AsianEthnicity, Volume 2, Number 2, September 2001

Towarda Synthesis of the Theories of Peripheral : AComparative Study of ’s and

YUAN-KANG WANG* (Universityof Chicago)

Thecentral question of this article is: whyis peripheral nationalism virtually non-existent inGuangdong but has long been a problemin Xinjiang? Existing theories are inadequate foranswering this question. They are tailored to either wealthy or poor regions, but not both.They also fail to explain the absence of nationalism in Guangdong and its presence inXinjiang. In this article, I proposea modelthat can explain peripheral nationalism in bothwealthy and poor regions. The model incorporates two key variables, elite status and nationalidentity. SpeciŽ cally, I arguethat if the elites from the peripheral region can hold importantpositions in the central government and if the peripheral residents identify themselveswith the core , the probability of nationalist movements will be low. On theother hand, if their elites are excluded from high ofŽ ce in the central government and iftheirpeople do notidentity themselves with the core nation, the probability of nationalist movementswill be high. The model succeeds in explaining the cases of Xinjiang and Guangdong.

Introduction At Žrst glance,China’ s XinjiangUighur Autonomous Region and Guangdong province lookrather different. In terms of economic development, Xinjiang is relatively backward, whileGuangdong is relatively advanced. More than half of the population in Xinjiang consistsof Muslim minority nationalities, while the majority of the population in Guang- dongis considered Han Chinese. However, both also have some similarities. They are locatedin the peripheral regions of China and far awayfrom centralpolitical leadership. TheMuslim nationalities in Xinjiang do notspeak the ofŽ cial language Mandarin, while the populationin Guangdong speaks Cantonese, a languagequite different from Mandarinin spokenform. Bothhave their respective regional identity. In short, these two regions have someethnic potential, though apparently Guangdong has less. Secessionismhas long been a prominentissue in Xinjiang. Uighur nationalists seek to establishan independent state free from HanChinese domination by resorting to riots and rebellions.In contrast, such secessionism is unthinkable in Guangdong. Cantonese have longconsidered themselves Chinese and Guangdong has always been part of China. Why isperipheralnationalism virtually non-existent in Guangdong but has long been a problem inXinjiang? Existingtheories of peripheral nationalism are inadequate for answeringthis question. Oneversion was developedto explain secessionism in economically dynamic peripheries,

*IthankDavid Laitin, two anonymous reviewers, andparticipants at the48th Annual Meeting of the Midwest Conferenceon Asian Affairs (1999)in the United States, for their thoughtful comments andsuggestions.

ISSN1463-1369 print; 1469-2953 online/ 01/020177-19 Ó 2001Taylor & FrancisLtd DOI: 10.1080/14631360120058857 178 Yuan-kangWang suchas the Basque Country in Spain, while the other was designedto explain nationalist movementsin economically backward regions, such as the Celts in Britain. They are tailoredeither to wealthy or poor regions, but not both. This restriction to a greatextent reducesthe applicability of both theories. Inorder to Ž llthis theoretical gap, this article proposes a newmodel that can explain nationalistmovements in both wealthy and poor peripheries. SpeciŽ cally, the model incorporatestwo key variables— national identity and elite status— in explaining secession- ismin the peripheral region. I arguethat if the national identity of the peripheral residents iscongruentwith the core nation, and if elites in theperipheral region can serve high ofŽ ces inthe central government, the probability for secessionismwill be low. On the other hand, ifthe peripheral people identify themselves as distinct from thecore nation, and if their elitescannot hold signiŽ cant positions in the centre, the probability for secessionismwill behigh. I thentest my argument in the cases of Xinjiang and Guangdong. I Žndthat nationalidentity and elite status greatly reduce the nationalist potential in Guangdong,while thelack of two such qualities gives rise to secessionist movements in Xinjiang. Cantonese believethat they are really Chinese, even though their spoken language is quite different. Localelites were ableto make it to the top of the country, such as Sun Yat-sen, as well assome key national Ž guresin the People’ s Republicof China (PRC). However, such hegemonicnotion and elite status were absentamong the Uighurs in Xinjiang. Thenext section reviews existing literature on peripheral nationalism and critiques its shortcomings.Then, I presentmy theory, explicate the two master variables— elite status andnational identity— and explain how they contribute to the emergence of peripheral nationalism.Finally, I testmy argument in a comparativecase study of Xinjiang and Guangdongand discuss its implications.

CurrentTheories ofPeripheral Nationalism Thereare currently two versions of the theory of peripheralnationalism, depending on the developmentlevel of the region in question relative to the political centre. The Ž rst deals witheconomically advanced regions. This version focuses on the interaction between the economicdevelopment of theperipheries and the centre’ s abilityto provide further growth. Accordingto Peter Gourevitch, when there is a lackof congruence between political leadershipand economic dynamism, the regions with ethnic potential are likely to develop peripheralnationalism. This non-congruence may result either from thefaltering of thecore inprovidingeconomic growth or politicalleadership, or from theimprovement in economic positionof the peripheral region relative to thecentre. 1 Inother words, if thepolitical centre isexperiencing economic or political decline or the ethnically distinct region in the peripheriesbecomes economically dynamic, the peripheral region is likely to develop ‘strong,politically relevant nationalism’ . 2 Gourvitch’s theoryfocuses on the economic function the centre plays vis-a`-vis the periphery.However, the logic of the theory is incomplete. The principal problem is that he doesnot establish any objective, independent criteria for regionswith ‘ ethnicpotential’ . He admitsthe difŽ culty of such a dauntingtask, but tries to circumvent it by asserting that ‘circularitycan be avoided by specifying (however arbitrary it may appear) which regions

1Peter A.Gourevitch, Parisand the Provinces: ThePolitics of LocalGovernment Reform in France (University ofCaliforniaPress, Berkeley,CA, 1980),chap. 10. 2Ibid.,p. 203. Synthesisof the Theories of Peripheral Nationalism 179 haveethnic potential and which do not’ . 3 Nevertheless,without any objective, clear and independentcriteria for ethnicdifferences, his theory still tends to be tautological. 4 Evenwithout clear criteria, it wouldbe hardto deny Guangdong’ s ethnicpotential. And aswe shallsee, it has the other ingredient needed for Gourevitch’s theory:economic dynamism.Guangdong’ s economicposition in the PRC beforethe 1978 reform was below average.During 1953– 78, its average annual GDP growthrate was 5.1per cent, 1 percent lowerthan the nation’ s average.This low level of growthwas duein large part to Chinese leadership’s emphasison developing the interior region. 5 After 1978,Guangdong became themajor testing ground for DengXiaoping’ s economicreforms. SpecialEconomic Zones were setup to attract investment. During 1979– 91, its average annual growth rate was 12.6 percent, 3 percent higher than the nation’ s average. 6 By1998, its GDP hadalready been thenation’ s largest,accounting for 10per cent of total national income, and per capita incomeranked Ž fth,after Shanghai, Beijing, Tianjing and Zhejiang province. 7 TheseŽ gures suggestthat Guangdong has become economically dynamic since China’ s reforms. Given itsregional identity and distinct spoken language, Gourevitch’ s theorywould predict peripheralnationalism in Guangdong. However, there is no nationalist movement there. Andas we shallsee later, careers of aspiring elites from Guangdongcould be aimedat the politicalcentre. Thus, we needto Ž ndout other factors that can explain the absence of peripheralnationalism in Guangdong. Thesecond version of peripheral nationalism deals with backward regions. Michael Hechter,Tom Nairn and Ernest Gellner attribute the rise of nationalism to the painful inequalitysuffered by the people of backwardregions resulting from theuneven pattern of development. 8 Wavesof industrialisation and modernisation hit different areas in uneven speed,engendering sharp and painful inequality in standards of living. Egalitarian expec- tationand non-egalitarian reality make the misery of those in peripheral regions seem all themore perennial, and intolerable. If theyare distinctly differentiated from thepeople of moreadvanced regions, such as by skin colour, religion or customs, we arelikely to see nationalistmovements there. 9 Independencebrings the prospect of being in charge of the nation’s ownaffairs andfreeing the region from exploitativerelationship with the centre. 10 Thisversion of peripheralnationalism presupposes industrialisation as the driving force ofnationalism. For Gellner,agrarian societies were composedof laterally insulated communities,while the ruling class was horizontallysegregated. Cultural differentiation

