetically when religion is at stake, at least in theory. For Orthodox believers, conversion of Muslim partners is considered the only acceptable solution, while Greek Muslims in the northwest tend to refrain from romance with Orthodox Greeks, or keep it secret, lest they be banished from their own communities (Petronoti and Papagaroufali 2006: 559). In practice, however, Greeks who do intermarry with Turks rarely request that Turkish partners convert to Christianity. Rather, couples manage religious differences by opting for a civil wedding and declaring that they are 'not religious' (Petronoti and Papagaroufali 2006: 574).

Another decisive factor controlling the frequency of intermarriage is history or, more specifically, historical memory of inter-group conflict in the form of war or occupation. As Barbara noted: 'Opinions on mixed marriages […] may be very different depending on the period in history through which a country has passed and on the time at which they are expressed' (1989: 28-9). The Second World War and its aftermath prevented Franco-German unions in Europe for many years (Barbara 1989: 25), and drove a 'wedge' between the races in the US, where non-Japanese would neither date nor marry *Nisei* (second-generation) Japanese Americans (Spickard 1989: 71). In China, the 'foreign imperialist' invasions of the late 1800s (the Opium Wars) meant that intermarriage between a Chinese and a Westerner was considered 'shameful for the individual and for the country' and liable to lead to 'national subjugation' (Dikötter 1992: 58). In contemporary times, Chinese girls continue to be routinely discouraged from marrying Western men based on past evidence that a 'foreign devil' cannot be trusted. A third example perhaps more comparable with the Xinjiang context is found in Algeria, where anti-colonial resistance precluded intermarriage with the French during the early twentieth century. In the Algerian case, French colonial policy opposed mixed unions (Barbara 1989: 25), whereas in contemporary China the official line on intermarriage between China's 56 nationalities is positive, centring on its role in the promotion of ethnic unity and territorial incorporation. A fourth case concerns Greece, with its history of conquest by Turkey. A Greek respondent in one study of Greek-Turkish intermarriage stated that he disliked Turks in general as 'invaders in his country', yet was prompted by business trips to Turkey and interac-

#### 318 chapter six

tions with Turkish artists while living in New York to question his social biases (Petronoti and Papagaroufali 2006: 565). Historical memory may be especially potent in preventing intermarriage when combined with adverse political, social, cultural, religious or economic conditions in the present. In contemporary Guyana, a history of racial conflicts has combined with ongoing economic rivalries to inhibit the development of intimacy between Africans and Indians (Shibata 1998: 84-6). In the US, a history of racial oppression combined with enduring political and socio-economic inequalities has impeded black-white intermarriage across three generations.

Finally, the politics of multi-ethnic societies may underlie minority group adherence to strict endogamy. Plural societies are normally characterised by the segmentation of culturally heterogenous groups, each of which is 'in a position of institutionalised inferiority vis-à-vis the dominant group'; this pattern may be the result of colonial expansion or large-scale migratory movements (Cerroni-Long 1984: 28). In the context of ethnic revitalisation movements—ethno-political incorporation—marriage with an out-group individual may be similarly out of the question (Cerroni-Long 1984: 41; 43).

So: which factors are decisive in outlawing Uyghur-Han intermarriage in contemporary Xinjiang? According to Mamet, Jacobson and Heaton, several mitigating factors operate to reduce the frequency of ethnic intermarriages between groups in China, including history of inter-group conflict, religion, language, culture, normative sanctions and geographic isolation. Among the Uyghurs specifically, they found that ethnic consciousness, defined in terms of ethnic pride and the desire to preserve Uyghur culture and resist Han dominance, was most important (Mamet, Jacobson and Heaton 2005: 187, 202). Chinese scholar Li Xiaoxia provides a broadly similar list of factors, comprising place of residence, history and traditions, language and religion (2004: 21). While he does not mention ethnic conflict explicitly, presumably owing to political sensitivity, one can assume that this issue is subsumed under the factor 'history'. Li categorises mixed unions according to the concept of 'intermarriage pools' (*tonghunquan*). Within the Northwest 'intermarriage pool' situated in Xinjiang, there are two primary triadic relationships, the first consisting of Uzbeks, Tatars and Uyghurs; the second of Uzbeks, Tatars and Kazakhs. Within this pool, religion is the most decisive factor when choosing spouse (Li Xiaoxia 2004: 26). Another study, which adopts intermarriage as a key variable in order to explore ethnic relations, concludes that obstacles to intermarriage in

Xinjiang include religion, language, tradition and culture (Yee 2003: 437). This echoes the observation by another Chinese research team that among the nationalities in Xinjiang 'there are many marital taboos that derive from different cultures, customs, and religions' (Ren and Yuan 2003: 99). Crucially, Yee's study also highlights the problems of mutual prejudice and distrust (Yee 2003: 437).

Among my respondents, interviewed between 1995 and 2004, all of the factors discussed in the general literature above emerged, to different degrees. In Xinjiang, race has not been a recurring theme when constructing boundaries; rather, criteria such as religion, cultural practices, socio-economic rivalries and ethnic/political loyalty (versus betrayal) have assumed principal importance. That said, some prejudices held by Uyghurs regarding Han people do involve a racial—or at least a physical—theme. The vast majority of respondents compared Uyghur looks positively against those of Han people, stating that Uyghur men and women were 'far more beautiful' than their Han counterparts. By contrast, almost all *minkaomin* and some *minkaohan* respondents claimed to find Han Chinese facial characteristics 'ugly' or 'repulsive'. Some went further in comparing those characteristics (in particular the nose) to the facial characteristics of pigs; here, the link to the 'unclean' Chinese diet is clear. Bahar, eighteen, the *minkaomin* daughter of a first-generation *minkaohan*, described in 2004 a mixed relationship she had observed at her high school in Ürümchi: 'There's this one Uyghur guy—he's really good-looking—going out with this ugly [Uy. *sät*] Han girl. I don't like it at all [Uy. *män bäk öch*]!' She pulled a face of disgust as she spoke. Gülhärä, twenty, a *minkaomin* student from Döngköwrük interviewed in the same year, ranked physical appearance third (after religion and cultural practices respectively) in her list of reasons why Uyghurs should not marry Hans: 'We don't like the way they look! It's offensive [Uy. *sät körünidu*]4 to see a Uyghur and a Han together!' She ended her speech with this rhetorical question: 'Come on, would *you* let your son or daughter marry a Han?' Later that same day, as we ate dinner in a Uyghur restaurant on the east side of the city, I observed her reaction after she spotted a mixed couple at a neighbouring table. The Uyghur woman sported a modern, urban appearance, while her male Han partner was apparently a rock fan, as he wore his hair very long. Gülhärä spent a considerable amount of time carefully ascertaining the nationality of each, before concluding with distaste that '*Uyghur* men never grow their hair that long, only

<sup>4</sup> This phrase translates literally as 'looks ugly'.

#### 320 chapter six

just to the shoulder'. The Han appearance was often the butt of Uyghur jokes. On another occasion in 2004, I held a conversation with two young men who came to eat lunch at a restaurant near the Nanmen mosque after emerging from midday prayer. I asked Äsäd, a *minkaomin* university graduate in his twenties, who had chosen to go into independent business in Ürümchi, what he thought of the increase in Uyghur-Han intermarriage. His reply was: 'What do I think of Uyghurs marrying Hans? I hate it! [Uy. *öch!*]' As a rejoinder, his friend Ablät (whose facial features were closer to those of Han than to those of Uyghur) joked: 'Do I look like a Han to you? I do, right? So he must hate me as well!'

As detailed earlier in this chapter, at times of relatively strict religious observance in Xinjiang, such as during Yaqub Beg's rule, willingness (or refusal) to intermarry has provided a yardstick by which to measure Islamic piety and moral standing in the community, particularly with regard to Uyghur women. Today, the factor most often cited to explain the 'impossibility' of Uyghur-Han mixed unions is again religion: the fact that Uyghurs are Muslims—although individuals may be observant to varying degrees while Han are perceived either as atheists (having no god) or as Buddhists (worshipping a multitude of deities rather than a single one). Yet it is noticeable that the way in which this barrier is articulated has altered over time. In interviews I conducted in the mid-nineties, respondents invariably placed emphasis on differences in Islamic social practices rather than on religious observance *per se*. One young Ürümchilik woman explained in 1995:

We believe in different religions. Before, there was intermarriage, but nowadays there is none. Once a couple gets older, they start to realise their customs and practices are different. And their religions are different. One person says one thing and the other says something else. They can't agree. It almost always ends in divorce.

In this way, Uyghur national customs were presented as being inseparable from Islam, and past experimentation with inter-faith marriage as proving that a mixed union with a non-Muslim can rarely succeed. Liu Lan, a 20 year-old mixed-heritage woman5 whose parents had divorced twelve years before, explained in 1996 how 'national sentiment had proved too strong' on the side of her Uyghur mother, causing her to separate from her Han

<sup>5</sup> In fact, she described herself as an *erzhuanzi*. This is the derogatory term designating children of mixed parentage, equating to 'half-breed'.

father. Many older Uyghurs had also observed mixed marriages fail in their lifetimes. A *minkaohan* woman in her forties from Ürümchi concluded, also in 1996: 'Two separate races of people still have areas that are very different at the end of the day. These differences cannot be resolved with love.' Thus, even though *minkaohan* are generally supposed to be linguistically and culturally closer to Han people, most nonetheless rejected the option of Uyghur-Han intermarriage.

In later interviews conducted in 2002 and 2004, most *minkaomin* respondents (including young, educated individuals, who are in other contexts more likely to consider mixed courtships) named Islam as the principal factor preventing intermarriage with Han people. Yet, more often than not, religion *per se* (observance of prayer, ritual) was now ranked above everyday religio-cultural practices in importance. This clearly reflected the general climate of religious renewal in the region (see Chapter 5). Uyghur respondents continued to mention divergent cultural practices, habits and customs as factors impeding intermarriage with Han Chinese, but these were now secondary concerns after religion itself. Adil, a 45 year-old male restaurateur, explained in the same year: 'Mixed marriage is OK for the first few years, but when you get older … our religion is different, temperament, diet, manners. And, when we die, we do it differently, there are different rituals … .' Patigül, a *minkaomin* music trader in her late twenties speaking in 2004, elaborated further on religion as the primary barrier:

We're not able to marry Hans because we're Muslims. That's the No.1 reason. Our religions are different. They worship the Buddha; we pray to Allah. You know, our religion and Sharia law entail a lot of religious rituals … you know, like the small wash and the big wash. The Han [partner] would have to become a Muslim completely; then it [marriage] would be fine; but not otherwise.

In this way, Patigül emphasised the distinction between mono-theistic and poly-theistic religions.

Others signalled the dietary requirements of Islam to explain their rejection of Uyghur-Han intermarriage. Yultuz, twenty, a *minkaomin* student working in a musical instrument shop in Döngköwrük, told me in 2004: 'I wouldn't consider marrying a Han. We Uyghurs don't marry with Hans because we're Muslims; they eat pork.' Similarly, Bahar, eighteen, the *minkaomin* daughter of a first-generation *minkaohan*, expressed revulsion at the sight in 2004 of mixed couples eating together in public:

#### 322 chapter six

I was at my favourite restaurant the other day and there was a mixed couple there, a Uyghur guy with a Chinese girl. They were making me feel sick, sharing food with each other … because she's not a Muslim.

Where some young people held less hard-and-fast views in respect of interethnic courtship, they were swiftly chastised and reminded of the religion factor by their peers. For example, Rabiyä, twenty, a *minkaomin* student speaking in 2004, initially stated that she did not oppose Uyghur-Han unions. However, she modified her view after her eighteen year-old friend Aygül(also *minkaomin*) protested: 'Hang on, though, at the end of the day, Uyghurs aren't really supposed to marry Hans because we're Muslims.' Under peer pressure, she conceded that the different diets of the two groups made intermarriage difficult. Yet the language employed by both women suggested a potentially higher level of tolerance. Another *minkaomin* respondent, Mälikä, a twenty year-old student who secretly worked as a hostess, admitted in 2004 that she had for some time considered marrying her Han 'manager' (pimp). This person had promised to become a Muslim if she agreed to the wedding, and she personally felt that Xinjiang-based Han people were quite used to Islamic dietary restrictions. She pointed out that many Hans who had been born and raised in the region ate only *qingzhen* (halal) foods. Were her partner to make these compromises, she had been prepared to consider marriage as a possibility. However, here again, a friend had talked her out of it, arguing: 'You *know* Uyghurs can't marry Hans.' Mälikä explained her eventual decision to separate with her Han partner as follows: 'Even if he says he'll enter the faith, you can never be sure whether he has really done so in his heart.' In this way, young Uyghurs on the whole continue to recognise religion as the primary barrier to Uyghur-Han intermarriage. This situation apparently differs from that in contemporary Tibet, where, according to one Chinese scholar, young Tibetans increasingly have worldly desires, pay less attention to religion, and enjoy independence from parents and relatives (Xie Lei 2006: 148). Clearly, religion constitutes an objective, defensible ground upon which to oppose intermarriage with non-Muslims. However, the situation is not as clear-cut as the above testimonies would suggest. For one thing, Uyghurs are prepared in some circumstances to waive the requirement to be a Muslim where a proposed intermarriage involves, say, an American or European rather than a Han; conversely, they remain loath to marry with Hui, even though the Hui might in theory be treated as Muslim brothers (see Section 6.7 below, on hierarchies of intermarriage). The extra ingredients serving to shore up

the taboo on Uyghur-Han—as well as Uyghur-Hui—unions are historical memory and politics (including socio-economic competition). I shall deal with these together below.

Arguably, many Uyghurs resisting intermarriage with Han people in Xinjiang today are doing so out of the desire to (symbolically) resist Chinese colonial policies against a background of historical injustices (older generations remember the draconian policies towards minority religions pursued during the Cultural Revolution with clarity) and a framework of contemporary inequalities: increasing encroachment on the territory Uyghurs without exception consider as their homeland via mass population transfer, escalating socio-economic competition, and exploitation of local resources. Chinese scholar Zhang Suqi is perhaps unique among his peers in recognising political and economic differences between groups (in addition to cultural ones) as lying at the root of group attitudes towards ethnic intermarriage (2005: 24). While few of my respondents were prepared to state openly that politics underpinned the current taboo on Uyghur-Han intermarriage, it was sometimes possible to glean this implicitly from their responses. Möminjan, a politically aware young man from Turpan studying in Beijing in 2004, was scathing when I suggested that, like some Uyghur students based in China proper, he might select a Han partner: 'Will I look for a Han wife now I've been in Beijing for so long? No, of course not, because Hans are not Muslims! That's the first and main reason.' However, he went on to add: 'There are many reasons … Secondly I don't *like* them!' I gently pushed him to elaborate, but he declined to do so, largely because he had become intensely paranoid about being overheard in the public eatery in which we sat. Even though we conversed in Uyghur, some time before a young Han male had sat down at a nearby table and rather pointedly asked me what time I made it. Möminjan had immediately suspected this man of trying to ascertain whether I used local Xinjiang time or Beijing time—and thus whether I was a local Uyghur or a foreigner. In the latter case, he might conclude that Möminjan was 'talking to a foreigner' and possibly revealing sensitive information (a politically subversive act). Uyghurs have been detained in recent times on precisely these grounds.

The next response, given in more secure surroundings, made the implicit issue explicit. Jelil, an observant university graduate in his late twenties, explained in 2002:

#### 324 chapter six

It's absolutely impossible to marry a Han. Why? Lots of reasons. First, customs and habits: there is a massive chasm between the Uyghurs and the Hans. We are completely different. Second, they are dirty, filthy people. Third, religion. The Hans have no religion. As such, they are very dangerous. Far more so than a believer of another religion, say, Christianity. They say they're Buddhists, some of them, but only a few are really Buddhists. The vast majority are faithless. And finally, there's [Han] prejudice; this is a major factor preventing intermarriage. Regardless of love, prejudice remains. And so, five or ten or even fifteen years down the line, the marriage will end in divorce.

Jelil's first reason covers differences in everyday customs and habits, which was a given for most respondents. His second highlighted the internalised stereotype that Han people are 'dirty' and 'filthy' (see Chapter 1 Stereotypes). That this insult was deliberately separated out from the first (cultural differences) and the third reason (religion) suggests that his perception was based on something more than either culture or religion. This emerges clearly when he states his fourth reason: Han prejudice against Uyghurs in the form of Great Han chauvinism in all its manifestations. Conceived broadly, this charge can be seen to encompass a wide range of ills connected to Chinese colonial practices in Xinjiang, including political domination (and political castration of Uyghurs), cultural assimilation (sinicisation), escalating socio-economic inequalities leading to Uyghur marginalisation, and appropriation of natural resources. Thus, it is Han chauvinism which is so intolerable, and which has led so many Uyghurs, particularly those with a sharpened political awareness, to resist Uyghur-Han intermarriage with such force.

#### 6.5. *Managing Intermarriage*

The general literature on intermarriage identifies various ways in which couples try to 'manage' intermarriage in difficult circumstances. Where religion is the main component in group identity, as often the case with Judaism and Islam, religious conversion can be the sole means through which intermarriage can take place. In Judaism, halachic law does not recognise marriage between a Jew and a non-Jew; on the other hand, conversion is a complicated process. Orthodox conversion can take up to seven years, and brings no guarantee of eventual acceptance by the Jewish community (Zemmel 1999: 71). Moreover, it is recognised even within Reform Judaism that interfaith marriage can place a great strain on a couple, leading to relationship breakdown and divorce (Zemmel 1999: 73).

In practical terms, conversion means that intermarriage may serve to augment the group and increase its influence, as opposed to bringing about its assimilation or disappearance. This fact was gradually recognised among Jewish Americans in the US. By the time the third generation came of age accompanied by a sharp rise in intermarriage rates—their parents had begun to reconcile themselves to mixed unions, especially where there had been a conversion. Instead of bemoaning the imminent extinction of the Jewish race, rabbis and lay people began to work to keep the intermarried connected to the faith, in the hope of reversing the perceived decline in the Jewish population (Spickard 1989: 193).

In Islam, cross-cultural marriage was traditionally welcomed where it simultaneously propagated the faith, and is thought to have been a 'significant agent of Islamization', especially in the outlying parts of the Muslim world which were not subject to military conquest (Yamani 1998: 153). This included the Eastern Turkestan region, where assimilation following intermarriage of Muslims and non-Muslims was an important factor in the advancement of the faith (Foltz, cited in Bellér-Hann 2008: 303-304). Since Islam works on the principle of patrilineal descent, it has been more important to convert an incoming male than to convert an incoming female. Thus, in Barbara's study, a Frenchman marrying into a Northwest African family was required to recite the *shahada*—profession of faith in Islam in order to be admitted to the religious group of his wife; this procedure was not necessary if a Christian or Jewish woman married a Muslim man (1989: 32). Based on the same principle, any offspring from the union must be raised as Muslims and take their father's Muslim name. For this particular Frenchman, conversion to Islam enabled deeper integration into his wife's society, and strongly influenced the way in which his children were brought up (i.e. as Muslims) (Barbara 1989: 14). This option may not be favoured by all potential spouses, but those who feel ill at ease in their own socio-cultural environment may be more prone to form bonds elsewhere (Barbara 1989: 14).

Among some Hui Muslim communities in China, the conversion through intermarriage of thousands of Han women and children to Islam was formerly viewed as a kind of 'peaceful *jihad* [holy war]' towards the achievement of the universal *Dar-al-Islam* (Pax Islamica: Islamic and non-Islamic territories under Islamic sovereignty) (Israeli 1984: 295). However, conversion must be thorough and complete if it is to be accepted. Writing on the Hundred Flowers movement (1956-1957), Bush notes that although

#### 326 chapter six

Muslims were able to coexist with Communism, certain boundaries—such as intermarriage with a Communist cadre—could not be transgressed:

Muslim imams were complaining about the anomaly of receiving Communist cadres into the Muslim faith [...] To be both a Communist and a Muslim was generally conceived to be an impossible contradiction. For a Communist to become a Muslim in order to fulfil a requirement for marrying a Muslim girl was regarded, therefore, as a mockery, especially since his marrying the girl was motivated in part at least by the Party's desire to blend with the Muslim community (1970: 283).

In contemporary Xinjiang, conversionary intermarriage has begun to be viewed as a relatively desirable outcome in an ambiguous climate where intermarriage trends are marginally on the rise against a backdrop of Islamic renaissance. For some Uyghur families, conversion has become the condition underpinning parental permission for a mixed union, so that prospective sons- and daughters-in-law have come under intense pressure to convert. Nurmämät, a *minkaomin* petty trader in his early forties speaking in 2002, knew of many Han husbands who had become Muslims. He explained that Uyghur parents would not agree to a Uyghur-Han marriage unless the Han partner consented to take the Islamic religion, expressed mainly in dietary terms:

We don't eat pork, so if the Han wife doesn't become a Muslim, neither the parents [in-laws] nor the husband will eat the food she cooks; similarly, if a Han husband doesn't become a Muslim, the Uyghur wife will end up becoming just like a Han.

In a direct reflection of the growing centrality of orthodox Islamic practice, the criteria affecting possibilities for Uyghur-Han intermarriage have undergone a distinct shift over the past decade. Ömär, eighteen, a *minkaomin* high school student, explained that, nowadays, intermarriage is allowed only if the Han partner becomes a Muslim. He identified religion as the influential factor controlling mixed unions, adding that whereas in the past Uyghurs had stressed differing customs and habits as the main obstacle, 'now it's all about religion'. Again, the growing importance of orthodox Islam as against the fading significance of contrastive cultural practices is noteworthy. This shift was also confirmed by Patigül, a music trader in her late twenties from Döngköwrük interviewed in 2004: 'We used to complain about how Hans used the same bowl to wash their feet and to wash vegetables [see Chapter 1 Stereotypes] … but not now. Now religion's the thing.'

Conversionary intermarriage is viewed by some Uyghurs as a cultural gain and may even be considered a symbolic, political victory. As a result,

Han partners who undergo conversion to Islam in earnest are often welcomed into the family and community. Ömär, eighteen, a male high school student, narrated in 2004 the story of one such union in his *mähällä* [neighbourhood] in central Ürümchi:

He [the Han husband] became a Muslim completely! He took a Muslim name—Jümäkhun—and prayed in the mosque, wore a *doppa*. Everyone respected him; no-one criticised the woman for marrying a Han. He wasn't a Han any more … he was just like a Hui Muslim! We have no problem about marrying other Muslims, an Uzbek, a Kazakh, a Hui, that's fine … but a non-Muslim is a different matter.

The adoption of a particularly religious name (Jümä or 'Friday', after this important day of communal prayer), the observance of prayer rituals and the respectful use of the appropriate headgear all underlined this man's commitment to the faith. As a result, he was respected by his Uyghur wife's community just as an observant Hui might have been, where a Hui is conceived as a person of mixed Islamic and Han Chinese cultural heritage. He however avoided the historical stigma associated with Hui, which paints them as untrustworthy 'go-betweens'.

However, where a religious conversion is merely superficial, relations in the community and between the couple are harder to manage. Zunun, a *minkaomin i*n his thirties working in the service industry, described in 2004 his brother-in-law's marital problems as follows:

My wife's older brother is married to a Han woman. But I think they may get divorced, because she hasn't learned the Uyghur language. The Uyghur parents wouldn't agree to the match at first, and the son cried and said that if he couldn't marry her, he would marry no-one. Yes, she has become a Muslim, or at least she feigns the appearance, you know, she says 'Amin' and all that … The other thing is that she won't go to Uyghur social events with him, when his friends invite him to weddings and so on. The parents don't want them to get divorced; they are urging him to teach her Uyghur properly.

# Zunun's wife later added in a follow-up interview:

My whole family will be very sad about this; they have tried hard to make it work. Though she's a Han, she's not a bad woman, she has a good heart. For it to work, she would need to attend weddings with him; when someone dies, she would need to go to *näzir* to pay respects with him. But she doesn't go anywhere with him, he always goes alone; and this causes him to lose face in front of his friends.

#### 328 chapter six

In this case, although the Han wife paid lip service to Islamic ritual in some situations, on the whole she had failed to integrate successfully within Uyghur culture and society, as evidenced by her inability to speak the Uyghur language apart from a few basics, and her failure to attend Uyghur social events and important life cycle rituals. Clearly, her reluctance to accompany her husband to life cycle rituals was closely connected to her linguistic limitations, since the *lingua franca* at such events is invariably Uyghur. In this way, religious conversion in Xinjiang requires not just religious commitments but broader linguistic and cultural commitments as well. In a second case, Patigül, a music trader in her late twenties from Döngköwrük, confessed in 2004 that she always felt uncomfortable when she visited her friend's [married to a Han male] house in Tieluju. Her anxieties centred on the feeling that the house was somehow unclean. She explained: 'I don't really feel like eating anything when I'm there; I don't know, the house looks like a Han house, you know, the décor … and so I don't like to stay long.' Thus, even where a Han partner shows goodwill and gives up pork, a sense of distrust may remain among relatives and friends. Fears that religious conversion may be incomplete led other respondents to reject the possibility of conversionary intermarriage altogether, as in the case of Mälikä, the 20 year-old *minkaomin* student/hostess who toyed with the idea of marrying her Han 'manager'.

The difficulties pertaining to religious conversion outlined above have also been documented in Turpan. In Zhang Suqi's study of ethnic intermarriage, in ten out of thirteen instances of intermarriage between a Uyghur or Hui and a Han, the Han spouse had converted to the Islamic faith. However, in many cases, the relatives of the Uyghur or Hui spouse continued to refuse to recognise the match, placing considerable pressure on the couple. More often than not, the couple ceased to socialise with the Uyghur or Hui family, and the Uyghur or Hui spouse ended by distancing him- or herself from the Islamic religion and/or the Islamic community (Zhang Suqi 2005: 25). In one case, a Han male respondent confessed that despite his conversion to the Islamic diet, it was almost impossible to maintain abstention since, as a businessman, he was routinely expected to dine with business partners in *Hancan* restaurants serving pork (an act not permitted within the Muslim community). His conversion was further constrained in that if he refused to drink alcohol or smoke in the company of these business partners, he would be accused of looking down upon them (2005: 25-26). This case underscores the importance of commensality in Han

Chinese culture and the penalties incurred for non-compliance, a situation perhaps first encountered by China's Buddhist vegetarian converts in the tenth century: despite renouncing meat, they were still expected to butcher meat for their daughters' wedding banquets, while certain social positions—such as that of high official—actually required the eating of meat on social occasions. To do otherwise was considered 'eccentric and inappropriate' (Kieschnick 2005: 204).

A second means of managing intermarriage is to conduct the romance at a distance from the community or communities which seek to supervise it, thus escaping public and private ethnic stereotyping (Breger and Hill 1998: 5). In his seminal study, Spickard observed that individuals from all three of his target groups—Blacks, Japanese Americans and Jews—intermarried more frequently when they lived away from their ethnic communities, and thus away from the power of community supervision (1989: 362). While he recognised scarcity of Jewish mates as a practical factor motivating American Jews to inter-marry, he argued that, equally, individuals felt free to make their own choices when 'unencumbered by the reservations of a Jewish community' (Spickard 1989: 198). Similarly, an American woman married to a Kurdish man observed that living far away from either set of in-laws had made their relationship much easier (Brown and Farahyar 1994: 180). Some couples find it easier to live in built-up areas of towns, where they can remain more anonymous (Barbara 1989: 29). In Guyana, for instance, many intermarried couples live in comparatively urban, mixed communities, which might be expected to be more tolerant of racial mixing (Shibata 1998: 97). To give some examples where national politics exert a definitive influence, many German-Jewish couples were forced to consider leaving Nazi Germany rather than be separated (Stoltzfus 1996: 50). Those who stayed found they had to adapt their 'public face' in order to survive. German wife Charlotte Israel took over from her Jewish husband as the 'public face of the business' at the height of the public boycott of Jewish enterprises. Re-registering the shop in her name and presenting it as an 'Aryan' business, it was she who greeted the customers and communicated with the Office of Business Affairs while her husband worked in the back out of sight (Stoltzfus 1996: 52). To cite another example, prior to the Good Friday Agreement of 1998, an Irish Catholic-Protestant couple might have lived happily in Paris, whereas they clearly could not have done so in Belfast (Barbara 1989: 29), where relationships were plagued by religious sectarianism, a difficult political history and ongoing communal conflicts. Finally, removal from a particular cultural context to a 'neutral'

#### 330 chapter six

one can assist spouses in learning to balance their respective cultures, rather than forcing one partner to acculturate wholesale:

… when one partner moves into the cultural milieu of the other, the pressure on the incomer to adapt to the foreign community is greater than for the other partner. It is unavoidable that one culture would predominate in such a marriage […] By changing their social context—either by moving away or establishing their own nuclear unit—couples in a cross-cultural marriage can […] neutralize to some degree the pressures they face (Yamani 1998: 167).

As noted by Mamet, Jacobson and Heaton, ethnically intermarried couples are still not widely accepted in China and frequently live separately and sometimes at great distances from their parents (2005: 192). In Xinjiang, the 'exile' of intermarried couples is a recurrent theme in both the past and the present. Tashmämät, an intellectual in his forties originally from Qäshqär, related in 2004 how destitute Uyghur women who courted and married Han Chinese army generals during the Guomindang period (1911- 1949) were forced to flee to Taiwan with their husbands following the accession to power of the Chinese Communists in 1949: 'They couldn't stay here because the people would have killed them … they called them ethnic scum *[Uy. milliy munapiq]*.' Meanwhile, the current demographics of Beijing mean that Uyghurs studying or working there are *structurally* more likely to intermarry with Han than in Xinjiang. The Han accounted for 97 per cent of the city's total population in 1990 (Mamat, Jacobson and Heaton 2005: 188), and this ratio remained high at 95.69 per cent in 2000 (Harbach, Lapeyre, O'Berry and Staley 2010: 11). As noted earlier, in instances of intermarriage between Uyghurs pursuing higher education in China proper and Han partners, Uyghur sons and daughters are often urged not to return to Xinjiang (cf. Rudelson 1999: 204). There are several reasons for this: firstly, Uyghur parents find it extremely hard to cope with such a match (for the religious, cultural and, ultimately, political reasons discussed above). Secondly, parents are keenly aware that the Uyghur community will roundly condemn the match, making the couple's public and social life extremely difficult. Thirdly, the intermarriage will reflect negatively on the family's social reputation by bringing 'national shame' upon its members. In 1996, while on the train from Beijing to Ürümchi, I interviewed a young Uyghur woman from the regional capital, who was dating a Han male whom she hoped to marry. This respondent explained that she had an elder sister who worked in Beijing (as a model), who also planned to marry a Han male. In the event, their mother buckled under the weight of public disapproval,

refusing to allow her younger daughter's marriage to go ahead in Xinjiang, even though the Han partner had given up pork and begun to learn the Uyghur language. She agreed, however, to her elder daughter's marriage, since the wedding would take place far away from Xinjiang and the couple's married life would be spent in Beijing away from public pressure.

In an interesting parallel to the situation of the German woman Charlotte Israel above, Osman and his Han wife chose to continue living in a difficult context, but adapted their public face at the Ürümchi outdoor market at which they worked. Because Osman's wife had refused to convert to Islam, the couple agreed that she would refrain from coming over to chat with him while he worked at his kebab stand, instead performing tasks inside the restaurant's kitchen, where she was not visible. They were obliged to follow this practice in order to avoid the potential for Uyghur customers to assert that the stand was not *halal* and decide to eat elsewhere.

In terms of internal management of a mixed union, studies demonstrate that intermarried couples can in some contexts learn to negotiate their different cultural expectations (Breger and Hill 1998: 5), rather than one spouse adopting the culture of the other in entirety. In Xinjiang, so far, this seems to happen only rarely. 'Others' tend either to be transfigured to one's own kind, or are expelled from the orderly world (Bauman 1997: 47). In most mixed households, one language and culture seems to prevail (usually the Chinese language and culture), and the offspring tend to identify more strongly with one parent's heritage (usually the Han). As outlined by Breger and Hill: '… the language in which [mixed couples] decide to communicate at home can be symbolic of the extent to which each partner is prepared to forego her or his cultural background and incorporate new elements' (1998: 21). They go on to assert that a lack of linguistic skills on either side must also affect power relations within a marriage, since choice of language involves elements of control and dependence (Breger and Hill 1998: 22). According to close relatives and friends, in most cases of Uyghur-Han mixed marriage, one partner had essentially agreed to adopt the other's language and culture, or had ended up doing so by default. Frequently, it was the Uyghur partner who had adopted the Chinese language and certain (though by no means all) Han customs rather than the reverse. In 2002, Zöhrä, a *minkaomin* service industry employee in her thirties, described the case of a male, Uyghur colleague married to a Han woman with whom he had raised two children:

#### 332 chapter six

The kids—you know, we call them 'half-breeds' [Ch. *erzhuanzi*]—are really beautiful. But neither speaks Uyghur, they've both turned into Hans. And the father is like a Han too, he doesn't speak Uyghur, always Chinese. This is the only way such a marriage can work; one partner must give in completely to the other.

