Music of The Uyghur Tribe

The Uyghur Twelve Muqam


Throughout the course of the 20th century, as newly formed nations have sought ways to assert and formalise their national identity, they have typically acquired a range of identifiable national assets. Thus we find in this period new musical canons springing up across the world. These canons, however, cannot be dismissed as arbitrary collections of works imposed on the public by the authorities. They acquire deep resonance and meaning, both as national symbols and as musical repertoires imbued with aesthetic value. This book traces the formation of one such musical canon: the Twelve Muqam (on ikki muqam), a set of musical suites which has come to mean a great deal to one little-known Chinese Central Asian nation.

The Uyghurs The Uyghurs might be introduced as one of China’s less well-known though more numerous minority nationalities (compared to, say, the Tibetans or the Mongols), or alternately as the only one of the major Central Asian nationalities (alongside the Uzbeks, Kazakhs, Kyrgyz, Tajik and Turkmen) who do not possess their own independent nation state. Culturally we might best regard the Uyghurs as a Central Asian people, although their homeland now lies within the borders of the People’s Republic of China (PRC), in the large desert and mountain region in China’s far northwest, currently known as the Xinjiang Uyghur Autonomous Region (XUAR).

There are also sizeable populations of Uyghurs living in the neighbouring Central Asian states of Kyrgyzstan, Kazakhstan and Uzbekistan. The Uyghurs follow Sunni Islam, and popular practice is strongly influenced by Sufi traditions especially shrine (mazar) pilgrimage. Their language belongs to the Turkic language family, as do the other Central Asian languages with the exception of Tajik, and is very closely related to Uzbek. Their music also displays much continuity with the folk and classical traditions of Uzbekistan and northern Tajikistan, where musicians use the same longnecked lutes and frame drums, and gather their music into large-scale suites, or cycles, called maqām. The term comes from the Arabic maqām but in contemporary Central Asia the concept of maqām, or muqam in the Uyghur pronunciation, is regarded less as a modal basis for improvisation and more as a fixed suite consisting of sung poetry and stories, dance tunes and instrumental sections. Probably the best known of these Central Asian maqām traditions are the six large-scale suites commonly known as the Tajik-Uzbek Shash Maqām. Rivalling this tradition in terms of size and complexity are the Twelve Muqam (on ikki muqam), the prestigious set of musical suites which have come to be emblematic of the Uyghur nation.

As in the better-known situation in Tibet, the relationship between Uyghur minority nationality and the Chinese state during the nearly 60 years of rule by the People’s Republic of China (PRC) has been marked by tension and sometimes violence. Throughout this period, the Twelve Muqam have been deployed as political emblems and tools by the state and by Uyghur nationalists. The Chinese state has invested large sums of money in a succession of projects to preserve and develop the Twelve Muqam, and it uses these projects to showcase the positive aspects of its minority policies on the national and international stage. These policies, and specifically the canonisation of the Twelve Muqam, inevitably meet with a mixed reception amongst Uyghurs, but, positive or negative, their assessments agree on the directly political nature of the canonisation project. To illustrate with two anecdotes: in 2006 I met one loyal old Uyghur cultural cadre based in a small town in southern Xinjiang, who was in ecstasies over the latest release of a full set of VCD (video compact disc) recordings of the Twelve Muqam. ‘Timur Dawamat [the then regional chairman of Xinjiang] did a great job with those VCDs,’ he told me, ‘better than liberating our region twice over!’ (Mulla Tokhti, interview, Qaratal, July 2006). During this same period a joke was circulating on the internet sites maintained by Uyghur exiles (which are blocked by China): A Uyghur meets a Chechen. ‘We have Twelve Muqam’, says the Uyghur proudly. ‘Twelve Muqam?’ retorts the Chechen, ‘huh, you’d be better off with twelve kalashnikovs!’

Existing studies of the Uyghur Twelve Muqam

In addition to the official transcriptions and recordings of the Twelve Muqam, there is a wealth of published studies in the Uyghur and Chinese languages, which I will draw on throughout this book. The most useful Chinese-language sources are the numerous books and articles by the musicologist Zhou Ji, former head of the Xinjiang Arts Research Unit, whose writings are based on many years of fieldwork and the experience of transcribing the whole repertoire as developed by the professional groups in the 1980s (Zhongguo 1996; Zhou 1995, 1998, 2001, 2005, 2006a, 2006b). There are some useful collections published in both Chinese and Uyghur (On ikki muqam 1992). Uyghur writers working within the Xinjiang region have tended to focus more on historical and textual aspects (see, for example, Teklimakaniy 2005), engaging in the internal polemics surrounding the repertoire (Imin 1980; Ötkür 1992). Uyghurs based in the Central Asian states have provided analysis and transcriptions of the shorter ‘Ili variant’ of the Twelve Muqam (Khashimov 1992) and have been more able to reflect on the political processes of their canonisation (Ärshidinov 2002).

Several Western-language studies are also available. Colin Mackerras provides one of the earliest English language introductions to the Twelve Muqam and other Uyghur performing arts (1985). The article is not written from a musicologist’s perspective but it provides an interesting reflection of the discourse of the period: the assumptions and attitudes which are widely shared in Uyghur and Chinese language The Making of a Musical Canon in Chinese Central Asia publications within China. Thus we read that the Twelve Muqam have ‘extremely ancient’ roots in the 4th-century music of the Buddhist kingdom of Kösän and that they are no less than the source of other maqām traditions across the Islamic world, and we puzzle over the strange dichotomy of the Uyghurs’ love of song and dance in spite of the proscriptions of their supposedly ‘anti-music’ Islamic faith.

During & Trebinjac (1991) provide the first Western-language attempt at detailed musical analysis of the Twelve Muqam, focussing on structure, rhythms and mode. Their analysis is based on an early set of Chinese transcriptions (Shinjang 1960), and recordings made in Uzbekistan in the 1970s and 1980s (Ministerstvo Kulturi SSR, no date). Their booklet provides a useful starting point for this discussion, but it is problematic in that it produces a reductive analysis of the Twelve Muqam based on primary sources which have since been criticised and quietly discarded. Musicologists in Uzbekistan and Xinjiang concur that the 1960s transcriptions often bear little discernible relation to the recordings on which they were meant to be based, still less to contemporary professional practice. Since the time of During and Trebinjac’s analysis they have been replaced by the plethora of new recordings and transcriptions which have appeared since the early 1990s. Trebinjac’s more recent and wide-ranging book (2000) provides further consideration of the formal aspects of Uyghur Muqam traditions and more contextual detail as part of a broad survey of Uyghur music, within her wider argument concerning the politics of appropriating minority musical traditions for Chinese modern composition.

Nathan Light’s PhD dissertation (1998) is an in-depth study of the Twelve Muqam texts, the ‘classical’ ghazal of the Chagatay poets and the many ‘folk’ beyit or couplets. As far as I am aware, the first scholar to discuss the Twelve Muqam in terms of ‘canonisation’, Light draws on interviews, historical sources and textual analysis to provide an illuminating account of the work of revising and fixing the Twelve Muqam over the past fifty years under the People’s Republic of China. I am delighted that the long-awaited revised version of Light’s PhD is soon to appear in print with Lit Verlag/Transaction Publishers. In a brief online publication, James Millward has contributed some interesting perspectives on the political implications of the project. Wong Chuen-Fung’s PhD thesis (2006a) takes a fresh look at issues surrounding the revising of the Twelve Muqam; these are usefully summarised in his article (Wong 2006b).

The present book is indebted to these earlier studies. It draws on these and other published sources in European languages, Uyghur and Chinese, interviews with musicians and musicologists, and field, archive and commercial recordings, towards an understanding of the Twelve Muqam as repertoire, juxtaposed with an understanding of the Twelve Muqam as discourse, or what might be termed the field of Uyghur muqamology. In the next part of this introduction I consider the nature of musical canons and the processes of their formation with reference to studies of the Western classical canon, 20th-century China and musical canons across the Islamic world. Chapter One provides an overview of Uyghur music, genres, instruments and contexts. Chapter Two presents a historical survey of the process of canonisation of the Twelve Muqam, bringing together personal accounts gathered during fieldwork and published material from Xinjiang and the Central Asian states. Chapter Three brings the focus to the personal level, with a biographical account of one actor in the canonisation process. This chapter is a revised version of the chapter ‘Abdulla Mäjnun: Muqam expert’, which appears in Lives in Chinese Music (Rees 2008). Chapter Four discusses the debates which are carried on within Uyghur professional musical circles regarding the canonisation process, and uses comparative analysis of published recordings of the Twelve Muqam to consider the nature of the repertoire, questions of modal character and variation between different versions. Chapter Five situates the Twelve Muqam repertoire within the context of maqām traditions across Asia, and more directly within the sphere of Central Asian musical traditions, considering musical structures and performance contexts. This chapter is a revised version of a previously published chapter which appears in the book Situating the Uyghurs between Central Asia and China (Bellér-Hann, Cesàro, Harris & Smith- Finley 2007). Chapter Six considers the canonisation process within the broader context of music-making across the region. Drawing on comparative analysis of professional and locally maintained traditions of Twelve Muqam, it discusses the impact of government-supported efforts at canonisation, and the impact of the independent recording industry. This chapter also considers new developments following the 2005 inclusion of the Uyghur Muqam within UNESCO’s third proclamation of masterpieces of oral and intangible heritage.

The book includes an appendix which contains brief notes on the accompanying CD, which was recorded by Abdulla Mäjnun during a brief residency in SOAS in 2003, supported by the AHRC Centre for Music and Dance Performance. As this residency took place before the structure of this book was fully conceived, the CD is not a straightforward illustration of the transcriptions provided in the book but rather a musical portrait of one individual musician and player in the canonisation process. As such it provides an audio complement to Chapter Three. Some of the tracks on the CD also serve as material for comparative analysis in Chapters Four and Six.

Musical canons

Across the world, and across historical periods, processes of canon formation are linked to political power, and especially to the rise of new forms of political power. In a historical survey of the Western classical tradition, William Weber offers a definition of musical canons as: ‘the presentation of old works organized as repertories and defined as sources of authority with regard to musical taste’ (Weber 2001: 339). Weber argues that the musical canon may take on moral, spiritual and civic force. He links the formation of the Western classical canon in the 18th century to the rise of the public as a political force independent of the monarchy, a period in which cultural life in general, and music in particular, played a central role in establishing new definitions of community (Weber 2001: 352). This focus on identity, and in particular national identity, has been central to studies of processes of canonisation in other parts of the world, where definitions of community involved particular and well-theorised processes. The formation of musical canons outside Europe, primarily during the course of the 20th century, has been widely linked to the rise of the nation state. Just as we are accustomed to thinking about nations and their histories in terms of Eric Hobsbawm’s model of invented tradition (Hobsbawm & Ranger 1983), so too it is common to approach their musical canons as repertoires whose purported completeness and deep historical roots are revealed on inspection to be contemporary constructions, or at least the fruit of new ways of imagining the past.

Philip Bohlman has argued that most canons are products of ‘bricolage’. He discusses the formation of canons as processes involving multiple agents who make multiple choices, selecting and revising according to certain (often contested) criteria, including or excluding particular melodies and texts. Through such processes the past is appropriated in the present and preserved for the future: ‘Models of the past are important, and where real models are not present, surrogates and imaginary models will do just fine’ (Bohlman 1992: 203–4). This understanding now informs contemporary approaches even to the Western canon. Weber cautions that ‘the ideological burden of the classical music tradition – its effort to enforce its authority – makes one think that there was a single, identifiable list, but upon inspection we find a great variety of practices at any one time in different contexts’ (Weber 2001: 347). More recent studies in ethnomusicology, informed by post-colonial theory, have sought to develop new ways of thinking about such tensions between tradition and modernity, impelling a closer focus on the actions of individual musicians, researchers and audiences as they negotiate between different modes of identity (local, ethnic, national, cosmopolitan or diasporic), and a focus on the social and performative spaces in which these actors move (Turino 2000; Rice 2003).

Musicologists working on the Western classical tradition, perhaps predictably, have been more interested in the repertoire itself, seeking to understand why and how certain musical sounds come to embody the moral, spiritual and civic force that Weber describes. Dismissing the fundamental claim that canons exhibit transcendental and objective values, Mark Everist asks: how are canons determined, why, and on what authority? Everist also probes the role of the audience and their reception of the canon, asking how the relationship between reception, canon and value works (Everist 2001: 389–93). Such questions have inherent cross-cultural interest, and we might fruitfully ask if there are any specific musical traits which are prerequisite for canons across different cultural contexts. If we can identify such traits – for example modes of reception, performance venues, the tendency towards large ensembles or fixed compositions – then to what extent do they draw on the Western classical canon, or rather on perceptions of this canon, as a model?

China, modernity and national minorities

Katherine Bergeron argues that ‘discipline’ is fundamental to canons, linking the ideology and practice of the Western classical tradition, such as playing in tune or practising scales, directly to broader notions of modelling behaviour and social control (Bergeron 1992: 3). The ideology of discipline is immediately apparent in the debates surrounding music in early 20th-century China, and it is here that we may find the seeds of the canonisation of the Uyghur Twelve Muqam being sown. During this period, equally under pressure from Western and Japanese military incursions, and inspired by Western and Japanese cultural models, China’s so-called ‘May 4th’ reformers and modernisers were calling for reform of virtually all aspects of Chinese culture and society. Music did not escape their attention, and numerous articles were published in China’s music journals introducing Western harmony and instrumentation, making comparisons between Western and Chinese music, and discussing China’s ‘national character’ as revealed in her traditional music. An excerpt from an article by the 1930s reformer Ying Shangneng gives an idea of the tone:

In days gone by China had her own great music. But judging from what we find of it nowadays, nothing remains to remind us of its past glory. … Her musical instruments … are quite crude and simple … no standard pitch can be found among them. No two flutes made by the same hand can be made to play in unison harmoniously. The scale steps are also found to vary … No wonder, therefore, in the annals of Chinese music, there is no Beethoven or Schubert. Andrew Jones argues that what these May 4th reformers demanded was the disciplining of Chinese musical life along Western or ‘scientific’ lines: ‘At the very moment that Chinese music became an object of study, classification, and rationalization along Western lines, it also came to signify Chineseness’ (Jones 2001: 40). This ideology, tinged red, carried over into the establishment of the People’s Republic (1949–), and culminated in the formation of China’s most famous revolutionary musical canon, the ‘model operas’ (yangbanxi) of the Cultural Revolution, while the search for a Chinese Beethoven produced such oddities as the ‘revolutionary composer’ or blind street musician A Bing (see Stock 1996). China’s minority nationalities were equally subject to this ideological approach. Under the systematising pressure of the Chinese state, in the 1950s each of China’s newly designated 55 minority nationalities developed one representative art form, singled out for state support and development. As Helen Rees argues:

The show-casing of these designated music and dance forms has played an important role in China’s formulation as a multi-cultural state. Yet (as Andrew Jones argues above) just as these forms were harnessed to perform difference on China’s national stage, they began to achieve sameness through being subjected to similar processes of reform. Subject to processes of ‘improving and ordering’ (jiagong, guifan), dance styles were transformed into group choreographies, songs were transcribed and fixed, scales and musical instruments standardised, and a nation-wide system of professional performers was put in place, trained in arts academies, and organised into state-sponsored performing troupes. Clearly not all of these designated minority musical forms with their faux-naïf folkloric appeal can lay claim to the weightiness which we tend to associate with canons, yet the processes of their creation bear many of the hallmarks of canon formation. It is important to see the Uyghur Twelve Muqam as one among these 55 designated minority cultural assets. However, the Uyghurs, as I have noted, are also culturally Central Asian Muslims and musically very much part of the Islamic world. By looking westwards beyond China’s borders to nation states across Asia and North Africa we can find closer parallels with the formation of their canon.ic

Processes of canonisation across the Islamic world

In her recent book, Ruth Davis describes the case of the Tunisian Ma’luf, a tradition whose canonisation ran along a similar timeline to the Uyghur Twelve Muqam, and in which it is possible to trace many correspondences with the Uyghur case (Davis 2004). The Ma’luf, a set of large-scale suites formerly patronised by the elite but widely performed in Sufi lodges, coffee houses and weddings, was first transcribed in the 1930s, impelled in part by the enthusiasm of the Frenchman Baron d’Erlanger to preserve and purify traditional Tunisian music. The establishment of the Rashidiyya Institute in 1934 marked the beginning of the canonisation of the Ma’luf. Supported by the French colonial government and modelled on the French conservatory, the Institute sought to replace the traditional small Ma’luf ensemble with a larger orchestra. It quickly became apparent that the musicians drafted into the orchestra were performing differing interpretations of the repertoire, and the move from small to large ensemble, remarks Davis, was marked musically by a move from heterophony to cacophony (Davis 2004: 51). Perceiving the need for notation to produce a unified performance, the Institute produced a composite version of the Ma’luf, drawing together the versions held by several different musicians. Davis argues that this was developed purely as a practical measure in pursuit of the creation of a large ensemble performance; the ideology of a sole correct version only took hold later (Davis 2004: 109).

Following independence in 1956 the tradition was promoted nationally, and successive layers of transcriptions and recordings were produced over the subsequent decades. Through this period Davis notes two competing ideologies at play. The first argues that the Ma’luf represents a unified national heritage, and notation should be deployed to define and restore the authentic tradition. In this reading, oral tradition results in superficial deviation from the norm over the course of centuries. The second more liberal argument holds that the multiple, regional orally transmitted traditions are legitimate; each generation defines its own interpretation, hence the need for regular revision of notations to keep the tradition alive (Davis 2004: 67).

We find a similar set of circumstances in 1960s Iran, under the rule of the Shah, when the government attempted to establish the definitive Radif, the repertoire which forms the basis for improvisation in Persian art music. Although they do not frame their discussions specifically in terms of canonisation, several major Englishlanguage studies deal with these issues in relation to the Radif. Bruno Nettl describes the Radif in paired concepts: it is both repertoire and theory on which performance is based; a contemporary version of a centuries-old way of making music throughout the Middle East and a coherent system developed recently by a small group of individuals. One may speak of the Radif, he argues, or hold that there are as many Radifs as there are master musicians, and each of these may have several variants (1987: 3).

The origins of the Radif are generally thought to lie in the 18th century, but little is known about musical practice in that period. Its canonisation owes much to one individual: Mirza Abdullah (d. 1917), who collected and classified the repertoire handed down by his father. This Radif of Mirza Abdullah is considered the basis of the contemporary mainstream tradition. It was not until the early 1960s that the Radif became subject to state intervention. Ella Zonis describes how a panel of the country’s leading musicians was chosen to prepare the official Radif. She opines:

Anyone who is familiar with Iran, or with any culture where values of individuality are prized over and above collective thinking and where artistic independence is the chief merit of artistic performance, would recognize that the chances of this group’s ever reaching agreement were remarkably slim. (Zonis 1973: 63)

This panel was indeed soon disbanded and replaced by another panel comprised of leading musicologists who, it was hoped, would be more scientific and objective. In fact there was even less agreement between them, and finally the task was given to one individual, the prominent musicologist Musa Ma‘rufi, whose completed transcription of the Radif was lavishly published by government in 1963. Zonis considers this version to be ‘solely the work of the transcriber’, and although it has acquired a degree of authority in subsequent decades, as in the Tunisian case, rival versions of the Radif are still fiercely contested by Iranian musicians today (Nooshin 1996).

The French musicologist Jean During has drawn direct parallels between the Iranian Radif and the Bukharan Shash Maqām (During 1993). These six prestigious ‘classical’ suites of Bukhara, the best-known Central Asian maqām tradition, are probably the closest model of canonisation to the Uyghur Twelve Muqam. In the Central Asian context the processes of canonisation brought to bear on this tradition are usually thought of as products of Soviet cultural and nationalities policies. A few existing studies of these repertoires, however, give us a sense of the historical depth 10 The Making of a Musical Canon in Chinese Central Asia of the processes of canonisation in this region. During reads the creation of the Shash Maqām as a political symbol, but one which predates Sovietisation and 20th-century nationalism, and which relates instead to the power of the city-state, in this case the 19th-century Emirate of Bukhara:

Le shash maqām boukhārien également est un monument incontournable, ‘mise en ordre’ par des musiciens qui en dirigeaient et obéissaient à l’émir, édifié à l’image d’un ordre central autocratique puissant. (During 1993: 35)

During’s use of the term ‘monument’ to describe the Shash Maqām is interesting, and impinges directly on questions relating to the preservation of ‘oral and intangible heritage’ now promoted through UNESCO. I will return to these questions in Chapter Six. It was this already monumental repertoire that the Soviet musicologist, Yunus Rajabi transcribed and revised in the late 1920s, selecting what he considered to be the ‘most authentic’ of numerous renditions, even synthesising his own versions from parts of different versions. Rajabi’s Shash Maqām transcriptions formed the basis of teaching in the Uzbek conservatory in subsequent decades. They have been preserved into the post-Soviet era and enshrined as an Uzbek national tradition, in a style which the American ethnomusicologist Theodore Levin has critiqued as ‘frozen music’ (Levin 1996: 47–51).

Continuing this theme of pre-20th-century canon formation, I turn now to Walter Feldman’s discussion of 17th-century Ottoman court music (Feldman 1996). Feldman argues that the development of cyclical concert forms in the Ottoman court during the 17th century distinguished this repertoire from the Iranian tradition which the Ottoman regarded as its forebear. By the mid 18th century all Ottoman classical genres were arranged cyclically, with separate courtly and Sufi cycles (fasil and ayin). Each cycle contained composed and non-composed sections. Rhythmic formulae (usul) played an important role in the composed elements of the cycles, and these became increasingly long and complex (Feldman 1996: 177–92). The development of this repertoire seems to bear many of the hallmarks of 20th-century canonisation. Feldman does not devote much space to a consideration of the social and political context which nurtured these musical changes, but he does comment briefly that the wealth of the state was a factor (Feldman 1996: 503). Reminiscent of Levin’s critique of the ‘frozen’ Shash Maqām, Feldman comments that by the early modern era this ongoing process of tempo retardation and melodic elaboration meant that many measured genres had acquired slow and ponderous rhythmic structures such that this became the hallmark of the Ottoman repertoire (termed vukur ‘dignity’ or aghir bashlik ‘seriousness’; Feldman 1996: 499).