3Ibid.,p. 202. 4See DavidLaitin, ‘ TheNational Uprising in the Soviet Union’ , WorldPolitics ,vol.44, no. 1 (October1991), pp.139– 77, fn. 13. 5Fearingstrategic encirclement, Chinalaunched the ‘ ThirdFront’ program from 1964 to 1971 by moving its industrialbase tointerior regions. See BarryNaughton, ‘ TheThird Front: Defense Industrializationin theChinese Interior’, TheChina Quarterly ,no.115 (September 1988), pp. 351–86. 6DavidS.G. Goodman and Feng Chongyi, ‘ Guangdong:Greater HongKong and the New RegionalistFuture’ , in DavidS.G. Goodman and Gerald Segal(eds), ChinaDeconstructs (Routledge,London, 1994), pp. 182– 3. 7State StatisticalBureau, Zhongguotongji nianjian (ChinaStatistical Yearbook )(Zhongguotongji chubanshe, Beijing,1999), pp. 62, 65. 8MichaelHechter, InternalColonialism: The Celtic Fringein British National Development, 1953– 1966 (Universityof California Press, Berkeley,CA, 1975); Tom Nairn, TheBreak-Up of Britain: Crisis and Neo-Nationalism (NLB,London, 1977); Ernest Gellner, Nationsand Nationalism (CornellUniversity Press, Ithaca, NY, 1983). 9ErnestGellner, Thoughtand Change (Weidenfeldand Nicolson, London, 1964), p. 167;and and Nationalism , pp. 73–4. 10Hechter called thisexploitative relationship ‘ culturaldivision of labor’ . Theadvanced core ‘seeks tostabilize and monopolizeits advantages through policies aiming at theinstitutionalization of theexisting stratiŽ cation system. Itattempts toregulate the allocation of social rolessuch that those roles commonly deŽ ned as havinghigh prestige are reserved forits members. Conversely,individuals from the less advancedgroup are deniedaccess tothese roles’. This‘ culturaldivision of labor… contributesto thedevelopment of distinctiveethnic identiŽ cation in the twogroups’ . Hechter, InternalColonialism , p. 9. 180 Yuan-kangWang prevailsover homogeneity between different communities as well as between different strataof the ruling class. The education of skills was veryspecialised and limited within communities.With the arrival of industrialisation,training became generic, less specialised andmore universally standardised in order to provide the mobility and interchangeablility ofskills needed for perpetualgrowth. The agrarian communities could no longer assume thistask on such a largescale, and the state took over and became the primary form of nationalidentiŽ cation. The nation-state will impose a nationallanguage associated with a highculture to facilitate communication. Seeking a betterlife, the rural folks, who speak adialectquite different from thenational language, migrate to the industrial centre. They learnthe national language to increasetheir job opportunities in the cities to whichthey had migrated.If theycan blend into the centre, without being noticed, and their leaders can get elitepositions there, the rural area will then be incorporated as part of the larger nation-state.However, if the mobility prospects of themigrant rural folks are blocked, even afterthe second generation, this exclusion will make these migrants begin to see higher economicpayoffs for havingtheir own nation-state in which their native dialect will be spoken.11 Thus,the driving force of nationalism in Gellner’ s modelis industrialisation. Xinjianghas always been a backwardregion in China. Separatist movements arose wheneverthe power of the Chinese Empire was onthe decline, when there was no industrialisationor modernisation as we knowit today. By positing a sharpdisjuncture betweenmodern and pre-modern societies, this theory cannot explain secessionist move- mentsthat had been happening in Xinjiang prior to industrialisation. 12 Themovements for independencewere nota modernphenomenon; rather, they have existed long before modernisationtook place in China. Therefore, the existing theories not only exclude possibleforms ofpre-modern nationalism but also fail to explain them. Tosum up, based on current theories, peripheral nationalist movements can occur in bothadvanced and backward regions. The respective mechanism of how they came into beingis very different. In the advanced regions, the inducement can be either the improvementof the economic position of the peripheries or the failure of the centre in providingfurther economic growth, or its political decline. In the backward regions, however,the inducement can be the seemingly everlasting misery suffered by the ethnically distinctpeople, who can no longer tolerate their plight and hope that independence will bringan end to their predicament. Theproblem with these two versions of peripheral nationalism, in addition to their failuresin Guangdong and Xinjiang, is that they are inherently limited to encompassing eitheradvanced or backwardregions, but not both. 13 Gourevitch’s theorycannot be applied toXinjiang, while Hechter’ s, Nairn’ s, and Gellner’ s theorycannot be applied to Guang- dong.Moreover, the cases they studied are mostly capitalist societies in theWest. One must becautiousin generalisingand extrapolating theories derived from Westernsocieties to the restof the world, such as China. 14

11 Gellner, Thoughtand Change ,pp.166– 8; and Nationsand Nationalism . 12Prasenjit Duara, ‘HistoricizingNational Identity, or WhoImagines Whatand When’ , inGeoff Eleyand Ronald G.Suny(eds), BecomingNational (OxfordUniversity Press, New York,1996), pp. 152– 3; AnthonyD. Smith has anexcellent account of this‘ modernists’vs. ‘ perennialist’argument. See Smith,‘ TheOrigins of Nations’, in BecomingNational ,pp.105– 30. 13In a studyof theinteraction between nationalismand immigration, Gershon ShaŽ r combinesthese versionsof nationalismto distinguishbetween hegemonicand corporate . However, ShaŽ r’ s studyfocuses on theinternal immigration in larger states’relatively developed peripheries inhabited by adistinctnationality, and is thusdifferent from this study. See Immigrantsand Nationalists: Ethnic Con ict andAccommodation in Catalonia,the Basque Country, Latvia, and Estonia (State Universityof New YorkPress, Albany,NY, 1995). 14See, forexample, Partha Chatterjee, TheNation and Its Fragments:Colonial and Postcolonial Histories (Princeton UniversityPress, Princeton,NJ, 1993),pp. 1– 13. Synthesisof the Theories of Peripheral Nationalism 181

The Model Iproposea modelof peripheral nationalism based on the interplay of two factors: elite statusin the central government and national identity. By elite status, I meanthe position localelites are able to hold in the political centre. If localelites can serve high ofŽ ce in the centralgovernment, without signiŽ cant discrimination from thosein the political centre withsimilar education and background, 15 thepotential for peripheralnationalist movements islow. On the other hand, if local elites are excluded from highofŽ ce in the central government,regardless of their similar education and background, we wouldexpect to see nationalistmovements in the peripheral region. National identity refers toa collectivesense ofwho we are,what we were, andwhich political community we belongto, and is often associatedwith an historic territory. National identity should be viewed as a uid, contingentand relational concept. 16 Itis constantly shifting over time, sometimes overlap- pingwith other identities. Language, culture, common history, customs, outer appearances andreligion can in uence the process of national identiŽ cation; nonetheless, they are sufŽcient but not necessary conditions for theformation of national identity. If thenational identityof peripheral residents is congruent with that of the political centre residents, we donot expect to see peripheral nationalist movements there. However, if their national identityis incongruent with that of the centre, we arelikely to see nationalist movements. Nationalidentity should be distinguished from ethnicidentity. A nationis usually composedof several ethnic ‘ elements’. 17 For instance,China is often said to be a multiethnicstate, with the Han as the ethnic core, coexisting with other ethnic groups. Althoughthese different groups have their respective ethnic identities, they have a common nationalidentity— Chinese. Thus, a Manchucan have an ethnic identity of ‘ Manchu’but a nationalidentity of ‘Chinese’; anAfrican American can have an ethnic identity of ‘African American’but a nationalidentity of ‘ American’. Whydoes elite status affect the potential of peripheralnationalism? To beginwith, high statusprovides the incentives for localelites to be co-opted into the political establishment atthecentre; 18 theycan use this status to advance their personal prestige and interests, and presumablypromote the welfare oftheirhome people. For theperipheral population, having aneffective voice at the centre ensures that their interests will not be ignored, and creates afeelingof belonging to the whole nation. Onthe other hand, low status or exclusionof localelites from thepower core provides few incentivesto stay within the political establishment, and the tendency to break away willbe high. Local elites see independence as the best way to enhance the well-being of themselvesand their people, and they hope that it willbring about the prestige and interests thatthey deserve but cannot attain in the old regime. The folks, seeing themselves as the under-privilegedpopulation and second-class citizens, are also likely to Ž ndindependence asawayto restore their rightful place, especially when the economy is doingpoorly. They feelthat they have been exploited by the centre and the fact that none of its elites is occupyingany high position in the central government only exacerbates their grievances. Consequently,we arelikely to see nationalist move ments, such as the revival of folk cultureand tradition, and the call for anindependent nation. Nationalidentity also plays a keyrole in nationalist movements. People identify with