It transpired that this Uyghur husband spoke Chinese rather than Uyghur even when outside the home and when conversing with Uyghur colleagues in the workplace, thus in key respects had willing adopted a quasi-Han identity. As a result, both his children had grown up as Chinese speakers and knew nothing of their mother tongue. Patigül, a *minkaomin* music trader in her late twenties, confirmed in 2004 that her friend, married to a Han male, spoke only Chinese in the home and decorated and furnished that home in a Han style. This Uyghur wife, however, did speak Uyghur at her place of work—a restaurant—since socialising with Uyghur customers formed a key part of her managerial role. This woman, then, had not adopted her husband's identity wholesale; rather, she attempted to negotiate the two identities available to her. That said, since the Chinese language dominated the home environment, and she was too busy at the restaurant to teach her son his mother tongue, the ten year-old child (described by a friend as 'looking more like a Han than a Uyghur') spoke not one word of Uyghur. In a third case, a middle-aged male of part Han Chinese and part Central Asian heritage (a Chinese speaker), who had married a Uyghur woman in 1983, had tried but failed to learn the Uyghur language. Since his wife was a first-generation *minkaohan* educated during the Cultural Revolution, she was herself more comfortable with the Chinese language. For this reason, the husband had not enjoyed the conditions necessary to practise speaking Uyghur and to master that language. The result was that the couple always communicated in Chinese, and raised their eighteen yearold son as a Chinese speaker. The household had thus become all Chinesespeaking by default.

The importance of a common language has been documented in another study of mixed marriages in Xinjiang, which focused mainly on Uzbeks and Tatars. Li Xiaoxia found that 'increased cultural proximity' normally in the form of common language use—facilitated ethnic intermarriage. The Uzbek and Tatar respondents in this study lived in villages dominated by Kazakhs, had grown up with Kazakh friends, and spoke Kazakh as their first language. As a result, they considered there to be no significant barriers to intermarriage with Kazakh peers (Li Xiaoxia 2005: 66). Interestingly, in one interview, a Kazakh-speaking Uyghur respondent

stated that while she would allow her son to take a Han wife *provided that the lady spoke Kazakh*, she would not permit her daughter to marry a Han because of the rule in Turkic cultures that ethnicity follows the father's line (Li Xiaoxia 2005: 67). Thus, on condition that the patrilineal system was not threatened, a common language was deemed more important than common ethnicity.

I heard of just two examples of Ürümchi households in which an embryonic form of cultural negotiation was being pursued. In the first case, related in 2004 by Ömär, eighteen, a *minkaomin* male high school student, a Han woman had previously been married to a Uyghur friend of Ömär's family, and had borne him a mixed-heritage daughter. Following her Uyghur husband's death from heroin abuse, this Han woman had re-married with a Han male. One might expect that the child would be raised as a monolingual Chinese speaker, given that her Uyghur father had died when she was only one year old. Yet Ömär confirmed that the seven year-old daughter spoke, in addition to fluent Chinese, 'beautiful' Uyghur, which she had learned from Uyghur neighbours living in her Han grandmother's neighbourhood. Every day, she played with local Uyghur children. As a result, she had learned to speak Uyghur exceptionally well, in spite of the fact that her closest living relatives—mother, step-father and grandmother—were all Han. My respondent, Ömär, was delighted at this outcome, suggesting that Uyghur-Han relations might be vastly improved were more Han residents to make the effort to learn the Uyghur language.

The second case was related in 2004 by Dilbär, also eighteen, a *minkaomin* female high school student. Her Uyghur uncle was married to a Hui (therefore Chinese-speaking) woman, with whom he was raising a small son. Although this child had initially been educated in a Chinese-medium kindergarten and so could not yet speak any Uyghur, Dilbär's family moved quickly to avert the perceived danger. She explained:

We've pulled him out of there now [the kindergarten], and our [Uyghur] relatives are looking after him instead, so that he learns Uyghur. Don't worry, we'll teach him! [*Ögitimiz!*] At present, we speak to him only in Uyghur and he can't understand! He asks 'What are you saying?' [Ch. *Nimen shuo shenme ne?*] But then we tell him the Chinese and the Uyghur version, so that he learns both.

Both cases above underline the growing recognition in the Uyghur community that the formative years will be crucial in determining whether or not a mixed-heritage child is 'lost' to the group.

#### 334 chapter six

# 6.6. *Hierarchy of Intermarriage Preferences*

Studies of intermarriage often reveal a hierarchy of marriage preferences beyond the endogamous parameters of the in-group: what Breger and Hill call a 'hierarchy of acceptable "foreign-ness"' (1998: 8). Within that hierarchy, 'not all groups of foreigners appear equally "strange"; some groups seem more familiar, their presence is more tolerated, their cultural practices perhaps even admired.' (Breger and Hill 1998: 8) This tendency to grade people outside of one's own ethnic group thus indicates differing degrees of perceived similarity and affiliation, and has an important bearing on possibilities for intermarriage (Spickard 1989: 16). Patterns of grading differ from group to group. Conceivably, an individual will normally prefer an exogamous partner who speaks a language from within the same language family over one who does not; a partner who follows the same religion over one who follows another; or a partner whose cultural norms are similar to one's own over one whose norms are different, and so on. Soviet scholars recognised that, within the Soviet Union, intermarriage was more likely between groups considered to be culturally alike, such as Russians and Ukrainians, or Uzbeks and Tajiks, while Muslim-Christian interfaith marriages were viewed as more difficult owing to cultural and linguistic differences and 'religious prejudices' (Edgar 2007: 588).

Yet even within sub-groups based on language, religion or regional culture, we often find further, cross-cutting ranking mechanisms. For example, first-generation Jewish Americans prioritised intra-religious differences, distinguishing carefully between orthodox and reform Jews, perceiving the latter to have betrayed their faith (Spickard 1989: 216-7). For second- and third-generation Jewish Americans, however, colour proved a more powerful barrier than religion. At the top of the list came White Protestants of liberal denomination and those who rejected religion (in other words, those perceived to be non-threatening because they had a weak religious identity or none at all); next were White Catholics or Baptists (believed to hold to their religion more tightly, thus more threatening); and lastly came those beyond the 'colour line', i.e. non-White Gentiles (Blacks and Asians) (Spickard 1989: 216-7). Yamani's study on international Islamic marriages demonstrates how regional and class factors can take precedence over a common religious foundation, with differences in language, customs and values becoming points of contention and differentiation (1998: 157; 159). Thus, modern descendents of mixed unions in Mecca (formerly characterised as the 'melting pot' of the Islamic world) are more intent on maintaining 'purity of blood' within the Saudi nation-state than were their ancestors (Yamani 1998: 165).

Hierarchies of marriage preference may enjoy continuity over time or may undergo change across generations, depending on environment. In Spickard's study of intermarriage between white Americans and black, Jewish and Japanese Americans, hierarchies endured over several generations (1989: 9). Yet Kohn's work among the Yakha of East Nepal revealed that rules and notions of hierarchy can sometimes be negotiated. Whereas the Yakha once avoided intermarriage with the Rai, deeming them to be of lower caste, at present Yakha-Rai unions are perfectly acceptable, producing what Kohn calls a situation of 'controlled abnormality' (1998: 70).

In Xinjiang, Uyghur individuals in mixed unions were intermarried with 47 different nationalities, including the majority Han, according to the population census of 2000 (Li Xiaoxia 2004: 21). Within the region's two main 'intermarriage pools' (consisting of Uzbeks, Tatars and Uyghurs, and Uzbeks, Tatars and Kazakhs respectively), Uzbeks and Tatars were found to intermarry with Uyghurs and Kazakhs at a rate of 10 per cent or higher, indicating that Uyghurs are perfectly willing to take Uzbek or Tatar spouses. There was however no such consonance between Uyghurs and Kazakhs, perhaps owing to historical differences in lifestyle (settled versus nomadic). Crucially, Turkic-speaking Islamic groups in the Northwest, including Uyghurs, Kazakhs, Uzbeks and Kyrgyz, were found to be the *only* minority peoples not to intermarry with Han Chinese to any significant level and, among them, Uyghurs and Kazakhs are the least likely to intermarry of any nationality in the PRC (Li Xiaoxia 2004: 23-4). Mamet, Jacobson and Heaton similarly report a 'clear pattern of endogamy' within the Turkic groups, noting that the Uyghurs in particular maintain a rate of in-group marriage of over 99 per cent (2005: 196). One study suggests that this remains the case even when Uyghurs are living in the diaspora. Despite being pursued by Turkish women for their 'pure Turkish blood', most Uyghurs and Kazakhs interviewed by Gladney in Istanbul continued to seek wives from within the in-group, no matter how hard this proved (2004: 224). A similar pattern has been reported for the Tajiks of the former Soviet Union, who are also highly endogamous. On the rare occasions that Tajiks marry exogamously, they are most likely to marry Uzbeks, rather than Russians or others of non-local origins (Atkin 1992: 60).

My interviews, conducted between 1995 and 2004, revealed a high level of consistency in marriage preferences among Uyghurs. Across this period, the most likely exogamous partners hailed from other Central Asian,

#### 336 chapter six

Turkic-Islamic groups, such as the Uzbeks, Tatars and Kyrgyz. Top of the list were Uzbeks, linguistically and culturally the closest of the Central Asian groups to the Uyghurs. In addition to pre-existing cultural affinities, many young Uzbeks growing up in Ürümchi speak Uyghur as their first language rather than Uzbek. This allows for a sense of mutual identification as well as ease of communication, reflected in the frequency with which Uyghurs and Uzbeks make friendships. Thus, intermarriage with Uzbeks is considered perfectly acceptable. In addition, there is some evidence to suggest that Uyghurs, Uzbeks and Tatars are united by a mutual cultural antipathy towards Han people. In Li Xiaoxia's study of ethnically mixed households in two villages in Xinjiang, none of the Uzbek or Tatar respondents had Han spouses (2005: 62; 64).

Next in the ranking came formerly nomadic Central Asian Islamic groups, including the Kazakhs and Kyrgyz. Uyghurs share Islamic practices with Kyrgyz and Kazakhs, but there are some perceived differences in national customs and greater differences between the respective languages. In the mid-nineties, I had found some evidence of successful intermarriages between Uyghurs and settled Kyrgyz in south Xinjiang, with the offspring invariably raised as Uyghur speakers and assimilated into the Uyghur culture.

One might expect to find a high level of Uyghur-Hui intermarriage based on the fact that both groups are Islamic believers; but this is not the case. As asserted in 2004 by a retired Hui lorry driver running a small grocery store in Ürümchi, Hui intermarry far more frequently with Han people than with Uyghurs. This is because negative stereotypes based on historical memory, together with linguistic and cultural factors, combine to make Uyghur-Han unions more complicated than one might expect, although not impossible. The problematic relationship between Uyghurs and Hui stems from a variety of sources. First, the Hui speak and write Chinese rather than a Turkic-Altaic language, although this difficulty is ameliorated if a Hui living in Xinjiang also speaks fluent Uyghur. Second, the Hui are widely dispersed across the territory of the PRC, and have developed a hybrid form of culture in which their ancient Arab-Islamic heritage merges with Han Chinese culture (including adoption of the Chinese language and script). Thus, Uyghurs' mistrust of Hui emanates in part from their perception of the latter as 'quasi-Hans'. Third, Uyghurs identify the Hui as a non-indigenous people who began migrating to Xinjiang at the same time as Han Chinese, and who have frequently allied themselves with Han Chinese against them in the twentieth century. Hui are therefore reviled along-

side Han as invaders and occupiers. Fourth, Uyghurs tend to believe that the Islamic faith among the Hui has been diluted after centuries of coexistence and intermarriage with Han Chinese; the two groups refrain from social interaction during Islamic festivals, and use separate mosques. Finally, the greater willingness among Hui to marry Han people provides further cause for Uyghurs to mistrust the Hui as 'collaborators' with the Chinese state. I heard of one case of Uyghur-Hui intermarriage in Ghulja; however, this marriage had already failed at the time of interview. Patimä, a *minkaomin* service industry employee in her early 30s, explained in 2002 the reasons for the breakdown of her marriage to a Hui male as follows:

I married a local Hui from my neighbourhood [Uy. *mähälla]* when I was 16. I didn't know anything then—I was too young. I never loved him. He didn't like going to see my relatives … though it was fine when we went to see his! He didn't like Uyghurs. It's always better to marry with your own nationality. Uyghurs do marry Hui, but in ten marriages, only one will succeed. The other nine will end in divorce.

According to her explanation, differing national customs and social norms had caused a rift between the couple, such that the Hui husband disliked prolonged interaction with his Uyghur wife's relatives. It is quite possible that, as in the above-described case of a Uyghur-Han marriage heading for divorce, the reason for his avoidance of family visits was his inability to communicate in the Uyghur language.

My findings are supported by two studies published by Chinese scholars. The first found that among the 840,000 Hui resident in Xinjiang in 2000, most were more likely to intermarry with Han Chinese than with fellow Islamic groups such as the Uyghurs or Kazakhs. The rate of Uyghur-Hui intermarriage in the region was confirmed at just 0.09 per cent (Li Xiaoxia 2004: 24-5). The second, a study of ethnic intermarriage in Turpan, revealed more than twice as many Hui-Han unions as Uyghur-Han unions, while Uyghur-Hui unions were few (Zhang Suqi 2005: 24).

Ahead of Han Chinese in the hierarchy of marriage preferences—and for some respondents, perhaps also ahead of Hui—came 'Western' partners (usually denoting Americans or Europeans). Such unions were considered more desirable than those with Han Chinese for a variety of reasons. Westerners are viewed as inherently more 'cultured' and 'civilised' than Han people, considered by most Uyghurs to be among the least cultured and least civilised peoples in the world. Nor is this perception solely based on their status as non-Muslims. Ghäyrät, a migrant worker in his thirties from Aqsu, explained in 2004:

#### 338 chapter six

In most cases of intermarriage, parents may not like it, but can usually be persuaded. For example, if a Uyghur wished to marry a Westerner, most would eventually agree. But few or none will agree to marriage with a Han. It's not only that they're not Muslims; Uyghurs think them culture-less [*mädäniyätsiz*] and bad-mannered [*ädäpsiz*].

Westerners are also perceived to be peoples who follow a 'religion of the book' (Christianity), making them significantly more acceptable as potential spouses than Han Chinese, who are perceived as atheists or polytheists (see Section 6.4 above). Several young Uyghur males admitted that they had considered the possibility of marrying a Western woman were they to undertake study abroad. Parental reaction was however variable. Jelil, an observant university graduate in his late twenties, stated in 2002 that his parents had already agreed to his marrying a Westerner should the opportunity arise. As a result, he joked (half seriously) that he was 'currently looking for a Western wife'. In one sense, a match with a Westerner was conceived as bringing concrete advantages. Jelil pointed out that he would have less trouble obtaining a visa to a Western country if he was married to a Western citizen. However, he gave three additional reasons for the enhanced acceptability of a match with a Westerner. First, he argued that Westerners 'present no religious problem' since they are Christians and 'have faith'. Second, he asserted that all Westerners are 'well educated'. And, third, like Ghäyrät above, he highlighted the 'high cultural level' of Westerners, describing them as 'a highly developed people'.6 In this way, he contrasted peoples of American and European origin favourably against the godless, poorly educated and uncultured Han people, as he perceived them. Other Uyghur parents had more guarded reactions to the suggestion that Uyghurs might intermarry with Westerners. When Ömär, an eighteen yearold *minkaomin* high school student, asked his parents whether he might marry a Westerner if he studied at a Western university, they reminded him that few Westerners are Muslims, the implication being that only a marriage to a Western Muslim is acceptable. Initially interested to learn in 2004 of Uyghur-British mixed unions in the UK, Ömär subsequently sought confirmation that the British partner had in each case converted to Islam, and that the couples had undergone the *nika* (Islamic wedding ceremony). Thus, some respondents deemed Uyghur-Western matches impossible

<sup>6</sup> At this point of the discussion, I could not resist pointing out that not all Westerners are highly developed, citing then US President George W. Bush as an example. Jelil was delighted, clasped my hand, and roared with laughter.

without full conversion of the non-Muslim partner. An interview with a *minkaomin* waitress in her twenties working in Ürümchi in 2004 revealed a definite double standard with regard to this dilemma. She told me that her aunt had lived in England for twelve years after marrying an Englishman. When I asked whether she thought it appropriate for a Uyghur to marry a Westerner, she confirmed that this presented no problem at all and seemed surprised. When I queried whether the spouse's religious status might be important, she revised her position, stating: 'Oh, yes, it's OK as long as the person's a Muslim.' When I asked whether her aunt's English husband was a Muslim, however, she was unable to confirm this one way or the other. Finally, I asked her whether it was acceptable for a Uyghur to marry a Han person. Without any hesitation, she replied: 'No, absolutely not.' When I asked her for the reason, she said: 'Because they're not Muslims! They believe in Buddhism, we believe in Islam, our religions are different.' In this way, the Islamic criterion, apparently unimportant when discussing her aunt's union with the English partner, suddenly became paramount when the question concerned Uyghur-Han intermarriage.

In the former Soviet Union, two schools of thought towards intermarriage competed fiercely among indigenous communists in Turkmenistan in the mid-1920s. One group argued that intermarriage with Russians was 'advantageous' for young men seeking to improve their Russian language skills and make a career within the Soviet Communist elite. The other feared that the practice would be 'detrimental to the Turkmen nation' in that it would produce a generation of 'Russified' offspring (Edgar 2007: 581). More broadly, social opposition to intermarriage with Russians came simultaneously from the poorest educated (where education designates a secular education, i.e. women, the elderly, and deeply religious people), and the best educated strata of society. Muslim intellectuals were 'especially prone', a phenomenon interpreted by Soviet sociologist Drobizheva as a means for national intellectuals to retain their national uniqueness through endogamy in opposition to an official discourse of homogenisation (cited in Edgar 2007: 591).

The disparity in the Uyghur waitress's narrative above could be partially although not wholly explained by the fact that some Uyghurs look upon monotheistic religions more favourably than polytheistic religions or atheism. Thus, the respondent may have assumed that her aunt's partner was at the very least a Christian and therefore acceptable, whereas a Chinese Buddhist was not. I would suggest, however, that for most Uyghurs (except strictly observant individuals) factors such as religion or culture

#### 340 chapter six

would not in themselves be insurmountable under conditions of comparative social equality. As observed by Li Xiaoxia, where there is a large gap in numbers (population), political power and social position, and groups are physically segregated to a significant degree, intermarriage can give rise to feelings of insecurity within the 'weaker' group, affecting ethnic relations with the dominant group (2006: 84). I would suggest that the reason why religion and culture are invoked with such passion in the case of Uyghur-Han mixed marriages, yet lose importance in other contexts of intermarriage, is that the vast majority of Uyghur individuals are *politically disinclined* to lose face by acceding to the hegemonic goal of nationality unity; instead, they choose to resist the state's attempts to achieve social harmony through a process of biological 'ethnic fusion'.

## 6.7. *Women as 'Culture Bearers'*

In situations of oppression of one social group by another, sex may become—and be seen as—a metaphor of domination. An obvious historical example is the seduction and rape of black people by whites in America, conceived by Hernton as a 'sexualization of the race problem' (1975: 150- 152). This 'racism of sex' is at its most extreme in the context of ethnic conflict or war, where fertile women committed to males associated with 'the other' are seen as potential vessels for the reproduction of that 'other'. At the same time, fertile women of the in-group may be fiercely protected against transgressions by out-group males, as in the Yugoslav wars, where Bosnian Serb leaders represented the reproductive potential of Serb women as a resource threatened by the proximity of Bosnian Muslim males (Hartley 2010: 236).

In China, sex—in the sense of sexual relations, seen as a man doing something to a woman—has long been a pervasive metaphor for relations of domination (Harrell 1994: 10). The state has routinely depicted peripheral peoples as women, 'feminising' them, and thus subordinating them within an official hierarchy of majority-minority representation (see Harrell 1994; Gladney 1994; 2004). Sexualised images of minority individuals are frequently found in art and cinema (Harrell 1994; Gladney 1994; 2004), with one particularly graphic instance explored by Dru Gladney (2004): Zhao Yixiong's oil painting *Awakening of Tarim* (1979). This painting depicts a nude female subject (recognisably Uyghur/Turkic), reclining against a collage of traditional and modern images; as oil rigs, airplanes and nuclear

installations spring up all around her, camel caravans and minarets emerge from between her thighs. According to Gladney's analysis, the painting made a dramatic link between nationality, woman, and tradition, and implied that Xinjiang could only be modernised by throwing off the traditional culture of Islam (2004: 80). One could take the analysis further, and propose that in contemporary times the painting may be viewed as embodying the sexual domination of Uyghur women by Han men (through mixed marriages and the hiring of Uyghur hostesses in karaoke bars) and, by extension, the political domination of the Uyghurs as a nation.

As argued by scholars of interracial and interethnic marriage, the 'sexual taboo' has traditionally been one of the strongest markers of group identity; as a result, the act of 'marrying out' equates to a violation of that taboo and thus an attack on the group's racial or ethnic identity (see Barbara 1989: 15). At the centre of the taboo we find women: the concept of a 'gendered ethnic centre' revolves around in-group ideologies and stereotypes of woman, which associate females with family, motherhood and love, and locate them in the emotionally charged symbolic role of 'mothers of our people'. These ideologies incorporate a 'sexual morality', so that in-group women are viewed as vulnerable to the aggressive sexuality of out-group men (Breger and Hill 1998: 15). As such, intermarrying women rather than men—tend to be the targets of criticism, harassment and abuse when crossing normative boundaries (Shibata 1998: 84; 96). For example, if an Indian woman in Guyana goes out with an African man, she is considered to have disgraced 'the whole Indian nation'; conversely, if an Indian man chooses an African woman, it is viewed rather as a matter of 'civilizing the Africans' (Shibata 1998: 90). Similarly, in contemporary Greece, endogamous principles are not uniformly dictated to men and women alike; men taking brides not belonging to the national group are tolerated far more easily than women who marry outsiders (Petronoti and Papagaroufali 2006: 559).

What can explain this gender discrepancy? It is partly explained by the organisation of kinship descent in a given society. In Judaism, for example, descent is matrilineal, meaning that children of Jewish mothers and Gentile fathers retain the Jewish identity (the 'rule of the womb'), while those of the inverse combination do not. As a result, Jewish women may marry out more easily than Jewish men (Spickard 1989: 225). The majority of societies, however, are patriarchal, so that the reverse is true: offspring retain the identity of the father. This is also true of Islamic societies. Since descent

#### 342 chapter six

in Islam is patrilineal, only children of a Muslim father retain the Islamic identity, making the protection of women—and their reproductive function—a matter of group honour for in-group males. Moreover, a Muslim husband ensures that the mother—the source of the child's religious and moral education—continues to practise Islam. In Islamic societies, notions of loss and restoration of male honour may be especially marked. For example, it is argued that, in the Arab nations, males experienced a double sense of humiliation in the latter decades of the twentieth century: firstly, national humiliation in the form of defeat by Israel in the 1967 Six Day War; secondly, social humiliation as a result of failing socio-economic policies and resultant class demotion. As Ayubi observes, this double humiliation and the loss of male dignity that it entailed catalysed a process in which women were made the main focus for 'restoration' of that dignity (1991: 40). Male honour—embodied in the sexual behaviour of a man's female relatives—was easy to enforce in the past, when female spaces were confined to the family courtyard or ritual visits to a local saint's tomb. In the contemporary urban space, where women are more 'exposed' (unveiled; less strictly controlled), male honour has been placed at greater risk (Ayubi 1991: 41). In this way, the removal of the veil, women's work outside the home, female independence, and the related stigma of 'promiscuity' came to be seen as principal causes of the evils of contemporary life in the Arab states, intimately related to the loss of Arab 'manhood', defined as personal dignity (Shukrullah, cited in Ayubi 1991: 40).

The dynamic of the 'gendered ethnic centre' is especially likely to play out in the context of minority oppression within a plural society; here, the social rules of the group—formulated by in-group males—can transform women into 'forbidden persons' (Barbara 1989: 15). Supervision of women can be prompted by a sense of social, political, cultural or economic impotence among in-group men; that is, where men lose their role as provider for the family, they may increasingly try to control and circumscribe their women (Buijs 1993: 5, 18). Abdulrahim (1993) illustrates this process among Palestinian refugees in West Berlin. In Palestine, although political organisations attached a patriotic meaning to the importance of the family's (thus the woman's) reproductive functions, they nonetheless recognised that it was essential to allow women to participate in the political struggle (Abdulrahim 1993: 57). Uprooted and relocated to Berlin, Palestinian men have reacted to male unemployment—and the associated feelings of disempowerment—by reconstructing the ideals of female seclusion and

the gendered segregation of space (Abdulrahim 1993: 67). In so doing, they reconstruct group identity around male honour and female shame, 'involving their increased control of female movement and behaviour, which in particular aims to restrict their contact with German [majority] men' (Breger and Hill 1998: 14-5). In this way, minority men make a last-ditch attempt to retain control over one final arena: the private sphere.

Throughout the course of my interviews in Xinjiang, the vast majority of male respondents were fiercely opposed to Uyghur-Han courtship and marriage; however, it was notable that their opposition was particularly targeted at Uyghur women. Nurmämät, a *minkaomin* trader in his early forties, remarked in 2002: 'If a mixed couple walks along in Döngköwrük, people will swear at them, even hit them! If it's a Uyghur girl, that is. But, if it's a boy, they say nothing.' He shrugged and smiled in recognition of the double standard. Male censure expressed itself in violent criticism and condemnation of females who violated the intermarriage taboo, coupled with an earnest defence of those Uyghur men who had been deserted for Han partners. When in 2004 I informed a long-term respondent named Shökhrät (male intellectual in his thirties) that I was researching Uyghur-Han unions, his reaction was deeply cynical:

Hmm. Yes, I'm also very interested in inter-ethnic courtship and marriage [grimly]. I'll tell you: there's really only one reason for it, and it's not romance, it's just money. That's all. The Chinese are loaded, the Uyghurs have no money. That's all they [Uyghur girls] think about.

He continued this diatribe by adding that such girls 'don't have brains' [Uy. *kalisi yoq*] or at least fail to use them, and that they 'sleep around' and are 'little better than prostitutes'.

Tashmämät, a male intellectual in his forties interviewed in the same year, was equally critical. He listed four reasons why Uyghur girls have begun to date Han men, as follows:

The first is money. Some girls will go with anyone if they have money, with Hans, with old men […] Second, these are girls who have grown up among Hans and so are closer to Hans than to Uyghurs. They are sinicised to a great extent. They are no longer full Uyghurs. Third, these are often girls who have slept around, girls who have one night stands on a regular basis. For these girls, it will be impossible to find a Uyghur husband. And so they can only expect to marry a Han. Fourth, these girls are living in their own world [Uy. *özining dunya*], a dream world [Uy. *khiyali dunya*].

#### 344 chapter six

Like Shökhrät, he attributed their actions first and foremost to an appetite for gold-digging; in his view, for a minority of young Uyghur women, wealth and status are more important criteria than ethnicity when selecting a potential spouse. This recalls the research findings from Kunming, where (albeit much larger numbers of) women of Hui, Yi and Bai nationality were prepared to take Han partners in the context of insufficient numbers of high-status minority males (Xing 2007). Secondly, Tashmämät cited a deep level of acculturation, suggesting that such women tend to be *minkaohan* at the more acculturated end of the spectrum. Having grown up in predominantly Han neighbourhoods and been educated with Han peers, these women are no longer 'purely Uyghur' in cultural essentialist terms. Their temperament is closer to that of Han Chinese, rendering them unattractive to Uyghur peers while encouraging romantic rapprochement with Han males. A blemished sexual history was the third reason given, a situation cited also in historical sources as likely to lead to intermarriage with a Han (Bellér-Hann 2008). On a separate occasion, this respondent had pronounced that the TV drama series *Xinjiang guniang* [Xinjiang Girls], in which a Uyghur daughter dates a Han male, should have been called 'Xinjiang Prostitutes'. It is instructive that both Tashmämät and Shökhrät hailed from Qäshqär; thus, despite aspiring to secular visions for Xinjiang's political future and bemoaning the Islamic revival, each was nonetheless permeated with Islamic male beliefs surrounding female shame, absorbed during a childhood spent in the deeply religious south. Only the fourth reason he gave contained an oblique reference to politics, implying a (perceived) lack of political awareness and ethnic self-esteem among females. It was interesting that Dilbär, eighteen, a *minkaomin* female high school student, later agreed with the first three reasons Tashmämät had given: lust for money; over-sinicisation; and bad sexual reputation. This suggests that some Uyghur women may be equally critical of their intermarrying peers.

The relative upward mobility of Han urban males was cited also by Abdukerim, an observant *minkaomin* graduate in his twenties from Aqsu speaking in 2002, who compared this with the high proportion of Uyghur males who are unemployed or under-employed. While allowing for the possibility of 'falling in love', Abdukerim characterised this as an act of 'empty-headedness', suggesting that like other male respondents, he perceived Uyghur women to lack an ethnic or political conscience. Another observant graduate in his late twenties, Jelil, similarly cited money as the

key motivation in interview in 2002, but added that some Uyghur women are also attracted to the educational background of potential Han partners. He pointed out that, while some Uyghur businessmen have accumulated wealth and entered the new minority middle class, these individuals tend not to have completed higher education, thus may be perceived to have a lower 'cultural level' than many Han males. In sum, social climbing (wealth and status) was most often cited as the impetus for intermarriage with Han men, with intermarrying Uyghur females characterised variously as brainless, sexually 'loose' and politically unaware.

While the causal factor underlying endogamy—politics—was rarely explicitly expressed, one young respondent openly invoked the Uyghurs' worsening political and cultural situation in connection with the intermarriage taboo. Ömär, eighteen, a *minkaomin* high school student, explained in 2004:

We cannot accept Uyghurs marrying Hans—unless they convert. It's too much to bear. You know, we think 'They took away our country [Uy. *dölitimizni eliwaldi*], our language and our culture, they made us like them. And now they want to take our women too.' We oppose that. All these things are linked, they cannot be separated. And so we oppose intermarriage. It's the final straw, the last thing we can resist. I can't bear the thought of reading in history books that there was once an ethnic group called the Uyghurs, but that we were assimilated through intermarriage with the Han. That's too humiliating, too much loss of face. I don't want us to end up like the Native Americans. The most important thing for Uyghurs to ensure now is that our language and culture are not lost.

Intermarriage to a Han is thus interpreted as the ultimate cultural loss in an environment where Uyghur language, culture and social and political status are already perceived to be under heavy threat. This brings me to perhaps the most charged of motivations for Uyghur endogamy: the perceived threat of cultural extinction.

# 6.7. *Endogamy as Cultural Survival*

At the turn of the twentieth century, Jewish rabbis in the US, fearful that the Jewish race was in decline, hoped that prohibition of intermarriage might create 'a social inclusiveness which is the most powerful of all preservative agents' (Heller, cited in Spickard 1989: 185). Similar concerns have acted to prevent exogamy in Islamic societies, where the female is often

#### 346 chapter six

viewed as the vehicle for the continuity of tradition in all its forms (language, culture, religion, ritual):

To allow a woman to move away is regarded as a loss for the group: a loss of capital and of future enrichment because of her reproductive potential […] it is she who will determine the continuity of the group: she produces and reproduces the life and traditions of the group' (Barbara 1989: 210).

For example, the Berber minority in Algeria view a woman who marries out as lost to the community and to the language; they therefore choose to exercise a certain degree of endogamy in order to ensure that the Berber language is safely passed on (Barbara 1989: 38). One contemporary Han scholar, apparently eager to support Chinese President Hu Jintao's political initiative of the 'harmonious society' (*hexie shehui*), argues that Tibetan-Han intermarriage was a historical 'necessity' which allowed the two ethnic groups to co-exist harmoniously (Xie Lei 2006: 147-8). In his paper, he states openly that intermarriage has 'sped the process of sinicisation and fusion of the different ethnic groups', while acting to suppress 'splittist behaviour' (separatism) and to uphold the peaceful unification of the Chinese nation (Xie Lei 2006: 149). It is perhaps not surprising, then, that intermarriage is frequently perceived in Xinjiang as a vehicle through which the state seeks to forcibly assimilate the Uyghur people.7 This has never been the case more than the present, with Uyghur individuals exerting every effort to maintain endogamy (albeit selectively) as an insurance against the perceived threat of group extinction. Cultural and religious factors and an increasingly politicised Uyghur ethnic consciousness combine with genuine fears for the survival of the Uyghur nation to ensure that the majority of the Uyghur population remains firmly opposed to intermarriage. Fears of group extinction stem from complex identity questions surrounding the future offspring of mixed Uyghur-Han unions: will they be Muslims? Will they understand Uyghur social etiquette? Will they know how to participate in important life cycle rituals? What kind of education will they receive? Will they speak and write the Uyghur language? And, most importantly, will they transmit the Uyghur culture across generations intact? Aynur, a *minkaomin* university graduate in her early twenties, phoned her male cousin, Ömär, in a rage after watching the final episode of TV drama *Xinjiang guniang* [Xinjiang Girls] in 2004. Evidently, in the denouement, the progressive Uyghur father had urged his daughter's Han suitor

<sup>7</sup> Judith Banister had already observed in the second half of the eighties that some people—particularly minority nationalities—viewed state encouragement of intermarriage as an attempt to assimilate minorities and therefore resisted such pressures (1987: 319).

to 'be patient until family obstacles to the match subside'. Ömär, an eighteen year-old *minkaomin* male high school student, interpreted her rage as follows:

She worries that the Uyghur people will disappear and become assimilated […] I used to say I would marry a foreigner and leave Xinjiang, but my classmates were furious and demanded 'What are you doing? If you marry outside the Uyghur group, you are speeding the process to our disappearance. Why can't you marry in the group? There are lots of Uyghurs!'

In this way, community supervision is applied among politically aware young urbanites to prevent mixed unions, particularly with the lowest group in the marriage hierarchy—the Han Chinese—but to a significant extent to prevent exogamy in general.

