The Art of Symbolic Resistance

Front matter

BRILL’S INNER ASIAN LIBRARY

UYGHUR IDENTITIES AND UYGHUR-HAN RELATIONS

IN CONTEMPORARY XINJIANG

Contents

Acknowledgements... 334 isansa rinane a nae i a is xiii

Preface ser oeras aa a a a a a aa XV PART ONE

INtrOdUCtiON:. insciescctexdperenewrrs T, 1 1.2 Language Policy: .s.ciaviveveitebavelertd nadie ninia 29 4:9 EduicatiOnieciiescds songs aa a aa acerca gags 36 1:4 EMployMent ...cc0ccdcacaiercacnaeneeeres tenes danmasenads 44

Part Two Sounds Of Discontent

2.4 Common Heritage: Foundations of Counter-stereotypes.... 92 Conclusion: Talk Breeds Conflict (or the Power of Oral

Complaint) Aciecidscuedadenteducaacktiin aan and 123

Making Culture Matter: Symbolic, Spatial, and Social Boundaries 130 viii

CONTENTS

$.9 Spatial Boundaries i. iciocscssiadacnisadeenesapdaniabees 147 9.4. Social Boundaries .:324%.Adeaka ites ttk EELEE EEIE EREA 159 3.5 The Consequences of Boundary-crossing: Bus Stories and

Street Fights: ic.40cheiawreeia ah ainisowienedugdaeieteeks 163 Illuminists: Popular Song and the Waking of the Uyghur Nation 173

4.2 Popular Singers as Illuminists: The Advent of ‘New Folk’ .... 184 4.3 Cassette Culture and Mass Mediation of Political Messages 189

Part Three ‘silence’: After The 1997 Ghulja Disturbances

Introduction EEE E T E 235 5.2 Globalising Forces: Border Trade, Pilgrimage, Study Abroad

5-3 Islam as Symbolic Resistance against Perceived Muslim

OppreSSiO m eiiean iaa a E Gude ade A cal 260 5.6 Islam as a Response to Frustrated Ethno-political

ASpitationS sricsriiieenrii Etes oiTa nse vad REAA 274 5.9 Internal Divisions within the Uyghur Muslim Population ... 284

CONTENTS ix

6.4 Perceived Barriers: Racial, Cultural, Religious, Historical,

POV GAM fet a0 ace A E lela huge A laa ale 314

Conclusion: Honour and Shame in a Context of National

IMPOtENCe rssi enp hE EEE T AE E 347

7.1 The First-generation Minkaohan: ‘Internalised Oppression’ 350 7.2 The Second-generation Minkaohan: Product of Urban

Instrumentalisin seers irete eono a N 354 7-3 Minkaohan Self-identity in Contemporary Xinjiang:

‘Ethnic: Anomaliés’?: 4.0 hiadiivs dav edeies daevtes ieee sans 359 7.5 The Han Wannabe: Prototype for the Third-generation

Minkähän is cabs ado a ER E U aes 375 7-6 Cosmopolitan Multilingualism: Uyghur Youth as ‘World

CIZEDS oo cesceeaed e paran E E a E Ratan week one 380

Conclusion: From Ethnic Anomaly’ to Subaltern Cosmopolitan 402

ConclüsionSs snaren a aaa a E ATE 408

E PEE E E E E E E EE E, 445

LIST OF ILLUSTRATIONS

Wood-turners at the handicrafts alley, Qäshqär Old Town (1996) Ever decreasing Uyghur fonts, Hongshan, central Uriimchi (2004) iitos i ccGiserseGs Heasks shade eee ada deena eoSans Supplicant at Kokhmarim mazar (Snake Mountain shrine), Khotiin (1996) siccicevetsiie rave teria ve be eente dees dee dare ces Threshing ground, Bay, Aqsu (1996) «0.60... cece eee eee eee Rural community around a 500 year-old walnut tree, Khotan (T096 ) 5 bes. c.s. ors ears Sstane E a t Saa ESARET

. “Han and minorities breathe the same air, share the same destiny, and their hearts are joined as one”, Ürümchi (1996)...

. Abdurehim Heyit, dutar virtuoso, Déngkéwriik, Ürümchi (2002) . Friday prayer at the new Yanan Road mosque, Ürümchi (2004) . Female modesty as national honour, Yan’an Road, Ürümchi

. Bilingual pop fusion—Poster of Uyghur ‘Guitar King’, Arken

. Convoy of People’s Liberation Army trucks, rural Aqsu (1996)..

ACKNOWLEDGEMENTS

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. Minawar Abdurehim at the Xinjiang Academy of Social Sciences, Uriimchi, 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 Ali, 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 Cesaro, 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 xiv ACKNOWLEDGEMENTS

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

I first came across the Uyghurs in 1988 while studying Chinese in Beijing. Across the road from the Beijing Foreign Studies University! (then Beijing waiguoyu xueyuan TESS NENE t) 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’ (Bockman 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.? Although researchers gained unprecedented access to the

1 This institution changed its name in 1996 to Beijing waiguoyu daxue (JESS NEMS KF).

2 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 ‘Qashqar 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 {F}, the collapse of Marxist-Leninist parties in Eastern Europe in the same year,

VT 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.? 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 uth 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 Uriimchi protest against state handling of the centre ‘Kashgar’ throughout the book, since it is closest to the original pronunciation in Uyghur.

3 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).

+ Foran 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.’ Anti-riot 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.> 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 Uriimchi (nine months), and in the southern oases of Kucha, Aqsu, Qäshqär and Khotän (three months). My research questions at that time were:

e How far have regional and social group identities among the Uyghurs been eclipsed by a broad-based national (including separatist) identity since independence of the Central Asian states in 1991?

e What is the nature of contemporary Uyghur-Han relations?

On subsequent trips in 2002 (Uriimchi 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:

e To what extent have stereotypes and boundaries deployed by Uyghurs in the mid-1990s remained in place?

e Which factors have caused some Uyghurs to turn back to Islam? Which factors have led others to accommodate—albeit to different degrees and in different contexts—to Han culture and/or state hegemony?

e How do Uyghur youth negotiate hybrid and multiple identities in their everyday lives?

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

5 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 (20114).

xviii PREFACE kexueyuan YIEE AF kit) 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),® 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.

6 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:

1. Chinese was—and remains—the main language spoken in most districts of Ürümchi, with the exception of the Uyghur-dominated Döngköwrük (Ch. Erdaoqiao) area in the south-east of the city;

2. Iwas based at an institution located in north-west Ürümchi, the opposite corner of the city in relation to Dongkéwriik;

3. Uyghur language textbooks were usually aimed at Chinese native speakers and produced in the Chinese language, meaning that non- Chinese students of Uyghur had to learn the third language through his/her second.”

Changes in local dialect between oases brought new problems, although interrogation of local vocabulary sometimes produced synonyms learned in Uriimchi; 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. 201: 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.

7 Tlearned 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 Engeszeth (2002).

XX PREFACE

Figure1: 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-a-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:

1. Making my personal identity and values explicit during data collection and analysis in a process of ‘active reflexivity’ (Mason 2002: 7).

2. Suspending for analytic purposes the views that I necessarily had to adopt in interaction with respondents.

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

8 Ätläs (Khan dtlds 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: 12). 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 Uriimchi 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-

XXiv PREFACE deng) and ‘nationality unity’ (minzu tuanjie).° 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-1ggos 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-iggos: 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

9 Even after the Ürümchi riots of July 2009, which demonstrated serious social cleavages along ethnic lines, Xinjiang Regional Chairman, Nur Bakri, 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’: Omarjan 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.