15David Laitin called thiskind of elite status‘ mostfavoured lord’ , aterm borrowedfrom the notion of mostfavoured nationin international trade relations.Laitin, ‘ TheNational Uprisings’ , p.143. 16Duara, ‘HistoricizingNational Identity’ , p.151. Also see Peter Sahlins, Boundaries:The Making of France and Spainin thePyrenees (Universityof CaliforniaPress, Berkeley,CA, 1989), pp. 270– 1. 17See AnthonyD. Smith, NationalIdentity (Universityof Nevada Press, Reno,1991), pp. 19– 42. 18Laitin, ‘ TheNational Uprisings in theSoviet Union’ , p.144. 182 Yuan-kangWang

Figure 1. EliteStatus, National Identity, and Peripheral Nationalism. anamedcommunity to distinguish themselves from the‘ Others’. Whetherthis process of identiŽcation is imagined 19 orreal should not obscure the fact that this is how each communityof people distinguishes itself from theothers. As PeterSahlins observes, ‘[n]ationalidentity is asociallyconstructed and continuous process of deŽning “ friend”and “enemy”, alogicalextension of the process of maintaining boundaries between “ us”and “them”within more local communities’ . 20 This‘ identityas opposition’ distinguishes us from the‘ Others’in our daily lives. A distinctnational identity is oftenassociated with the demandfor aseparatepolitical entity. When the people of a peripheralregion identify themselvesas the same nation as that of the centre, we arenot likely to see nationalism therebecause these people see themselves as culturally homogeneous with the core nation. However,if they identify themselves as a distinctnation, the potential for nationalist movementswill be high, because the fact that the rulers of the region belong to a nation otherthan the majority of those being ruled constitutes an intolerable situation for the peripheralnationalists. 21 Bothelite status and national identity must work together in order to generate enough supportfor nationalistmovements. Low elite status by itself is not sufŽ cient to cause peripheralpeople to want an independentstate of theirown. They must also have a different nationalidentity. In other words, national identity is an antecedent condition that enables elitestatus to have an effect on theprobability of peripheralnationalist movements. 22 Figure 1summarisesthe relationship between elite status and national identity. The model makes twomajor predictions. First, if anethnic group in the periphery identiŽ es itself as having thesame national identity as the core nation, and if its elites enjoy high status in the central government,the probability of peripheral nationalist movement would be low (Cell I). Secondly,if anethnicgroup in the periphery has a differentnational identity from thecore nation,and if its elites are excluded from importantpositions in thecentral government, the probabilityof nationalist movement would be high (Cell IV). CellsII andIII depicta situationin which both elite status and national identity point inopposite directions. Cell II refers toa peripheralregion with high elite status but populatedby minority nationalities that do not identify with the dominant nation. In this

19See BenedictAnderson, ImaginedCommunities: Re ections on the Origin and Spread of Nationalism (Verso, New York,1991). 20 Sahlins, Boundaries, p. 270. 21 Gellner, Nationsand Nationalism , p. 1. 22An antecedent condition is ‘[a] phenomenonwhose presence activates ormagniŽes theaction of acausal law or hypothesis.Without it causation operates more weakly… ornotat all’. See StephenVan Evera, Guideto Methods forStudents of PoliticalScience (CornellUniversity Press, Ithaca, NY, 1997),pp. 9–10. Synthesisof the Theories of Peripheral Nationalism 183 situation,the potential for nationalistmovements would be moderate because high elite statuscould offset some of the nationalist demands aroused by different identities. Cell III representsa situationin which the peripheral elites have low status in the political centre butthe majority of the population identify themselves as belongingto the same core nation. Inthis case, the nationalist potential is from lowto moderate because there would be very few incentivesto seekseparation from theirfellow countrymen (Cell III). Inherentin these twosituations is that national identity seems to play a moreimportant role than elite status. After all,as Ernest Gellner states succinctly, ‘ nationalismis a politicalprinciple, which holdsthat the political and the national unit should be congruent’ . 23 BydeŽ nition, nationalismrequires a distinctnational identity. However, relying solely on nationalidentity toexplain peripheral nationalism seems a bitobvious, and tautological. This is where the inclusionof elite status can illuminate our grasp of the problem.

Ethnic Conicts in Xinjiang

Xinjiang’s relationshipwith China dates back in history. As earlyas 60 BCE, the Han dynastyrecognised the strategic importance of Xinjiang as the ‘ westernfrontier’ and establishedgarrisons in keyoases there. As theChinese empire waxed and waned, Xinjiang experiencedperiods of revolts and independence. Most of the time, its relationship with Chinaproper was loose.It did not become a provinceof China until the late nineteenth century.During the Republican period, Xinjiang was dominatedby a handfulof Chinese warlords.Its provincialstatus was abolishedafter the Communist Revolution of 1949. To reect the diversity of ethnic minorities in Xinjiang, the PRC governmentestablished an autonomousregion on 1 October1955. Thereare 56 nationalities in China, and according to the 1995 sample census over 91 percent of the total population are Han. Nearly two-thirds of China’ s arepopulated by minoritiesand abundant in natural resources, such as petroleum, minerals and timber. Mao Zedong,when addressing the relationship between the Han nationality and the minority nationalities,reminded his cadres that:

Wesay China is a countryvast in territory, rich in resources, and large in population; as a matterof fact,it isthe Han nationality whose population is large and the minority nationalities whoseterritory is vastand whose resources are rich, or at least in allprobability their resources underthe soil are rich. 24

Minoritypolicy has always been an important aspect of China’ s domesticpolicies. The ofŽcial policy stresses the equality, solidarity, autonomy and co-prosperity of nationalities. 25 TheChinese government established several autonomous regions and counties across the country.In Xinjiang, it divided the region into several autonomous prefectures and counties accordingto thedwelling places of nationalities, and recruited minority cadres into the local government.Minority languages are used in newspapers, broadcasts, and government documents.For instance,the Xinjiang Television Station broadcasts in Mandarin, Uighur andKazak, and the XinjiangDaily ispublished in the above three languages plus Mongolian,to reect their relatively larger population. At present, there are 47 nationalities amongXinjiang’ s 17.5million population (1998). TheUighurs are the largest nationality in Xinjiang, constituting nearly 50 per cent

23 Gellner, Nationsand Nationalism , p. 1. 24‘ Speechat anEnlarged Meeting of the Political Bureau of theCentral Committee oftheChinese Communist Party (April25, 1956)’ , in Selected Worksof MaoZedong ,vol.5 (ForeignLanguages Press, Beijing,1977), p. 295. 25Huang Guangxue, Xinshiqi minzu wenti tansuo (Inquirieson the Problems of Nationalities in the New Era ) (Zhongyangminzu xueyuan chubanshe, Beijing, 1989), pp. 19–22. 184 Yuan-kangWang

Figure 2. EthnicComposition in Xinjiang 1949 and 1995. Source:JinYunhui (ed.), Xinjiangjingji yu shehui fazhan gaishu (AGeneralSurvey of Economicand SocialDevelopment in Xinjiang )(Minzuchubanshe, Beijing, 1998), pp. 205– 6. ofthepopulation; separatist movements are mostly associated with them. More than 80 per centof the Uighurs live in southern Xinjiang. The Uighur language belongs to the Altay Turkiclanguage, and is writtenusing the Arabic alphabet. 26 TheHan population in Xinjiang was smallbefore the Communist Revolution, making up 7 percent of the population. The newCommunist government recognised the importance of this peripheral region and systematicallyencouraged and forced Han migration into this region. As aresultof this policy,Han population grew rapidly,reaching nearly 40 per cent at present (see Figure2). Thereare also a largenumber of Kazaks, Hui (Chinese Muslims), Kirgiz and Mongols, alongwith a muchsmaller number of Xibes, Tajiks, Uzbeks, Manchus, Daurs, Tatars and Russians.After 1949,the (CCP) establishedcadre schools and systematicallytrained minority cadres in Xinjiang. Most of the CCP-trained cadres were recruitedin the local government in accordance with the principle of national autonomy; few were ableto berecruitedto the centre. In 1997, 48.9 per cent of thecadres in Xinjiang areminorities, an increase of 2.6 times from 1978and 23 times from 1954. 27 Heads of autonomousprefectures, autonomous counties, and nationality townships are all minority cadres.A policysimilar to ‘ afŽrmative action’ was inplace where minorities dwell to ensurea certainproportion of minority cadres in local governments, parties, and the military.28 Despitethe high percentage of minority cadres in the local government, there exists a sharpdiscrepancy in their living standards and educational level compared with the Han. Accordingto the 1990 census, the infant mortality rate for minoritieswas 73.6per cent, 3.6 timeshigher than Han infants born in Xinjiang (20.4 per cent). Life expectancy among the minoritieswas 62.9years, while among the Han it was 71.4years. The Han were alsomore