Feldman’s work also allows us to see how canons which are closely tied to the old order may suffer under the new. The rise of Turkish nationalism and the establishment of the Turkish state in the early 20th century led to the sidelining and direct criticism of the musical canon of the fallen and discredited Ottoman Empire. The influential Turkish nationalist writer Ziya Gökalp argued in his 1923 book that the ‘Eastern’ (in other words, foreign) music of the Ottoman had remained confined to the elite while the Turkish lower classes got on with creating an authentically national popular music free from outside polluting influence. Between 1935 and 1945 the Ottoman classical repertoire almost disappeared from radio and public performance, and teaching opportunities were severely limited (Feldman 1990: 98–100).

These ideological attacks in Turkey have direct echoes in Soviet Uzbekistan where, during more extreme periods of Sovietisation, the Shash Maqām were attacked as ossified examples of foreign music, imbued with bourgeois-feudal ideology (Levin 1984). Yet the tradition survived these attacks, and gained new state support as a symbol of the Uzbek nation following the fall of the Soviet Union. In Tajikistan, however, the Shash Maqām were too closely associated with one region (the northern power-holders) to survive the ensuing power struggle between regional factions. When the other (southern) regional faction took power in the late 1990s, the Shash Maqām were duly sidelined.

More recent developments around the Ottoman repertoire demonstrate that canons may have staying power beyond the direct patronage of the state. In recent years the Ottoman classical repertoire has enjoyed a revival, quite independent of state support, which suggests the strength of the aesthetic and other values which canons imbue. It is interesting that the style of this revival, tending to small ensembles and a greater use of heterophony, is in direct contrast to the weightiness and dignity described by Feldman. Similar developments in the 1990s have been described by Davis in Tunisia (2004: 105), and noted (and indeed actively promoted) by Levin and During in post-independence Uzbekistan.

The historical studies of Feldman and During remind us that processes of canonisation are not solely the product of 20th-century nationalism, nor necessarily a reaction to the West. Instead, we might read the impulse to canonisation, including that which took place in Europe in the 18th century, more broadly as part of the political process of centralisation and consolidation. We have seen that some canons survive the rise and fall of regimes, while others are too strongly linked to the old regime and are dropped or attacked by the new. Several musical themes, briefly mentioned, are recurring – size and complexity of form and ensemble, a tendency towards fixity and ponderousness – yet these are by no means present in all the traditions surveyed here; indeed, these canons run from open-ended collections of compositions to a repertoire (the Radif) a significant proportion of which is not performed as such but serves as the model for improvisation.

Although the phenomenon of canonisation stretches back beyond the 20th century, we can identify tendencies that are shared by canons produced during the 20th century which are less clear-cut in earlier models. The Radif, the Ma’luf and the Shash Maqām in the 20th century all share the common features that they are understood as national heritage, they exist in multiple variations, they are the product of oral tradition, and that the thrust of the canonisers’ work has been to unify and fix tradition. Another recurring theme in these accounts is the degree to which the canonising projects are contested from within. Most importantly, the 20th-century canons, unlike their court-based predecessors, are much more widely promoted through state institutions: they are performed by state-supported orchestras or troupes whose members are trained in state colleges and conservatories, and disseminated through live performance, TV and radio, publications and recordings. They thus have more direct impact on the practice of music beyond the state-supported sphere. All these themes will be further explored in my account of the canonisation of the Uyghur Twelve Muqam.

Notes: 1-For influential critiques of China’s system of minority nationalities (shaoshu minzu) see Gladney (1991: introduction) and Harrell (1995). The Uyghur population in Xinjiang was given in the 2000 census as 8,139,458. Since then the region’s population figures have not broken been down by ethnic group.

2-Chagatay is the literary Turkic language of medieval Central Asia.

3-James A. Millward, ‘Uyghur Art Music and the Ambiguities of Chinese Silk Roadism in Xinjiang’ (accessed 10 July 2007).

4-Copyright 2008 by the Board of Trustees of the University of Illinois. Used with permission of the University of Illinois Press.

5-The landmark May 4th 1919 demonstration against an unequal treaty following the First World War gave its name to the subsequent reform movement in China.

6-Benjamin Z.N. Ing, ‘Music Chronicle’, T’ien Hsia Monthly 1 (January 1937), 54. Quoted in Jones (2001: 39).

7-See Rees (2000: 19–27) and Harris (2004: 1–15) for more detailed treatment of this subject.

8-‘The Bukharan shash maqām too is an inescapable monument, “put in order” by musicians under the direction and control of the emir, built in the image of a powerful centralised autocratic order.’

9-Feldman prefers the term ‘cyclical form’ over ‘suite’. The concept of cyclicity (tsikl’nost’) was developed by the Soviet musicologists Rajabi and Karamatov in relation to the Shash Maqām (Feldman 1996: fn 60).

 Author: Dr Rachel Harris

(Department of Music, SOAS, University of London, UK)

(Introduction and Chapter One and Two)


Putting the Uyghur Muqams in Historical and Cultural Context

In order to understand the Uyghur muqams and their performers, we must understand Turkic and Uyghur cultural history within its Central Asian context. In this dissertation I explore the recurrent ideological issues that Turkic and Uyghur authors and performers have expressed in their art. In particular, I examine how these artists expressed their ideas about shared identity and culture in their art.

In Chapter One I introduce the geography and history of Eastern Turkistan, now known as Xinjiang, and my own first encounters with the Uyghur muqams. I give a basic introduction to the musical and verse structure of the Uyghur muqams themselves, and briefly compare them with other maqâm traditions. I introduce the conceptual categories scholars and politicians impose on the muqams to help construct a shared Uyghur culture.

In Chapter Two I describe the ambiguous distinctions between Uyghur and Turkic culture through historical narratives. This establishes the background for my story of the course of Turkic literature.

There is a neat symmetry to the history of Turkic literature. The early encounters of Turks with Islamic literary culture stimulated Turkic authors’ attempts to create distinctively Turkic written literature. Over eight hundred years later, encounters of Muslim Turkic authors and performers with “modern” ideas about ethnic identity are stimulating their efforts to integrate a long tradition of Muslim internationalism with new ideas about shared Uyghur linguistic, literary and musical culture. Uyghurs now are working to create a strategically modern ethnic nation within China. The Uyghur muqams are useful as evidence of a shared ethnic history and culture, entirely separate from the Chinese, and heavily engaged with cultures from Central Asia, South Asia and the Middle East.

The recurrent problem that I discuss across this vast span of literary history is how individual performers and writers express and define collective culture. Mahmûd al-Kâshgharî does this with a dictionary of the Turkic language in which he hopes to define and preserve culture that he perceives as authentically Turkic in the face of the strong Persian influence brought by Islamization. After Kâshgharî, Turkic authors use more and more Persianate literary forms, but continue to infuse them with identifiably Turkic styles.

Although Turkic literary tradition remains distinctive in many ways, most Turkic authors after the eleventh century do not make much of an issue of ethnic identity in their poetry. The major authors of Turkic literature from the twelfth through the nineteenth century are more concerned with expressing Islamic and Sufi ideas and images than distinguishing Turkic culture from that of other Muslims.

Although none of the early poetry from Kâshgharî, Yûsuf Khâss Hâjib or Ahmad Yüknäki appears in the Uyghur muqam songs, in Chapters Three and Four I consider these three authors in detail to show how the politics of Turkic culture shaped early Muslim Turkic literature, and to show the development of Muslim themes in Turkic literature. These authors participated in the spread of the Muslim beliefs and knowledge that become the core of Turkic literature for the next eight hundred years. Even after modern editors’ changes these themes remain central to the Uyghur muqam songs.

In Chapter Five I discuss the Sufi poets from Ahmad Yasavî to Khoja Nazar Huvaydâ whose poetry dominates in Central Asian Turkic literary tradition. I discuss the Sufi imagery that these authors all use, as well as looking at the lives and polemical writings of poets such as Mîr `Alî Shîr Navâ’î and Bâbâ Rahîm Mashrab. These two poets are at opposite ends of the spectrum in terms of lifestyle, yet their poetry requires careful reading to find the distinctions. I analyze Navâ’î’s writings in which he defends Turkic poetic style and language to show that there is an ongoing but subdued sense of Turkic identity. I show how both Navâ’î’s work and Sufi oral poetry use Turkic didactic proverbs.

In Chapter Six I return to the theme of ethnic national culture in my consideration of how music has come to be seen as a cultural gift from Uyghurs that has circulated throughout Eurasia over the past two thousand years. I discuss how historical narratives in several media have been very important for dividing up the past and linking it to present collective identities as heritage. More than anywhere else, it is in these discussions of music history that Uyghur and Chinese ideas about “modernity” are expressed. Public representations are part of a subtle but important debate over who has rights to the past, who can claim to be culturally prior to others, and how history legitimates rights to resources and identities in the present.

I use texts and other artistic and discursive representations to reveal ideological issues for the authors and original audiences, and the ways meanings have changed in the modern context. In Chapter Six, I begin with close contextual readings from the contemporary narrative artworks of Ghazi Ämät, and end with an analysis of the relationship between an 1854 hagiography of musicians and the 1983 drama based on it written by former Xinjiang Chairman Säypidin Äziz. Likewise, all the literary works that I analyze are not simply historical artifacts, but have ongoing significance as within Uyghur understandings of their historical and cultural traditions.

Moving from texts to living people in Chapter Seven, I have the good fortune of learning about creating the muqams from Ömär Akhun, one of the few traditional performers. His descriptions of the muqams as performing art gives life to the literary texts and musical forms I have described, revealing the muqams as creative acts. I compare Ömär Akhun’s conflicts with the ideologues who want to modernize, secularize and nationalize the muqams, to Mahmûd al-Kâshgharî’s resistance to the non-Turkic cultural influences that came with Islamization in the tenth and eleventh centuries.

In Chapter Eight I discuss how the muqam texts were transcribed and published, and the ideas of the editors who have transformed them into the modern Twelve Muqams. I compare the recordings of Turdi Akhun–the traditional performer whose Kashghar muqams are the primary source for the present version of the Uyghur Twelve Muqams–with my own recordings of Ömär Akhun. My results show that both performers had similar techniques for creating muqam songs in performance from memorized ghazal verses and couplets. I also compare Turdi Akhun’s methods of singing literary ghazals to his rather different way of performing a cycle of folk couplets.

Through comparison of Turdi Akhun’s texts with those in the modern Muqams, I am able to further detail the work of the editors. Like my readings of historical representations, my reverse editing of these texts refines my understanding of the ideological goals of the modernization process.

Finally, in Chapter Nine, I discuss the Dastan and Mäshräp sections of the muqams as Turdi Akhun sang them, and as they have been edited. I show that here the editing has eliminated many of the repetitions that resulted from Turdi Akhun’s method of composition, and has reduced the religious references. I do a close comparison of Turdi Akhun’s use of quatrains from various editions of the Gherip-Sänäm dastan to the quatrains used in the Ili muqam tradition, and show that in the editing process most of Turdi Akhun’s songs have been replaced by other verses, many of which were originally sung in the Ili muqams. My analysis also shows one of the ways that dastans were orally performed.

Chapter One

Turkic History and the Uyghur Muqams


Mountain ranges, rivers and deserts divide Eastern Turkistan, now known as the Xinjiang Uyghur Autonomous Region, into three distinct regions–northern (shimaliy), eastern (ghärbiy), and southern (janubiy)–and shape the ecological areas within these regions. The political subdivisions of modern Xinjiang are more complex and fragmented than these basic regions: nominally autonomous districts and counties are associated with the Qazaq, Qirghiz, Hui, Mongol, Tajik, and Xibo nationalities. Because some of these autonomous areas in fact include highly mixed populations and ecosystems, they do not offer the conceptual clarity that makes the division of Xinjiang into regions such a wide-spread way of thinking.

Also important is the ecotype division of Eastern Turkistan into grassland and oasis: Qazaqs, Qirghiz, Mongols, and Xibo have adapted their cultures for life within the mountain pastures and steppe grasslands, while the oases and agricultural river valleys are occupied primarily by Uyghurs. More recently, the Han and Hui ethnic groups have come to dominate in the more industrialized cities and large government-run farms, mines and petroleum operations. All these areas are ethnically mixed, but pastoral, agricultural and industrial economic activities and urban or rural habitation are important aspects of ethnic and personal identity in Xinjiang.

Northern Xinjiang comprises the area north of the Tian Shan mountains. Most of this is the arid steppe grasslands of the Jungarian basin, separated from Mongolia and Tuva by the mountains of the Altai range to the east and north, and from Kazakstan by the Tarbaghatay and Alatau ranges to the west. Although the foothills of all these ranges include significant towns, the southern edge of this area, in the northern foothills of the Tian Shan, contains the most important agricultural areas and cities: southwest of the Jungarian region is the fertile Ili river region, between the Alatau and Tianshan mountain ranges. Ghulja is the main city of the Ili region. To the south of the Jungarian basin is the capital city of Xinjiang, Ürümchi, at the main crossroads between the southern, eastern, and northern regions.

Eastern Xinjiang includes the cities of Turfan, just over the Tian Shan mountains from Ürümchi, and Qumul (Hami), roughly four hundred kilometers to the east. The entire region south of the Tian Shan is desert, but mountain drainage supplies water to Turfan and Qumul and a number of other oases that form a chain around the desert.

Southern Xinjiang consists of the oases surrounding the Taklamakan Desert in the Tarim Basin.(1) Known as Kashgharia or Altishahr (Six Cities), this region has long supported settlements with close cultural ties to India and Iran, as well as to the nomadic peoples of the Central Asian steppes. Turkic speakers settled and became dominant in this region towards the end of the first millenium C.E. The major towns and cities include Khotan, Keriya, Niya, and Cherchen along the southern edge against the Kunlun range, and Kashghar, Yarkand, and Artush on the west on the edge of the Pamirs. Along the northern edge of the Tarim basin, in the foothills of the Tian Shan are the cities of Aqsu, Kucha, and Korla.

Today, Uyghurs make up a little less than half the population of Xinjiang. Han Chinese in Xinjiang are almost equal to the Uyghurs in number. The other important ethnic groups include the Qazaq, Qirghiz, Mongol, and Hui. In south Xinjiang, the dominant population and language is Uyghur, with several local dialects that are rapidly disappearing. Most people of other ethnic groups in south Xinjiang, including the Han Chinese, rely on Uyghur as the lingua franca.(2)

The Ili region has a large Uyghur population dating back to agricultural colonists brought there in the late eighteenth century after the Manchu Chinese conquest of Xinjiang drove Qalmaq Mongols from the area. Until the twentieth century these Uyghurs were known as Taranchi, from the Mongol word for ‘farmer.’ There are Uyghur communities in the towns throughout north Xinjiang, but they make up only about one-sixth of the population, while the Han and Qazaq each make up around one-fourth, and the Qirghiz, Hui and Mongol ethnic groups each make up less that ten percent. Many Uyghurs have moved from throughout Xinjiang to the capital city, Ürümchi, as well, but they still only make up about one-sixth of its one and one-half million people.

Uyghurs have long lived in east Xinjiang but there are large Han and Hui communities as well, and these areas have stronger historical connections to inland China. In Qumul, Chinese has become the dominant language, although Uyghurs still make up a considerable proportion of the population. Turfan is still a largely Uyghur city, although culturally distinct from the cities of south Xinjiang.

South and north Xinjiang have long moved in and out of Chinese influence. The first Chinese conquests in this region occurred during the Han dynasty (206 B.C.E.-220 C.E.), which politically unified China in defense against the Xiongnu nomadic empire. The Han dynasty established Chinese military settlements in the Tarim Basin and Jungaria to guard its border regions, which along with the states of Central Asia, became known in Chinese as Xiyü ‘the Western Regions.’ After the fall of the Han dynasty, China lost control of Anxi, ‘the pacified West,’ and had to establish their border posts closer to the Chinese heartland. Although the Chinese had extensive ongoing tribute and trade relations with the kingdoms of the Western Regions, only in the Tang dynasty (618-907 C.E.) did Chinese forces again invade parts of the Tarim and Jungaria. The Tang dynasty faced continual struggles with Tibetans, Turks and Uyghurs over the Western Regions.

After the Tang Dynasty, it was not until the Mongol conquests that China was again a part of the same empire as the Western Regions. Under the Turko-Mongol descendants of Chingiz Khan, most of the Tarim and Jungaria were part of the domains ruled by Chaghatay and his descendants until the mid-fourteenth century, when the more conservative Moghul rulers of the eastern Chaghatayid domains broke away from the sedentarizing Mongols of the Ulus of Chaghatay. Moghulistan remained under the rule of Turkic-speaking Mongol princes until the seventeenth century when Qalmaq Mongols took control of northern Xinjiang (Jungaria), and helped establish the leader of the White Mountain (Aqtaghliq, Afaqi) Khoja clan as the ruler of Kashgharia.

In the 1750s, China under the Manchu Qing Emperor Qian Long again invaded the Western Regions and established political and military control. Many exiles and refugees from the Qing conquest settled in the Qoqand Khanate of the Ferghana region west across the Pamirs from Kashghar. These exiles supported frequent rebellions against the Qing Empire during the 1800s, when Eastern Turkistan became the object of power struggles between Qoqand, China, Russia, and Britain. In 1864 when the Qing rulers were preoccupied with rebellions and imperialist encroachments in the heartland and coastal regions of China, Muslim rebellions largely organized by the Tungan (Hui) minority began in cities throughout Kashgharia and Ili regions. In 1865 Alim Quli, the Khan of Qoqand, sent Ya`qûb Beg at the head of a small army to put an Aqtaghliq Khoja leader on the throne in Yarkand as a puppet governor for Qoqand. Between 1865 and 1867 Ya`qûb Beg managed to take control of the cities of Kashgharia from the Qing, Tungan and Turki Muslim forces. Under his rule this region was fairly prosperous, but when he attempted to extend his rule north into the Ili and Jungarian regions, and east to Turfan and Urumchi, he was far less successful, and the resulting wars devastated these regions.(3)

Russia took control of the Ili region for a period in the 1870s. After regaining control of the Ili region and Kashgharia in the early 1880s, China established closer economic and political links to the Tarim and Jungarian areas. The Western Regions were renamed Xinjiang (‘New Territory’) and united into a province in 1884, and stronger Chinese military control was imposed. But China’s political weakness continued, and in the early twentieth century, Xinjiang was largely ruled by Han and Tungan warlords. In 1932-34, Uyghurs in Kashgharia rebelled against their nominally Chinese rulers and briefly established an independent Islamic republic, while in the Ili, Tarbaghatay and Altai districts of northern Xinjiang people of several nationalities rebelled against Chinese rule and founded the Eastern Turkistan Republic (1944-1949).

When Xinjiang was incorporated into the People’s Republic of China in 1950, autonomous counties and districts were established for areas with dominant ethnic groups. In 1955, Xinjiang itself was made into the Xinjiang Uyghur Autonomous Region. Under policies promoting economic, agricultural and military development of China’s border regions, hundreds of thousands of Han Chinese settlers came or were sent from “inland” (neidi) China to Xinjiang. This began around 1955, but intensified in the 1960s with the railroad reaching Qumul (Hami), and during the Cultural Revolution when many urban youths and intellectuals were sent to the countryside.


My first trip to Ürümchi, the capital of the Xinjiang Uyghur Autonomous Region, was in May of 1989, during the massive student protests that changed the attitudes of many younger Chinese towards the government from one of docile obedience to activism and hope. On my way to Ürümchi, I visited former students of mine in Chengdu who had been suddenly galvanized by confidence in their ability to change China’s entrenched bureaucracy and end its grip on their destinies. They were optimistic and excited, no longer willing to think of China as inferior to other countries. Later in Ürümchi when I heard the news that the Chinese government had crushed the student protests in Beijing, Chengdu and many other cities, I saw my students’ hopes and confidence shatter, their dream of Chinese political progress collapse. The oligarchs were not willing to give up their power over ordinary Chinese lives. But in Ürümchi, news reports of the military suppression of the “counter-revolutionary student movement” and renewed attention to loyalty to the Communist Party and the People’s Liberation Army were seen as just one more official campaign.

In my third week in Ürümchi, my friend Hairet took me to a Uyghur wedding celebration at the Agricultural College. The party was in a large dining hall rented for the occasion. Men sat in groups of nine or ten around tables along the wall near the windows, while women sat at tables along the inside wall. The dining hall employees began to serve the main courses in the late afternoon. These consisted of polo (pilaf made with mutton, rice, carrots, onions, and raisins) and manta (steamed dumplings filled with mutton and rice), along with endless dishes of pastries, nuts and dried fruit, nan (flat bread), and bowls of tea. Spicier stir-fried food was set out towards the end of the meal, as an accompaniment for drinking the grain liquor aq haraq, more commonly known by the Chinese term baijiu.

While the drinking was still going on, musicians began to perform. Hairet pointed out that some of the musicians were Han Chinese, which he feared would mean that the music would not be appreciated by Uyghur wedding guests. In the end he decided that the music was good anyway, since the singer had been a Uyghur, and since the Chinese learned music from reading it, rather thanaurally as Uyghurs do. The instruments included a guitar and electronic keyboards, and the songs were popular waltz tunes. But it was dancing that was the main attraction for the guests.