In highlighting the need to preserve the Uyghur language and culture, many respondents cited the decline of the Manchus or the Mongols, emphasising that these peoples had formerly been great nations in their own right, but were now outnumbered in their own lands owing to Han inmigration and intermarriage with the Han. In 2002, Ghäyrät, a migrant worker in his thirties from Aqsu, observed sadly:

There aren't any Manchus now … hardly any. Those who remain are intermarried with Hans, or they're half-breeds (Ch. *erzhuanzi*) with one Manchu and one Han parent. They used to be a great nation with their own kingdom, but now their language and script have disappeared …

Other respondents offered the example of the Native Americans (see above). Zunun, a service industry employee in his thirties, despite being unusually tolerant towards intermarriage among his peers, insisted in 2002 that he would never allow his own child to marry a Han, even though he planned for him or her to receive a Chinese-medium education: 'That would mean that we would have no descendents, no *Uyghur* descendents […] We need to make sure that the Uyghurs continue as a nationality. Otherwise, we will simply disappear.'

# *Conclusion: Honour and Shame in a Context of National Impotence*

Women are the symbolic markers of the nation and the reproducers of the nation and its future people (Kaufman and Williams 2007: 31).

As Hartley observes, the extent to which minority individuals marry members of the dominant socio-economic class is highly situational, with eval-

#### 348 chapter six

uations of status gain or loss measured by the dynamics of cooperation or competition between the two groups (2010: 228). Around 8 per cent of marriages in the former Yugoslavia are estimated to have been mixed, while the offspring of such unions often described themselves as 'Yugoslavs' rather than using ethno-national or ethno-religious designations. Fusion of this type did not present a problem until hostilities broke out in 1991, following which mixed couples and individuals came under attack, and the percentage of people ascribing to the Yugoslav identity dropped dramatically (Hartley 2010: 229-230).

In contemporary Xinjiang, the vision of Uyghur women dating and marrying Han men is for most Uyghur males the painful culmination of a series of increasingly humiliating blows, from the political (invasion and colonial domination of the Uyghur homeland) to the socio-economic (marginalisation through discriminatory policies and practices), to the cultural (repression and forcible assimilation of the Uyghur heritage—language, socio-cultural practices, religion). In this context, the responsibility for reproducing the Uyghur nation and transmitting the Uyghur culture down the generations lies squarely with the woman. Ultimately, it is Uyghur females who are expected to both avoid further (national) shame and rebuild (national) honour. Women are the 'culture bearers', to be protected from Han male encroachment in order to sustain Uyghur male dignity as much as the Uyghur national future. Will endogamy then come to embody the sole remaining vehicle for Uyghur national resistance in Xinjiang? In the final chapter, I consider the Uyghur youth, those perceived by some to be most susceptible to linguistic and cultural 'dilution', and the acquisition of a 'Han temperament'.

#### Chapter Seven

hybrids: identity negotiations among the urban youth 349

# hybridS: identity negotiationS among the urban youth

this final chapter focuses on the uyghur urban youth, including *minkaohan* (uyghurs educated in Chinese-medium schools), *minkaomin* (uyghurs educated in uyghur-medium schools) and those in between, that is, young people who have received a combination of uyghur-medium and Chinesemedium education. it begins by considering the negative self-identity ('internalised oppression') common among first-generation *minkaohan* educated between 1958 and 1978. i then compare this with the more positive identities maintained by second-generation *minkaohan*, whose parents opted, for pragmatic reasons, to have them educated at Chinesemedium schools in the 1990s. While these young people have been—at least initially—more successful in the labour market than their *minkaomin* peers, the consequences for cultural self-identification can be significant. interviews with both *minkaohan* and *minkaomin* youth yield an intricate stratification of types of *minkaohan*, who may be located along an 'accommodation spectrum' ranging from 'pure' (fluent in the uyghur language and observant of islamic practices) to 'largely acculturated' (unskilled in mother tongue and uyghur cultural mores). the strength of relationships between *minkaohan* and *minkaomin* individuals depends to a very great extent on where a *minkaohan* locates him- or her-self on this spectrum, as i will show below. negotiations among the youth can be framed within a social constructionist theory of authenticity, whereby the 'authentic' is organised, performed and ultimately agreed upon by group—and/or subgroup—members. in this way, the boundary marking involved in negotiating the authentic becomes a flexible yet powerful process of evaluation with implications for the shaping of in-group identifications (vannini and Williams 2009). i present a case study of a rare example of a highly acculturated *minkaohan*, before ending with a discussion of the potential for all young urban uyghurs to become internationalist multilinguals: 'world citizens' able to negotiate hybrid identities according to time, place and necessity. in this last respect, i begin to explore the combined empowering effects of education and globalisation on uyghur youth identities, with reference to the ethnic politics of international sport and pop music pro-

#### 350 chapter seven

duction and consumption. In conclusion I suggest that, given the imposition of Chinese as sole language of instruction since 2002, we might expect urban youth studies—where youths are defined as multi-lingual and multicultural—to emerge as a major future trend in Uyghur studies research, with the *minkaohan-minkaomin* distinction becoming increasingly blurred.

# 7.1. *The First-generation* Minkaohan*: 'Internalised Oppression'*

As a people caught between China, Central Asia, the Middle East and the West, the Uyghurs are a fascinating subject for the study of identity formation, negotiation and hybridisation. Yet perhaps the most acute example of Uyghur in-between-ness—the self-identity of the *minkaohan* (Uyghurs educated in Chinese-medium schools)—to date remains relatively unexplored (exceptions are Taynen 2006; Smith Finley 2007a; Chen 2010). In Xinjiang, there have been three distinct 'waves' of *minkaohan*, emerging within different policy environments. The first was produced during China's 'twenty lost years' (1957 onwards), a period including the Great Leap Forward (1958-60) and the Cultural Revolution (1966-76). During this time, a split emerged within China's top Party leadership between those who favoured concessions to minority traditions and those who favoured immediate assimilation (an extension of the 'Red' versus 'Expert' dichotomy) (Dreyer 1968; 1976). In 1957, some Uyghur parents chose to educate their children in Chinese-medium schools as a means of 'ensuring their children's advancement in the new Chinese-dominated system' (Benson 2004: 195). Later, during the Cultural Revolution, most Uyghur-medium schools were closed, and all Uyghur children were obliged to attend Chinese-medium schools. There they were taught in the Chinese language, which subsequently became their first language if not their mother tongue. Among such individuals one can easily observe the linguistic and cultural legacy of a Chinese-medium education. Their level of spoken Uyghur varies, but rarely reaches native proficiency. At meetings, they may be seen taking notes in Chinese characters. In the company of other *minkaohan*, they frequently speak in Chinese or code-switch between Chinese and Uyghur even though no Han are present (cf. Taynen 2006: 48). In the words of one respondent named Räwiä in 1995: 'We usually speak Uyghur until we can't express something we want to say; then we get impatient and break into Chinese'. In many cases, a first-generation *minkaohan* learned the Uyghur language only after going to university, or well into adulthood following the relaxation of language and education policies in 1980. Räwiä's brother,

for instance, learned the Uyghur alphabet from their mother only at the age of 26, and taught himself how to speak and read Uyghur with the help of Uyghur-language newspapers and television channels.

A weak command of the Uyghur language often affects the ability of first-generation *minkaohan* to participate effectively in Uyghur social life. While conducting fieldwork in Ghulja in 2002, I watched as Polat, a *minkaohan* taxi driver in his forties, was unable to join in with Patimä, Kamil and friends in their singing and playing at the Shatliq restaurant. While his younger companions joined in heartily, Polat—the sole *minkaohan* present—remained silent. This did not however prevent him from declaring his pride concerning the musical affinities of group members: 'When it comes to music, one in every five Uyghurs sings or plays the dutar!' Nor was he chided by any of his companions. However, in other cases, frequent lapses into the Chinese language while attending Uyghur social functions can create a social barrier, feeding existing mistrust among *minkaomin*  towards *minkaohan*. Worse, a lack of familiarity with Uyghur dance can result in the tearful exit of a deeply embarrassed *minkaohan* guest (Taynen 2006: 49, 56). For Räwiä, the social strain attached to attendance at Uyghur social gatherings (*olturush*), which invariably include musical performance, meant that she usually preferred to stick to her inner circle of friends in 1995-6, which was composed entirely of *minkaohan*. As a child, she had not learned Uyghur dance (*usul*) at home or at her Han school. This, coupled with her imperfect grasp of the language, meant that she could not perform 'correctly' in traditional social settings. Probably for the same reasons, Räwiä disliked the Uyghur habit of 'house-hopping' (dropping in on friends and neighbours unannounced). This perceived tendency among *minkaohan* is one reason why southern and rural Uyghurs often claim that Ürümchi folk 'shut themselves behind closed doors' and are inhospitable.

However, it was notable that where ethnic attachment could be expressed in other ways, it was often clearly communicated. I observed that both Räwiä and Polat set their watches to 'Xinjiang time' (local time) rather than 'Beijing time' (imposed by the centre). Polat explained in 2002: 'Most people still use Xinjiang time, even those working in train stations, etc. But some other people have begun to use Beijing time now—this is another indication of assimilation. It is mainly the *minkaohan*, it goes along with speaking in Chinese and working with Hans in state work units.' While Polat worked independently from the state work unit system and could use Xinjiang time without any practical or social consequences, the fact that Räwiä chose to use Xinjiang time despite working in a unit where Chinese

#### 352 chapter seven

was predominantly spoken tells us much about the depth of her ethnic allegiance. To give another example, in spite of not knowing how to pray and feeling a general lack of fit with Islamic religious practices, Räwiä nonetheless maintained a sense of religiosity. This could be observed in her adherence to fatalism, so that she would often frame events—past, present and future—in terms of Allah [Uy. *Khuda]*. Describing her pregnancy and subsequent miscarriage in 2002, she said: *Khuda bärmidi … bärdi … yänä eliwaldi* (At first, Allah didn't give, then he gave, but finally he took away).

As suggested by the above, it was common for first-generation *minkaohan* to experience a sense of ambivalence towards ethnic group membership. Räwiä in her youth had evidently internalised the negative official portrayal of Uyghurs as 'backward' (in relation to the 'advanced' Han), and confided that she had wished she had been born a Han. When I first met them in 1995, both she and her brother made derogatory remarks about Uyghurs from time to time, in an apparent echo of Chinese state and media representations. On one occasion during that year, Räwiä warned me not to fraternise with Uyghurs from a certain work unit in Ürümchi, on the grounds that they were 'low-quality' (*suzhi di*); on another, her brother approved the official prohibition of foreigners to visit the countryside, and stated that 'some rural Uyghurs are savage' (*yeman*). Of relevance here is Helena Jia Hershel's description of the socio-psychological process of 'identity wounding'. Following earlier scholars (e.g. Epstein 1978), Hershel characterises identity formation as simultaneously an individual and collective process, influenced throughout one's life by personal and public relationships, and by broader social, cultural and racial attitudes. She argues that belonging—the fundamental core of human survival—becomes especially important when a person experiences exclusion. Exploring the case of biracial (also bicultural) individuals, she notes that lack of acceptance may emanate from one or both racial categories (cultural groups) at different times. Where there is tension between the two groups, the bi-racial (bi-cultural) individual may become internally conflicted because of 'the need to protect one's self from social or racial discord' (1995: 172). Or, as Thomas Hyland Eriksen puts it, they may find it psychologically and socially difficult to 'bet on two horses' (1993: 63). Where status is constantly and arbitrarily determined by the two main categories, or denied by both, an individual may begin to doubt their sense of self. One response is to try to pass as a member of the majority, although this carries dual risks. On the one hand, the individual may be condemned by the minority group for 'identifying with the oppressor'. On the other, while identifying with the

privileges associated with the majority group, the individual may also be influenced by its projections of inferiority, a process Hershel calls 'internalised oppression' (1995: 179).

Yet, as Breger and Hill (1998: 5) note in their study of mixed marriages, common to the experience of bicultural children is the way in which feelings of belonging and identity change, not only as they grow up and move through significant life-cycle phases, but also in response to changing relationships at and outside home. Later in life, Räwiä developed a strong sense of pride towards Uyghur cultural traditions. In her mid-thirties, she was alive to the social and economic injustices experienced by Uyghurs, an awareness partly fuelled by her own experiences of Han chauvinism over a decade of working with Han colleagues. By 2004, rampant employment discrimination against Uyghurs and the resulting high level of unemployment in the Uyghur community had led to Räwiä's almost complete realignment with her ethnic group ; she was now in her forties. This change was also translated in her physical appearance. Where she had once worn shoulder-length dark hair, muted colours and plain fabrics (more reminiscent of Han than Uyghur), she now streaked her hair red with henna and wore a dress made from a sparkly textile. She told me: 'As a person gets older, she has to find herself. But notice that the dress is knee-length and that I wear it sleeveless!'

One notable consequence of a Chinese-medium education was that it had provoked some first-generation female *minkaohan* to reflect on the respective gender roles and norms within the Uyghur and Han societies. In 1995-6, Räwiä sometimes spoke frankly about the pressures exerted on women by Islamic social norms, often revolving around the expectation that women 'give face' to their men-folk. She would have regular squalls with her husband about issues such as the female responsibility to provide hospitality for guests. On one occasion she came to my house in floods of tears after he became furious with her for declining to get out of bed at 1am to cook for him and his friends.

Some first-generation *minkaohan* embarked on marriages with Han Chinese, a scenario almost unthinkable at the present time (see Chapter 6). In some cases, this may have been driven by an internalised inferiority complex (shame of being Uyghur) engendered by the state intolerance of minority cultures prevailing at that time and a corresponding desire to associate with the Han culture. This phenomenon has also been observed in other societies during (or following) particular historical periods when an ethnic group was ostracised or reviled. For instance, in the wake of their

#### 354 chapter seven

incarceration during the Second World War, Japanese Americans were rejected as possible marriage partners by the white community, with the result that some *Nisei* (second generation) individuals began to display ambivalence towards their Japanese-ness as well as a desire to emulate White culture, which sometimes extended to interracial romance. In that context, the trend strengthened in the third generation, so that *Sansei* men and women carried 'exaggerated images of Caucasian attractiveness' and 'artificial notions of Japanese American inferiority' into their romantic encounters with whites (Spickard 1989: 44-46). In the context of blackwhite relations in the US in the 1970s, Beigel noted that (black) individuals who felt their personal value to be threatened by the alleged deficiencies of their group sometimes sought confirmation of their self-worth in interracial marriage: 'One of the compensations in interracial relations is represented by the elevation of self-value by having impressed or conquered a member of the culturally leading race' (1975: 82). Some first-generation *minkaohan* in Xinjiang may similarly have sought to enhance their sense of self-worth through marital union with the dominant group.

# 7.2. *The Second-generation* Minkaohan*: Product of Urban Instrumentalism*

June Dreyer (1976: 64) once argued that the earlier trend of minority nationality CCP members opting to school their children in Beijing or Moscow reflected deep ties to the urban intellectual elite—that which had also produced Han communists—and little consideration for the minority group. In contrast, the decision taken by some Uyghur parents in the 1990s to send their offspring to Chinese-medium schools represented a conscious strategy to increase the life chances of young Uyghurs (higher education, employment, social status, economic stability) in a context of spiralling employment discrimination. This strategy was not restricted to elite social groups with close ties to white-collar state work units or the Chinese Communist Party. Indeed, my observations suggest—although there exists no published quantitative data to support them—that the trend was more common among the merchant class (independent traders) and families where one or both parents worked in blue-collar roles, in short, among *minkaomin* (parents who had themselves attended Uyghur-medium schools).

### hybrids: identity negotiations among the urban youth 355

Märyäm, a migrant divorcee in her thirties from Ghulja, working in the music trade in 2002, confirmed that her daughter was studying at a Chinese-medium school: 'There's no way around it [Uy. *Amal yoq*], she has to be able to make it in this society … we are at the mercy of the Chinese [Uy. *Ularning qolida*]. Everything comes to us through them.' She added that she would teach her daughter the Uyghur language herself, and to that end maintained a strict 'no Chinese' policy in the home. Patigül, another music trader in her late twenties native to the city, presented a similar argument in 2004: 'I have two boys. They're both at Han schools to give them the best chance in life … otherwise they'll have no possibilities open to them.' Like Märyäm, she confirmed that she intended to teach her sons Uyghur at home. However, she acknowledged that the strategy had yielded mixed success; while her older son spoke Uyghur at home, the younger one often got stuck. At such times, she tried to help him find the right word in Uyghur, but he inevitably ended up code-switching between the two languages. Helil, a kebab trader in his early forties, told me in 2002:

My son's 5 and he already knows the Uyghur language really well. He knows the odd bit [Uy. *anchä-munchä*] of Chinese too. He didn't go to kindergarten; he was cared for by us, his relatives. That's why his Uyghur is strong. But we're going to put him in a Chinese school, because the Uyghur language is of no use [Uy. *paydisi yoq*]. I studied at a Uyghur school for 9 years, so did my wife. And now we're selling kebabs! It's impossible to get a job without Chinese. Yes, we make more money selling kebabs than people in the state work units (Uy. *khizmätchilär*) make, but we can't sell kebabs when we're old!

Zöhrä, a woman in her thirties working as a chambermaid in an Ürümchi hotel, expressed a similar view in 2002: 'We'll send our kid to a Han class. Because the Uyghur language doesn't function in society [Uy. *ishlimäydu*, Ch. *yongbushang*]. It's no longer needed. You can't get a job speaking Uyghur, you must be able to speak Chinese, and speak it well.' She added that when she came to the city (from the outskirts) to find a job, the first thing employers wanted to know was whether she could speak Chinese, regardless of the fact that chambermaid services require few frontline communication skills. Because she did not speak Chinese fluently [Uy. *toluq ämäs*], she had been forced to resort to nepotism in the form of an introduction by a relative in a senior position at a particular hotel. She stated that the use of contacts (a method omnipresent in China proper) was the only way for Uyghurs to get work in Ürümchi, if not proficient in Chinese. Another important motivation for choosing a Chinese-medium school was the desire that her children have the opportunity to study English, since she had

#### 356 chapter seven

heard that English proficiency had eclipsed the importance of Chinese proficiency in the job market. At a Uyghur-medium school, her children would be compelled to learn Chinese as their second language, while at a Chinese-medium school the second language taught would be English. Patimä, a divorcee in her early thirties and dancer in a Ghulja hotel, explained in 2002:

My older boy is going to a Han school—he is sharp [Uy. *ötkür*]. He's doing well; his teacher says he is very intelligent. But my younger one is going to a Uyghur school. He's most interested in cows and sheep! He doesn't like the Hans; says he won't go to a Han school. I want the older one to get a good job, not end up like me … but he is also learning the Quran, and prays five times a day!

This was a very interesting case. I noted that the older son liked to speak the Chinese he knew, and took a sense of pride in his linguistic achievement. At the same time, his grandmother (who said she 'hated' the Hans) had made the decision that he would simultaneously learn the Quran.1 Although Chinese law did not permit him to pray at school, he nonetheless frequently attended the mosque or came home in order to pray. I will explore more examples of the possibilities for hybrid practice in Section 7.6. The younger son, meanwhile, had been permitted to attend a Uyghur school, partly because he showed less academic potential than the elder but also because he had expressed aversion to the idea. This suggested that parents would not compel children to attend Chinese-medium schools, even if they felt that option to be the better strategy. Some parents had gradually changed their approach to education over time, in response to changing labour market conditions. For instance, Gülnur, a *minkaomin* kebab trader in her early thirties, explained in 2002:

Our eldest is at a Uyghur school. He's really intelligent and loves studying. He gets marks like 93 all the time, not like those kids who only get 40 or 50 … he won't play with them, he'd rather read. And we tell him not to mix with them. But we're thinking of sending the younger one to a Chinese school, because that's what employers want now. I know, it's a threat to the Uyghur language and culture; that's what my husband said when we were deciding what to do with our first child.

<sup>1</sup> Bellér-Hann and Hann (1999: 17) reported a similar situation while doing fieldwork in rural Xinjiang. While a rural official's wife explained that she had never been properly taught how to pray, they suspected that the couple's children were receiving religious instruction from the maternal grandparents.

In this way, despite being cognisant of the risks involved, this mother had decided to put her younger son's socio-economic prospects before the native language and culture.

On the other hand, by the turn of the century, a backlash had begun among the Uyghur urban community against Chinese-medium schools. Often, parents who reverted to placing their children in Uyghur schools had themselves studied in Chinese-medium schools during the Cultural Revolution. One example is Anarkhan, a first-generation *minkaohan* who was working as a chambermaid in an Ürümchi hotel in 2002:

I originally wanted my kids to go to a Han school. I thought they could learn Chinese well there and achieve a high cultural level. Then they would be able to find better jobs. But, in the end, I worried that I wouldn't be able to help them with their homework. You see, I only speak Chinese; I can't read or write it. So we sent them to a Uyghur school. And they can learn Chinese there all the same, and English, and computing.2

While Anarkhan echoed the desire of other respondents to increase employment prospects for her children, her assumption that they would somehow obtain a 'higher cultural level' at a Chinese-medium school implied an internalised inferiority complex, such as that observed among other first-generation *minkaohan* (see Section 7.1) She went on to explain:

I did learn Chinese at school but wasn't able to learn it well. I have learned it here at the hotel; they have special classes in Chinese for the Uyghur staff. I have nothing in my life. I just work here all day, cleaning, and then go home to the cooking and the children. I want my kids to do better for themselves. They have a private tutor who comes in every week to teach them Chinese language and Maths. I want them to go abroad to study.

She seemed glad of the opportunity to improve her Chinese proficiency, and spoke warmly of her Han colleagues.

The low self-esteem pervading this respondent did not characterise all *minkaohan* parents opting for Uyghur-medium schools. Räwiä was furious to hear in 2004 that some group members had denigrated the Uyghur language as 'useless', arguing:

The mother tongue has to be the base from which to learn other languages. Who are these people who say "Uyghur is of no use"? What kind of talk is that? If all Uyghurs sent their children to Uyghur schools, then the Hans

<sup>2</sup> This respondent stated that many Uyghur schools in Ürümchi offered English and computing classes in 2002, which represented an advance on the situation in the past. Certainly, I observed that a huge array of Uyghur-English dictionaries was on sale at the city's official Xinhua bookstore.

#### 358 chapter seven

would have to introduce policies that prevented discrimination against *minkaomin*; they couldn't simply stand by and watch all Uyghurs become unemployed! But Uyghurs—especially *minkaomin*—send their kids to Han schools and so contribute to the gradual disappearance of the mother tongue. And the kids grow up fluent in Chinese and have this sense of prestige [Uy. *äbroy däp oylaydu*] because they speak Chinese, while their parents don't, or speak it poorly.

She thus confirmed my impression that *minkaomin* parents were more likely to opt for Chinese-medium schools. Polat, first-generation *minkaohan* in Ghulja, similarly asserted in 2002 that it was paramount for young Uyghurs to learn their mother tongue—and know their ethnic group—before doing anything else: 'Kids can learn Chinese in Ürümchi as they grow up—it's all around them.' He added that many Uyghurs had begun to think in this way, and confirmed that his two sons would go to Uyghur-medium schools. While he acknowledged the desire of some parents to safeguard their child's future and ensure a good standard of living, he claimed that the negative effects of a Chinese-medium education were already in evidence: 'Some Uyghurs are forgetting the Uyghur language because they speak Chinese all the time.'

Young respondents also reported the trend for *minkaohan* parents to revert to Uyghur-medium schools. Dilbär, an 18 year old *minkaomin* student, described one *minkaohan* couple living in her neighbourhood in 2004:

They have a child, but they've put her in a Uyghur school. They don't speak Uyghur very fluently, and fall back into Chinese as soon as they can't find a word. But they absolutely won't speak any Chinese at home in front of the child! They say 'Don't let her become like us' (Uy. *bizdäk bolmisun däp*).

Like Anarkhan, she argued that a Uyghur-medium education no longer disadvantaged pupils since they could study both Chinese and English from the age of seven:

Three languages have to be better than two, right? And parents are starting to realise that it's dangerous to send kids to Han schools. They see a new generation of Uyghur youth that can't speak its mother tongue […] In any case, if a company doesn't want you [i.e. as a Uyghur], it won't hire you regardless of whether you are *minkaohan* or *minkaomin.*

Perhaps the most progressive approach to the education dilemma was that proposed by Ghäyrät, a petty entrepreneur in his thirties originally from Aqsu. He explained in 2002 that he intended to send his daughter first to a Uyghur-medium school for two years, to enable her to get a solid grounding in Uyghur. Then, he would have her start primary school afresh from Year

1, this time in a Chinese-medium class: 'Because she has to be fluent in Chinese, right? Or she won't be able to get a job anywhere in Ürümchi. But it's important for her to know her mother tongue.' He added that he had thought about this issue a lot, but had ultimately decided that although his daughter could learn both Chinese and English at a Uyghur-medium school, she would not reach as high a level in either compared with an education at a Chinese-medium school.

# 7.3. Minkaohan *Self-identity in Contemporary Xinjiang: 'Ethnic Anomalies'?*

Given that Chinese-medium education is increasingly a fact of life in the urban capital, what can be said about the self-identification of the *minkaohan*? Do they enjoy better relations with Uyghur *minkaomin* or with Han peers? Alternatively, do they feel excluded from both groups? Elsewhere, I have ruminated on the term 'ethnic anomalies' (Smith Finley 2007a), as applied to groups of individuals who find themselves 'betwixt and between' two cultures, for example, the offspring of mixed-race couples, or children of immigrants in Europe (Eriksen 1993: 62). In the past, children born to mixed-race couples were frequently classed in pejorative terms such as 'half-caste' or 'half-breed', although initiatives in anti-racist education have more or less successfully re-cast them as individuals of 'mixed race' or 'mixed heritage'. As for second-generation immigrants in Europe, these individuals are often bilingual in their mother tongue and in the national language of the host country and, as such, find themselves caught between their native culture and the host culture. Each group may experience conflicting cultural loyalties, finding themselves caught between the dual possibilities of assimilation to the dominant group or ethnic incorporation (resistance). Drawing upon Mary Douglas's concepts of anomaly and ambiguity, I have considered whether the term 'ethnic anomalies' might also be applied to contemporary Xinjiang to characterise children from the Uyghur nationality (a minority group at the national level) educated in the language of the culturally distant (Han) coloniser. Here, the term anomaly is defined as 'an element which does not fit a given set or series', while ambiguity constitutes 'a character of statements capable of two interpretations.' (Douglas 1966: 38) Assuming that cultures will normally react with anxiety to the advent of anomaly, Douglas suggests that we may choose to

#### 360 chapter seven

ignore anomalies, or to condemn them. On the positive side, we may allow a new pattern to evolve in which the anomaly has a place.

In early childhood, *minkaohan*—like *minkaomin*—are cared for by Uyghur parents, who will normally give them a traditional family education in Uyghur socio-cultural practices and life cycle rituals, many of these influenced by Islam.3 Upon reaching school age, however, they attend Chinese-medium schools, with the result that Chinese gradually replaces Uyghur as their first language, and where they are exposed to contrastive Chinese notions of culture (*wenhua*). Finally, they enter adult life to find that the two worlds they span are fundamentally divided by inter-ethnic tensions, while their partial sinicisation earns them only partial entrance to the sphere of Han privilege. Jennifer Taynen described *minkaohan* as 'inhabiting an ill-defined and uncomfortable middle ground'; as 'contending with levels of isolation and discrimination not encountered by individuals who remain within the linguistic confines of their ethnic communities' (2006: 46); and cited a *minkaohan* respondent who characterised herself as a perpetual outsider observing other people's cultures (Taynen 2006: 56). The source of their personal discomfort and social ostracisation derives precisely from their anomalous position, which spans Turkic-Islamic language and culture on the one hand and the Chinese language and Han culture on the other. Notions of anomaly and ambiguity figured prominently in the self-analyses of my *minkaohan* respondents. In the summer of 2011, I received a letter from a young woman whom I will call Tursungül. Introducing herself as 'a Uyghur girl from Xinjiang', she explained that she had come across my online profile while searching the internet for information about 'Uyghur *minkaohan*'. She was in a state of great excitement. I will quote her letter (which was written in English) at length to allow her to tell her story:

I am a 100% 'minkaohan'. I was sent to a Han kindergarten when I was 2 years old; there were only a few Uyghur children went to Han kindergartens or schools back then. Fortunately, I learned Chinese so quick, so well, and I also did an excellent job in school. I had even been treated like a little star since primary school, all my teachers praised me and my parents were so proud of me. I went to Shenzhen city for high school in 2000, I worked so hard and in 2004 I entered University of International Business and Economics in Beijing, majored in accounting. After graduation, I worked for one of

<sup>3</sup> Uyghur socio-cultural practices are here conceived as one branch of Central Asian practices, within which most scholars consider the intertwining of religion and culture to be axiomatic.

#### hybrids: identity negotiations among the urban youth 361

the 'big four' [transnational accountancy firms] for 2 years. I missed my family and I wanted to come back to my hometown, Xinjiang, and I did.

I've been back to Ürümchi for one and a half years; now I'm working in the finance department of […] a famous NGO. I really like what I do now, because I've always wanted to do something for my hometown, for my people; though I don't directly participate in the project implementation, I still feel good that I can help manage the funds to be used properly. However, nothing is as easy as I thought it would be, because I have left home too long ago. I'm not used to how things are done here, how people think here; the worst is my own family doesn't understand me well.

This woman represents the first cohort of 100 per cent *minkaohan* graduates, having been educated in Chinese from kindergarten through to university level. She excelled at school, quickly mastering the Chinese language and earning the respect of both teachers and parents. In 2000, she was among one of the first groups of students selected to pursue their high school education in coastal cities of China proper, as part of the *Xinjiangban* or 'Xinjiang class' project.4 There she spent four years—an initial year spent improving Chinese language skills followed by three years in high school. Her diligence in her studies won her a place in a good university in the Chinese capital and a first post in an international accounting firm. Yet despite these successes in the spheres of education and work, and the positive self-image they must have fostered, Tursungül still felt that she missed her 'hometown'. After returning to Xinjiang, she once again secured a high-level post, this time with an international non-governmental organisation, and derived a heightened sense of reward from her new-found ability to 'do something for her people'. Yet at the same time she felt as though she did not quite belong to her natal community. She did not understand how to behave in different social situations, could not get used to the way in which other Uyghurs thought, and worried that her family could

<sup>4</sup> The 'Xinjiang class' project, initiated in 2000, constitutes a national policy to run four-year boarding high school classes (Inland Senior High School Xinjiang Classes, *Xinjiangban*) in local Han majority schools in eastern cities of the PRC. The stated policy aims are to foster a cohort of young Uyghur talents through the national curriculum and to enhance ethnic integration. Since 2005, the estimated total enrolment has exceeded 20,000 individuals in 500 classes across 24 inland cities (Chen 2010: 2). The project might be compared with the practice employed by the American government in the late nineteenth century, whereby Native American students were separated from their communities and educated at boarding schools run by religious missionaries. At these schools, which acted as a 'civilising' tool or means of assimilation, students were prevented from speaking their own languages or following their own religions (Vasquez and Wetzel 2009: 1572).

#### 362 chapter seven

not relate to her. Being of an intellectual mindset, she set out in her letter to try to understand her predicament:

So I started to think what caused this […] I think of the following reasons: I have been taught in Chinese since I was a little kid: at first, I didn't speak Chinese at home, but gradually I had to say some words in Chinese to make myself clear when talking to my family members, because I couldn't speak my mother tongue well; it's a shame.

All of my *minkaohan* friends communicate in Chinese instead of Uyghur language: even though I insist that we talk in mother tongue, they prefer Chinese because it's more convenient through communication.

As to the cultural reasons, we know too little about our own history and culture, all that we were taught was the Chinese history, all we have received was the Chinese way of thinking.

We are more open to the world, this is a good aspect I think, we have a strong desire to learn about the outside world, we embrace the different cultures, we welcome different opinions, and all of these make us different from other Uyghur people.

As you may have known, Uyghur *minkaohan* are called the '14th ethnic group of Xinjiang'. Sometimes we are isolated by Uyghur people, sometimes we cannot be understood by others; I am confused, and I am seeking for some answers.

Tursungül identified the start of her troubles as a gradual deterioration of her grasp of the Uyghur language and greater ease of expression when using Chinese. This infused in her a sense of shame, but it was difficult to reverse the process because her *minkaohan* peers found it more convenient to converse in Chinese. The combined sense of discomfort, loss and humility surrounding an incomplete grasp of the mother tongue can also be understood in the following two examples. The first consists of an online post in 2005 to an English-language discussion forum focusing on the Uyghur language5:

I am Uyghur boy, I love to understand myself and my language […] I am also so glad to see such a group on Uyghur language, but actually I must confess that I sometimes cannot communicate with you because some of you speak in Uyghur and I do not know so many details about that language. If you care for this mishap, I will be more thankful to you, group members.

The second example concerns the reaction of a young female *minkaohan* in Ürümchi, when asked by an American-born Uyghur postgraduate student to express her opinions on Uyghur culture *in Uyghur*. According to

<sup>5</sup> Abdulla Adilä, 'It is a great achievement to have like this program', UyghurLanguage @yahoogroups.com, 6th October 2005.

the interviewer, an expression of fear rapidly spread across the young woman's face, before she admitted that her spoken Uyghur was 'really bad' because she had studied in Chinese schools all her life (Eri 2008: 76).

Returning to Tursungül's letter (2011), she goes on to state that a Chinesemedium education has a deep impact on how a person situates him- or herself historically, as well as on the way in which a person thinks and behaves (i.e. social and cultural identity). For these reasons, she often feels excluded or exiled as part of an anomalous group, sarcastically dubbed the '14th minority nationality'. On the other hand, her experience is not entirely dominated by anxiety. She also enjoys a sense of self-esteem from the positive qualities she feels she possesses in comparison with Uyghurs educated at Uyghur-medium schools, characterising *minkaohan* as more adventurous, more broad-minded, and more tolerant of difference.