26Foreign Affairs OfŽce ofthe People’ s Governmentof the Xinjiang Uighur Autonomous Region (comp.), Xinjiang: AGeneralSurvey (New WorldPress, Beijing,1989), p. 28. 27Liu Min and Wusman Niyazi, ‘ Bianjiangshaoshu minzu diqu ganbu renshi zhidu gaige chutan’ , (‘AStudyof FrontierMinority Regions’ Cadre andPersonnel System Reform’ ), Xinjiangdaxue xuebao (Journalof Xinjiang University),vol.27, no. 4 (December 1999),p. 6. 28A. DoakBarnett, China’s FarWest: Four Decades ofChange (Westview Press, Boulder,CO, 1993), p. 366. Synthesisof the Theories of Peripheral Nationalism 185

Figure 3. EducationalLevel of EmployedMinority Nationalities in Xinjiang1990. Source: TongYufen, Reyila Maimaiti, et al.,‘Xinjiangshaoshu minzu renkou de suzhiyu jiuye’ , (‘The QualiŽcations and Employment of Xinjiang’s MinorityPopulation’ ), Xinjiangdaxue xuebao , vol.27, no. 3 (September1999), p. 17. educatedthan the indigenous nationalities (see Figure3). Although the educational level of thenationalities has improved in absolute terms since 1949, they are still signiŽ cantly behindthe Han. More than 30 per cent of the Han work in industries that require higher skills,such as the industrial and construction sectors, while over 80 per cent of minorities arein agriculture, forestry, animal husbandry, and Ž shing. 29 Inorder to minimise friction, there are some institutional designs to separate Han areas from minorityareas, such as separate restaurants, cinemas, and marketplaces. However, Hanand other minorities still harbour negative stereotypes of each other. The Han’ s habit ofeatingpork, the way they use well water, and methods of food preparation are considered uncleanby the Muslims, who in turn are viewed by Han as ‘backward’and ‘ superstitious’. Neitheris completelyconvinced that their separate facilities are equal. Ethnic relations were relativelyquiet in theearly years of thePRC untilthe Anti-Rightist Campaign and the Great LeapForward in1957– 58, when members of Xinjiang’ s minorityelites were declared guiltyof ‘ localnationalism’ and sentenced to reform throughlabour. The famine after the GreatLeap fuelled the  amesof ethnic hostility, escalating into a majorriot in 1962. The CulturalRevolution was evenmore devastating; some Muslims were force-fedpork, mosqueswere burned,and there were somepublic burnings of the Koran by the Red Guards.30 In1975, Muslims rebelled when ordered to work on Fridays,their holy day. Riots brokeout in 1979 after two Han beat a Uighurto death. Xinjiang’s ethnicproblems were exacerbatedafter 1989 when fundamentalist Muslims beganto protest. A politicalgroup, the Islamic Party of East Turkistan, had called for a jihad (holywar) towipe out Chinese rule in Xinjiang. Ethnic issues were evenmore complicatedafter the dissolution of the Soviet Union; the independence of former Soviet centralAsian republics seemed to have encouraged the separatists. Some kind of terrorist actwas carriedout. In 1990, about 50 people died when the government put down an uprisingby religiousextremists in a townnear Kashgar. In February 1992, six people were killedand 20 injured in an Urumqi car bomb explosion that was linkedto Muslim separatists. 31 Separatistsassassinated several people in 1996, including two policemen and apro-BeijingMuslim cleric. In early May, nine Muslim rebels were reportedlykilled in a

29Tong Yufen, Reyila Maimaiti, et al.,‘Xinjiangshaoshu minzu renkou de suzhiyu jiuye’, (‘TheQualiŽ cations andEmployment of Xinjiang’ s MinorityPopulation’ ), Xinjiangdaxue xuebao ,vol.27, no. 3 (September1999), pp. 16–19. 30June Teufel Dreyer, ‘ThePLA andRegionalism in Xinjiang’ , ThePaciŽ c Review ,vol.7, no.1(1994),pp. 42–3. 31Ibid, pp. 49– 52. 186 Yuan-kangWang gunbattle with police while trying to plot bomb attacks. 32 InFebruary 1997, about 1,000 Uighurseparatists rioted in theregion’ s north-westerntown of Yiningand at least10 people were killed. 33 Incontrast, Guangdong is much more integrated into China dating back to 224 BCE whenthe Ž rst emperorof China,Qin Shihuang, uniŽ ed the whole country. It has remained partof the country’ s administrativesystem ever since. 34 Ina word,Guangdong has been successfullyincorporated in China’ s nation-buildingprocess, while Xinjiang has remained onthe fringe of the Chinese nation and separatist movements have been a constanttheme theresince its incorporation. Guangdongwas thegateway of the Western powers to Chinain the nineteenth century. Its capital, (Canton), was oneof theŽ rst portsopened to the West. Commercial activitiesand trade brought economic dynamism to this region. Guangdong was the Žnancialbase of the Nationalist Government in Guangzhou when the Northern Expedition was launchedin 1926 to unify the country. While the decade 1926– 36 witnessed intermit- tentGuangdong , the movement was aform ofprovincialism and did not seek independencefrom China. 35 Theprovince became economically dynamic again after China’s economicreforms in1978. Based on Guangdong’ s distinctdialect and cultural traits,its long distance from Beijing,and economic dynamism, current theories of periph- eralnationalism would predict strong nationalist movements in that province. However, thereis none in Guangdong. Whyare there nationalist movements in Xinjiang, but not in Guangdong? The next sectionswill address this question in light of elite status and national identity.

Elite Status Sinceour interest lies in the elite status in the central government, we needto examine the percentageof elites in high positions in the centre. In the PRC, the CCP isthe only dominantpower in the state. Virtually all key positions in the central government are held byCCPmembers.Even at the local level, party membership is anecessaryrequirement for anyoneinterested in pursuing a seriouscareer in politics. The presence of theparty at every levelof thestate is a widelyrecognised phenomenon in China.Hence, in assessing the elite statusof any particular province or region, we cananalyse the percentage of its elites in highpositions of the party and come to a fairlygood estimate. Theorganisation of the CCP consistsof four major party organs: The National Party Congress,the Central Committee (CC), the Politburo (also called the Political Bureau), and theStanding Committee of the Politburo. The frequency of meetings and sizes of membershipdetermine the order of importance of these four organs: 36 TheNational Party Congress has the largest membership, over 1,500 in recent Congresses.It meets infrequently (generally once every Ž veyears),and because of its size, inpractice, it becomes an organ for endorsingmajor decisions made by the leadership. TheCC alsomeets infrequently (once or twice a year),but has a smallermembership (approximatelythree hundred in recent years, including alternate members), selected

32Anthony Davis andSangwon Suh, ‘ China:The West Gets Wilder’, Asiaweek,vol.22, no. 24 (14June 1996). 33 TheNew YorkTimes ,11February 1997. 34For a detailedhistory of Guangdongup to1949, see JiangZhuyuan and Fang Zhiqin (eds), JianmingGuangdong shi (AConciseHistory of Guangdong)(Guangdong renmin chubanshe, Guangzhou, 1993). 35See JohnFitzgerald, ‘ Increased Disunity:The Politics and Finance of Guangdong Separatism, 1926–1936’ , ModernAsian Studies vol.24, no. 4 (1990),pp. 745– 75. 36Kenneth Lieberthal, GoverningChina: From Revolution Through Reform (W.W.Norton, New York,1995), pp. 159–61. Synthesisof the Theories of Peripheral Nationalism 187