Between the women’s side and a men’s side was a large dimly lit dance floor. As soon as a song began, the men would get up from their tables and move rapidly across the room to find a woman to dance with. For a woman to refuse to dance was considered rude, but since there seemed to be fewer women than men, the men had to compete for a partner. With each song, the center of the hall was flooded with waltzing couples rotating counter-clockwise in a great circle. After each song, everyone would sit down again. Hairet told me that it was inappropriate for the same couple to dance together more than a couple of times.

Briefly during the evening was the waltzing, called tans oynash (literally, ‘playing dance’) replaced with ussul oynash (‘playing ussul’), in which pairs would dance traditional Uyghur ussul. While men and women, old and young could all dance ussul, it was most commonly two young men who faced each other and performed a dance that consisted of rotating their arms and wrists, softly snapping their fingers, and sometimes gently stamping their feet on the floor. In dancing ussul the partners would revolve around each other in a circle, remaining closely parallel in their movement and usually keeping their eyes on one another. Hairet told me that this music was a Uyghur form known as “muqam.”

In a recorded conversation with Hairet not long after the wedding, I briefly followed up on our earlier discussions about music.(4)

I asked him, “Can you tell me about different styles of Uyghur music? The traditional, popular and foreign?”

Hairet responded with a commonplace about the Uyghur love of music and dance that can be found in many descriptions of Uyghur culture, “Uyghurs, by custom, are a people who are very enthusiastic about song and dance.”(5)

He then explained the differences between traditional music and popular music in terms of the instruments used by musicians.

In general, Uyghurs consider such instruments as the dutar,
tämbur, rewap, and ghejäk,
and then another kind, we call it the çañ,(6)
as the most traditional instruments.

From long ago until now, people would greatly enjoy,
enjoy playing these instruments.

But now, in the present, there has been a change in this.

The young in general have begun to like electronic instruments,
instruments that are played electronically.

Or they take these earlier instruments such as dutar, rewap, and tämbur,
and make them electronic
and play them in sorrowful [mungluq] songs, in other . . . ah, music, in other songs.

So, coming from abroad, such as the guitar, an instrument from abroad,
the young artists of our people have popularized it some.

Now among the young, they are enthusiastic about the guitar,
they play the guitar a bit more.

Following up on this theme, I asked Hairet, “What songs or music, or what places does the music come from that most people listen to now? That they dance to? You know, that’s on the radio?”

He told me about how the music was changing because of people’s changing tastes in dance.

In the past, among the Uyghurs, music or songs were played on the dutar,
or on the tämbur, or the rewap, or other instruments,
music was played separately, or songs,
songs that went with the music would be sung,
and people would enjoy listening to them.

When someone liked to dance an ussul, they would dance the ussul.

Now this is gradually changing.
Now young people dance together with middle-aged and old people.
They all dance tans together.

Because of this, in the music that had been played on the dutar, tämbur, and rewap,
those instruments are being replaced by electronic instruments.

And all kinds of new, modern songs are played on instruments
such as electronic pianos, electric guitars, and drums.
When the songs suited ussul people who liked to dance ussul would dance ussul,
but most of the time people would dance tans.

Uyghurs were changing to dancing waltz as part of a larger cultural change. Although Hairet does not mention it, other Uyghurs I talked to also commented that they liked waltzes more because of the body contact between men and women. There was also an increasing interest in foreign instruments and music, especially the guitar. In 1992 when I was back in Ürümchi, the musical instrument factory that I often visited had turned to making pool tables and guitars as its dominant products, rather than specializing in traditional Uyghur instruments.(7) And the husband of my Uyghur teacher in 1992 had recently become a travelling salesman of imported electronic keyboards.

Although I became fascinated by the characteristically Uyghur music and texts, dance was more important in many social gatherings. During my research in 1992-93 I observed that while traditional Uyghur musicians and musical aficionados were often concerned about musical skill and authenticity, other participants at celebrations were ready to dance to all kinds of music. Dancing was multicultural, while music was often used to characterize ethnic uniqueness.

Dance has long been a significant mode of intercultural entertainment in Central Asia. The Turko-Sogdian commander An Lushan, although supposedly so heavy that he needed support to walk, was famous in the Chinese court for his skill at the “barbarian whirling dance” (huxuanwu), which he performed for the Tang Emperor.(8) Dances were a common way to pay homage on a visit to the court of powerful figures. A few years after An Lushan’s rebellion, a report from the history of Chinese-Uighur relations describes a young Chinese prince who went on an embassy to the Uighur court in 762, but refused to perform a dance for the Uighur Qaghan. The prince was whipped but was not executed because of his youth.(9)

The custom was widespread under Mongol rule as well, in which both singing and dancing were considered ways of showing respect, especially when drinking a toast to an authority figure. In 1604 when the Christian missionary-explorer Benedict Goës appeared in his Aqsu court during a dancing celebration, the twelve-year old Turko-Mongol prince of Aqsu asked him to demonstrate how “the people of his country used to dance? And so Benedict, not to be churlish with a prince about so small a matter, got up and danced himself to show the way of it.”(10)

In 1869 the British explorer Robert Berkeley Shaw was treated to tamâsha (‘recreation, fun’), by his host the Yuzbashi Muhammad Ya`qûb.

The Yoozbashee took care that my evenings should not be
dull.  He generally fetched me away to his own rooms for
tamâsha, as he called it, when one of his men would play on
a two-stringed guitar [dutar], while others danced in
Toorkee fashion.  The movements are slow; they wind in and
out with an alternate chassé of each foot, the dance, on
the whole, having a soothing and somnific effect (169-70).

The Yuzbashi sought to draw Shaw’s Indian servants into the recreation, and all broke into “a storm of applause” when one danced a “Guddee” dance for them. The Yuzbashi himself, “before a select circle,” danced an Andijani dance for Shaw.(11)

Later in his stay in Kashghar, another of his official hosts, the Mährämbashi, demonstrated some local dances for Shaw. Unlike the Yuzbashi, the Mährämbashi was a native of Kashghar, and resented the strictness of Ya`qûb Beg’s regime.

He complained, “Formerly there was much amusement, but now it is all
Islâm.”  And then he represented the grave demeanour of a
devotee taking out his rosary, and mumbling constant
repetitions of the name ‘Allah’. . . .  “This is what we
all have to do now,” he said, “each alone by himself; no
merry meetings or ‘tamacha [sic]’ now.”(12)

Even in Ürümchi in 1993 people frequently distinguished between religious leaders who accepted music and dance, and others who felt they were unacceptable for true Muslims. At least for traditional musicians such as Ömär Akhun and Qawul Akhun, this was a major concern, and they happily described mullas and ishans who showed their support for musicians by inviting them to perform or teaching them new poetic texts to sing.

Visiting Kashghar in the days before Ya`qûb Beg’s rule, Chokan Chingiz Valikhanov comments, “Wherever you go, you cannot be a invited to a party without female dancers who sing amazingly delicately the epic of Laili and Majnun that is loved in Kashgar.”(13)

The ussul dancing I saw at parties and weddings is still a way to show friendship and respect. One dancer begins by directing their dance at another, inviting someone seated to get up and dance with him or her. They dance together for a while, until one or the other, or sometimes both, turn away to find another partner with an invitational dance. This results in a sequence of people who each in turn take the floor, dance together for a while, and then the new dancer finds someone to join him or her. Men and women often dance together in this way, but there is usually only one pair of people dancing at any one time.

Lady Macartney describes a similar pattern at women’s “gala entertainments” which she attended in the early 1900s. “A small orchestra of four or five professional female musicians, who took their turn at dancing, provided the music.” Judging from her description, they played one or two bowed satars, a couple of rewaps, the zither-like qalun, and a dap.

When the music struck up, one of the professionals started the dancing and was followed by the guests, who, one by one, got up and gave a solo dance very slowly and gracefully, with swaying movements of the body, and arms, which were held straight out with the long sleeves hanging over the hands; the dancer sometimes simply marking the rhythm of the music with the feet, while the body remained still with arms outstretched and eyes closed. The waving and winding of the arms in perfect time to the music played a great part in the dance, and was wonderfully graceful and in keeping with the dreamy music. (1931: 122-23)

Under Manchu rule during the 18th and 19th century, Chinese writers described similar music and dance performances among Uyghurs, and as I discuss in Chapter Six, the Uyghurs have become proud of the repute their music and dance has earned throughout China.


During my stay in Ürümchi in 1989, all I learned about the Twelve Muqams was that they were the most important traditional Uyghur music. I did not meet anyone who could describe them in more detail, but people referred me to a few books that had been recently published.

When I returned to Ürümchi to do research on Uyghur song in 1992 and 1993, I had decided to focus on the Uyghur muqams. I was not able to gain access to the busy performers from troupes such as the Xinjiang Opera Troupe and the Xinjiang Muqam Ensemble in order to interview them about their lives and skills as muqam performers, so I focused on the more public apects of muqam performance, and the media presentation and popular understanding of the muqams as Uyghur national culture and as historical artifacts. I was fortunate to be able to talk to a number of the people who recorded, transcribed and edited the muqams to establish a fixed canon of texts and musical notations for large-scale staged performances. In addition, a number of semi-professional performers invited me to their muqam performances at weddings.

Heytäm was the chief craftsman at the musical instrument workshop near the main Uyghur market area known as Erdaoqiao. I visited his workshop often, and Heytäm always chatted with me and demonstrated his skills on the rewap he had made thirty years before in his apprenticeship. Several times he pointed out that he would never sell this rewap because it was something his mother had seen when she was alive.

Knowing that I was studying the Twelve Muqams, Heytäm one day told me that I and my wife Lynne should come with them to a bride’s side wedding celebration where he and a group of his friends were going to perform. He was organizing this group because the bride’s father was a friend of his from high school. His only request was that I avoid drinking because he felt alcohol always led to trouble.

A Uyghur wedding usually consists of two parties in different parts of town, with little mixing between them. The bride’s side party begins in the late morning, and the friends and relatives of her family occupy the main hall. The bride and her female friends have a somewhat quieter party in a smaller room off of the main hall. The main room is the center of the celebratory eating and dancing where the bride’s parents are hosts for their friends, while the bride’s group is more subdued and does not leave their room.

Usually, the groom arrives in the late afternoon with a large group of male friends to take the bride off to his family’s party. As he arrives, the door of the main hall is barred to the groom and his friends, and after humorous negotiations between the groom’s friends and the family of the bride, they are finally let in with much banter. At the party I attended with Heytäm, one of the groom’s friends played loudly on an accordion as he led the groom’s group into the hall. Similar negotiations took place at the door to the bride’s room, where her friends barred the door and tried to prevent the groom’s entry.

After the bride was brought out into the main room, the groom and his friends remained briefly while the bride said goodbye to her parents and a mulla read a blessing over her and the groom. In this case, some of the bride’s women friends and family members went with her to the groom’s party, but they stayed there for only a half-hour before returning.

Although both are all day events, the bride’s side party began winding down with the departure of the bride, while the groom’s side party is generally fairly low key until her arrival. At the bride’s party, we had a meal at noon consisting of the usual manta and polo, with many side dishes. During the meal, Heytäm and his friends began to play muqam songs. The group he had organized consisted of mostly conservatory-trained musicians who were doing this performance for pay. A few worked as professional musicians and music teachers, while others were only part-time performers. There were two ghejäk players, three rewap players including Heytäm, a dap player, and a çañ player. Everyone sang together.

After the noon meal, the musicians performed off and on throughout the afternoon. At this party, because the musicians were playing only muqams and other traditional songs, the dancing consisted of ussul. Only after the bride and her party had left– while the musicians were taking a break–was a rather small tape player used to play more popular music for tans, and many of the guests formed couples and danced counter-clockwise around the room.

Lynne and I sat with the musicians, enjoying the extra food and drink that musicians are given. As Heytäm had pointed out, drink did cause trouble, and one of the performers became so intoxicated that he could barely sit in his seat to play his ghejäk. I managed to fend off offers of baijiu and drank only beer. Except for Heytäm all the other musicians were drinking the strong liquor.

While some of the guests were dancing tans, one of the rewap players somewhat belligerently challenged them, asking them why they were not dancing Uyghur style. The exchange did not last long, but the upshot was that the musician wanted people to dance ussul, and not resort to the tape player, while the dancers wanted to enjoy themselves, and liked dancing together as couples. They saw dancing as entertainment while the musician was more concerned with preserving the characteristically Uyghur music that he was skilled in performing.

At the two groom’s side wedding parties that I attended, traditional Uyghur music was played during and after the afternoon meal, and in the evening a group of popular musicians using keyboards and amplifiers along with some traditional instruments would play. These bands still play distinctively Uyghur songs, but the loudness and rhythm was appropriate for dancing tans.

Dance was also a central aspect of the Thirty-Fifth Anniversary celebrations at the Xinjiang Art College where I lived. Preparations for this event were extensive and the performances were enormous. The younger dancers did outdoor group performances with their classes, while older ones did a special evening dance program. There were also many spontaneous music and dance gatherings as alumni and teachers joined in the celebrations.

At the evening celebration that began as an organized outdoor dance performance on the basketball court, faculty and even the school president joined into the dancing around two bonfires to the music of naghra and surnay. The long line of dancers revolved around the two bonfires, doing a dance that in Uyghur scholarship and popular media has been identified with the primitive origins of Uyghur culture. One Uyghur scholar of dance argues, “Since the fire dance [ot ussul] is based in the earliest myths and legends of the Uyghurs, it reflects the way of life, the customs and traditions, and the religious beliefs of those times.” He says the dance preserves the traces of Uyghur fire worship customs across several thousand years, and thus represents the superstitions of Uyghur primitive culture.(14) Despite unusual pouring rain, students continued the fire dance late into the night, until campus police had to tell them go to bed.

At one point late in the evening, someone put Michael Jackson music on the public address system, which initially got a small group of people performing under the street light and speaker pole across the basketball courts from the bonfires. When the naghra drummers and surnay player took a rest, the mostly student crowd began to drift towards the people dancing to Michael Jackson’s music. Unlike the popular music at the wedding, this music did not attract as much participation, but students were eager to see the dancing of a graduate of the school who was known to be a great dancer. Some other people joined in the dancing, but eventually interest waned, and people returned to the more participatory dancing of the naghra-surnay music. No one that I heard made moral judgments about the music: people seemed more interested in enjoying themselves than in selecting music based on ethnic origins or traditionality.

While Uyghur dancing thus has a social importance, and encourages collective participation, traditional Uyghur music has taken on a political importance, especially the classical music of the Uyghur Twelve Muqams. But even traditional muqam performers felt that the more delicate, sophisticated, slower muqam songs could find few appreciative listeners because they required mature tastes, and the texts were drawn from the works of difficult classical poets such as `Alî Shîr Navâ’î.

Because of their many difficult songs and musical pieces, in the 1950s the muqams began to be promoted as a vehicle for creating a classical national culture that would exemplify and reflect the important cultural history of the Uyghurs. From their origins in local professional traditions passed down from master to apprentice, usually within families, the muqams have been made into an extensive canon of rigorous and difficult music that no one can perform in its entirety. Muqam performers need institutional support for training and practice, and through these institutions this Uyghur cultural canon has become closely tied to political projects of creating an ethnic culture.


Since already in 1989 most Uyghurs identified the Twelve Muqams were a great achievement of their traditional culture, I decided to make them the focus of my research on the Uyghurs. In my background research I came to appreciate the Twelve Muqams as literature in performance, and I became fascinated by the poems full of the Sufi images and ideas that made up Islamic literary tradition. I was fascinated by the complexity of the Muqams as an oral tradition based in composed texts, a Turkic performance tradition using Persianate `arûz (quantitative) poetry, and the great esteem that the illiterate Muqam performers had for the difficult Turkic ghazal poems of the fifteenth century. It seemed to offer an excellent opportunity to challenge “great divide” theories of the distinction between oral and literate traditions and the associated mental capacities. I decided to research how traditional performers learned and used their poems, how their ideas influenced their singing, and how they shaped performances in response to their audiences.

But I was not able to experience muqams in traditional settings: the politics of research in Xinjiang made it difficult to work with local performers in Khotan, Kashghar or Kucha in southern Xinjiang. Instead, I remained in Ürümchi, taking advantage of connections I had with scholars, writers, editors, and a few musicians, to explore how the muqams were being researched, revised and recreated for presentation in the media and on stage. In addition to Heytäm, these musicians included Ömär Akhun, an eighty-year-old muqam performer from Kashghar, Qawul Akhun, the son of Turdi Akhun, the great muqam performer from Kashghar who had been recorded in the 1950s, and Shir Mämät, a tämbur performer from Ili.

I have already briefly described the performance context of the muqams when they were a part of social celebrations such as weddings. The best professional musicians were also part of muqam ensembles that performed stage concerts in which muqams were the main or sole attraction. These were the Twelve Muqams in their full glory. They were available on recordings of an hour and a half to two hours in length, and these recordings were often played on the radio on Sunday afternoons. The larger cities in south Xinjiang had muqam ensembles that specialized in these muqams, with certain performers who were particularly famous.

In Ürümchi the better known male singers were Osman Ämät, Ghuppar Akhmät, Akhmätjan Akhtäm, Rishit Qasim and Abliz Shakir, while the women were Turnisa Salahiddin, Risalät Hapiz, Aygül Yasin and Chimängül, all of whom are appear in the documentary Uyghur On Ikki Muqami that I discuss in Chapter Six. Local people would mention them with great pride, while people from cities of south Xinjiang would praise their own local singers, comparing them favorably to Osman Ämät or Chimängül.

The officially sanctioned Twelve Muqams were based on the Kashghar tradition recorded from Turdi Akhun. The only other muqams that had been officially institutionalized in Xinjiang were the somewhat simpler Ili Muqam tradition. The Ili Muqams were closely related to the Kashghar Muqams because a number of musicians had gone back and forth between the Kashghar and Ili regions in the 1800s. However, the dominant tradition in Ili as recorded from Rozi Tämbur in 1951-54 did not include most of the complex songs of the first part.(15) The Ili muqams are thus generally seen as a local version, while the Kashghar-Yarkand version is considered the Uyghur Twelve Muqam tradition, preserving the fullest and most cultivated form. Local traditions from Kucha, Aqsu, Turfan, Qumul, are generally seen as simplified folk versions of the classical Twelve Muqams because the former tend to be shorter, have fewer than twelve muqams and use more folk verse. Nonetheless researchers believed these local versions preserved some elements lost or forgotten in Turdi Akhun’s tradition, so they were collected and used as alternate pieces to fill in what researchers saw as gaps, repetitions and other flaws in Turdi Akhun’s muqam performances.

Because of the musical structure of the Kashghar and Ili variants of the Twelve Muqams is the complex result of collection and creativity, they are hard to present in a compact form. Exhaustive musical characterizations of the different parts have been provided by others, and I will only describe them in general terms so that I can refer to them later.(16) The difficult characteristic aspects of muqam music such as mode and rhythm are still aurally taught, and the various transcriptions are attempts to describe or represent these forms on the page, but there is considerable debate about the best way to do this.

Each of the Twelve Muqams is a suite of songs and instrumental pieces. According to Ömär Akhun the Twelve Muqams are played in five different tunings, but some performers reduce these to three. But within these tunings modal structure is flexible since the strings on the instruments used to play the melody in the muqams–satar, tämbur, dutar, qalun–can be readily adjusted by finger pressure on the strings passing over frets. Electronic measurement of the intervals of the scales in Turdi Akhun’s muqam scales shows that each muqam is distinctive despite similarities among the modes for some pieces within two different muqam suites. In working out his systematic analysis Zhou Ji has tried to characterize the Mushavräk, Nava, Ushshaq, and Çargah muqam scales according to sequences of half tones, three-quarter tones, whole tones, and five-quarter tone intervals, but in doing so has reduced the complexity found in the measured scales, and not reflected performers’ ideas.(17)

Only in the Çäbbiyat and Nava muqams do most of the songs remain within a stable characteristic mode. In the other muqams most songs modulate from one mode to another, although usually starting from and returning to the same basic mode in each song within the muqam. Zhou Ji and Jean During and Sabine Trebinjac point out that while most of the songs in the muqams are heptatonic, some are more pentatonic or hexatonic.(18) I discuss the various interpretations of the scales and modes more fully in my chapter on Ömär Akhun.

The melodies of muqam songs were probably originally improvised according to a characteristic mode associated with each muqam.(19) This is still the case in many traditions in the Muslim world, such as the makam tradition of Turkey and the dastgah tradition of Iran. However, in Central Asia, the modern Tajik-Uzbek Shashmaqom tradition and the Uyghur Twelve Muqam traditions are based on fixed tunes with little or no improvisation.(20) This is part of the institutionalization of the tradition, and its projection onto the past. Improvisation would leave the tradition open and productive, an ongoing technique for creating song and music, rather than a fixed corpus. Its elimination facilitates unison performance by ensembles, and enables institutional teaching of the songs and music without teaching how to create new ones. When I talked to him, the traditional performer Ömär Akhun claimed he could readily create new songs, and said he had offered to fill in the gaps of the institutionalized tradition, but his offer was rejected. Instead he was employed to teach the standardized muqams, rather than being encouraged to create new ones.

Each muqam melody consists of many repeated motifs, following a basic progression from low register to climactic moments in a high register somewhere after the mid-point of the song, with a return to low pitches at the end. The point or points where a muqam song reaches its highest notes are called avj ‘peak, summit.’ There can be up to three of these in a single song, but Qawul Akhun pointed out that this was more common in Turdi Akhun’s longer songs. The recent edited versions are shorter, since, according to Qawul and others, modern audiences are less willing to sit through long performances. An avj can reinforce the sense of the song when it is sung with words that express the poet’s despair or passion. The progression up to and down from the avj is usually gradual, but when reaching the avj there are sometimes leaps in pitch of a fifth or even a entire octave.(21) Like the song itself, each line tends to begin with a rise in pitch, and descend towards the end. In the slower and less dramatic sections of songs, in many measures the melody moves in only a narrow ambit, but these measures alternate with measures that are predominantly climbing or descending.