In a second letter I received in the same year, Tursungül described her relationships with peers. Regarding her interactions with Han classmates and colleagues, she wrote:

I enjoyed living in Beijing. During the four years studying in Beijing, I made a lot of good friends (both Han and Uyghur); we got along very well back then and still keep in touch. After graduation, I started my first job with one of the biggest auditing firms, and that was also a valuable experience. I met many outstanding colleagues from different places in China and some from Hong Kong and Malaysia. I learned a lot while working with them, and some of us became good friends as well. Actually, most Han do not know much about Uyghur culture or Uyghur people. I think I played a role as a bridge between these two cultures; I introduced Uyghurs to them, and they changed the misconceptions they had in mind before.

The above narrative demonstrates that, removed from the everyday cultural and political tensions that characterise life in Xinjiang, it is possible for an acculturated Uyghur (one familiar with the Chinese language and culture) to form not only good working relations but also friendships with Han peers. Moreover, such a person has the capacity to overturn typical Han prejudices towards minority nationalities and the Uyghurs in particular. That said, Tursungül conceptualised her working experiences in Beijing less within Chinese national boundaries and more within the *international* sphere, highlighting both her position at a global firm and the multinational nature of its workforce.

Despite her positive experience in Beijing, Tursungül eventually decided that she did not 'belong' there, and acceded to parental pressure to return to Xinjiang. Yet over the course of a year living in the regional capi-

#### 364 chapter seven

tal and interacting with her family for the first time since graduating from middle school, she became acutely aware of the uniqueness of her social, cultural and political position as a *minkaohan*. She writes:

I want to share an insider's view […] toward the subtle relationship between the dominant Han culture and the Uyghur sub-culture. *Minkaohan* is a special group in between the Uyghur culture and Han culture. *Minkaohan* are planted in the Uyghur culture, which affects us in all aspects, even when we are far from home, e.g. studying in inland cities or living abroad. We magically attract each other, because that magical thing is in the blood.

Being educated in Chinese from an early age has led to the inveteracy [deep-rootedness; habit] of Chinese culture in *minkaohan*, which also makes us stand at the margin of these two different cultures. We absorbed the superiority of the two cultures, but we are confused when we encounter cultural conflicts regarding identity issues. The mother culture-Uyghur culture is in the blood; we cannot eliminate the impact of our own culture. The formation of [cultural] characteristics, the acquired behaviors: all are ironed [conditioned] by Uyghur culture.

Thus, while *minkaohan* internalise the foundations of Uyghur socio-cultural practices when very young (in a process so seamless and natural as to be described as 'magical' and 'in the blood'), they inevitably go on to acquire learned behaviours from Han peers during their school and university years. This leaves them standing at the cusp of the two cultures, making them open to the best features of each but also vulnerable to the conflicts generated by difference. During their formative childhood years, *minkaohan* tend to interact quite amicably with Han peers. Yet as they mature and gain deeper insights into the ways of the world, they gradually move apart from Han school friends and try to realign themselves with the Uyghur community (a phenomenon also observed among first-generation *minkaohan*, see Section 7.1):

We get along very well with Han, we have many good Han friends, but we unconsciously tend to make Uyghur friends as we grow up. We just unwittingly come closer to Uyghur culture and the Uyghur community, and try to set up a small community of our own. When we meet cultural conflict, we are the first to feel it, and it is more obviously reflected upon us. The [Uyghur] sub-culture might have fallen behind in some manners, but we were taught like that, and we believe that is the right thing to do, e.g. we were told that marrying Han was wrong. We admit that facing the strong merging [assimilatory] power of the dominant culture, we stay instinctively vigilant. This is reflected in *minkaohan* in the way we reject the popular culture of Han, such as Chinese music, movies or dress and personal adornments. In this way, *minkaohan* show their reluctance to accept the Han culture 100%.

In this section of the letter, we find Tursungül in personal conflict. Brought up to believe instinctively that intermarriage with a Han Chinese is 'wrong', she is later confronted with charges of 'backward' thinking by Han peers who see intermarriage as a symbol of modernity (having themselves internalised government discourses which promote ethnic harmony through social engineering). Her case also demonstrates that *minkaohan*—like *minkaomin*—may experience a powerful counter-reaction against the attempts of the centre to forcibly assimilate the peripheries. Though they speak Chinese, *minkaohan* may nonetheless express an alternative cultural identity through the avoidance of Uyghur-Han intermarriage and consumption of alternative forms of popular culture.

In the final section of the letter, it becomes clear that Tursungül conceives of her identity not simply as one straddling two cultures within the Chinese polity (Uyghur and Han) but as one capable of projecting itself well beyond China's national boundaries:

The strike [impact] of the dominant culture opened a new window for *minkaohan*. We needed new knowledge to fill [meet] the needs of new culture as Uyghur culture expanded. At this time, we chose to embrace the Western culture, while [using] the Han culture as springboard. So we *minkaohan* possess the advantages of the three cultures.

As to the relationship with Han, I want to say that we are all human, we are all equal, we do not recent [resent] anyone; we just pursue peace all over the world. We have good Han friends, and I think we play some kind of role as a bridge between Uyghur and Han.

Tursungül's remark about *Xinjiangban* students acting as a 'bridge' between the two groups implies on the surface that the policy aim of improving ethnic integration may be succeeding; it evokes Ben-Adam's earlier conceptualisation of Uyghur school teachers as 'agents of acculturation' for the state, who help to ameliorate inter-ethnic relations by straddling China's national curriculum and local Uyghur culture (1999: 205). On the other hand, Jennifer Taynen's study suggests that while *minkaohan* possess the potential to facilitate reconciliation, this has so far been achieved only by the first cohort of *minkaohan*, and only at a time when Han were present in small numbers and generally opted to adapt to local cultural mores (2006: 50). I would argue that, in stressing equality between all individuals of the world, Tursungül appears to advocate 'internationalism', going beyond the Uyghur-Han duality and seeking to occupy a boundless space as 'world citizen'. She does not equate Chinese modernity with global modernity; rather, she explicitly underlines her instrumental use of the Chinese

### 366 chapter seven

language and culture as a 'springboard' to a less bounded, more complex and diverse modernity. Under the present climate in Xinjiang, however, there is little room for cosmopolitan tolerance towards Han people, and so Tursungül has begun to ask herself some tough questions. Her letter continues:

What kind of life do I really want? I have encountered some difficult times during the year, I am confused. I have had the idea of studying abroad for a long time, but never really taken any actions. However, recently, I reconsidered this and started my preparations.

In this way, like some mixed Uyghur-Han couples (see Chapter 6), Tursungül is contemplating whether the only way to combat her sense of double exclusion is to leave Xinjiang altogether.

# 7.4. *The* Minkaohan-minkaomin *Divide*

Lucian Pye, writing in the 1970s, predicted that 'the reaction of minorities to […] increased pressure for political conformity is likely to be uneven. Power relationships within each minority community may shift as a consequence of Han pressures and the result is likely to be greater internal ethnic tension' (1975: 508). Particularly in the wake of the drive towards Chinese-medium education and the resulting linguistic and cultural diversification of the Uyghur community, this has indeed occurred, with an internal split emerging between *minkaomin* and *minkaohan*. According to some Western observers, minority nationality students who received a Chinese-language education in the 1990s tended to speak, dress and act like Han students. While often considered a source of prestige from the perspective of Han society, this was an embarrassment from the viewpoint of local ethnic identity (Dwyer 2005: 38). However, not even all Han considered the alleged 'raising of cultural level' to be a positive development. Susan Blum cites a Han friend for whom a 'real' (*zhen*) minority was one whose identity was 'full' (*shi*) (of visible differences) as opposed to a 'false' (*jia*) minority whose cultural identity was 'empty'(*kong*) (2001: 6). Just such a perceived division was observed among the Uyghurs of the Kazakhstan borderlands. There, 'sub-ethnic boundaries' were constructed among separate migratory streams of Uyghurs, first between *yärliklär* ('locals', i.e. Uyghurs born in Kazakhstan) and *kegänlär* (Uyghurs who migrated to Kazakhstan during the fifties or sixties); then, in the 1990s, between *khitayliqlar* (temporary sojourners from China, in Kazakhstan on business) and

*sovetliklär* (Uyghur citizens of Kazakhstan, i.e. the *yärliklär* and *kegänlär*  combined). Each sub-group vied with one another for status as the true representatives of a 'pure' Uyghur culture (Roberts 1998a: 512-513, 519-519).

This simple binary model of 'pure' versus 'fake' was frequently evoked by *minkaomin* respondents during my fieldwork in 1995-6. When making introductions, *minkaomin* often classified their companions (and themselves) in terms of whether they had attended Uyghur- or Chinese-medium schools, the implication being that *minkaohan* constituted a separate category. In private, many *minkaomin* claimed that *minkaohan* had developed cultural characteristics more commonly associated with Han Chinese. Most significant among these was the suggestion that *minkaohan* often acquire the 'Han disposition' (*Khänzu mijäz*) (see also Chapter 2 Stereotypes).6 *Minkaomin* respondents often communicated a deep sense of suspicion towards *minkaohan* peers, believing that since *minkaohan* had grown up alongside Han schoolmates, they must be prone to political collaboration (cf. Taynen 2006: 57). Consider the way in which one *minkaomin*  respondent introduced me to his good friend in 1996: 'This is X. Don't worry, although he is *minkaohan*, he hates Hans as much as anyone else!' In this way, a 'black sheep effect' could be observed, whereby members marginalise a sub-section of the group and declare any negative traits associated with the group as a whole as associated only with that sub-section (Marques and Paez 1994). This process of internal stigmatisation recalls Henri Tajfel's theory that stereotypes of other groups (in this case, a subgroup) allow the in-group to maintain 'positive distinctiveness' (summarised in Cinnirella 1997: 44-47). Interestingly, Uyghur citizens of Kazakhstan articulate similar prejudices against the Xinjiang Uyghurs, claiming that their passage through the Chinese state education system has resulted in their assimilation into the Han culture and willing sinicisation as they attempt to 'climb the ladder of success' (Roberts 1998a: 520-521).

*Minkaomin* suspicions were further fuelled by the fact that *minkaohan* tend to belong to state work units, where there is ample potential for collaboration with the majority group. Räwiä, a first-generation *minkaohan* based in Ürümchi, found herself in just such an awkward position, as she

<sup>6</sup> Such attitudes towards Chinese education were already in evidence in Xinjiang in the late nineteenth century. Hossain Bay Bacha, an advocate of reform of Muslim education, was critical of Chinese educational reforms on moral grounds, claiming that Chinese methods made boys disrespectful to their elders and more likely to gamble (Bellér-Hann 2008: 334).

#### 368 chapter seven

was frequently asked to translate English-language articles on the Uyghurs so that Han officials might know their contents. While protesting in 1996 that if she refused the work, someone else would do it all the same, she nonetheless confessed to 'feeling like a tool.'

Yet as Uradyn Bulag has observed with regard to contemporary Inner Mongolian identities, 'hybrids at the geopolitical borderlands are people upon whose bodies and consciousness national states map their power, often in contradictory or antagonistic ways' (2002: 20). Put differently, sinocentrism, Han nationalism and, more recently, Chinese-medium education can and do produce diverse results among peripheral subjects, ranging from near-complete acculturation through partial acculturation to ethnic incorporation and symbolic (or violent) resistance. This diversity is also reflected within the *minkaohan* and *minkaomin* sub-groups so that, by the early 2000s, *minkaomin* reactions to *minkaohan* had become more discerning. Now, opinion towards *minkaohan* was governed by an onlooker's perception of the extent to which an individual desired and strove to maintain a distinct Uyghur cultural identity. Thus, the distinction was less about cultural attributes *per se* and more about a person's level of self-awareness and strength of in-group loyalty. Linda Benson points out that some minority cadres in Xinjiang use their positions to assist their relatives, and view their jobs as a means of helping fellow minority individuals (2004: 211). Arienne Dwyer distinguished between *minkaohan* who sacrifice aspects of Uyghur culture for the sake of practical convenience and those who reject non-Uyghur behaviours 'as a matter of ethnic identity'. She gives the example of computing input methods, noting that while some find the Chinese *pinyin* transliteration system more economical and intuitive when inputting Uyghur words into a computer, others carefully avoid it: 'For some, *pinyin* smacks of Chineseness' (Dwyer 2005: 23). This preference has proved so strong in Ürümchi that since the mid-1990s, Uyghur software engineers have formed numerous commercial enterprises to market addon keyboarding utilities for Arabic script-based Uyghur, desktop publishing tools, modified operating systems, translation tools, and so on (Dwyer 2005: 24-5).

In the early 2000s, many of my respondents continued to react scathingly to *minkaohan* whom they considered to have no ethnic awareness or national conscience. Adil, a 45 year-old male restaurateur, was visibly disgusted by Gülshäm, a highly acculturated 20 year-old *minkaohan* woman who visited his establishment one evening in 2004. By communicating with her father solely in Chinese and treating him as if she were his equal, Adil

felt that she displayed social behaviour entirely inappropriate of a young Uyghur girl:

That girl's no good! She doesn't understand [Uyghur] social mores! She and her father are just like Hans, they speak Chinese all the time and act as if they were friends … parents should be parents, and children, children. Uyghurs should go to Uyghur schools; otherwise they pick up Han habits, the Han temperament. They come home from school and don't ask their parents how they are … they're peevish, moody … they slam out of the house in the mornings.

A young woman who understood—or, perhaps more importantly, tried to understand—Uyghur social practices would have shown respect to her father by addressing him in the mother tongue and by accepting his authority as a father and an elder. The situation was exacerbated when Gülshäm proceeded to pick up her pet dachshund and kiss him. The restaurateur mimed the act—with great sarcasm—and immediately pronounced it 'un-Islamic' (the dog being considered *haram* in Muslim culture): 'Just like a Han! It's no good, Uyghurs shouldn't do that; Muslims shouldn't do that.'

Abdukerim, an observant *minkaomin* graduate in his twenties from Aqsu, declared in 2002 that the vast majority of *minkaohan* were 'brainwashed', but conceded that the odd few 'had a brain', citing a former girlfriend as an example: '*Minkaohan* like to find *minkaomin* partners. It's because they feel they've lost something; they're afraid that if they marry another *minkaohan* they will never have the occasion to speak Uyghur or interact with Uyghurs … they will speak Chinese all day at work, then come home and speak it there too.' Abdukerim thus echoed the notion expressed by the Han respondent in Blum's (2001) study in stereotyping *minkaohan* as empty and lacking, and in implying that *minkaomin* were better, richer, fuller. He added that he had hoped to marry this girl, but her parents had not agreed to the match because he was not attached to a state work unit. This demonstrates the importance of professional status, wealth and economic security in *minkaohan* households. What followed next was very interesting. Abdukerim and I had been chatting in one of Ürümchi's transnational fast-food outlets (reputedly *halal*) when two Uyghur teenagers came in, speaking to one another in Chinese. At least on the surface, they appeared to be deeply acculturated. Abdukerim was visibly repulsed, declaring them examples of 'brainwashed' *minkaohan*, and condemning the fact that they were conversing in Chinese, despite being Uyghurs. When they sat down at the next table, he took the opportunity to harangue the girl about the importance of loyalty to her ethnic group. At first, she was

#### 370 chapter seven

clearly baffled but remained polite. Eventually, however, she began to defend herself, protesting: 'You have no right to tell me I have no "ethnic spirit" [Uy. *milliy roh*] … it's no business of yours if I decide to speak in Chinese. Anyway there's no advantage in learning Uyghur.' In this way, she echoed the arguments of some parents who chose to place their children in Chinese-medium schools in the 1990s.

Abdukerim next asked her who she intended to marry. She replied that she planned to marry another *minkaohan*, demonstrating an awareness of the liminality of her sub-group vis-à-vis the Uyghur cultural community. Her response reflected an emerging pattern in Xinjiang whereby *minkaohan* individuals frequently seek marriage partners from among other *minkaohan*, a tendency which suggests that the *minkaohan* sub-group may, as some predicted, have begun to form a 'third community' with shared linguistic and cultural habits, shared loyalties, and shared political views (Taynen 2006: 46, 59; cf. Smith Finley 2007a: 234). Studies show that there is a tendency for marginalised individuals to be attracted to marriage partners who are marginalised in similar ways. While some women in Khatib-Chahidi et al's (1998: 52-3) sample were initially attracted to the (cultural) 'difference' in their partner, they were on the whole attracted by individuals who had similar personal experiences of marginality, be it within the family, community or ethnic group. In this way, feelings of marginality served to maintain the bond between fellow 'outsiders' (Khatib-Chahidi et al 1998: 62). A similar process could be observed among separate migratory streams of Uyghurs in the Kazakhstan borderlands during the fifties and sixties, with *yärlik* (local-born) and newly-arrived Uyghurs refusing to allow their children to 'inter-marry' with the other sub-group. Yet intermarriage between *yärlik* and second-generation immigrant Uyghurs became possible later, as boundaries broke down with the passage of time (Roberts 1998a: 515).

While Abdukerim debated with the older sister, her younger brother, aged 15 or 16, told me that their parents were both *minkaohan*, and that his mother taught Chinese as a second language in a city middle school. The natural result of such a family background and home environment was that the two had grown up with very little Uyghur language proficiency. The brother explained: 'I can understand Uyghur if spoken slowly, but I always speak Chinese at home. I don't speak in front of my grandparents, though, as a kind of respect.' Thus, despite finding it hard to speak or understand Uyghur, the boy nonetheless understood and maintained the important social rule requiring young Uyghurs to address older Uyghurs in

the mother tongue. This aspect is mentioned also by Vasquez and Wetzel, who cite a young Mexican American woman as stating that maintaining one's native language is 'a really important way of not forgetting your roots' and of 'valuing where your ancestors came from' (2009: 1563).

Dilbär, an 18 year old *minkaomin* student, described in 2004 a conversation she and a *minkaomin* peer had held with a *minkaohan* middle school pupil they knew:

We bumped into him the other day on the street and were talking about music. He asked us whether we liked Zhou Jielun [a popular Han rap artist]. We said we preferred the *On Ikki Muqam*![The Twelve Muqam, a fixed suite consisting of sung poetry and stories, dance tunes and instrumental sections; associated with the Uyghur nation, but played in different forms across Central Asia and the Arab world]. He was astonished! He couldn't believe it. He said he would never get into that … We told him he just didn't understand it yet, that it was too deep. I said 'If you don't come to like it in a few years' time, I'll bring you my head!' Eventually, his expression changed and he looked less certain, and asked 'Really?' He had a Zhou Jielun CD in his pocket …

This *minkaohan* teenager perhaps represents one of the more deeply acculturated young Uyghurs in that he has embraced not just the Han language and behaviours but even Chinese material culture in the form of popular music. Jennifer Taynen has suggested that this level of acculturation is common among *minkaohan* children, describing them as 'irresistibly drawn to Chinese movies, TV shows, comic books, and music' (2006: 48). In fact, as I have shown earlier, many *minkaohan* draw the line at Chinese popular culture, deliberately drawing on alternative sources of music, fashion, and art in order to state their difference. In this instance, Dilbär and her friend did not condemn the degree of the boy's acculturation; rather, they first established their own cultural superiority and then set out to convert him, or at least to persuade him that he would be converted once he had reached maturity.

For most *minkaomin* respondents, more disturbing even than a lack of proficiency with the Uyghur language or Uyghur social behaviours is a perceived lack of religiosity. This may take any form from strict adherence to orthodox practice (prayer five times a day) to the fatalistic assumption that everything that happens is directed from above. Gülhärä, a 20 year-old female *minkaomin* student from Ürümchi's Döngköwrük district, explained in 2004: 'We do mix with *minkaohan* at college. There are different kinds of *minkaohan*: some behave more like the Chinese, especially if they've

#### 372 chapter seven

always attended Han schools; some less so. Those with a very strong Han temperament, and who speak very little Uyghur, we don't talk to as much. They also tend to be the ones who don't pray.' Dilbär, an 18 year old female *minkaomin* student interviewed in the same year, reiterated this distinction: 'There are degrees, you know, with *minkaohan.* Some are alright … and some aren't. If a *minkaohan* student refers to Allah and looks to him for support, uses phrases like God has given [Uy. *Khuda bärdi]*, then we feel comfortable with them; if not, we tend to steer clear.' She then provided an example of her preferred type of *minkaohan*, as follows:

I have one *minkaohan* friend, who has a close Han friend, and she's converted the Han to Muslim food! The Han says that after eating in 'pure and true' [Ch. *qingzhen; halal*] restaurants for three years, she's become *qingzhen*  herself. This *minkaohan* is very religious, her faith is strong. But still she prefers to mix with other *minkaohan*; she gets on better with people with a *minkaohan* temperament. And she says she'll look for a *minkaohan* boyfriend. Some *minkaohan* pray and some don't. Some have never learned how, so they can't. Again, it all depends on what their parents teach them (or don't teach them). Family education is crucial.

Dilbär presented this particular friend as an acceptable *minkaohan*, on the grounds that she had been responsibly educated in the family home and had consequently developed a strong attachment to Islam. This reflects other studies, such as that of Vasquez and Wetzel, which highlight the family as 'the critical venue to transmit values and cultural knowledge' in marginalised cultural communities (2009: 1566). Dilbär's friend was apparently so confident of her Muslim identity that she had even succeeded in effecting the dietary conversion of a Han Chinese friend (conversion normally being the sole condition under which a Uyghur-Han mixed marriage may take place). Yet, when it came to friendship and marriage, even this 'acceptable' *minkaohan* felt more comfortable interacting socially with other *minkaohan* than with either *minkaomin* or Han friends. This again underlines this sub-group's experience of liminality. Dilbär's 18 year-old male *minkaomin* classmate, Ömär, elaborated further on the relationship between religious faith and social trust:

Those who pray don't really mix with those who don't. The first have faith in common, they have nothing to talk about with people who don't pray. And they don't trust them readily. Introducing someone in terms of whether they pray is not a new thing. We have always done this … it just shows that a person is a good person, a better person (Interview, 2004).

In this way, the decisive factor governing intra-group social relations in 2004 seemed to be not so much whether a person's first language was Chinese or Uyghur but whether or not they prayed or demonstrated some other sign of religiosity. Religious faith was then automatically linked to high personal morality and a good conscience (*wijdan*).

*Minkaohan* could also be deemed acceptable by *minkaomin* peers if they voiced dislike or mistrust of Han Chinese, based on their everyday interactions with, and deeper familiarity with, Han classmates or colleagues. Jelil, an observant *minkaomin* graduate in his late twenties, explained in 2002:

The *minkaohan* are getting better! Now they understand the Hans much better than they did before. But some go the other way and become completely sinicised [Ch. *hanhua le*]. Your friend [Abdukerim] was right to curse those two he met in the fast food joint … I would have had a go at them too! […] Five criteria are needed for a nation to exist: language; territory; religion; distinct customs and habits; distinct facial characteristics. Of these, language is the most important.

He thus linked Uyghur language proficiency, religiosity and social mores directly to the nationalist cause: only by protecting and preserving these unique cultural attributes—and by recognising the Han goal of destroying them—can the Uyghur nation be saved. By implication, those *minkaohan* who fail to make efforts to nurture their Uyghur language proficiency, their religious faith and their knowledge of Uyghur social practices are 'disloyal' to the nation. As Märyäm, a divorcee and music trader in her thirties from Ghulja, put it in 2002: 'How can Uyghurs unite when some like the Han? If only we could unite [in dislike of the Han], we could kick the Hans off this land!'

Being fluent in the Chinese language, *minkaohan* enjoyed better employment prospects than *minkaomin* in the 1990s (although this situation had begun to change by 2002, see Chapter 1 Inequalities). By extension, popular mistrust of *minkaohan* was extended also to employees of state work units or private Han-run companies *[danwei-lar*7*; khizmätchilär]*. Qurban was an 18 year-old *minkaomin* youth, who was teaching himself English at an evening class in Ghulja in 2002. He described three 'types' of *khizmätchilär*:

You would find it really revealing to speak to the *khizmätchilär*. You never know with them. Some are 'Yes-men' [Uy. *maqulchi*]; if you say anything to

<sup>7</sup> This hybrid word is formed using the Chinese term for 'work unit' (*danwei*) plus the Uyghur plural pronoun suffix *-lar*.

#### 374 chapter seven

them about the Chinese, they will tell them. They will side with the Chinese! Others have low self-esteem and think they should be grateful to the Hans. They have little awareness of the injustices of Uyghur social exclusion, or at least they rarely voice it. The third group is loyal to the ethnic group, but they say nothing out loud for fear of losing their jobs and livelihoods.

In this way, Qurban categorised Uyghurs in work unit or company posts as either social climbers (willing to compromise all in order to gain further privileges), victims with an inferiority complex vis-à-vis the Han Chinese, or loyal pragmatists. The first category recalls singer-songwriter Ömärjan Alim's castigation of indigenous officials in the 1995 song 'Äwliya dostum' ('My sly friend'). The second evokes the 'internalised oppression' characterising first-generation *minkaohan* as well as some poorly educated *minkaomin* working in the service industries [cf. Zunun; Anarkhan]. The third describes perhaps the majority of *danwei-lar* with whom I have come into contact, at least in the regional capital where ethnicity is more politicised. *Danwei-lar* must conform linguistically (communicate in Chinese) and culturally (abstain from prayer; remove head coverings) on a daily basis purely in order to protect their job security; they have chosen to hold their tongues and work from within the system. A similar trend was identified in the former Soviet Central Asian states, where 'the difference between the way of life of those who have any access to privilege […] and those who have none is sufficiently great to give "affirmative action" [representation of minority nationalities in the Communist Party, government, professions, and skilled trades] attractions despite its flaws' (Atkin 1992: 53-4). As Atkin put it, some Tajiks and other minorities preferred to obtain feasible gains within the system rather than hold out for the unattainable. However, the appearance of collaboration means that 'workplace pragmatists' may find themselves kept at arm's length by observant peers, most of whom prefer the (less stable) self-employed business route. Most suspect of all are Uyghurs occupying official government posts. The scale of their transgressions is reflected in the gravity of the accusations levelled against them, as when Shökhrät, an intellectual in his thirties, asserted in 2004 that such persons ignored Islamic dietary prescriptions: 'Some Uyghur cadres eat pork now! If they eat pork this pleases their Han superiors, and they are more likely to get promoted.'

To sum up, the prevailing view at the current time is that the first responsibility for linguistic, socio-cultural and religious education of *minkaohan* children lies with the Uyghur family. As Ghäyrät, migrant worker in his thirties from Aqsu, explained in 2004:

## hybrids: identity negotiations among the urban youth 375

*Minkaohan* don't necessarily have to 'go bad' [Uy. *buzulup ketish;* become overly acculturated]; they can be really strong, stronger than us *minkaomin*. They can be really bright, retain a strong ethnic spirit, yet know when to speak fluent Chinese and play the game. That kind of Uyghur is the most effective Uyghur. The Chinese would be happy to see us retreat into the Uyghur language and have us fall further and further behind. We can't do that, we have to take them on. It's all a question of family education [Uy. *ailä tärbiyisi*]. Some people don't educate their children properly, and so their children end up going bad; they are totally sinicised, have no sense of ethnic loyalty, no ethnic spirit. Such people embarrass me, I look down on them.

In this way, it is considered incumbent on Uyghur parents to establish a sense of pride and belonging in their sons and daughters during early childhood in the home. Beyond that, *minkaohan* alone are responsible for the way in which they are received by the Uyghur community, and this depends on: 1) the effort they spend developing and maintaining Uyghur language and socio-cultural skills; 2) their depth of religious faith; and 3) their display of ethnic loyalty.

# 7.5. *The Han Wannabe: Prototype for the Third-generation* Minkaohan*?*

As outlined in Chapter 1, from 1995 onwards the Chinese government began to test its future plan for an exclusively Chinese-medium education system in Xinjiang. Since 2002, this 'bilingual education' strategy (*shuangyu jiaoyu*—a complete misnomer) is being rapidly rolled out at all levels from higher education institute down to kindergarten. This means that although many school-age children experienced a combination of Uyghur-medium and Chinese-medium education prior to 2002, a steadily increasing proportion will now be taught solely through the medium of Chinese. What effect this early exposure to the Chinese-language curriculum will have on the identity formation of children of kindergarten age remains to be seen. Can Uyghur families continue to foster the important foundation in Uyghur linguistic and socio-cultural skills in the home, or will these efforts be compromised? This is difficult to predict, although we can try to envision a rough 'prototype' of the third-generation *minkaohan*—those educated exclusively at Chinese-medium schools—by looking at the self-identification, social behaviour and consumption patterns of existing respondents (second-generation *minkaohan*) whose parents elected for them to receive an all-Chinese education.

#### 376 chapter seven

Gülshäm, twenty, a Uyghur woman brought up in Ürümchi by *minkaohan* parents, had attended Chinese-medium schools throughout her life. She asserted in 2004 that she felt much more comfortable speaking in Chinese, and was proud of her teacher's pronouncement that her Chinese was 'much better than her Uyghur'. She did not seem to experience a deep sense of linguistic or cultural 'lack', as had many first-generation *minkaohan*. She admitted that although she had many Uyghur friends with whom she spoke in Uyghur, she got lost whenever they began to talk in any depth. Yet this reality did not seem to trouble her: 'At those times, I feel regret … because I'm a Uyghur! But it's not so bad.' Although she was aware of the social rule that Uyghur young people should not speak in Chinese to their elders 'out of respect' (and informed me of it), she consistently addressed her *minkaohan* father in Chinese and it emerged that the two spoke Chinese almost exclusively in the home. When asked which language she *liked* to speak, she replied without hesitation that she preferred Chinese, characterising Mandarin as an international language:

Chinese is the way to go if you want to advance and develop … if you want to achieve internationalisation, be an international person. And it's crucial when you're looking for a job. No-one will look twice at you if you don't speak Chinese fluently.

While the latter assertion was expressed by all respondents in Ürümchi, I had never heard a Uyghur respondent express the notion of Chinese being the 'road to internationalisation'. Rather, most respondents identified English as the 'world language', asserted that Uyghur is closer to European languages than to Chinese (because of the large number of foreign loans among other reasons), and observed that Uyghurs learned English more rapidly than Chinese. Gülshäm, however, was unconvinced by these arguments and openly displeased when I ventured that English (rather than Chinese) is currently the 'international language'. Asked what 'time' she used—Beijing time or local Xinjiang time—she responded: 'Of course I use Beijing time! And so do all my friends. Only old people use Xinjiang time now.' Absent here was the heavy symbolism surrounding the time question that had characterised the mid-nineties and that prevails today among *minkaomin*, old and young. For Gülshäm, Beijing time symbolised modernity and progress: 'You see, it's the time we use to interact in society … the clocks at school read Beijing time.'

Significantly, Gülshäm was sometimes openly critical of the Uyghur in-group. For instance, she avoided going to the Uyghur-dominated Döng-

köwrük district in south-west Ürümchi in 2004 because, as she put it, 'there are lots of bad people there, and it's chaotic' [Ch. *luan*]. Such statements are more usually found on the lips of Han residents and likely reflected majority representations and stereotypes, unconsciously absorbed through everyday contact with Han peers in a Chinese-language environment. However, the most remarkable example of Gülshäm's antipathy towards the 'Uyghur authentic' was her reluctance to contemplate a romantic partner from the in-group. During a month of interviews, it emerged that she yearned for a Han boyfriend. Indeed, she displayed a complete disregard for endogamous norms. This state of mind, associated with minority groups in an advanced stage of acculturation, occurs in some individuals, who may increasingly identify with the majority group, accept their values, and come to see intermarriage as a strategy towards integration (Cerroni-Long 1984: 37). Gülshäm's preference for (potential) Han suitors was partly derived from a strong sense of familiarity with Han classmates and neighbours. She explained in 2004: 'I would prefer a Han boyfriend... I just feel more at home with them. I suppose it's because everyone in my street is Han, and I've always gone to Han schools since I was small … so all my friends are Han.' Gülshäm revealed that while she was too young (at twenty) to have a boyfriend, some of her friends were already dating Han Chinese. She shrugged off the suggestion that Islamic dietary prescriptions might create problems for such unions, replying nonchalantly: 'At the end of the day, they [Han boyfriends] go home and don't eat pork in front of us.' Her attitude—highly unusual among young Uyghurs—evokes studies connecting the secularising effects of rapid modernisation and globalisation to changes in marriage attitudes (for example, see Shibata 1998 on Guyana). Such studies show that where young people have not been raised to adhere strictly to religious practices, the potential for ethnic prejudice is muted, and they are increasingly free to base opinions of potential friends or suitors on observations made of peers—in-group and out-group—at school or work. That said, Gülshäm's comment about Han boyfriends not eating pork in front of their Uyghur girlfriends highlights an awareness of the important distinction between possibilities for courtship and marriage.

According to Dilbär, eighteen, a female *minkaomin* high school student speaking in 2004, some *minkaohan* girls who mixed with Han classmates at school and perhaps also grew up in Han neighbourhoods, may find direct communication with Han men easy and familiar while mistrusting the less direct approach of Uyghur males:

#### 378 chapter seven

They think that Uyghur guys will mess them around, say one thing while thinking something else. I understand the *minkaohan* point about Han lads being gentler; I like that about them too. But Uyghur lads do hold your hand, it's just that they make light of it … they may not say romantic things, but they think them and feel them *here* [presses two hands against breastbone]. But the *minkaohan* like straight talking. The Hans are much more direct, whereas the Uyghurs go round the houses to say things, it's all implied.

In Tibetan culture, we find a similar male reluctance to express love in verbal and tactile ways; there, a truly masculine man would never show affection to his wife, but should only love her in his heart (Hillman and Henfry 2006: 264).