Table 1. EliteStatus in the CC byEthnic Origins

GuangdongXinjiang a

7thCC (1945–56) 5/77 0 8thCC (1956–69) 7/170 1/170 9thCC (1969–73) 9/279 2/279 10thCC (1973–77) 12/319 3/319 11thCC (1977–82) 8/333 3/333 12thCC (1982–87) 11/348 2/348 13thCC (1987–92) 10/285 3/289 14thCC (1992–97) 9/319 4/319

*Denominatorsrepresent total membership; numerators rep- resentnumbers of members in each CC. BothŽ guresinclude alternatemembers. a AllXinjiang elites are from theUighur nationality, except for oneKazak in the 14th CC. Source: LijieZhonggong zhongyang weiyuanhui renming cidian (BiographicalDictionary of VariousCC Membersof theCCP) (Zhonggongdangshi chubanshe, Beijing, 1992); Zhonggongdi shisijie zhongyang weiyuan minglu (BiographicalDictionary of the14th CC Membersof the CCP) (Zhonggong dangshi chubanshe,Beijing, 1993). bythe Party Congress according to a listmade by the Politburo. Article 15 of the CCP Constitutionstipulates that ‘ [o]nlythe Central Committee of the Party has the power to makedecisions on major policies of a nation-widecharacter’ . Inreality, with few exceptions,it discusses and announces policies handed down by the Politburo in plenary sessionsrather than decides them. Nevertheless, all of its members hold other important positionsand have access to some social privileges and inside information of the Party. ThePolitburo typically has 14– 24 members, selected by the CC. Itfunctions as asmall butpowerful committee. Politburo members are among the top power elites of China. TheStanding Committee of the Politburo is the most powerful organ of the party, and thusof the state. It has four to six members, selected by the CC, andmeets frequently (normallyonce a week). Table1 showsthe degree of representation of elites from Guangdongand Xinjiang in theCC. As isclearly seen, Guangdong elites enjoyed a signiŽcantly higher status than their Xinjiangcounterparts. Although Guangdong’ s populationis much larger than Xinjiang’ s, its favourableposition is still re ected in the fact that Uighur membership in the CC was occupiedby only a handfulof the same people in different terms, while Guangdong membershipwas changingconstantly, with new faces coming in. Only a totalof Ž ve Uighurswere ableto serve in the CC from 1957to 1992, compared with 37 Cantonese duringthe same period. 37 Thedata here suggest that the role of Uighurs in the centre was moresymbolic than real. Ananalysis of elitestatus will not be complete without looking at the micro level. Three CantoneseCC members,,Liao Chengzhi and Xie Fei, were ableto serve in the

37 See LijieZhonggong zhongyang weiyuanhui renming cidian (BiographicalDictionary of Various Central Committee Members ofthe CCP) (Zhonggong dangshi chubanshe, Beijing, 1992). 188 Yuan-kangWang

Politburo.Ye evenserved on the Standing Committee of the Politburo (10th– 12th), the mostpowerful organ of theparty. Ye joinedthe CCP in1927,and had served high positions inthe Red Army. He was acommanderin the Long March and became mayor of Beijing (thenPeiping) in 1949. He was madeMarshall in 1955 and was Ministerof Defense in 1975.He hadbeen member of the Politburo since 1966 and had remained a keynational Žgureuntil his death in 1986. Liao Chengzhi joined the CCP in1928, served in the propagandaand united fronts departments, and was electedto the CC (alternatemember in 7th,member in 8th, 10th– 12th). He becamea memberof the Politburo in 1982, but died ayearlater. Xie Fei joined the CCP in1949. In his youth, he was incharge of the propagandadivision of his home county in Guangdong. He was promotedCCP secretary ofthe Guandgong provincial committee in 1983. Later on, Xie was electeda CCmember in1987 and 1992, respectively. He becamea Politburomember in 1994, was re-electedin 1997,and has remained in that position to date. Historically,Guangdong elites have long been able to serve high ofŽ ces in the central government.During the Ming Dynasty (1368– 1644), for example,several Cantonese held importantpositions in the central government. Although the percentage of Cantonese elites inthe central government was lowerthan that of the northern provinces, this re ected a generalbias in favour of northernelites rather than exclusion. Elites from distantprovinces suchas Fujian, Yunnan, Guangxi and Sichuan, ranked lower as well. 38 Thesituation was similarduring the Qing Dynasty (1644– 1911). Guangdong elites were ableto make it to the topof thenation. Nearly 4 percent of theQing Dynasty’ s jinshi degreeholders came from Guangdong(obtaining a jinshi degree was a sine qua non for peopleinterested in governmentpositions). In terms of the number of jinshi producedper prefecture, Guang- dong’s Guangzhouranked Ž fthin the nation. 39 Cantoneseelites were alsoable to lead nationalmovements. The leaders of the late Qing’ s threemost important events— the TaipingRebellion (1851– 64), the Hundred Days Reform (1898), and the Republican Revolution(1911)— were allCantonese. Therelatively high status of Guangdong elites substantially curtailed the incentives for secession.Career aspiring individuals knew that, by working hard, they could advance towardsigniŽ cant positions in the central government. Having an effective voice at the centre,Cantonese could rest assured that their interests would not be neglected by Beijing andthat their grievances would be heard and remedied. They thus saw littleincentive to havea nationof their own. Theonly Uighur to be elected to the Politburo was SaifudinAzizov (Seypidin Aziz); neverthelesshe was merelyan alternate member. Saifudin joined the Communist Party of theSoviet Union in hisyouth, but was allowedto transferhis membership to theCCP after MaoZedong negotiated with Stalin on his behalf in 1950. He becamedeputy governor of Xinjiangand later chairman of the Autonomous Region. He helpedimplement Chinese policiesin Xinjiang, and in returnwas awardedwith positions in the CC (alternatemember inthe 8th, member in the 9th– 13th) and in thePolitburo (alternate member in the 10th and 11th).40 However,Saifudin was neverfully entrusted with top CCP, government or military posts.41 Thepositions he maintained were moresymbolic than real.

38James B.Parsons,‘ TheMing Dynasty Bureaucracy: Aspects ofBackground Forces’ , inCharles O.Hucker(ed.), ChineseGovernment in Ming Times: Seven Studies (ColumbiaUniversity Press, New York,1969), pp. 182– 3. 39Ping-ti Ho, TheLadder of Success inImperial China: Aspects ofSocial Mobility ,1368–1911 (Columbia University Press, New York,1962), pp. 228, 247. 40 LijieZhonggong zhongyang weiyuanhui renming cidian ;alsosee WolfgangBartke, BiographicalDictionary and Analysisof ChineseParty Leadership 1922– 1988 (K.G.Saur, New York,1990), pp. 395–6. 41Donald H. McMillen, ChineseCommunist Power and Policy in Xinjiang, 1949– 1977 (Westview Press, Boulder, CO,1979), pp. 36– 7. Synthesisof the Theories of Peripheral Nationalism 189

Thepolicy of co-opting ethnic leaders into the power structure dated back to the later daysof the Republican era (1912– 49), when separatists proclaimed the establishment the EastTurkistan Republic (ETR) in1933 under Soviet in uence. It was latercrushed, and re-establishedin 1944. The then Nationalist government was preoccupiedwith Ž ghtingthe Japaneseand the Communists, and, in an attempt to deal with the ETR, sought to co-opt itsleaders into the local power structure. After thePeople’ s LiberationArmy (PLA) marchedinto Xinjiang in 1949, it became clear to its leaders that in the long run military resistancewould be useless,and it was alsoclear to theCommunists that resistance, if any, wouldbe bloody and enduring. The CCP thensought to incorporate Xinjiang leaders into thepower structure. 42 Itappointed (a Tatarborn in Aksu, Xinjiang, and former NationalistParty functionary who had participated in the ETR) governor of the province,and Saifudin deputy governor. Burhan was essentiallya minoritynationality Žgureheadin Xinjiang. However, he did condemn the separatist intentions of his fellows andstressed the unity of Xinjiangwith China. 43 He was replacedby Saifudin in late1950s. ETRmilitaryleaders were alsoincorporated into the PLA, thoughsome were laterreplaced by Han.44 However,the chairman of the autonomous government has always been of the Uighurnationality. These Uighur cadres had learned to speak Mandarin and provided Beijingwith their networks of inuence and patronage in their home region. In return, they were rewardedwith ofŽ cial positions and privileges. Thelow status of Uighurelites in the centre was seenas areection of Han domination. Somefelt that the Han was exploitingthe rich natural resources (particularly cotton and petroleum)of thisregion for thedevelopment of China’ s coastalregion without giving the Uighursadequate compensation. A recentproposal for theconstruction of a $14billion naturalgas pipeline from Xinjiangto Shanghai exacerbates suspicion of exploitation. 45 In addition,the Chinese nuclear test site at Lop Nor raisedconcerns among the indigenous people.Within the region, there was asensethat Xinjiang was allowedto fall behind the coastalprovinces in the race towards industrialisation and economic development. 46 In the meantime,there were noUighurelites in the centre powerful enough to defend the interests oftheirhomeland. Uighur separatists are thus tempted to seeindependence as awayto free theirhomeland from Handomination and Ž nallyput themselves in charge of their own affairs.