The now standard Twelve Uyghur Muqams are the following: Rak, Çäbbiyat, Mushavräk, Çargah, Pänjigah, Özhal, Äjäm, Ushshaq, Bayat, Nava, Sigah, and Iraq. In the past, the names and sequence of the muqams, like the names for their parts, varied from performer to performer.(22) Historical and literary sources from Eastern Turkistan mention several other muqams as part of the tradition, and Ömär Akhun often said that there should be thirteen or fourteen, rejecting the idea of a canonical list of Twelve Uyghur Muqams.(23)

The historical shift from maqâm as mode to muqam as particular song or suite of songs has profoundly changed the concept of the twelve maqâms. For examples of using maqâm to characterize particular pieces of music, particularly the beginning and ending modes, see Mavlânâ Najm al-Dîn Kavkabî (d. 1532) (Jung 1989:93-94) and Signell’s discussion of the Turkish makams (1977). Historically in Eastern Turkistan, the idea of new compositions based in the maqâm can be found in the Tavârikh-i mûsîqiyyûn describing maqâm practice in the fifteenth and sixteenth centuries. I discuss this work in Chapter Six.

The names of the Uyghur muqams are closely related to names from other maqâm traditions. For instance, the Bukharan Shashmaqom tradition consists of the maqoms Buzruk, Râst, Navâ, Dugâh, Segâh, and Irâq, but includes parts known as Çârgâh, `Uzzâl, `Ushshâq, and Bayât, among others (Jung 1989: 172-74). According to During and Trebinjac’s analyses, however, the parts with similar names in these two traditions differ in scale and structure (1991).

Except for Çäbbiyat and perhaps Rak, all of Uyghur Twelve Muqam names are known to be used in other maqâm musical traditions. `Ushshâq, Nawâ, and `Irâq are mentioned already in the Qâbûs nâmâ. Ärshidinov and others argue that Rak is derived from the Hindu musical term râga.(24) This seems supported by Ömär Akhun’s pronunciation of this name as ‘Rägh,’ but if we seek an origin within the maqâm traditions, the name Rak most likely comes from the shu`bâ rakb listed by `Abd al-Qâdir ibn Ghaibî (1350-1435).(25) Çäbbiyat is from çap ‘left, dischordant'(26) and Bayat.

During and Trebinjac claim that Mushavräk is derived from Arabic shavrak.(27) In an error prone article the usually thorough scholar N. A. Baskakov derives it from mushâvir and aq, claiming it would mean ‘sign, signal.'(28) But the East Turkistani poet Molla Sabir binni Abdulqadir (1840-1920) gives us a clue when he calls this muqam Nish Aväräk (Bulaq 1988, #2, 88 and 123). There is a shu`bâ named Nishabûrak (‘from Nishapur’) that is fairly common since `Abd al-Qâdir ibn Ghaibî first noted it (Jung 1989:75 and passim). The changes [b] > [v] and [n] > [m] are common in Uyghur.

Sigah, Çargah and Pänjigah mean third, fourth and fifth position in Persian and are first noted by Qutb al-Dîn Shîrâzî (1236-1311) (Jung 1989:67). Ärshidinov claims Özhal means ‘to descend,’ but it should probably be interpreted as from `uzzâl which is a common maqâm also first noted by Qutb al-Dîn Shîrâzî (Jung 1989:70).

Äjäm means ‘non-Arab, Persian’ and like the muqam named after the Iraq region, accords with the common use of names of places or peoples for maqâm names. Äjäm is not mentioned until the end of the fifteenth century and remains uncommon in later discussions of maqâms, but it does appear in the Turkish tradition (Jung 1989:88 and 112, Signell 1977:24).

Ushshaq comes from the Arabic `ushshâq ‘lovers.’ Bayat is said by some to refer to the name of a Turk tribe, or to be the early Turkic word for God.(29) Baskakov derives it from Arabic bayat ‘poem, couplet’ but this should be bayt (Wehr: 84), and the maqâm name is common as bayât and bayâtî since being first noted as a shu`bâ by `Abd al-Qâdir ibn Ghaibî (Jung 1989:75). It should probably be interpreted as ‘grief, anxiety, care’ or ‘passing the night, doing anything in the night [implying going sleepless?]’ (S:213). Nava is a common maqâm name and comes from navâ ‘song, tune, music.’

Early descriptions of the maqâms classified them according to which ones were appropriate for older and younger audiences, for happy and sad moods, and for particular times of the day.(30) The Qâbûs nâmâ goes into detail about when to play heavy and light modes: For old men and those of serious character, one should play heavy modes, while for the young play in light modes. Then the delicate tarâna are for “women, children and men of more effeminate taste.” In addition to character, there is music appropriate to people of each of the four humors:

When you are seated with a company, look about you.  If the
audience is composed of men of ruddy and sanguine
complexion, let your music be largely on the bass strings;
if the audience is pale and bilious, let the music be
chiefly on the short strings; if your hearers are pale-faced, obese and large, play mostly on the bass; if they
are dark-complexioned, lean and melancholic, play on the
lute for them.

For elderly men “let most of your songs then be about old age and scorn of the world.” For young men and boys

sing melodies picturing women or in praise of wine and
wine-bibbers.  If you see that the audience is military or
given to living by freebooting, sing them quatrains of
Turkestan (Transoxiana), songs about about battle and
bloodshed and in praise of an adventurous life.”(31)

Because of ideas about the proper time to play maqâms, and the division of the twelve muqams into roughly thirty songs each, with each muqam taking roughly two hours to play through, some Uyghur performers and scholars feel that muqam structure should be synchronized with the division of the year into months, the month into days, and the day into hours. But it is clear that these idealized abstractions do not guide practice: instead the slower and longer songs of any muqam are felt more suitable for more mature listeners, while younger people are said to like the more rapid jula and sänäm dance tunes that come later in a muqam suite.

While the scales and modes are generally distinctive for each muqam, and melodies are distinctive for each song (although often related through the reuse of motifs and even tunes in different muqams), rhythmic patterns are highly consistent from muqam to muqam. Each muqam suite of songs occurs in a specific order, and each named song is characterized by a rhythmic pattern that–with a few exceptions–is very similar for each song of the same name in the other muqams. In each of the three sections, the songs generally move in a sequence from slower to faster rhythms (the kiçik säliqa and the täkit are exceptions to this rule), and each song itself also gradually increases in tempo from the beginning to the end during play. In Turdi Akhun’s performances, he slowed the Mäsräp section down towards the end, but this is not done now. The organization of muqams into songs according to rhythm and tempo provides the basic structure of muqams.

The list below gives the basic parts of the Kashghar muqams according to the publications based on the 1950s recordings of Turdi Akhun. Besides bash and çüshürgä all of these names are from Arabic or Persian, but most of these words had little meaning to the Uyghurs I talked to. After each part I give a brief description of its characteristics and relationships to other pieces in the muqam system. Ömär Akhun calls each song and its märghul a shakh ‘branch’ or shöbä (from the term shu`bâ used in maqâm theory), and conceives of them as the combination of an ahañ (‘tune’) and bäyt (‘poem, song text’). Ärshidinov argues that each muqam should consist of 12 shöbä with 2 shakh each (a song and its intrumental märghul), although in the past there were 24 shöbä.(32) He says that each shöbä now takes about ten minutes, so a muqam should take about two hours. If such a system ever existed, it was probably never performed in such a regular form, since as even Ärshidinov admits, performers were constantly learning, adding and forgetting songs, and changing the tradition.

First Section: Çoñ Näghmä (‘great songs’)
This includes the basic songs of the muqam tradition, while the Dastan and Mäshräp sections were added to the muqams from other traditional performance forms. Qawul Akhun described to me how his father Turdi Akhun learned dastan and mäshräp songs from other performers and added them to his Muqam repertoire. The fact that very few song texts are shared between these three sections supports the idea that they originate in socially distinct traditions and were not usually performed by the same people. Ömär Akhun performs exclusively the Çoñ Näghmä.

The Çoñ Näghmä begins with a slow free-rhythm song and moves through progressively more rapid songs, which the performers play through without pausing between pieces. The slower songs have up to 12 measures per line of verse with syllables melismatically drawn out, while the dance songs that use qoshaq folk verse can have as few as four measures per line. Each song can have verses drawn from more than one poem in order to fit the length of the music. Often single couplets will be drawn from several ghazals. Certain poems and couplets are particularly popular for these songs, and occur in more than one song in the same Çoñ Näghmä and often even in those of several muqams. These texts almost never appear in the Dastan and Mäshräp sections. The çoñ säliqä does not seem to share verses with other songs in this section. The päshru and täkit only occasionally share a verse with other songs.

Bash. ‘Head.’ Also known as bashlinish ‘beginning,’ more formally as muqäddimä ‘introduction,’ and very occasionally as sarakhbar.(33) Ömär Akhun simply called this section “muqam.” This is the initial slow free-rhythm song in which a solo muqam singer accompanies him- or herself on a lute, such as satar, dutar, or tämbur. Because there is only one accompanying instrument, the muqam beshi (possessed form of bash) has the most clearly audible text, and thus makes the most verbal impact, especially in the revised version in which the other songs are sung in chorus. This ranges from 8 to 34 lines of poetry in the texts from Turdi Akhun’s performances. Qawul Akhun explained that the shorter texts generally belonged to a set of alternative introductions known as shahzadä (‘prince’) that were more lively and interesting for younger audiences.

Täzä. This name may be from tarjî` ‘repetition’ which is the piece immediately following the maqâm in the Khwarazmian Shashmaqom. During and Trebinjac (1991:13) suggest it is from täz- ‘to arrange in order’, which should be tüz-, while Ärshidinov (n.d.) suggests täzä comes from an Arabic word meaning ‘pure, new’.

The täzä is the beginning of the measured rhythm songs. Here the traditional performer would be joined by an apprentice playing a stable rhythm in 3/4 or 6/4 time on a small frame drum known as kiçik dap. In the present standardized Uyghur ensemble form, the täzä is the first song sung by a chorus, which then sings in unison for the rest of the muqam suite. In the published texts, this song ranges from 14 to 58 lines of ghazal poetry by Navâ’î and others.

Märghul. ‘Curled locks, decoration, sweet melodious voice’ (S:1217). Kaufmann (1990:474) gives marghûla as ‘repetition of the same note to avoid any pause in the melody.’

This is an instrumental piece based on the preceding song. Always in the same rhythm. Usually ends with a repetition of the final lines of the preceding song. These lines are known as the çüshürgä or çürigä ‘tail.’

Mustähzat. Baskakov derives this from the Arabic word for ‘a poem in which there is rhythmic augmentation’ (1990:22). Steingass has ‘kind of poem’ (S:1230). It is in 5/4 time. Appears in Çäbbiyat only, apparently taking the place of the nuskha, which does not appear in Turdi Akhun’s Çäbbiyat.


Nuskha. ‘Archetype, model’ (S:1400). Turdi Akhun apparently only sang these in the Rak, Mushavräk, Pänjigah, Özhal, Ushshaq, and Nava muqams, but they since then have been included in other muqams as well. There are three basic rhythmic patterns performed in the nuskha. According to Zhou Ji, the tätür (‘reversed, stubborn’) nuskha is 2.5/4 + 4/4, which can also be thought of as 6.5/4 or 5/8 + 8/8. This nuskha appears in Mushavräk and Nava Muqam, although Mätruzi Tursun also gives an example from Rak Muqam (1992:406). The oñ (‘correct’) nuskha is 4/4, and appears in Ushshaq and Özhal Muqam. The täzä (sometimes known as tokur ‘lame, crippled’) nuskha form is 5/4 or 3/4 + 2/4. It appears in Rak, Pänjigah and now in the Bayat and Chargah Muqam.

Mätruzi Tursun has apparently renamed the Çäbbiyat muqam mustähzat as nuskha to give a fourth kind (1992:403, 416). Although the Pänjigah muqam as annotated by Wan Tongshu did not include it, the present Pänjigah has two nuskha, and in Qawul Akhun’s enumeration of nuskha rhythms there are five kinds, including the second nuskha of Pänjigah which is unique. Qawul listed the four other different kinds as appearing in the Pänjigah, Nava, Ushshaq, and Rak nuskha (T 4/7/93). Zhou Ji’s transcription shows it as being in 6.5/4 time. Nuskha songs range between 14 and 32 lines of poetry.

Märghul. (does not appear after nuskha in Rak.)

Kiçik Säliqä. ‘Little decoration’ (S:695). In 4/4 or 2/4 time. Ranges from 20 to 32 lines, sometimes simply called säliqä, as in Äjäm muqam. Often contains the changes of rhythm known as özgirish. In some muqams the kiçik säliqä follows after the jula and sänäm, or there is a second one that follows the çoñ säliqä.


Yarim Saqi. ‘Beloved cupbearer’ or possibly ‘half’ saqil and thus related to the nim saqil ‘half saqil (‘heavy, slow’), of the Shashmaqom.(34) In 7/8 time. Only found in the Ushshaq muqam. 22 lines.


Jula. ‘Bright, lively,’ Ar., but see Steingass for ‘presenting bride,’ (S:369) and Sänäm (‘idol, beloved, beautiful’ Ar., S: 794). In 4/4 and 2/4 time respectively, these are shorter dance pieces with strong rhythms. No märghul follows these. Ranging from 4 to 14 lines, the jula often include folk verse and easier ghazals. Turdi Akhun used the same tune and text for jula and sänäm in several different muqams. This is perhaps explained by Qawul Akhun’s description of the jula, sänäm and çoñ säliqä as dance pieces that are meant to wake people who are sunk into muñ (‘pensive sorrow’) caused by the more contemplative pieces that precede them (F 6/21/93). These more lively pieces were perhaps not unique, integral parts of a muqam suite, but pieces included for social reasons more than musical ones. In addition, the jula and sänäm often share verses. The distinctive rhythm is characteristic of that used in music for weddings and other toy (‘celebrations’), usually played on the naghra and suonay. The sänäm tunes are also used for a very popular traditional Uyghur dance also called sänäm. In their book on Uyghur dance Himit Mehrulla and Lätipä Qurban claim that the traditional Sama (samâ`) ussul is now danced to the jula muqam song.(35)

Çoñ säliqä. ‘Big decoration’. In 5/8 time which Zhou Ji felt was best described as 3/8 + 2/8. This piece has no märghul. Turdi Akhun apparently did not perform this in the Çargah, Äjäm, Sigah, and Iraq Muqams. The çoñ säliqä ranges from 8 to 16 lines of folk verse or easier ghazals, rarely uses the same text as any other verses. It also uses the same rhythm as the music for the Sama ussul which is performed as part of Sufi worship.

Päshru ‘Preamble, overture’ (S:252). In 4/4 time, sometimes 5/4. Zhou Ji felt that Ushshaq Päshru should be annotated as 2.5/4. This only appears in Turdi Akhun’s Rak, Mushavräk, Pänjigah, Özhal, Ushshaq and Bayat. It ranges from 8 to 22 lines of poetry, often including folk verse and easier ghazals.


Täkit. ‘Affirmation, emphasis,’ Najip 301 claims is from Arabic, but I can not identify the original term. In 6/8 time. Zhou Ji says it can be divided into three thirds or two halves, like Uzbek and Azerbaijani music. This is the short concluding piece for the section. Ranging from 2 to 6 lines long, rarely the same text as in other songs. According to the transcriptions, Turdi Akhun did not perform the täkit in the Iraq muqam, and in four of the muqams he used the same couplet by Navâ’î for the täkit:

Noh omrigä baqa yoqtur, Sulayman mülkigä,
Äy Navayi badä iç, aläm ghemi bihudädur.

Noah’s life was not eternal, nor was Solomon’s kingdom,
Oh Navâ’î, drink wine, and become unconscious to the suffering of the world.

Second Section: Dastan
The Dastan consists of up to four progressively more rapid songs in a generally fixed rhythmic progression: 1) 2/4; 2) 7/8, sometimes better as 3/8 + 4/8; 3) 6/8 or 9/8, sometimes better as 3/8 + 3/4 or 3/8 + 6/8; 4) 6/8. Many of Turdi Akhun’s versions have been replaced by those from the Ili variants of Rozi Tämbur and Yüsüp Ämät. The texts include romantic dastan epics from written sources, and some ghazal poems. These verses were not used in the Çoñ Näghmä or the Mäshräp sections until some of the editors of the revised muqams borrowed texts and even tunes from one section to put into another.

A line ranges from three measures to ten. For most of the dastan songs the vocal, instrumental and dap rhythms all start simultaneously in each line, and there are no özgirish rhythm changes in the middle of a song. As pointed out below when I discuss Zhou Ji’s analysis of a dastan song, it is common for the first line of a couplet to take a different number of measures than the second. The songs range from 12 to 32 lines in length. Each one has a fairly brief märghul following it. The dastan are simply called first, second, third, and fourth.

Third Section: Mäshräp
Mäshräp means ‘drinking, social gathering,’ and implies ‘intoxication,’ thus suggesting the Sufi poet Mashrab, pronounced Mäshräp in Uyghur. This section consists of up to six shorter and progressively more rapid songs in fixed rhythmic progression: the first and sometimes the second songs are 7/8 or 3/8 + 4/8. These also use the rhythm of the religious samâ` dance, but not the same one as that of the çoñ säliqä. The second and subsequent songs are in 2/4 time. However Zhou Ji found that Turdi Akhun was not very consistent in following this rhythmic sequence. There are rhythmic changes in the middle of some mäshräp songs. The mäshräp songs have no märghul.

Turdi Akhun performed folk verses and ghazals by Mashrab, Navâ’î and even the twelfth-century Sufi poet Ahmad Yasavî in the mäshräp songs. These verses are rarely used in the other two sections. The songs range from 8 to 26 lines long, and like the dastans are referred to by ordinal numbers.

Turdi Akhun’s son Qawul Akhun believed that Turdi Akhun had learned these songs from Sufi worshippers in their rituals. The accelerating rhythm of these songs follows the accelerating pace of the ecstatic chanting and group movement that is the most common form of Sufi worship in Xinjiang.

Although most of the names of the Uyghur muqams and their component songs exist in other maqâm systems, the identification of traditional usul rhythms that has been so important to maqâm theorists since Fârâbî seems to be absent in the Uyghur muqams. Instead, in Uyghur, the word ussul refers to dance. The terms udar (from Russian ‘strike, beat’) and takt (through Russian from German ‘time, beat, measure’) are now used to describe the dap rhythms in the muqams and other music.

Uyghur musicians use rhythmic patterns to identify songs, but the rhythms are conventional and fixed rather than part of a system of variations. Classes of similar rhythms such as the jula and sänäm or the three or four kinds of nuskha rhythms are recognized. But the use of özgirish rhythm changes suggests an earlier more systemic understanding of muqam rhythms.

Three Uyghur names for kinds of tunes seem to be derived from the names of maqâm rhythms, but the rhythms do not match those used elsewhere. The rhythm of päshru in the Uyghur muqams only generally resembles that of the peshrav of the Shashmaqom. The mustähzad appears only in the Çäbbiyat muqam, and does not resemble the Shashmaqom mustazod.(36) But there are other connections between the rhythms of the Uyghur muqams and the Shashmaqom. The characteristic rhythm of the sixth song in the Mäsräp section of the Kashghar muqams appears in the Bukharan Shashmaqom where it is called the kashqarça usul (Jung 1989:181-2). None of the other rhythms are identical, although aqsaq ‘limping’ rhythm characteristic of Shashmaqom parts such as talqin have a general similarity to the Uyghur muqam aqsaq rhythms in nuskha, sänäm, çoñ säliqä, and päshru.(37)

Different scholars have different ways of representing and characterizing the complex rhythms of the muqam songs. During and Trebinjac describe the creation of a hemiole within a hemiole, and point out that this is unique to Uyghur music, and describe the result as “a sort of aqsaq of an aqsaq” (1991:18). My teacher on muqam music analysis, Zhou Ji, used a complex notation to show the effect of hemioles on the meter. He annotated the Nawa tätur nuskha rhythm as 2/4 + 2.5/4 + 2/4, but pointed out that the rhythm was ambiguous, because it was not clearly divided into measures and could be annotated as 2.5/4 + 4/4, or 6.5/4. An example of the complex variations in the rhythms in Turdi Akhun’s muqam performances can be found in his third dastan song in Nava Muqam: Zhou Ji found that for the first line of each couplet would be in the rhythm 4/4 + 3.5/4, while the second line of each couplet would in the rhythm 4/4 + 3.5/4 + 3/4.(38) In her transcriptions of the free-rhythm muqam introductions, Tamara Alibakieva (1988) uses the rather complex method of putting time signatures such as 6/8, 9/8, 12/8, and 15/8 at the beginning of each measure to give a sense for its rhythm.

Another source of rhythmic ambiguity is that in each line of song the dap rhythm starts a few beats ahead of the melodic rhythm. In transcriptions, the melody is usually taken as determining the beginning of the musical measures, so a measure beginning with two or three rest beats has to be annotated to show the beginning of the percussion.

Mätruzi Tursun gives a detailed discussion of the rhythmic complexities of the muqam parts, especially the offset between the melody and usul. He worked together with Zhou Ji, and their analyses are similar. They transcribe the rhythms in greater detail than for instance Shao Guangshen, who seems to base his discussion of the usul on the transcriptions of Wan Tongshu rather than on the muqams as performed now.(39) Offsets of usul and melody are also common in Shashmaqom.