Gülshäm stated in 2004 that she found Han faces more physically attractive, although she giggled and placed her hand across her mouth as she said it, aware of breaching a social taboo. This preference had apparently given her the idea of relocating to the Han-dominated city of Shihezi.8 For Räbigül, her *minkaomin* friend, on the other hand, Uyghur men were self-evidently more handsome than Han men. Their different opinions probably reflected their different status as *minkaohan* and *minkaomin* respectively; none of my *minkaomin* respondents agreed that Han boys were aesthetically pleasing. Both women signalled the casual fashion sense of Han men as a 'pull' factor, although they admitted that they found young Uyghur men who don casual clothes similarly attractive. Essentially, their preference for this particular aesthetic seemed to reflect a desire for the 'modern'. Although neither respondent articulated that notion explicitly, it emerged clearly in the observations of Dilbär, eighteen, a *minkaomin* high school student interviewed in 2004:

I know how the *minkaohan* think … I hear them talk about it a lot. They like those [Han] guys who dress in leisure gear (Ch. *xiuxianfu*); they attach importance to the way guys dress. Uyghurs will usually dress up as if they're going to a party. The guys will wear Western-style suits, shirts and dress shoes. The *minkaohan* like guys who dress down. The other night, Gülshäm told me how she didn't like guys who 'dressed like her grandfather'!

This desire for the modern reflects in the consumption patterns of some young Uyghur women, leading them to purchase clothing or other material items from Chinese merchants and thereby provoke community disapproval for 'going Chinese' (Bellér-Hann 2001a: 57).

<sup>8</sup> The proportion of Han Chinese within the local population of 'New Town' Shihezi in 1998 was 95 per cent (Toops 2004: 257).

#### hybrids: identity negotiations among the urban youth 379

It is important to remember that the sketch I have developed above represents a very rare example of a deeply acculturated *minkaohan*: a fluent Chinese speaker, raised by *minkaohan* parents in a Han-dominated district of Ürümchi; an individual desperate to take part in modernity (as bestowed by the Han), and fully acquiescent with the Chinese state's 'civilising mission'. There are currently many Uyghur intellectuals and professionals who fear that this may be the shape of the Uyghur future. Already in 2002, a male intellectual in his thirties, Shökhrät, had begun to sound alarm bells:

The biggest threat to Uyghur culture now is assimilation, especially in Ürümchi. Most Uyghurs in this city are no longer speaking pure Uyghur—some don't know the language, others are just too lazy. Since 1996, it has got worse year by year, with more and more people switching between languages. This is the result of the imposition of the Han language.

The same sense of hopelessness coloured a conversation I had with three work unit employees from Ghulja in the same year. They told me: 'Our mother tongue is going to disappear. There's nothing to be done about it [Uy. *Amal yoq]*. Our children have to go to Han schools; otherwise they will have no future, no social position. And so they forget their own language; even if we make sure to speak only Uyghur in the home.'

Certainly, increasing numbers of Ürümchi dwellers were code-switching between Uyghur and Chinese by the early 2000s—and not just *minkaohan*.9 During a social visit to the home of a *minkaomin* male in his thirties, I observed that he spoke continually in Chinese for the duration. Initially thinking that he did so for my benefit, I realised after a time that he had simply developed the habit of speaking Chinese. He is the sole Uyghur employee in an industrial work unit in the Ürümchi suburbs, and speaks Chinese all day to his colleagues in an overwhelmingly Han-dominated workforce. When I spoke to him in Uyghur, he replied in Chinese; he even chatted in Chinese with Zunun, the Uyghur friend who had introduced me into the man's home. Moreover, he seemed to assume Han mannerisms at the dinner table, barking 'Eat! Drink!' [Ch. *Chi! He!*] in contrast to the gentle Uyghur-language 'Please take some' [Uy. *Eling*…]. Zunun apparently felt the need to scold his friend several times for not speaking in Uyghur. His wife, for her part, spoke mostly in Uyghur, but she too sometimes reverted

<sup>9</sup> Code-switching can even occur within the same word, so that neologisms are created—words 'in-between' the two languages. For instance: *minkaohan-*lar [*minkaohan* individuals], *yue*-liq qilish [to arrange a meeting], *danwei-*lar [employees of state work units].

#### 380 chapter seven

to Chinese, apparently when the latter language came to her more quickly. This trend has been observed also by Bovingdon, who cites informants as chiding one another for speaking in Chinese, with the comment: 'Infidel! Don't go turning into a Han!' [Uy. *Kapir! Khänzu bop kätmä!*] (2010: 89). Yet the question remains: do these developments *necessarily* mean that the Uyghur language, culture and, indeed, the entire race are going to die out? Do we assume, following the 'purists' who resist any modification of 'tradition', that acculturation or hybridisation is necessarily a bad thing?

# 7.6. *Cosmopolitan Multilingualism: Uyghur Youth as 'World Citizens'*

Experience in a globalising world suggests otherwise. There are plenty of contemporary examples of positive incarnations of ethnic and cultural ambiguity. Stuart Hall notes a shift away from identities constructed *in response* to the ethnic/racial classifications imposed by majority groups and institutions. For him, the hybridisation of different cultural forces in a world of migrated peoples must produce a new era in which everyone will live with impurity, with the fact that 'there are no pure origins left' (Hall 1996: 134). Thus, by the end of the 20th century, previously 'reactive' Afro-Caribbean identities in Britain had evolved into a 'double or multiple consciousness', which allowed situational deployment of dual or multiple identities. In China, '*minzu* [nationality] building' has similarly not been simply a state project, but one embraced in diverse ways by the nationalities themselves, or at least fractions thereof (Bulag 2002: 10). A young generation of urban Uyghurs, if not exactly *embracing* the system, is at least learning how to deal with it. Unlike the first-generation *minkaohan*, who were raised under the repressive conditions of the Cultural Revolution, second-generation *minkaohan* grew up in reform China amid more conciliatory minority policies. Unlike their predecessors, their attendance at Chinese-medium schools represented—at least prior to 2002—a strategic choice made by urban parents to increase their children's life chances. Uyghur families were thus able to exercise a degree of social agency. As a result, fewer of these young people manifest the deeply embedded inferiority complex ('internalised oppression') which characterised the previous generation.

Having originally intended to present separate sections on the *minkaomin* and *minkaohan* as 'world citizen', I combine them below, after acknowledging that their respective features are more alike than dissimilar. In order

to shape him- or her-self as a 'citizen of the world', a young, urban Uyghur typically strives to be—in no particular order—multi-lingual, a good Muslim, loyal to the Uyghur ethnic group, well versed in Uyghur social and cultural practices, politically 'savvy' (aware), and (selectively) inter-cultural, that is, having a global outlook, but one which eschews Han culture and transcends Chinese national boundaries. To begin with language proficiency: it has been correctly noted that Uyghur youth are 'more eager' to learn Chinese than the older generation, and that a young generation of bilingual Uyghurs is now emerging in Ürümchi (Yee 2003: 446). However, it must be said that the desire to learn Chinese is more often linked to pragmatic aims such as the wish to access higher education or to get a white-collar job (see Chapter 1), and not to the joy of language learning *per se*. Many Uyghurs dislike the study of Chinese, which they perceive as merely a means to an end, and their dislike has been heightened by the post-2002 policy that makes Chinese-only tuition mandatory. At the same time, the majority of Uyghur households will endeavour to ensure that their sons and daughters grow up fluent in Chinese *and* the mother tongue, regardless of which medium they are taught in at school. Take Nurgül, a female *minkaohan* student, as an example. While she was educated entirely at Chinese-medium schools, she informed me in 2004 that she could read the Uyghur (Arabic-based) script, and that she had learned it through selfstudy. Meanwhile, she maintained strict divisions between language domains, speaking Chinese at school, but only Uyghur at home. This practice had been encouraged by her parents in order that she would not 'forget' her Uyghur; the effect was to produce a fully bilingual, modern young woman.

To suggest that young Uyghurs are doomed simply to absorb Han culture in a one-way process of 'sinicisation' would be grossly over-simplistic. In fact, they experience, negotiate and appropriate a variety of cultures, including Uyghur local culture, Han national culture, and a myriad of global cultural flows (Central Asian; Russian; Turkish; Indian; Arab; American; European) in a process which inescapably broadens their horizons. In 2003, one of the most popular acts with urban youngsters in Xinjiang was Dervishes, a band composed of four mainly Russian-speaking Uyghur musicians trained in the Almaty Conservatory in Kazakhstan, who produced a heavy rock sound infused with samples from Uyghur folk, Brazilian birembau and London breakbeat (Harris 2005: 631). Thus, in Xinjiang, as elsewhere, reflexivity is a central tendency among contemporary subjects under the conditions of global modernity (Kim 2005: 448). Two concepts within the broad theory of globalisation seem especially appropriate to the Uyghur context: firstly, the notion of 'accelerating interconnectedness' (the intensification of global social relations and consciousness of world society); and, secondly, the notion of 'action at a distance' (when actions of social agents in one part of the world come to have significant consequences for distant others) (Kim 2005: 446). In the former case, we see young Uyghurs informing themselves of events and trends around the world via the mass media, including television, film, domestic and international news agencies, and the internet. In the latter, we see Uyghurs identifying with fellow 'victims' around the world (Palestine, Chechnya, Iraq, Afghanistan) and strengthening their sense of a borderless Islamic solidarity (cf. Smith Finley 2007b). Unfortunately, we also find them *being identifiedwith* radical Islamic extremists such as the Al-Qaeda activist network, with grave consequences for minority representation and regional policy.

Young, urban people strive to understand and respect Uyghur social mores, including customs and traditions, modes of social interaction, and so on, and to integrate these into their daily behaviours. The pressure to conform to these expectations issues initially from Uyghur parents and/or grandparents and is later reinforced and reproduced by peers (see Chapter 6). Such pressures are partly related to the strong reaction among firstgeneration *minkaohan* embodied in the phrase 'Don't let them end up like us' [Uy. *Bizdäk bolmisun]*. Take the example of Räwiä: because she had felt uncomfortable with traditional Uyghur performing arts throughout her own life, she purposely encouraged her small daughter to learn Uyghur dance from a young age. In 1995-6, she sent her regularly to her grandmother—a retired teacher of Uyghur language—for coaching in the mother tongue (which Räwiä herself was unable to provide). The grandmother's house was a textbook example of the perfect environment in which to build Uyghur cultural awareness, its décor of carpets, mirrors, crochet and lace typically Central Asian. Räwiä's efforts paid dividends. By 2004, her daughter had grown into a young *minkaomin* woman with perfect command of the Uyghur language and excellent Chinese. She prayed each day at home, and disapproved of the inter-ethnic courtships she observed at her school. In pragmatic terms, the two generations were fully aware of their ability to share their respective attributes, so that the *minkaomin* daughter taught her parents new Uyghur vocabulary, while the *minkaohan* parents coached their daughter in Chinese.

Interviews with young, urban Uyghurs reveal an increasingly internationalist outlook and an intellectual preparedness to engage with the

Figure 14: Carpet dye sellers, Qäshqär (1996)

'Other' reminiscent of that identified elsewhere in East Asia, for example, among young, middle-class, educated women in South Korea (Kim 2005: 456). Yet even as young urban Uyghurs transition to a multi-lingual and multi-cultural status, the vast majority remain deeply proud of their Uyghur national origins. A survey by Herbert Yee (2003) revealed that 90.6 per cent of urban Uyghur respondents said they were proud to be Uyghur, compared with only two thirds of Han respondents who felt proud to be Han. Uyghur respondents were apparently equally proud of being a Chinese citizen, although the survey did not require them to rank the two

#### 384 chapter seven

levels of identity, and some may have supplied positive answers owing to perceived political sensitivity (Yee 2003: 437). Some urban young people do appear to take pride in the Chinese nation to a degree, encouraged by its growing importance within the international community, and seeing that it is possible (if difficult) to succeed on a national scale. In the early 2000s, a number of celebrity youth figures provided positive role models for the Uyghur youth. One such figure was Adil, the tightrope walker (five times a Guinness World Record Breaker), whom Dilbär, an 18 year old *minkaomin* student, described in 2004 as 'our pride'. When asked whether she meant 'China's pride' or 'the Uyghurs' pride', she responded confidently: 'It means both at the same time. He is a star on a national scale and at the level of nationality. In fact, he's famous all over the world, isn't he?' She thus saw him as inspirational at local, national and global levels.10 The implications are that young urban Uyghurs have become more adept at negotiating and exploiting their local (Xinjiang) and national (PRC) identities. As Räwiä, a first-generation *minkaohan*, put it in 2004:

This next generation will far super-cede our generation. They will be much more effective than we were. *Minkaohan* don't have to be quasi-Hans; they can be both things at once. And, in that way, they are better able to challenge the Hans.

By the early 2000s, Chinese language proficiency—as a necessary form of cultural capital—had been overtaken by a perceived need to learn English (a trend also observed across China proper).11 Jelil, an observant graduate in his twenties, confirmed in 2002 that most schools in the regional capital now offered English classes, although this opportunity was not widely available in other parts of Xinjiang. He noted however that the Ürümchi schools lacked native English teachers, unlike those in Beijing. Asked why they wished to learn English, some Uyghur respondents stated—like most Han young people—that knowledge of English would help them 'get an even better job'. This recalls Kim's finding that transnational culture in contemporary South Korea is tied to the job market; and that English, as a 'language of global modernity' represents a means of making a living and attaining

<sup>10</sup> A book has been published in Chinese documenting Adil's exploits, titled 阿地力: 高空王子 [*Adili: Hero of the Heights*].

<sup>11</sup> Ben-Adam (1999: 196) reported that Uyghurs in Ürümchi were studying Russian in the hope of taking advantage of ties with the Central Asian republics, while Hans put more emphasis on English. This was no longer the case in 2002. Although Russian and Japanese remained popular foreign languages for Uyghur students, English was considered by far the most important target of second language acquisition.

economic independence (2005: 455). However, significantly more respondents linked English language proficiency to the associated opportunities to study abroad. For Anarkhan, a first-generation *minkaohan* who worked as a chambermaid in an Ürümchi hotel in 2002, this represented a chance to experience personal freedoms not enjoyed in Xinjiang. She told me: 'I'm doing my best to educate my two kids. I want them to go abroad to study [Uy. *chät'älgä barsun däymän!*] Because it's more free over there. Our education system is so inflexible, with teachers always telling kids what to do. I want them to have fun, see the world.' Also speaking in 2002, Polat, a firstgeneration *minkaohan* in his forties from Ghulja, stressed the potential of study abroad to build awareness and confidence in young Uyghurs:

Education here is no good. Now, lots of parents want to send their kids abroad to study. You can be as educated as you like, but it won't do you any good if you've been stuck in the immediate narrow environment all the time. Uyghurs need to get out, have a look around, broaden their minds. Then they can really start to build their self-confidence! Uyghurs have no self-esteem. That's the biggest problem to my mind. Not just money; and not just education.

Jelil, an observant *minkaomin* graduate speaking in the same year, also underlined the importance of English, although, for him, the goal of 'getting as many Uyghurs abroad to study as possible' evidently formed part of a bigger national cause. He explained that, in order to further that cause, some Uyghur 'big businessmen' (*chong sodigärlär*) are now willing to fund study abroad for friends, relatives and acquaintances, or for their sons and daughters.12 Significantly, all respondents assumed the return of these individuals to Xinjiang following completion of study. Study abroad was thus not considered a permanent 'exit strategy', but a way of increasing young Uyghurs' self-confidence and political awareness. Return to the motherland was treated as a given since, as one high-profile Uyghur musician remarked in 2002, 'If they don't come back, what hope will be left in Xinjiang? What will remain?'

Linked to the desire to learn English (and by extension to access global internet communications technology) in the early 2000s was a scholarly debate surrounding Uyghur scripts. Shökhrät, an intellectual in his thirties from Qäshqär, explained in 1996 and again in 2002 that the Arabic script

<sup>12</sup> One of my respondents was able to study in Hungary for a year thanks to the generous patronage of an older friend, a successful international trader whom he affectionately called '*Akam*' (my big brother). This trader planned in the long term to open a school or private university in Ürümchi, to support the training and development of young Uyghurs.

#### 386 chapter seven

represents a hindrance and a barrier when using computer programmes, since one Arabic letter uses two bytes of memory while one Latin letter uses only one byte. He noted too that computers find it hard to cope with the formatting of right-to-left script, and repeatedly switch back to left-toright format. Such flaws waste time, and cause problems where documents include more than one type of script or format, especially scientific or technical documents. As a result, some younger members of the XUAR Language and Script Committee (*Yuyan wenzi weiyuanhui*) had advocated returning to the Latin script (*yengi yeziq*; New Script), or using a dual script system whereby the Latin script was taught alongside the Arabic script (*kona yeziq*; Old Script) at primary schools.13 Older members, however, feared that further script changes would cause yet more cultural discontinuity, and that Uyghur education and literacy levels might be set back. Still others were concerned that a new system would bring insupportable financial costs. In 2004, I asked Dilbär, an 18 year old *minkaomin* student, her views on this debate. She commented that the New Script would indeed be helpful for learning English, and that it came in useful when conversing in online chat-rooms such as QQ (where knowledge of Chinese pinyin is essential); but confessed that she ultimately preferred the Old Script, even with its technical disadvantages.

Young urban Uyghurs consider it particularly important to have a strong sense of religiosity. However, this can be expressed in a variety of ways and need not require strict adherence to the five daily prayers. Across the Central Asia region, a syncretic style of religion has long been a feature of popular religious practices, and ideological coexistence (Islam and Communism) has frequently been achieved (Faure, cited in Bush 1970: 281; Carrere d'Encausse 1981: 280; Voll 1987: 138). In the Chinese context, the pragmatic approach of Islam allowed many Hui Muslims to become prominent within the state system, and these persons may have been able to intervene on behalf of the masses (Israeli 1984: 300). The key to living in a multi-cultural and religiously diverse society as a Muslim minority is to 'discover ways to resolve the duality, the twoness that resides in the minority condition' (Abedin, cited in Voll 1987: 133). For many young urban Uyghurs, it is simply not possible to perform prayers throughout the day owing to the ongoing prohibition on mosque attendance for employees of state work units, individuals in full time education, and minors below the

<sup>13</sup> The Latin alphabet was adopted in post-independence Kazakhstan, where it has become 'a symbol of national independence' for many Kazakhs (Haghayeghi 1994: 192).

age of eighteen (Fuller and Lipman 2004; Mudie 2006). It is however increasingly common for a school pupil to slip out to the mosque during the Friday lunch break or perform his prayers at home in the evening. The most important thing is to believe, and be seen to believe. As Ömär, an 18 year old *minkaomin* student, elaborated in 2004:

We don't know when the Day of Judgement will come, but we know it will come one day … it could be tomorrow! We tend to think that *those who pray are better than those who don't*. The first will certainly go to Heaven, but we think the latter will almost certainly go to Hell. I think everyone believes this' [my emphasis].

Beyond the prayer ritual itself, Islam is also 'embedded in many […] daily activities, such as greetings and exclamations, mode of dress, scrupulous personal cleanliness (including the consumption of *halal* food), and a sense of solidarity with other Muslims in the world' (Dwyer 2005: 3). This last phenomenon sees contemporary urban Uyghurs identifying with Central Asia, Turkey, and the Arab world (cf. Harris 2005: 633 on belly-dancing in Ürümchi). 18 year old *minkaomin* students Ömär and Dilbär related the following interesting example of pan-Turkic identification in 2004, a story centring on international football:

During the last World Cup, it happened one time that China had to play Turkey. We were watching it in the big hall at school and, of course, China lost! It was really funny; every time Turkey began to do well, we cheered them, and our Han classmates looked askance at us and got really irritated!

As we chatted, the two sat eating Turkish biscuits and chocolate, now widely available across the city, thus expressing their cultural, religious and political (Pan-Turkic) affinities through the consumption of international sport and *halal* foodstuffs.14 Another study mentions a similar scenario which played out during the same World Cup match, involving Uyghur spectators' glee at the vision of the blue flag of Uyghur independence, which someone had managed to drape behind the Chinese goal (Boving-

<sup>14</sup> Chris Hann noted the frustration of many young Uyghurs at the limited opportunities for competition at the Ethnic Games held in Xinjiang in 1986. Han did not take part at all, while the minority nationalities were not allowed to compete on a group basis lest this intensify inter-ethnic rivalries. Instead, regions were asked to compete with regions. Uyghur participants were greatly disappointed, stating that they would welcome the chance to form a pure Uyghur team, take on the Han, and prove their worth (superiority?) by defeating them. It is instructive that these same respondents displayed great delight in the fact that the top table tennis player in the region at the time was a Uyghur, who had beaten the Han at their own traditional game (Hann 1991: 233-234).

#### 388 chapter seven

don 2010: 102). Such incidents fully demonstrate the importance of the notion of 'alternate centres' (Becquelin 2004a: 377) in stressing an alignment away from Beijing and towards Turkey (among others).

Arfiya Eri, an American Uyghur working in the neighbourhood of Yan'anlu in Ürümchi in summer 2005, describes how this Uyghur-dominated area 'was becoming a cultural hotspot influenced primarily by the Turkic West rather than the Han Chinese East' (2008: 77). Eri's Uyghur students listened not to Mandopop (Mandarin-language pop music performed by Han artists), but to music from Uzbekistan, Kazakhstan, Russia and Turkey. Featuring Uyghur supermarkets packed with halal foodstuffs, newly built apartments inhabited by the emerging Uyghur middle class and Central Asian expatriates, and institutes of education with a high ratio of Uyghur students and academics (Eri 2008: 77-78), the Yan'an area has begun to emerge as a locus for a 'subaltern cosmopolitanism', an alternative modernity to that proposed by the Chinese state. As such, it has become a magnet for the city's Uyghurs, who have reputedly moved there in droves following the Ürümchi riots of 2009, in tandem with a sharp exodus of Han residents from Uyghur-dominated areas.15 A multi-lingual environment, in which Uyghur, Chinese and Russian are represented in both the oral buzz and the local signage, it reflects the 'current social expectation for young Uighurs to be fluent in Mandarin and also proud to be Uighur' (Eri 2008: 78). This example demonstrates that popularisation of the Chinese language among the youth need not lead to negative incarnations of acculturation or 'sinicisation' but may instead evolve into a bold contemporary pride in hybridity. Indeed, the Yan-an district might be perceived as a microcosm of a broader 'spurning' of the notion that Uyghur interests are 'indissolubly bound with those of China' (Bovingdon 2010: 84).

Nurgül, an 18 year-old bilingual student I met in 2004, provided an interesting example of a young urban person. A *minkaohan*, she was a selfproclaimed 'punk', who wore a skull bracelet, and listed American popular singers Eminem and Pink among her music tastes. As such, she reflected the way in which cultural referents at the global level, such as hip hop culture and associated discourses of equality, may be appropriated as 'tools for self-definition' at the local level, and across national cases (Lamont and

<sup>15</sup> Personal communication, Dr Ablimit Baki, University of Manchester, May 2012. In two new interviews conducted in 2013, I learned from a long-term Uyghur respondent that despite the initial rearrangement of residence patterns directly following the event, everyday life in the city has now begun to normalise. The city 'once again feels as it did prior to the riots, that is, peaceful on the surface, but tense underneath.'

Mizrachi 2012: 369). Nurgül confessed in 2004 that she could not dance *usul* [Uyghur traditional dance] well; yet at the same time she strongly doubted that it would die out:

Uyghurs have it in their blood somehow; they hear the music and just have to dance. It's not necessarily our parents who pass [the tradition] on. Once the music starts, we just have to move to it in that way. And we see it on the TV from an early age.

I later watched as she and her school friends—including both *minkaohan*  and *minkaomin*—danced a modern adaptation of *usul* at a birthday disco party, whereby the dancers moved around in constant circles, boys following boys, and girls following girls. This form of *usul* seemed to be far less about courtship than about good fun. Also of note was the fact that this 18th birthday party was held at 3pm in the afternoon because, as Nurgül explained, 'we are still young.'

One feature of urban youth identity is (arguably) more observable among *minkaohan* than *minkaomin*, and that is flexibility. Tursungül, a 100 per cent *minkaohan* and graduate of the first *Xinjiangban* cohort in China proper, considered herself well placed to analyse the differences between the two sub-groups (although, as an 'insider', she was clearly at risk of bias). In her view, *minkaohan* have a greater capacity for intercultural communication, in particular, a higher tolerance threshold towards Han culture. She argued in 2011:

Though *minkaohan* and *minkaomin* are from the same culture, we really differ a lot. *Minkaomin* are slightly inflexible, [culturally] exclusive and isolated. As *minkaomin* are generated from the more pure Uyghur culture, their perceptions towards cultural conflict [difference] are more direct and stronger. *Minkaohan* can handle this better because they are in the middle of the two cultures. Therefore, *minkaomin* believe that *minkaohan* are weak and tend towards compromise […] *minkaohan* think of *minkaomin* as [un?] sophisticated and not easy to control their temper [quick-tempered].

In other words, while young *minkaohan* are open to—or at least able to negotiate—all cultures, including the Han culture, young *minkaomin* are more open to some cultures (Central Asian; Turkish; Arab; Western) than others (Han). This is a proposition partly supported by some of the data in this study, for example, the findings on inter-ethnic courtship.

Above, I have described some of the positive features of an urban youth identity in contemporary Xinjiang, an identity which has (so far) remained possible even in a climate of Chinese-medium education and continuing Han in-migration. There are, however, other impacts of rapid urbanisation

#### 390 chapter seven

and modernisation that sit less well with the middle and older generations, many involving the changed behaviour of young urban women. They include the flouting of curfews by daughters; the growing reluctance of urban women to bear children; female consumption of alcohol; female participation in hostess work (I deal with this separately elsewhere, see Smith Finley, forthcoming); the waning of the hospitality ritual; and the increase in social atomisation and decline of community life. In 2004, some respondents lamented that, increasingly, young women are to be found outside the home after dark. One popular venue for this transgression was the internet bar. According to my observations during that year, most Uyghurs who entered internet bars did so in pairs or groups of boys, or as a boy-girl couple. Occasionally, however, pairs or groups of modern-looking teenage girls dressed in casual Western clothing could also be found in these venues. Unaccompanied (i.e. un-chaperoned by male friends or relatives), they sometimes paid visits at times as late as 8.30pm Xinjiang time (10.30pm Beijing time). Since they spoke to one another in Uyghur, it may be assumed that these girls were *minkaomin*, although it is certainly possible that they were *minkaohan* speaking in Uyghur so as to avoid in-group disapprobation. Several female respondents, including both *minkaohan* and *minkaomin*, reported that they sometimes went to internet bars, although all admitted with a smile that their parents had forbidden them to do so. Others disapproved of this activity. Gülhärä, a female *minkaomin* student who lived in the Uyghur-dominated district of Döngköwrük, stated in 2004 that she refrained from visiting such places, since it was well known that this was where Uyghur boys and girls went 'to meet and sit together'. The image of the internet bar as a dating venue was also suggested by Dilbär, an 18 year old *minkaomin* student, who talked in 2004 about possibilities for virtual romance through the popular instant messaging computer programme QQ16: 'With QQ, it's always girls writing to boys and vice versa; never the same sex! You can choose who you want to write to, and select them. Then, if they also select you, you write to each other, and the computer tells you when a message has arrived from one of your selected "partners".' In this way, young urban Uyghurs are beginning to do something unheard of in traditional Uyghur society: they are using modern communications technology to make free, un-chaperoned contact with the opposite sex.

<sup>16</sup> As of July 2011, there were 812.3 million active QQ user accounts, making it one of the world's largest online communities.

#### hybrids: identity negotiations among the urban youth 391

Increasing numbers of young urban women have also begun to drink alcohol. This practice has long been frowned upon for females, and is outlawed for both sexes in those sections of the community which have recently returned to orthodox Islam. Like the flouting of curfews, this trend cannot be attributed especially to one or the other sub-group. Several of my female respondents had experience of drinking. Gülshäm (*minkaohan*), although making protests to the contrary, apparently considered draft beer to be better than bottled beer, and was quick to recommend it in 2004. Her friend, Räbigül (*minkaomin*), admitted to drinking red wine more often than beer, but added that both made her feel unwell. Zunun told me in 2002 about the *minkaomin* wife of a friend, who had clearly been a tearaway in her youth, frequenting the city's first nightclub Daniya, where she drank beer. The couple's 'best woman' at their wedding was also mentioned as being 'famed for drinking *aq* [Ch. *baijiu* or sorghum wine]' and was apparently considered quite impressive. Dilbär (*minkaomin*) admitted in 2004 that she too drank alcohol sometimes, usually at birthday parties attended by both sexes, but stated that boys would frequently drink more, especially when gathered in all-male groups. She asserted however that she had never become drunk, and that *minkaomin* never drank or smoked in front of their parents out of respect. On the other hand, she claimed that 'some *minkaohan* behave as though their parents were their friends and drink with them.'

This emerging trend towards female emancipation was further suggested by the attitude of some urban women to childbirth (cf. Clark 2001). Following the birth of her first child, Aliyä, a highly educated health professional in her late twenties, mused in 2002:

I'm not sure if I'll have another one. I'm so tired, and I want to focus on my work. Your society becomes very small when you have a child, you don't go out much, and you can't invite friends around because you can't host them properly. The baby needs attention all the time. But I'm very proud to be a mother. It's very rewarding!

She thus appeared to be in two minds: pulled between motherhood and career, and between family and social life. Significantly, her husband—also highly educated—was certain that he wanted more children. In terms of the Uyghur tradition of hospitality, the couple did indeed fall short of normative expectations. On my first visit to their home in 2002, it was some time before Aliyä absent-mindedly asked their nanny (a young girl from southern Xinjiang) to 'bring the table up for the guest'. Her husband produced melon and peaches, but failed to invite me to eat with the all-im-

#### 392 chapter seven

portant '*Eling*!'. They made tea, but it was not poured until several hours later, when the couple noticed that I was about to take some tablets. Despite having been my close friends for 8 years, they were too tired to cook for me on this as well as my second visit, and we dined in a local restaurant instead. They explained their busy circumstances—juggling two careers and childcare—but did not give a lengthy and elaborate apology as would have been usual practice. More urgent than hospitality was their burning need to discuss with me Uyghur (and world) politics, a subject which consumed both visits, each of around 5 hours' duration.

The waning of the hospitality ritual among busy urban youngsters was linked to a second phenomenon, that of an increasing preference for private family life at the cost of community life. As Dilbär, an 18 year old *minkaomin* student, explained in 2004:

Actually, my parents like to live apart from other Uyghurs. You know, if you live with a lot of other Uyghurs in the building, people gossip … We've got a Hui [neighbour] opposite and a Han downstairs. And if there's a funeral, we don't worry, because there are still Uyghurs close at hand. It's more relaxed this way. My Dad has one very good Han friend; he's an electrician in our work unit. Most of the Hans in our neighbourhood are pretty good.

Her description suggested a tendency among newly urbanised Uyghurs to desire an increased level of freedom, independence and anonymity which is not possible within the traditional, close-knit Uyghur community, where an effective system of moral supervision is in place.

These behavioural developments in the regional capital have been denigrated by less urbanised Uyghurs—those based in rural and southern areas of Xinjiang. Most of all, the changing behaviour of young women has come in for heavy criticism, especially by Uyghur males, both rural and urban. Above, I mentioned the wife of Zunun's factory worker friend, who had frequented the nightclub, Daniya, in her youth. When I visited the couple again in 2004, the wife talked about wanting to join an evening class to study computing skills in order that she might improve her employment prospects. Perhaps because of her slightly wild past, however, her husband was unenthusiastic about this plan, since the activity would mean her going out alone at night. The attempt by men-folk to restrict women's movements in the evenings was a pattern observable across the city in the early 2000s, although it was not always successful. Dilbär, an 18 year old *minkaomin* student, confirmed in 2004 that she rarely went very far at night, preferring to remain within the work unit complex or the street directly in front of it. Being a respectful daughter, she did this in order to ensure that

her parents did not become anxious about her whereabouts: 'If I come home too late [after 9pm Xinjiang time], my parents worry. You know, because Ürümchi is chaotic these days.' One night, the two of us went out for an evening meal to chat about urban youth issues. Sure enough, when I escorted her home at around 9.40pm Xinjiang time, her father was stood at the local bus-stop awaiting her safe return, despite the fact that she had informed him of her movements by mobile phone. Some other Uyghur wives and daughters were apparently less 'obedient'. Shökhrät, an intellectual originally from Qäshqär, was convinced in 2002 that the general moral decay of Uyghur urban society (as he saw it) was directly attributable to the excessive female reaction to modern urban life. He unleashed this polemic:

Why do so many Uyghurs take drugs now? Because they are incited to do so by others—often, young women. Young women are completely carried away by modernity at the moment. They go out in big gangs to restaurants, discos, the Holiday Inn, leaving the fathers at home to watch over their children. I knew one guy who committed suicide because his wife did that!

Hearing this, I was reminded of Siapno's account of Acehnese male reactions to the Javanese feminist campaigner Raden Adjeng Kartini. As Siapno concludes, the campaigner is essentially 'de-legitimised' as an improper heroine: 'Kartini is despised by Acehnese men because she is perceived to be "not Islamic enough" and perhaps too because her emancipatory ideals threaten the Islamic patriarchy' (2002: 150). Shökhrät's bitterest remonstrations were aimed at Uyghur hostesses: young girls who earn a living by accompanying Han businessmen and officials in karaoke bars (drinking, smoking, singing and dancing).