WithWhom Do They Identify? Identityis anelusive concept,  uidand constantly changing over time. Since conducting a reliablesurvey of identity is extremely difŽ cult at this point, the data presented below will draw largelyfrom thewritings and discourses of intellectuals and ethnic entrepreneurs as wellas from themedia. By carefully examining the discourses, we canstill obtain a fairly goodestimate of the national identities in Xinjiang and Guangdong. Note that this article

42Dreyer, ‘ThePLA andRegionalism in Xinjiang’ , pp.41– 2. 43McMillen, ChineseCommunist Power and Policy in Xinjiang , p. 33–4. 44Most of theother ETR leaders were killedin a mysteriousplane crash enroute to Beijing. See Dreyer, ‘ThePLA andRegionalism in Xinjiang’ , p42;Barnett, China’s FarWest , p. 373. 45John Pomfret, ‘ China:Go West, Young Han; Beijing Urging Dominant To Resettle, DevelopRestive Regions’, TheWashington Post ,15September 2000. 46Beijing embarked upon a massive campaignin March 2000 to develop the western regionspartly due to this concern,planning to spend as muchas $12billion on 70projects during the next Ž veyears inXinjiang.These projectsinclude airports, telecommunications, agriculture and other infrastructures. The goal is todevelop Xinjiang’s economyand strengthen its ties withChina. See SusanV. Lawrence, ‘Where BeijingFears Kosovo’, FarEastern Economic Review ,vol.163, no. 36 (7 September 2000). 190 Yuan-kangWang doesnot attempt to offer anexplanation of how identity is formed. Instead, I ammainly interestedin showing what the identity is rather than explaining how it came into being.

Xinjiang TheQing dynasty’ s controlof Xinjiang since the late eighteenth century did not stem separatistmovements. Many Uighurs viewed Qing rule as illegitimate and an onslaught on theirway of life. 47 Severalrebellions took place in theregion, but were quicklysuppressed. Thelongest and most important of these rebellions was ledby Yakub Beg from the mid-1860sto early 1878, who remained a legendaryŽ gurein thearea. These rebellions and Russianpenetration into the region led to the Qing court’ s decisionto proclaim Xinjiang (literally‘ theNew Frontier’) aprovinceon 17 November 1884. It introduced several militaryand political reforms inan attempt to make Xinjiang become a regularprovince. Throughoutthe Republican era from 1912to 1949, the Chinese government made clear itspolicy on minority nationalities: to preserve China’ s nationalunity. Sun Yat-sen, the founderof the Republic of China, delineated this policy in his Theory of Nationalism, the Žrst ofhis Three Principles of the People. According to Sun, the Chinese people are composedof Ž vemajor nationalities— Han, Manchu, Mongolian, Hui (including Uighur), andTibetan. Together they made up the Republic of Five Nationalities ( Wuzu gonghe). The initialRepublican national  aghad Ž vecolours, representing each of theŽ venationalities. Sunadvocated assimilation of different nationalities into ‘ asinglepowerful nation’ . China, inhisopinion, should follow the examples of Switzerland and the United States and ‘ satisfy thedemands and requirements of allraces and unite them in a singlecultural and political whole’.48 Reecting Sun’ s view,the ofŽ cial policy of the Kuomintang (Nationalist) government regardingminority nationalities was, theoretically, equality for allnationalities and the right ofself-determination. 49 Inpractice, Xinjiang was dominatedby Hanwarlords or local elites duringthis Republican era. The harsh policies of the warlords had a verynegative impact onhow the minorities perceived the Han Chinese. Viewing Chinese rule as oppression of aforeignpeople, Muslim Turks in Yili, located in far north-westernXinjiang, rebelled againstthe Chinese government and re-established the ETR in November 1944. The rebels successfullydrove out Chinese troops from alltowns and border posts in the three districts oftheprovince and advanced toward the provincial capital Dihua (U ¨ ru¨mqi).One early rebel pamphletdescribed why they were Žghting: Whoare we? Who and where are our near and far relations? … .[A]nyman who seeks the truthand whose heart is right cannot fail to say that the root of our nation and soul is not in China,but in Central Asia … .Ournative place is … .Ofthe fourteen nations livingin East Turkestan, the 10 nations accounted the most numerous have had no national, racial,or cultural relationship nor any community of bloodwith the Chinese, nor did any ever exist… .Ourfathers and grandfathers were for hundreds of years oppressed by the savage Chinese… .EastTurkestan belongs to the real masters of the territory, the Uighurs, the Taranchis,the Kazaks, the Kirghiz, the Tatars, the Uzbeks, together with all those who live amongthem in peace and friendship and who alike suffer Chinese oppression. 50 Apparently,the rebel pamphlet exhibited a distinctnational identity from theChinese. They

47Colin Mackerras, China’s Minorities:Integration and Modernization in the Twentieth Century (OxfordUniversity Press, HongKong, 1994), pp. 35– 6. 48Sun Yat-sen, Memoirsof a ChineseRevolutionary, A Programmeof NationalReconstruction for China (AMS Press, New York,1970), pp. 228– 9. 49Linda Benson, TheIli Rebellion: The Moslem Challenge to Chinese Authority in Xinjiang 1944– 1949 (M.E. Sharpe,Armonk, NY, 1990),p. 12. 50Cited in Ibid., ‘ WhyAre We Fighting?’, pp.200– 5. Synthesisof the Theories of Peripheral Nationalism 191 deniedany ties with China and viewed the Han Chinese as oppressive foreign colonists and urgedtheir people to resist Chinese rule. They sought to demonstrate that Muslim Turks were anethnically distinct nation from theHan Chinese. The rebellion would not have succeededhad there not been enough popular support. The abuses perpetuated by success- iveChinese regimes, culminating in the rule of the warlord Sheng Shicai, had made the peopleready to acceptyet another attempt at self-rule. This period witnessed the growth of resistanceorganisations aimed at ousting Han Chinese from Xinjiang. 51 They looked forward toan independent country composed of Muslim Turks, putting an end to the long ruleof the ethically distinct Han Chinese. TheChinese central government was stillat war withthe Japanese at thistime and was unableto defeat the rebels militarily. It sought a negotiatedoutcome with the Muslim leadersand in January 1946 reached a peaceagreement promising self-government. The ensuingcoalition government did not fulŽ l thatpromise because the Chinese military interferedfrequently in the region’ s civilaffairs andthe Han bureaucrats were reluctantto sharetheir power. 52 Thecoalition lasted only one year and collapsed in 1947. The rebels maintainedindependent control of the three districts in Yili, while the Nationalist Party controlledthe rest of theprovince, both until Communist take-over in 1949.Chinese failure inXinjiang during this period can be attributed to two factors: the insistence on Xinjiang behavinglike other provinces of China and continued interference by Chinese military in civilaffairs. 53 Nationalitiesin the PRC were state-sanctioned.From 1953,the PRC governmentsent outidentiŽ cation groups, including linguists, archaeologists, historians and ethnologists, to checkthe validity of claims being made by numerous groups for statusas minority nationalities.Applying Stalin’ s fourcriteria for anationality—common language, common territory,common economic life and common culture— and  exiblyadjusting for the ‘nationalwill’ , thestate approved the status of some50 nationalitiesout of over400 groups whohad made such a claim. 54 Thebasic policy of thePRC towardminorities is that all the nationalitiesshould be treated equally and have the right to autonomy. According to the constitution,the PRC isa ‘unitarymultinational state’ , and‘ [a]llthe national autonomous areasare inseparable parts of the People’ s Republicof China’ . 55 Secessionis not to be tolerated.As aresultof thisnew policy, the status of Xinjiangwas changedfrom province toautonomous region in 1955. Theinitial PRC policies in Xinjiangwere generallymoderate. Taking into account local conditionsand peculiarities, Beijing carefully implemented its policies to achieve the goals ofstabilisation and consolidation of its power in the region. The granting of regional autonomyin 1955 put minority cadres such as Burhan and Saifudin in nominal leadership. Thereal power however still resided in the hands of Han cadres such as and WangEnmao, along with their colleagues in the PLA. 56 Ethnicrelations were gradually stabilisedbut began to deteriorate after the Great Leap Forward andthe later Cultural Revolution.The excesses of thesetwo campaigns, such as starvation,burnings of mosques, andpublic burnings of the Koran, resulted in several ethnic riots. Politicaland cultural differentiation from theChinese provided Uighur nationalists the

51 Benson, TheIli Rebellion , pp. 34–5. 52Mackerras, China’s Minorities , p. 67. 53 Benson, TheIli Rebellion ,pp.179– 82. 54Mackerras, China’s Minorities ,pp.140– 3; ThomasHeberer, Chinaand Its NationalMinorities: Autonomy or Assimilation? (M.E.Sharpe, Armonk, NY, 1989),pp. 30– 4. 55 Constitutionof the People’ s Republicof China (ForeignLanguage Press, Beijing,1961), p. 9. This policy is stipulatedin Article 3oftheŽ rst constitutionadopted in 1954. 56Donald H. McMillen,‘ Xinjiangand Wang Enmao: New Directionin Power, Policy, and Integration?’ , The China Quarterly,no.99 (September 1984), p. 570. 192 Yuan-kangWang raw materialsto constructtheir argument. The historical intermittent Chinese control of the area,they contend, is notsufŽ cient to establish a territorialclaim. The formal incorporation ofXinjiangin Qingdynasty ‘ cannotbe viewedas anassertion of Chinesedominion because theManchus were notChinese’ . 57 Preferringto call the region ‘ EastTurkestan’ instead of theChinese name of ‘ Xinjiang’, Uighurnationalists see their homeland as a country occupiedby a colonialforeign government. They stress a deepcultural, religious, and linguisticgap with China and resist what they called ‘ forcedassimilation’ by the Chinese government.Isa YusufAlptekin, a Uighurleader living in Turkey, was quotedas saying: ‘At present,the peoples of Eastern Turkestan are forced to choose between national extinctionand a mortalstruggle of heroicresistance to defend their cultural identity against aChineseCommunist policy of assimilation’ . 58 Somelocal people also resent Chinese migration.One Uighur university student was reportedlyquoted as saying: ‘ We havetoo manyChinese here— too many, and more keep coming … TheChinese have no religion; theyare inŽ dels. We lookto Turkey and Iran andArabia. We donotlook to Beijing.Islam is our law’.59 Therevival of Islam, which was harshlyrepressed during the Cultural Revolution, accompaniedChina’ s open-doorpolicy in the post-Mao reform. Hundredsof new mosques were built,new Islamic colleges and Koran schools were opened,and interest in the hajj (pilgrimage)was renewed,accompanied by a revivalof Islamic architecture and renewed useof Arabic in place of Roman script. As AnthonyDavis points out, ‘ theIslamic revival hasbeen more a reassertionof a culturaland ethnic identity deŽ ned in religious terms than anyconscious embrace of an ideology of Islamism. Yet, the rediscovery of religion has unquestionablyprovided fertile recruiting ground for thosewho have adopted a speciŽcally Islamistagenda and are committed to armed struggle ( jihad)againstan “ occupying” communistpower.’ 60 Increasingly,more people are wearing Islamic dress, refusing to speak Mandarin,and defying Beijing’ s efforts toimposea uniformtime zone across China. Some evenfound links with Kosovo and celebrated its liberation from thenon-Muslim Serbs and fantasisedabout a similar‘ rescue’in Xinjiang. 61 Theemergence of the newly independent Islamic states in Central Asia after the Cold War encouragedUighur separatists to seek a nationof their own. One remarked that ‘ [w]e Uighurswill Ž ghtfor ourliberty and independence and we wantto be as free asthe other Asianrepublics of the old Soviet Union’ . 62 Theproximity of these republics to Xinjiang alsomade them an ideal place for sourcesof weapons, money, training, as well as places ofrefuge.Some Uighurs have reportedly joined the Taliban in Afghanistan,and some even havelinks with alleged Saudi terrorist Osama binLaden. In order to deal with Uighurs’ increasingcontacts with Islamic groups, Beijing has requested assistance from Russia, Kazakhstan,Kyrgyzstan and Tajikistan to help put down separatist movements, terrorism, andreligious extremism. 63 Todate, religion has increasingly become a powerfulsource of nationalidentity among Uighur nationalists. Byemphasising their distinct identity of Turkic origin, constructed mainly through

57(No author) ‘ Is Eastern Turkestana ChineseTerritory?’ , EasternTurkestan Information Bulletin vol. 3, no. 1 (February1993). [ETIB is availableon theInternet at http://www.taklamakan.org/etib/] 58‘ Appealto the Free World’, EastTurkestan Information Bulletin ,vol.3, no.6 (December 1993). 59Jeffrey Taylor,‘ China’s WildWest’ , TheAtlantic Monthly ,vol.284, no. 3 (September1999), pp. 22– 9. 60Anthony Davis, ‘ XinjiangLearns toLive With Resurgent Islam’ , Jane’s IntelligenceReview ,vol.8, no. 9 (1 September1996), pp. 417ff. 61Elisabeth Rosenthal, ‘ DeŽant Chinese Muslims Keep TheirOwn Time’ , TheNew YorkTimes ,19November2000. 62Patrick E. Tyler,‘ InWest China,Tensions With Ethnic Muslims Erupt in Riots,Bombings’ , TheNew YorkTimes , 28February1997. 63Susan V. Lawrence, ‘Where BeijingFears Kosovo’; AhmedRashid and Susan V. Lawrence, ‘JoiningForeign Jihad’, FarEastern Economy Review ,vol.163, no. 36 (7 September 2000). Synthesisof the Theories of Peripheral Nationalism 193 history,religion and language, Uighur nationalists have appealed to their people and the worldfor self-determinationand independence.

Guangdong Separatedby the high mountains of the Nanling range from therest of China, Cantonese havelong considered themselves somewhat distinct from otherChinese ethnic groups. They speaka languageunintelligible to Mandarin speakers unless written in characters, maintain theirunique cultural style, and have their own regional identity. For centuries,the area Lingnan,the ancient name for Guangdongand neighbouring Guangxi, was considereda backwardbarbarian region by northern China. Although Chinese in uences reached this areaas early as the Qin Dynasty (221– 206 BCE),thenatural barriers of highmountains had prohibitedcontacts with the north until the Tang Dynasty (616– 907 CE)whenimprovements inriver and land transportation brought the region into closer economic and cultural contactswith northern China. The SiniŽ cation of Lingnan was alongprocess, fuelled by repeatedHan migration from thenorth. The in ux of Han brought the language of the north andmixed with the local language in what now becomes known as Cantonese. 64 Cantonese cultureis now accepted as part of the Chinese subculture. Some Cantonese are still proud ofthe fact that their language was morelike ancient Chinese than was Mandarinand that theywere purerrepresentatives of Chineseculture. 65 Today,as theresult of continuousHan migrationfrom thenorth, most of the Cantonese population is related to Han Chinese. Whenthe Westerners came to China in the nineteenth century, there had been a high levelof resistance and xenophobia among those who were accustomedto a Sinocentric worldorder. Compared with northern China, the idea of yixia zhi fang (‘protection against barbarians’) was muchweaker in Guangdong. Though still met with some resistance, Westernideas and technology got their Ž rst footholdin this southern part of China in the latenineteenth century. Patriotism characterised the so-called ‘ LingnanCulture’ , manifest- ingitself in poetryand action. 66 LiangQichao, a Cantoneseintellectual who was oneof the mastermindsin the Hundred Days Reform in late Qing period, wrote:

Beautiful,my young China, eternal as the heaven; MagniŽcent, my Chinese youth, boundless as the country!’ 67

Others,lamenting the weakness and corruption of Chinaat thattime, resorted to rebellions, reforms, orrevolutions. Three important modern Chinese movements originated in Guang- dong:the Taiping Rebellion (1851– 64) led by Hong Xiuquan, the Hundred Days Reform (1898)led by Kang Youwei and Liang Qichao, and the Republican Revolution (1911) led bySun Yat-sen which overthrew the last dynasty of China. All of the above leaders were Cantonese,who had no doubt that they were Chineseas well. Though having their own differentspoken language and regional identity, the Cantonese still identify themselves as Chineseand they are no less patriotic than northern Chinese. Guangdong’s traditionof regionalismgrew notout of anylocal movements for political separationbut out of itsdistance from variousChinese capitals in the north. There had been 15or so independent or de facto governmentsin Guangdong since its incorporation into

64Jiang and Fang eds, JianmingGuangdong shi, p. 12. 65Leo J. Moser, TheChinese Mosaic: The Peoples and Provinces ofChina (Westview Press, Boulder,CO, 1985), pp.206, 209. 66Hua Min,‘ Lingnanwenhua yu minzujingshen’ (‘ LingnanCulture and the National Spirit ’),Guangdongshehui kexue (SocialSciences inGuangdong ),vol.2 (1995),pp. 117– 18. 67Quoted from Zeng Xianjiu, ‘ LiangQichao shiwen de hexinsixiang ji yishu tese’ (‘ TheCore Thinking and Artistic Characteristic ofLiang Qichao’ s Poetryand Articles ’), SocialSciences inGuangdong ,vol.5 (1995),p. 133. 194 Yuan-kangWang

China in 224 BCE.However,almost all of these were alternativenational governments opposedto the north rather than movements for anindependent country. Guangdong’ s reputationfor regionalismremained after the establishment of the PRC in1949, but this regionalismis concerned with diversity and individualism rather than a drivefor political separation. 68 TheŽ rst goalof the CCP leaderswas tobuild a strongstate. Guangdong’ s unique sub-culture,dialects, distance from Beijing,and history of treatyports helped prompt tighter centralcontrol. 69 Securityconcerns led Beijing to move China’ s industrialbase from the coastalregion to the inland, leaving Guangdong’ s economysluggish. This situation began tochange after China entered the reform erain 1978, as special privileges were grantedto thisprovince for thegoal of economicdevelopment. As aresultof thispreferential policy, Guangdong’s economygrew dramaticallyand became one of the richest provinces of China.The province’ s GDP, nationalincome, total retail sales, and exports were the nation’s largestin 1991. Half ofChina’s foreigndirect investment poured into Guangdong, andone-third of the nation’ s totalexports originated in that province. 70 Guangdong’s rapideconomic development has given the province a newcultural status inChina. Recognising the commercial power of the south, some people in the north began tolearnthe Cantonese language and culture, and some migrated into that province in search for jobs.The anti-imperialist capital of Beijing, on the other hand, was ridiculedby the Cantoneseas a backwardcity of mere talkerswho live off thewealth and proŽ t ofits people.Southern virtues spread to the north and gave new meaning to what is meant by beingChinese. 71 For theCantonese, it is very hard to imagine themselves not being Chinese. Although theyspeak a differentdialect, common written characters provide the essential tools of communication,as they do elsewhere in the dialect-rich China. Guangdong’ s literacyrates rankedamong the highest in China: according to ofŽ cial sources, illiteracy was ‘basically eliminated’in the province by the early 1980s. 72 For thecadres, the ability to speak Mandarinis a badgeof elite status and a necessaryrequirement for anyoneinterested in furthercareer advancement. When one wishes to speakwith the authority of thestate or the Party,Mandarin is thebasic qualiŽ cation to do so. On the other hand, when one wishes to defendlocal interests against the state, local dialects are the preferred means. 73 Thisreminds usofEugen Weber’ s celebratedstudy of modernisingrural France, in whichpeasants spoke patois todiscuss local politics but switched to French when discussing national issues. 74

Conclusion Thecases of Xinjiang and Guangdong illustrate the inadequacy of current theories of peripheralnationalism. Gourevitch’ s theoryfocuses on economically advanced regions and attributesperipheral nationalism to the noncongruence between political leadership and

68Goodman and Feng, ‘ Guangdong:Greater HongKong and the New RegionalistFuture’ , pp.179– 80. 69Peter Tsan-yinCheung, ‘ TheCase ofGuangdong in Central-Provincial Relations’ , inJia Hao andLin Zhimin (eds), ChangingCentral-Local Relations in China: Reform and State Capacity (Westview Press, Boulder,CO, 1994),p. 211. 70Ibid., p. 230. 71See EdwardFriedman, ‘ AFailedChinese Modernity: China in Transformation’ , Daedalus,vol.122, no. 2 (22 March,1993). 72Glen Peterson, ‘ TheStruggle for Literacy inPost-Revolutionary Rural Guangdong’ , TheChina Quarterly , no. 140(December 1994),p. 926. 73Ibid., p. 935. 74Eugen Weber, Peasantsinto Frenchmen: The Modernization of Rural France, 1870– 1914 (StanfordUniversity Press, Stanford,CA, 1976), pp. 67– 94, 266– 9. Synthesisof the Theories of Peripheral Nationalism 195 economicdynamism in ethnically distinct areas. His theory,however, fails to predict nationalistmovements in Guangdong, an economically dynamic region with some ethnic potentiallocated in China’ s peripheryfar awayfrom thepolitical centre. Gellner’ s theory dealswith backward regions and presupposes industrialisation as the driving force of nationalism.Nevertheless, his theory ignores possible forms ofpre-industrialisation nationalism.Separatist movements had been a salientissue in Xinjiang prior to China’ s industrialisation.In addition to theaforementioned drawbacks, the applicability of thesetwo versionsof theory is inherently limited to either advanced or backward regions, but not both.Gourevitch’ s theorycan not be applied to backward regions, while Gellner’ s cannot beapplied to advanced regions. Themodel proposed in thisarticle broadens the scope of applicability.It can be applied tobothadvanced and backward regions. By examining elite status and national identity, the modelsucceeds in explaining what current theories fail in Guangdong and Xinjiang. It directsour attention to the importance of the positions local elites are able to occupy in the centralgovernment and whether local people identify with the core nation in affecting the emergenceof peripheral nationalism. In Xinjiang we foundfew localelites in the central government,and the separatists identify themselves as a distinctnation from theChinese. Theopposite is truefor Guangdong,where local elites enjoy ‘ most-favoured-lord’status in Beijing,and its population has no problem identifying with the Chinese. The low values of thesevariables— elite status and national identity— explain the presence of separatist movementsin Xinjiang, while their high values explain the absence of such movements in Guangdong. Bothvariables in our cases point in thesame direction; i.e. either both have a highvalue atthe same time or a lowvalue. This raises an interesting question in cases where elite statusand national identity move in opposite directions. In Figure 1 (p.182), Guangdong issituated at Cell I, towhich we mayalso add Fujian province, where the majority of the populationspeak the Minnan dialect and its elites have access to high positions in the centre.Xinjiang is located at Cell IV, towhich we mayalso add . Although further researchis required for theother two cells not covered in this study, we canspeculate on whatwould possibly happen when one of thevariables has a highvalue but the other does not.Cell II describesa regionwith high elite status but different national identity. In this case,the potential for nationalistmovement is likely to be moderate because high elite statuscould reduce the nationalist demands resulted from differentidentities. Real world casesare few inthisregard. One possible example is InnerMongolia Autonomous Region, whoseelites had been powerful Ž guresin the centre and has so far nosigniŽcant incidences ofirredentism or separatism. For provincesor regions with low elite status but similar nationalidentity (Cell III), thenationalist potential is likely to be from lowto moderate becausethere would be very few incentivesto seek separation from theirfellow Chinese countrymen.However, a reidentiŽcation process could tip the situation toward Cell IV. Possibleexamples are Yunnan, Guizhou, Gansu, Qinghai and Ningxia.