In most songs each line of vocals, melody and percussion do not stop at the same time. The melody and percussion end a measure or two after the vocal line. This phenomenon is even more pronounced at the transition from one song to the next, when the dap player changes the rhythm while the melody and even the vocal part is still finishing the song before. This also happens in the özgirish changes when the dap rhythm changes in mid-song, and the vocal rhythm changes immediately afterwards. In recordings of Turdi Akhun, it appears that he did the opposite, changing his song, with the dap player following him. This makes sense since the dap player was his apprentice.

The songs in each muqam section run together without pauses. Qawul Akhun pointed out to me that Turdi Akhun would go back to the beginning of the Çoñ Näghmä, Dastan or Mäshräp section to start over if there was a mistake or a pause in performance. But between sections dap rhythm and melody stop at the same time. This gives the dramatic simultaneous end of chorus, melody and dap rhythms in mid-measure, with a brief silence followed by the solo vocalist performing the slow free-rhythm çüshürgüsi that ends the Çoñ Näghmä and Mäshräp.

Traditional muqam performers seem to have had considerable flexibility in correlating song text and musical melody in their songs. Although each poem has a fixed line length, ranging between seven and sixteen syllables of ghazal poetry, differences in sung melody and repetitions of parts of lines, as well as different song tempos makes for a vast range of ways and times for singing identical lines. In singing the sixteen-syllable ghazal lines of the free-rhythm bash muqam, a line usually takes around 20 seconds to sing, but can take as little as eight seconds or as many as thirty. In Ömär Akhun’s Ushshaq Täzä some lines take as much as forty seconds to sing, while others are half as long. The lines in the revised Uyghur Muqams täzä songs are a little less long, and the pieces that follow the täzä have lines that vary less in length, have fewer repetitions and melismas, and are progressively shorter, down to about four seconds. In contrast to the flexibility of Turdi Akhun and Ömür Akhun’s performance methods, the modern conservatory-trained muqam singers memorize a rigid text-melody relationship because almost all the songs are sung in unison.

The relationship of meter and melodic rhythm in the muqam songs is complex. Although many muqam songs use accentual folk verse, vowel quantity plays an important role in singing. But it is based in flexible principles, not adhering to the `arûz quantitative meters derived from Persian prosody. Vowel quantity is determined by tune and the performer’s interest and skill in elaborating sounds. The Uyghur language has final stress in most words, but there are also long vowels in words borrowed from Persian and Arabic. In muqam songs melismatic extension of a single vowel sound is an important melodic technique, and the syllables thus extended are usually the final syllables of Turkic words or the long syllables of words borrowed from Persian and Arabic. Sometimes the extended sound is an -i sound affixed to the end of a word. The melismatic vowel is always the last sound in a melodic measure, and can take up more than one measure. The musical significance of melismatic vowels can be seen in the fact that they take up more than half of the time in most lines, even though only one, two or three syllables out of sixteen are melismatic. In one twelve-second sixteen syllable line sung by Ömär Akhun, he makes a single melisma take eight seconds.

Most maqâm scholars argue that `arûz is fundamental in singing the poetic texts. During and Trebinjac cite approvingly the argument of Alibakieva that the Uyghur muqam musical rhythms are patterned after the `arûz quantitative meters of classical Chaghatay verse.(40) But they discuss the relationship of musical meter and prosody without making a clear distinction between dap rhythm and melodic rhythm. `Arûz rules do seem to cause the first syllable in a prosodic foot to be scanned short, but melody determines which syllables are sung long, and how long they are sung.

Especially in the early slower pieces in a muqam, significant variations of scansion can occur when words or phrases are repeated in a different melodic context. And the scansion of ghazal poems also varies when they are used in different songs. These variations show that while the ghazal poetry itself fits quantitative meters, these only establish preferences about which syllables to sing long. In practice most syllables can be long or short, and the melody does not follow `arûz meters closely. Singers may prefer to sing long and melismatic syllables on long vowels of Persian and Arabic words and on endings of Turkic words, but these same vowels are sung as short when the melodic rhythm requires it.

Theodore Levin closely examines the relationship of poetic `arûz meter and melodic rhythm in the Tajik-Uzbek Bukharan Shashmaqom tradition. He finds that there is a “very close correspondence between prosodic quantities and melodic rhythm” with short values of each coinciding, such that an eighth note or less accords with prosodically short vowels.(41) Levin’s examples show that the dominant scansion rule is that open syllables are the only ones that can be scanned short, while they and all others can be scanned long. However, he also finds that overlong syllables are respected and scanned as a long and a short together, corresponding to a single quarter note. The sixteenth notes that would correspond to a short syllable are eliminated in such cases. He also finds cases of adding a short vowel -i to words so that a line will have enough morae (142-43).

In contrast to Levin’s analysis, Mark Slobin demonstrates that the verse meter and musical rhythm in Tajik maqâm songs have a more flexible relationship, in which the singer seems roughly to follow the rules of Persian verse, but nonetheless uses melody to emphasize syllables expressively. The singer is not bound by strict rules about vowel quantity, although poetic feet and melodic feet do remain closely alligned.(42)

Slobin’s description is apt for the Uyghur muqam songs as well. In the free-rhythm Çäbbiyat bash muqam as sung in the revised Twelve Muqams, the melodic lines and poetic lines are organized into metric feet which coincide, but there are only general preferences for how many syllables can be in a foot, and which vowels can be long or melismatic (I discuss this verse in Chapter Eight). Even though the ghazal poem in this song consists of four equal feet of four syllables in every line, the melodic rhythm and the number of syllables in each melodic foot do not remain constant from line to line. Each melodic foot includes three to eight syllables, spanning one or two poetic feet. The long and melismatic syllables can be the second, third, fourth, seventh or eighth syllables. Any two of these can be melismatic in a single melodic foot. Uyghur muqam singers avoid singing the first syllable in a poetic foot as long, and the only place two adjacent syllables are long is when the first is overlong, as in the Persian guft’äm ‘if I say.’ But even overlong syllables such as this one or the ones in the words türki and parsi [fârsi] can be sung short. The limits of melodic feet almost always coincide with the limits of poetic feet, although sometimes a two-syllable word–such as yüzüñ in line two of the poem–can be used as the ending of one foot and then repeated as the beginning of another foot.

This kind of relationship of melody and poetry holds true for the bash muqam and täzä songs in the modernized stage form, and as performed by Ömär Akhun, and as near as I could tell from poor recordings, by Turdi Akhun as well. It also sounds like Turdi Akhun sang each line in täzä more rapidly and with fewer repetitions of words and phrases within each line than in Ömär Akhun’s performances, while the revised muqam täzä songs seem to be performed more like Ömär Akhun’s versions. This can be explained by the fact that Ömär Akhun worked at the Art College in the 1970s and 1980s training new performers in muqam techniques.

In the measured songs that come after the bash and täzä, syllables that are scanned long are fixed even when parts of lines are repeated. In the revised version, the choral songs have very little melisma, perhaps because of the difficulty in singing it in unison. When there is an extended vowel it is usually sung on a single pitch. The dap rhythm in most of the faster songs does stay synchronized with the lines of melody and song, and keeps time with the verse meter in the verse sections where both melodic and poetic feet consistently contain four syllables. On the recordings, Turdi Akhun sang many of his faster songs so that melodic emphasis coincided with stressed dap beats.

Ömär Akhun also repeated parts of poetic lines with more variation than in the present standardized versions, in which the repetitions are fixed and sung in unison in most of the songs. Ömär Akhun actually varied the wording on some repetitions to make the meaning clearer. In the penultimate stanza of his Ushshaq täzä he sang the phrase at aqsaq-u yol tayghaq ‘the horse is lame and the road slippery’ and then repeated it as etim aqsaq yol tayghaq ‘my horse is lame and the road slippery,’ giving at ‘horse’ the possessive suffix (along with the usual Uyghur vowel change at > et). In the first two stanzas of this same song, Ömär Akhun also repeats part of one line with the final words of the preceding line, which is never done in the new standardized muqams. This seems to be further evidence that he integrates music and verse with more flexible and improvisational techniques, while performers of the revised muqams have to remain within pre-established edited forms. The repeats are fixed for the standard versions and are explicitly indicated in many of the oliographed texts that are now used for training the muqam singers.


In addition to presentation on stage, the muqams are presented, taught, analyzed and interpreted by performers, scholars, media figures and politicians. The performers themselves have little interpretive authority or access to public media. Instead, political and scholarly authority is used to craft an image of the Uyghur Twelve Muqams as a treasured artistic canon from the Uyghur past that belongs to all Uyghurs. They are variously represented as having been preserved and transmitted without essential change by Turdi Akhun, or having been rescued and restored to their former greatness by the scholars of the Muqam Research Group (Muqam Tätqiqat Gruppisi) and related organizations. As a widely-accepted shared tradition of Uyghur culture, the Uyghur Twelve Muqams are interpreted to give weight and support to ideologies about the Uyghur past. Because they encompass so many songs and musical styles within a single large and complex system, ideas about the muqams’ past shape ideas about Uyghur history more generally. The Twelve Muqams are a monument that makes a collective Uyghur past prominent and easily available as a concrete material presence that serves as foundation and guide for ethnic consciousness and pride.

The institutionalization of the muqams has involved the recording of performers throughout Xinjiang, transcribing and correcting music and texts, comparing and collating musical pieces, and integrating selected pieces into a dominant Kashghar tradition. This work has been promoted by politicians and carried out by scholars collaborating with musicians, singers and poets. The Uyghur Twelve Muqams now have an important place in the mass media, on stage, and as a cultural showpiece in political events and discourse within the Uyghur context, the Chinese context, and internationally. They have been presented as the core and canon of Uyghur cultural heritage, a tradition shared by all Uyghurs that links them with a coherent past, and a container of representations of Uyghur customs and literary, musical and dance genius.

In order to create this canonical tradition and perpetuate it within an official institutions, politicians and scholars had to redefine and rename the content, and institute new ways of understanding what the muqams in terms of history, society, ethnic identity and culture. While the muqams are in some contexts still performed by individual performers who rearrange them to please themselves and their audiences, but in the urban context where I was, the institutional form of the muqams was accepted as standard, while other forms were “local” or “individual” variations. The idea of shared culture has brought the idea of fixed form. Although the muqams have been transformed from divergent fluid traditions to standardized canon, they are presented as an unchanged traditional art that has always been Uyghur, now restored to its original shine and put on stage by professional performers. In the rest of this chapter I discuss how and why the muqams were rearranged, how this work was justified, and what methods were used to reshape and reinterpret the materials collected from traditional performers to make the modern (zamaniviy) Uyghur Twelve Muqams.

The idea of the Uyghur Twelve Muqams as a “folk classical” (khälq klassik) tradition is already established in the title of the first major publication about the muqams in 1960.(43) Here the use of khälq (‘folk, people’) to characterize ‘classical’ is somewhat ambiguously balanced between a description of musical culture that is ‘folk, not institutional’ and musical culture ‘belonging to the national ethnic group.'(44) In ongoing usage the name Uyghur On ikki muqam ‘Uyghur Twelve Muqams’ continues to specify that these muqams are shared by, or belonging to the whole Uyghur ethnic group, while the terms khälq and klassik continue to be used as common qualifiers to suggest that they are both ‘classic’ and ‘folk, popular, traditional.’

Uyghur intellectuals and politicians with Soviet education such as Säypidin Äziz were doubtless influenced by Soviet ideas about ethnic traditions. But the “folk classical” idea never became dominant in Soviet maqâm scholarship. In the 1930s, V.M. Beliaev determined that the Uzbek Shashmaqom were derived from Arabic and Persian music and so could not be considered an ethnic folk tradition, while in the 1950s Iunus Akbarov included Shashmaqom in collections of “Uzbek folk music.”(45) Definitions of the maqâm traditions were refined in a debate in the early 1960s, from which it was concluded that the Shashmaqom are native professional music that has evolved from folk music (Levin 1984: 84, 88-89). Since then scholars writing in Russian on Uzbek music, have consistently distinguished narodnaia ‘folk, popular, national’ from professional’naia ‘professional’ music of the Shashmaqom.

In his Soviet publications on the Uyghur Twelve Muqams, Batur Ärshidinov describes the poetry as classical, and the music as folk (BÄ1970, BÄ1987). Other important studies of the muqams in the Soviet/Russian sphere term them “oral traditional professional” music.(46) It appears that in the Soviet sphere native Uyghur and Uzbek scholars tend to connect the cosmopolitan professional art music more with ethnically distinct folk music, while non-natives are more likely to reject ethnic distinctiveness and acknowledge international influences. The stigma attached to acknowledging foreign influences in the present territory of China has prevented strong statements about Persian origins by any scholars, and in China scholars tend to view the Uyghur muqams as autochthonous.

Rather than denoting elite art music and institutions, the term klassik is used to identify the muqams as an old, enduring musical tradition that can be compared to the great works of European, Chinese or Japanese tradition. The Soviet Uyghur composer Quddus Khojamyarov (Kuddus Kyzham’iarov) drew on this understanding of klassik when he described his struggles to dissuade the Muqam Research Group editors from changing the music and texts.

Quddus told me,

The muqams are treasures of the Uyghur past, and like
Chinese opera or Japanese gagaku music, they should
be preserved without changes.  They are monuments like
the Taj Mahal, or Zhong Nanhai in Beijing.  They must be
protected and respected as records of Uyghur culture.  One does
not say Faust is old and put new texts into his works.

Quddus said that one may borrow musical elements from the muqams–as he himself has done in some of his compositions–but should compose in European forms, in modern style.

Do not take old music and put in new style texts.  Preserve the
medieval philosophical poetry of the muqams.  If one wants socialist art,
it should be done entirely modern, not by changing the old” (F 11/4-11/11/92).

For Quddus, a classical ethnic tradition should preserve its roots unchanged, and be incorporated into internationally approved European style music through new compositions, not the revision of the old. Classical and monumental culture is most potent and valuable as an unchanged treasure of the past, not a staging area for creating modern ethnic culture.

Although muqam songs were being performed as a popular form on stage and by ensembles in the 1930s, it appears that only in 1945 did nascent Uyghur cultural nationalism in the Eastern Turkistan Republic lead ETR official Äkhmätjan Qasimiy to promote the study and transmission of Rozi Tämbur’s muqam tradition in the Ili region (BÄ1970:7). In the 1950s the ETR’s offical sponsorship of the Uyghur muqams seem to have stimulated Säypiddin Äziz–another former member of the ETR government who became Chairman (ra’is) of the Xinjiang regional goverment (his leadship extended from the mid-1950s until the mid-1980s)–to promote the recording and transcription of the muqams under government sponsorship. Already in the 1950s, Säypiddin Äziz’s goal in these recordings was not only to collect and preserve the most monumental and archaic form of the muqams as an important part of Uyghur cultural heritage, but also to revise them to suit contemporary political goals.

He wrote in the foreword to the 1960 publication of the muqam musical transcriptions: “The collection, editing, and first complete publication in history of the Uyghur people’s famous classical music ‘Twelve Muqams’ is a great event in the development of Uyghur classical literature and art.” For him, this classical art form expressed the life and ideas of working class ancestors of the Uyghurs, but rather than simply being preserved, he writes it must be researched, edited, made complete, and its rich form and contents continuously developed as part of the political project of creating a new society and culture in China: “We must develop new modern literature and art and give full expression to the new socialist life and socialist relations of the people (khälq) of every nationality.” He went on to states that art and literature must express the new realities of socialist life and contribute to the development of socialism and communism, but it must also be in an appropriate national form. The people of each nationality must develop its own heritage of classical literature and art.

The Twelve Muqams are a rich national musical tradition,
so in the development of the Uyghurs’ modern music it is of great
significance to study deeply and correctly carry on the artistic
successes of the Twelve Muqams” (vol. I, p. 1-2, introduction
also reprinted separately in 1992).

Obviously Säypidin Äziz’s words reflect the Soviet and Chinese communist doctrine that arts should be national in form and socialist in content. But as with Quddus’s comments above, they also reveal the difficult conflict between the goals of researching and preserving a past tradition while also using the same material to make a new socialist tradition. These contradictory goals continue to make editing the muqams difficult. Because the Uyghur Twelve Muqams have been valorized or even sacralized as unchanging monumental art, it is difficult to change them without losing some of their force as a symbol of Uyghur history, culture, and ethnic identity. While there have been efforts to create new music based in the muqams, the symbolic importance of authenticity and the political risks inherent in allowing living artists to gain authority over a politicizes tradition have meant little attention is given to new creations.

Despite the prominence of the muqams as cultural symbols, it has proved difficult to give them convincing socialist content because it would mean creating entirely new poetry, entirely different from the traditional style. This was attempted briefly during the Cultural Revolution but given up. Instead, the texts of the muqams have been edited to “modernize” them, reducing the importance of religious ideas, replacing archaic words, and using more recent and distinctively Uyghur poetry. Because of this compromise, many politicians and scholars see the texts as ideologically contaminated, and the texts are largely ignored. The Sufi imagery in them is interpreted as love songs, and since most of them are sung in chorus they are barely possible to make out.

What Säypidin Äziz described as a tradition of classical music and literature is now generally considered a tradition of classical music. Neither the traditional texts nor the revised texts have been officially published and when I told Uyghurs I was studying the Twelve Muqams they almost always assumed I was studying the music. The texts that are well-known are different in emphasis from those sung by Turdi Akhun. The most commonly sung and recorded bash muqam uses a text that was not in the traditional Kashghar muqam repertoire, but fits the idea of “national in form” better than most muqam songs would. It is a comparison of Arabic, Persian, and Turkic words for attributes of the beloved, and thus reinforces the ethno-nationalist cultural distinctions that are so important in socialist states. But even so it uses a Persian verb which would not be known to most speakers of modern Uyghur, and it suggests that Uyghurs speak Türki which raises the spectre of a pan-Turkic identity that Chinese authorities try to suppress (see Chapter Eight for my discussion of this verse).

The editors of the modern Uyghur muqam texts justify their work as a pragmatic, populist reformation of a native tradition that has been contaminated by foreign religious culture. No one I talked to was trying to make the muqams socialist in content, but they were cleaning up the tradition to prevent active contradiction of socialist public culture. Qawul Akhun explained that the muqams pre-dated the coming of Islam, but singers had had to include religious songs in order to overcome restrictions by Muslim authorities. The religious references now had to be removed if the songs were going to be broadcast on state radio, since the state could not support religion. Ömär Imin, a member of the Muqam Research Group who was editing the texts, justified the substitution of archaic words and replacing Navâ’î’s fifteenth-century Chaghatay ghazals with nineteenth-century “Uyghur” ghazals by saying these changes would make the tradition comprehensible to popular audiences.

Once the muqams were edited, they became the subject of a major effort to disseminate them to local performers from all over Xinjiang. Mämät Ili, a retired math teacher from Khotän told me of being brought to Ürümchi in the late 1980s as part of a special program to teach provincial performers a few of the revised muqams. This is the completion of the policy of taking popular culture from the masses, reforming it, and returning it to them, which I discuss further below.

The idea of the muqams as folk classical music is constant in publications in China, appearing again in the title of Abdushukur Muhämmät Imin’s Uyghur khälq kilassik muzikisi ‘On ikki muqam’ häqqidä, published in 1980. This is the first historical analysis of the muqams that appeared in China, and reflects the author’s understanding of them as a continuous classical tradition. Abdushukur is concerned primarily with establishing the autochthonous origins of the Uyghur muqams. He argues that the muqams predate the coming of Islam, and thus do not derive from the Arabic maqâms. The use of Arabic words as Uyghur terms for musical instruments and musical genres and works is the result of Islamization, and did not mean that the contents had been Arabicized. He also suggests that the greater length and elaboration of the Uyghur muqams shows that the similar Turkish, Azerbaijani, Uzbek and Tajik “muqams” are all derived from the Uyghur muqams. He says Beliaev’s work on the Tajik Shashmaqom (six maqâms) shows that they developed later, and they lack the Çoñ Näghmä, Dastan and Mäshräp of the Uyghur Twelve Muqams. In a footnote, he enlarges on this relationship, attributing the dissemination of the Turkic muqams from Central Asia to Anatolia to Seljuq rule during the eleventh century. He concludes his sweeping historical argument:

Therefore, the Twelve Muqams are in no way a foreign
heritage or ‘foreign love songs,’ but rather are an
achievement of the Uyghur people, and an important treasure
in the musical wealth of our multi-national homeland, the
People’s Republic of China.(47)

Abdushukur’s arguments are not universally accepted among Uyghurs, but people did frequently tell me similar accounts of the history of the Uyghur muqams. The muqams were widely seen as unique classical music: an important and enduring part of Uyghur cultural heritage, distinctive and untouched by outside influences, although borrowed widely. It was felt to be a comprehensive collection, representative of the literary and musical attainments of the Uyghurs, and often described as an encyclopedia.(48)

As I discuss in Chapter Six, some scholars and musicians did not feel that acknowledging non-Turkic elements in Uyghur musical and literary traditions threatened cultural significance or autonomy. One scholar I talked to recognized that all languages are mixed, and compared the Turks to Europeans who used Latin or English, Germans and Russians who used French at court, arguing that such strong linguistic influences were not a problem. Most Uyghur scholars working outside of China did not seem so emphatic in describing the tradition as purely Uyghur, although Tamara Alibakieva and Quddus Khojamyarov both did claim that the tradition began in the Tarim region in the first millenium.