# 7.7. *The Beginnings of Rapprochement?*

In the wake of Uyghur urbanisation and modernisation, and the emergence of an—albeit small—educated Uyghur middle class, some scholars had begun tentatively to muse on possibilities for Uyghur-Han rapprochement prior to the Ürümchi riots of July 2009. David Wang, echoing Colin Mackerras, claimed in a paper delivered in 2004:

Currently, more and more Uyghurs from the younger generation are more prosperous than their parents. Although still being Muslims, they are more interested in getting along with the Han people rather than in rising with arms against them. A Uyghur middle class is emerging, even though with a slow pace (Wang 2004).

#### 394 chapter seven

Israeli scholar, Nimrod Baranovitch, speaking at a conference in Haifa in 2006, similarly made the case for a growing acceptance among the Uyghur youth—and even a sense of pride—regarding their incorporation within the Chinese nation (cf. Baranovitch 2007b). I personally observed the appearance of a sense of urban Uyghur pride in being one of the nationalities of the PRC during the Silk Road International Costume and Dress Festival (*Sichouzhilu guoji fuzhuang fushi jie*), held on the 20th July 2004. The festival involved a long parade through the streets of Ürümchi, featuring a troupe of dancers to represent each of Xinjiang's minority nationalities, each clad in national costume. Slogans like 'The great unity of China's nationalities! [*Zhonghua minzu datuanjie*]' adorned the streets, and the city traffic was stopped for several hours as the parade passed through the centre. Bystanders of all nationalities and their children seemed thrilled and transfixed by the spectacle. I was reminded of Dru Gladney's earlier observation that: 'Patriotism need not be exclusionary […] In Ürümchi, Uyghur residents cheered the triumphs of the 1996 Chinese Olympic gold medallists as loudly as the Han' (1997: 290).

This book has been largely concerned with illustrating the changing means by which Uyghurs of all ages exercise social agency to demonstrate difference and distance from Han people along religious, cultural, socioeconomic and political lines. Nevertheless, there are some glimmers of light; chinks in the armour, if you will. Urban attitudes towards mixed marriage may provide an example. While the vast majority of individuals in the Uyghur community continue to oppose Uyghur-Han intermarriage, there are nonetheless some individuals who stand aside from the tide of general opinion. In other national contexts, the proportion of those who could accept the advent of intermarriage grew as the frequency of intermarriage increased across generations. This was the case with Japanese Americans. While most *Issei* [first-generation] Japanese Americans opposed intermarriage, a minority of individuals saw in it hope for the eventual resolution of racial antagonisms and ethnic prejudices, for real friendships as well as social and economic harmony (Spickard 1989: 62). Similarly, while most within the African American community ruled out intermarriage on socio-cultural or political grounds, a third strand of black thought (comprising mainly middle-class, educated readers of *Ebony* magazine) espoused a universalist vision and was 'proud to be included on equal and intimate terms with non-Black people' (Spickard 1989: 305).

In the early 2000s, Uyghurs opposed intermarriage on the whole, and all the more fiercely for the fact that the trend had shown a slight increase.

However, I did interview a number of respondents who demonstrated a heightened tolerance of, or at least a capacity for empathy with, mixed Uyghur-Han unions. Most hailed from the young generation, and might be described as 'internationalists'. Rabiyä, twenty, an English-speaking *minkaomin* student (the 'punk' mentioned earlier in this chapter), stated in 2004 that while she herself 'didn't much like inter-ethnic courtship', she nonetheless considered it inevitable in the Han-dominated urban environment of Ürümchi:

If a person falls in love with someone else, what can you say? […] It's a natural consequence of their growing up with Han children […] their way of thinking comes to resemble that of Hans […] So I'm not against it as such [Uy. *qarishi ämäs*].

When reminded by her *minkaomin* friend, Aygül, eighteen, that Uyghurs are not really supposed to marry Hans 'because we're Muslims', she acknowledged the dietary problem, but said she regretted the fact that most Uyghur parents will oppose a Uyghur-Han match even if the two people really love each other. On another occasion in 2004, I witnessed Gülshäm, the *minkaohan* respondent whose case I considered in Section 7.5, reassure her divorcee father (also a *minkaohan*) that she would approve of his taking a Han Chinese wife 'provided that he believed she was the one for him'. In this way, empathy is sometimes possible where onlookers recognise that a couple's feelings for one another are sincere.

While most Uyghur men tended to be unforgiving towards inter-ethnic courtships, I nonetheless interviewed several respondents who displayed a greater tolerance and understanding. Elsewhere, I document how Tashmämät, a *minkaomin* intellectual in his forties originally from Qäshqär, remarked in 2004 that the TV drama series *Xinjiang guniang* ('Xinjiang Girls', 2004), which dealt with the issue of Uyghur-Han intermarriage, should have been called 'Xinjiang Prostitutes' (Smith Finley, 2013). When I described this reaction to Zunun, a *minkaomin* service industry employee in his thirties, he was quick to criticise it as 'very extreme', reasoning: 'The two types of woman [prostitutes and intermarriers] are completely different. One is selling her body for money; the other is in a romantic relationship with a man which is heading towards marriage! How can you compare the two?' This man had grown up in a district of Ürümchi where Uyghurs were comparatively few, and worked in an environment where the percentage of Han co-workers was high. He had witnessed several mixed unions in his workplace, although each had collapsed under the strain of 'community supervision' from the Uyghur side. Possibly relevant to his case was

#### 396 chapter seven

the fact that he had received only a minimal education owing to his family's financial circumstances, and was one of the least politically aware of my respondents.

That said, a strong political awareness does not necessarily rule out inter-ethnic tolerance in certain situations. In one case, a respondent who otherwise maintained a highly developed political consciousness (and antipathy towards Han Chinese) was surprisingly empathetic where a mixed union involved a close friend. That Ghäyrät, a migrant worker from Aqsu in his thirties, placed more importance in 2004 on his friend Osman's happiness than on community taboos was evident in the way in which he demonstrated compassion and concern for his friend's marital situation:

I don't think they [Osman and his Han wife] can be completely happy because, if so, they would have had a baby by now. Sometimes I ask him: 'Aren't you going to have a baby together?' or 'Is your wife pregnant yet?' He gives no reply. And it's not that they can't afford to bring up a child, he's made plenty of money for himself […] There was one time when he said his wife had run away and left him. But, later, they were back together again […] They seem to be OK now. I think he likes her. Because sometimes, when he comes here, and I ask him to sit with me for a while, he always says: 'No, I must be off, or my wife will be angry.' It's not that he's a *saymakhun* [a man who fears his wife]; it's because he likes to be with her. At the end of the day, I don't ask him why, because it's his decision and nothing to do with anyone else, right?

Evidently, there may have been medical reasons underlying the fact that Osman and his wife had not yet conceived a child; or the couple may have had concerns about how a mixed-heritage child would be received in such a polarised society. But the key point to be made concerns Ghäyrät's attitude, which displayed none of the condemnation or disgust expressed by other respondents; just a simple recognition that his friend's relationship made him happy and a conviction that no outsider had the right to interfere.

Some opponents of Uyghur-Han intermarriage attributed the tolerant outlook of these individuals to youth and naivety, and predicted a change in attitude with age. Yet it was clear too that some Ürümchi parents are beginning to come round to the idea. Bahar, eighteen, *minkaomin* daughter of a first-generation *minkaohan*, was in 2004 vehemently opposed to it. However, she did report a number of 'mixed relationships' at her high school, and observed—with some amazement—that in some cases parents had agreed to the courtship: 'It seems there are some who say 'yes'!' [Uy. *Hä däydighanlarmu bar okhshaydu!*]. While some of these parents were

perhaps afraid to put up opposition through fears that the couple would elope, or that they might be excommunicated, it is at least possible that a percentage holds a more enlightened view towards intermarriage and boundary crossing in general.

Prior to the Ürümchi riots of 2009, then, there were some discernible signs of *potential* for improved ethnic relations in the regional capital. An interview with a Han resident in his thirties while en route from Ürümchi to Ghulja in 2002 suggested that the same may not however be true for other towns. He observed: 'Basically, relations are OK in Ürümchi. Wait until you get to Ghulja, though. Then you'll see some hatred.' This respondent was a demobbed PLA soldier from Ürümchi turned taxi driver, who lived in Ghulja, and made frequent trips between the two locations; he was therefore in a good position to make the comparison. It is unsurprising that ethnic relations were more strained in Ghulja following the 1997 disturbances and the intensive 'Strike Hard' campaign that followed, involving the arrests of one in ten Uyghur males, according to local reports. However, it could also be argued that relations in Ürümchi simply *seemed* less strained than elsewhere in 2002 because: a) educated, urban Uyghurs no longer made open criticisms of Han (aware that the charge of 'local nationalism' now extended to verbal slander); b) urban Uyghurs were acutely aware of the need to 'play the system' if they were to obtain secure, whitecollar employment; c) increasing numbers of 'acculturated' (Chinese-educated) Uyghurs were employed in work units, and these individuals automatically appear less alien to Han residents; d) inter-ethnic relations in work units were amicable on a superficial level. However, interviews with Uyghur residents of the city in 2002 and 2004 suggested that ethnic boundaries remained very much where they had been in the mid-nineties: urban Uyghurs interacted with Han colleagues as necessary in the working environment, but interaction ceased when it was time to go home. Zeynagül, a *minkaomin* grocery store assistant in her early thirties, lived in 2004 in a residential district with an approximately 80 per cent Uyghur and 20 per cent Han population. While she seemed pleased with her newly acquired Chinese language skills, learned over the preceding two or three years, and pronounced the Hans she worked with to be 'OK', she confirmed that the two groups nonetheless maintained separate social lives: 'They keep to themselves and we keep to ourselves.'

This separation appears to develop gradually as young Uyghurs move from childhood to young adulthood. In 2004, Dilbär, an 18 year old *minkaomin* student, readily confirmed that Uyghur and Han children mixed

#### 398 chapter seven

freely at that time, both within her neighbourhood and at Han schools, and pronounced this 'quite normal'. I too observed such interactions personally, as when a Uyghur girl sat chatting on the grass with two Han friends in Xidaqiao park; or when a teenaged Uyghur girl and Han boy clutching violins stood talking and smiling on Youhaolu [the aptly named Friendly Road], having apparently emerged from a music lesson. However, Dilbär's experience suggests that relations become strained as individuals become conscious of ethnic politics (including socio-economic stratification) in early adulthood. During a discussion between Dilbär, Ömär (male classmate) and myself in 2004, she explained:

I used to like Hans well enough through primary and middle school; I grew up playing with them in our neighbourhood. I spoke more Chinese than Uyghur, and that's why my Chinese is pretty good. But something changed when I went to high school and spent more time with Uyghur students, *minkaomin*... Our high school didn't have a Chinese class. I began to think differently. Now I don't really know how to talk to the Hans I grew up with; usually, I just say Hi in the street and pass them by. Another reason why I learned Chinese so well is that my parents both speak Chinese at their work unit. But they don't like the Chinese, because they say we [Uyghurs] are not equal with them. When I was small, I didn't understand that part; it was only when I got older that I saw the inequalities between the Han and the minorities.

Her description of changing relations indicates, like my interviews about ethnic boundaries conducted in 1995-6, that patterns of interaction and segregation continue to be dictated in large part by socio-economic inequalities, regardless of the various explanations for ethnic conflict (often religious or cultural) that respondents may give on the surface. Interestingly, on a separate occasion when Ömär—her male classmate—was not present, Dilbär gave a different reason for segregation. This time, she replied absent-mindedly that the reason why she spoke little more than 'Hello' to Han Chinese in her neighbourhood was that she was so busy these days. This indicated that she had not made a deliberate decision to selfsegregate and rather *did not have time* to socialise with Han neighbours. While it is probably true that, in a situation where she lacked the time to interact with both, she had ranked Uyghur friends over Han, it seems possible that Dilbär had emphasised the politics of segregation when speaking in front of peers, while referring only to practical considerations when not under conditions of 'community supervision'.

Of especial note in the above case is that Dilbär's mother was from a *minkaohan* background, while her father was *minkaomin*. Despite their

differing levels of exposure to the Chinese language and culture, what united them in their antipathy towards those they referred to as 'the Chinese' (a nomenclature that by default separates Uyghurs from the Chinese nation) was the continued problem of Han chauvinism in the workplace and Han privilege in wider society. Thus, while the picture was not one of total mutual isolation, there were real limits on which interactions could be tolerated. Dilbär explained that when the family's Han neighbours came to visit them during the Qurban festival, a careful etiquette had to be observed by both sides:

You know how we Uyghurs help ourselves to a slice of the Qurban roast? Well, the Hans don't do that. They know, you see … if they cut a slice off the roast, then we Uyghurs won't eat it after that … because we think it's unclean (Uy. *haram däp*). We cut slices off for them [Han visitors], and put them on a separate plate.

In this way, Han neighbours were symbolically kept at a distance through an emphasis on religious differences, while the base reason for poor relations was at least partly socio-economic. This dynamic was suggested explicitly by Dilbär's response to my question about Han attendance at Uyghur weddings:

Yes, it's true that we invite Han colleagues and neighbours to our wedding parties, and the Hans invite us to theirs … but Mum and Dad do feel there are inequalities. In some areas, things are OK; but, in others, they feel that Uyghurs are discriminated against.

Again, the implication is that, beyond the obvious dietary obstacles, it is the fact of enduring ethnic discrimination that prevents closer interactions between the two groups. This situation was mirrored in a 2004 interview with Zunun, a *minkaomin* in his thirties working in the service industry. He confirmed that a considerable number of Hans had attended his wedding in the early 2000s, including his Han boss and all his Han colleagues. Given that most employees in his work unit were Han and that his (Uyghur) social circles outside of work had shrunk, this was perhaps inevitable. As he put it: 'When you get married, you have to invite the people you work with! But the Hans sit at their own tables. *People sit with their own*, with the people they know.' In the wedding video Zunun showed me, his boss and several Han colleagues could be observed attempting—rather painfully to perform Uyghur dance, as others looked on and clapped. The overall impression, then, is one of invitation by necessity: Han colleagues and neighbours are included on the wedding list out of a need to maintain appearances and, more importantly, employment connections; but they sit alone, eat alone, and dance alone.

As mentioned earlier, Zunun was one of the least politically aware of my respondents when I met him in 1995. At that time, he was among a tiny minority of individuals who refrained from criticising Han people during our conversations. Seven years later in 2002, Zunun admitted that he too had begun to 'resent' Han Chinese in some respects, and related these feelings to his experiences of ethnic discrimination in a Han-dominated workplace. In his case, discrimination took the form of promoting two Han colleagues to posts above his own, while he remained in the same position despite having joined the company years earlier than either. My reading of this case is that while Zunun claimed he had always disliked Han Chinese but had not voiced that dislike for fear of being overheard (we conversed in Chinese in 1995), his dislike was in fact retrospective, formed in the wake of negative experiences of 'equalities' guaranteed—but not produced—by the state (cf. Taynen 2006: 51-52).

Two years later, Räwiä was one of several respondents to report a worsening of employment opportunities for Uyghurs, partly as the result of inadequate government legislation to prevent employment discrimination in the private sector (a problem also affecting other groups in the PRC, such as women). She told me in 2004: 'A decade ago, private Han-run companies wouldn't consider employing *minkaomin*, only *minkaohan*; now, many won't consider Uyghurs full stop.' This development may also be seen as the consequence of heightened Han suspicions towards Uyghurs following the Ghulja disturbances in 1997, state coverage of which led many Han residents to discard the entire group as 'ungrateful' separatists (cf. Smith Finley 2011a). It is in this political climate that young Uyghurs have realised that mere survival in urban Xinjiang now depends upon playing the game according to the state's rules, including developing a (reluctant) capacity to maintain a semblance of ethnic unity to the extent that this allows them to access education and employment opportunities and make a secure living. When I pointed out a *minkaohan* Uyghur sat chatting with friends in an Ürümchi park in 2004, Aynur, a *minkaomin* graduate in her early twenties, responded: 'We *have* to be united with the Han now; we have no choice. Unity is a requirement of this society. This is the nature of our situation [Uy. *wäziyät ashundaq*].' Ömär, an 18 year-old *minkaomin* student, nodded in agreement and rejoined: 'If we don't learn Chinese, do things their way, get along with them, they will say "They're not the same as us, they're backward" and simply throw us away.' On another occasion in 2004,

when I asked Ömär whether he believed Uyghurs and Han Chinese could ever be genuinely united, his reaction was again resigned and practical. Head on one side, he ventured slowly: 'We're *getting* united (Uy. *Boliwatimiz*) …'. Like Aynur, his response implied that young people feel obligated to humour the state's goal of ethnic unity on the surface even as they nurture a quite different identity beneath it. Interestingly, Uyghurs who had spent some time in Beijing demonstrated greater capacity for a nuanced consideration of Han people as individuals. In 2004, I asked Möminjan, a student in his early twenties from Turpan, whether he liked Han Chinese in Beijing. He clicked his tongue (the Uyghur negative) and shook his head. The same question, this time about Han Chinese in Xinjiang, provoked the same response. But then he remarked: 'But there are some good people among the Hans, some who are sympathetic to Uyghurs. Every nationality has good people and bad people, right?' A high-profile Uyghur musician had made a similar observation during that same year. In essence, while most Uyghurs do not *like* Han Chinese (for religious and cultural reasons), increasing numbers recognise that interaction can be managed where Han counterparts have the correct attitude towards the Uyghur situation, i.e. where they acknowledge policy failures regarding social (ethnic) equality, and where they respect Uyghurs' rights to linguistic, cultural and religious expression. In the light of the preceding arguments, it seems worth noting that the only group of *minkaohan* to date who were able to act as a 'bridge' between the Uyghur community and the Han administration was the oldest cohort. These people had enjoyed respected positions in the community during the 1950s and early 1960s, and their experience of working with the small number of Han then settled in Xinjiang was positive because many Han attained functional fluency in Uyghur and nearly all abstained from practices that would have caused offence to the local population (Taynen 2006: 50-51). In other words, *minkaohan* were able to act as arbiters during this period because on the one hand they were properly rewarded with high-status positions for their mastery of the Chinese language, while on the other early Han settlers were willing to adapt to the local lifestyle, including religious and cultural mores.

Writing in the early nineties, Marsot said of the Hindu-Muslim conflict in India that 'in fact the more affluent Muslims have reached a fair level of accommodation with the Hindu majority', although he went on to warn that continuing religio-ethnic strife in India, combined with the Islamic resurgence worldwide might simultaneously increase the political role of Islam within the Indian Muslim community' (1992: 166). I would argue that the same may also be true for Xinjiang, that is, even if the ethnic gap in

#### 402 chapter seven

language status, cultural status, education, employment and income were bridged, there would remain some individuals who would continue to oppose Chinese state hegemony on religious, cultural or purely political grounds. However, socio-economic equality must surely provide a starting point for improved social stability. Uradyn Bulag has asked:

How […] can Manchus, Mongols, and other non-Chinese peoples be accommodated as equal national siblings in unity against a racial imperialism from without? […] especially when those now styled as minority nationalities persistently refuse to be confined to this newly imagined sinocentric community (2002: 7).

This point is key to the future improvement of ethnic relations in border regions of the PRC; without equal rights, it will be extremely difficult to convince Uyghurs and others that they should perform their civic 'duties' of ethnic unity (*minzu tuanjie*), ethnic harmony (*minzu hexie*) and the safeguarding of the unification of the motherland (*baozhang zuguo tongyi*) (for an expansion of this argument, see Smith Finley 2011a). Lucian Pye warned more than three decades ago that the 'levelling of cultural differences' (i.e., the gradual acculturation of minority nationalities through for instance a Chinese-medium education) could in fact lead to greater ethnic tensions as economic and political power considerations are elevated in relative importance. He predicted too that the narrowing of economic circumstances could in turn lead to a greater minority concern with the power status and political efficacy of their communities (Pye 1975: 509-511). It could be argued that in Ürümchi today, Uyghur urban identity has commenced on just such an evolution.

# *Conclusion: From 'Ethnic Anomaly' to Subaltern Cosmopolitan*

With regard to Uyghur *minkaohan*—especially those of the first generation—an analogy may be drawn with Eriksen's second category of 'ethnic anomaly': the children and grandchildren of immigrants (1993: 62-64). In Eriksen's sense, these are individuals whose parents or grandparents immigrated into another national space and into the contrastive culture of its majority group. In Xinjiang, the term may be applied in reverse: that is, *minkaohan* children grow up in a space previously dominated by their parents' or grandparents' cultural norms but now colonised by a contrastive majority culture. In other words, the new culture has immigrated into the physical territory of the indigenous culture and over time established its

dominance (where perceptions of indigeneity are relative). Though the migration dynamics are different, the resulting systems of majority/minority classification and politics of socio-economic stratification are surely comparable. Furthermore, ethnic discrimination and social inequality are all the harder to bear when suffered in one's own homeland (as perceived by Uyghur protagonists).

During the mid-1990s, *minkaohan* resembled biracial or culturally micro-diverse individuals to the extent that they did not fit neatly into either main culture—Central Asian or Han. As a result, they were sometimes identified—and identified—with their majority status (as quasi-Han) and sometimes with their minority status (as Uyghurs). Much of the time, they experienced exclusion from both and found it easiest simply to mix with other *minkaohan*. At least four methods for dealing with ambiguity (as categorised by Douglas 1966) were observed in Xinjiang during this decade:


Yet while in-group reaction to the ambiguous category of *minkaohan* was quite negative during this period, it was evident to an outsider that over time, *minkaohan* individuals often developed a greater capacity to know the enemy, and became deeply cynical towards state policies and their Han colleagues. This pattern was also noted in the context of (former) Russian Turkestan where, for indigenous elites, socialism was initially seen as a tool and technique for underground activism and mass action, a means of gaining foreign support, and a possibility of equality in a situation where independence could not be achieved. Their hopes expired soon after the

#### 404 chapter seven

integration of Turkic nations into the Soviet Union (Micallef and Svanberg 1999: 154), and the realisation that Soviet 'internationalisation', while sounding good in theory, involved a strong element of Russification. For many of Xinjiang's first-generation *minkaohan*, childhood involved studying in a Chinese-medium school alongside Han classmates, identifying with the privileges of the oppressor, and internalising Han projections of Uyghur inferiority. Yet during the course of many years spent in a Han-dominated urban workplace (then normally a state work unit), many developed a strong sense of injustice regarding the unequal treatment afforded Han Chinese and minority nationalities, and their attachment to their own ethnic group strengthened in response. In other words, such individuals developed a *reactive* identity. Yet despite recognising and living with ethnically informed social injustice, at the same time they remained unable to participate fully in Uyghur cultural life (owing largely to their limited linguistic capabilities) and continued sporadically to make negative observations about the Uyghur ethnic group that suggested enduring pockets of internalised oppression. As I have described in detail elsewhere (Smith Finley 2007a), such individuals learned over time to situationally manipulate their Chinese linguistic capability and relatively deeper degree of acculturation in order to win high-status positions in their work unit and better life chances for their children. Most raised their own children to be proud of the Uyghur language and the Uyghur culture, and some also instilled in their offspring a strong sense of religiosity by orchestrating frequent interactions with the child's grandparents. As a result, the sons and daughters of first-generation *minkaohan* frequently grew up having a perfect command of both Uyghur and Chinese, and thus the potential to be Uyghur in some situations and Chinese in others, as circumstances required.

In this way, across the course of the 1990s, many *minkaohan* began to develop a double consciousness, an ability to use their dual identity situationally; they emerged, if you like, as modern Uyghur survivors. But would the inherent ambiguity of their position bring unbearable emotional and social consequences? Or could human beings, as Douglas theorised, allow new patterns to evolve in which anomalies had a place? Writing on mixedrace experiences, Alcoff warned that 'cast off from all communities, the individual has no historical identity, and thus is unlikely to value the community's future' (1995: 274). If Chinese-educated Uyghur peers were to be retained as loyal members of the in-group, it was crucial that a change in attitude take place within the Uyghur community, allowing for a height-

ened degree of tolerance. Only then could the liminal *minkaohan* shift from the negative experience of 'neither-nor' to a positive hybrid consciousness of 'both-and' (Eriksen 1993: 62), and only then could the broader Uyghur community prevent the loss of those individuals to the group.

In fact, this process may be said to have begun in Ürümchi and other urban spaces from the end of the nineties and early 2000s, aided partly by the ever higher proportion of urban youth being educated in Chinesemedium schools, but also by the diversity of global flows making their way in and out of the regional capital. Gradually, the former 'anomaly' has begun to turn into the norm. Instead of being represented and rejected by both communities as 'lacking' and 'suspect', *minkaohan*—like mixed-race individuals—may gradually come to enjoy positive, new identities as border-crossers, negotiators and bridges between communities (Alcoff 1995: 276). Alcoff, drawing on the writing of Anzaldua (and echoing the work of Hall), suggested that such a future requires a new, transformative way of thinking that celebrates 'identity without purity' (1995: 257; 276-78). This way of thinking is evident, for example, in the progressive views found among *Xinjiangban* respondents like Tursungül, who claims to 'possess the advantages of the three [Uyghur; Chinese; Western] cultures'. In such a context where individuals have learned to celebrate hybridity, one might expect to see increasingly tolerant attitudes towards intermarriage; yet in reality the embrace of diversity remains highly selective. Indeed, *minkaohan* may remain some of the fiercest opponents of mixed Uyghur-Han unions, based upon their often greater awareness of ethnic inequalities in the environment in which they were raised.

On the other hand, aesthetic cultural forms as well as deeper cultural values from alternate centres like Turkey, the Middle East, Central Asia and the Mediterranean are being eagerly grasped by a willingly hybrid youth. Writing on processes of cultural globalisation, Motti Regev describes how contemporary 'aesthetic cosmopolitanism' in the form of appreciation of pop rock fusion was preceded in two countries—Argentina and Israel—by a form of musical ethno-nationalism that featured a quest for 'essentialist distinction'. Initial lack of appreciation for rock fusion was derived from a perceived lack of 'authenticity', where new musical forms were viewed as a threat to national purity. Yet in each case, the earlier emphasis on purism gave way over time to fluidity, an openness to 'otherness', and a media discourse that stressed the need for local culture to keep pace with broad, global artistic innovations (Regev 2007: 326, 328). In the 1990s, Uyghur traditionalists such as the *dutar* player, Abdurehim Heyit, sought to preserve

Figure 15: Bilingual pop fusion—Poster of Uyghur 'Guitar King', Arken Abdulla, Döngköwrük, Ürümchi (2004)

and transmit Uyghur musical traditions to the next generation in a bid to ensure the continuity of the Uyghurs as a group. In the 2000s, young Uyghur musicians are instead opening Uyghur culture up to *selected* outside influences and, in doing so, they reshape the new generation as part of a confident, unfettered global youth. By aligning themselves with Mediterranean and North African Sufi cultures, flamenco guitarist Arken Abdulla's fans shift the boundaries and 're-map their lifeworld', positioning themselves as confident, passionate, outward-looking 'world citizens' (cf. Jaffe and Sanderse 2010: 1570; Stadler 2011); as 'subaltern cosmopolitans' capable of border crossing and connecting in a way that transgresses the established order (Gidwani 2006: 19), in this case Chinese national (political and cultural) borders. While Eri frames this process in terms of the 'economic and

# hybrids: identity negotiations among the urban youth 407

social forces of globalization currently aiding the survival of the Uighur nation' (2008: 78), I would suggest that as young Uyghurs reach out to culturally sympathetic spaces around the world, they do not simply maintain Uyghur traditions; they enable the ongoing development and transformation of a modern Uyghur culture that orients itself towards the Turkic, Arab and Mediterranean (Spanish/North African) West.

# ConClusions

408 conclusions

# *Mainstreaming of Uyghur National Identity in the 1990s*

From the end of the eighties and particularly after 1991, uyghur identities experienced radical change. The 1990s saw the mainstreaming of a strong uyghur national identity that reached beyond urban, intellectual circles, as well as the development of ethno-political aspirations among certain though not all—social groups. in some respects, this dream of uyghur independence was a new phenomenon. none of the independent regimes established in Xinjiang during the second half of the 19th century and first half of the 20thcentury can be described as 'pure' uyghur nationalist movements since, as outlined earlier, the rebellions were rarely initiated by uyghurs, while the regimes in question represented a unification of Turkic Muslim groups or, in the case of the secular East Turkestan Republic (ETR, 1944-1949), of Muslim and non-Muslim groups. At that time, uyghurs had not yet begun truly to consider or represent themselves as a political entity separate from other indigenous groups.

What catalysed identity change at this precise historical moment? in this volume, i have characterised the period 1991-1997 as a 'run-up to an imagined independence' in the minds of young males: a period of growth in ethno-nationalist sentiment, punctuated by widespread sounds of discontent. This discontent resulted from a combination of internal (domestic) and external (international) factors. The internal factors stemmed almost exclusively from the ill-conceived policy of Han in-migration to Xinjiang, which has had three visible effects on life for local people. Firstly, as the number of Han settlers has grown, pressure on fragile inter-ethnic boundaries has increased, making religio-cultural differences harder to manage and creating more instances of boundary crossing. secondly, Han in-migration has led to increased socio-economic competition for education, employment and the region's resources, and to uyghur perceptions of ethnic inequality. Finally, it has had a profound effect on the psyche of Han migrants themselves. As the Han community in Xinjiang has expanded, new settlers increasingly feel confidence in numbers, and no longer see the necessity to adapt to local cultural traditions and norms (Ch. *ru xiang sui su*), as previous generations of migrants had. This has caused uyghur

communities to complain of escalating ethnic discrimination as a result of Great Han chauvinism.

Parallel to the increase of Han in the regional population, the Chinese language has been slowly but systematically institutionalised in the multiple spheres of education, employment and administration: a policy trend which should be viewed as the founding stone of the income and status gap between local Uyghurs and Han migrants. Uyghurs who wish to obtain a higher education and gain professional, white-collar employment in state organs or Han-managed companies are obliged first to attain fluency in Chinese. Conversely, as Chinese gradually replaces Uyghur as the urban if not the regional—*lingua franca*, new Han settlers are increasingly unwilling to learn the local tongue, perceiving it to be of little use in the Han-dominated urban environment in which they live and work. In adopting this attitude, they alienate themselves from local peoples as well as from first-generation Han settlers for whom it was only natural to assume the language and ways of the territory.

Yet these internal factors were already starting to come into focus in the 1970s. By that time, the proportion of Han settlers in the regional population had grown to 40 per cent, while the 1980s saw the rapid expansion of the region's industry and commerce mainly for the benefit of Han settlers. These factors did not initially give rise to separatist aspirations, and they were not alone responsible for the dramatic identity transformations witnessed during the 1990s. Rather, what triggered this response at this particular moment was the vision, broadcast around the world by the mass media, of collapsing Marxist-Leninist politics and a series of multi-lateral interventions made on behalf of some—though not all—of the world's 'stateless nations'. As a result of these international spectacles, some Uyghurs—in particular young males—began to dream that they too might achieve political independence for Xinjiang, given the right conditions.

Prior to 1989, though there had been intermittent Uyghur protests against ethnic discrimination, including a first call for 'equality between nationalities' in 1988, public demonstrations had been peaceful and participants had not demanded secession from the PRC. The first calls for Xinjiang secession coincided scarcely by accident with the 1989 pro-democracy movement centred on Beijing's Tian'anmen Square. One of the student leaders involved in the movement, Örkäsh Dölät (Chinese name, Wu'erkaixi), was a young Uyghur man. Proud that a person of Uyghur nationality should have evolved into a major figure in the protests, Uyghurs back in Xinjiang began to wonder whether opposition to contemporary

Chinese rule had now become possible. After all, here were the Han people protesting against their own government. Although the 1989 pro-democracy campaign failed in China, 1989 brought a series of revolutions in Central and Eastern Europe against the continuation of communist one-party rule, beginning in Poland (February), and spreading to Hungary (August), East Germany (August), Bulgaria (October), Czechoslovakia (November) and Romania (December). These movements would eventually culminate in the fall of the Berlin Wall and German reunification (1990), and all but one involved popular campaigns of civil resistance. Only the Romanian regime was overthrown violently.

In Xinjiang, some Uyghurs began to joke that although the Chinese government once said that only socialism could save China, it now believed that only China could save socialism (Rudelson 1997: 14), a joke also popular among Han Chinese during that period.1 However, it was the disintegration of the USSR and formation of independent Central Asian states in 1991 that sowed the definitive seeds of independence in some Uyghur minds. Following the establishment across the border of the new Kazakhstan, Uzbekistan, Kyrgyzstan, Turkmenistan, and Tajikistan, the Uyghurs, the Tatars, and the Salars became the only Muslim Central Asian groups in Xinjiang without a nominal homeland.2 The vision of politically independent neighbours subsequently encouraged the consolidation of Uyghur national identity as well as the notion of Uyghur (rather than Turkish-Islamic) self-determination.

# *Eclipse of the Old and Foregrounding of the New*

In this all-new political context, long-standing Uyghur identities relating to region, oasis and social group, previously central in ordering in-group relations, were suddenly less salient. While some individuals among the educated and the wealthy maintained superior attitudes towards groups of lower social status (see Chapter 4 Illuminists), and Uyghurs from one oasis continued to poke fun at those hailing from another, class and oasis differences became far less significant than the escalating conflicts with

<sup>1</sup> Personal communication, Professor Gregor Benton.

<sup>2</sup> The Salars are said to have originated from a Turkmen tribe (Schwarz 1984: 39-40), and therefore might be said to have their own country in Turkmenistan. The Tatars and the Salars in Xinjiang numbered only 4,821 and 3,660 persons respectively in 1990 (Hoppe 1992: 360).

the Chinese state and its representatives, the Han people. In illustrating these transformations in Uyghur identities, I have found Sartre's theory of 'we-hood' (shared experiences among the in-group) and 'us-hood' (ingroup loyalty through enmity towards the out-group) invaluable. On the basis of this model, I have argued that Uyghurs had long shared a sense of 'we-hood,' expressed in common patterns of social, cultural, religious and economic practice. This was expressed most poetically by Kulbhushan Warikoo, who wrote: '[Xinjiang's] oases were like beads on a string separated by desert sands and inter-connected by caravan routes' (1985: 78). While the sense of 'we-hood' was sometimes undermined in the past by competing, divisive identities (north-south, inter-oasis, urban-rural, social group/class), increased contact with the Han coloniser, combined with unprecedented global political change, engendered in Uyghurs a new and complementary sense of 'us-hood'. Uyghur 'us-hood', constructed in relation to the external threat of the Chinese state and the Han migrant population, fused with the pre-existing sense of 'we-hood' in the context of re-nationalisation processes taking place in the neighbouring Central Asian states. As a result, the Uyghur national identity was rendered whole. Now, Uyghurs from different social backgrounds began to draw upon the unifying core that had been present among them all along: the body of cultural and religious practices governing their everyday, taken-for-granted social interactions for more than five centuries.

At the same time, several new identities emerged. The identity of the *minkaohan* (Chinese-educated Uyghurs) was potentially divisive, and involved often damaging in-group judgements of linguistic and cultural status as 'pure' and 'impure', 'real' and 'fake'. The second, an orientation towards Western culture, was neutral and potentially transcendent of the Uyghur-Han binary and Chinese national borders. The third, a broad-based Uyghur national identity, saw Uyghurs increasingly identifying themselves as a distinct ethnic entity. And it was this latter that created the potential for certain social groups to develop an ethno-political identity and aspirations to Uyghur independence in the context of changing domestic and international circumstances.

#### *Factors Affecting Uyghur Solidarity in the 1990s*

Yet despite these developments, the successful mobilisation of a unified movement for Uyghur independence proved difficult if not elusive. State

Figure 16: Convoy of People's Liberation Army trucks, rural Aqsu (1996)

restrictions on freedom of speech and association, combined with the heavy military and paramilitary presence in the region, forced aspiring separatists to carry out their acts of resistance underground, in guerrilla fashion (assassination; sabotage). Meanwhile, significant fissures in attitude remained across social groups. To start with, views on a possible secession differed greatly across generations. Male youths formed the core of new activists seeking political and administrative independence from the People's Republic. However, the elderly and the middle-aged, who bore the scars of persecution during the Cultural Revolution, proceeded with greater caution. Like Han families in China proper in the wake of the 1989 Tian'anmen incident, they wanted to protect their homes and livelihoods, and feared the return of social and political instability.

At the same time, the young generation itself faced internal divisions. Across the 1990s, increasing numbers of urban Uyghurs had been educated

partly or wholly in Chinese-medium schools. These *minkaohan* often struggled to be accepted as 'true Uyghurs' among their *minkaomin* peers, and were ostracised as somehow lacking cultural authenticity.

There was also a significant disparity between urban and rural attitudes, owing to the contrastive experiences of city-based Uyghurs and farmers. Most urban Uyghurs, while acknowledging that the Chinese state had opened up the urban labour market, felt that Uyghurs had benefited little from economic development. Although their overall standard of living was substantially higher than prior to 1949, urban dissatisfaction now stemmed from income and status gaps in relation to Han in-migrants. The situation of rural Uyghurs was very different. In the mid-nineties, the number of Han settlers in the countryside was comparatively few, and thus Uyghur farmers had little experience either of Han chauvinist attitudes or of socio-economic inequalities. For them, the key issue was less the broad question of Uyghur independence and more the narrow need to resist state birth control policy. Finally, Uyghur males during the 1990s were often loath to include their female counterparts in political discussions, rejecting the opposite sex as too focused on marriage and motherhood, and therefore lacking the political awareness necessary to counter the external threat. In adopting this position, they missed the opportunity of harnessing the support of one half of the Uyghur population to their cause.

### *1997: Enter the Silent Era*

If the period prior to 1997 was a time of unprecedented consolidation of a broad-based Uyghur national identity, the Ghulja disturbances of February 1997, their military suppression, and the ensuing state crackdown should be viewed as a definitive turning point. These disturbances represented both a climax and an anti-climax in ethno-political terms. In their wake, Uyghurs suffered from indiscriminate state repression, combined with the shock realisation that external political support would not be forthcoming in their case. The ensuing 'era of silence' appeared to engender a polarising effect on Uyghur society. From then on, Uyghurs were increasingly confronted with a bi-polar world in which they could resist and face socioeconomic marginalisation, or accommodate to ensure socio-economic survival. If they chose the latter, they risked linguistic and cultural extinction. Yet, as Gardner Bovingdon has asked: can Chinese officials really hope to stop the use in Xinjiang of innuendo, symbolism, allegory, and other

strategems as vehicles through which to resist the Chinese national project (2010: 81)? He rightly concludes that while China has moved from erasing troublemakers themselves to an attempt to erase troublesome *ideas*, Uyghurs have simply 'refused to give those ideas up' (Bovingdon 2010: 84). In response to changed political conditions in which verbal criticism—direct or metaphorical—of another ethnic group was punished as an attempt to 'split the motherland' (and thus as state subversion), methods of Uyghur resistance transitioned almost seamlessly from voiced critique to silent statement in the form of religious renewal, endogamy and subaltern cosmopolitanism.

That for some protagonists symbolic resistance now took a purely religious form should not come as a surprise. Throughout the modern history of the Northwest, opposition to the Chinese polity has been mobilised along religious lines. One inspiration for the 1864 Muslim Rebellions was the resurgence of Sufism to the west, while the Turkish-Islamic Republic of Eastern Turkestan (TIRET, 1933-34) advocated the introduction of Shari'a law, though it showed willing to 'adapt aspects of Islamic custom in accordance with political and social conditions' (Forbes 1986: 113-114). Even the protagonists of the secular East Turkestan Republic (ETR, 1944-49) used Islam strategically as a spiritual focus in the initial stages of their campaign. Since the events of 9.11, the Chinese government has attempted to capitalise on the global 'war on terror' by alleging the existence of links between groups in Xinjiang and international Islamist organisations such as Al-Qaeda. Yet to what extent is this charge really justified? During the 1990s, one heard anecdotal reports from time to time that weapons were being supplied to Uyghur separatists by organisations in Pakistan, Afghanistan, Pakistani Kashmir, Iran or the Central Asian republics. Yet my own interviews conducted in Xinjiang in 1995-6 yielded no evidence that respondents had begun to adopt an extremist Islamic identity. On the contrary, in 1996 many Uyghurs expressed reservations about the social behaviour of Pakistani traders in Qäshqär, and distinguished carefully between Uyghur and Pakistani culture. There were certainly no indications that respondents advocated radical Islamism for Xinjiang, or even that they favoured a return to Islamic traditionalism.

Since 1997, Islam in Xinjiang has indeed undergone a renaissance among some sectors of the population. However, as I hope I have shown in Chapter 5 (Reverts), this has been a complex and multi-faceted process, by no means solely attributable to the deployment of Islam as a tool of symbolic resistance. For many individuals, the return to orthodox practice does con-

stitute a form of local opposition to state oppression of religion; and by 2002, many Uyghur respondents expressed a firm sense of solidarity with oppressed Muslims around the world (Chechens; Palestinians; Iraqis, Afghans) (see Smith Finley 2007b). Yet for most respondents, the return to prayer was far from one-dimensional. While some have struggled with what they see as failed policies of regional development, others contrast widening socio-economic inequalities negatively against Islamic—and sometimes Maoist—ideals of egalitarianism. Whereas some have abandoned politics for religion in the wake of frustrated aspirations to independence, others are reacting against negative social—especially gender-related changes brought about by modernity. For still others, who fear that the current miserable situation of the Uyghurs is a punishment from Allah, orthodox Islam is a vehicle for personal and national reform. The reasons for the current Islamic renewal in Xinjiang are complicated and diverse; but what can be stated with certainty is that this renaissance is currently taking a symbolic, peaceful and cathartic form.

While religious renewal may strengthen unity among observant and newly observant Uyghurs, it also threatens to create a new in-group divide between those who continue to resist (in symbolic religious form) and those who choose the nominally secular path of pragmatic accommodation. Observant Uyghurs often express distrust of pragmatists, rejecting them as 'disloyal' or 'sinicised'. Yet the experiences and self-ascriptions of the pragmatists speak otherwise. Interviews suggest that many if not most accommodate themselves to Han-dominated urban society only insofar as they draw benefit. As I argue in Chapter 7 (Hybrids), young Uyghur people growing up in urban centres may be located at a range of positions along an 'accommodation spectrum' ranging from the extremes of 'pure' (Uyghur as mother tongue; adhering to orthodox Islam) to 'deeply acculturated' (unskilled in mother tongue and culture; desirous of full assimilation into Han society). However, an individual's personality may be composed of a set of diverse features drawn from different positions on that spectrum. Not only this, but many people have become adept at negotiating multiple, hybrid identities, so that they re-position themselves at different points on the spectrum according to time, place and necessity. While pragmatists may be obliged to temporarily forego aspects of their Islamic identity (head coverings; daily prayers) at school or in the work unit, this does not necessarily mean that their Muslim identity is less important. It simply means that for many it plays out in the private sphere of the home rather than in public. A study conducted in Ürümchi in 2003 clearly showed that, when

asked to choose identity ascriptions, urban Uyghurs of differing educational levels, gender, age, and cadre/non-cadre status emerged as a 'homogenous and cohesive ethnic group', both in their core ethnic identity and in their sentiments towards Chinese government policies. The study concluded that Uyghurs are 'apparently held together by a strong centripetal force—a common language, religion and culture' (Yee 2003: 448).

Moreover, for some young pragmatists, hybridity itself has become a form of symbolic opposition, as demonstrated in the interviews with *Xinjiangban* graduate, Tursungül (see Chapter 7). The twin forces of higher education and globalisation—involving daily exposure to mass media, and mass production and consumption—are exerting an individualising and empowering effect on today's Uyghur urban youth. Faced with myriad choices concerning what food to eat, what clothes to wear, and what music to listen to, many opt to transcend Chinese national boundaries through the products they consume. In so doing, they utilise means that are not explicitly political to neatly sidestep multi-ethnic membership of the bounded Chinese nation. Through a free and boundless form of 'subaltern cosmopolitanism', they align themselves instead with alternative cultures with which they feel a deep affinity: Turkey, the Middle East, the Mediterranean and North Africa. And, when it comes to the question of marriage to a Han partner, the articulated hierarchy of intermarriage preferences mirrors, with few exceptions, that of their observant Uyghur peers.

Given the imposition of Chinese as sole language of instruction since 2002, we can expect the emerging trend in urban youth studies—conceived as studies of multi-lingual and multi-cultural hybridity—to predominate in future Uyghur studies research. As Alexander, Kaur and St Louis point out, recent studies of identity have taken a celebratory form, focusing on issues of diversity, multiplicity, hybridity, fragmentation and change (e.g. mixed race identities; global popular culture). At the same time, however, these authors rightly warn that empirical studies must aim both to 'give voice' to those whose lives are silenced or 'ventriloquised', and to (re)place culture and identity as 'sites of struggle, constraint and resistance' (2012: 4).

#### *From Independence to Federation*

Given the identity changes that have taken place over the past two decades, what then of Uyghur dreams of independence? In the mid-nineties, this question was already troubled by competing political visions. Some south-

ern Uyghurs, predominantly those in Qäshqär and Khotän, articulated a pan-Turkic vision and advocated the (re-)establishment of a state of 'Eastern Turkestan', a notion that first resurfaced during the 1990 Baren riots. However, some northern Uyghurs, particularly the highly educated based in Ürümchi and Ghulja, had begun to speak of a nation named 'Uyghuristan',3 a choice of name which reflected the strengthened Uyghur national (as opposed to pan-Turkic) identity, but risked alienating Kazakhs and others in the north.

Since the 1997 Ghulja disturbances, aspirations to independence have receded to a significant degree. Yet the core debates remain. In the context of the Islamic renaissance occurring in the early 2000s, some southerners who formerly advocated the establishment of a pan-Turkic 'Eastern Turkestan' now envision that polity as an Islamic state. This ambition is however viewed with deep suspicion by urban intellectuals. Shökhrät, a scholar in his thirties from Qäshqär, commented: 'The *taliplar* [religious students] go on about creating an Islamic State of East Turkestan, but they have no idea what form this new state would take!' His discomfort reflects that provoked in the Middle East, where theology students have sought to win popular support through the use of vague, emotional appeals in place of detailed policy proposals (Anderson 1997: 25). Both Shökhrät and Tashmämmät, another scholar in his forties, rejected the possibility of an Islamic state outright, describing the students' vision as 'ill-thought out and irrational'. Aliyä, a health professional living in Ürümchi, also signalled a pan-Turkic attitude in interview in 2002:

I often wonder what it would have been like if we had still been governed by the Guomindang. You know, was it Sun Zhongshan [Sun Yat-sen] who categorised the Chinese people into 5 groups: Han, Manchu, Mongol, Turks,4 Tibetan? That was a much better system, I think. Why did the Communists divide the Turkic people into such tiny groups? Kazakhs, Uyghurs, Uzbeks,

<sup>3</sup> According to Bovingdon, the notion of a 'republic of Uyghuristan' was originally invoked at a conference in Ghulja in 1951, at which prominent Uyghur leaders and others raised grievances over the failure of regional autonomy to allow them authority over their own affairs (2010: 82). In the mid-nineties, it circulated also among the Uyghurs of the Kazakhstan borderlands, who in the wake of the dissolution of the USSR had become 'acutely aware of their liminal status as a stateless people in a new Central Asia where political representation has become linked to nation states' (Roberts 1998a: 524).

<sup>4</sup> A point of interest is that in Sun Yat-sen's original formula, the group Aliyä refers to above as the 'Turks' was in fact designated 'Hui', according to the older use of the term which included both Tungans (Hui Muslims) and Turkic Muslims. Yet the Hui did not figure at all in Aliyä's version of Sun's formula, perhaps reflecting the ambivalent attitude of many Uyghurs towards their co-religionists.

Tajiks, Kyrgyz … we can all talk to one another, and understand! The language is 60-70% the same. And yet they categorise all of the Han as one group, while Beijingers can't understand the Shanghainese, and the Cantonese speak a totally different language!

On a separate occasion in 2002, this respondent pondered over what life might have been like for Uyghurs under the Japanese, a highly inflammatory notion given the historical and contemporary anomie which characterises Sino-Japanese relations. Notably, in neither case did she touch upon the possibility of Uyghur self-rule; her musings were always framed in the context of *being ruled by someone else*.

Respondents who had dared to dream of an independent Uyghuristan back in the mid-1990s were quite disillusioned by the early 2000s. As noted by ethnomusicologist Rachel Harris, the preoccupation with sending Uyghur offspring abroad to study reflected 'a new mood among Ürümchi intellectuals, most of whom had replaced the nationalism of the mid-1990s with an increasingly pragmatic and global outlook' (2005: 637). While my respondent Shökhrät continued to support the notion of Uyghur self-determination in 2002, the spectre of an irrevocably altered regional population composition gave rise to serious doubts as to whether this was possible. He explained:

Suppose in future China proposed to split into a federation of states, and Xinjiang's fate were being decided. At that time, if the Hans outnumbered the Uyghurs in the region, Han votes to join the federation would outweigh Uyghur votes to secede.

He very likely had the situation of neighbouring independent Kazakhstan in mind. The large number of ethnic Russians in that country has made its proposed repatriation into a larger Russia a recurring theme for ultra-nationalist factions in ethnic Russian political circles (Micallef and Svanberg 1999: 149-50). Two years later in 2004, Shökhrät no longer considered independence to be a sensible course, this representing a dramatic change of direction for a man once the staunch supporter of a secular, nationalist secession. His thinking appeared to have been influenced by political developments (especially compromised democracy) in the neighbouring Central Asian states. He explained:

If Uyghurs did gain independence, it would be like a tail with seven, eight or nine heads: the *taliplar* would go and do their own thing; the *minkaohan* [Chinese-educated Uyghurs] would go one way, the *minkaomin* [Uyghurs educated in the mother tongue] another; the secular intellectuals yet another way, and so on. The situation would be unmanageable, and in the end a

dictator would have to emerge in order to control it. This is what has happened in the Central Asian states. The chaos would result in the economy crashing … on all fronts, it would not be a good solution for the Uyghurs […] In a federation, Xinjiang would be like one tail with one, two or at most three heads. There would be an existing legal framework to prevent the people from descending into chaos and disunity. People would have to adhere to the law. The economy would remain stable.

Several other respondents echoed this view, citing the saying: '*Otturidä burun bolmisa, ikki köz bir-birini yäp ketidu*' [roughly translated as 'Without the nose (the Han Chinese) in-between, the two eyes (different sectors of the Uyghur populace; different ethnic groups in Xinjiang) would go and eat one another'].

The potential for internal fractiousness was neatly reflected in a conversation held in 2002 between Abdukerim and his friend, two observant university graduates raised in rural Aqsu. Each argued that an independent state should be called 'East Turkestan', since the name 'Uyghuristan' would cause unhelpful friction between Uyghurs and Kazakhs. Indeed, they proposed that ethnic categories be abolished in the new state and replaced by the term 'Turks' to refer to all Turkic groups (cf. interview with Aliyä above). However, they could not agree on where the capital of the new state should be located. The friend favoured Ghulja in north Xinjiang 'in order to keep control of the large Kazakh contingent up there'. This strategy reflected the fact that Kazakhs would become the second largest minority group under independent Uyghur governance (Han Chinese being the largest, assuming that independence did not lead to mass Han exodus). Abdukerim, on the other hand, favoured Qorla on the grounds that the capital 'must be in the religio-cultural heartland of the south', although he rejected Qäshqär on the grounds that it was 'too close to Pakistan'. Both rejected the current regional capital, Ürümchi, which they apparently considered a lost cause owing to the high proportion of Han residents.5

<sup>5</sup> A comparable case of the importance attached to the location of a capital city is available in neighbouring Kazakhstan, where the national capital was moved from Almaty in the south, a region where mainly Kazakhs are concentrated, to Astana in the northcentral area, which is populated by a mixture of Kazakhs and ethnic Russians (Faranda and Nolle 2011: 633). In enacting this move, the Kazakh political elite presumably hoped to reassure the sizeable Russian population now stranded in the newly independent state. At the same time, they probably perceived this area to have benefited from Russian influence in terms of modernisation, and therefore to be most suitable as the economic, cultural and political centre of the new state.

By 2004, both Shökhrät and Tashmämmät were in agreement that the best solution for Xinjiang at the present time would be to form part of a Chinese federation. The changing mood was reflected also in the conclusions of several academic studies conducted in the early 2000s. Three years earlier Colin Mackerras had written: 'Xinjiang is likely to remain part of China for the indefinite future […] Most Uyghurs with whom I have discussed this issue, both in Xinjiang and outside, recognise this reality. Though few are happy about it, some see it as preferable to most realistic alternatives' (Mackerras 2001: 302). Similarly, in Yee's study of ethnic relations, 35.8 per cent of respondents 'strongly agreed' that separation would be harmful, while another 61.6 per cent either 'agreed' or replied that it was 'hard to tell' (this latter reflecting the political sensitivity of the question) (2003: 438-439). Taken at face value, these data suggest that, currently, more Uyghurs see secession as potentially damaging than see it as helpful. At the very least, they show that many do not believe that it makes sense to pursue secession at the current time. However, as Mackerras's above statement makes clear, acceptance of current realities is not the same thing as abandoning one's dream of independence. As individuals, many Uyghurs continue to hold on to the possibility that the Chinese empire must one day fall. As Häsän, a secondary school teacher, pointed out in 2002, no empire has lasted forever, citing the dissolution of the Abbasid (1258), Ottoman (1918), Spanish (1950s), British (1960s), Russian (1991) and Yugoslav (1992) empires.

Yet despite feeling that federal governance was the best political option Xinjiang could hope for at the current time, few respondents expected this to come to fruition any time soon. For one thing, there are considerable differences between the former Soviet Union and the PRC in terms of numerical strength of the dominant ethnic group (Hann 1991: 225). At the time of dissolution of the USSR in 1991, ethnic Russians comprised only 50 per cent of the Soviet population, in stark contrast to the latest figures for China which confirm an ethnic Han proportion of 91.51 per cent. Furthermore, as Shökhrät noted in 2004, although Chinese leaders have apparently considered the option of federation, especially with reference to Taiwan, they presently rule it out, ostensibly on the grounds that patriotic feeling among Han peasants would prevent the majority from accepting that political solution. Other respondents listened open-mouthed in 2002 as I described the political process of devolution then taking place in the UK. Aliyä, a *minkaomin* health professional in her thirties, and Häsän, a secondary school teacher, responded indignantly: 'And yet the Chinese

government is doing the opposite! Power is more and more centralised, and we have less and less control'. In the same year, a male bank employee in his thirties in Ghulja had scoffed: 'The UK may have introduced devolution for Northern Ireland, Wales and Scotland, but don't forget that the English have a *brain* [*kalisi bar*]! The Hans will never let us go, at least not for a very, very long time.' Each ideally wished to see something resembling devolution for Xinjiang, including the decentralisation of government, the introduction of independent regional elections and policy-making structures, the reinstatement of local languages and the revival of literature departments working in those languages, and genuine freedom of religion. These points are also mentioned by Han dissident author, Wang Lixiong, in his 2010 call for the introduction of a high degree of autonomy for Xinjiang and Tibet within the framework of a Chinese federation (Veg 2009b).

# *Policies towards Conflict Resolution*

At the end of Chapter 1, I asked whether it was sustainable to bring Uyghurs limited economic advancement without granting them equal empowerment. Three CCP-approved objectives published in the *People's Daily* in August 1980 had laid out the new direction advocated by Deng Xiaoping, Zhao Ziyang and (especially) Hu Yaobang as follows:


Yet by 1990 these objectives had sadly been repudiated by Party hardliners, who identified them as the cause of a spate of protests occurring at the end of the eighties and start of the nineties. While the current Party leadership recognises the dangers of contemporary perceptions of socio-economic inequalities (it frequently stresses the need to promote economic development in order to promote social stability), thus far it has taken little effective action to address the problem. Instead, it has sought through the Great Western Development campaign to speed the process of Han occupation of the land, and forcible linguistic and cultural assimilation of non-Han peoples. Ironically, the strategy echoes the one-time conviction of the Chinese nationalist party (Guomindang), as outlined by Lucien Pye:

The confidence of Chinese nationalist officials in the all enveloping powers of assimilation was fully the match of the faith of European colonialists […] or the once undaunted confidence of Americans […] they were acting with the not surprising expectation of those who are better off materially that those who are less well off will welcome change. Evidence of resistance to Han culture was ignored […] simply more time and more resources would be necessary to accomplish the inevitable (1975: 492-3).

That the current leadership continues to make the same mistake, believing that improved material conditions can somehow substitute for social justice and equality, is wholly unfortunate. The continued use of (military) stick and (economic) carrot to achieve regional stability and development (cf. Shichor 1994: 81) will not work unless carrots of equal size are offered to all. Economic development *per se* will change nothing so long as Uyghurs perceive development to be unequal.

It has been noted that the Chinese government considered its policy of cultural accommodation in the 1980s to have been the *cause* of ethnic unrest rather than a solution to it (Dwyer 2005: xi). In fact, as Dwyer eloquently argued, the CCP began to lose the 'tepid' Uyghur acceptance of its rule only after it began its hard-line campaign against religion in 1996 (Dwyer 2005: 4). This approach has 'shifted the line so dramatically […] that many peaceful people cannot help finding themselves on the wrong side' (Bovingdon 2010: 88). On the other hand, if Chinese leaders only have the courage to abandon their failing hard-line policies and allow a truly progressive society to develop in Xinjiang, in which diverse groups maintain cultural distinctiveness while approaching a degree of ethnic equality, it has the opportunity to win the Northwest once and for all. Given that former colonies which adopted, for example, English, French, or Christianity, became not less but more determined to break with the coloniser (Tam 2011: 356), the forcible mainstreaming of Han language and culture seems unlikely to succeed in bringing about Uyghur docility. Rather, as Wang Lixiong argues, it may simply push the Xinjiang problem into a phase of 'Palestinisation' (Veg 2009a).

Aggressive behaviour between ethnic groups is ultimately grounded in fear: 'fear of losing economically, materially, reproductively' (Hartley 2010: 235). The ethnic riots which occurred in Ürümchi in July 2009 highlighted the urgent need for policy change, and should have been viewed as a wakeup call. In a climate increasingly characterised by mutually dehumanising ethnic stereotypes, reinforced patterns of ethnic segregation and endogamy (known to preserve and enhance the probability of violent conflict, see

Chirot and McCauley 2006: 118), there is a danger that society may slip into long-term, chronic unrest, for which there are horrifying precedents. In the former Yugoslavia, 'acceptance of exaggeration [of negative ethnic traits] […] fuelled ethnic polarization and socially legitimised local vendettas' (Oberschall 2000, cited in Hartley 2010: 230), culminating in unspeakable acts such as the Bosnian genocide of 1995. In the early 2000s, schools remained segregated for Croat and Serb students in Vukovar, an area formerly known for its high rate of ethnic intermarriage, while young people on both sides remained deeply affected by fear and prejudice (Freedman *et al*, cited in Hartley 2010: 231).

In 1996, I had asked one respondent in Khotän whether she thought that religio-cultural differences alone could have led to violent ethnic conflict between Uyghur and Han people. She replied firmly in the negative:


Thus, until relatively recently, it had been possible to *manage* religious and cultural differences in such a way that verbal and physical conflicts were normally avoided. In the wake of increased social marginalisation, inequality and injustice, this appears no longer to be an option. To reverse this social deterioration, the central and local governments would be wise to revisit cultural, economic and (in the long term) political policies in Xinjiang and effect a thorough overhaul. The first place to look must be migration policy, since it is Han in-migration and the changed regional population composition which has led to the problems listed in Chapter 1 (Inequalities). While it is common knowledge that a majority—if not all—Uyghurs recognise a multiplicity of ill-effects linked to this policy, it is important to note that local leaders –Uyghur and Han—had also objected to the policy of 'unlimited migration' during the 1990s, and that local Han residents are themselves opposed to a further influx (Sautman 1998: 102). In this context, Mackerras's argument that further in-migration should be curtailed on the grounds of reducing resource competition and easing population pressure seems eminently sensible (2005: 16).

The second policy area needing urgent attention is language and education. Instead of assuming that Uyghur-medium education fosters 'splittism'

and subversion, the state should consider reintroducing a progressive language policy, which allows for equal recognition of diverse languages and cultures in the region—in other words, a *truly* bilingual or multilingual education policy. Only by doing so can it dispel Uyghurs' perception that the state is attempting to systematically exterminate their mother tongue (cf. Dwyer 2005: 39). Once Uyghur-medium schools have been re-established, it should then strive to ensure that these receive equal funding and resources compared with Chinese-medium schools. As Benson tentatively suggests, a portion of the income from the sale of the region's oil and gas might usefully be earmarked for this purpose (2004: 206).

Third, stringent legislation must be introduced and enforced to prevent ethnic discrimination in the region's labour market, even if this means the initiation of further preferential policies, such as a requirement for private enterprises to hire a certain quota of minority nationality employees. Tellingly, all three of these areas have been identified by Han anthropologist, Ma Rong, who pinpoints reform of the economic management system, the migration system and the educational system as the basic minimum required to ease local ethnic conflicts (2003: 122).

In new interviews conducted with a long-term Uyghur respondent in 2013, I learned to the contrary that policy has changed for the worse since 2010. As published in the Chinese news media, cities in Xinjiang have been twinned with—and are now subsidised by—east coast cities such as Shanghai, Guangzhou, and so on. In practice, however, this has not translated into a redistribution of wealth for the benefit of impoverished Uyghurs. Rather, it has led to the establishment of branches in Xinjiang of the core Chinese national banks, coupled with the construction of new buildings and shopping centres. In my respondent's view, these facilities are obviously intended for the use of Han settlers, since Uyghur farmers cannot afford to patronise them. While new jobs have been created, these target Han newcomers in the south and not Uyghur locals, so that the redistributed money is 'being poured straight into Han pockets'. The XPCC is set to play a key role in the further opening up of Xinjiang's poorest southern regions via a new division of the Xinjiang *bingtuan*, established in the countryside surrounding Khotän. The plan is to divert water from the Yäkän river (a headstream of the Tarim river), the Irtysh river (source in the Altay mountains) and the Indus River (source on the Tibetan plateau) to support the new settlement. Given that the problem of competition over water is already acute in Xinjiang, these diversions are bound to create further

shortages (although these will not affect members of the *bingtuan*) and new conflicts.

My respondent also confirmed in 2013 that Han settlers in south Xinjiang are now permitted—and encouraged—to bear two children rather than the one outlined in official family planning policy, on the grounds that there are 'too few' Hans in this area. At present, the status of this new policy remains *neibu* (restricted to government circles, and unpublished). According to both this source and a group of Uyghur migrant workers I interviewed in Sichuan province in 2013, new Han migrants are buying up tracts of land from local Uyghurs in Khotän, and compensating them with a monthly 'allowance' of just 250 yuan per family. At the same time, the influx has caused the market price of mutton to double, from about 35 yuan per kilo in 2010 to about 70 yuan per kilo in Ürümchi and 75 yuan per kilo in Qäshqär in 2013.

A recent study on identities in post-independence Kazakhstan produced the interesting finding that Kazakhs living in the North—a traditional Russian settlement area—identify more closely with ethnic Russians than they do with either Kyrgyz or Uzbeks, despite the fact that the latter are fellow Turkic Muslims (Faranda and Nolle 2011: 627, 634, 639). This surprising finding could be explained by a number of factors. First, it could result from the fact that Kazakhs in this area are among the most 'Russified' of the Central Asian, Turkic-Islamic groups, being largely Russian-speaking and substantially acculturated to Russian culture. Second, it could be because Russians living long-term in northern Kazakhstan have acculturated to, respect, and (to an extent) understand local practices. Third, it could result from an inherent admiration among Russified Kazakhs for certain aspects of Russian modernity, an admiration that they now have less reason to hide, since they are the titular group in their own country and enjoy the right to form their own government policies. Might such a scenario evolve in Xinjiang if it were allowed to develop into a truly autonomous part of a federal China? Either way, a long-term solution to the Xinjiang problem will require a *truly joint effort* by intellectuals, leaders and legislators hailing from *all* ethnic groups in Xinjiang (Li Yuhui 2009: 15) [my emphasis]. Only by enjoying equal participation can Uyghurs begin to believe that they have a genuine stake in forming part of a multi-ethnic Chinese nation.

426 bibliography


Banister, Judith (1987) *China's Changing Population*. Stanford: Stanford University Press.

Baranovitch, Nimrod (2001) 'Between Alterity and Identity: New Voices of Minority People in China.' *Modern China* 27 (3): 359-401.




Bovingdon, Gardner (2010) *The Uyghurs: Strangers in Their Own Land*. New York: Columbia University Press.


Cerroni-Long, E.L. (1984) 'Marrying Out: Socio-cultural and Psychological Implications of Intermarriage.' *Journal of Comparative Family Studies* XVI (1): 25-46.



Dashefsky, Arnold ed. (1976) *Ethnic Identity in Society*. Chicago: Rand McNally.



Erkin, Adilä (2009) 'Locally Modern, Globally Uyghur: Geography, Identity and Consumer Culture in Contemporary Xinjiang. *Central Asian Survey* 28(4): 417–28.

Esposito, Bruce J. (1977) 'The Militiamen of Sinkiang.' *Asia Quarterly* 2: 163-172.


Esposito, John L. (1991, 3rd ed.) *Islam and Politics*. Syracuse: Syracuse University Press.


Esposito, John L. (2003) *The Oxford Dictionary of Islam.* Oxford: Oxford University Press.











Strauss, Neil (1999) 'In a Far-Flung Corner of China, A Folk Star.' *New York Times*. 7 February.


*terns, and Perceptions of Cross-Cultural and Interracial Marriage*. Lanham: University Press of America, 147-154.


*The Epoch Times* (2007) 'Hong Kong Reports Mainland Chinese Eating Infants', 29th March. http://www.theepochtimes.com/news/7-3-29/53482.html (accessed 16 February 2012). *The Guardian* (1997) 'China's troubled frontiers.' 11 February.



Wade, Nicholas (2010) 'A Host of Mummies, A Forest of Secrets.' *New York Times*. 15 March.



#### *Discography*

Alim, Ömärjan. *Pärwayim Peläk* [Why Should I Care?]. Censored: no publication details.