But many Uyghurs did resent the contamination of their language with foreign words. Persian forms in the ghazal songs disturbed many people, and people usually attributed difficulties in understanding the muqam songs to the foreignness of the words, rather than to archaic language. But even though `Alî-Shîr Navâ’î’s Persianate language was often questioned, he was recognized as a key figure in Uyghur literary history. Several scholars objected to Navâ’î being described in Uzbek literary histories as a founder of literary Uzbek.(49)

Although largely implicit, the conceptual problems in defining possession of past culture and history loom large: Persian language is seen as part of Persian culture, and Uyghurs who use Persian culture lose the purity of their Uyghur identity. Likewise, accepting Arabo-Persian maqâms as the source of the Uyghur muqams denies the music its autochthonous essence. On the other hand, the use of Arabic and Persian names for people and for musical forms could be more easily explained away: these were merely new labels that were adopted without changing the Uyghur essence. Although a song’s or person’s name was considered foreign, the person or song could still be autochthonous. Only when a song or person used foreign words did this foreign cultural content imply foreign essence or identity.

Even so, some Uyghurs advocated using native Uyghur words for foreign inventions. Others sought non-foreign etymologies to explain words that appeared to be borrowed, or invented new words to replace foreign names for things. My tämbur teacher Shir Mämät derived the word tämbur from Uyghur roots, and gave up the Persian word nahun for plectrum, preferring the invented Uyghur word simtir instead.(50)

The Uyghur situation can be illuminated through comparison with Ziya Gökalp’s discussion of how to create Turkish national modernity. He argues that much is at stake politically when he insists on clearly distinguishing foreign and native national cultural characteristics, and we find again the identification of the nation, the people and the folk in his use of the word halk. In his 1923 article “Halka Doghru” (“Towards the Folk”), Ziya Gökalp defends the nascent Turkish nation, and justifies rejecting Persianate Ottoman culture. Although the Ottoman elite tried to keep the Ottoman Commonwealth together, he argued that inevitably the component nations would seek independence and their own “homogeneous, genuine, natural social life.” The Ottomans were like somnambulists: “While their real language was Turkish, they talked an artificial language in their delirium. In poetry, they put aside their own metre and sang in artificial metres copied from the Persians.” But in waking them to the “sacred word Turk,” Turkists taught the Ottoman elite “the name of the nation” and “the beautiful language of the nation.”(51)

Throughout his writings, Gökalp distinguishes folk culture–the harmonious and unified expression of ethnic emotional life–from elite civilization–a mixture of intellectual institutions composed by borrowing ideas that freely circulate across national boundaries. For example, Turkish music “arose spontaneously from the people” while Ottoman music “was originally adapted by Fârâbî from Byzantium.” The Turkish music was culture, “composed of sentiments which cannot be developed artificially, and cannot be transmitted from nation to nation.” It was melodies without stereotyped rules, that were the sincere expression of the emotions of the Turkish people. Ottoman music was civilization, “the sum total of the concepts and techniques developed according to certain methods and transmitted from nation to nation.”(52)

Unlike the Uyghur argument that Turkic music theory was taken to the Arabs by Fârâbî or by the Seljuqs, Gökalp consistently attributes Byzantine origins to the music that Fârâbî converted to Arabic style. Thus Gökalp can argue that music is part of the elite civilization that spread among the upper classes of the Arabs, Persians and Turks, but was not taken up by the folk. “Turkish masses continued to play their older music which they had when they belonged to the civilization of the Far East, and produced a folk music out of it. The Arab and Persian peoples did the same thing” (1959: 272-3). Uyghur nationalist scholars in China view all Uyghur musical traditions as folk, although they describe some folk music as “classical.” They claim Fârâbî as a culture hero who took Uyghur music abroad, while Ziya Gökalp finds two traditions of music in Turkey, and selects native traditions of religious music performed in mosques and tekkes for promotion while rejecting the foreign rule-bound musical traditions of the courts. The Uyghur scholars distinguish elite and folk strata of literature but not of musical culture, while Gökalp dichotomizes both music and literature into folk and elite. But even Gökalp identifies certain “practical arts” as solely the work of common people: he argues that architecture, calligraphy, dyeing, weaving and joinery have thus remained genuine Turkish arts (1959:107-108, 263).

Like Uyghur scholars, Gökalp believes that Turkic culture influenced the origins of Western civilization, and uses this to explain why he accepts Western cultural practices while he rejecting those of Byzantine and Iranian civilization. He goes back to early history, claiming Turanian peoples founded Mediterrean civilization, and explains that these Turks were driven “temporarily towards the Far East” by Semites and Aryans. “Much later, Muslim Arabs, Persians, and Turks again improved this civilization and became the teachers of the uncivilized Europeans. . . . We are connected with Western civilization through several contributions, and thus have a share in it” (1959: 267).

Unlike Gökalp and many Uyghur scholars, Soviet scholars such as Plakhov(53) and Shklovskaia (1991) who describe all maqâms as art music are not concerned with the connection of folk arts with national origins and destiny. Their ideas about tradition are shaped by the economic relations among social classes: they do not recognize a strong relationship between tradition and cultural identity.

The ideological discourses about the social and cultural situation of the muqams are only one aspect of how they are used in political contexts. The scholars’ arguments over the muqams in which “the poor muqams are pulled this way and that, played this way and that” (in Quddus Khojamyarov’s words), provide the background for the political uses of the muqams in speeches and media representations.

To suggest how important a role muqams can play in the political imagination of Uyghur history, I will discuss some presentations at the Second Scientific Convention on the Uyghur Twelve Muqams held in Yarkand in August, 1992. I begin by paraphrasing some of the speech by Tömür Dawamat, Xinjiang Chairman.

The Twelve Muqams are the jewel of Eastern music.
They are the treasure of the industrious Uyghur
people’s intellect, and express all aspects of Uyghur
life in a capacious artistic encyclopedia.  The Twelve
Muqams are a priceless part of China’s cultural

Professor Anderson [Bakewell] said that Xinjiang
is the cradle of the musical wealth of Asia, and no
maqâms exist elsewhere that are as large as those of
the Uyghurs.  The Twelve Muqams are a brilliant
representative of Uyghur culture and are imbued with
the ancient Uyghur spiritual world, social life and
psychology.  They have inspired people to a passion
for social justice, moral relationships, and admirable
customs and life. . . .

We must remember that before Liberation only a few
muqam performers remained.  After Liberation the Party
and the People’s Government stimulated and developed
this great musical wealth.  Our artists and scholars
have spent forty-one years in difficult but successful
research, collection, editing, performing, recording,
staging and publishing.(54)

Tömür Dawamat goes on to describe that more collecting, research, analysis and publishing needs to be done.

At the same conference Ismayil Ähmäd lauds recent achievements in muqam research, and stresses the need for further international exchange and cooperation, but concludes with his hope that “since the muqams appeared in this country, research here should stay in the front ranks, and results remain at the highest levels, in the world” (ibid., 1-3).

Finally, the former Chairman of Xinjiang, Säypidin Äziz, who was an important impetus for muqam research in the 1950s, published a small book in 1992 that presents his opinions about how to continue the work that has gone into them. He reviews the history of the muqams since the first historical records, and then encourages stabilization of the muqams, and use of them to show the greatness of Uyghur cultural development under Communist rule.

But Säypidin goes further in explaining the work that still has to be done, including the composition of new poetry in the modern Uyghur language to make some of the muqams more comprehensible and popular. He mentions that he wrote some poetry with Abdushukur Muhämmät Imin and a few other poets for the Çargah and Rak muqams. He is also concerned that the muqams be presented in international form, using “international notation” (staff notation) and modern research methods.(55)

The Uyghur Twelve Muqams exist in political modernity dominated by ideas about collective cultural property and national self-presentation. They are part of political strategies for claiming world importance in the history of civilization, and national importance as a valued minority within China. The practices of modernity arise from internationally disseminated ideas about how to create a national people, territory, culture, and history. These ideas impose conflicting demands for national uniqueness combined with international integration. All that is unique about a people must be explained in internationally accepted ways. In the rest of this dissertation I discuss and analyze the literary and historical background to the muqams, the verbal techniques of muqam performance, and the transformations that have turned the muqams into the Uyghur Twelve Muqams.

Chapter Two


“[The fact that the Qur’ân includes extensive history] is
the clearest proof there is of the excellence of the
science of history, and upon the benefits of this sciences
all nations are agreed.  Most nations, indeed all the
people of the world, practice this science and tell stories
and tales of their forebears, and they use them as proofs
and justifications, particularly the tribes of the Türks,
who in all exposition and affairs, indeed even in most of
their every-day conversations, tell both stories [hikâyât]
and fables [rivâyât] of their ancestors and forebears.”

Mirzâ Haydar Dughlat.  Târîkh-i Rashidi; A
history of the Khans of Moghulistan.  English
translation and annotation by Wheeler M.
Thackston.  Sources of Oriental Languages and
Literatures #38.  Harvard University, Department
of Near Eastern Languages and Civilizations,
1996, p. 3.  Translation modified.


One of the foundational events for modern Uyghur identity is the revival of the Uyghur name in Tashkent in 1921.(56) European historians, drawing on the work of V.V. Bartol’d, stress that oasis dwellers would use local city or village names as ethnonyms, and only with the re-creation of the Uyghur identity did a shared regional identity become recognized by members of these diverse communities.(57) But despite scholars reluctance to accept it, local people also were aware of a shared Turkic linguistic and cultural identity. Although the political term Eastern Turkistan was not a local term until after its use in the Russian Empire, Türk tili and later Türki were the commonly accepted terms for the language spoken.(58) Identity, like historical continuity, depends on symbols, discourses, ideology and actions, as well as on the social and political context of expression. There are many interacting sources of identity for people in this region, and Bartol’d’s focus on toponyms as the basic source for popular identity denies them their complex self-understanding. To understand how people in the Tarim region thought of themselves, we must observe the many different cultural practices that contributed to their identities.

Uighurs ruled the area of Turfan between the mid-ninth century and around 1130, when they submitted to the rule by the Qara Khitai. But in 1209 the Uighur ruler known as Iduq qut (literally ‘holy fortune,’) rebelled against the Qara Khitai and submitted voluntarily to Chingiz Khan. The Uighurs were one of the first non-Mongol peoples to join the nascent Mongol empire, thus earning a respected priority in the Mongol bureaucracy, particularly as the literati (known as bakhshi, from the Chinese boshi ‘doctor’), who organized the Mongol chancellery and created the first Mongol alphabet.(59)

Although Uighurs continued using their name as an ethnonym under the Mongols, the Uighur Iduq qut’s rule was intermittent as Mongol politics changed. Chingizid Mongols ruled during some of the thirteenth and fourteenth centuries, and all of the fifteenth century. With the progressive conversion to Islam of the Mongol and Turkic peoples in Central Asia and the Tarim Basin, Uighur became a term for Buddhist Türks or infidels, and was largely forgotten by the sixteenth century when most of the Buddhist Türks were converted to Islam.(60) The name “Uighur” was known in Eastern Turkistan before 1921, but not used as an ethnonym. Nineteenth century European-trained scholars who studied the history of this region concluded that the modern people were related to the Uighurs.(61)

A comment by the sixteenth century Moghul historian Mîrzâ Muhammad Haydar that is widely interpreted to mean that the name ‘Uighur’ was no longer known when he wrote seems to be misunderstood.(62) Wheeler Thackston recently re-translated Mîrzâ Haydar’s discussion of Juvaynî’s definition of Moghulistan in the Târîkh-i Jahângushay: “‘On the east it is bounded by Cathay; on the west it abuts the land of the Uyghur; on the north it is bounded by the Kirghiz and the Selenga river, and on the south by the Tang’ut and Tibet.’ . . . At present Cathay is clearly known, but what is called the province of the Uyghur is not now known at all” (1996: 225). It is not that the Uighurs were not known, but that their territory was no longer recognized. In Mîrzâ Haydar’s day Moghulistan was not to the east of the old Uighur territory, but to the west, since the Uighurs had ruled in the region of the Turfan and Besh Baliq. It would thus be difficult for Mîrzâ Haydar to understand this description, even if he had some sense of where Uighuristan was.

The Uighur name had another trajectory in Chinese. The name Huihu developed as a transliteration of Uighur, but later came to be used to refer to all Muslims, both within China and from abroad. Huijiao, or “Hui religion,” meant Islam, and the Turkic Muslims were distinguished from Chinese-speaking Muslims by calling the former Chuantou Hui “turbaned Hui” and the latter Huihui, simply reduplicating the name.(63)

The distinction and connection of Turkic peoples and Uighurs has been an issue throughout their history. Identities are not fixed and exclusive, but ambiguous and flexible, in order to be useful in changing historical circumstances. Franke points out that Mongol strategies for legitimating their rule are “like so many different garments which they could wear according to what seemed appropriate for the particular occasion.”(64) Likewise, histories and genealogies and names are multiple and flexible. Dominant or synoptic versions exist only through the exertion of institutional power, laying claim to the high ground of truth from which history can be told from a master perspective.


The beginnings of Uighur recorded history can only be followed in Chinese sources. They were part of the Turkic tribe called the Gaoqü in Chinese (Gaoqü means ‘High Carts’ and reflects their mode of steppe transportation). The Gaoqü were part of the Tiele, a weak but widespread group of nomadic tribes who lived mostly north and west of the Rouran, but appear to have been involved in relations with many of the different northern Chinese kingdoms that existed in the fourth century. In the first decade of the fifth century, the Gaoqü segment of the Tiele came under Rouran rule. In 429 when the Tabghach inflicted a major defeat upon the Rouran and their Gaoqü vassals, the Gaoqü revolted against the Rouran and submitted to the Tabghach. Other Tiele tribes revolted against the Rouran later in the same century, and also joined the Tabghach in attacks on the Rouran, while the Hephthalites seem to have allied with the Rouran in efforts to destroy the Tiele. In 447 the Tabghach settled 3000 families of the Tiele, Dingling and Gaoqü in the capital Pingcheng.(65)

In 486 one of the twelve sub-tribes of the Gaoqü escaped from Rouran rule in the Orkhon area and established a Qaghanate based in the towns of Besh Baliq and Qocho, and ruling in Jungaria where later the Uighurs would establish themselves. With Tabghach support they managed to resist Rouran and Hephthalite attacks in this area until 540. Their close relations with the Tabghach included extensive tribute and gift exchange, including the custom of the court orchestra, which used a set of 80 instruments and some musicians given by the Tabghach.(66) The Chinese dynastic history for 554 describes the Tiele territory as stretching north from this region.(67)

In the history of the Sui Dynasty (581-617) the Tiele are described as stretching far to the West, but in the area of the Tola and Selenga rivers (now in northern Mongolia) they included the Uighur, Tongra, Bayirqu, Buqu, and Fule tribes. The union of these was called the Uighur as well. Recent scholarship has shown that several groups among the Uighur tribes appear in Byzantine history during the fifth to eighth centuries. Eventually the tribal structure was nominally stable with nine or ten Uighur tribes.(68)

The Tiele were defeated by Bumin Qaghan of the Türks in a battle in 546, and after the division of the Türk Empire into two parts around 582 the Tiele were incorporated into the West Türk Empire. They and other tribes revolted against the West Türks in 605 but only remained independent until 611.(69) After the fall of the East Türks in 630, the Western Regions entered a period of chaotic conflict between several different powers: the Tibetans in the process of forming an empire, Tang China reasserting its power, and the various tribes of the Türks.

The Uighurs struggled with the Syr Tardush for control of northern Mongolia, and in 646 the Uighurs (Huihe) brought tribute to the Chinese court at Changan, and were rewarded for their successes in fighting the Syr Tardush. The whole of the Mongolian area was under Chinese control, and the Uighurs lived in the Hanghai mountains (where they would later establish the capital of their empire).(70) In 659 the Tiele (identified differently as Huihe or Nine Named Tiele depending on the Chinese source) in the Tian Shan area responded to Chinese attacks with a widespread rebellion against Chinese rule, and were only pacified by 661. The Tibetans and the Türks allied and drove the Chinese out of the Tian Shan area around 663, but the Türks submitted to Tibetan rule in 667 (Beckwith 1987: 28-33).

In 688, the Uighurs were badly defeated by the rising Second Türk Empire under Ilterish Qaghan. Based on his analysis of the Uighur monument inscriptions, Kliashtornyi argues that this marks the end of a Uighur empire in Northern Mongolia (ca. 605-688) that has been ignored by scholars because it did not gain recognition from the Chinese.(71)

The Türk inscriptions composed between 714 and 735 suggest that the Uighur confederation of north Mongolia had fallen apart and the Uighurs themselves became a minor tribe within the Türk Empire. They are only mentioned once (BQ E37). The Toqquz Oghuz had been the enemies of Ilterish Qaghan (KT E14), but were incorporated into the Empire, so much so that the Qaghan considered them “my own people”(72) although he also distinguishes the Türk and the Oghuz as the two main peoples of the Empire (KT E22).

The resentment of the Tiele, Toqquz Oghuz, Qarluq and Basmil tribes towards the East Türks grew as a result of the latter’s constant attacks upon them, and the Tang Chinese were able to gain the nominal submission and alliance of many segments of these tribes. After the death of Bilgä Qaghan in 734, the East Türks fragmented into rule under several successors, and were overthrown in 742 by an alliance of Qarluq, Uighur and Basmil, who divided the rule among themselves until 744 when the Uighurs and Qarluq forced the Basmil from power. Then in 745 the Uighurs pushed aside the Qarluq as well, and unified the empire under their own command.(73)

When driven out by the Uighurs, most of the Qarluq migrated to Central Asia where they became a ruling power. Initially allied with the Chinese, at the Battle of Talas they switched sides and joined the Arabs, resulting in a major defeat for the Chinese. They later took control of the Suyab (Chu) and Talas river valleys around 766, and eventually became central to the Qarakhanid Qaghanate founded around 840.(74)

Despite the dominance of Omeljan Pritsak’s conclusion that the Qarakhanid originate from the Qarluq tribe, just as with the uncertainty around the Uighur-Toqquz Oghuz identity, the fluid identities and conflicting historical sources allow other interpretations. Modern Uyghur scholars such as Änvär Baytur, Khäyrinisa Sidiq, and Haji Nurhaji argue that the the Yaghma were an important founding tribe of the “Orkhon Uighur” Qaghanate from 745-840 and with the fall of the Orkhon Uighurs, they were one of the first tribes to show up in Qarluq territory, where by 880 they held the city of Balasaghun and become dominant within the Qarluq political structure. They go on to become the founders of the Qarakhanids as well, and their descendant Sultan Satoq Boghra Khan converted the Qarakhanids to Islam. While they mention some of the sources that suggest alternative narratives, Änvär Baytur and Khäyrinisa Sidiq manage to explain them away and end the story of Satoq’s conversion to Islam with the summary, “Thus the Uighur Yaghma tribe established the first Islamic government in history, the Qarakhanid dynasty came into existence.” To make their point even clearer they write that the Qarakhanid ruling family “should be understood as a continuation of the rulers of the Orkhon Uighur Qaghanate.”(75)

On the other hand, the Soviet historian, politician and publicist Mashur Ruziev, in his study of Uighur ethnogenesis, offers a slightly less strictly linear connection for these two Qaghanates: “The basic nucleus of the [Qarakhanid Empire] was formed by the Qarluq and Yaghma tribes, who were kin to the Uyghurs. . . . In the beginning the Qarakhanid dynasty lay within the realm of the Uighurs. . . . The process of the Qarluq and Yaghma tribes mixing in the Qarakhanid state, and the Qarluq and Uighur tribes mixing in the Turfan State led to the formation of a common language.” His conclusion is still towards continuity in Uighur/Uyghur history: he says that the linguist Shcherbak has proven “the continuous historical development of culture and literature of the Uighurs from the Orkhon [Qaghanate] to Eastern Türkistan.”(76)

As we will see, despite Mahmûd Kâshgharî’s rejection of ties to the Uighurs, many of the proverbs documented in the Uighur kingdoms are similar to those popular among the Qarakhanids as well, while those of the Türk Empire inscriptions show less continuity with later forms.


The Uighurs legitimated their claim to the Qaghanate by occupying the sacred Ötükän refuge and the Orkhon valley, and by taking up the symbols of empire, including the title of Qaghan and other Türk titles.(77) Like those on the Türk Empire, the Chinese sources on the Uighur Empire are limited to external relations, primarily those with Tang China. When the Uighurs were succesful in taking control of the Qaghanate and the Qaghan took the title Qutluq Bilgä Köl Qaghan, the Chinese court appointed him Huairen (‘Cherishing the benevolence [of the Chinese emperor]’) Qaghan.

Especially in the early years of the Uighur Qaghanate, the Chinese were mostly concerned with using the Uighurs to help put down the rebellion of general An Lushan, which began in 755. Uighur military support for the Tang arrived only in 757, but led to heavy involvement of Uighurs in Tang Chinese affairs for years after. It was perhaps encounters with Sogdians in the Tang capital, and the usefulness of a religion that was distinct from that of the Chinese, that led the Uighur Qaghan Bögü to convert from Buddhism to Manicheanism in 762. Eventually most of the Uighur elite converted to Manicheanism.

The relationship between the Chinese and Uighurs was full of friction arising from the radically different economic, political and military cultures of the Uighurs and the Chinese. Military and civilian contact in the Tang capital and elsewhere, and the large gifts that the Uighurs extracted from the Emperor, motivated continual incidents. Relations deteriorated to the point that the Uighurs invaded Taiyüan in 778, but they were eventually driven off by the Chinese. In 780 these tensions and a rift between Sogdians and Uighurs following the accession of Tun Bagha Qaghan to power in 779, led to a Chinese plan to eliminate most of the Sogdians and Uighurs in the capital.(78)

Chinese relations with the Uighurs remained strained until 788 when the threat of Tibetan expansion in the area of Qocho (Turfan) and Besh Baliq(79) led the Qaghan and the Tang court to exchange embassies and princesses and to renew their alliance. The new alliance did not last long, but there appear to have been no active hostilities between Uighurs and Chinese. In 821 a wedding established a marriage alliance, and a Chinese princess was sent to the Qaghan. Silk and horse trade resumed, and the Chinese again had to deal with incursions of Uighurs into the empire as traders and mercenaries, and often refused the soldiers entrance to China even when Chinese sorely needed the trade goods (Mackerras 1972: 112-123). Nonetheless, allies or not, the Uighurs and Chinese fought frequently over territory along the north and east of the Tarim and Turfan Basins, and in the Gansu corridor both fought against Tibetans and their Qarluq allies between 789 and 823. In 823 treaties signed by Chinese, Tibetans and Uighurs ended these conflicts. These held until the destruction of the Uighur Empire by the Qirghiz in 840 (Beckwith 1987:164-168).

The Qirghiz did not take over the whole territory of the former Türk Empire when they defeated the Uighurs in 840, and the Qarluqs filled the power void. Mas’udi and other Arab writers describe the Qarluqs as the supreme leaders of the Türks (Golden in CHEIA, 349-350).


The early Uighurs were politically distinct from the Türks, but drew heavily on Türk methods of rule and legitimation. After the fall of the Türk Empire in the 730s, Türk became a cultural and linguistic identity, shared by Uighurs, Oghuz, Qarluqs, and others. By the time of the Iduq qut Uighur kingdom in the Turfan area, the Uighurs could readily accept Türk identity to describe their language. One of the most important works of Uighur Buddhism is a translation from Chinese of a work describing the trip of the seventh century Chinese pilgrim Xuan Zang. He describes his journey through the Western Regions to India to get Sanscrit training and collect Buddhist scriptures.

In the colophon of this early to mid-tenth century work, the Uighur monk Singqu Sali Tutung of Besh Baliq (the Idiqut capital on the north side of Boghda mountain from Turfan) explains that this work was translated from Chinese (Tavghaç tili) to Turkic (Türk tilinçä ävirmish). The same author translated the Suvarnaprabhasa where he states that the work was translated from Chinese into ‘Türk Uighur’ (Türk Uyghur tilinçä ikiläyü ävirmish).(80) Many other “Uighur” translations identify the language as Türk in the colophons.(81)

Hamilton argues that Türk served as the name for the language and was applied to all related languages, even that of their enemy, the Uighurs. He disagrees with von Gabain who says Türk is used only by the descendants of the Türk Empire.(82) In the next chapter I will refer to as Turkic all those speakers who use türk or türki for their language, and I will refer to as Türk those who Kashghari describes as pure Türk.

Nonetheless Uighur genealogical and political identity remained distinct, and is mentioned in titles and, as we have seen, qualifications of the kind of Türk language. The On Uyghur (Ten Uighur tribes) political structure is mentioned in the fairly early work Maitrisimit (Meetings with Maitreya), which was translated from from Hindi into Tokharian and then ‘into the Türk language’, probably in the eighth or ninth century.(83)

For most of the period of the Uighur oasis kingdoms, Uighurs were a genealogically, politically and religiously defined group, while Türks were most strongly identified by their language. Some Uighurs appear to have had a tradition of genealogical or political descent from Türks, but more importantly the shared language enabled extensive cultural contact and a sense of shared cultural identity that is reflected in oral literature, such as proverbs. But the Uighur tradition of alliterative verse was never shared as far as can be seen from available sources.

The next chapter puts the relationship between Uighurs and Türks in the more complex context of Türks within the Islamic world. I discuss this first in terms of mutual attitudes writers exhibited towards ethnic groups, moving towards a close reading of a few “monuments” of Turkic literature to show how these Turkic authors created and explained generic vehicles for their political and cultural programs. I argue that these authors are attempting to establish distinctively Turkic literary traditions as a way to select and disseminate what they find valuable in Turkic and other cultural traditions. Literary texts are one of the expressive means through which cultural boundaries are negotiated, since they are symbols of the past and of tradition that can be used to control culture, identity and cultural exchange.

Return to the table of contents.

End Notes

1. Despite some rather silly etymologies offered for the word “Taklamakan,” it obviously derives from the Turki name “tärkli makan” or “dry place” that occurs in a number of Turki manuscript sources from the 16th century C.E.

2. 2 See the articles Yuan Qing-li “Population Changes in the Xinjiang Uighur Autonomous Region (1949-1984)” CAS 9:1 (1990) 49-73; and Li Yuanqing “Qianlun Xinjiang renkou de jixie zengzhang” XASS 1989:3, 123-124. These are probably the same author: Chinese names often get scrambled in foreign publications.

3. 3 The best synthetic narrative of the nineteenth century history, taking into account local and Chinese sources as well as Russian and Japanese analyses, is Ho-dong Kim’s The Muslim Rebellion and the Kashghar Emirate in Chinese Central Asia, 1864-1877. Dissertation, Harvard University, 1986. He discusses in some detail the Islamic ideology that motivated the 1864 rebellions, but finds there is not enough evidence to understand its roots. He concludes that the rebellions can be loosely defined as “collective political violence unleashed by the Muslim population in Sinkiang against the exploitative system imposed by foreigners whom they conceived as infidels” (87). Other useful studies of the history of this area have been written by O. F. Akimushkin, L. I. Duman, D. A. Isiev, D. I. Tikhonov, and O.V. Zotov.

4. 4 These conversations were part of a broad survey of Uyghur culture and language. Since I was still learning Uyghur, I would ask a question in English, and Hairet would translate the question into Uyghur and present a rather general overview of a response.

5. 5 Uyghurlar, adättä, nakhsha usullar nahayiti qizghindighan bir millät.

6. 6 For a brief description of these instruments, see Appendix I.

7. 7 For further discussion of this factory, see Stanley Toops, The Tourism and Handicraft Industries in Xinjiang; Development and Ethnicity in a Minority Periphery. University of Washington, Ph. D. Dissertation, 1990, pp. 137-140).

8. 8 Biography of An Lu-shan, Howard Levy, trans. Berkeley: University of California Press, 1960, 34. An Lushan is famous for his rebellion against the Tang dynasty almost overthrowing it in 756-58. The Tang court relied on Uighur military aid to put down An Lushan’s rebel forces, which left the Uighurs with great power and influence within the Chinese Empire.

9. 9 Mackerras, 1972:72-75 (October 762).

10. 10 “The Journey of Benedict Goës from Agra to Cathay” in Cathay and the Way Thither, Sir Henry Yule, trans. and ed. New Revised Edition by Henri Cordier. Volume 4, p. 229.

11. 11 Robert Barkeley Shaw. Visits to High Tartary, Yarkand and Kashgar. London: John Murray, 1871. Reprint with introduction by Peter Hopkirk. Hong Kong: Oxford UP, 1984, pp. 169-70.

12. 12 319. Writing in the first years of the twentieth century, the historian Mulla Musa Sayrami describes similar resentment of authoritarian Islamic leadership under Ya`qûb Beg. Tarikhi Hamidi. Änvär Baytur, ed. Beijing, 1986, 522-23.

13. 13 Gegel’ M. Iskhakov. Ethnograficheskoe Izuchenie Uigurov Vostochnogo Turkestana Russkimi Puteshestvennikami vtoroi Poloviny XIX veka. Alma-Ata: Nauka, 1975, 47, citing Valikhanov’s “O zapadnom krae Kitaiskoi imperii.”

14. 14 Inayitilla, Uyghur ussulliriniñ tarikhiy örnäkliri, Ürümchi: Shinjañ Yashlar-Ösmürlär näshriyati, 1987, 276-279. The fire dance also appears in the 1993 documentary film On Ikki Muqam.

15. 15 Overviews of the life of Rozi Tämbur and his involvement in the Ili performing arts scene can be found in Tursun Qahhariy, “Uyghur muqaminiñ miraskhori. Khushnava bulbul ediñ. Rozi Tämbirniñ tughulghinigha–100 zhil.” Kommunizm Tughi. 29 Dec., 1990, and Ämätjan Äkhmidi, “Ataqliq khälq muqamçisi Rozi Tämbur,” article in the author’s Uyghur muqamliri toghrisida. Ürümchi: ShKhN, 1992, pp. 145-55.

16. 16 The most complete musical notations are Uyghur khälq klassik muzikisi On Ikki Muqam. 2 vols. Beijing: Yinyue Chubanshe and Minzu Chubanshe, 1960. (Chinese title: Weiwuer zu minjian gudian yinyue Shi Er Mu-ka-mu) and Tamara Alibakieva, Dvenadtsat’ Uigurskikh mukamov; Rak mukam, Chabayat mukam. Alma-Ata: Önär, 1988. Mätruzi Tursun and Zhou Ji have produced unpublished transcriptions of several of the muqams as presently performed by the Muqam Änsambli in Ürümchi. I have not seen the transcriptions of Rak, Çäbbiyat, and Mushävräk in N. T. Razieva, Uigurskie Mukamy. Alma-Ata: Gylym, 1991.

17. 17 Music Research Department of the China Art Research Institute, and the Xinjiang Art Research Department. “‘Xinjiang Weiwuer zu yinyue yuelü yü diao shi wenti taolun hui'” ceyin gongzuo baogao” [Report of the musical measurements of the Xinjiang Uyghur nationality conference on questions of musical scale and mode], Xinjiang Yishu, 1986, #5, pp. 47-51. Zhou Ji, “Shilun ‘Nawa Mukamu’ de jiegou” Xinjiang Yishu, 1988, #1, pp. 19-26. “Weiwuerzu chuantong yinyuezhong de ‘Nawa diao,'” Xibei Minzu Yanjiu, 1991, #1, pp. 227-236. Jean During and Sabine Trebinjac attempt an even more reductive analysis of muqam scales into half and whole tones: Introduction au Muqam Ouïgour, Papers on Inner Asia, #17. Bloomington: Indiana University Research Institute for Inner Asian Studies, 1991, pp. 32-34.

18. 18 Sabine Trebinjac’s dissertation Musique Ouïgoure et collectes musicales en Chine, Université Paris-X, Nanterre, Jan. 1994, remains inaccessible to me. During and Trebinjac (1991) make broad comparisons among local versions and even different maqâm traditions, and thus elide the complex differences among songs even within a single muqam. Other traditions derived from or influenced by the maqâm concept are analyzed in Karl Signell, Makam: Modal Practice in a Turkish Art Music. Seattle, 1977. Jean During, La musique traditionelle de l’Azerbayjan et la science des Muqams. Baden-Baden, 1988. Theodore Levin, The Music and Tradition of the Bukharan “Shashmaqâm” in Soviet Uzbekistan. Thesis, Princeton University, 1984. Jürgen Elsner, “Zum Problem des Maqam” Acta Musicologica xlvii/2 (1975) 208-30. Habib Hassan Touma, “The Maqâm Phenomenon: An Improvisation Technique in the Music of the Middle East,” Ethnomusicology, 15 (1971), 38-48. Excellent historical material is presented in Angelika Jung, Quellen der traditionellen Kunstmusik der Usbeken und Tadshiken Mittelasiens. Hamburg: Verlag der Musikalienhandlung, Karl Dieter Wagner, 1989; and Owen Wright, The modal system of Arab and Persian music, A.D. 1250-1300. Oxford: Oxford UP, 1978, and Wright, “Aspects of Historical Change in the Turkish Classical Repertoire” Musica Asiatica 5 (1980) 1-108.

19. 19 The system of twelve maqâms comes from ten modes or parda with Arabic and Persian names that are first mentioned in the late eleventh century C.E. by Kai Kâ’ûs ibn Iskandar in his Qâbûs nâmâ. In the thirteenth century Safî al-Dîn wrote the first systematic treatise on the twelve maqâms, and includes eight with the same names as in the Qâbûs nâmâ (Jung 1989, Wright 1978). Harold S. Powers discusses the development of the concept of mode and its application to maqâm and râga music. The New Grove Dictionary of Music and Musicians, v. 12, pp. 376-450, at 422-36. In Jean-Claude Chabrier’s article “Makâm” (EI2, v. 4, 96-104) he discusses the Arabic, Persian and Turkish modes in general terms.

20. 20 See Theodore Levin, The Music and Tradition of the Bukharan Shashmaqâm in Soviet Uzbekistan. Dissertation, Princeton University, 1984.

21. 21 The progression from low register to avj and back to low register is characteristic of most maqâm-derived traditions, including the dastgâh of Iran.

22. 22 Some of the other names that have been used in the past two hundred years in the Kashghar and Ili muqam traditions are Husäyni, Saba, Dugah, Rukhsarä, and Naläsh (see BÄ1970, BÄ1987). The Dolan Muqams are Zil Bayavan, Ärzal Bayavan, Dolan Rak, Dolan Mushaviräk, Sim Bayavan, Bom Bayavan, Dolan Jula, Khudäk, Bayavan, Dugamät Bayavan. Mähämmät Äziz, Dolan Toqquz Muqami, Ürümchi: ShKhN, 1992; Ömär Imin, “Dolan vä Dolan mäshräp-muqamliri häqqidä” Shinjañ sän’iti 9:3 (1989) 59-74 and 9:4 (1989) 44-58. The Qumul muqams are Dur, Ulugh Dur, Mustähzat, Çarigah, Khupti, Çäbiyat, Mushaviräk, Uzhal, Du’a, Dolan Mushaviräk, Iraq and Rak. Eli Isma’il, Qumul näzmiliri, Urumçi: ShKhN, 1991, pp. 1-343.

23. 23 Batur Ärshidinov finds the first modern mention of the muqams in a 1924 article from an Almuta (Alma-Ata) almanac Birinçi Çamdan in which the author, Yüsüpjan Ghapparov, describes thirty-two muqams that are popular at the courts in the Ili region. They are described as bäglär näghmäsi (‘songs for begs [officials]’), and the author goes into some detail describing how they are being gradually forgotten. This fits with comments from several Uyghurs I met who said that there used to be more muqams but they were taken to other places, other countries.

24. 24 Batur Ärshidinov, “On Ikki Muqam Terminlargha Izakhlar,” Kommunizm Tughi. n.d. Article part of a collection of clippings loaned to me by Tursunmuhämmät Sawut, for which I am very thankful. H. Powers does point out that râga is pronounced râg in North India, “Mode”, p. 429.

25. 25 Jung 1989:75-77. A less likely possibility is that the Rak Muqam is derived from the Raghi songs that Mark Slobin identifies as named for the town of Ragh in Badakhshan, or from the style of playing that he terms reig about which I can find no other information. Music in the Culture of Northern Afghanistan. Viking Fund Publications in Anthropology, Number 54. Tucson: University of Arizona Press, 1976, pp. 81 and 230.

26. 26 S:387. John Baily describes chap as a weak stroke in playing the dutâr, opposite to the strong râst stroke. “Recent Changes in the Dutâr of Herat” Asian Music 8(1976) 29-64, at 38.

27. 27 1991:29. Shawrâk appears in Walter Kaufmann’s Selected Musical Terms of non-Western Cultures. Warren, Michigan: Harmonie Park Press, 1990, as a Persian mode that uses the Hijaz scale and is sung slowly, p. 640.

28. 28 N.A. Baskakov, “Mukamy v razvitii muzykal’noi kul’tury tiurskikh narodov.” Sovetskaia tiurkologiia, 1990, #1, pp. 17-25 at 23. Baskakov’s hit and miss identitifications show the problems of not referring to historical sources on music to find which words were actually in use. As long as we are speculating on derivations, it is possible that a mispronunciation of the Arabic mushâwarat ‘consulting, deliberation, advice’ S:1243, motivated the name Mushavräk.

29. 29 See Abdushukur Turdi, “Muqam tätqiqati vä muqam namliriniñ tarikhiy qatlimi toghrisida,” Uyghur On ikki muqami häqqidä, Ürümchi: ShKhN, 1992, p. 76.

30. 30 For an overview of the sources for theoretical ideas about the maqâms in Central Asia, including relationship to astrology and Greek influences on Arabic music theory, see see T. Vyzgo, D. Rashidova, “Muzykal’nogo-teoreticheskoe nasledie velikikh Sredneaziatskikh myslitelei,” in Voprosy muzykal’noi kul’tury Uzbekistana. Sbornik statei vypusk II. Tashkent: Izd. Khudozh. Lit. im. Gafura Guliama, 1969, pp. 162-183. They cite the 16th century theorist Najmuddin Kaukabî who elaborated Ibn Sînâ’s theories: “in early morning, at dawn, perform Râhawî, and all that suits it. At sunrise, perform Ushshâq, and then Râst after that but before noon; at noon perform Irâq. As soon as the sun begins to decline, listen to Buzurg, and halfway to sunset, Bûsalîk. At evening perform Zangûla, and at the hour of the evening prayer perform Zîrâfkand, and Isfahan at the final hour. At midday [mistake for midnight?] perform Hijâz” (181). See also S. Veksler, “Uzbekskie Makomy” in Voprosy Muzykal’noi Kul’tury Uzbekistana. Tashkent, 1961. pp. 72-99. Tamara Älibakieva suggests that youthful audiences prefer the lively and natural Çäbbiyat, Mushavräk, Pänjigah, and Ushshaq, while older people like the peaceful, dignified (shevkätlik), and sad (ghämkin) Rak and Nava muqams. “On ikki Muqam,” Kommunizm Tughi, May 9, 1987, p. 4. See also Johann C. Bürgel, The Feather of Simurgh; The “Licit Magic” of the Arts in Medieval Islam. New York UP, 1988, pp. 91-97.

31. 31 Kai Kâ’ûs ibn Iskandar, Prince of Gurgan. A Mirror for Princes; The Qâbûs Nâma. Reuben Levy, trans. New York: Dutton, 1951, pp. 186-190. He also warns against drinking when performing, and gives advice on how to deal with drunken audience members.

32. 32 Ärshidinov, n.d. He speculates that the twelve shöbä should be as follows: muqäddimä, täzä, nuskha, kiçik säliqä, jula, sänäm, çoñ säliqä, päshru, täkt, mustäzad, yarim saqi, abu çäshmä. This series is thus pieced together from different songs that do not actually appear in the same muqam.

33. 33 The term sarakhbar was not used when I was there. It appears to have been in vogue only briefly in 1987-88 when the Xinjiang Art Research Institute (Yishu Yanjiu Suo) was organizing events, and the muqams were officially described as a part of the larger maqâm tradition. Sarakhbar is also used in the Tajik-Uzbek Shashmaqom tradition as the name for the introductory song. See the history and description in the program Mukam Music and Dance Ensemble of Xinjiang (China). 12th Festival of Asian Arts, Presented by the Urban Council, Oct. 14-Nov. 2, 1988, Hong Kong.

34. 34 Viktor M. Beliaev, Central Asian Music. Greta and Mark Slobin, ed. and transl. Middletown: Wesleyan UP, 1975 [1962], pp. 214, 331. In Persian music, the Sâqi-nâma is a “classical composition based in Hâfiz’s famous poem,” and appears in the Azerbaijani muqam tradition. During, 1988: 47.

35. 35 Inayitilla describes sänäm dances from Qäshqär, Kuça, Dolan/Märkit, Qaghiliq, Korla, and Qumul with extensive song texts, but no musical transcription, in his Uyghur ussulliriniñ tarikhiy örnäkliri, pp. 41-223. A brief collection of texts and music for Sänäm dances can be found in Uyghur khälq sänämliri, Iskändär Säypulla, A. Ömär, eds. Ürümchi: ShKhN, 1984. For the Sama dance in jula see Himit Mehrulla, Lätipä Qurban, Uyghur ussul sän’iti toghrisida. Ürümchi: ShKhN, 1991, 37.

36. 36 Akbarov and Kon show that even in the Khwarazmian Shashmaqom the same usul varies from performer to performer. Khorezmskie makomy, vol. VI of Uzbekskaia narodnaia muzyka. Tashkent: Gos. izd. khudozh. lit., 1958, pp. LIII-LIV.

37. 37 I use aqsaq in the absence of local terminology for this characteristic kind Uyghur rhythm. Theodore Levin points out that the term aqsaq is not used in native rhythm theory in Central Asia (personal communication, 1998).

38. 38 This song is not now used in the Uyghur Twelve Muqams. See also the analysis and table of rhythmic patterns by Wan Tongshu, Uyghur khälq klassik muzikisi On Ikki Muqam, 2 vols. Beijing: Yinyue Chubanshe and Minzu Chubanshe, 1960. Chinese title: Weiwuer zu minjian gudian yinyue Shi Er Mu-ka-mu, vol. 1, pp. 22-23, vol. 2, “Ilavä” [Appendix, notes], pp. 1-15.

39. 39 Mätruzi Tursun, “Uyghur klassik muzikisi on ikki muqam vä uniñ muzikiliq qurulmisi toghrisida,” pp. 389-426; Shav Guañshen [Shao Guangshen] “On ikki muqamniñ udar shäkli vä küy tärtipi gurulmisiniñ ähmiyiti,” pp. 309-326; both in Uyghur on ikki muqami häqqidä (tätqiqat maqaliliridin tallanma), XUAR On ikki muqam tätqiqat ilmiy jäm’iyiti tüzgän, Ürümchi: ShKhN, 1992.

40. 40 1991: 18. Although it does not appear in their bibliography, they are apparently citing Alibakieva’s “Ladovye printsipy uigurskoi mukamnoi monodii,” Izvestiia AN KazSSR. Seriia filologicheskaia. No. 3. 1987.

41. 41 Levin 1984:129, 148-50. However, though melodic rhythm fits `arûz exactly, he finds that any usul rhythm can be applied to any `arûz. “Instrumental interpolations as well as held sung notes provide sufficient metro-rhythmic license in the melody line to assure metric solution for the convergence of any pair of usûl and `arûz cycles.”

42. 42 “Rhythmic Aspects of the Tajik Maqam,” EM, 15 (1971) 100-104, with supplemental note p. 253.