### INDEX

'1995 theories' 9, 274 *see also* civil unrest, Ghulja disturbances (1995) Abdulla, Arken 179n12, 208, 210, 406 Aceh 72, 274n52, 275n54, 277n59, 393 affirmative action policy 39, 374 agriculture 24n3, 25, 38, 41, 45-48, 53, 57, 61-64, 67, 69, 75, 92, 95, 171, 413 Al-Azhar University (Egypt) 251 Algeria 77, 173, 260, 287n72, 317, 346 *see also* North Africa Alim, Ömärjan 185, 188, 190-196, 202, 204- 206, 209-210, 216-217, 221-229, 274 Aqsu xv, xix, 41-42, 57, 90, 142-144, 156-158, 164, 237, 419 assassinations 10, 17, 20, 202, 231, 412 Atush, armed conflict in 10 barbarians 85, 95, 126, 295 Beauty of Loulan, the 195 Beg, Yaqub xiiin2, 16, 18, 101, 309 birth control policy 134, 142-145, 197, 413, 425 birth rate *see* population trends Bortala 26, 45 boundary/boundaries construction and maintenance 12, 129, 131-133, 136, 167, 184, 201n46, 230, 294, 312, 319, 326, 349, 408 crossing and transgression 130, 163, 296, 302, 306, 313, 341, 370, 381, 397, 408, 416 ethnic segregation 10, 174, 296, 366, 397, 398 food 153-154, 168 linguistic 135-136, 139 political 21 psychological 8 religious 87, 141, 158, 235, 278, 288-289 social 155, 159-163 spatial 147-150 symbolic 12, 134-5, 146, 151, 171, 235 Buddhism 3, 5, 94, 96, 101, 131, 239, 289, 320, 324, 329 *see also* religious

cadre-people relations 20, 76, 87, 197, 199-203 Central Asia Chinese state policy 32, 288 comparison between Uyghurs and xiii, 4, 9, 13-14, 31, 125, 132n2, 314, 360n3, 384n11, 387, 389, 403, 405, 417-419 effects of independence 20, 169, 245, 410-411 local time 140 music 161, 207-208, 371 religion 246n8, 248n10, 262, 290-292, 386 Soviet period 316, 374 trade 64, 245-246, 269n43 Uyghur exiles 252n17 Central Institute for Nationalities (*Zhongyang minzu xueyuan*) 34, 37, 71, 197, 199, 200 Chechnya 262, 275, 290, 382, 415 China National Petroleum Corps (CNPC) 64-65 Chinese civilising project 6-7, 8-9, 84, 121, 132, 277, 379 federation 22, 70, 74, 416-421, 425 human rights record 232, 253 ,275n55 national unity xv, 11, 19, 36n19, 77, 128, 165n15, 175, 203, 228, 317, 340, 394, 400- 402, 421 nation-state 1-6, 8, 9, 19, 23, 60, 74, 77, 105, 117, 131, 175, 185, 189, 205, 229- 230, 254, 260, 267, 278, 291, 334-335, 367, 413 official discourse 2, 19, 28, 50, 76, 87, 152, 165, 254, 339 Chinese Communist Party (CCP) 27, 70, 71-2, 85, 86-87, 200, 354 Chinese Muslims *see* Hui civil and martial virtues (*wen-wu)* 94, 126 civil unrest 1980s demonstrations 9n6, 19, 142, 409 Andijan incident (2003) 290 Baren riots of 1988 xiv, 9n6, 19, 72, 89, 417

index 445

civil unrest (cont.) demonstration against religious repression xivn4 Ghulja disturbances (1995) xvn5, 11-12, 20, 89, 117, 231, 237, 260, 265, 270, 274 riots in India 51 Ürümchi riots (2007) xiv-xv, xx, 76, 90, 128, 179, 203-204, 388, 397, 422 *see also* terrorism; resistance, violent Classified Document No. 175 11-12, 12n12 class, social 14, 58, 76, 94, 96-97, 194-195, 223-224, 260, 266, 271n46, 345, 347, 393 collaborators mistrust of Hui as 337 mistrust of *minkaohan* as 367, 403 indigenous officials as 10, 105, 175, 185, 196, 200, 204, 231, 374 state informers 196, 205, 215-216, 231, 251, 270, 292n77 *see also* pollution, concept of; stereotypes, parasitism/insect-based colonialism 50, 197, 260, 275n54, 317, 323- 324 commensality 152-153, 328-329, 159 Commonwealth of Independent States (CIS) xiv, 9 Communist Youth League 71n42, 181 community pressure 83-84, 111, 158, 215, 244, 249, 283, 296, 304, 305-312, 322, 329, 343, 396, 398 confiscation of land 61, 63 Confucianism 36, 84-85, 96n5, 99, 125-126, 151, 162, 230 corruption 50, 201, 224, 290 cosmopolitanism 13, 295, 349, 363, 365- 366, 376, 380, 382, 388, 395, 402, 405, 406, 414, 416 cross-border trade 57, 58, 67, 245 Cui Jian 185 cultural alienation 267, 276, 309, 409 anomaly and ambiguity 109, 359-360, 380, 403, 405 authenticity 182, 208, 276, 280, 286, 349, 377, 405, 413 autonomy 22, 31 censorship 12, 185, 187n25, 189-190, 192, 193, 196, 205-206, 209, 216-217n59, 221, 228, 235, 261

commonality/ shared practice 8, 164, 166, 170, 294, 334

cultural (cont.) co-optation 100, 175, 184, 188, 230n69 cross-cultural misunderstanding 49, 106, 125 differences/ conflict 7, 8, 11, 92, 97, 98, 113, 130, 131, 134, 147-148, 166, 169-171, 186, 314 , 324, 336, 364, 389, 402, 408, 423 genocide/ extinction 36, 44, 261, 267, 299, 345, 413 'lack' 126, 348, 376 multi-culturalism 125, 155, 383, 386 norms 74, 92, 93, 98, 101, 124, 167, 229, 289, 295, 334, 402, 408 repertoire/ toolkits 7, 124 survival 13, 235, 261, 293, 296, 345, 413 transmission 60, 208, 210-211, 346, 348, 372, 406 transmission, oral 173-174, 190, 191 Cultural Bureau (*Wenhuaju*) 175 Cultural Revolution 15, 25, 37, 72, 87, 116, 137, 155-156, 230, 232, 240n4, 241, 243n5, 265, 323, 350, 380, 412 culture 4, 13, 208, 405, 416 high 223-224, 228-229, 230 Mediterranean 406, 407 Middle Eastern 245, 246n8, 300, 350 Turkish 181n16, 207, 210n54, 381, 289 Daghestan 249, 262, 290 Dai 8-9 Daoism 125, 126 *see also* religious death rate *see* population trends dehumanisation 85, 124-125, 422 demonstrations *see* civil unrest Deng Xiaoping 9, 81, 421 direct action (individual) 10, 22, 91 domestication 28, 260, 261, 277 Döngköwrük xvii, 54, 136, 148, 157, 311, 343 *Dongtu* 263 drug abuse 95, 195-196n40, 216, 225, 249, 281-282, 333, 393 East Turkestan 19, 120, 256, 275n55, 325,

417, 419 aspirations for Islamic State 276n56, 417 East Turkestan Republic (ETR) xiiin2, 17, 37, 408, 414

East Turkestan (cont.) Turkish-Islamic Republic of East Turkestan (TIRET) xiiin2, 17, 414 economy 50-51, 56-58, 75-77 economic development 28, 73, 267, 413, 421-422 infrastructure projects 27n11, 46, 61 modernisation 6, 173, 266, 276, 277-278, 286, 377, 390, 393 private enterprise 46, 166, 201, 246, 269- 271, 374, 385, 424 *see also* equality, socio-economic education bilingual education 7n5, 36-44, 150, 299, 333, 350, 354-360, 366-370, 375-376, 381, 404 empowerment through 42, 215 family/home 11, 60, 243, 360, 372, 375 Jadidism 16, 215, 262, 286 literacy 38, 41-42, 94, 137, 187, 188, 386 marketisation of 39, 42, 268 minority language education 38, 43-44, 349-350, 354-359, 363, 375, 423-424 school mergers 44 state policy towards 36, 37, 43, 44, 424 study abroad 245, 250, 252, 288, 338, 385 university 14, 39-44, 46, 52, 154, 155, 168, 243, 261, 268n41, 269, 273, 385n12 *see also* language elites 6, 39, 77, 95, 106, 286-287, 354 cultural/ intellectual 173, 175, 184 minority 5-6, 200, 297, 403, 419n5 Emet, Tursunjan 12n13, 231-232 Employment 41, 44-55, 127, 190, 373, 397 discrimination 51, 55, 59, 118, 120, 353- 354, 400 minority quota 49-50, 424 redundancy/ under-employment/ unemployment 51, 54, 268-269, 353-354 environment damage to 19, 61, 66, 69-70, 127, 261 exploitation of natural resources 26, 46, 47, 49, 61-70, 120, 122, 127, 171, 177, 324, 340-341, 424 land reclamation 24, 60, 61, 63, 69n40 pipelines 61, 66, 122 equality/ inequality gender 186n23, 273, 294-295, 297 in education 37, 39, 49, 52, 171 Islamic egalitarianism 266, 271-273, 291, 415 Maoist egalitarianism 57, 415

equality/ inequality (cont.) national/ethnic 20, 22, 51, 73, 74-77, 85, 86, 203, 409, 421 redistribution of tax revenues 65 regional 29, 50, 90 socio-economic 19, 46, 47, 55-59, 119, 167, 171, 258-259, 266, 318, 323, 398- 405, 422-425 *see also* economy; gender essentialisation 6, 99, 124, 344, 405 ethnocentricity 155 ethnic accommodation 12, 74, 135, 204, 235, 304, 349, 401, 402, 413, 415 acculturation 12, 13, 14, 294, 314, 344, 365, 368, 371, 377, 380, 388, 402, 404 assimilation 29, 30, 38, 135, 140, 171, 260, 294-296, 325, 345-348, 359, 364-365, 379, 415, 422 dichotomisation 8, 15, 124, 167, 190 differentiation 7, 21, 38n21, 39, 57, 76, 83, 96, 128, 131-132, 187, 278, 334 distinctions 8, 93, 132, 168, 177, 405 diversity 77, 189, 277n59, 422 harmony (*minzu hexie*) 152, 165, 189, 365, 402 incorporation xiii, 3, 8, 29, 91, 105, 131, 132, 318, 359, 368, 394 integration 2n1, 29, 294, 297, 315, 361n4, 365, 377, 404 loyalty/ disloyalty xiv, 8, 17, 189, 196, 198, 202, 204, 216, 227, 262, 319, 368, 369, 374-375, 381, 404, 411 pride/shame 7, 35, 52, 91, 296, 297, 311, 314-315, 317, 318, 330, 347-248, 353, 375 symbiosis 166 *see also* identities; minority nationalities; Uyghur; Uyghur-Han relations ethnicity, relative 3, 8 family planning policy *see* birth control policy

feminisation 2, 340

filial piety 85, 109, 110, 300, 369, 370, 376, 391

'five stages of history' theory 86


GDP 59

gender female emancipation 254n22, 351, 393 'gendered ethnic centre' 340-342 patrilineal descent 316, 325, 333, 342 segregation 161, 164, 342 sex as metaphor for domination 340- 341 subversion of traditional roles 179n11, 295 traditional roles 59, 353 *see also* equality; Islam, gender roles generational changes/ differences 32, 82, 109, 215, 229-230, 237, 264n36, 323, 335, 390, 412 Ghulja 4, 11, 17, 20, 95, 102, 122, 142, 148, 186n23, 216, 222-223, 231, 302, 397, 417 *see also* civil unrest, Ghulja disturbances (1995) globalisation 9n6, 14, 245, 252, 254, 349, 377, 380-382, 416 government incentives 25, 28, 44, 53-54, 61 Great Britain *see* UK Great Western Development strategy (*Xibu dakaifa*) 28, 43, 47, 50, 66, 75, 90, 260, 266, 277, 421 Guomindang (GMD) 17, 27n11, 52n29, 85- 86, 214, 221, 417 *habitus* 124, 133 *see also Relevanzstruktur* Han chauvinism 11, 17, 30, 34, 84-87, 89, 95, 125, 130, 324, 353, 399 *see also* nationalism, Han Heyit, Abdurehim 188, 190, 206-214, 221- 224, 228n66, 228-229, 232, 406 historical memory 300, 306, 316, 317-318, 323, 336 Hong Kong 9, 81, 115-116, 180, 274 hostess industry (*sanpei xiaojie*) 225n63, 304, 322, 328, 341, 393 household registration (*nopus /hukou*) 25, 26, 47, 51 household responsibility contracts 63 Hui 6, 11n11, 16-18, 48, 72, 132n2, 151, 277, 288n74, 299, 325, 417n4 Chinese Muslims 85n1, 247n9, 248-249, 250, 253, 262n34 Han-Hui relations/ comparisons 128, 147, 153, 162, 296, 328 Uyghur-Hui relations/comparisons 3, 39, 92, 139, 140-141, 154-155, 157n12, 256, 298, 322-323, 327, 336-337

humanitarian intervention 20 humour 97, 105, 120, 122-123, 165, 174, 320 Hundred Flowers movement 71-72, 262n34, 325-326 hybridity *see* identities, hybrid identities/ identity alternative 6, 83, 249, 258, 286, 365, 416 ethnonational 3, 4, 5, 7, 9, 84, 91, 92, 101, 124, 128, 131, 132, 135, 174, 196, 216, 294, 318, 346, 368, 408-411, 416, 417 formation 1, 3, 6-7, 44, 132, 186, 196, 277, 293, 332, 343, 350, 352, 375, 380, 404, 405, 410 globalised 14, 349 group 4, 7, 83-84, 117, 341, 363 hybrid xxi, 4, 13, 221, 236, 293, 295, 349- 350, 356, 365, 368, 380, 388, 404, 405, 415-416 identity crisis 274n52, 280, 364 oasis 1, 3, 92, 93, 132, 174, 223, 228, 410, 411 pan-Turkic 32, 33n15, 387, 417 personal 349, 350, 416 symbols of 7, 90, 131 urban youth 389, 401, 402 'we-hood and us-hood' 1, 7-9, 289, 411 *see also* ethnic; Islam; minority nationalities; Uyghur Imin, Bahargül 176-181, 188 independence declining support for 416-418 economic 385 'imagined independence' 81, 408-411 influence of Central Asian independence xiii, xiv Islam and 254n23, 257, 260, 264, 274, 276, 287, 415 pro-independence sentiment 10, 65, 68, 74, 124, 171-172, 174, 229-230, 231, 253 Uyghur independence movements xv, 9, 19-20, 22, 292 *see also* separatism inequality *see* equality Islam/ Islamic abstinence 13, 15, 87, 89, 109-111, 208- 209, 226, 244, 249, 278, 281-283, 285, 328, 391 *bäsh waq*(five daily prayers) 13, 208, 226, 242, 248, 282, 386

Islam/ Islamic (cont.) conversion 3, 5, 101, 152, 169, 289, 305, 324-328, 331, 338-339, 345, 371, 372, 396 Day of Judgement 111, 237, 264, 275, 283- 284, 387 dietary prescriptions 74, 87, 88, 101, 103-105, 123, 127, 130, 133-134, 151-159, 160, 164, 168, 169, 270n45, 280, 305, 319, 321-322, 326, 328, 331, 369, 372, 374, 377, 387-388, 395 divisions within 284, 286 education 41, 235, 248, 250, 251, 268n41, 276, 277, 278n62, 280, 342, 367n6, 374 extremism 90, 236, 246n7, 285, 290, 382, 414 fatalism 143, 168, 216, 352, 371 festivals 98, 109, 164, 192, 285, 305, 337, 399 fundamentalism 237-238, 254, 255, 281, 291 funerary practice 111-112, 161,162, 192, 240, 285, 327 gender roles 98, 126, 241-242, 243, 246, 278, 289, 290, 309, 316, 342, 344, 345- 346, 353, 391 *hajj* (pilgrimage to Mecca) 245, 248-250, 252, 271n47, 283, 288 harm, notion of 110, 112, 283 Islamic state 285n68, 256-257, 417 Islamism 86, 252, 254-258, 263, 266, 268n40, 272, 275, 276n57, 281, 284, 287, 289, 292, 414 Literature 247, 248n10 *mädris* (Islamic schools) 41, 265 merit , earning of (*sawab elish*) 264, 282, 283 mosque construction 237, 238, 246, 271 Muslim identity 3, 32, 168, 249, 252, 254- 7, 260, 287, 288, 291, 342, 372, 414-415 Muslim Rebellion (1863) 17, 18, 414 oppression/ perceived oppression 236, 258, 260, 266, 272-273, 280-281, 288, 290-291, 414-415 radicalism 250, 255, 262, 290-292, 382, 414 reformist 14, 18, 237, 241, 250, 272, 284, 285 renewal 12, 13, 14, 111, 138, 235-238, 241, 245, 248n10, 250, 253, 255, 258-259, 263-269, 271, 274, 276-282, 284, 286- 293, 321, 344, 391, 401, 414-415

Islam/ Islamic (cont.) ritual purity 105n16, 106, 108, 112, 126, 127-128, 154, 192 Salafism 260, 272, 281 Shiism 272 state control of imams xxiiin10, 201- 202, 251-252, 262, 269n42, 286, 292n77 Sufism 13,17, 67, 181n16, 250, 269n43, 281, 284-285, 287n72, 414 traditionalism/ orthodox practice 12, 13, 111, 208, 226, 235-237, 254, 258, 266, 267, 272, 278, 280-286, 290, 326, 371, 391, 414-415 *umma* (global Islamic community) 248, 249, 252, 288, 382 Wahhabism 285, 290 *see also* religious Israel *see* Palestine/ Israel Jadidism *see* education

Kazakhstan 5, 32, 37, 46, 92, 102, 140, 205, 332-5, 417, 425 allegiances with 17, 19, 107-108, 245-246, 388 independence 169, 386n13, 410, 418 Kazakhs in Xinjiang 156, 417, 419 mixed marriage in 295, 298, 310, 318, 327, 335-337 Uyghurs in 101, 186n23, 208, 240n3, 366- 367, 370 Kelimu 175, 180-183, 184, 188 *Khänzu mijäz* (Han disposition) 93, 100- 101, 367 Khotän 28, 41, 53-54, 59, 67, 74, 142, 150, 157, 160n13, 168, 176, 191, 292n77, 417, 424-425 Kyrgyzstan 17, 19, 37, 92, 169, 246, 335-336, 410, 418, 425 language xxii, 30, 32-34, 43, 73, 83, 134-135, 167 code-switching 135, 350, 355, 379 dominance of Chinese 34, 35, 39, 43, 51,

52, 76, 126, 136-139, 155, 261, 269, 277, 331, 332, 373, 384, 388, 397, 401, 409 Law on Regional Autonomy 36


language (cont.) script changes 30-32, 137-138, 187, 210, 385-386 Turkic-Altaic 33, 38 *see also* education life cycle rituals 98, 127, 328, 346, 360 literature 12, 37, 98, 421 historical novel 174, 187, 221, 289 Islamic 246, 248n10, 272, Uyghur-language 31, 33, 137, 261 *mähällä* (neighbourhood) 162, 327, 333 marginality (liminality) 370, 372, 403 marriage between ethnic groups 220, 317, 423 between faiths 322, 324-326, 328 elopement 301, 302, 304, 305, 310, 397 endogamy 12, 134, 146, 167, 237, 294-306, 312, 315, 317, 318, 334-335, 339, 341, 345- 348, 377, 414, 422 exogamy 179, 298-300, 306, 309, 334, 335, 345, 347 hierarchy of intermarriage preferences 146, 296, 318, 334-336, 416 mixed-heritage children 359, 396, 404, 297, 320, 333, 359 opposition to intermarriage 13, 294, 303, 305, 306, 308, 309, 316, 330, 343, 345, 346, 394-397, 403 Uyghur-Han intermarriage 38, 146, 147, 169, 296, 302, 303, 309, 310, 318, 320, 323, 340, 344, 365, 394-397 *mäshräp* 20, 98, 160n13, 196n23, 232 mass media dissemination of political ideas 82, 174, 189-190 dissemination of stereotypes 82, 121 influence 50, 143, 170, 252, 253, 254, 288, 382, 416 Manchu Qing empire 16, 17, 18, 62, 85, 200, 296, 347 Ma Zhongying ('Big Horse') 17 Miao 5-6, 296 migration early Han settlers 24n4, 25, 54, 74, 75, 88, 123, 147, 170, 401, 409 Han in-migration 19, 23-29, 45, 49, 61, 67, 69n40, 148, 260, 389, 408, 423 rural-urban 46, 124, 190, 277, 286 temporary/ 'floating' migrants 24, 25, 26, 28, 45, 49, 240 voluntary migration 23, 24n4, 25, 26, 46, 54, 118

*see also* population trends

*minkaohan* attitudes of other Uyghurs to 154, 351, 366-369, 371-373, 374-375, 391, 400, 403, 405, 413 attitudes to Han 28, 319, 364, 371, 373, 376, 378-379, 403 attitudes to intermarriage 146, 303-305, 312-314, 321, 344, 353, 354, 370, 395, 396, 405 comparison with other Uyghurs 12, 39- 40, 44, 52, 54, 58, 141, 146, 150, 154, 235, 349, 351, 359, 364, 366-370, 371-372, 400, 401, 411 dual/ hybrid identity 349, 360, 363-364, 365, 380, 391, 402-403, 404-405, 416 negative self-identity (internalised oppression) 349, 350, 352, 357, 360, 362, 374, 404-405 language use 135-137, 332, 350-351, 358, 361, 362, 373, 376, 379, 382, 390, 403, 404 religion 14, 243-244, 252, 353, 372 *minkaomin* attitudes to Han 319, 344, 378 attitudes to intermarriage 303-305, 319- 322, 326-328, 331-332, 336-339, 343, 344, 395 attitudes to *minkaohan* 351, 367, 369, 371-75, 375, 391, 400, 403, 413 comparison with *minkaohan* 39, 40, 54, 154, 349-350, 354, 359, 366, 368-369, 372, 390 disadvantages 39, 267, 349, 358, 373, 374, 389, 400 language use 135, 379, 382, 385, 386, 390 religion and 14, 326, 371-372, 395 minority nationalities 1-2, 8, 27, 29, 38, 40, 48-51, 69, 72, 73, 99, 394 agency 5, 6, 50, 289, 402 media representation 6, 289, 352 minority cultural events 6, 99, 394 state treatment/ classification of 3, 29, 33, 36, 37, 38-39, 49, 50, 70, 87, 88, 404, 424 state representation of 75, 96, 175 modesty code (*namähräm*) 161 Mongolia xiii, 5n3, 28n11, 29n14, 38n21, 39, 43, 69n40, 70-71, 92, 178, 194, 220, 296, 299, 347, 368, 417 morality 11, 84, 107, 110, 115-116, 216, 224- 225, 277n60, 341, 373 music production and consumption 184, 189-190, 349

*see also* popular song

Muslim

*see* Islam

national enlightenment 205, 213 nationalism xiv, 7, 16, 187-188, 199, 230-231, 260, 286, 418 Chinese 17, 24n2, 66n37, 71, 85, 86, 89, 221, 368, 421-422 musical 189, 190, 230-232, 405 religious 169, 173, 292 Uyghur xv, 9, 11, 17, 18, 76, 81, 84, 86, 123, 131, 165-166, 174, 175, 186, 187, 201, 213, 229-231, 240n3, 252, 261, 274, 289, 373, 397, 408, 418 *see also* Han Chauvinism *Nawruz* 12 nepotism 50, 53, 355 New Towns/ Old Towns 45, 58, 69, 99, 106, 121, 130, 148, 150, 159, 168 *nopus see* household registration North Africa 74, 86, 260, 281, 416 *see also* Algeria Old Towns *see* New Towns/ Old Towns origin, notions of 8, 156n11, 194, 288n74, 380, 383 Pakistan 169n16, 246, 251, 257n27, 278, 292n76, 414, 419 Palestine/Israel 273, 274n52, 275, 283, 291- 292, 342, 382, 415 Pan-Arabism 274n52, 292 People's Liberation Army (PLA) 24, 31, 176, 179, 181, 255, 397 peripheral peoples/ peripheries xiii, 2, 5, 7-9, 13, 24, 72, 76, 77, 81, 84, 86, 100, 106, 132, 173, 175, 294, 296, 340, 365, 368 placentophagia 113-114 political aspirations, ethno-political xxiii, 19, 84, 124, 166, 174, 187, 236, 266, 274-276, 288, 291, 292n76, 408, 411, 415 autonomy 22, 32, 70, 417n3, 425 impotence 195-196n40, 267, 297, 342 international context xxiiin11, 9, 12, 20, 81, 170, 174, 238, 254, 274, 287, 408, 409, 411 martyrdom 206, 227-228 mobilisation 18, 189, 215, 229 representation xxi, 23, 70-74, 417n3

pollution, concept of 88, 101-102, 105-107, 108, 127, 134, 155, 169, 319, 324, 328 *see also* collaborators; stereotypes, parasitism/ insect-based popular song catharsis in 188, 192 domestication of 183 illuminists 173, 184-189, 205 musician-comprador 175-184 new flamenco 208, 406 poetry of witness 189 song and dance troupes (*gewutuan*) 98, 99, 175, 180, 181, 186, 207, 209, 228n66 *see also* music production and consumption; individual artists population trends 23, 26, 28, 38, 75, 135, 148, 150, 167, 261, 330, 418, 423 birth rates 26, 141-142, 143, 145 death rates 55, 56, population pressure 24, 61, 423 *see also* migration poverty 29, 39, 42, 55-59, 68, 76, 90, 107, 171, 285, 290, 413 practice theory 133 PRC constitution (1952) 70 preferential policies 39, 40, 45, 50, 58, 63, 200, 297, 424 primordialism 7-8, 133 pro-democracy movement xiv, 20, 409, 412 propaganda 23, 27, 35, 58, 75, 125, 176, 194, 206, 249, 253n17 purity, concept of 127, 236, 259, 276, 278, 279, 281, 315, 335, 405 Qadir, Räbiyä 204, 232 Qaramay 45, 46, 47, 55, 56, 64, 66, 120, 177 Qäshqär 16, 17, 18, 26, 29, 41, 46, 49, 55-57, 63, 64, 72, 90, 141n6, 142, 150, 156, 168, 206, 223, 237, 241, 248, 251, 272, 283, 285, 292n77, 298, 309, 344, 414, 417, 419 Qumul rebellion 17, 18, 309 Radio Free Asia 253, 275 Rebellion of the Three Regions (*Üch Wilayät Inqilabi*) 17, 214 *Relevanzstruktur* 83, 124, 133

*see also habitus* religious conflict 49

conversion 296, 317, 324-328, 339, 372 religiosity 109, 174, 175, 208-209, 237, 241, 242, 245, 291, 292, 371, 373, 386, 404

religious (cont.) repression xxiiin10, 19, 128, 172, 238, 257, 261-262, 264-265, 290-291 syncretism 250, 284, 285, 386 *taliplar* xivn4, 14, 255, 417, 418 *see also* Buddhism; Daoism; Islam Republic of China (ROC) 85 resistance 13, 14, 17, 23, 36, 87, 135, 145-146, 153, 167, 170, 201, 221, 257, 289, 291, 348 collective 10, 215, 299 non-violent 10 'poetry of' 173, 184 Popular 86, 153 sabotage 10, 412 symbolic 6, 9-13, 16, 81, 131, 171, 174, 235- 236, 240, 256, 258, 260, 263-265, 271, 280, 296, 368, 414 violent 10, 255, 257, 368, 412 *see also* civil unrest; terrorism Russification 247n8, 295, 339, 404, 425 Salars 169, 170n18, 250n13, 270n44, 278n62, 410 Sawut, Osmanjan 210, 211, 222 secularism 14, 18, 41, 244, 255n23, 256, 260, 283, 284, 285-286, 344, 377, 415, 418 separatism 10, 11, 12, 50, 55, 95, 186n23, 229, 237, 247, 248, 255, 286, 291, 400, 409, 412 ethnic 17, 21, 231, 254n23, 76, 88, 202, 232, 284, 346, 414 political 22, 265, 275 territorial 21 *see also* independence Shanghai 25, 26, 45, 66, 114, 122, 142, 240n4, 418, 424 Shanghai Co-operation Organisation (SCO) 288 Shaoguan factory fight *see* civil unrest, Ürümchi riots (2007) shared history 5, 167, 92 sheep mentality 105, 194, 255 Sheng Shicai 11n11, 17, 37 Shihezi 26, 45, 46, 47, 55, 61, 62, 69, 142, 378 sinicisation 13, 24, 32, 86, 261, 267, 283, 324, 344, 346, 360, 367, 373, 375, 381, 388, 403, 415 sino-centrism 84, 85, 131 Sino-Soviet split 31-32, 286

social agency xxii, 6, 7, 131, 167, 252, 380, 394 exclusion 8, 124, 128, 132, 153, 204, 352, 359, 363, 366, 374, 403 interaction xvi, 4, 7, 101, 162, 164, 167, 337, 382, 411 sovereignty 8, 20, 22, 325 Soviet Union See USSR 'splittism' (*fenlie zhuyi*) 11, 185, 192, 201, 209, 231, 237, 289, 346, 423 Spring Festival 168, 183 Stalin, Joseph 71 state mediation 6 state violence 77 steppe Uyghur kingdom 5n3, 85n1, 95, 220 stereotypes 'backwardness' 39, 73, 87, 352, 400 counter-stereotypes 13, 81-82, 84, 90-93, 97, 99, 101, 110, 124, 128, 139, 230 dogs 105, 197, 205 ethnic and national 81, 83, 90, 329, 422 invasion 10, 119-120, 122 parasitism/ insect-based 118-120, 122- 125, 197, 201 social and cultural 82-84, 128, 314 social causality 82, 83, 127 *see also* collaborators; pollution, concept of sterilisation 145, 168 *see also* forced abortion Strike Hard campaigns xiv, 10n7, 11, 20, 237, 247, 255, 397 subaltern scholarship 6 voice 173, 175 Taiwan 9, 71, 77, 81, 183, 330, 420 Tajiks 37, 139, 169, 246, 255, 285n68, 334, 335, 374, 410, 418 Täklimakan highway 26 Taliban 10n6, 251n16, 257 Tang dynasty 85, 295 Taranchis 16, 37 Tarim river 61, 182, 424 Tatars 37, 169, 170n18, 295, 318, 332, 335- 336, 410 territorial assimilation 29, 72, 124, 317 terrorism aftermath of 9/11 50, 256n26, 263n35, 265, 276, 291

terrorism (cont.) Al-Qaeda 382, 414 conflation of Islam with 254 Uyghur/Han reactions to 9/11 113, 258 'war on terror' xiv, 12, 20-21, 238, 414 Tian'anmen Square *see* pro-democracy movement Tianshan (Heavenly)mountains 5n3, 19, 41, 45, 55, 151, 177, 190, 216-217, 246, 311 Tibet xiii, 5, 27, 38n21, 43, 57, 70, 71, 87, 88, 90, 94, 99, 104n14, 298, 345n6, 322, 346, 378, 417, 421 tourism 67, 70, 89, 158, 238, 240, 246 Tungans *see* Hui Turkmenistan 169, 170n18, 295, 339, 410 Turpan 3, 5n3, 16, 18, 32, 33, 40, 64, 67, 72-73, 98, 101, 113, 142, 159, 178, 184, 213, 268n41, 282, 298, 302, 309, 328, 337 Twelve Muqam (*On Ikki Muqam*) 207, 223, 371 UK 64, 274-275, 338, 380, 420-421 urbanisation 69, 266, 389-390, 392, 393 urban-rural divides 55, 90, 411 Ürümchi 12, 15, 17, 25, 33, 36-38, 40-41, 45- 47, 52, 54-56, 59-60, 69, 90, 99, 104, 106, 109, 121, 130, 135-140, 142, 148, 152, 157- 166, 190, 194, 209-210, 236-244, 257, 264, 271-272, 280, 299, 310, 331, 333, 351, 355, 358-359, 368, 377, 379, 381, 384, 388, 395-397, 402, 405, 416, 418-419 *see also* civil unrest, Ürümchi riots (2007) USSR xiii, xiv, 5, 9, 17, 18, 20, 71, 92, 169, 248, 249, 262, 290, 295, 334, 339, 404, 410, 417n3, 420 US 20-21, 70-71, 81, 91, 135, 274-275, 276n56, 294, 300, 301, 306-306, 314, 317, 318, 325, 345, 354 Uyghur activists 10, 117, 171, 237, 412 ethnonym 3-4, 92 hospitality 159, 178, 390, 391, 392 middle class 58, 76, 271n46, 345, 383, 388, 393 progressives 18, 19, 358, 405 youth 16, 35, 51, 84, 109-110, 113, 138, 146, 153, 179n12, 195n40, 275, 281-282, 300, 304, 312, 322, 338, 343, 347, 349-407, 416

Uyghur (cont.) *see also* ethnic; identity; Uyghur-Han relations Uyghur, Abdukhaliq 213 Uyghur-Han relations competition 39, 41, 43, 46, 62, 91, 122, 124, 127-128, 130, 136, 148, 166, 167, 261, 308, 319, 323, 348, 408, 423 host-guest dynamic 178, 180, 193, 206- 207, 221-222 hostility xxii, 1, 3, 11, 82, 88, 125, 134 inferiority complex 82, 90, 353, 357, 374, 380 internalised oppression 349, 350, 352, 353, 374, 380, 404 juvenilisation/ sibling dynamic 2, 174, 402 master-servant dynamic 279 positive relationships 74, 170, 344, 363, 393 prejudice/ racism 50, 82, 86, 147, 170-171, 306, 307, 315, 319, 324, 340, 363, 367 state representations of 2, 340 *see also* ethnic; identity; Uyghur Uyghuristan 20, 417-419 Uzbekistan 18, 37, 101n9, 139, 169, 189, 197, 227, 246, 255, 290, 318, 327, 332, 334, 335-336, 388, 410, 408-418, 425 Wang Zhen 151 *wen-wu see* civil and martial virtues *Xinjiangban* (Xinjiang Class Project) 361, 365, 389, 405 Xinjiang Production and Construction Corps (XPCC) 24, 26, 45-47, 60-64, 424

*see also* environment, land reclamation Xinjiang time 140, 323, 351, 376 Xinjiang University 37, 38, 40, 44, 52, 68n38

Yang Zengxin 11n11, 43, 173 Yunnan 140, 147

Zhangqi 45 Zhou Enlai 86 Zuo Zongtang, General 36