43. 43 Uyghur khälq klassik muzikisi On Ikki Muqam, 1960. This transcription is generally considered to be solely the work of the musicologist Wan Tongshu, although he is not assigned authorship in the book. In fact, Säypiddin Äziz and Turdi Akhun are the only people whose names appear in the text. These transcriptions are detailed and represent an enormous amount of work, but not they are not very accurate according to most people who refer to them.

44. 44 In Uyghur, when referring to cultural practices khälq is used to specify those that are oral, traditional, or communal in some way, as opposed to institutional, individual, or literary. In the titles of all collections of Uyghur oral folklore, khälq is used to modify the name of whatever genre is collected. In order to specify ‘national’ the word milliy is usually used, but not often as an attribute of culture. In discussing the folk music, the term älnäghmä ‘popular/folk song’ was also common.

45. 45 Uzbek khalq muzikasi, Vols. V and VI, Tashkent: 1958 and 1959. V.M. Beliaev and F. M. Karomatov both characterized Tajik and Uzbek Shashmaqom pieces as klassik music (Beliaev, Shashmaqâm, vol. I, Moscow: 1950, p. 11, cited in Levin 1984: 81; Karomatov, Shashmaqâm, vols. I-VI, Tashkent: 1966-1975).

46. 46 A. S. Shklovskaia, Uigurskaia professional’naia muzyka ustnoi traditsii (na materiale mukama). Alma-Ata: Gylym, 1991. Tamara Alibakieva, Dvenadtsat’ Uigurskikh mukamov; Rak mukam, Chabayat mukam. Alma-Ata: Önär, 1988. Tamara Alibakieva relies on scholarship from China, but ignores the “folk” label that scholars in China attach to the muqams. In her brief discussion of the relationship between Uyghur professional and folk music traditions, Alibakieva defines folk as “nonprofessional, oral musical culture of the people.” “K voprosu o sviazi uigurskoi narodnoi i professional’noi muzyki ustnoi traditsii” in Aktual’nye problemy sovetskogo Uigurovedeniia, Alma-Ata: Nauka, 1983, p. 162.

47. 47 Abdushukur Muhämmät Imin, 1980, pp. 1-12, and pp. 197-98n6. Alibakieva draws on his argument and reinforces it with newly published Chinese sources, arguing that “in the begining of the eighth century the prototype of the çoñ näghmä was created” in the city of Kucha. But while she claims this is the prototype, she reports historical events, such as the origins of Arabic music theory in the writings of al-Fârâbî (870-950) and Ibn Sînâ (980-1037). She does not explain how these events are connected to earlier and later forms of the “Uyghur muqams.” Despite a historical narrative connected only by juxtaposition, she makes stronger claims about the historical priority of Uyghur music when she says that V. M. Beliaev “truly describes its genesis” when he characterizes an early twentieth century Uyghur music recording as “transporting us into an atmosphere of austere and hoary antiquity, into an atmosphere of the pre-music of Central Asia” (1988:22-23).

48. 48 For instance Ämätjan Akhmidi distinguishes five classes of Uyghur muqams, among which the Twelve Muqams are first as “the encyclopedia of Uyghur khälq classical music” (Uyghur khälq klassik muzika qamus), while the other five classes are the muqams from Qomul, Dolan, Turpan, and Kucha. (“Uyghur muqamliri vä uniñ muzikiliq qurulmisi toghrisida tätqiqat” in Uyghur on ikki muqami häqqidä (tätqiqat maqaliliridin tallanma), XUAR On ikki muqam tätqiqat ilmiy jäm’iyiti tüzgän, Ürümchi: ShKhN, 1992: 286-308 at 296). Himit Mehrulla and Lätipä Qurban express a more developed version of this idea: “The Twelve Muqams are an artistic encyclopedia of great capacity that describe all aspects of Uyghur life using the language of song, dance and music.” Uyghur ussul sän’iti toghrisida. Ürümchi: ShKhN, 1991, 36.

Soviet Uyghur authors often cited a memorable metaphor from the Qazaq S. Mukanov, who compares the complex composition of the muqams to the Ming-Oy (literally ‘1000 rooms,’ meaning Caves of the 1000 Buddhas found in several oases in the Tarim area), each different in size and time of creation, but united into a single architectural whole. “The Twelve Muqams are an encyclopedia, containing all the song and musical riches of the Uyghur people from a to z.” Shagi Velikana. Alma-Ata: 1967, 232. Cited from Alibakieva, 1988: 22. The idea of traditional ethnic cultural encyclopedias could be found already in the nineteenth century: Ch. Ch. Valikhanov assessed the Qirghiz Manas epic as “an encyclopedical collection of all the Kirghiz mythological tales and traditions.” Cited in Eugene Schuyler, Turkistan; Notes of a Journey in Russian Turkistan, Khokand, Bukhara and Kuldja. Second ed. London: S. Low, Marston, Searle & Rivington, 1876, Vol. II, p. 138.

49. 49 In Uzbek Literary Politics (The Hague: Mouton and Company, 1964, pp. 46-87), Edward Allworth discusses the dangers and benefits of nationalization of Navâ’î as a literary culture hero. Since the modern nation is inextricably linked to a shared national history, language and literary tradition, early authors both establish the language and the nation. But founders must be political models as well. During the Stalinist purges, when there was no allowance for historical context, Navâ’î was politically incorrect because he supported feudalism, so he could not be acknowledged as a founder of the Chaghatay language.

50. 50 See Chapter Six for a discussion of scholars who created Turkic derivations for the names of muqams. Levin claims that in editing the Uzbek Shashmaqom songs, Yunus Rajabiy reduced the foreign contamination of Uzbek culture by avoiding breaking up Uzbek words according to the rules of Persian `arûz meters (1984: 150-160).

51. 51 Turkish Nationalism and Western Civilization, Selected Essays of Ziya Gökalp. Niyazi Berkes, trans. and ed. London: George Allen and Unwin, 1959, pp. 259-62.

52. 52 “Culture and Civilization” 1959:104-109, originally published as “Hars ve medeniyet” in 1923. In fact, in two 1913 articles “Halk medeniyet” (“civilization of the folk/people”) and “An’ane ve Kaide” (“Tradition and Formalism”) he said every nation has two civilizations (medeniyet), and divided them into authentic, natural, traditional folk civilization and artificial, borrowed, rule-governed formalism (kaide). He argues that it was a folk elite that created the Ottoman Empire, and only with the rise of formalism in the education system does the Empire begin to decay. Rules are “always devoid of creativity and growth” while tradition “means creativity and progress. . . . Tradition is something growing and creating by itself, and, moreover, giving life to the borrowed innovations grafted on itself in such a way the the foreign elements do not dry out and become rotten, as happens in ordinary [formalist] imitation.” (1959:89-96).

53. 53 Iu. N. Plakhov, Khudozhestvennyi kanon v sisteme professional’noi vostochnoi monodii. Tashkent: FAN, 1988.

54. 54 “Shinjañ uyghur aptonom rayoni boyinçä Yäkändä ötküzülgän 2-qetimliq uyghur on ikki muqami ilmiy muhakimä yighinida sözlängän söz,” Shinjañ uyghur aptonom rayonluq ikkinçi qetimliq uyghur on ikki muqami ilmiy muhakimä yighinniñ materiyalliri. Ürümchi: Shinjañ Shinkhua basma zavut, 1992, pp. 5-7.

55. 55 Uyghur muqami toghrisida. Beijing: Millätlär näshriyati, 1992, pp. 7-30.

56. 56 I. V. Zakharova, “Materialnaia kultura uigurov Sovetskogo soiuza” in Sredneaziatskii ethnograficheskoi sbornik. Trudy instituta ethnografii imeni Muklukho-Maklaia. Moscow, 1959, 216, although she gives no source for this fact. Joseph Fletcher also describes the revival of the name Uighur in 1921. His is the account referred to in most English discussions. “China and Central Asia, 1368-1884” in The Chinese World Order, John K. Fairbank, ed., Harvard UP, 1968:363n96.

The use of the name spread among the literati from Eastern Türkistan (Xinjiang) who travelled and had contact with Uyghurs in Western Türkistan (Soviet Central Asia). In the early 1930s, Sheng Shicai the warlord who ruled Xinjiang began using the name Uyghur officially in the areas of Urumchi and Ili. But it was not until after the suppression of the short-lived Turkish Independent Republic of Eastern Türkistan based in the Kashghar region that “Uyghur” officially used in South Xinjiang. See Gunnar Jarring, Prints from Kashghar: the printing-office of the Swedish mission in eastern Turkestan. Stockholm: Swedish Research Institute in Instanbul, 1991, pp. 95-7.

57. 57 See Nazif Shahrani, “‘From Tribe to Umma’: Comments on the Dynamics of Identity in Muslim Central Asia” CAS 3:3 (1984) 27-38, for a critical discussion of the complex bases and expressions of ethnic identity in “Turkistan.” He shows the limits to Bartol’d’s oft cited comments on this topic.

58. 58 See Golden IHTP, 115, where he critiques Bartol’d’s claim that the Türk ethnonym was disseminated primarily through Islamization.

59. 59 Thomas Allsen, “The Yüan Dynasty and the Uighurs of Turfan,” in China Among Equals; The Middle Kingdom and its Neighbors, 10th-14th Centuries, Morris Rossabi, ed. Berkeley: University of California Press, 1983, 243-280. Peter Jackson “Bakshî” EIr (1989) Vol. 1, 535-536. Victor Mair “Perso-Turkic Bakshi = Mandarin Po-shih: Learned Doctor,” in Festschrift for Richard Frye. JTS 1992, 117-21.

60. 60 Oda Juten “Uighuristan” AA (1978) 22-45 and Morris Rossabi “Ming China and Turfan, 1406-1517” CAJ 16 (1972) 206-225, describe the later history of Turfan under Moghul rule. Although in early 1400s Buddhist temples were much in evidence and the Turfan area was still considered a Buddhist center, the Mongol rulers were Muslims and by the 1450s all of Turfan was ruled by Muslims. They describe the diplomatic and military contest of Chinese, Moghul, and Oirat rulers over trading rights and the Chinese-controlled city of Hami, throughout he fifteenth century. It is finally conquered by Mansur in the early sixteenth century. China aquiesces to this conquest of its nominal Uighur vassals by the Moghuls.

The abandonment of the name ‘Uighur’ is similar to the suppression of the name Ellinikos among Greeks who converted to Christianity, since their former ethnonym came to mean ‘pagan.’ The advent of European modernity and European intellectuals’ great interest in Ancient Greece motivated Greeks to retrieve a Hellenic past for their national origins. This gave them a uniquely valuable national identity at a time when Christian religious identity was no longer as valuable. They could thus claim cultural priority over Europe, which compensated for the contaminating influence of Ottoman rule and marginality in European modernity. See Michael Herzfeld, Ours Once More: Folklore, Ideology, and the Making of Modern Greece, Austin: University of Texas Press, 1982.

I distinguish Uyghur and Uighur because many Western scholars object to the revival of the Uyghur ethnonym after its long oblivion. Obviously we are not talking about the same Uyghurs, but it is difficult to know how we can talk about the same Greeks, the same Britons, except merely as a convention that must be understood in terms of historical particulars.

61. 61 Julius Klaproth Abhandlung über die Sprache und Schrift die Uiguren. Paris: 1820. Reprint, Hamburg, 1985. According to R.B. Shaw in JRASB (1877) p. 243, A. Rémusat in Langues Tartares already identified Eastern Turki with the language of the Uighurs. But European scholars and explorers used a remarkable variety of names for people in this region. The people were identified as Turki, Taranchi, or Chuantou, or named for the place, variously called Eastern Turkistan, Altishahr (Six Cities), Little Bukhara, or Machin. G.M. Iskhakov, Etnograficheskoe izuchenie Uigurov vostochnogo Turkestana russkimi puteshestvennikami vtoroi poloviny XIX veka. Alma-Ata: “Nauka,” 1975, pp. 34, 52, 67. Fletcher mentions a Hami prince who wrote letter to Tao Mo in 1890s claiming a family tradition of descent from “Uighur Muslims,” thus showing that the name had not been lost as a term and idea (1968:364).

62. 62 Golden IHTP, 409, and Dru Gladney, “Ethnogenesis of the Uighur,” CAS 9 (1990) 1, 1-28 at 9, are examples of the many uses of this citation.

63. 63 See Gladney, 1991: 15-21, for a discussion of the derivation of the Hui ethnonym.

64. 64 From Tribal Chieftain to Universal Emperor: The Legitimation of the Yüan Dynasty. München: BAW Press, 1978, p. 79. Since the work of Edmund Leach and Fredrik Barth, most anthropologists have understood ethnicity as relational and situational, expressed through genealogy and history and a variety of cultural markers within a variety of social contexts.

65. 65 Golden IHTP 75-79, 93-95, Esin 1980: 73-78, 204-205.

66. 66 Esin, 74, 226n152 and 227n155.

67. 67 James Hamilton “Toquz-Oghuz et On-uyghur” JA 1962: 55n26, citing the Wei Shu, which says the territory begins north of the Tanhan (Bogda) Mountain, which is north of Turfan.

68. 68 Chavannes, 1903:89, 94. Hamilton 1962, pp. 23-63.

69. 69 Esin 1980: 76 gives 536 (but Sinor CHEIA 295 gives 546 as the date for Bumin’s attack on the Tiele to forestall an attack the latter planned on Bumin’s allies the Ruanruan), Kliashtornyi, 1985, citing von Gabain “Das Fruhgeschichte der Uiguren, 607-745” in Nachrichten des gesellschaft für Natur- und Völkerkunde Ostasiens Hamburg, 1952, No. 72, pp. 18-32. Hamilton 1962:28

70. 70 Chavannes, 1903: 90, Kliashtornyi 1985: 151. Hamilton 1962:29 suggests that the Syr Tardush and the Uighurs divided north Mongolia peacefully from 630-645.

71. 71 1985:151. Kliashtornyi tries to use Uighur historical narratives from stone inscriptions to establish the reality of this empire. But these Uighur narratives use elements of Türk historical tradition within their own tradition, thus undermining the argument that they had two empires before the one from 744-840. They give Bumin as one of their founding Qaghans (Terkhin inscription E1), and confuse the Orkhon and Selenga rivers, (Shine Usu N2 ‘Selenga’ and Terkhin E3 ‘Orkhon’). While the Chinese sources do indicate that the Uighurs lived at Hanghai (which fits with Terkhin E3, ‘Ötükän’) in the seventh century, they also describe them as subject to the Türk Empire, which Kliashtornyi explains saying that the Tiele were in continual revolt since about 600 to 630. Esin describes the founding of a Qaghanate in Jungaria and Besh Baliq and Qoço 486-540 by Tiele tribes, as mentioned above. The distinctions and ties of Uighurs and Türks remain unclear, a vague object that becomes a project, a projection, clarified into historical discourses for use in the present. In this case, the parallelism of Uighur and Türk generates a logical ambiguousness about whether relations are of similarity or hierarchy or succession or causality.

72. 72 KT N4, describing the revolts that occurred when the empire split into two parts in 716 upon the death of his uncle Qapaghan Qaghan, and “heaven and earth were in disorder,” N3-N4. Kül Tegin was over 31 at this point and he died in 731 at 47. See the NE side of inscription. “My own people” may just mean that he ruled them directly, since there are many mentions of revolts by the Toqquz Oghuz.

73. 73 W. Samolin East Turkestan to the Twelfth Century. The Hague: Mouton, 1964: 72n2; A. von Gabain, “Fruhgeschichte,” 1952: 30-1.

74. 74 Hudud, 286-7. O. Pritsak “Von den Karluk zu den Karachaniden” 270-275. Despite the several migrations of the Qarluq to the west that are recorded in the sources, around 789 some of the Basmil and “Three Qarluq” were still under Uighur control, according to Sima Guang, but then they allied with the Tibetans after the Uighurs took Beshbaliq. Mackerras in CHEIA, 321, citing TCTC 16:7520.

75. 75 Änvär Baytur, Khäyrinisa Sidiq, Shinjañdiki Millätlärniñ tarikhi, Beijing: Millätlär Näshriyati, 1991: 643-652. In their first statement they must be implying “in Turkic history” or “in this history”, since there were other Islamic governments. Haji Nurhaji also points out that this is the first Muslim Turkic Khanate, Qarakhanilarniñ qisqiçä tarikhi, Ürümchi: Shinjang Khälq Näshriyati, 1983: 61-66.

Pritsak also argues that three clans of the Qarluq were the Chigil, Yaghma, and Tukhsi, “Von den Karluk…,” 271, but this only adds to the conflict, since the Uyghur scholars argue that the Yaghma are a Uighur clan, not Qarluq. Peter Golden briefly discusses the historical sources pertaining to this issue, and makes it clear that while evidence points to Yaghma involvement in the Qarluqs, there can be no definitive resolution to the sparse and contradictory data, IHTP, 201, 214-215.

76. 76 Vozrozhdennyi Uigurskii Narod. Second Edition, expanded and corrected. Alma-Ata, 1982, 23-25, 113.

77. 77 Reportedly the Uighurs already used the Türk titles for their own officials by around 647. Jiu Tang Shu 195, 1b. The early Uighur Qaghans used titles and names that were similar to those of the Türks, but eventually extended them to include many more epithets than the Türks seem to have used.

78. 78 Mackerras 1972:86-89. This is despite the fact that Tun Bagha Qaghan was apparently opposed to the plans to attack China that Sogdians had been suggesting.

79. 79 These areas were willing to cooperate with Tibetan conquest due in part to resentment of Uighur taxation. See Beckwith, 1987:153, who uses the writings of Moriyasu Takao to settle the debate presented in Mackerras 1972:162n192.

80. 80 Tugusheva, 1980:4-5

81. 81 For example, Ingrid Warnke “Ein Uigurisches Kolophon aus der Berliner Turfan-Sammlung” in Scholia. Beitrage zur Türkologie und Zentralasienkunde. K. Röhrborn and H. W. Brands, eds. Wiesbaden: Otto Harrasowitz, 1981, pp. 214-220, at 217.

82. 82 “Autour du Manuscrit Staël-Holstein” 115-153, TP t. XLVI, at 142. Von Gabain’s arguement is in Maitrisimit 1961, II, 11. See also Hamilton’s demonstration that Uighur texts often present Turkic proverbs with the phrase Türk savinta bar ‘among Türk proverbs there is…’ (1986: 67-68.)

83. 83 Qadimqi Uyghur yeziqidiki “Maytri Simit”. Facsimile, transcription and translation by Israpil Yüsüp, Abduqeyum Khoja, Dolqun Qämbiri. Ürümchi: Shinjañ Khälq Näshriyati, 1987, pp. 114, 141, and 192 of the transcription. The date is much contested, but I am basing the above on the assessment of Geng Shimin, cited in ibid. 5. See also Zieme 1991: 23, for a discussion of the dating of this work. Cf: Geng and Hamilton, 1981: line III:47, pages 21, 29 and 37, where they find a possible mention of the tradition of the On Uighur in a fourteenth century inscription.

© 1998-2001 Nathan Light

Appendix :

Uyghur Muqam of Xinjiang

Inscribed in 2008 on the Representative List of the Intangible Cultural Heritage of Humanity (originally proclaimed in 2005)


Uyghur Muqam of Xinjiang© ICH Protection and Research Centre, Xinjiang Uyghur Autonomous Region of China

The Xinjiang Uyghur Muqam is the general term for a variety of Muqam practices widespread among the Uyghur communities, which form one of the largest ethnic minorities of the People’s Republic of China. Throughout its history, the Xinjiang region has been marked by a high degree of cultural exchange between East and West, due in particular to its central location along the Silk Road.

Xinjiang Uyghur Muqam includes songs, dances, folk and classical music and is characterized by diversity of content, choreography, musical styles and instruments used. The songs vary in rhyme and meter and are performed solo as well as by groups. The lyrics contain not only folk ballads but also poems written by classical Uyghur masters. Thus, the songs reflect a wide range of styles such as poetry, proverbs, and folk narrative, bearing witness to the history and contemporary life of the Uyghur society. In Muqam ensembles, the lead instruments are made from local materials and vary in form (they may be bowed-stringed, plucked or wind instruments). The dancing skills involve unique steps, rhythms and formations as well as figures such as flower-picking-bymouth, bowl-carrying-on-head and imitation of animals in solo dances. The Xinjiang Uyghur Muqam has developed four main regional styles, namely the Twelve Muqam, Dolan Muqam, Turpan Muqam and Hami Muqam.

Today, community festivities, such as meshrep and bezme in which everybody would participate in the Muqam, are held less frequently. The responsibility for passing on the tradition to new generations of practitioners has fallen on the shoulders of folk artists, but the interest of young people in Muqam is gradually declining. Several Muqam pieces are no longer performed, in particular elements of the “Twelve Muqam”, which in all consists of more than 300 pieces and runs over 20 hours in twelve instrumental and vocal suites.

About RAM Chandrakausika राम च 51

Ram51 is a researcher in the various fields of Musicology, Philosophy and History as well as old languages. One of his first topics is the wide scope of Indo-arabic cultures as represented in various art-forms religion and history. Below a list of selected Research topics which sum up partitionally the task of anthropological Frameworks in totaliter : Sanskrit Hinduism and Mythology Hindustani Music, The Muqhal Empire Gharanas from North India Kashmir Sufiyana The Kashmir Santoor Traditional Folk Music from USA Philosophy in Orient and Okzident Genealogy of musical instruments Ethnomusicology, Arabic Maqams, No Theatre fromJapan, North american poetry, Cultural heritage of mankind and Islamic architecture... View all posts by RAM Chandrakausika राम च 51

You must be logged in to post a comment.

%d bloggers like this: