Mixing and Unmixing Languages: Romani Multilingualism in Kosovo 9780367860578, 9781003016694

Citation preview

Mixing and Unmixing Languages

Mixing and Unmixing Languages uses the politics and practices of language to understand social hierarchies and social change in a post-conflict and post-socialist context. The book focuses on Roma in Prizren, Kosovo, where the author conducted long-term ethnographic fieldwork, using language learning as a central method. Shifts in language practices among this highly multilingual group have reflected the demise of Yugoslav socialism, the rise of ethno-nationalist politics and conflict, and the post-war reversal of power relations in Kosovo. Roma in Prizren nostalgically narrate a past of cosmopolitanism and employment in contrast to the present. Their position today is complex: while they stress their relative integration, this position is fragile in the face of nationalist politics and imported neoliberal economic policies. Within this context, Roma NGO workers have found an economic niche working on projects to protect multiculturalism and minorities, funded by international aid agencies, centred on Romani language. This book discusses the historical trajectory and current configurations of a Romani organisation in the town, the standardisation of Romani and the hierarchical organisation of linguistic forms and language learning, the self-representation of Roma and the ‘gypsy’ image through Romani-language drama, and attitudes to purism, mixing and cosmopolitanism. Mixing and Unmixing Languages is suitable for academics and students in the areas of linguistic anthropology and linguistic ethnography, Romani studies, South-East European studies and sociolinguistics. Amelia Abercrombie has recently worked as a postdoctoral research associate looking at language learning among marginalised migrants in Manchester, UK. She completed a PhD in the department of Social Anthropology at the University of Manchester, UK, in 2017. Before that she was at University College London, School of Slavonic and East European Studies, UK, completing MRes East European Studies in 2011, and BA Serbian and Croatian Studies in 2008. She has also worked as a support worker for mental health and learning disabilities.

Routledge Studies in Language and Identity Series Editor: Reem Bassiouney

The Routledge Studies in Language and Identity (RSLI) series aims to examine the intricate relation between language and identity from different perspectives. The series straddles fields such as sociolinguistics, discourse analysis, applied linguistics, historical linguistics and linguistic anthropology. It aims to study identity and language by utilizing novel methods of analysis as well as ground-breaking theoretical approaches. Titles in Series: Arabic in Israel Language, Identity and Conflict Muhammad Amara Identity and Dialect Performance A Study of Communities and Dialects Edited by Reem Bassiouney Arabic Translation Across Discourses Edited by Said Faiq Language, Identity, and Syrian Political Activism on Social Media Francesco L. Sinatora Mixing and Unmixing Languages Romani Multilingualism in Kosovo Amelia Abercrombie For more titles, please visit www.routledge.com/languages/series/RSLI

Mixing and Unmixing Languages

Romani Multilingualism in Kosovo

Amelia Abercrombie

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 Amelia Abercrombie The right of Amelia Abercrombie to be identified as author of this work has been asserted by her in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record has been requested for this book ISBN: 9780367860578 (hbk) ISBN: 9781003016694 (ebk) Typeset in Times New Roman by Deanta Global Publishing Services, Chennai, India

Contents

Note on orthographies Acknowledgements Introduction PART 1

vi viii 1

Roma, Prizren and language

17

1

Terzimahalla, Durmish Aslano and me

19

2

Durmish Aslano in Prizren: From partisan to NGO

33

3

Intangible culture and tangible employment after socialism

54

PART 2

Purity, mixture and representation

73

4

O Romano Teatro and gypsy theatricality

75

5

Standardisation: Learning linguistics in the bath

95

6

Before the war: A nostalgic speech genre

114

Conclusion: Between two fires? Dissonance, diglossia, disorientation Bibliography Index

133 145 163

Note on orthographies

Throughout this book I have kept examples of writing as close to their originals as possible, and used local orthographies for transcriptions. I use Ro to denote Romani, specifying further if there is a distinction between variants, Shq to denote Albanian, BCS to denote Bosnian/Croatian/Serbian and Tu to denote Turkish. While Cyrillic alphabet is rarely used in Prizren, the Bosnian/Croatian/Serbian Latin orthography, the standard (Tosk-based) Albanian orthography, Turkish orthography and the local Romani orthography are all used. They are often mixed, and words are often spelt according to local pronunciations rather than standard forms. The following table is a list of orthographical differences between the languages. Romani

Albanian Bosnian/Croatian/ Turkish IPA Serbian

*

**

**







/*** / ***

**

**

/ /





/ł/ /ʎ/ /dʒ/ /dʑ/ /tɕ/ /tʃ/ /j/ /ʃ/ /y/ /ə/

*

/ł/ and /ʎ/ are differentiated in Romani speech but not in this orthography. ** These graphemes represent phonemes which are distinct in the standard variants (and as such in orthographies), but not in the local variants. Many people therefore mix them up. *** This sound is present in Romani, but is not included in this orthography. Romani has additional aspirated consonants which are written: , , and .

These orthographies are often mixed spontaneously, which can be especially problematic with names, as people often write their name in several different

Note on orthographies

vii

ways. For example Durmish Aslano (Romani) could also be written Durmish Asllano (Albanian), Durmiš Aslano (Bosnian/Croatian/Serbian) or Durmiş Aslano (Turkish). This is further complicated by the fact that Albanian surnames take a definite article, so change again. For example, the Albanian surname Kryezi would have a definite nominative ending –u making it Kryeziu. While in the Kingdom of Yugoslavia there was a trend to translate Albanian names into Serbo-Croat (Kryeziu, meaning ‘black head’ became Crnoglavić), in Socialist Yugoslavia it was more common to write Albanian names without the article, and with the Serbo-Croat orthography, giving Kruezi. To avoid taking a stance on what is essentially a political debate, I have kept people’s names in the form that they would normally use, regardless of their ethnicity and the origin of their surname.

Acknowledgements

Research for this book was carried out with ESRC funding via the North West Doctoral Training Centre, Language-Based Area Studies pathway. The book was completed as part of the AHRC OWRI (Open World Research Initiative) project. I would like to thank both my PhD supervisors for their constant and very different types of support: Stef Jansen has been a source of inspiration, giving invaluable insights on anthropology, south-eastern European studies, writing and general survival, and Yaron Matras has provided rigorous feedback and inputs, drawing on his depth of knowledge on Romani studies and linguistics, and has included me in other events and opportunities. I am also grateful to my examiners, Tanja Petrović and Soumhya Venkatesan for their thorough feedback and constructive input. I would like to thank everyone who supported me and helped me learn Romani in the field. First, to those who are no longer with us: Nexhip, who orchestrated my stay despite everything, as well as Kujtim and Ibrahim. Inshallah ano xheneti. Also: Fatmira and family for giving me a home; Enver and Drita for their kind welcome; Saverd and Moni for providing a friendly working environment; and Vera for being a true friend when I needed one most as well as an excellent cook. Edis, Erxhan and Denis, thank you for always being friendly and supportive, and for always having time to help me. More generally I would like to thank Roma in Prizren for showing me another way of living. Ov saste savorenge: tumen siklilen mange thay o yaver qhib, thay o yaver jivdipe ano tumaro diz. But kamlipe thay sastipe tumenge. I would also like to thank Saara for always providing a haven in Prishtina and Misha for unfailingly providing a warm welcome in Skopje. Thanks also to the many people who have helped me learn languages over the years, in particular Jelena at SSEES for teaching me Serbian, and Milena and Ivan for making me speak it, Ahmet at the Yunus Emre centre for teaching me Turkish, Hasan and Mirkena for Albanian classes. I am also grateful to Danial Abondolo and Ger Duijzings for encouraging me to undertake a PhD. I have received useful feedback from a variety of seminars and conferences. I would particularly like to thank Gëzim Krasniqi for inviting me to take part in the South-East European Studies seminar series at SSEES, UCL, UK; Kimmo Granqvist for inviting me to take part in the Centre for Baltic and East European Studies seminar series at Södertörn University, Sweden; the organisers of the

Acknowledgements

ix

‘The Politics of Teaching and Learning Languages’ conference at SSEES, UCL, UK; the organisers and participants of the ‘Romani Linguistics’ conference at Södertörn University, Sweden; and Frances Pine for her feedback on the ‘Postsocialism and anthropology: theoretical legacies and European futures’ panel at EASA in Milan, Italy. I would also like to thank Madeleine Reeves for comments and suggestions on my early drafts, and my fellow PhD students at Manchester for academic and moral support. Thanks also to the anonymous reviewers for their fresh perspective and helpful suggestions in making the thesis into a book, and to my colleagues at Multilingual Manchester for their input, but more so for their kindness and humour in this period. I would like to thank my family for awakening and never questioning my interest in this part of the world, and Perry for providing ample distractions while I was completing the book. And last but not least William for his insights, patience and his constant support.

Introduction

Preamble Prizren is an Ottoman-style city built along a river, with a fortress on the hillside, winding streets and Turkish-style houses. There are many mosques, a sufi Tekke, Orthodox churches and a Catholic cathedral. It is also the second-largest urban centre in Kosovo, after the capital Prishtina. According to the OSCE (2015), the population of the Prizren municipality is 177,781, including 82% Albanian, and only 0.1% Serb, with larger minorities of Bosniacs, Turks, Gorans, Roma, Ashkali and Egyptians. Albanian and Serbian are currently the official languages of Kosovo, while Albanian, Bosnian and Turkish are official in the Prizren municipality. There is also a substantial number of international aid workers, and English, German and Italian are spoken as second languages. The ethnic composition of Prizren is a legacy of a long-lasting Ottoman, and more recent Yugoslav, past. In Yugoslav times, Kosovo had varying levels of autonomy, though it was always part of the Republic of Serbia. During the eighties it came under military rule and its autonomy was revoked, leading in the nineties to the establishment of an Albanian parallel state and the Kosovo Liberation Army (KLA), both supported financially by the large Albanian diaspora. In 1999, following a campaign of ethnic cleansing against Albanians, vast numbers fled (to return after three months), and NATO intervened against Yugoslavia. This was followed by the KLA victory, and Serbs fled to Serb enclaves and ‘inner’ Serbia, the part of Serbia outside of Kosovo. From then until the declaration of independence in 2008, it was under the protection of the United Nations Mission in Kosovo (UNMIK). However, Kosovo’s independence is still not recognised by many states (Serbia still claims the territory), and there is a continued foreign civil and military presence. At the 2011 census, the population of Kosovo was estimated at 92.9% Albanian, 1.5% Serb and 5.6% other (ASK 2015). The figure for Serbs is likely to be significantly higher, as many boycotted the census. The ‘other’ ethnic groups include Turks, Gorans, Bosniacs, Croats, Roma, Ashkali and Egyptians. The postwar spatial segregation of ethnic groups persists, and Serbs live almost entirely in the north, and in Serb enclaves. In this respect, Prizren is no different, and very few of the pre-war Serb population of 10,000 remain. Prizren is unusual

2

Introduction

in remaining ethnically, and linguistically, mixed. Within the well-known narratives of Serbs versus Albanians, the other ethnic groups in Kosovo receive less attention, often being labelled as ‘other,’ or as ‘communities.’ In Prizren itself, however, many are keen to stress the cosmopolitan, multilingual composition of the town. It was less damaged by the war than some other parts of Kosovo, though many Albanians fled in 1998–1999, and other groups fled in 1999 and 2004, when Potkalaja (the Serbian quarter) and many of the Serbian churches were set on fire. Roma, suspected of collaborating with Serbs, fled either to Internally Displaced People’s camps or abroad, though many also remained, or have since returned. There has also been an influx of Albanians from deprived rural areas, and parts of Kosovo which were more heavily affected by the conflict. Prizrenites portray their city as cosmopolitan and harmonious, which is symbolised by the continued use of several languages from Ottoman times to the present. The cosmopolitan element of Prizrenite identity allows for long-term Roma residents, present since Ottoman times, to feel integrated into the town, while also belonging to the wider sphere of the Romani nation. However, narratives of multiculturalism and integration are punctured by the everyday experiences of this group in post-conflict, independent Kosovo. The perspective which posits the Roma of Prizren as part of a wider Romani nation, while also a composite nation of Prizren’s cosmopolitanism, sees language as both an objective characteristic and an innate possession of the ethnic group. Here language becomes more than communication: it becomes symbolic, as part of a wider Romani movement. Synchronic distinctions between Roma and non-Roma require the work of activists to embed them in the narrative of Indian origins (Gay y Blasco 2008). In this introduction I elaborate the constellations of ethnicities in Prizren and the wider region, before locating the position of Roma within them.

We speak four languages: malleable identities and porous borders ‘Ja znam ciganski, srpski, šiptarski i turski’ (‘I know gypsy, Serbian, Albanian [derog.] and Turkish’), an elderly Roma man told me in BCS (Bosnian/Croatian/ Serbian).1 People were proud of their ability to speak several languages, and often enumerated them to show how cosmopolitan they were as Prizrenites, and how integrated they were as Roma. ‘Romski, a ne ciganski’ (BCS: ‘Romani, not gypsy’)2, corrected my host, Nexhip, who was showing me around the Romani quarter, Terzimahalla, for the first time. I had contacted Nexhip as he was the director of the NGO Durmish Aslano. I originally tried to contact him in English, as I was unsure about using Serbian in an Albanian area of Kosovo,3 and my Albanian was limited. Nexhip told me that he didn’t know much English, but once he realised I spoke BCS he was relieved that we could communicate, and told me that he spoke romski, bosanski, albanski i turski (BCS: ‘Romani, Bosnian, Albanian and Turkish’). This referred to the

Introduction 3 same four languages as the elderly man above, but with all except Turkish being called something different. The elderly man used the word ciganski, seemingly unaware that many deem it derogatory. Nexhip’s correction alluded to a pride in Romani language and culture, which he wanted to put on a par with the other languages. While the shift away from derogatory terms for Romani has been going on for some time, in Kosovo it has recently become more widespread owing to the growth of westernfunded, multiculturalist NGOs. The shift in allegiances and ethnic identification processes has also played a part in the shift in language names shown above, and in Romani the new labels are often seen as ‘standard’ terms. As such, they form part of a move to standardise Romani in Prizren, as will be discussed in Chapter 5. In Romani, while gajikani, korohani, and gaunani (Ro) are used in informal speech to refer to ‘BCS,’ ‘Turkish’ and ‘Albanian’ respectively, boshnyakuni/ srbyuni, turkyuni and albanyuni are used in formal settings where the standard is required. However, this Romani standard is local to Prizren. Despite various attempts to institute one, there is no centralised Romani standard. Romani cultural organisations have been active in eastern Europe for the past hundred years, and an inter-state national movement has been growing since the seventies. While discrimination against Roma in much of eastern Europe has increased in the postsocialist era, there is also now a wealth of NGOs and other organisations working with Roma. Their funders have sought to valorise and celebrate Romani language and culture, and this move is linked to the use of ‘Romani’ (and its translations in other languages) as a neutral term. In calling his language ciganski, the old man was clearly not up to date with this shift. The old man had also used šiptarski, which is a derogatory term for ‘Albanian.’ Nexhip avoided this, using albanski, the neutral term for the language of the majority nation. The use of šiptar and šiptarski in BCS refer to ‘Albanians’ and ‘Albanian language’ respectively. Despite coming from the Albanian shqiptar, a neutral term for ‘Albanian,’ the BCS usage is offensive, and often associated with Serb nationalist repression and violence. These terms are thus rarely used in Albanian areas, but remain common in Serb areas. This elderly man probably used the offensive term out of habit, not quite aware that such attitudes are no longer acceptable in the area. There is a similar use of parallel terms gaunani and albanyuni to refer to Albanian language in Romani, with albanyuni being the formal, neutral variant. Gaunani etymologically refers to the language spoken in the villages, which would historically have been Albanian. This term may also carry negative connotations of being rural, backward, especially given the lower socio-economic status of many Albanians prior to the war, the ethnic distinction being inseparable from socio-economic ones. The old man, having lived his whole life in Prizren, spent the majority of his life in Serbia. As such he said he spoke ‘Serbian,’ rather than ‘Bosnian’ or ‘Croatian.’ The exodus of Serbs after the war, and the de facto rejection of Serbian language by the current authorities, led Nexhip to use the more acceptable bosanski. In Prizren (and elsewhere), there has been a shift away from calling BCS

4

Introduction

‘Serbian,’ or indeed ‘Serbo-Croat.’ Srpski is still sometimes used in informal contexts, but the vast majority of the time, in public, it is referred to as bosanski or bošnjački. The distinction between bosanski, relating geographically to Bosnia, and bošnjački, relating to the Bosniac ethnic group, is irrelevant in the context of Prizren, and both are taken as referring to the Bosniacs4 in the Prizren municipality. While in many Albanian parts of Kosovo, BCS is not spoken at all, and even considered unacceptable, in Prizren its use is legitimised by relabelling it and detaching it from its link to ethnic Serbs. This usage is paralleled in Albanian. In formal settings, BCS is often referred to as jugosllovenisht, boshnjakisht, kroatisht. Jugosllovenisht (‘Yugoslavian’) acts as a catch-all, avoiding any suggestion about the ethnicity of the speaker (though clearly distinguishing Slavs from Albanians), while boshnjakisht (‘Bosniac’) and kroatisht (‘Croatian’) are more clearly used to avoid accusing someone of being a Serb. As there are no Croats in the area, kroatisht is used solely to avoid linking BCS with Serbs, while boshnjakisht refers to the fact that the vast majority of people speaking BCS in Prizren are ethnic Bosniacs. Before the war, serbisht would have been common, and many people continue to call it this in informal settings, especially when telling people mos fol serbisht (Shq: ‘don’t speak Serbian’). A more offensive term, shka, is also used in Albanian to talk about Serbs and BCS. For me, as a BCS-speaking foreigner, the choice of language was tricky to navigate: I didn’t want to offend people I didn’t know by speaking BCS, but equally my Albanian was limited. I tried to circumvent this issue by saying boshnjakisht (Shq: ‘Bosniac’), but I generally found most people in Prizren were open to talking to me in BCS under any name, and most people preferred this either to me struggling in Albanian, or to them struggling in English. The shift in the use of labels for languages in both Albanian and BCS since the war thus stems from political transformations. Terms that may have been neutral (or at least more widely accepted) before the war have come to be offensive, and mostly used in private, or by older people that history has passed by. Here the names used for languages are symbolic in the sense that they represent ethnic groups, groups which have been subject to change. This seemingly minor difference in the way Nexhip and the old man described their languages thus speaks volumes about the political and social changes in Kosovo since the war. Along with language names, linguistic repertoires have also changed, not just in the aftermath of the war, but throughout living memory. Later in my fieldwork, my Romani-speaking participants told me: ‘I went to school in Serbian. My father in Turkish. And now my children go in Albanian.’5 People also pointed out to me that there were a variety of dialects of Romani in Prizren, as the different Romani areas had different contact languages. Despite initiatives to teach Romani in schools (from the seventies up to the present), Romani has remained primarily a private language, and Roma have studied in the surrounding languages. Language of education relates to shifting state structures, and variation in the imperative to learn a particular language at a particular time, hence these generational shifts. People also regularly contrasted life ‘under the Serbs’ with life ‘under the Albanians,’ paralleling the shift in language use. Here too, large-scale historical transformations are inextricably linked to language politics.

Introduction 5 In other parts of Kosovo that were previously mixed, now only Albanian or BCS are spoken. The ongoing linguistic mixture in Prizren relates to its ethnic make-up and the relative lack of violence. Romani is spoken mainly in Terzimahalla, and among groups of Roma in other parts of town, and borrows heavily from other languages. Albanian is now the closest to a default language, and is commonly used for official purposes. Turkish is spoken in the streets, shops and cafes, and is deemed symbolic of being a true Prizrenite. BCS is less spoken, but can still be heard in the streets. There is also generational variation: young Roma rarely learn BCS, unlike their parents, for whom it was a necessary language for interaction with institutions and Serbs. As we shall see, Turkish was promoted by the state after the Second World War but has since declined, though not so dramatically as BCS. Albanian is now the state language, and many Roma children are brought up as monolingual Albanian speakers. I wanted to work with Roma as they generally spoke the most languages in the town. This was particularly true of the older generations. Within this broader context, my intention in going to Prizren was to focus not on ethnicity and conflict, but on people speaking four languages: Romani, Turkish, BCS (Bosnian/ Croatian/Serbian) and Albanian. I hoped to investigate how languages are mixed and ‘unmixed’ in a situation of societal multilingualism. Duijzings (2003) uses the term ‘unmixing’ to describe the way ethnic groups were reconstituted in postwar Bosnia. My aim was to understand how language boundaries are made and crossed, and how people’s understanding of languages as distinct entities related to ethnic groups affects their speech practices. Ethnicity and language are often seen to be unproblematically linked in the region, and thus language politics is as contested as ethnicity. However, there is a clear discrepancy between the essentialised, ethnicised language ideologies used by politicians, and actual language use. Prizren is linguistically unusual: although it is an Albanian majority area, BCS is still fairly widely spoken. Turkish is widely spoken both by Turks and others, and Romani is spoken by the Roma population. Some of the traditional mahallas (monoethnic neighbourhoods or quarters) have remained, including Terzimahalla, the Romani quarter near the centre. Despite defining my fieldsite with regard to a specific ethnic group, I want to move away from the study of ethnicity, identity formation and ethnic conflict, and to problematise the link between language and ethnicity. Rather than seeing Romani as the language of the Roma, Turkish of the Turks, and so on, I want to see how Roma use and value these languages. This discussion of language names and shifting linguistic repertoires shows that my idea of wanting to research people who speak four languages was more complex than it first appeared. But it also opens avenues of enquiry about the relationship between language and social and political change.

Multiethnic Prizren Most days I walked to the NGO Durmish Aslano, where I volunteered during my fieldwork. It was situated just outside the city centre, and housed Radio Romano Avazo, the Romani-language radio station run by the NGO. I walked through the

6

Introduction

winding streets, passing the Serbian Orthodox church Crkva Bogorodica Ljeviška (Our Lady of Ljevish). This church had been set on fire in 2004, but most of the outer structure has remained.6 It was surrounded by barbed wire, and when I approached it, I could see the blackened remains of the frescoes inside. Opposite, in front of the Ottoman clock tower, stood a police checkpoint. There were occasionally workmen doing something inside, but unlike other destroyed churches in central Prizren, there was very little visible improvement here during my stay. I did however find its name on tourist posters dotted around the city, which proudly displayed the multicultural heritage of Prizren – but these posters did not mention that the churches had been destroyed precisely because they were symbolic of a non-Albanian heritage. Some of the tourist posters were defaced. No one I met used the church’s BCS name, but referred to the church as Xhuma Xhami, meaning ‘Friday Mosque.’ Unusually for Prizren, a town with tens of mosques, there were no mosques nearby. When I asked people how it got its name, they told me that it wasn’t really Serbian, and had previously been a mosque, but they didn’t know the details. There are many stories of mosques turning into churches and vice versa in the region. Indeed, the Sinan Pasha Mosque, the largest in Prizren, is rumoured to have been built with stones from the nearby Monastery of the Holy Archangels. Xhuma Xhami pre-dates the arrival of the Ottomans, and most likely was then turned into a mosque, before being turned back to a church again, explaining where it got its name. This history means that, while it is currently protected as an object of the Serbian Orthodox Church, many claim that it was originally Muslim, or originally Albanian. Some complain that the laws on the protection of Serbian heritage further divide what they see as Kosovo’s heritage into Serb and Albanian, and that Albanians should be allowed to enter protected Serbian Churches.7 Debates over originality and ownership are part of a broader preoccupation in Balkan historiographies with ‘who was there first,’ which assume that proof of ‘being there first’ gives legitimacy to being there now, and hence the exclusion of others as newcomers.8 These narratives gain strength and support not on account of their historical accuracies, but their ability to mobilise in the present (Verdery 1991; Esbenshade 1995; Niedermüller 1998; Brunnbauer 2004). In the case of Xhuma Xhami, an essentialist notion of national groups has been anachronistically grafted onto the past to explain the unusual name of this partially destroyed church. Nowadays, people see Serbs as Orthodox, and Albanians as Muslim. The name Xhuma Xhami is therefore assumed to show that the church is really Muslim, and therefore really Albanian. As a scattered and relatively powerless group, Roma tend to be less involved in claims of being the first in the Balkans, but such claims still exist. For example, Acković (2009) emphasises the early arrival of Roma in Belgrade, even suggesting the term Singidunum (a Celtic name for the settlement that became Belgrade) may come from the word Sinti.9 These conflicting projections of history, used to construct and contest contemporary identities, are far from unusual. Duijzings (1995) discusses the complexity of ethnic affiliations and contestations among Kosovo’s Croat population: Croats

Introduction 7 claimed that Muslim Albanians are really Croat Catholics who were Islamicised and Albanianised in Ottoman times, while Serbs claimed the Croats were Catholic Serbs, and native Croats in the Slavonian villages where the Kosovo Croat fled as refugees referred to them using the derogatory šiptari (BCS: ‘Albanians’) (Duijzings 1995). This shows not only that their identities shift, but also that such shifts can be interpreted as inauthenticity, and used by one group to deny the identity of another (Petrović 2008). This contestation of, and slippage between, ethnic and religious categories is also reflected in the names of languages and ethnic groups. The elderly Roma man cited earlier used different names from Nexhip for all languages except Turkish. Turkish has kept its name in both Albanian (turqisht) and BCS (turski). It is an exception, perhaps because it is the most neutral language in the town, and is less tied to its ethnic ‘owner,’ being spoken as a first language by some Bosniacs, Roma and Albanians, as well as ethnic Turks, and seen as a distinguishing feature of a true Prizrenite. In Romani, there is a distinction between korohani and turkyuni again, but this seems far less politicised than terms for Serbs and Albanians. Korohani refers to the fact that Turks are seen as Muslims.10 The conflation of Turk and Muslim is common in the Balkans, but in Prizren ‘Turk’ is not a negative term as it is, for example, when used by Serbian and Croatian nationalists in Bosnia. The overlap between religious and ethnic identification is a result of the shift from the Ottoman millet system (in which people were organised as religious groups, rather than on a territorial basis) to the western nation-state model (Misha 2002; Duijzings 1999). In this context, the shifting names and distinctions of ethnic groups is part of a wider problem of the imposition of modernist classifications of peoples in the Ottoman region, and more recently political crisis and ethnic war (Duijzings 1995). As ethnic distinctions have solidified after the war, so have ideas about religion. While Albanian nationalists often say ‘their only religion is Albanian,’ there has been an increased investment in mosques and madrasas from the Middle East. For some, this means a shift towards Islamism, and away from Balkan Islamic practices, as will be discussed further in Chapters 3 and 4. These Balkan practices are increasingly seen as ‘pagan,’ and therefore haram, forbidden for religious reasons. On the other hand, some of the new practices being adopted are seen as fundamentalist, and this has been used to arrest purported terrorists. The crackdown on terrorism seems to be a way of pleasing the United States, and also of asserting power at home.11 Prizren has a reputation in Kosovo for being ‘more’ Islamic, which for some means a greater interest in Islam, while for others this is seen as radicalisation. The ease of slipping between ethnic and religious categories, along with the projection of modern categories onto the past in south-east European historiography, explains the complexity of ethnic, religious and linguistic boundaries between groups. Within this regional reconfiguration of ethnicity, shifts in belonging and in the naming of groups extends also to Roma, Ashkali and Egyptians, as I discuss below.

8

Introduction

The formation of Balkan nations can be seen in relation to (and often competition with) one another (Kitromilides 1989). Roma narratives of origin, and in particular the tendency to treat certain groups and labels as fictitious, can be seen as part of wider Balkan ideas of nationhood. If all nations are fluid and constructed (Gellner 1983; Hylland Eriksen 1993), it becomes necessary to ask why, for example, the existence of one nation is treated with far greater scepticism than another. In the past, Macedonian nationhood has been equally contested: Macedonia and its population have been claimed by its Slavic neighbours Serbia and Bulgaria at various points in its history, due to the lack of a clear linguistic border, and did not receive any official status until towards the end of the Second World War (Poulton 2003). This is not just a result of stronger neighbours trying to take land, but rather the mutability of Macedonian identity itself during this period (Yasamee 1995). Today, these lands, though not their inhabitants, are contested by Albanian nationalists, and the name of the country is contested by Greece, which rejects the use of the name ‘Macedonia’ except for with reference to Greek lands. At the time of my fieldwork the official name of Macedonia was FYROM (the Former Yugoslav Republic of Macedonia). It has since been changed to North Macedonia. These terms are rarely used in everyday speech. The centrality of the name debate shows the importance not just of land, but also of symbols (Danforth 1993). Roma, Ashkali, Egyptians In Prizren, Roma from Terzimahalla differentiate themselves from those who live in Jeta e Re (Shq) / Nevo Jivdipe (Ro), ‘New Life.’ Jeta e Re is located along the disused railway, in the area which had formerly been called Dushanovo (BCS), the industrial zone in Yugoslav times. A children’s centre Iniciativa 6, where I volunteered for a time, was then housed in Jeta e Re. The houses there were built informally, after 1999, and have little infrastructure: there is no tarmac, no streetlamps, intermittent water and electricity supplies, and some houses have no piped water or sewerage. They are far more spread out than the houses in the centre, and the land around them is used for scrap metal collection, wood-chopping, chickens, burning rubbish, and so on. Many, but not all, of the inhabitants are monolingual Albanian speakers, having moved from Albanian villages after the war, and most identify as either Roma or Ashkali. Roma from Terzimahalla, unlike the inhabitants of Jeta e Re, claim that they arrived with the Turks, and were given the right to remain without fear of attack as they were blacksmiths and made weapons. There is a ritual surrounding a flag given by the Turks (described in Chapter 3), which ensures everyone is aware of this myth of origins,12 and many people from Terzimahalla, when asked where they came from, will mention a specific town in Turkey. The relationship to the Turks is used to show Roma as an integral part of Prizren. The centrality of Terzimahalla, in contrast to Jeta e Re, means that inhabitants of Terzimahalla have non-Roma friends, and are more able to mix with others in the town. Although

Introduction 9 the town is small, and Jeta e Re is not so far from the centre, this spatial difference indexes a far greater social difference. As such, the distinction ascribed to those from Jeta e Re by those from Terzimahalla is primarily a socio-cultural one, related to levels of education, religion and prejudice against post-war migrants. Roma are often called cigani (BCS) or maxhup (Alb), and these derogatory terms are also used as a catch-all, an external term to describe all groups thought of by others as Roma, as well as sets of behaviour and activities stereotypically associated with these groups. In Kosovo, this includes those who now identify variously as Roma, Ashkali and Egyptian. Those who define themselves as Roma, Ashkali and Egyptian tend to be grouped together by international organisations as ‘RAE,’ three groups which share many of the same problems but are constitutionally recognised as distinct groups. On the whole, the groups who speak Albanian and live in Albanian areas are far more likely to identify as Egyptian or Ashkali, while those who live in Serb areas identify as Roma. Prizren is thus unusual: while many people identify as Ashkali, there are also many who identify as Roma. On the whole, those who call themselves Ashkali are post-war migrants from Albanian villages, and both Ashkali and Egyptians are Albanian-speaking, and live in Albanian areas. However, they inhabit different parts of Kosovo. In Gjakova, there is a large population of Egyptians, while in Prizren, those from the RAE groups who do not identify as Roma identify as Ashkali. The way these social categories have shifted will be discussed further in Chapter 3. The boundaries between the three groups (Roma, Ashkali and Egyptians) remain fluid, and many people shift in the way they identify. Roma often say that Ashkali and Egyptians were Roma before the war but now deny their identity. Given that Ashkali and Egyptians rarely speak Romani, it is likely that they were previously referred to as maxhup (Shq), rather than roma. This is not to imply that they have never considered themselves part of the same group, or are not somehow related to Roma. Rather, it seems plausible that in most cases those in rural, Albanian-speaking areas were never exposed to ideas of Romani nationhood before the war, and as such they were unlikely to have developed national consciousness until the changes that came with the war, by which time identifying as Roma in Albanian areas would have seemed increasingly unappealing.13 Here, the rise in national consciousness was linked to the rise in linguistic consciousness (see also Friedman 2002a). This indeed could be described as an incidence of ‘ethnic unmixing,’ whereby the trauma of war, rather than being caused by ethnic difference, in fact creates ethnic difference (Duijzings 2003). These shifts in identification are also influenced by politics. The emergence of new ethnic categories can be seen as a re-identification encouraged by Milošević in order to reduce the size of the demographic identifying as Albanian (Duijzings 1997), and the distinction between Ashkali and Egyptians is primarily regional, and defined by local political loyalties, rather than by cultural difference (Lichnofsky 2013). Indeed, these new ethnic denominations can be seen as part of a wider ethnicisation of social stratification, which is shaped by regional and local

10

Introduction

loyalties. In addition, international forces’ attempts to ensure minority representation in Kosovo have accelerated the distinction of Ashkali from Roma, Egyptians and Albanians (Marushiakova & Popov 2001). While the difference between these groups is structured in terms of social distinctions and allegiances to other groups, it is legitimised through different myths of origins. The Terzimahalla narrative of Turkish origins forms part of a different discourse to the Indian myth of origins of the Roma. The Indian myth of origins is fully accepted by most that identify as Roma, and has been mythologised by intellectuals in Prizren, who emphasise connections to present-day India.14 Egyptians claim that they are from Egypt, not India, and are therefore distinct from Roma. While some may have identified as Egyptian prior to 1990 (Marushiakova & Popov 2001), the first organisation of Albanophone Muslims, identified by others as gypsies, into a political group occurred in Ohrid in 1990, and shortly afterwards in Prishtina. Trubeta (2005) argues that this is part of a process to deconstruct gypsy stigma, while constructing a modern national identity. Similarly, Ashkali claim to have originated in Persia,15 and therefore also see themselves as distinct from Roma, though those who I know do not deny being related to Roma. Many Roma activists see these new formations as deceptive, and born out of prejudice against Roma. They are problematic for Roma: by reducing the numbers of the Roma ‘community,’ they are reducing their bargaining power. The term ‘community’ in Kosovo is used to refer to an ethnic group, usually a minority ethnic group. In Prizren, ‘the communities’ (zajednice [BCS] / komunitetya [Ro] / komunitetet [Shq] / topluluklar [Tu]) refers to Turks, Bosniacs, and Roma and Ashkali, while at the level of Kosovo, it would also include Serbs, Egyptians and other native minorities, but not of course members of the large international ‘community’ in Kosovo. In Kosovo, ‘communities’ are entitled not just to funds, but also to political representation.16 The Assembly of the Republic of Kosovo has 120 members. 100 are directly elected, while the remaining seats are reserved for official minorities, including four seats for Roma, Ashkali and Egyptians. This means that, regardless of the number of votes they receive, one of the Roma parties will get at least one seat. In one sense, this means there is a vested interest in getting support for Roma parties, while on the other, it is often thought to make parties complacent. Roma activists also complain that the political representatives of Ashkali and Egyptian parties encourage people to identify as Ashkali or Egyptian. The debate about the origins of Roma, Ashkali and Egyptians does not revolve around scientific facts, but is rather based on the varying present-day interests of these groups, reinforced by state and non-state structures. Matras (2013) argues that in addition to being scientifically indisputable, the Indian narrative explains the unusual situation of Roma in Europe, as a scattered nation. The acceptance of the common Indian origins of Roma, then, has the potential to act as a powerful unifying force. In Kosovo, however, the violence, both symbolic and actual, that these groups have faced in recent years are of far greater concern than both scientific proof and pan-Roma unity.

Introduction 11 My research does not concern the origins of, or ethnic distinctions between, RAE. Rather, I am interested in discourses about these distinctions and foundation narratives, and how they are mobilised by my informants in relation to language. Here, local and transnational interests in belonging compete: for some, the unity of the international Romani movement is vital, while for many in Kosovo, their immediate security takes precedence. More generally, origins are contested when people’s presence is threatened. The origins of Serbs tend only to become important when they are contested by Albanian claims to have been in the Balkans before them. For Roma, the need to establish a myth of origins relates primarily to present concerns: their need to find a political unity in the face of threatened existence, coupled with a powerlessness to find this unity in the form of a claim to lands in the region they live. Roma from Terzimahalla told me they are from Turkey when they wished to show allegiance and integration in Prizren, and that they are from India to show they are part of a wider group, and there was no contradiction between the two. This is not to suggest that many of the ancestors of Roma today did not come from what is now India, nor that Romani is not genetically an Indic language. Rather, this is part of a broader ontological debate about originality, related both to historical evidence of origins and to the construction of collective identity through a myth of origins.17 From this perspective the construction of a history and a myth of origins may be, but certainly does not have to be, based on historical or linguistic research, and, as all genealogies are selective, they cannot be completely neutral. The categorisation of communities, and the political manipulation of identities, is not merely a fallacy, but under certain circumstances can affect the way groups behave, feel and interact. In this context, language is not an objective attribute of a pre-existing group, but rather language names, linguistic repertoires and language ideologies both construct, and are constructed by, broader politicaleconomic forces. Identities are always politically constructed, but constructed within certain restrictions, whereby some have a greater ability to construct themselves as authentic than others. I do not wish to imply that the designation of Egyptians, or indeed Macedonians, or any other ethnic group are somehow more constructed or less valid than others on account of their relatively recent genesis (or national ‘revival’). Rather, the construction and political manipulations of forms of identification is clearer among certain groups, and with less powerful groups, this leaves them open to accusations of being made up, or masking their true identity. The war increased the salience of ethnic categories, and solidified boundaries between groups, meaning that discrete national categories became the primary way of categorising people. For Serbs and Albanians, this led to an increased salience of myths of origins as a discourse of legitimisation, a claim of originality. For minority groups this meant less flexibility in terms of identity (they could no longer be just ‘Albanian-speaking gypsies,’ for example), along with an imperative to align with one side or the other. The way origins are mythologised in an ‘attempt to establish a link between past and present by asserting historical continuity and

12

Introduction

objectifying this through the concept of descent’ (Trubeta 2005:85) is as clear for Roma, Ashkali and Egyptians as it is for Serbs and Albanians. Again, this context has led to an increase in the saliency of myths of origins, and the contestation between them. While Ashkali and Egyptians use Persian and Egyptian origins respectively to delimit their groups, and position themselves on the side of Albanians, in opposition to Serbs and ‘gypsies,’ Roma in Albanian areas try to envelop all of those seen externally as ‘gypsies’ into the Romani nation. The contestation, appropriation and malleability of ethnicity in the region is by no means limited to Roma; the malleability of ethnicity, the rigidification of ethnic boundaries and the shift from social and religious groups to ethnic ones have all been written about extensively with regards to the Balkans, and beyond (cf. Erikson 1995; Daniel 1996; Sundar 2004). Violent wars and changing state borders in the nineties meant growing interest in these issues in the region (Campbell 1998; Gow 1997; Kent 1997; Stokes et al. 1996; Naimark & Case 2003). Within this focus on ethnicity, language tends to be treated as an objective characteristic of the ethnic group,18 and something which, unlike other parts of culture, is beyond human control and manipulation. I intend to show that this is not the case: languages, while being symbolic of ethnic groups, are far from objective, and are open to pressures and manipulations.

Mixture and purity in linguistics and anthropology Work on multilingualism in the fields of both sociolinguistics and language contact focuses on the content and structure of language. There is extensive research on the broader effects of language contact on language structures (e.g. Matras 2009; Friedman 1984, 1985b, 2002b). More specifically, theories of codeswitching focus on the details of speech practices, examining how and when people with a multilingual repertoire alternate between languages, as well as the social function and uses of mixing languages (Heller 1988; Myers-Scotton 1988). There is also research on the way social structures and ideologies affect speech practices in the short term, and language structure in the long term (Friedman 2001). These literatures inform my understanding of the social functions of language, and attitudes towards language. However, I depart from these approaches in that I use language to explore deeper social structures. I build on research on language ideologies (Gal 2005; Kroskrity 2000; Irvine & Gal 2000), in particular work on the ideological forces behind language standardisation and ideologies which support (and impose) monolingualism (Jaffe 1999; Milroy 2001; Silverstein 1996; Blackledge & Creese 2010; Woolard 1998). While this field of research broadens the scope of knowledge about how people understand and talk about language, I wish to go further, by asking what the dissonances between ideologies and practices can divulge about the social relations of speakers. A focus on these dissonances, and the political economy in which they are embedded, reveals the role of language and languages within social relations. Some studies of language socialisation combine the former approach, focusing on the link between society and language, and the latter focusing on language

Introduction 13 ideology, by looking at how language ideology and speech practices are acquired. The acquisition of a multilingual repertoire is dependent on the ideology of, and attitude towards, the relevant languages. People don’t just learn to speak; they also learn communicative competence (Hymes 1972) and speech genres (Bakhtin 1986): they learn that certain speech practices are more appropriate than others in certain contexts, and these contexts are understood hierarchically, with nesting distinctions of public and private. In this context, specific speech practices come to be identified as inferior, and this inferiority is then legitimised within broader transformations in social relations. If the mixture of languages is metonymic of the mixture of cultures, it is also necessary to look at the way the mixture of cultures is ideologically constructed. While linguistics literature looks at the way mixture is enacted through speech practices, anthropological work on mixture has focused on the relationship between ideological understandings of purity and mixture and social structures. Literature on Balkanism discusses how the Balkans are represented as too mixed, in contrast to the ‘modern’ and categorisable nations of the west (Todorova 1994, 1997; Fleming 2000; Green 2005; Petrović 2008). Here mixture is popularly understood as negative, messy and uncategorised, in contrast to purity. Discourses of cosmopolitanism, on the other hand, posit mixture as a positive feature, in contrast to monocultural backwardness. Cosmopolitanism has been theorised as an ideological stance which is overtly inclusive, while covertly excluding certain groups, namely those of a lower social class (Jansen 2005; Stefansson 2007; Spasić 2011; Grant 2010). As such this cosmopolitan ideological perspective may celebrate mixture, in contrast to purity, but also certain forms of mixture compared with others (Ballinger 2004; Calhoun 2002). I build on this literature to show how ideologies of multilingualism are embedded in power structures, and how multilingualism can be represented both positively and negatively according to who controls the representation. Recent political shifts in Kosovo have led to a purist language ideology being imported into Romani, which has resulted in the development of an unequal diglossia (see Chapter 5). This purism sits alongside a nostalgic cosmopolitanism, related to the use of a mixed variant of Turkish, as I explore in Chapter 6. Here, the celebration of (lost) cosmopolitanism and the imposition of linguistic purism are not dichotomous: rather, speech practices are understood according to categories of purity and mixture, and subsequently celebrated or denigrated. The hierarchical organisation of speech practices is connected to the hierarchical organisation of social groups, and shifts in the organisation of these groups affect ideological understandings of speech practices, and speech practices themselves. Speech practices have a social life; they shift in relation to ideologies and power relations, and their meanings vary according to context. My contribution to these fields of research is informed by the work of early Soviet theorists Bakhtin and Voloshinov. Bakhtin introduces speech genres as follows: ‘(e)ach separate utterance is individual, of course, but each sphere in which language is used develops its own relatively stable types of these utterances’ (Bakhtin 1986:60). I analyse the way speech genres are formed, and changed by a particular group.

14

Introduction

Voloshinov argues that ‘(i)t should be one of the tasks of the study of ideology to trace [the] social life of the verbal sign’ (Voloshinov 1973:27). By taking not just the sign, but also its historical trajectory, we can see how meaning is formed within social relations. The multiaccentuality of the sign, a sign’s potential to have multiple meanings, and the space for conflict within this multiplicity, can be used to elicit ‘a little arena for the clash and criss-crossing of differently orientated social actions’ (Voloshinov 1973:41). These theories are crucial, not only because they incorporate diachronic and synchronic aspects of language – the historical trajectory of signs – but also because they examine these trajectories in terms of structural inequalities. In order to understand language, we need more than an understanding of the social context or background: the ethnographic study of language can open up a wealth of ways to study social life itself. By combining and incorporating approaches from linguistics and anthropology, especially theories about the effect of ideology on speech practices and ideological interpretations of speech practices, I hope to further knowledge about how language is used within unequal power structures.

Book layout This book argues that language can be used to study social relations. Rather than looking at how the study of language enriches the study of society, or how a knowledge of society informs the study of language, I look at how society can be studied through language. Language and society are not just contexts of one another, but are in fact mutually constitutive. As such changes in speech practices are closely connected to changes in social relations. This leads me to ask how social hierarchies are articulated through linguistic ideology and speech practices at different scales. In the first part of the book I focus on the relationship between changes in linguistic repertoires and political and economic changes in Kosovo since the founding of socialist Yugoslavia, relating this to contemporary social relations. Taking Roma in Prizren as my starting point, I examine how broader political forces affect smaller scale social relations and language politics. Chapter 1 describes how I set about investigating these issues, gives details of my fieldwork and explains how I used language learning as my main method. Taking the idea of the ethnographer as a novice or learner, I look at how the processes of language acquisition and language socialisation can be used by the ethnographer to probe further into the hierarchies of speech practices and language ideologies. Chapter 2 asks how language politics has shaped Roma and Durmish Aslano, the Romani organisation I volunteered with in Prizren, diachronically. The chapter discusses the conditions in which the organisation came to be formed, and how it has been reconstituted over time, locating this trajectory within the literature on Yugoslav and Kosovo history, language policies and theories of memory and ethnicity. Chapter 3 investigates the contemporary role of Durmish Aslano, focusing on a project to protect Romani intangible cultural heritage. This chapter discusses how language and

Introduction 15 policies are translated and transformed as they flow from western funders to eastern projects, and how this relates to the politics of recognition and redistribution. The second part of the book focuses on how language ideologies and discourses of purity and hybridity relate to speech practices. Given the place of Roma amidst external forces, and the social inequality discussed in the first part of the book, this part looks at how Roma represent themselves both linguistically and extralinguistically, and how this reflects and reproduces a Balkanist worldview. Roma are widely perceived as in-between and out-of-place, and below other groups. When this perception is internalised, it is both reproduced and changed. Chapter 4 looks at the Romani dramas performed in Prizren. Paradoxically, the dramas which aim to valorise Romani identity also perform a negative image of Roma. By examining the representation of Roma and the ‘gypsy’ image in the dramas, and exploring how this shifting representation of self and other is enacted off stage, we can see how this representation forms part of a broader Balkanist schema of representation, performed dramatically through the use of the tragicomic form. Chapter 5 moves on from Roma in representation to Romani speech practices, looking at how certain practices are perceived to be better than others, how the distinction between good and bad speech practices is produced and how a Romani standard is being produced and legitimised in Prizren. As such it investigates the symbiotic relationship between ideology and practices. Chapter 6 focuses on another paradox: in parallel to the process of standardisation discussed in Chapter 5, there is also extensive positive valence afforded to the mixture of languages. Purism, ridding Romani of alien words, is seen as crucial, but at the same mixture, in the form of cosmopolitanism and multilingualism, is valorised. People acknowledge the local variant of Turkish as mixed, and embrace this mixture, within the confines of the speech genre of nostalgia. These speech practices used to be neutral within the town, and excluded those from outside, but since the war and major shifts in power relations, they have come to signify nostalgia for this pre-war sedentarist urban cosmopolitanism. The concluding chapter draws together the discussion of language and change in the first part, and the relationship between ideology and speech practices in the second to question the way Roma are positioned and perceived as being ‘in-between,’ not quite fitting into any category.

Notes 1 The language (or group of variants) formerly called Serbo-Croat is now known as Bosnian, Croatian, Serbian or Montenegrin, depending primarily on the ethnicity of the speaker rather than any particular linguistic features. 2 I use ‘gypsy’ as the closest translation for cigan, and similarly derogatory terms in other languages. In the Balkans, it does not have the same implication of a travelling lifestyle that it does in the UK, but rather refers to a negative image of the ethnic group. In addition to using it as a translation, throughout the book I will use the term gypsy, not capitalised, to indicate a negative image of the group, in contrast to Roma, a neutral term for the ethnic group. 3 Where possible I use English versions of place names to avoid choosing either Albanian or Serbian versions.

16

Introduction

4 Bosniac (bošnjak) is a relatively recent term for Muslims Slavs. It replaced the ethnonational designation Bosnians Muslim in Bosnia, and has spread throughout the region. 5 This is true for those who attended school: many, though by no means all, of the men I spoke to had attended school and were literate, but this was rarely the case for women. In the pre-war era, boys were sometimes kept back from secondary school in order to learn to be musicians. Most Roma women of this generation were kept at home to help with housework, childcare and care of the elderly. The practice of keeping girls at home now seems less common in Terzimahalla, but continues in other mahallas. While firm statistics are elusive, my impression is that this is similar to the situation among Albanian girls. 6 Since some churches were built in Kosovo during the Milošević era as a sign of Serbian hegemony (for example in Prishtina), this was used to legitimise retaliations against all Serbian churches (new or old) after the war. While more accept the presence of older churches, there is an ongoing debate over whether new churches, built in the Milošević era, should be allowed to remain (see Kraja 2016). 7 The organisation Ec ma nryshe is one of the main proponents of ending this protection of Serb monuments, suggesting it would be better to focus instead on preventing the destruction of other buildings which makes way for new developments. (See for example Ec ma ndryshe 2015.) 8 This is not, of course, unique to the Balkans, but rather takes on a particular form in the Balkans. 9 A group of Roma living primarily in Germany. 10 Matras (2014) relates the etymology of Koroxaj for ‘Turkey’ to Romani contact with the Karakhanid dynasty in the tenth and eleventh centuries, suggesting that this was the first time they encountered Muslims, and they subsequently labelled all Muslims with derivations of Karakhanid. 11 See for example an article in The Guardian by the then Prime Minister and former KLA leader Thaçi (2014), where he refutes the idea that Kosovo breeds jihadis and promises to repress all jihadi movements. 12 I use myth to refer to the way historical events, factual or fictional, are constructed and narrated as something of importance to a particular group of people, rather than in the lay sense of a story which is fictional or untrue (Kolstø 2005; Duijzings 1999; Duijzings 2005). 13 For a comparative view of Vlachs, a group who, despite certain distinctive features, never mobilised into a national movement see Winnifrith (1992). 14 This takes the form of women and girls wearing saris for Romani events, using the imagery of the Taj Mahal in publications and drawing comparisons between Indian soap operas and their own lives. 15 Some also claim Roman and Palestinian origins (Marushiakova & Popov 2001). 16 They also have their own star on the flag of Kosovo. 17 See Clifford (1988) for a discussion of this predicament with regard to Mashpee Native Americans. 18 Discussions over the division of Serbo-Croat are an exception in that this division is widely debated. However treating the distinction between some languages as natural and others as constructed in fact reinforces the idea that languages can be objective defined (cf. Hodges 2016; Greenberg 2004; Kordić 2010).

Part 1

Roma, Prizren and language

1

Terzimahalla, Durmish Aslano and me

Introduction I chose Terzimahalla as my original fieldsite on account of the linguistic repertoire of its inhabitants. Not only are several languages spoken in the town, but I was also told there are different dialects of Romani: the dialect spoken in Terzimahalla is influenced primarily by Serbian and Turkish, while other mahallas have more Albanian influences, and in some cases other differences too. I spent the summer of 2013 in Prishtina, hoping to learn some Albanian, before starting my fieldwork in September 2013. I was based in Prizren from then until March 2015. The majority of my participant observation of language took place in Terzimahalla, and at the NGO Durmish Aslano. I had planned to stay in Terzimahalla for the whole period, but this did not work out. When I had originally explained that I would like to live in a Romani area, such as Terzimahalla, Nexhip reacted positively and said he would help me find a žensko mesto (BCS: ‘a female place’). In my next visit he explained that I could either stay with a family with two unmarried daughters, but that they had no bathroom, or in an annex to his house where his cousin used to live. Thinking I would have my own space, and preferring the use of a bathroom, I opted for the latter. I only stayed there a short time, however. I was not aware at the time that an unmarried woman having her own space was looked upon with great suspicion in the mahalla. My lack of a clear position in a family structure was problematic, and I was not allowed to sleep in a room alone, but was always accompanied by one of Nexhip’s daughters. Subsequently, I moved three times. First, I was moved next door into Nexhip’s family home. In addition to a downstairs kitchen and shower room, he had three rooms upstairs, and I shared one of them with his daughters. Along with his two teenage daughters, he had two teenage sons. His father also lived with them, as did two brothers with learning disabilities. His wife had a daughter from another marriage (this was her third marriage, a fact which was often commented on) who was married and lived with her in-laws. Nexhip had an elder brother who lived abroad, and a half-sister, also abroad, who he had only found out about in adulthood. The members of the household spoke to each other in Romani, and to me in both Romani and BCS, though the children, born after the war, did not speak

20

Roma, Prizren and language

BCS. All the children were attending school in Albanian except the eldest daughter, who had just finished school; Nexhip was hoping she would study pedagogy at the university. Turkish soap operas played constantly in the background. Then I was moved into his sister Fatmira’s house; despite my sharing a bedroom with girls, people continued to gossip about my place in the household. Fatmira’s house was two rooms just outside the mahalla, where she lived with her husband, Taylor, and two teenage children. She worked cleaning the hospital, while he had no permanent employment. They spoke Romani at home, while many of their neighbours, also Roma or Ashkali, spoke only Albanian. In their street there were far more infrastructural problems, and most of the time we were without running water. During my stay at Fatmira’s, the family were very careful to take care of me, which from my perspective often felt like I was restricted in my movements, and had little time to myself. People were not just suspicious of me wanting to be alone, but also expressed surprise that I had wanted to sleep in a room alone, saying they would be scared to do so. What seemed to me as an oppressive hierarchical family structure was far more commonly viewed as protective, especially for women. While I was living with Fatmira, I would often speak to her in BCS. Her daughter, who didn’t know (or want to know) the language, would tell her ma vakar gajikani (Ro: ‘don’t speak gajo language’). At first I assumed that she wanted us to speak in Romani, out of pride in the language, and so she could understand. Later I realised she wanted to avoid BCS. While in many other variants of Romani, gaje refers to all non-Roma, in Prizren it refers to Serbs. Gajikani then refers to BCS, also called either srbyuni or boshnyakuni. The younger generation were particularly concerned with distancing themselves from Serbs, in part due to the schools which displayed Albanian nationalist pictures. She therefore wanted to correct her mother, and prevent her from speaking BCS. Similarly, Nexhip’s son Fatmir would often tell us not to speak in BCS, so people didn’t think he was a gajo, a Serb. While it would have been obvious from his appearance that Nexhip was not Serbian himself, he was concerned with appearing pro-Serb. I was at Fatmira’s until Christmas 2013, when I returned home briefly. As the same rumours and suspicions continued to cause problems for myself and my hosts, when I came back, I moved to private accommodation. This was the upper floor of the house of an Albanian family, away from the mahalla, and I had this space to myself. This was more comfortable, and made life easier for my key informants, but was a huge disadvantage in terms of my language immersion. I thus shifted my focus away from language use in Terzimahalla towards more institutional forms of language use, in particular the Romani used at the NGO Durmish Aslano and the other NGO I volunteered with briefly (Iniciativa 6), as well as the Romani language radio and Romani school classes. In Autumn 2014 a substantial number of Terzimahalla Roma, as well as other groups, began to leave Kosovo to seek asylum in the west. This was due to a combination of their ongoing frustrations with the political and economic situation in Kosovo, and a cheap smuggling route which opened up through the Balkans as a result of the war in Syria. To my knowledge they were not successful in seeking asylum and

Terzimahalla, Durmish Aslano and me 21 many were returned in the spring. I completed my fieldwork in March 2015 and therefore did not witness their return or hear about their experiences. There was a variety of ethical problems during my fieldwork, the specifics of which I have decided not to write about in order to protect the privacy of those involved. While it would be unlikely for those who participated in my fieldwork to ever read my work, I have tried to write it as if they would read it. This does not mean that I have written it in order to please them – my attempts to deconstruct social hierarchies would not be popular if participants were to read it, and my views on gender and sexuality diverge so much from those of most of my participants that it would not be possible to write about these issues with integrity without displeasing them. However, I have made a conscious effort not to divulge anything which was told to me in secret, or anything which may cause shame or embarrassment. As will be discussed in Chapter 4, people were keen for me not to represent them in terms of the negative stereotype of ‘gypsies.’ I have tried to comply with this wish, while at the same time not avoiding or masking certain contentious issues. As such I have tried to avoid either wholly positive or wholly negative representations of the group, without divulging information that may be damaging to individuals.

NGO Durmish Aslano During my preliminary visits, I had been in contact with Durmish Aslano, the Romani NGO, and met the staff of their radio station. I had originally planned to volunteer there in exchange for access to Terzimahalla, and also so that I would have a role in my stay there, something to do and talk about on a day-to-day basis. Over time, the NGO became the central site of my fieldwork. When I first arrived, Nexhip was excited at the idea of me learning Romani, and was also glad to have someone who spoke both a local language (BCS) and English to help with project applications. All the staff spoke Romani, which was the main language used for interactions, both on and off air. The staff were all keen for me to learn Romani, but spoke to me in BCS when they wanted to clarify something, or just to chat. When non-Roma came to visit, they generally spoke to them in Albanian, unless they were Bosniac, in which case they would use BCS. If there were visits from internationals (such as people from other NGOs or KFOR troops), they brought their own Albanian translators. For much of this time, I worked from Monday to Saturday as a volunteer at the NGO and radio. I continued to volunteer at the NGO until December 2014, but began working three days a week in the spring. At this point I also started to volunteer at Iniciativa 6, based in Jeta e Re. The director was Osman Osmani. I stopped volunteering for Iniciativa after a few months, as they had some management and legal problems – I was told they had previously been raided by the police due to an accusation of holding weapons in their house, and later, the owner of the house, now living abroad, decided he did not want them to use it any more. This was followed by a complicated dispute over rights of the board with regards to the NGO and its funds.

22

Roma, Prizren and language

Durmish Aslano is a small NGO. While I was there, aside from Nexhip, the director, Saverd was the technician, and Moni and Vera worked as journalists. Their hours and wages varied greatly, as will be discussed in Chapter 3. Saverd, like Nexhip, was from Terzimahalla. He was the only one among his seven siblings that had stayed in the mahalla. His sisters had married, and most of his brothers had migrated, apart from one who had died, and the eldest brother Hysni, who had moved to another area but visited both the mahalla and the NGO fairly regularly. Saverd lived with his father, wife, three daughters and one young son. Like most members of the mahalla born before about 1990, he grew up speaking Romani, and also spoke BCS, Turkish and Albanian fluently. Moni was not from the mahalla and had grown up speaking Albanian. He had learnt Romani and BCS as an adult. He lived with his wife and two (later three) young children in Ortakoll, one of the ethnically mixed areas of town. He had close connections to Terzimahalla, and seemed to know all the gossip from the mahalla before its inhabitants. Vera also lived in a mixed Romani and Albanian area outside Terzimahalla, but had grown up speaking Romani in a mixed Romani and Serbian area. Like most women, she now lived with her husband’s family, some of whom spoke Romani, while the younger family members generally only spoke Albanian. She had an adult son who worked with her husband at a supermarket, and two teenage daughters. She was one of the few Roma women of her generation to receive an education, which was in BCS (then called Serbo-Croat or simply Serbian), and she often spoke of the Serbian friends she had had at school. There were several people who did not have formal roles at the NGO, but stopped by when they felt like it, a kind of secondary staff. These included Gjengiz, a young man who, despite speaking Romani, often called himself Ashkali rather than Roma. He had left Terzimahalla, as his wife did not get on with his family. Young women often struggle with the strict hierarchy of their in-laws’ homes, as they are expected to be quiet and subservient. In most cases, their only option is to put up with this, or to leave their husband and return to their own family. In the case of Gjengiz’s wife, however, her father had the means to build them a house elsewhere. She was the daughter of an Ashkali politician, hence Gjengiz identifying as Ashkali, and they spoke only Albanian at home. She was also a nurse, one of the few RAE (especially RAE women) to qualify as a medical professional. Bajram Galushi (also Galjuš, Galjuši), a cousin of Vera’s, was a middle-aged man who had grown up in the mahalla but moved away due to lack of space in the family home. He (and his family) were relatively wealthy. He had previously been more involved in the NGO, but was having various health problems during my stay so was not always present. He had several sons, and they spoke Turkish at home. His youngest son, Ertan, also helped out, especially towards the end of my stay when Saverd had left. Three of his nephews, Erxhan, Edis and Denis Galushi also worked on various projects at various points and were generally involved in Romani activism. The Galushi family were well-respected, and many of them were well-educated and had good jobs. Like Nexhip, Kujtim Paqaku was considered a local Roma intellectual. He was an in-law of Vera’s who was involved in the NGO only sporadically, but

Terzimahalla, Durmish Aslano and me 23 was often talked about, and, as such, features heavily in this book. He was from another Roma mahalla in Prizren, and spoke a different dialect of Romani, but he had been active in the international Romani movement and had spent some time at INALCO (Institut national des langues et civilisations orientales, ‘National Institute for Oriental Languages and Civilizations’) in Paris. He wrote poetry in Romani. As such he had earned great respect and admiration, and I was repeatedly referred to him when I told people I was learning Romani. During my stay he also entered politics, and was made a Deputy in the Assembly. There were two other Roma men who were locally considered to be Roma intellectuals. Ibrahim Elshani (known as ‘The Professor’) had translated works such as The Iliad and some Indian epics into Romani, had written a variety of other books and was editor of the local Romani magazine Yekhipe, ‘Unity’ (cf. Halwachs 2011). The connection to India was clearly important to him. He was very well-respected, and there were many copies of his books around, particularly at the radio station. By the time of my stay, he was elderly and unwell, so my main knowledge of him is second-hand. His daughter Indira lived with him, and was the only Roma person working for Caritas, another NGO working with Roma in the area. She did not speak Romani, and they did not live in a Romani mahalla. Selahadin Kruezi (also Kryezi, Kryeziu), a Romani author/journalist from Prizren, who had lived in Denmark for a long period, was well known locally for some Romani children’s books he had written, as well as some other articles and pamphlets on Romani orthography and the Romani standard language. His most recent (self-published) work was a large book about Roma and Romani language in Romani and English (Kruezi 2014), in which he claimed Roma are really from Egypt, not India, which was met with outrage by many Roma in Prizren. Women were less involved in the NGO because it was deemed a public space, and thus less suitable for women, as we will see below. Aside from me and Vera, the two women who visited regularly (and unaccompanied) were Fatka and Nehalla – both unusual within Romani society. Fatka was in her late thirties, and divorced, while Nehalla was unmarried and in her late twenties. I had met both in a café with Nexhip and other men on one of my preliminary visits to Prizren, an encounter which had led me to believe that gender relations were less strict than they in fact were. Women in the public space are generally highly unrepresentative, and most women I knew were far more restricted in their movements. Other people involved in the organisation would come and go, depending on whether there was work, or whether they needed something. In this sense, the radio station also had a social function – people would drop in, have a coffee, smoke, sometimes sing and dance, at other times fill out job applications, scan documents, use the internet and do a range of other things they did not have access to at home. Most of these people who used this space were in some way connected through family to the staff, and each other; accordingly, although there were arguments, most people respected age and gender hierarchies. On meeting me there, people didn’t ask who I was, but rather kaski siyan? (Ro: ‘whose are you?’). It was not clear how I fitted in.

24

Roma, Prizren and language

My role in this hierarchy was thus ambiguous. For a foreigner from a western European country to come to the mahalla was flattering and interesting, but for an unaccompanied woman to do so was dangerous and threatening. People’s attitudes to what they called e Bare Evropake Rashtre (Ro: ‘the Big European Countries’) was a mixture of admiration and fear: admiration for the perceived wealth and modernity of these countries, evidenced by the large diaspora who returned with cars, electrical goods and other desirables; fear for what was seen as moral corruption, and lack of family values. Nexhip on one occasion told me that people in the west don’t need their family, because they have money. It was unclear to me whether he saw this financial independence as positive or not. Families were strictly patriarchal, and the workplace hierarchy at the NGO replicated, to some extent, the family structure. Vera sometimes referred to me as i terni (Ro: ‘daughter-in-law’ or ‘young woman’), especially when asking me to make coffee or wash-up, and when my partner visited, they referred to him as o jamutro (Ro: ‘son-in-law’), and thanked him for allowing me to work there. My role as youngest female meant I was expected to follow orders, and to make and serve coffee. I found the level of subservience expected of me while working in the NGO extremely difficult at times, in contrast to the freedoms afforded women in British middle-class life. During my fieldwork, I consoled myself with the knowledge of the limited timeframe of my research, but this left me feeling extremely uneasy with regards to my female friends, colleagues and other informants. I could leave, and the knowledge of my own agency with regards to being there made the experience more bearable, but the women there did not have this. However, I did not see contestation of these hierarchies; while it would be easy to attribute most women’s acceptance of the situation simply to false consciousness, I would suggest that there is a more complex structural issue at stake here. In Prizren, I was very aware of getting unwanted attention from Albanian men, but not from Roma men. I did not understand the reasoning behind this difference, and simply tried not to see it as some kind of essential difference, until an incident at the radio. A young Roma man who wasn’t involved with the NGO and didn’t speak Romani came to get a lift with Kujtim, Nexhip and Osman. As I left the building, he was asking Kujtim how to say something in Romani, then turned to me and said but laqi siyan (Ro: ‘you are very pretty’). The situation was hardly threatening, but, having overheard, Nexhip (semi-jokingly) slapped the back of his head and pushed him away. This incident was playful rather than serious, but even so, it was very clear that there were certain boundaries which were not meant to be crossed, and that for me, Nexhip was responsible for policing these boundaries. I was a kind of surrogate daughter-in-law to the NGO ‘family.’ Here, by describing the NGO as a family, I do not wish to romanticise it. Rather, I am referring to it as a strict hierarchical structure: I was obliged to follow orders and to serve people, but this also offered me a level of protection that was imperceptible for much of my time there. As director of the NGO, Nexhip acted as a kind of patriarch to the staff: he held almost absolute control, but this also entailed a level of responsibility, albeit one that was limited to the time and space of work. As such, the acceptance of patriarchal structures by women is better described

Terzimahalla, Durmish Aslano and me 25 in terms of an imperceptible structure of protection and dependency than simply false consciousness. I had not realised that my subordinacy to these structures was in exchange for this protection until someone questioned it.

Language learning as method Whenever I attempted to explain what I was doing in Kosovo, that I was an anthropologist, interested in multilingualism and in Romani language and culture, people took from it that I had come to learn Romani. Many people weren’t familiar with ‘anthropology’ and ‘multilingualism,’ but they were impressed that I was interested in Romani language. Concerned that they might be giving consent to participate on the basis of a misunderstanding, I tried to explain multilingualism by saying that I wasn’t just interested in Romani language, but other languages they spoke too. But people were dismissive of me wanting to learn languages other than ‘their’ language. I tried to explain anthropology by saying I was also interested in Romani culture, at which point I was normally referred to Kujtim Paqaku, as a poet and someone who is thought of as knowing about Romani culture, in the sense of music, folklore and poetry. I learnt quite quickly that most people weren’t that interested in the intricacies of my research, and that it was easier to simply say I was there to learn Romani. If they had time, and were interested, then I explained that it was rather more complicated than that. It was tricky to explain to people that my aim, in learning Romani, wasn’t to write a linguistic account of Romani, but rather an ethnography of language use. While some people knew what linguistics was, they had not heard of ethnography or anthropology.1 Simply saying I was learning Romani not only had the advantage of being comprehensible, but also seemed to evoke the feeling that I respected Romani in relation to other languages, and Terzimahalla Romani in relation to other variants, which made people willing to help me. But it also meant people wanted to teach me to speak in a certain way. The way people taught me to speak was often different from the way they spoke, and as such the learning process became my central research method. I participated in and observed speech practices, as well as metalinguistic discussions. In addition to language learning, I used participant observation more broadly. Spending time with people in Prizren, and especially at the radio station, gave me a wealth of data about implicit social structures, and sociality. I also spent a considerable amount of time writing project proposals, which gave me an insight into the workings of international funding, in particular minority and multiculturalist funding, in Kosovo. In addition, I conducted interviews towards the end of my stay. Having pinpointed issues of interest for my project, I used these interviews to elicit responses about specific issues. These interviews were conducted mostly in Romani, and for some of them I enlisted a research assistant (Erxhan) to help with translations and organising the interviews. I interviewed the staff of the radio station on more personal questions, such as their involvement with Romani language and Romani activism; I also conducted a group interview about Romani in the

26

Roma, Prizren and language

media. I interviewed the staff of the Romani language section of Radio Television Kosovo (RTK) about their language use, as well as intellectuals, and others involved in language standardisation. Finally, I interviewed an Albanian official from the Ministry of Education about minority languages in education. Using language learning as participant observation was successful, though not quite in the way that I had anticipated. Before starting my fieldwork, influenced by anti-essentialist theories of language and multilingualism (Blommaert 2003; Heller 2003), I had hoped that by living in Terzimahalla and interacting with people on a daily basis I could both observe and participate in speech, and learn how people talk and mix. Theories of translingualism have questioned the idea of languages as discrete entities, focusing instead on theorising language mixture as the norm, rather than a transgression (Canagarajah 2013; García & Li Wei 2013). With this in mind, I had hoped that by entering the field without an expectation of distinct bounded languages I would be able to learn to mix languages in the way my participants did. I was aware that I would not learn through formal learning, such as language classes, but rather through a more informal, or ‘natural,’ process. My aim was to use participant observation to watch and listen to the way people interact, and crucially, to participate in these interactions, and learn what things can and can’t be spoken in what ways. In this sense, I wanted to learn language in the way that children learn language – that is, by trying certain combinations in certain contexts, and by making mistakes. Ethnographic methods have been compared to children’s learning (Burnett 1974) and as such the ethnographic process can be seen as a process of acquiring culture, and similarly of acquiring language. Both the child and the ethnographer undergo a process of socialisation in order to function in a society, learning the rules of language and culture.2 Learners do not just absorb rules, but play an active role in their acquisition. With regards to language socialisation, linguistic anthropologists Schieffelin and Ochs (1986:165) argue that: socialization is an interactive process […] the child or novice (in the case of older individuals) is not a passive recipient of socio-cultural knowledge but rather an active contributor to the meaning and outcome of interactions with other members of a social group. They argue that the way ‘children come to understand this social context is through (minimally) exposure to and (maximally) participation in verbally marked events and activities’ (Schieffelin & Ochs 1986:167). In this sense, participant observation is very much like language socialisation, and other forms of learning which go on throughout life (Jenkins 1994). Learning requires both participating and observing: the ethnographer aims to be exposed to and participate in the events and activities under study. In discussing ‘socialising affect,’ Schieffelin & Ochs argue: ‘[a]s a part of becoming socially and linguistically competent members of particular social groups, children must learn how to appropriately convey their feelings to others as well as to recognize the moods and emotions that others

Terzimahalla, Durmish Aslano and me 27 display’ (Schieffelin & Ochs 1986:178). This is again similar to the ethnographic method. Children and ethnographers learn through the reactions and responses of others, through experience. Learning culture and becoming part of a society are what the anthropologist Garrett (2008) defines as ‘the human developmental process whereby a child or other novice (of any age) acquires the knowledge, skills, orientations, and practices that enable him or her to participate in the social life of a particular community’ (Garrett 2008:189). The process of learning these rules is what children and adults in new situations go through continuously, and what the anthropologist aspires to. However, the analogy of the anthropologist to the child or novice has its limitations. These relate not just to the development of the child, but also to the fact that children are learning a way of being and interacting with the world around them, for the first time. An anthropologist is more similar to an adult learning a new culture than a child learning culture for the first time (Schwartz 1981). The anthropologist is like a child, in that they are gradually socialised into the society, and learn how to function as an adult individual in that society. They are different, however, in that they are not starting from a blank slate, but come with various ideas, assumptions and presuppositions, and are (at least partially) understood to belong somewhere else. In addition, the ethnographer does not, at least for the most part, learn unconsciously. In fact, they must be acutely aware of the way they are learning, and the way they are being taught. The anthropologist Wolcott (1982:100) emphasises that ‘nothing is more unnatural and unchildlike than the ethnographer learning a culture, since the ethnographer is trying quite self-consciously to learn “another” culture and to make that learning explicit.’ While a child or adult learner may play an active role in their socialisation, they differ from the anthropologist in not seeking to make what they are learning explicit. While it is clear that the ethnographic experience can be understood as a learning process, as socialisation into a group, learning the language of the group under study is often not seen as part of this ethnographic process. This is not to say that anthropologists don’t draw on language or indeed language learning in their ethnographies, but rather that this is not usually explicitly stated as a method. There is, on the other hand, some work which promotes the use of ethnography in language learning. While coming from a different angle, and having different aims, this work is of interest because it argues, first, that language is inseparable from culture, and, secondly, that in highlighting the differences between languages and between cultures, ethnography helps us to better understand another language. Understanding the complex webs of meaning that make up social relations can help us to speak another language, as by making the implicit explicit we can understand difference, and thereby highlight the nodes of language that are alien and cannot be translated into a language we are more familiar with. An ethnographic approach to other languages and cultures can help learners understand some of the different representations of realities encapsulated in different languages (Byram 1997; Agar 2009; Byram & Esarte-Sarries 1991; Roberts 1997). From this perspective, culture is the context you need for learning a language. I wish to take this further by suggesting that culture is more than just the

28

Roma, Prizren and language

context for language – the relation is more profound. By using language learning as a method, I would argue that it is possible to come to a deeper understanding of the way language is used and understood in a particular fieldsite. I do not treat the cultural differences between myself and my informants as necessary background knowledge for learning Romani, but rather, by treating language as an intrinsic part of that culture, I hope to better understand both. Language is often treated by ethnographers as a skill necessary prior to fieldwork, rather than a part of the ethnographic process, and anthropologists often treat their language acquisition as complete and perfect, ignoring the learning process (Moore 2009; Phipps 2010).3 By focusing on how I negotiated my own position in the field as a language learner, while using what I learnt to understand people’s use of and attitudes towards language, I chose instead to investigate this process. Studies of socialisation tend to focus on general acquisition of language and culture, rather than treating language as a specific skill to be learned (Pelissier 1991). Anthropologists often see their task as understanding the non-linguistic and implicit in culture (see for example Bloch [1991]). As such, language is treated neither as an intrinsic part of the culture under study, nor as a specific skill to be studied by anthropologists in the way that, for example, basket-making may be. In this sense, anthropological understandings of language often treat language as a metaphor for culture, rather than a metonym of culture. Instead, I propose, by seeing language as metonymic, language learning can become a method for opening a new understanding of culture. The role of the anthropologist is a gradual process of becoming, of learning culture. If language is metonymic of culture, language learning can also be a part of this process. Anthropologists, unlike children, are learning from the perspective of a specific personal, social and historical background. In this respect, they are not just learning language and culture, but learning difference, and as such are likely to be more acutely aware of the nodes in language and culture that are significantly different to their own. At the same time, they are not learning this difference in order to erase it, but rather to make it explicit. In this sense, the cultural and linguistic background of the anthropologist acts as a negative imprint against which a contrast can be drawn.

Learning Romani While trying to learn ‘authentic’ mixed Romani, I repeatedly ran into the same problem: people didn’t want me to learn it. When I tried to explain my desire to learn everyday speech, most people were resistant to this idea. I was struggling to learn what I considered to be the way people really talked, while they were keen to teach me what they thought was the most appropriate form for a foreigner to learn, their standard (as discussed in Chapter 5).4 Later, I realised that rather than trying to learn what I perceived to be the ‘authentic’ way of speaking, I should instead focus on the process of learning, and use people’s resistance to my learning the everyday variant, their insistence on my learning to speak a particular way, as material in itself. Rather than attempting to

Terzimahalla, Durmish Aslano and me 29 overcome the language barrier, it became necessary to focus on it. I shifted my aims from learning ‘authentic’ speech to using participant observation to focus on differences in speech. In sociolinguistics, Bucholtz (2003) discusses how participant observation can be an attempt to overcome the ‘observer’s paradox,’ by becoming quasi-community members. Learning involves becoming, and overcoming difference, and the aim is to become a competent member of the group in order to research their language use, without them feeling the need to perform in a certain way for the researcher. For Bucholtz, the aim of overcoming difference is not simply to fit in, but to use the insider perspective to deploy cultural critique. In my case, reaching this stage of intimacy, where I could go unnoticed, might have allowed me to access the ‘authentic’ speech of Terzimahalla. However, this would have missed an important element of ethnographic research. Aside from the question of whether speaking fluently and authentically is ever completely possible, it would require focusing on language as a product of language socialisation. By focusing on the product of knowing a way of speaking, rather than the process of coming to know, one of the main strengths of participant observation is lost, namely, the focus on the processual nature of human knowledge. In wanting to subvert boundaries, by learning how to mix, I had originally focused on the idea of a mixed speech variant as a product of the learning process, in contrast to the product that they wanted me to learn, which turned out to be a local Romani standard. Once I stopped focusing on the product of language learning, whether the mixed way of talking I had hoped to learn or the standard pure form they wanted to teach me, it became much clearer how and why different ways of speaking mattered in different contexts. I needed to learn gradually to be like them in order to be accepted by them, as Evans (2006) describes in positioning the ethnographer as novice, in an unequal relationship with the participants. However, learning gradually to be accepted by participants is not just a practical issue: it also highlights differences between the researcher and the participants, and therefore is part of the method. Rather than the ultimate aim of full language acquisition, overcoming differences and becoming embedded in the fieldsite, I was observing the way people spoke and participating in conversations. Using language socialisation as a central method allowed me to focus on the relationship between language ideology and speech practices. I was not trying to record natural or ‘authentic’ speech, nor was I trying to analyse language structure. This would have involved very different methods. Once language is abstracted into a structure, it can be described and analysed by linguists. It can also subsequently be prescribed, in the sense that certain variants and structures are defined as correct and incorrect, for example for the purpose of teaching to those who don’t speak the language, or for the more political purpose of valorising a particular way of speaking at the expense of another. From this perspective, language data are a product to be analysed. The linguist here needs to be relatively invisible, and objective, which is to say to minimise their own impact on the data they are collecting. In much contemporary anthropology, on the other hand, the role of the researcher is positively embraced, and the process of becoming (to a greater or lesser extent) part of the group is what is treated as data. For

30

Roma, Prizren and language

this reason, anthropologists who work ‘at home’ often speak of the need to make otherness, to make things foreign, before being able to understand things afresh (e.g. Messerschmidt 1981). This does not mean there is no overlap between linguistics and anthropology, or that one area or method is preferable for the study of language. I expect that many researchers from both disciplines may engage in similar processes in fieldwork. However, they approach them from different perspectives. Treating language learning as participant observation enabled me to gain a deep understanding of the ideals, inconsistencies and hierarchies in language ideologies and speech practice among a relatively small group of multilingual Roma in Prizren, as well as learning what was at stake when it came to the various projects and funding opportunities aimed at promoting Romani language and culture. By contrast, many different questions could be asked about Romani language and multilingualism in Prizren which would require different methods, and entail focusing on the product of language learning. For example, I often got the impression that the Romani spoken in Terzimahalla had stricter boundaries between languages than that spoken by the Roma from the enclaves, who slipped fluidly between Romani and BCS, and often seemed unsure of a distinction between the two languages. This created problems in understanding between some members of the two groups, as some Roma from Terzimahalla did not know BCS. This ethnographic observation made me wonder whether the generational shifts in their main other language among Roma from Terzimahalla meant that they had something close to a monolingual mode. This is a term coined by Grosjean (2008) to describe a mode of speaking among bilinguals where one language is deactivated, in contrast to more fluid mixing where both languages are considered active. With grandparents speaking Turkish, parents BCS and children Albanian, the ability to speak Romani distinctly from other languages, which may be termed a monolingual mode, seemed to exist within the households of these families. However, while this impression may be informed and supported by ethnography, ethnographic data cannot prove empirically that the shift in main second languages through the generations is what has caused this state of affairs. This would require a different, more systematic approach to data collection. What participant observation in language learning did elicit on this issue is that while some people from Terzimahalla complained about the Roma from the enclaves mixing too much, they also felt that they themselves mixed too much. Their criticisms of both the bilingual mode of the enclaves and the monolingual mode of Terzimahalla stemmed from a purist ideology which categorised both the use of recognisable borrowings (in Terzimahalla speech) and bilingual mode (in enclave speech) as mixed, and therefore impure. A different methodology could analyse language data to show differences in bilingual and monolingual mode with regards to Romani, and the cause of this difference, but participant observation showed what practices occur and how people understood them. Rather than being frustrated with not being able to learn ‘authentic’ language, in my case the everyday talk of Terzimahalla, I found that language learning as a methodology was best for eliciting the learning process, which therefore became

Terzimahalla, Durmish Aslano and me 31 central to my project. From this, I came to understand the ways in which my role as a learner, a foreigner and a researcher affected what they wanted to teach me. Unlike the way a child learns, I came to the field with certain ideas about what I wanted to learn, which didn’t match what they thought I wanted to learn. Similarly, unlike a child, I needed to be constantly aware of the process of teaching and learning and to use it as a method.

Conclusion As an unmarried woman from a western European country I was seen as fulfilling a specific role in both the mahalla, and the NGO. As someone from the west I was seen as someone in a position of power, and someone with the necessary cultural capital to help the NGO gain funds. On the other hand as a young, unmarried woman in a patriarchal society, particularly within the private space of the mahalla, I was treated as someone who should be subservient, or even someone who was dangerous by not having a position in family relations there. In this sense rather than being an outsider who gradually became an insider through a process of learning, I became increasingly peripheral to the private space of the mahalla, and spent most of my time in the more public space of the NGO. The role of the anthropologist has long been compared to a novice, learning another culture. Here I have sought to use learning a language as a form of participant observation, to better understand the way social relations are enacted with regards to language. The anthropologist is in some ways akin to both a child learner, learning for the first time, and an adult, learning in relation to the things that are already known, but is also distinct in that by using learning as a method I am seeking to make explicit the implicit rule of culture and language. My positionality, described above, affected the way I learnt both in terms of what I already know, and in terms of how people perceived me and what they wanted to learn. People are socialised into cultures and languages in different ways at different stages in their development, and as learning builds on other learning, their understandings of what they have learnt will be informed and changed by what they already know. The process of acquiring ‘cultural intimacy’ that the anthropologist aspires to is, in most cases, also one of acquiring linguistic intimacy. Herzfeld describes how ‘cultural intimacy often consists of “what everyone knows”. This is inadmissible knowledge before the court of official opinion, and thus is often conveyed in idioms that escape the textual, verbocentric, and legalistic preoccupations of bureaucracy and formal education’ (Herzfeld 2009:134). I would argue that it is these things that everyone knows, that go without saying, which in part make up communicative competence (Hymes 1972). In this sense by using language learning as a method I was investigating the process of acquiring cultural intimacy with reference to my own positionality in the field.

Notes 1 This is not specific to Prizren but a general feature of trying to explain academic terms to non-specialists.

32

Roma, Prizren and language

2 Enculturation refers to roughly the same process of learning the rules to act in a certain society, though there seems to be some disagreement about whether the distinction is disciplinary (with socialisation coming from sociology, and enculturation from anthropology; see sociologist and linguist Mehan & Griffin [1980]), or whether there is a conceptual difference (with socialisation referring to child-rearing practices, and enculturation referring to the active participation of the child, see the anthropologist Schwartz [1981]). For the sake of simplicity I will use the more common term, socialisation, to refer to the process by which a person learns to act in a way that is culturally appropriate. 3 This perspective is often reinforced by funding bodies who require researchers to have the requisite language skills before embarking on fieldwork, or provide a specified period at the start of fieldwork for language training. For example, I was awarded an additional six months for ‘difficult language training’ before starting my year of fieldwork proper. In practice, my language learning continued throughout, and there was no such distinction between language training and fieldwork. 4 Some young people were an exception. While Nexhip’s sister, Fatmira, would tell me I should learn properly from Nexhip, her daughter Gynesh told me not to worry, and just say things the way that they say them.

2

Durmish Aslano in Prizren From partisan to NGO

Introduction The NGO Durmish Aslano, where I spent a substantial amount of time during my fieldwork has a long history in the town. It was founded as a KAM (Kulturuno Artistikano Malipe [Ro: ‘Cultural Artistic Society’]) in 1969, and continued through the period of decentralisation of Yugoslavia, through the Milošević era and the rise of violent ethno-nationalism, to the conflict in 1999 and the eventual victory of the KLA and international forces. Named after a Roma partisan from Prizren, the organisation was initially engaged in music and folklore performance. After the 1999 conflict, it was reconfigured as an NGO, with the aim of meeting the new needs of Roma in post-war Kosovo, and started to work in education, human rights, media as well as running other projects in line with the objectives of international funding organisations that will be discussed in Chapter 3. In 2007, the radio station Radio Romano Avazo (Ro: ‘Radio Romani Voice’) was set up, with funding from the American embassy. The organisation also hosts the theatre group which will be discussed in more detail in Chapter 4. People are proud of the long history of the organisation, citing it as evidence that Prizren Roma have always been cultured and educated. The people who work there today often discuss the organisation’s past and the involvement of their uncles, fathers and even grandfathers. By unpicking the history of the organisation we can shed light on the biographies of individuals, the position of Roma in the town as well as state ethnicity policies. This informs the social hierarchies implied in language standardisation as will be discussed in Chapter 5. The scale of the organisation and that of state politics are inextricably linked, and the linguistic ideologies and practices of those I worked with are, and have always been, contingent on these broader shifts. Thus, language politics have shaped Roma and the Romani organisation in Prizren. As such, by showing the relationship between the organisation’s history and the history of the town, and the region, we can start to see how historical change structures shifts in language ideologies, as will be discussed in Chapter 6. This chapter uses radio interviews recorded for Radio Prishtina in 1972 (shortly after the inception of the organisation) as well as monographs produced by the NGO about its history, secondary historical sources and ethnographic data to describe these different scales of history, and how they affect the contemporary

34

Roma, Prizren and language

position of Roma in Prizren. The first section describes the historical context before the founding of the organisation, including language and ethnicity policy, and socio-economic divisions. The second section goes on to discuss the period of decentralisation, and the growing Romani movement. The third section uses historical sources to describe the economic decline of the 1980s followed by nationalist division and conflict in the 1990s, and the fourth section uses ethnographic material to describe the position of Roma during and after the conflict.

1969: KAM Durmish Aslano The NGO has produced two monographs with Nexhip as the author (Menekshe 2004; Menekshe 2009) which describe the history of the organisation, and with it the history of Roma in Prizren. They contain photos (past and present) and pictures of various activities, and discussions of Romani identity. The Indian origins of the Roma and Indian (or seemingly Indian) aspects of their identity today are emphasised, but their longstanding integration into the town, and their connection to Turkey and the Ottomans, is also central. The history of the organisation is therefore used both to celebrate Romaniness and forge a link with India, and also to show the place of Roma in Prizren. The organisation was founded at a time of rapid change in Yugoslav government and society. According to an interview published in this book with Shaip Menekshe, a relative of Nexhip and one of the founders of the organisation, and its second director (from 1971 to 1976), this was part of a wider move to accept and affirm national identity in Yugoslavia.1 It was thus one of several ethnic ‘cultural artistic societies’ set up around this time, and while these organisations performed the culture of a particular ethnic group, many of the Roma mentioned in publications on the history of the organisation were also active in the Albanian KAM (Agimi) and Turkish KAM (Dogri Yol) in Prizren. In a recording from 1972 from Radio Prishtina (Misini 2014b), which I was told about while on fieldwork, an Albanian journalist interviewed those who were active in the organisation. The interview was about the KAM, which had then been active for three years, and performed music and folklore. The interview was in Albanian, Romani and BCS, interspersed with Romani music performed by the interviewees, described as këngët autentikët të romve (Shq: ‘Authentic songs of the Roma’). To begin with, the questions were asked in Albanian, with the responses in Romani, though the interviewer often had to repeat the questions in BCS so respondents could understand, especially the children. In some cases, this led to the whole interview switching to BCS. The interviewer introduced the programme in Albanian, and the first interviewee was Shaip Menekshe. He declared in Romani: Romalen, Srbiya, Albayna, Korhaya … Angluno descende sin olen but bari pharipe. Pharipe sine vash oleske kay na sine kote to kerol buti, na sine amen lovo, garderoba, instrumentya. Odole

Durmish Aslano in Prizren 35 kedime vahteste, hem but pomushka amen sah amare mala, a hem but amare prezedenti, angluno o Nusret Sehari, o … ov hito profesor albanolojike. Oleya sa o mala, o Roma phik phikeya amaro malipe adine. Roma, Serbians, Albanians, Turks … The first members had many difficulties. The difficulties they had were that they didn’t have anywhere to work, we didn’t have money, costumes, instruments. With time, we gathered these things, and we were greatly helped by our friends, and even more by our president, the first one, Nusret Sehari, he … he is a teacher of Albanian language. With him and all the friends, the Roma, came together arm in arm with our society. This was clearly a well-prepared speech, developing a formal register in Romani, suitable for socialist-era radio announcements. He narrated the organisation as struggling against hardship and succeeding through cooperation and unity. The interviews with the children were less prepared. The children were interviewed in Albanian, which they often struggled to understand and to speak. Sometimes they seemed to understand, but didn’t know how to respond, and on other occasions it sounded like someone may have been translating quietly and prompting them. The interviewer asked about the organisation, the daily life of its members who were being interviewed and about their music: Interview with Skenderi: I: Kush ka qen Durmish Asllano? S: Durmiš Aslano … bio narodni heroj iz Prizrena. I: Jel ima on nekog živog ovde? S: Da, ima oca živi. Još je živ. Ima i sestru. I: Çka punon ti, Skender? S: Ja radim … sobaslikar. I: A fiton pare mjaft? S: Dosta zarađujem, samo sezonski. I: A punojnë gjith Rom te profesiona të ndryshme sikur ti, për shembull? S: Po, punojn … I: (Shq) Who was Durmish Aslano? S: (BCS) Durmish Aslano was … a national hero from Prizren. I: (BCS) Does he have anyone alive here? S: (BCS) Yes. He has a living father. His father is still alive. He also has a sister. I: (Shq) What do you do Skenderi? S: (BCS) I work … painter and decorator. I: (Shq) Do you earn enough money? S: (BCS) I earn enough, but it’s only seasonal.2 I: (Shq) And do all Roma work in various professions like you, for example? S: (Shq) Yes, they work …

36

Roma, Prizren and language

Skenderi sounds young, and speaks little Albanian, but tries to use it in the last sentence. As he stumbles in BCS, it is hard to tell whether his BCS isn’t that fluent, or he is just nervous. His BCS is clearly better than his Albanian. The interviewer persists in questioning him in Albanian, and may be reluctant to switch the interview completely to BCS. His interest in Durmish Aslano the man, and showing him to be a Roma partisan, switches to an interest in Skenderi’s work, and showing that Roma work. Durmish Aslano himself is not described as Roma in this interview, but rather as narodni heroj, a ‘hero of the people.’ The fact that Durmish Aslano had fought and died for Tito’s partisans made him a hero in the eyes of the government and the local Roma population.3 Having a Roma partisan, and naming the organisation after him, can be seen as a way of displaying belonging to the socialist Yugoslav partisan myth through being Roma. The founders navigated the establishment of an ethnically based organisation in politically turbulent times through the link to the partisan past. The partisan myth was used to legitimise Communist power at the end of the Second World War. Prior to this, in the Kingdom of Yugoslavia, Serbo-Croat was in official use in the region of Kosovo, and there was also Serb and Montenegrin colonisation of the region. Albanians who were educated in this period were educated either in Serbian-language schools, or religious institutions which used Turkish and Quranic Arabic (Kostovicova 2002; Poulton 2003). During the Second World War the region was under the control of Italian and, later, German forces. While for Slavs this was generally seen as an occupation, many Albanians treated it as a liberation from Yugoslav rule (Marmullaku 1975; Marmullaku 2003). Under Italian fascist rule, Albanian-language schools, administration and newspapers were set up on a large scale (Poulton 2003).4 The policies in Italian areas were not so severe as the genocidal policies in German-occupied areas. However, while Jewish, Serb and Roma populations may have been relatively safe from genocide, there were still retaliation attacks by Albanians against Serbs for the treatment of Albanians in the Kingdom. After the war, despite the very different form of the new Yugoslav state, there was resistance among Albanians to its return. Partisan forces found it necessary to impose martial law in 1945 until resistance in the Drenica region was crushed (Malcolm 2002). The treatment of Albanians in what was now the Republic of Serbia, Yugoslavia, was also dependant on relations with the newly formed People’s Socialist Republic of Albania. While the relationship between Tito’s and Hoxha’s forces was initially good, there were still revenge attacks on Albanians in Yugoslavia (Marmullaku 2003). In the three years following the Second World War, Yugoslav Turks were treated with some suspicion as Turkey was seen to be supportive of the United States. Tito planned to build a Balkan Federation, to include Albania and Bulgaria (and thus unite territories inhabited by Albanians without taking them away from South Slavs). This plan failed, and Yugoslavia was ejected from Cominform in 1948 (Lampe 2000). This meant a rift between Tito and Hoxha, and the closing of the border between Albania and Yugoslavia, which in turn meant a shift in political orientation: while Yugoslavia eventually opted for non-alignment in the Cold War, Albania remained loyal to

Durmish Aslano in Prizren 37 Stalin. Borders between Albania and Yugoslavia remained closed, and Albanians in Kosovo (rather than Turks) fell under the suspicion of the Yugoslav government and security services. The Turkish language, long neglected in the Kingdom of Yugoslavia, became accepted and even used as a language of instruction in schools in areas such as Prizren, where it was widely spoken. There were also a series of unpopular campaigns to disarm Albanians, as well as ultimately unsuccessful collectivisation campaigns which disproportionately affected rural populations, primarily Albanians (Marmullaku 1975; Lampe 2000:Ch.8; Malcolm 2002). At this time, as the Communist Party consolidated power, nationality was downplayed by the authorities, in favour of ideological affinities. The distinction between communists and fascists was more important than that between Serbs and Albanians. The fact that this division often ran along ethnic lines was overlooked. The role of the partisans in the war, as anti-fascists and revolutionaries, was mythologised, to legitimise the new regime. However, while state organs privileged ideological purity – anti-fascism – over the idea of national purity, nationality was still treated as a mark of the political impurity of certain groups, in that certain nationalities were suspected of sympathising with ‘enemies of the people.’ In this way, Yugoslavist, anti-nationalist centralism paradoxically implied the repression of national minorities, in this case Albanians. There was complex wrangling over the removal of colonists and the Serbianisation of Albanian names in Kosovo, as well as the use of Albanian in public institutions. In the hope of distancing themselves from the interwar state, the government gave way on some of these issues, including allowing Albanian-language schools. However, due to the deprivation of Albanian education in previous years, and the primarily rural and uneducated situation of Albanians in Kosovo, there were very few Albanian-speaking teachers or indeed literate professional Albanian speakers of any description (Malcolm 2002). From 1956, there were increasing opportunities for those who might otherwise have identified as Albanian to identify as Turkish (Malcolm 2002; Marmullaku 1975; Poulton 2003).5 While some Albanian-language primary schools existed, schools which taught in Turkish were increasingly set up, encouraging a shift to Turkish. In addition, an agreement between the Yugoslav and Turkish governments allowed for the migration of ethnic Turks to Turkey.6 Census data on ethnic groups shows the Albanian population falling in this period. Seeming shifts in populations suggest that people’s choice of declaring themselves as Turkish or Albanian was by no means constant, and that for many, ethnic affiliation was decided by political pressures. Conversely, Tanasković (1991) argues that as Turks and Albanians generally inhabited the same areas, it was common for census declarations to shift, and for Turks to declare as Albanian; he thus argues that many Albanians are ‘really’ Turks. However, given the malleability of ethnic self-identification and the particularly porous border between the designation of Turks and Albanians, it seems likely that the shift goes both ways, and in certain historical circumstances and under certain political pressures many people would also switch their allegiances the other way. Indeed, they may well have seen

38

Roma, Prizren and language

no contradiction in identifying as both. There is no precise correlation between language and ethnicity, and figures on language use are problematic, owing to bilingualism and also to the fact that many declare their main language as one other than that of their ethnic group (Bugarski 1991). In Prizren, this meant that Albanians and Roma who attended school, as well as ethnic Turks, would attend in Turkish. The choice of education in Turkish may seem strange in a centralist Slavic state, but it seems that this was intended to discourage the use of Albanian. However, the option of studying in Turkish, and declaring as an ethnic Turk, was not open to all. Ideologically, the authorities considered class more important than ethnicity, and, in theory at least, people were not excluded on the basis of their ethnicity. The emphasis on workers’ rights, however, led to another form of exclusion: the exclusion of rural people. Self-management meant that work, wages and welfare were all organised by the factory committee, or other workplace. As agriculture had not been successfully collectivised, those in rural areas were not directly attached to a self-managed institution, and faced a variety of problems. Stubbs (2009) explains the localised, self-managed social welfare system which was built in Yugoslavia, explaining that ‘(f)rom the late 1950s the experiment of workers’ self-management went alongside a need for professionally trained social workers responding to social problems and engaged in local social planning’ (Stubbs 2009:1). The self-management of social work meant that local issues would be dealt with more effectively, but also that there was regional variation. Stubbs goes on to explain that Yugoslavia had ‘well-developed social insurance systems, albeit within a continued dual social structure favouring urban industrial workers over those reliant on agriculture’ (Stubbs 2009:1). While everyone was offered access to basic services, most benefits were work-related, meaning they were accessed via the workplace. While in other urban areas housing was a principal benefit both of being employed in self-managed organisations and of holding a higher position in these organisations (Archer 2016), this would not have affected the inhabitants of Terzimahalla and much of central Prizren, as the housing was privately owned. The problem with this structure of benefits was that it excluded those not working in factories, especially those in poor rural areas. On this topic, Deacon and Stubbs (2007:10) argue that ‘modernisation linked to industrialisation and rapid urbanization created a kind of dual social structure, with those employed in the socialist industrial sector having work-based benefits which small farmers did not have.’ On the other hand, employment took priority over economic efficiency; as Stubbs and Maglajlić (2012:5) put it, ‘emphasis on full employment led to factories being built in some locations not solely for economic reasons, but also due to social and political factors.’ While the socialist era provided work based, at least in part, on the needs of workers, rather than the needs of the market, the emphasis on providing work and benefits in urban areas further entrenched the socio-economic divide between urban and rural areas, and this divide also ran along ethnic lines, as most of the Kosovan rural poor were Albanians. Urban Roma who had factory jobs, such as those from Terzimahalla, were certainly not part of the lowest social class in Kosovan society of this period. Many of the

Durmish Aslano in Prizren 39 men had access to education, followed by jobs, and therefore access to services. Rural, primarily Albanian, populations lacked these benefits.7 As such, lack of access to Yugoslav modernity correlates with a later anti-socialist movement (Allcock 2000). In addition to self-management, another peculiarity was that Yugoslavia, unlike all other socialist countries, arranged for workers from a variety of social backgrounds to legally go and work abroad. Seasonal migration has a long history in the region, and was prevalent through most of the Yugoslav period, peaking in the early seventies (le Normand 2016; cf. Daniel 2007). While remittances were a major boost to the Yugoslav economy, work abroad left skilled workers with expectations of living standards which Yugoslavia could not fulfil, while increasing the disillusionment of unskilled workers with the inequalities within the Yugoslav state.8 Economic inequality between the wealthiest republic, Slovenia, and the poorest part, the province of Kosovo, also continued to grow throughout the socialist era, due to ‘the politically encouraged growth of regional disparity and disconnection’ (Lampe 2003:194). The relative lack of industrialisation and the large rural population (primarily ethnic Albanians) led to an increase in inequality at every scale. While for many, self-management and non-alignment provided greater access to prosperity than was apparent in other socialist countries, it also created different divisions and inequalities, and a class of people who had no way to access this prosperity other than by becoming migrant workers (Ströhle 2016).

From workers to nations: changing Yugoslavs In the 1972 interview with members of Durmish Aslano described above, the interviewer also asks the youngest member, Dashni Veshal (aged around 11)9, about his schooling: I: Gilavav o romenge, titullove të këngët qe interpreton Dashni Veshal, solisti me i ri i ksaj shoqerie. Dashni, n’cilën klas shkon ti? DV: Ja … ja idem …. I: Koji razred? DV: Peti razred. I: Jel si dobar đak? DV: Da. I: A … kako ti se zove učitelj, nastavnik? DV: Učitelj se zove H … Hurmet Šeju. I: Ti ideš na Turskom jeziku, slušaš … nastavu … DV: Jest. I: I šta misliš, da bi te bilo lakše da slušaš nastavu na romskom jeziku? DV: Da. I: Jel cepaš knjige? DV: Ne.

40

Roma, Prizren and language

I: Jel se igraš lopte? Pause. Voliš da igraš lopte? DV: Da, volim. I: U ulici tamo sa dečacima sa … DV: Volim kad imam vreme. I: Nemaš vremena, jel? DV: Je … I: Dobro, ajde! Music. I: (Ro) I sing about Roma. (Shq) The title of the song as interpreted by Dashni Veshal, the youngest soloist in the society. Dashni, which grade are you in? DV: (BCS) I … I go … I: (BCS) Which grade? DV: Fifth grade. I: Are you a good pupil? DV: Yes. I: Err … What is your teacher called? DV: The teacher is called H … Hurmet Sheyu. I: You go in Turkish language … you take … classes. DV: Right. I: And what do you think, would it be easier to take classes in Romani language? DV: Yes. I: Do you tear your books? DV: No. I: Do you play ball? Pause. Do you like to play with a ball? DV: Yes, I like … I: In the street with the other boys … DV: I like it when I have time. I: You don’t have time? DV: Yes … I: Ok, come on! Music. In this interview, the interviewer realises that the child doesn’t speak Albanian, and switches completely to BCS, which the child knows, although he goes to a Turkish-language school. Again, the interviewer seems to want to undo prejudices about Roma by showing that Roma children don’t tear books, and also to assert Romani identity and link it to language by showing it would be easier to go to school in Romani language. The child seems to know the ‘correct’ answer to both these questions (it is possible the questions were rehearsed), but flounders

Durmish Aslano in Prizren 41 on the question about playing with a ball. Presumably he wasn’t sure whether this would be approved of or not, and eventually decided to say he only played when he had time. This discussion of Romani language in schools took place in a context of shifting policies, which increasingly allowed issues such as language rights and prejudice against minorities to be discussed more openly. The shifting between languages is indicative not so much of societal multilingualism (there are clearly comprehension problems), but rather of the confusion caused by an abrupt shift in language and nationality policy. Serbo-Croat was no longer the lingua franca between ethnic minorities, as Albanian was becoming dominant in Kosovo, but there had not yet been time for everyone to learn Albanian. The founding of the organisation, and this interview, took place in a period of significant turmoil in Yugoslavia, entailing a paradigm shift in nationality policy. Rights of ethnic minorities were increasingly recognised, and real power was devolved; but, at the same time, growing decentralisation increased economic inequality across the federation. Both processes were especially relevant to Kosovo, as a province with an Albanian majority and as the poorest part of the federation. This shift started with the purge of Ranković in 1966 and culminated in the new constitution of 1974. Ranković had been Minister of the Interior from the end of the war until his removal in 1966, ostensibly for bugging Tito’s private rooms. He had implemented strict controls on Kosovo Albanians, which were relaxed after his removal. This moment is described by Jović (2003) as a shift from Tito’s idea of a centralised Yugoslav state, built on the Stalinist model, to Kardelj’s10 decentralised federal model. Jović defines this shift as a new interpretation of the ‘socialist + nationalist formula,’ changing from the statist view, to the socialist non-statist one, with the latter including the devolution of power to the smaller nations (cf. Unkovski Korica 2015). 1968 saw social movements and protests across eastern and western Europe. While in parts of the Soviet sphere, protesters called for a relaxation of stringent controls and a move away from totalitarianism, in Yugoslavia many of the protesters were asking for a more orthodox Marxism. This was problematic for the Belgrade government, in that protests by Albanians were also calling for self-determination, and this was seen as a desire to move closer to the variant of Marxism practised in Hoxha’s Albania, which had by this time cut off ties with Moscow. This was followed by the Croatian Spring of 1971, a series of protests by students and others which focused, amongst other things, on the status of the Croatian literary language. For many, Belgrade’s insistence on a unified SerboCroat was metonymic of wider centralist policies and, as such, Croatian language and orthography became a rallying point. The negotiation of these challenges to the regime eventually led to a relaxation of repressive measures against Albanians in Kosovo, and ended in 1974 with the new constitution, which included new ethnic designations.11 This shift also took place around the time of the rapprochement between Yugoslavia and Albania, as Hoxha had lost the support of the Soviet Union. As the threat of a Soviet invasion of Yugoslavia via Albania receded, strict controls

42

Roma, Prizren and language

on Albanians were relaxed. In 1967, Tito visited Kosovo, and also announced that the offensive Serbian term for Albanian šiptar was no longer acceptable (Malcolm 2002). Albanian symbols and the Albanian flag were no longer prohibited, and in 1969, the University of Prishtina was set up, to teach in Albanian as well as SerboCroat. Teachers were brought from Tirana to set up the course, using textbooks from Albania. At the same time, the Albanian literary language was finally being codified in Tirana, on the basis of Tosk, a process completed in 1972. Gheg speakers in Kosovo who called for more language rights in this period were not just calling for the right to use the Albanian they spoke, but also the right to speak the new standard, and be part of a broader Albanian imagined community. By 1974 there was also Albanian-language television in Kosovo, and in the 1974 constitution Kosovo was made a Socialist Autonomous Province. While the unrest in this period is now commonly understood solely in nationalist terms, the issue of economic decentralisation, rather than ethno-national sovereignty (of the republic, or autonomous province), was also crucial. The protests were defused, and the criticisms of the protesters were to a certain extent taken on board. The 1974 constitution shifted nationality policy away from the idea of a Yugoslav national identity, such that ‘Yugoslav’ became an umbrella term for a group of nations, nationalities and ethnic groups; but it also decentralised power away from Belgrade. In this period Albanians were granted increasing rights, and, while Kosovo was never made a de jure republic, its autonomy meant it was de facto equal to the republics. Increased recognition of Albanians was welcomed by many. However, economic decentralisation in the subsequent period also led to a rise in inequality between the republics and provinces. More independence from Belgrade meant that Kosovo had more control, but less resources. Meanwhile, the self-management system which tied benefits to worker status continued to increase the rural/urban divide (Stubbs 2002; Ströhle 2016). Roma in socialist Yugoslavia were defined as an ethnic group, rather than a nation (with a ‘homeland’ inside Yugoslavia) or a nationality (a national group who had a ‘homeland’ outside Yugoslavia). Their lack of a ‘homeland’ outside Yugoslavia meant that policies towards them did not change according to shifts in international relations in the way that those towards Turks and Albanians did. Despite these ethnic and national designations, boundaries between the groups remained porous. The musicians who set up Durmish Aslano also played for Turkish, Albanian and Serbian musical groups. Being Roma did not preclude being part of these other organisations, and did not seem to be a problem for the organisations themselves. The shift in nationality policy did not mean that ethnic groups were understood as bounded, essential categories. While Roma did not have a territorial claim to either a republic or a province, this period of change in nationality policy facilitated the emergence of a number of Roma organisations within Yugoslavia, at the same time as the international Roma movement was growing – the first World Romani Congress was held in London in 1971. Roma organisations were formed in Belgrade and Skopje, the former headed by Slobodan Berberski, a member of the Central Committee of the Yugoslav Communist Party, who was to preside over the World Congress. Roma

Durmish Aslano in Prizren 43 from Prizren were also connected to the international movement. The book published to celebrate the 40th birthday of Durmish Aslano in 2009 states: Ano angluno bershuno kedipe e K.A.M. “Durmish Aslano” misafirya sine o Slobodan Berberski thay Rayko Juriq, kolencar sinemen beshipe thay yekh solmyuno habe ano hotel Turist. (Menekshe 2004:63) At the first annual meeting of the KAM Durmish Aslano, the guests were Slobodan Berberski and Rajko Đurić, who held a seminar, and had a ceremonial dinner at hotel Tourist. Nusret Sehari (also Nysret Seharsoj), the first president of Durmish Aslano (mentioned in the interview with the second president, Shaip Menekshe) was especially involved in the international movement. According to Seharsoj’s biography in the book, he was originally from Terzimahalla but went to study in Belgrade, then worked as a journalist on Albanian-language programmes at Radio Belgrade, and taught Albanian language at high school. He had also taken part in the 1971 London Congress, and so would have been engaged in issues surrounding Romani nationalism, at the same time as various local, national and federal initiatives were becoming possible. The constitutional changes outlined above thus led to the recognition of minorities, while at the same time the move in nationality policy away from the idea of a Yugoslav nation meant an increase in freedoms of expression of nationhood in the form of arts, culture and language. While Roma were not recognised as a nationality (just an ethnicity), a number of cultural and artistic societies were founded at this time. Meanwhile, Roma continued to work in factories in Prizren, as well as being involved in other manual work, traditional crafts and playing music in the summer. While in other countries the forced inclusion of Roma into the workforce, based on the stereotype of Roma as workshy, was an attempt at assimilation (Stewart 1999; cf. O’Keeffe 2013), in Prizren my informants do not remember it in these terms. Roma in Prizren now had a cultural and artistic organisation as well as membership of the urban workforce which gave them access to benefits and social support.

Nations to nationalists The peaceful (yet increasingly unequal) prosperity of the 1970s is remembered fondly in many parts of former Yugoslavia. The 1980s saw a period of economic decline and increased nationalist tensions across the country. Following Tito’s death in 1980, 1981 saw the start of student protests in Prishtina about conditions in Kosovo. As the decade progressed, the target of discontent shifted from the (by now primarily Albanian) Kosovar elite onto the Yugoslav state itself. At the same time, Milošević rose to the head of the Serbian Communist Party, and ethnonationalism increased across the federation, eventually leading to the

44

Roma, Prizren and language

break-up of Yugoslavia. As unemployment rose in Kosovo, many Serbs began to leave. Serb nationalist narratives claim this was due solely to ethnic harassment by Albanians, while worsening conditions in the province compared to other parts of Serbia were clearly a central cause. By 1984, unemployment was 29.1% for Kosovo, compared to the national average of 12.7% and 1.8% in Slovenia (Poulton 2003:130). Albanian opposition became increasingly nationalist, and the regime became increasingly oppressive. Kosovo was de facto placed under military rule from 1981, and de jure had its autonomy reduced in 1989. The unrest and repression of Albanians in Kosovo throughout the eighties had reached crisis point by 1990, when Albanians began to set up a parallel system. Increasingly nationalist politics had meant that fewer Albanian children, especially girls, were attending school (Zymberi 1991). Kosovo was a particularly emotive issue for followers of Milošević, and the gradual increase in repressive measures against Albanians culminated in the retraction of Kosovo’s autonomy in 1991. Albanians resigned en masse from state institutions, and parallel institutions were set up. The Albanian police were suspended in 1990, and Albanian deputies and functionaries were dismissed, causing them to found an underground government in Kaçaniku (Kačanik). By 1992, Yugoslav schools had stopped teaching in Albanian, continuing in BCS (officially Serbian, or Serbo-Croat) and in some places Turkish (Shatri 2010). Albanian teachers were fired. Classes in Albanian continued in people’s houses, and with the tacit agreement of the Yugoslav authorities. As well as political institutions, the LDP (Democratic League of Kosovo) set up a parallel welfare system, including schools and health clinics, and a taxation system (Malcolm 2002; Cocozzelli 2009). The LDP, leading the passive resistance, was mostly made up of urban intellectuals, many of them former communists. Later, especially after the looting of arsenals in Albania in 1997, the KLA (Kosovo Liberation Army) gained strength; it was formed mostly of young men from poor rural areas, especially Drenica, the area with the highest rate of poverty in Yugoslavia (Cocozzelli 2009; Ströhle 2016). As essentialised ethnic differences were becoming increasingly salient, the Yugoslav system ceased to offer even basic or promised benefits to the Albanian population. Regarding the dual system, Cocozzelli (2009) describes a complete segregation during the nineties, with neither system functioning fully. The Mother Teresa Society (MTS) set up a series of local health clinics with volunteer workers, paid for by taxation, in particular from the diaspora. Following the emigration of Serbs to ‘inner’ Serbia in the eighties, there was now massive emigration of Albanians to western Europe. For many Kosovo Albanians, conflict was the culmination, rather than curtailment, of the Yugoslav regime. Within this larger scale, the experiences of Roma and other minorities in Kosovo are often lost. Systemic socio-economic inequality along ethnic lines in Yugoslavia turned into outright ethnic segregation in Kosovo in the nineties. Economic stratification of ethnic groups meant that many urban Roma (as well as other minorities) carried on working within the Yugoslav state system after Albanians had set up the parallel system. This later left them open to recrimination: they were accused of siding

Durmish Aslano in Prizren 45 with the Serbs. The Yugoslav state12 itself was in disarray in this period: after the economic decline of the eighties and the conflicts of the nineties, although the promise of work-based benefits remained for many people, in practice they were not able to access them. There was a shift from the socialist era where prices were low and pensions were high, and where unemployment was high but surplus labour was absorbed by emigration and social enterprises, to the situation in the nineties where ‘(f)ormal rights were not cut; there was no attempt to adjust the system to the new economic reality’ (Arandarenko & Golicin 2007:168). Benefits went into arrears, and many turned to informal labour, which was condoned, if not encouraged. For minorities in Kosovo, this chaotic state of unfulfilled promises still offered more than the alternative. While Roma continued to survive inside the system, Albanians were divided between the former urban elites, who formed the passive resistance under Ibrahim Rugova, and the rural Albanians who had never been fully incorporated into the redistributive side of self-management socialism. Roma (and other minorities) continued to access state services, while Albanians mirrored the state either as part of the parallel state (schools, hospitals) or the army (the KLA). As severe repression of Albanians worsened and as the passive resistance became less credible, many fled abroad, while many others joined the KLA. The guerrilla activities of the KLA were treated as terrorism by the Serb authorities, and used to justify martial law, followed by mass expulsions, persecution of civilians and eventually massacres. In 1998 Adem Jashari, a KLA leader, was killed, along with 40 members of his family, precipitating an escalation in violence on both sides. While the Yugoslav Army (by this time called Vojska Jugoslavije) still had the upper hand in terms of weaponry, the collapse of law and order in Albania meant the KLA were increasingly well-armed. An international attempt to forge a ceasefire at Rambouillet failed, and eventually NATO intervened, bombing Serbia and Kosovo (both military and civilian targets) in an attempt to force Milošević into ending these atrocities. The legality of the bombings and their role in the eventual fall of Milošević are hotly debated. Milošević did, however, sign an agreement with NATO to withdraw troops from Kosovo, on condition that the bombings would end. Many have challenged the notion of ancient ethnic hatreds which is popularly attached both to the image of the Balkans and to the wars in Yugoslavia (e.g. Bracewell & Drace-Francis 1999; Verdery 1993). In particular, scholars have been critical of the idea that Balkan forms of violence are more irrational, primordial and therefore crueller than the violence carried out by states across the world (Sorabji 1995; Mueller 2000; Mazower 2000). There is a contradiction in this image: while the violence committed was seen by some to be caused by irrational primordial hatred, in fact its aim, ethnic purity, is in many respects the epitome of modernity and rationality. In certain respects, the post-conflict segregation enforced and reproduced by international agencies continued the same process, diminishing these many complexities and ambiguities of identification and appellation (Duijzings 2003). In Kosovo this resulted in institutions that are either one-sided or international:

46

Roma, Prizren and language without policy-making institutions that are representative of the Kosovo public, the province is left to be buffeted by powerful global trends as articulated by international actors such as the World Bank, or donor governments such as the USA or the UK. (Cocozzelli 2007:218–219)

Pre-existing ethnicised economic differences are in some respects reinforced by the use of a ready-made template for social policy, coupled with privatisation policies (to be discussed further in Chapter 3) which pay little or no attention to rising unemployment. The causes and consequences of this war are highly contested, as are figures relating to the numbers of people killed and forced from their homes. In general, Serb nationalist readings of the events describe the KLA as a terrorist group while not mentioning any crimes committed against Albanian civilians. NATO is portrayed as an aggressor in breach of international law, and mindlessly bombing civilians, many of whom were actively protesting against the Milošević regime. The Albanian nationalist narrative, on the other hand, focuses on the persecution of Albanian civilians, from imprisonment of protesters in the eighties through to massacres and systematic rape in the nineties. From this perspective, the KLA are heroic freedom fighters, supported by their close ally, the United States (i.e. NATO). The Albanian passive resistance is often forgotten, and reprisals against civilians (Serbs, other minorities and Albanians perceived to be collaborators) are ignored, or dismissed as being the results of the traumas faced by fighters at the hands of Serb forces. Surprisingly, the Serb and Albanian narratives do not completely contradict one another, but rather emphasise events where each group can be seen as victim, passing over any responsibility they themselves may have. These narratives, and their accompanying symbols (such as flags and monuments), mark both Serbian and Albanian territories in Kosovo, and it is almost always clear whose territory you are in. These two distinct ways of memorialising the war leave little room for the experiences of ethnic minorities, or indeed those harmed by the forces of their ‘own’ ethnic group. Those whose experiences don’t fit with either of the two narratives are generally silenced.

Forgetting the war: Roma narratives In my research, I found that the nineties are not talked about much by Roma, except when people talk about their personal lives, or deny their collective involvement in war crimes. With regards to Durmish Aslano, it is also difficult to gain information that isn’t viewed through the lens of the present. In both the books produced by the organisation (Menekshe 2004, 2009), there is a passage which says that from 1990 to 1994 the organisation was inactive due to lack of accommodation. The president in this period was Sami Koka, an elementary school teacher, and a clarinettist who also played in Agimi (the Albanian organisation) and Dogri Yol (the Turkish organisation). The regular mention of other ethnic organisations seems to be part of a wider insistence on the integration of Roma generally,

Durmish Aslano in Prizren 47 and their involvement with Albanians specifically. In 1994, Nexhip Menekshe became president, and remained so throughout my fieldwork. Nexhip (who also authored the book) states that in the years 1994 to 1998, the organisation began to work with great enthusiasm, despite the lack of a premises. It seems plausible that there was an element of competition between the two presidents. Dashni Veshal, the child from the 1972 interview, reappears, this time as a prominent musician. For Roma, the organisation had remained constant amid shifting nationality policy, language policy and ideology, as had the use of Romani language. This is not to say that Roma have been unaffected by, or isolated from, large-scale political changes. In fact, in many respects they have been highly vulnerable to turmoil brought about by political upheavals. Although people did not openly express this, their vulnerability came across in some of the discussions I had. While visiting Vera’s house after a day at the radio station we started to discuss cooking. She was a very good cook, and when I told her this, she started to talk nostalgically in BCS about her youth in the late eighties and early nineties: I’m so glad you’re my friend. It’s very hard to find real friends these days. I used to have lots of Serbian friends when I was at school. I was very spoilt as a child, because my brothers were older they were already getting married when I came along. I was free to go to school, and I used to invite my friends round for food and drink. We girls didn’t go out for coffee in those days because people would think we were sluts, but I invited friends round. We used to make such nice food for them. My mother was a good cook. I couldn’t cook at all then because I never needed to. I didn’t learn till I got married. No one believed I would stay with my husband in a Roma family because I had gone to school, and spent all my time with Serbs. But I stayed. I have a lot of respect for my husband. It was unusual for Roma women of her generation to have gone to school, and, as she mentioned, this related to her family’s peculiarity: before sons are married, daughters must share the burden of the housework; once sons are married, daughters-in-law are expected to do the majority of housework. As Vera was so much younger than her brothers, the daughters-in-law arrived before she was old enough to start doing housework and cooking, so she was exempt, and was able to complete medical high school and work at the hospital pharmacy before getting married. When she was young, her family was relatively well-off, and lived in a Serbian area, which is why she went to school mostly with Serbs, and had primarily Serbian friends. After the war the Serbs fled, and rural Albanians increasingly moved into urban areas, changing the town’s demographics significantly. Many other Roma also mentioned having Serb friends and colleagues before the war, even if they avoided association with ‘the Serbs’ as a group. Because such experiences of minorities don’t fit well into either of the main narratives, they are often silenced. Their experience of the past is still clearly shaped by the war though, and the rupture of the war means that narratives of the past are divided into two frames: before and after. As such, the war can be seen

48

Roma, Prizren and language

as an ‘interpretative turning point,’ a point in a biographical narrative which leads to a reinterpretation of past, present and future (Van Boeschoten 2000:137). The war did not just profoundly affect their lives from then on, but also affected their understanding of what went before. When people talk about ‘before the war’ they often do so fondly, with an idealised nostalgia for the old days, but on other occasions they have a more matter-of-fact approach to life before. Most Roma I spoke to rarely mentioned socialism or Yugoslavia. When they did mention Yugoslav times, they would usually describe it as pod srbima (BCS: ‘under the Serbs’) implying that now they are ‘under the Albanians,’ and they would tell me it was better under the Serbs because there were factories. This relates to perceptions of present-day unemployment (see Chapter 3). While factories are talked about unemotionally, with the straightforward notion that employment is better than unemployment, when people talk about ‘before the war,’ or simply ‘before,’ it tends to be described in much more affective and localised terms of peace and harmony. In such narratives, Prizren was a religious, cultured place of multi-ethnic harmony. These narratives are common not just among Roma, but also Turks and Albanians, and tend to lack any exact reference to a historical period. They refer more to a generic time when everything was better, and act as a point of comparison for the present. This nostalgia will be discussed in more detail in Chapter 6. Palo maribe (Ro: ‘after the war’) was a phrase I learned quickly in Romani as it is used very commonly. While it literally translates as ‘after the war,’ in most contexts when it is used a certain causality is implied: it often signifies ‘following the war,’ or even ‘because of the war.’ While few during my fieldwork stated explicitly that Roma were accused of involvement in Serb atrocities, it was common for Roma to react to these unspoken accusations by telling me that they were peaceful people, that they didn’t get involved in wars, that they used to sell cigarettes to both sides. ‘After the war’ is often used to describe the period of fear of political violence in the immediate aftermath of the war, in comparison with more recent severe economic problems. For example, people would say that a greater number of people are emigrating to western Europe in search of a better life (because of lack of work), than the number who fled because of (or after) the war. Discourse surrounding the changes since the war also focus on a demographic change, the influx of rural migrants following the war and the perceived change in urban life. Palo maribe is also a phrase used to explain the reconstitution of the organisation Durmish Aslano in the form of an NGO in 2000, aimed at meeting the postconflict needs of Prizren’s remaining Roma population. The organisation was reformed ‘after the war’ to help Roma, who were now in a more difficult position. The organisation’s emphasis now fell on the areas being funded by international organisations in Kosovo, in particular education, ethnic reconciliation and the promotion of culture. This allowed them to continue their pre-war activities such as folklore, drama and music, reframing them in terms of the liberal-democratic principles of multiculturalism and equality. When my informants did talk about the war itself, they would almost always say there was no war in Prizren. However, there were occasions when people

Durmish Aslano in Prizren 49 discussed the war, and these were generally while walking around, and something about the landscape would dredge up memories. One of these occasions was while I was volunteering at the children’s centre, and went for a lunchbreak. As I left the children’s centre in Nevo Jivdipe/Jeta e Re with two friends, Edon and Erxhan, we walked along a bumpy mud track next to the remains of the railway line, heading towards the town centre. ‘I hate this road,’ Erxhan said to me in English. They liked to practise their English with me. ‘It reminds me of the war,’ Edon responded. I wasn’t sure exactly what the link was to the war – I had been told that the passenger railway had not been used for a long while before the war (though possibly some freight trains had run until 1999) so assumed they were referring to the state of the road. ‘It was horrible during the war,’ Edon said. I was only eight. A bomb landed on my street. They had seen some light which they thought was Serb soldiers, but it turned out to be a Serb family. They were a good family. When I looked out I could just see shattered glass, but it was completely silent. I was not exactly sure what the light referred to, but assume he meant that NATO had seen something that they identified as Serb soldiers, therefore bombed them, but it turned out to be a Serb family in their home. I didn’t want to interrupt him to ask for clarification. ‘Yes, it was horrible,’ added Erxhan. But it was better here than in other places. My cousins were in Peja with my aunt and uncle. They tried to leave for Montenegro on foot, even with children. They were scared. Then I remember my mother seeing on the news that the refugee convoy leaving Peja had been captured by Serb forces, and everyone was massacred. She was crying, we were all crying. Only later we found out that our family was in the first convoy and they had got through. It was the second one that was massacred. So they were ok. But we were worried too, although there wasn’t war in Prizren, we lived in a Serbian area, by the military base, and the street next to us got bombed by NATO so we all ran away to Terzimahalla to stay with Nexhip. We stayed inside there until we thought it was safe. One day my father took me by the hand into the centre. He wanted to see if they had burned down our house. Everything was smashed and shops had been looted. There was glass and trainers and jeans everywhere. I asked my father why we didn’t take things from the shops, but he said ‘you never know who is watching you.’ As we left the track and crossed the rails onto the road that runs alongside it, Edon said: ‘I was standing here when the Serbian army left. They left along this road, although the road has changed now. We couldn’t believe what we were seeing. I had a camera but I was too scared to take pictures.’

50

Roma, Prizren and language

On another occasion, as I was walking to a meeting with a group of colleagues, one of them said: ‘Look, all these houses are new.’ Assuming we were getting into another discussion about house-building, I complimented the houses and carried on. ‘A bomb landed here,’ he added. I was surprised: having been told repeatedly there was no war here, I didn’t realise there was bombing so close to the centre of town and asked if many people were hurt. He answered that they were, unfortunately, and that even a pregnant woman was killed. Even while discussing being bombed, tanks in the streets and walking down the street where this happened, they still stuck with the narrative ‘there wasn’t war in Prizren.’ To a certain extent, this is relative: despite bombing in 1999, and the Serb quarter being burnt down in 2004, Prizren is still tolerant and peaceful when compared, for example, with Gjakova – but it also relates to their image of the town. In such conversations, the war is not discussed in terms of blame, at least not overtly. Victims are talked about as individuals who have suffered, not as a group being victimised. In a sense, in these instances the war was narrated to me as things that happened, and resulted in individuals being harmed. This is very far from the general overarching narratives of the Kosovo war which describe Serbs attacking Albanians, or vice versa, leaving no room not only for people who are neither Serb nor Albanian (with Roma often being subsumed into the Serb side), but also no room for individual suffering. From this perspective, the experiences of Roma in Prizren do not fit into the dominant binary narrative of the war; where they do, they fit in with the Serbian narrative, now unacceptable in public in Prizren. No one ever openly discussed KLA attacks on Roma in my presence. The experiences of Roma immediately following the war are in many ways similar to that of Serb civilians, the main difference being that it was often easier for Serbs to flee to Serbia. There was a range of incidents during and immediately after the conflict in Kosovo where Roma were targeted alongside Serbs in revenge attacks (Human Rights Watch 1999; Memorandum of the European Roma Rights Centre [ERRC] 2005). Some Roma fled temporarily to other parts of the former Yugoslavia, where unsegregated camps meant they were sometimes threatened by Albanian refugees, and there was one recorded incidence of an attempted lynching in a camp in Macedonia (for details on this and other ways in which Roma refugees were adversely affected, see Perić & Demirovski 2000). Many also fled further abroad. The ERRC (2005) collected testimonies of Roma who were attacked by the KLA in the aftermath of the war, finding evidence of extrajudicial killings, torture, imprisonment and rape of Roma. Many of those attacked were accused of collaborating, either in the rape and murder of Albanians during the war, or in supporting the JNA (Yugoslav People’s Army) in some other way (for example being employed in a firm which supplied the army). Some were also accused of looting. Many of those testifying say that they were afraid to report the attacks, and it seems likely that many attacks were never reported, especially in the case of rape, due to the stigma attached to victims of sexual violence. Some of the reports mention people from Terzimahalla being taken by the KLA and tortured. Given the nature of the mahalla, all of its inhabitants would have known

Durmish Aslano in Prizren 51 about these attacks, and they may well have been related to the victims. But I never heard such stories from my informants. As it is implausible that they could have not heard about this, nor forgotten it, it would seem that not talking about these events is a result of trauma, and a continuation of the fear produced by the attacks. Roma are aware that their future in the town, and in Kosovo, rests on silence about these issues, rather than either reconciliation or retribution. The KLA in part legitimised their attacks on Roma on the basis that a Kosovo Roma man, Luan Koka, went to the Rambouillet negotiations between Serb and Albanian forces as part of the delegation sent by the Serbian government. Milošević, in part, legitimised his actions by mobilising narratives of multi-ethnic Yugoslavism against what was treated as Albanian nationalist terrorism, and used minorities to give credence to this claim. Nowadays, monuments to the KLA, pictures celebrating KLA martyrs in schools and public administration offices, and nationalist displays on public holidays, all act as a constant reminder of who is in charge, and Roma who have remained in these areas do not mention the war. If they wish to stay, and to maintain a level of stability in their otherwise precarious existence in Kosovo, Roma must accept the dominant nationalist paradigm. In certain respects, the lack of narratives and commemoration of atrocities amongst Roma is related to the fact that the threat is perceived to be ongoing. This is in some ways similar to what Stewart (2004) describes as lack of commemoration among Roma in Hungary: he argues that Roma don’t overtly commemorate the Holocaust because in a sense it never really ended. It is remembered in their reactions to threats (real or perceived) against them, and the ongoing nature of violence against Roma because they are Roma means that the Holocaust is not perceived as a finite event.13 In the case of the Kosovo conflict, Roma do not mention the atrocities committed by those now in power, treating them as an ongoing threat. In addition, Roma were targeted not just as Roma, nor just as pro-Serb, but also as informers. Luan Koka’s testimony at Rambouillet meant that he was guilty not of committing war crimes, but of talking about war crimes committed against him, and thereby incriminating the KLA. Equally, in the eyes of the international community which had intervened on behalf of the KLA, the KLA had to appear as the lesser evil. In these circumstances, it would have been clear to Roma that silence was the safest option. For those whose traumatic experiences during and after the conflict fit into the dominant narratives of either of the two opposing sides, such narratives and official commemorations may be meaningful; personal memories may thus be open to manipulation and political instrumentalisation.14 In the cleavage between official memory and silenced experiences there lies the potential for political intervention. Official modes of remembering were used throughout Yugoslav times to celebrate those who, like Durmish Aslano, had suffered in the Second World War battle for national liberation, but this ‘hid the much more complex picture of many competing memories that remained present in Yugoslav society as long as the Party controlled the historical discourse’ (Hoepken 1998:225). These blank spots in memory, the things which did not make up the official discourse, are precisely what allow history to be so easily mobilised and manipulated for political

52

Roma, Prizren and language

gain (Hoepken 1998; cf. Temple 1996; Mazower 1995). For Roma in Prizren, and other groups who are trying to navigate a place for themselves in the new Albanian-dominated parts of Kosovo, silence acts as a defence mechanism. Not only do their experiences not fit with the overarching war narratives, but silence shows an allegiance to, and fear of, the current regime.

Conclusion The way people narrate and silence the complex layers of the past has meant that the Romani organisation Durmish Aslano has come to be seen in terms of continuity despite its changes over the years. The KAM was set up during a period of change in Yugoslav ideology and policy. The organisation used the partisan myth to tap into Yugoslav ideology, while at the same time taking advantage of a shift towards the recognition of national and ethnic minorities to forge a space for Romani culture. The organisation was set up at a time of relative prosperity, and of hope. As Yugoslavia was federalised, the Yugoslav sphere of belonging was increasingly envisaged as a collective of distinct nations, nationalities and ethnicities. Roma could be part of a Yugoslav sphere of belonging by virtue of being Roma. Meanwhile the international Romani movement, with strong involvement of Yugoslav Roma, was becoming increasingly organised, meaning that Roma could claim to be part of a wider imagined community (Anderson 1983), a Romani nation in their own right. On a local scale, Roma in Prizren were workers, and therefore part of the urban, cosmopolitan sphere of belonging which differentiated itself from the rural. Their position as part of a Yugoslav collectivity has been lost as a result of the violent destruction of Yugoslavia. This has meant an increasing importance of other spheres of belonging. Identifying with urban cosmopolitanism previously meant incorporation in the Yugoslav state, its services and employment, but Roma are now vulnerable to accusations of collaboration because of their involvement in this sphere. At the same time, the dramatic demographic shifts since the war, the departure of the Serbs (along with some Turks and Roma) and the influx of rural migrants (Albanian and Ashkali) have meant that the urban groups which excluded the peasants are no longer at the top of the hierarchy, and Roma themselves are now excluded. There has thus been a shift from the social exclusion of the Titoist period, to ethnic exclusion of the Milošević period, to new forms of economic exclusion. Roma in Prizren are now living in economically precarious circumstances, without the pre-war prestige of being urban cosmopolitans, or the support of the Yugoslav factories. This kind of economic exclusion makes people more dependent on NGOs, in this case Durmish Aslano, which has increased the importance of the organisation, and made ethnic identification increasingly important.

Notes 1 This is mentioned in Menekshe’s 2004 publication. It is likely given the tense post-war situation that it was heavily self-censored. 2 People have their houses decorated in the spring.

Durmish Aslano in Prizren 53 3 Despite, or perhaps because of, the instrumentalisation of Second World War history by nationalists on all sides during and after the break-up of Yugoslavia, it was rarely discussed during my research. People were aware of the genocide of Roma and Serbs alongside Jews at Jasenovac, but they said that this did not affect ‘their’ Roma in Prizren. 4 For details of the history of schooling in Albanian, see Kostovicova (2002). 5 It has also been argued that the reverse was true during the Second World War, with Italian forces promoting Albanian identity at the expense of Turkish identity in order to bolster anti-Slav feeling (Blumi 2002). 6 Malcolm estimates that between 1956–1966, 246,000 people left in this way (Malcolm 2002:323). 7 These were not solely Albanians, but also Muslim Albanophone people seen as gypsies, i.e. those who would later identify as Ashkali. 8 Cf. Baučić (1971) for details of the effects of migrations. 9 The young musician in this section of the interview is a brother/cousin of Nexhip's brother in law, Taylor Veshal, whose family I lived with for two months. 10 Edvard Kardelj was a leading politician and ideologue in Yugoslavia. 11 For a general history of Yugoslavia, see Lampe (2000). For an in-depth history of Kosovo, see Malcolm (2002). 12 In its reduced form, now consisting only of Serbia and Montenegro. 13 Kapralski (1997), by contrast, finds that there is commemoration of the Holocaust among Roma, and describes this as part of the nation-building process. 14 See Sorabji (2006), however, for memory not being wholly manipulable.

3

Intangible culture and tangible employment after socialism

Introduction NGOs in Kosovo have proliferated in the aftermath of the war, but they are not solely a product of the post-socialist era: many came into being in the socialist period (as with Durmish Aslano) or in the nineties (as with MTS, the Mother Teresa Society). The shift in organisational goals and funding since the conflict has meant these organisations have changed aims and objectives, but the ways primarily western funders target their funds take on local inflections and interpretations. The concepts that underpin funding priorities often mean different things to the funders and the funded. The use of these concepts differs in official and unofficial discourses, and vary between scales, and this relates to the politics of recognition and redistribution. By looking here at a project to protect intangible cultural heritage (ICH) in Kosovo, I seek to understand how western-funded ethnic recognition politics translate into local realities forming and altering social and economic group boundaries. In the first section I describe a Council of Europe project to protect Romani culture, going on in the second section to describe how meanings are made and changed in the course of their transfer from west to east. The emphasis of western funders on ethnic minorities and their culture leads to the ethnicisation of NGOs which will be discussed in the third section, and the fourth section goes on to discuss how this relates to the politics of recognition and redistribution.

Intangible culture The majority of my time volunteering for Durmish Aslano was spent translating and writing funding applications at the radio station. The radio station is part of the NGO and is set in an office consisting of a small studio, the director’s office and an anteroom, as well as a kitchen which isn’t used much in the winter. The office is furnished with broadcasting equipment and several computers by the American embassy, who provided the grant to set up the radio station in 2008. A project proposal I had written for the protection of Romani ICH was accepted in spring 2013. This was part of a joint project of the EU and the Council of Europe

Culture and employment after socialism 55 (CoE) to ‘Support the Promotion of Cultural Diversity,’ within the framework of the obscurely labelled ‘Local Development and Heritage Component.’ At our first meeting with our partners/donors, the CoE office in Prishtina, we arrived late. Nexhip (the director), Taylor (his brother-in-law) and I had driven from Prizren that morning, and while the staff at the CoE seemed to have adopted a ‘western’ attitude to timekeeping, we had not. This created a slight awkwardness, which Nexhip responded to by joking, though he was clearly concerned that appearing unprofessional may affect our chances of receiving funding. The CoE have multi-ethnic staff, and we met with Branislav (Serb, male, in his thirties), Harmonije (Albanian, female, around 40) and Elfete (Albanian, female, early twenties). Branislav did not speak Albanian and Elfete did not speak BCS, so they communicated primarily in English. Since Nexhip did not know English, there was no common language to hold the meeting in, so the two Albanian women left and we held the meeting with Branislav, in BCS. He explained what intangible cultural heritage was, using complex jargon, and stressing that what was important was not artefacts but ovi elementi (BCS: ‘these elements’). Nexhip nodded sagely, but when Branislav left the room he whispered to me, asking if I understood any of it, and explained that he understood about half of what had been said. When Branislav returned, I asked him to explain further, but Nexhip told me loudly not to worry as he understood. As we left the building, he whispered to me that they had mentioned UNESCO, so there could be a lot of money in this project. This was a joint project funded by the European Union and Council of Europe which aims to: Contribute to increased intercultural dialogue, social cohesion and economic development through an integrated and inclusive approach for long term sustainability of cultural and natural heritage in Kosovo. Furthermore, the project aims at facilitating the development of viable heritage planning and management in Kosovo in accordance with European norms and standards with a strong emphasis on community well-being through the active participation of all stakeholders and civil society. (Terms of Reference for project application: private correspondence, February 28, 2014) The project involved researching and documenting ‘intangible cultural heritage,’ with the possibility of a particular ‘element’ of heritage being protected by UNESCO. UNESCO defines ICH as follows: Cultural heritage does not end at monuments and collections of objects. It also includes traditions or living expressions inherited from our ancestors and passed on to our descendants, such as oral traditions, performing arts, social practices, rituals, festive events, knowledge and practices concerning nature and the universe or the knowledge and skills to produce traditional crafts.

56

Roma, Prizren and language While fragile, intangible cultural heritage is an important factor in maintaining cultural diversity in the face of growing globalization. An understanding of the intangible cultural heritage of different communities helps with intercultural dialogue, and encourages mutual respect for other ways of life. (UNESCO 2015)

Rather than arguing for the importance of ‘culture’ for its own sake, both extracts emphasise the importance of culture in building a diverse yet cohesive society. It is notable that UNESCO use globalisation in negative opposition to diversity, despite their existence as a global organisation. While the Terms of Reference also mention ‘economic development,’ they give no hint as to how this can arise from the preservation of intangible cultural heritage. This is part of a move to protect grassroots communities, by shifting the emphasis ‘from the protection of the object to safeguarding the process enabling its production’ (Bortolotto 2007:26). As such, this turns habits into traditions, exoticising them by focusing on preservation and authenticity rather than on creation. This conceptualisation of culture clearly differs from contemporary anthropological ideas of culture in that it refers only to explicit forms of culture, not those which are implicit or taken for granted. The selection of certain practices for protection implies other practices do not qualify as culture. There is a tension between the need to protect certain practices undertaken by disadvantaged or powerless groups, and the process of protecting them, which necessarily involves some kind of transformation and reification. The overt aim of such projects, then, is to define and categorise certain practices as ‘heritage.’ The role of Durmish Aslano in this project was to find and collect this heritage, i.e. ‘these elements.’ The requirements of the project were to hold community meetings in order to get a bottom-up view of how communities viewed their culture, which the CoE could then analyse before deciding whether certain elements qualified as ICH or not. Branislav told us that these must include Roma from the enclaves and Ashkali and Egyptians, despite the official understanding that these groups are distinct. Nexhip was slightly resistant to working with these other groups, especially in the enclaves, but Branislav insisted. The next stage involved writing up ‘Inventories on intangible cultural heritage elements.’ My role in the project was filling these inventories, along with other reports, and running or attending the community meetings. Nexhip had already decided which practice would be put forward as RAE (Roma, Ashkali and Egyptian) intangible cultural heritage at the stage of project application, but we still had to go through the process of conducting the community meetings. The example he had chosen was the ceremony of Bayrakosko dive (Ro: ‘The Flag Day’), held in Terzimahalla. This flag is considered to be sacred. It is green and white, with an inscription in Arabic script (see Introduction). People described it as a passport, given to them as protection as they were metalworkers, and made weapons for the Turks. Once a year, the sheikhs1 bring the flag out and parade it through the mahalla, after which there are prayers and chanting, and people make wishes under the flag. At the same time, the women of the mahalla perform a circle dance,

Culture and employment after socialism 57 and the male musicians play drums and pipes. The next day, the flag is taken to the shrine of Baba Umer in the hills, in an abandoned Albanian hamlet called Lez, around 40 minutes from Prizren by car. Before the war, I was told, they used to walk the whole way, and were offered refreshments along the way by Albanian villagers. After the war, they were too scared to make the journey on foot, and now only a minibus and a few cars full of people attend. The village itself is abandoned. I was not told why, but it is likely that people moved either to cities or abroad during and after the war. We conducted ‘community meetings’ with several different groups (Egyptians in Gjakova, Ashkali and Egyptians in Rahoveci, Roma in the enclaves and Roma in Prizren) to ask explicitly what elements of their culture they considered important, and unique. These interviews were heavily directed by Nexhip’s ideas about what qualifies as culture. In Terzimahalla we interviewed the older men, including the bayraktar (Ro: ‘the flag-keeper’). The meetings, led by Nexhip, were held in the mahalla. Nexhip’s son, as an assistant on the project, took pictures, while I recorded, took notes and asked questions. The men explained the origins of the flag: no one was sure of the dates or exactly where it came from, whether the writing was Persian, Ottoman Turkish or Arabic, or whether it was a reproduction of the original. While these details were of interest for the project, for the men being interviewed, the most important aspect of the flag was that it had been given by the korohaya (Ro, Terzimahalla: ‘Turks’). They were more certain of the recent history of the flag: who had looked after it in living memory, and why it had to change families. There were a variety of reasons given for the flag changing hands, primarily migration, or lack of religious dedication. There was some tension, as it seems to have passed between some of the families present. We interviewed the mahalla women in the house of Nexhip’s sister, Fatmira. This time his daughter was taking the pictures. When I arrived, there was an older woman I had met several times before. She said to me hoću da idem u hajm (BCS: ‘I want to go to seek asylum’). I didn’t understand the use of the German-origin word hajm for ‘asylum,’ so I asked what this meant. She was surprised, asking: ‘How can you work for an organisation, and not know what hajm means?’ My role in the organisation meant I should have been able to understand, and to help her. More women gathered, and they put out their cigarettes when Nexhip arrived, and some put on headscarves when they saw the camera. The discussion grew heated, perhaps more so than the meeting with the men. Our aim was to find out about the women’s traditions on the day: the women were the only ones who dance on Bayrakosko dive, and they knew about the specific clothes and dances. The older women described the importance of traditional dress (emphasising they wore their best dimije),2 and complained that women no longer dressed up. They added that this used to be the one day they could take their headscarves off, whereas now it is the other way around. The conversation turned again to the contentious issue of whose family had looked after the flag, and why it changed hands. Nexhip turned the conversation back to dimije, saying that there could be a lot of money in this project if UNESCO recognised the day, and that they could all get new dimije.

58

Roma, Prizren and language

In the early stages of the project, there had been some debate about the date of the ceremony. I was told that originally it had to be a specific date, but then in the times when they had jobs in factories (that is, in Yugoslav times) they moved it to a Saturday. In Yugoslav times such rituals were tolerated, but were not officially celebrated in the way that folklore, for example, was. Nowadays, the day seemed to vary, and one Thursday when I arrived at work I was told to go to the mahalla to see the flag. Nexhip would come separately with the employees of the CoE, who were driving from Prishtina. People in the mahalla (and those from other mahallas) heard on the day by word of mouth, and probably social media, when the ceremony would be held, and as very few had formal employment there was no need for advanced planning. For the CoE and their camera crew coming from Prishtina, this was a different matter: they wanted to protect the traditional format of the ceremony, but also needed to fit it into their busy schedules in advance. It was tricky for them to protect it without a specified day, and a standardised, recordable format. Having compiled inventories based on the interviews, we needed to record the ceremony. I went separately, with Moni, and arrived before the CoE representatives, so we started recording the event. When we arrived, the street was full of people. We visited a house at the bottom of the street where an elderly lady held a flag over me (not the flag – the significance of this flag remains obscure) and gave me water to drink. We continued through throngs of people taking pictures on smartphones to where the flag was brought out. I was given strict instructions by the other NGO staff on where to stand, and what to record. The flag was brought out by the sheikhs, chanting, while the women performed a circle dance. The flag was carried to the top of the mahalla, where people made wishes and drank water under it. This was all accompanied by a young and very enthusiastic drummer (who I was told was the son of the best drummer in the mahalla), and pipe-players. At some point the camera crew from the Council of Europe arrived, late, and as the flag was paraded back down the street they invited us for coffee nearby. For the CoE employees running the project, the recognition and documentation of ICH was crucial: it needed to be framed before it could be codified. This is what Kirshenblatt-Gimblett (2004) calls the ‘metacultural,’ rather than cultural, outcomes of ICH. If all cultural knowledge is transmitted and intangible, then ICH can be seen more as a way of recompensing minority, indigenous and nonEuropean peoples for their lack of UN officially recognised ‘culture.’3 Culture in this sense is cast as the traditional attribute of minorities; it contrasts strongly with notions of culture as civilisation, contrasted to nature, which will be discussed in Chapter 4. Cultural heritage also roots people through a sense of past and continuity (Kuutma 2009). However, the necessity of deciding whether this qualified as heritage or not was up to the CoE employees, and their decision dictated whether it would receive any further financial support. Roma from the mahalla were less interested in it being defined as heritage – the ritual was set to continue with or without funding. But they were far more concerned about the question of whether this was tradition or religion. Many Roma, especially educated young men, are becoming increasingly interested in what they

Culture and employment after socialism 59 see as true Islam, in contrast to Balkan Islamic practices (see further Chapter 4). They argued that, as a religious ceremony, this, and other celebrations such as Ederlezi,4 are haram. Older residents disputed this, saying it could not be haram as it was tradition, rather than religion; there were clearly different issues at stake for the different groups involved. In terms of the project, the last section on the form was crucial, as it required the ‘interest of the practitioners/participants to extend the traditional element.’ If younger people were increasingly abandoning it for religious reasons, there was no justification for its protection. These religious objections to the ceremony were therefore ignored in conversations with funders. The existence of ICH projects, and the insistence of funders on minority culture, has a particular salience in post-conflict Kosovo. While the project was not specifically described as minority protection or an attempt at ethnic reconciliation, it was aimed at the protection of cultural diversity. The vulnerability of minorities (including the vulnerability of their culture) provided the impetus for the project: if their ICH was not at risk, it would not need protecting. Although it did not receive governmental funds, this type of project also worked in favour of the government, as it implied the acceptance of minorities. This parallels Adams’ (2013) findings in Uzbekistan, where ICH projects aim to protect local culture for the sake of global humanity, while also being absorbed by the national government: the Tajik minority history of the songs deemed to be ICH was not taken into account, but was rather taken as Uzbekistani, and therefore also formed part of a nationalist narrative within Uzbekistan. This left a gap between the actual nationalist appropriation of cultural heritage, and the official rhetorical proclamations of multiculturalism. In some respects, the predominantly Albanian authorities in Kosovo are in a similar position: it is important for them to show protection of minorities to international observers, while maintaining exclusionary nationalist control of power internally. This leads to a situation where minorities have extensive formal rights while not always being able to exercise these rights, and the communities of expertise which seek to protect minorities become tied up in these power relations (Kuutma 2009). Heritage, when perceived as distinct and as possessed by a particular group with a particular genealogy, is always embedded in inequalities of power between the group and others. For Roma in Kosovo, these projects at least nominally serve to protect ‘their’ culture as distinct from the culture of others, as distinct from ‘tangible’ heritage, and from the other more tangible but less elevated needs of Roma. For the CoE on the other hand, this is also a site of scholarly knowledge production: the purported aim of documenting this type of heritage is to acknowledge, record and valorise local and minority knowledge, but it is the members of the CoE who are in charge of deciding what knowledge constitutes ICH, not the ‘community.’ Their use of language, often incomprehensible to ordinary people, made it tricky to explain the purpose of the project, and at the same time, Roma knowledge and rituals had to fit with their forms, requirements and schedules. Their official aim of protecting grassroots elements of intangible cultural heritage was couched in terms so alien to most people’s idea of what culture is and should be that they in effect had a monopoly on deciding what could be included and excluded. Nexhip was very

60

Roma, Prizren and language

keen to buy into this expertise, and to fulfil the necessary criteria, so ordinary people were chosen and consulted on this basis. The increasing number of Roma who believe this day to be haram was therefore overlooked for the purpose of the project, as were people’s other, more material, needs. There was no follow-up to the project, and the inventories were never taken to UNESCO. But the project was still considered a success. Both Durmish Aslano and the CoE had fulfilled what was required of them, and everyone was satisfied.

Rights as discourses For the NGO, my presence as a foreigner, and someone who could write project proposals, was not just of practical value, but also something to be proud of, lending a kind of cultural capital to the organisation. In addition to writing project proposals, I translated or interpreted between Romani, BCS, English and sometimes Albanian. In this context, trying to explain the concept of ‘intangible cultural heritage’ to people was tricky. Official translations were meaningless for many. On one occasion Osman, the director of Iniciativa 6, asked me if it meant duhovna kultura (BCS: ‘spiritual culture’). The official BCS translation of ‘intangible culture’ was nematerijalna kultura. Duhovna kultura does also relate to intangible, or immaterial culture, but has been used for much longer, and generally relates to art, music or poetry, rather than the practices and processes that formed ICH. Intangible cultural heritage was not the only project which meant different things to the different interested parties. One German-funded women’s project, which aimed at preventing early marriage among Roma girls across eastern Europe, pointed to the central role of parents in educating their daughters. As part of this project, one young Roma woman in Prizren used an educational radio broadcast to suggest that parents should be responsible for their daughters, keeping them at home rather than allowing them to be involved in extracurricular activities; this, she suggested, would prevent them from marrying early. Roma (and other) girls who marry early in Prizren are seen as being not properly brought up or controlled by their parents, since they have evidently engaged in activities which involve meeting boys rather than staying at home with their family. The young broadcaster’s logic here was straightforward: preventing early marriage means restricting the (already very limited) freedom of girls. This was of course not the intention of the funders, but a result of the uncritical transference and translation of discourses. Crucially, this is not simply a mistranslation, but rather the insertion of concepts into the new context of a complex social system; in this new context, people interpret the new concepts in light of their existing ideas and assumptions. As funds are transferred from west to east via international funding bodies, they are accompanied by policies, ideas and representations (Sampson 1996). These transfers often ignore the local context of what went before, such as socialist welfare policies, and can thus have highly unpredictable and often unintended results. Donors often blame this on lack of implementation, rather than local reinterpretation (Stubbs 2002, 2009; Deacon & Stubbs 2007; Stubbs & Maglajlić 2012; van

Culture and employment after socialism 61 Baar 2012). While local NGOs do not just passively receive donors’ concepts, it is still necessary for them to master their symbolic resources (Sampson 1996). However, during my time at Durmish Aslano, the receipt of funds seemed at least as dependent on social capital, namely personal connections with potential funders, as on well-written proposals (cf. Mikuš 2015). Successful projects therefore required both a mastery of the language of the funders, and a good relationship with the funders. These personal connections were not seen as corruption. Rather, a good relationship between funders and organisations was considered vital, and once this had been built it was in both their interests to prolong it. The necessities of knowing the requisite language and of knowing people working in funding bodies often preclude grassroots or radical initiatives, privileging individualistic or conservative organisations (Jeffrey 2007; cf. Helbig 2010). The aims of NGOs often relate more to the preservation of the NGO itself, rather than to their overt goals of protecting rights and promoting culture. Organisations are, on paper, expected to cover their own overheads and to ensure that project goals continue beyond the phase of implementation. This leads to a paradoxical situation, whereby the organisation needs to show that the project has permanently solved the problems it set out to solve, but at the same time must show the need for further projects of the same type.5 Working within this paradox (based on good relations between the funders and the funded) requires an acknowledgement of the flexibility of certain terms and criteria. I found that the terms used by NGOs and funders often showed an elasticity of meaning in translation. ‘Civil society,’ for example, was commonly used solely to refer to NGOs, sometimes to the extent that several NGOs would be called ‘civil societies.’ This stems in part from confusion over the word society: as in English, the Romani, BCS and Albanian terms can refer both to a society in the sense of an organisation of some kind, and society more generally. Durmish Aslano had shifted from being a Cultural Artistic Society to an NGO, but could still be referred to as a society. It is easy to see how this could slip in to calling it a civil society. But I would suggest it also stems from a fundamental reconfiguration of the idea of civil society, alongside notions of society and state. If western notions of civil society posit it as an organisation of citizens outside state institutions, such as grassroots initiatives or non-state activism, in Prizren, ‘civil society’ is understood solely as NGOs. This way of understanding civil society is far from unique to Prizren (Jeffrey 2007; Goldstein 2013; Deacon & Stubbs 2007). In this case, the shift is the result of both the linguistic and conceptual translation of the term. While many in the west expected that the end of the Cold War would mark the flowering of civil society in post-socialist countries, what many people actually experienced was a sense of chaos and loss of privileges (Brandstädter 2007). The importation of western understandings of civil society, which treat the state and society as distinct entities reaffirm the state/society divide, contrary to anthropological theories which emphasise that the state itself only comes into being through social processes (Hann 1996; cf. Vetta 2009; Ferguson 2006). The state is never a tangible entity, but in contemporary Kosovo processes of state formation are particularly contingent and incomplete. Even before

62

Roma, Prizren and language

independence, throughout the nineties, a kind of Albanian ‘parallel state’ existed in opposition to the Milošević administration. In contrast to other post-socialist countries where there was a radical transformation from an unequal though (partially) functioning welfare state to a neoliberal state relying heavily on NGOs for social welfare, Kosovo actually had a long period of services being provided by quasi-state organs, much more like NGOs. One of these quasi-state organs was the MTS (Mother Teresa Society). After the war, the UN administration set up humanitarian aid, and eventually a new healthcare system through MTS. Rather than distinguishing between state and society, the question here was which state was perceived to be legitimate, and by whom. Cocozzelli describes how ‘(i)n the space of less than 10 years it [MTS] had gone from being a fiercely oppositional and highly politicized dissident organization to a massive local NGO with operational partnerships with some of the world’s largest international NGOs and IGOs’ (Cocozzelli 2009:54). The series of locally run initiatives which in the nineties fell under the name of ‘the Republic of Kosova’ were smoothly transformed into humanitarian NGOs under the auspices of the UN in the aftermath of the conflict. Indeed, foreign governments and international organisations have played a decisive role in setting up the state of Kosovo, writing its constitution, and, given the continued military presence in Kosovo, they still play a central role in the course state formation takes. Hence, the influence of western governments goes far beyond the provision of small grants; a ‘compartmentalized, neoliberal, approach to social policy’ (Cocozzelli 2007:218) was employed in Kosovo after the conflict, controlled by world powers in conjunction with particular new local elites. The embassies of the USA and the EU countries, who have substantial influence and power, have rolled back the state and supported privatisation. The state itself has sparse funds to provide basic services, which are thus provided by NGOs. NGOs receive funds from the government of Kosovo, as well as foreign governments, the EU and international organisations, so the distinction between governmental and non-governmental is, in many cases, merely nominal. ‘Civil society,’ then, is not distinct from government, but embedded in processes of transnational governmentality, which reproduce power relations (Ferguson & Gupta 2002). However, understandings of civil society are malleable. In Kosovo today, civil society slips between its original, more general meaning of non-state organisational forms to specific NGOs. At the same time, it loses its semantic opposition to the state when transferred into a situation where the boundaries of the state are understood very differently, owing to the long history of state welfare provision. International involvement in Kosovo through project funding requires the transfer of concepts and policies, and with this the translation of terms. This involves reincorporating them into new settings, with pre-existing worldviews and understandings. The formal translations of ‘citizen,’ ‘community,’ and ‘human rights,’ very familiar to those working in NGOs, take on local inflections when they are used informally, in the specific context of Roma NGOs in Prizren, and as they replace or reshape local ideas of the state and citizenship. They are translated into standardised forms, without the acknowledgement of the contrasting unofficial

Culture and employment after socialism 63 speech practices which make up the diglossia described in Chapter 5. At Durmish Aslano, discussions about rights formed part of project applications, interviews, ‘seminars’ and so on, but were less used in informal conversation, except perhaps jokingly or disdainfully. For example, when Nexhip’s father gave money to Nexhip’s sons but not his daughters, Nexhip joked with them all that the girls had rights too. In this context, this notion was amusingly nonsensical: girls are rarely allowed to go out, and when they do, they are accompanied by a male relative who pays for them. The idea that women and children should have rights overtly clashed with patriarchal organisation of families. However, those working at NGOs still needed to be aware of the discourse of rights, and to know how and when to use it. As part of a project to ‘monitor the implementation of the strategy for RAE’ in Prizren municipality,6 Durmish Aslano conducted a series of visits and radio interviews with ‘governmental and non-governmental officials’ and ‘community members.’ The interviews were conducted mostly in Albanian, with some in Romani and one or two in BCS. Summaries were written in Romani, which were then translated into BCS, Albanian and English. I was in charge of the English translations. There were certain topics I struggled to translate more than others, one of which was ‘rights.’ Rather than discussing rights as something abstract, which should belong to every human being, interviewees talked about their rights to concrete things like welfare payments, and a proper sewerage system. On the whole, these were things which were previously provided (or promised) by the state, and now had to be accessed through an NGO. Here rights are not universal, but rather discourses which can be mobilised in a particular context (Cowan et al. 2001). At the same time, they can become transformed and enmeshed in the logic of this discourse, and from this perspective, we can see that the discourse of rights is not just co-opted from the source context, but also reinterpreted in the new target context. As such, the discourse of rights, as promoted by international sponsors, was locally understood in terms closer to the state entitlements of the socialist era (cf. Anderson 1996). In Kosovo, rights have come to be understood as the entitlements and services lacking in the post-socialist era. Here, like the understanding of underage marriage mentioned above, the notion of rights has not simply been transferred, nor has it been rejected, but has rather taken on new meanings and logics in its new context. This predicament of translation exposes a gap between the rights the state purports to provide and people’s lived experience. Despite the legal provision of rights for all, the state is dominated by the ethnic majority, and rights are unequally distributed (Krasniqi 2013). This is in part based on differing ideas of justice; for western funders, a just distribution of rights tends to be grounded in the notion of equality, while for Kosovo Albanians it is more commonly understood in terms of guilt and suffering. If Albanians were the greatest victims, they therefore deserve the greatest rights (Beha 2014). In theory, rights are granted according to western protocols, yet no provision is made for their implementation. While the Serbian authorities have an inherent suspicion of western ideas and funding (Mikuš 2013), Kosovo’s government

64

Roma, Prizren and language

rather aims to please powers abroad, while maintaining hold on an ethnonationalist state at home. In both cases however, western concepts translated by funders do not simply replace local concepts, but rather they co-exist uneasily. The discourse of rights is potent at a central, formal level, but this discourse does not correlate with informal local practices, and the dissonance between the two reinforces diglossic speech practices (Abercrombie forthcoming). The presence of the ‘international community’ means that informal, private or critical ideas of rights are rarely expressed overtly. Rather, the polysemous nature of terms allows for dissonance in meaning, an implicit non-congruence between the de facto and de jure state of affairs. The relocation and translation of terms and policies without regard for local inflections and understandings has led to a discrepancy between official western democratic values, and the unofficial reality of minorities in Kosovo. In Kosovo, then, the process of targeted funding has resulted in a dissonance between official, western-influenced discourse and unofficial ideas of how things ‘really’ are. There is slippage between the de jure use of terms and concepts and the more private, everyday understandings, resulting in the awkwardness I described when translating these usages back into English.

Work, kin and community The ICH project budget included two paid positions, a coordinator and an assistant. Nexhip wanted me to be the assistant. I objected on the basis that I was there as a volunteer, and that the other members of staff needed the wage more. He was very insistent, arguing that we were only awarded the project because of me, and that it would look bad if I was not paid for it. As it wasn’t acceptable for me to argue with him, I told him that as I received a scholarship, I was not allowed to receive a wage as well. He laughed and said he wasn’t allowed to receive more than one wage either, but it is fine as long as you do not call it a wage, and allocate it as volunteer expenses. Alternatively, he suggested paying it to my husband.7 Eventually I persuaded him that I would call the CoE personally and explain that I would work on the project without payment, and that another member of staff would also help and receive the assistant’s wage. I thought I was behaving ethically towards my colleagues, but the other members of staff were annoyed. They said that if I really wanted to help them, I should have accepted it and given it to them. Later, there was an argument because Saverd was not receiving a wage for his work while Bayram, who was unable to work for health reasons, was still given the assistant’s wage. For most of my time there, four people were considered permanent staff in the organisation. They worked both for Durmish Aslano and Radio Romano Avazo, depending on what grants we received. In practice, this meant that they turned up every day and obeyed Nexhip, not that they had a contract or received regular payment. The NGO survived on a range of small grants, and knowledge of how to juggle expenses so that there was generally enough for electricity and internet, and a small wage for some of the employees. The rent for the office and radio station had not been paid in several years due to some legal wrangling over

Culture and employment after socialism 65 ownership of the building. Staff were often expected to share or redistribute their wages, and there was a clear pecking order of who should earn what, an order which was obviously gendered. On a project which didn’t include a wage for Vera, Nexhip asked Moni and Saverd to each give a portion of their wages to her, while he would also give a lesser amount. Moni objected: he thought this was not his responsibility, especially since his salary was already very small and his wife was expecting another child. They then complained that Nexhip often diverted part of their salary towards the upkeep of the office. Wages caused disputes, as did the level of secrecy over who got what, with a general assumption being that those at the top were taking too much. ‘Volunteers’ sometimes helped on projects, and they treated their role as a form of employment as they received per diems, food and travel costs that are actually more like a kind of wage, especially in the absence of more permanent work. In fact, those connected to the organisation often didn’t see a clear distinction between volunteers and employees: this was more a matter of pay scale. People who attended regularly were more likely to be given the position of ‘coordinator’ on a project, which meant better pay. Wages on projects were generally very low, and local NGOs were treated with suspicion by funding organisations, whose anti-corruption policies dictated that project workers should receive no more than an average wage in Kosovo, which is scarcely enough to live off. While those in charge of organisations sometimes put themselves on several projects to increase their wage (as well as prioritising their own family for employment opportunities to increase their household income), those further down in the organisation could not do this. Similarly, budget constraints made it preferable to employ more people for a lower wage. The precarity of project work thus affected people at the radio unequally (cf. Baker 2012). There was a reserve of people vaguely involved with the NGO who would be called upon when necessary for volunteer roles. Volunteer payments would often be used as a way around rules about employees only receiving one wage, meaning they were able to coordinate several projects alongside other roles. While many were disadvantaged by this precarity, Nexhip’s position lent him the symbolic capital to navigate it to his advantage. He was very proud of his role as director, schoolteacher and coordinator of several projects, but the precarity of all these roles meant he could also declare himself as unemployed when necessary, for example when applying for bursaries for his children. Such tactics are necessitated by the context of extremely high unemployment. ASK (Kosovo Agency of Statistics) estimates the unemployment rate at 32.9% of the working age population, finding it to be much higher (around 60%) among young people. ASK also estimates that 62.4% of the population are ‘inactive,’ meaning they have no formal employment, and are not seeking work, so are not classed as unemployed. This figure is much higher for women (81.9%), who are usually expected to care for children and ill, disabled or elderly relatives (ASK 2016). The precarity and informality of work of course means that figures are unreliable, but it is clear that RAE are disproportionately affected by unemployment. During my stay, I met a few Roma who had formal jobs as cleaners, as well

66

Roma, Prizren and language

as one judge and an Ashkali nurse. Some still worked as blacksmiths (though not in Terzimahalla), others in garages, one drove a taxi and a few people still sold cigarettes. Selling cigarettes was apparently very common among Roma during the war, but is less so now. Some told me they had previously had professional jobs (a teacher, and a health worker for disabilities), but they had both lost their jobs, seemingly for collaborating with the Yugoslav state in the nineties. I was also told that before the war Roma would travel to Turkey and even Syria to buy cheap goods and gold to bring home and sell. The exact period was never specified, but I didn’t meet any Roma who still sold goods like this. The majority of the work people engaged in was either low-paid or precarious, so people often supplemented their work by painting houses in the spring, playing music in the summer, and most people relied heavily on remittances. Remaining involved in the NGO despite the lack of adequate remuneration meant people were more likely to receive payment in the future. But there were also other advantages. These include the social aspect of being at the NGO, the feeling of having a purpose and a focus and the symbolic capital of being attached to projects. The notion of work was still valued: even without a salary people liked to be able to feel, and to tell others, that they were employed. My colleagues regularly told me it was better before the war because they had factories: ‘everyone’ had work. They put unemployment today down to privatism8 and nepotism, both concepts that tend to evoke nostalgia for a better time before the war. Such nostalgia (see also Chapter 6) focuses on perceived easy access to labour before the war. As Petrović states with regards to a factory in Jagodina, ‘(n)ostalgia in the worker narratives is not a longing for the past but a way of making sense of the present through symbolically retaining lost values, status, stories, and memories’ (Petrović 2010:148). The use of informal networks, and the negative perception of them as ‘nepotism,’ is likely to have existed before the war. However, today pre-war employment is perceived nostalgically, as free from nepotism. The present is thus associated with job scarcity in privatised firms, in comparison with the factory work of the socialist era. Roma complain that the people who now run institutions and own companies are Albanian, and as they seek, and indeed may be obliged, to help their relatives, they tend to employ other Albanians. Before the war there were several large factories producing textiles and shoes and, I was often told, there were jobs for ‘everyone.’ The idea that ‘everyone’ had work cannot be taken at face value: in addition to being shaped by nostalgia, it is likely to refer only to the adult, male friends and family of the person I was speaking to, when compared with now. Unemployment existed in Yugoslav times as well, and in the later years rural Albanians in particular suffered extreme deprivation. But for Terzimahalla Roma, the perception that ‘everyone’ had work is juxtaposed to a present where ‘everything’ has been privatised and bought by Albanians, who only employ other Albanians.9 At first glance this parallels complaints about privatisation in other post-socialist settings (Arandarenko & Golicin 2007; Kalb 2009; Henig 2012; Gal & Kligman 2012; Živković 2015; Grdešić 2015), yet here it is distinctly ethnicised.

Culture and employment after socialism 67 Despite this impression that Albanians had privileged access to employment, in other contexts my informants also discussed ‘nepotism’ among Roma. On a smaller scale, they resented people privileging their own family both at Durmish Aslano and the Roma department of RTK (Radio Television Kosovo). On the other hand, the use of personal connections to find work was also treated as a moral obligation: Vera’s husband had managed to get a job for their son at the shopping centre where he worked, as he had a good reputation there. Her son in turn was obliged to work diligently for his father’s sake. Finding work through informal connections was often perceived externally as ‘nepotism,’ and internally as an obligation. However, since there are fewer Roma in a position to find employment for others in the first place, the reliance on kinship networks to find work reproduces ethnic inequalities in employment. The dependence on kin networks, and the small number of ethnically mixed families10 means that it is common for firms (and NGOs) to be mono-ethnic. However, while people commonly complain about the lack of job opportunities, they do not cast this as an ethnically discriminatory practice. Verdery (1996) observed that during a shortage of hair dye in Romania, Hungarian hairdressers privileged their own, Hungarian, kin, leaving Romanian women with greying hair. Here too, dependency on kinship networks in a context of scarce employment leads to a seeming ethnic bias. While de Swaan (1995) describes how widening circles of identification mean that family ties have gradually evolved into wider ties in the form of national identity, here it is in fact the narrowest ties which are most important, creating a situation which can be interpreted as an ethnic bias. Albanians, for the most part, are not discriminating against non-Albanians per se; rather, those who are in a position to help others find employment assist those who they have closest ties with, who will in turn be obliged to be a reliable worker. In the discourse of both governmental and non-governmental organisations, citizen is used to refer to the people of Kosovo regardless of their ethnicity, as opposed to community, which refers to an ethnic group. The term community can be slippery, and while people sometimes talk of ‘the Albanian community,’ this term is more often reserved for ethnic minorities, also called ‘non-majority communities.’ When people talk about the need to integrate communities into society, they discursively construct communities as something separate or outside society. Elsewhere, the term community is used to mean a small, simple group defined by certain traditional cultural traits, in contrast to the more complex social formations of modernity (Cohen 1985). In Kosovo it has come to refer not to any group seen to have these traits; rather, such qualities are attributed solely to ethnic minority groups, in contrast to the Albanian majority. Across the world, the term is invariably used positively, to suggest autonomy, homogeneity and solidarity, and it is deployed by states as well as NGOs to suggest the right to representation (Creed 2006); in the case of Kosovo, ‘community’ rights are constitutionally enshrined. The notion of community, ascribed to an ethnic group, and with certain formal rights attached to it, shapes the group in a particular way. Within the local discourse of community rights, those in a less powerful position need

68

Roma, Prizren and language

the ‘community’ in order to exercise their ‘rights,’ but they can only do this via organisations such as NGOs. While the funding of communities means that people can access certain benefits on the basis of their ethnicity, the political system also reinforces ethnic distinction. This ethnic bias is reinforced even more strongly by minority NGOs, and hiring is based on pre-existing forms of solidarity (cf. Pétric 2005). Reliance on NGOs to access services and employment, coupled with the mono-ethnic makeup of most NGOs, leads to the process of ‘ethnicization’ and ‘minoritization’ (Helbig 2010; cf. van Baar 2012). In a context in which state services have been withdrawn under the supervision of international organisations, the direction of funding for NGOs towards ethnic groups rather than citizens creates an unintentional bias in service provision. This does not, however, lead to the privileging of whole ethnic groups, but rather an elite. Similarly, on the larger scale, Kosovo Albanians are privileged, but many ordinary Albanians still live in extreme poverty. Despite funders’ emphasis on ‘grassroots’ organisations, the system of grant allocations to NGOs privileges various types of pre-existing elites and reinforces social hierarchies. The move to create ethnically based NGOs can be seen as part of a wider shift from the politics of redistribution to that of recognition: in recent decades, as the state has been rolled back across the globe, identity politics, drawing on the language of culture, heritage and ethnic communities, has displaced the politics of redistribution, aimed at tackling socio-economic inequality (Fraser 2000). To prioritise culture over economic inequality precludes viewing cultural groups within socio-economic structures, making it impossible to address disparities in wealth. As such, working-class solidarity is displaced by a language of culture (Edwards et al. 2012). When class is obscured in this way, it can also lead to the reproduction of class formations, as NGO workers tend not to come from the lowest classes, and tend not to employ others from the lowest classes (Mikuš 2015). To return to the project with which this chapter opened, the ICH funders, while seeking the ‘community’s’ perspective on their culture, overlooked their main concern: unemployment. In this sense, it could be argued that by funding Romani ICH, the CoE was displacing the economic problems of Roma, and their lack of access to employment. However, I would suggest that the situation is more complex. Displacement may explain what is happening at the level of official discourse, but the existence of projects aimed at Romani culture, coupled with the dependence on kin networks for work, meant that some sort of employment is provided to Roma. NGOs do not write project applications with the explicit goal of providing employment for their families, and would be highly unlikely to receive funding for this. However, in practice, such projects do provide employment for Roma. Although Bayrakosko dive was not, ultimately, recognised by UNESCO, the ICH project was deemed a success by my colleagues because it had served its purpose by giving the staff a period of paid employment, however precarious. This form of redistribution is clearly insufficient and highly unequal, but it is not entirely absent. Rather, it is instrumentalised by those with the necessary social and cultural capital to

Culture and employment after socialism 69 facilitate access to employment. Recognition at an official level translates into redistribution at an unofficial level.

Unequal redistribution The redistribution that resulted from the projects described above is far from a continuation of a socialist ethos of redistribution, resisting the discourse of recognition promoted by western funders. Rather, there was an uneasy tension between the two. This tension, I would suggest, frames an ongoing process of class formation in the post-socialist era, which contrasts sharply with memories of the socialist era. In one of the Romani-language classes, Nexhip used the term albanyuni qhib (Ro: ‘Albanian language’). He explained to me that the more common word is gaunani, and then told me that before the war you only heard Albanian on Wednesdays. He saw I was confused, and explained that gaunani is from the word gav, meaning ‘village.’ The informal Romani word for an Albanian, gauno, also translates as ‘peasant’ or ‘villager.’ Wednesday is market day in Prizren, so the city is full of people coming from the villages to sell vegetables, cheese and so on. While it may have been an exaggeration that there were no Albanian speakers except on market days, it seems that Albanian was much less widely spoken in public in Prizren before the war; Turkish and BCS were the main languages, Albanian being considered rural, backward and uneducated. The term gauno, which previously referred to a peasant, has shifted from being a primarily social category to an ethnic one.11 Class, like ethnicity, is not a predefined entity with certain characteristics, but rather a social process shaped by its historical trajectory (Kalb 2014). At the same time, class is always defined and delineated from a certain perspective, and these varying perspectives are intrinsically unequal (Carrier 2015). The powerless lack the possibility of defining themselves, an issue further complicated in the case of Roma by the easy slippage between socio-economic and ethnic definitions of Roma (Tiefenbacher 2013).12 In Prizren there has been a shift in the power to define groups, which has impacted on the way Roma define themselves with regards to others. A central factor in explaining current class configurations in post-Yugoslav space is the legacy of self-management socialism.13 While Ladányi and Szelényi (2003) argue that Roma who had been working class in socialist Hungary became an underclass in the post-socialist era, the situation in Yugoslavia was more complex. Sardelić (2013) argues that the exclusion of Roma from the ranks of the working classes in socialist Yugoslavia meant that in the post-socialist era they became excluded. This does not apply to Prizren: Roma from Terzimahalla did work as blacksmiths and musicians, but they also worked in factories. Urban factory workers had access to a number of services, and most lived in reasonable conditions, in contrast to independent peasants. Regarding class and social stratification in Yugoslav times, Ströhle (2016) argues that while Yugoslavia was forming a ‘new class’ of bureaucrats from the urban Partisan elite (as defined

70

Roma, Prizren and language

by Đilas14), an underclass of poor, excluded and primarily Albanian peasants was also being formed. By working as gastarbajteri (BCS, from German: ‘guest workers’), the position of this underclass gradually improved in terms of material goods, health and education. However, this improvement was interrupted by the crisis in the early eighties (see Chapter 2). This group was disproportionately affected, becoming even more economically deprived. In Yugoslav times, many Roma in Prizren, as urban workers, were thus on the right side of the dual social structure which privileged urban workers over the independent peasants (Deacon & Stubbs 2007; Stubbs 2009). This is reflected in their contemporary attitude to rural Albanians and Albanian-speaking Ashkali who still live in extreme poverty. In Prizren, those who now identify as Ashkali, including those who have recently moved to the city, are seen as ‘gypsy’ peasants. In this case the social divide (peasant/urban worker) has been reconfigured as an ethnic divide (Ashkali/Roma). For Roma in Prizren, the post-war period brought a significant drop in living standards. Factories had provided their workers with access to services such as health insurance, social support and other benefits, and factories were situated where employment was needed (Stubbs 2002; Stubbs & Maglajlić 2012). State provision of factory work and workplace provision of services are no longer present in much of the former Yugoslavia. Indeed, they are sorely missed: in a study of textile workers in post-Yugoslav states, Bonfiglioli found that ‘(t)he social rights connected to industrial employment during socialism are often compared to the lack of social rights experienced in post-socialist times’ (Bonfiglioli 2015:62). The legacy of self-management can therefore be seen as heightened expectations, resulting in a sense of loss of the privileges of former urban workers, in contrast to the independent peasants. The social inequalities created by self-management socialism have affected current intersubjective understandings of class and ethnicity, but this is further complicated by the legacy of the conflict, which has reconfigured ethnic hierarchies. Broadly speaking, the most disadvantaged group in Yugoslav times was the independent peasants, many of whom readily radicalised and joined the KLA (Ströhle 2016). Thaçi and other KLA leaders now form the ruling party, the PDK. In other post-socialist countries, the most disadvantaged socio-economic group was often further disadvantaged by the shift to capitalism, while in Kosovo the conflict meant the leaders of this group came out on top. This left them in a position to distribute patronage, perceived by others as ‘nepotism.’ Urban Albanians, Turks and Roma who were formerly part of a local elite are now experiencing a reversal of these roles. This is not to suggest that these groups have been affected equally: urban Albanians were also persecuted in the nineties and went on to form the passive resistance (see Chapter 2). Serbs, Turks and Roma, many of whom continued to work and live in the state system in the nineties, faced discrimination after the war. Furthermore, while Thaçi and other KLA leaders may have gained vast amounts of wealth and power, much of the Albanian peasantry remains in extreme poverty. But the crucial point here is that pre-war configurations of class and culture as perceived by Prizrenites have been overturned, leaving a feeling of

Culture and employment after socialism 71 fear, but also resentment about the new state of affairs and the concomitant loss of privileges. With this historical background in mind, the redistribution of employment which occurs through the NGO can be seen as a process of class differentiation on a micro-scale. This differentiation does not result in the formation of distinct groups, but is processual: economic and social patterns of differentiation are being re-formed and delineated. As former urban workers, some Roma are carving out a new niche for themselves within the limits of the ethnically based NGOs that make up a substantial part of contemporary Kosovo’s employment sector.

Conclusion Western-funded projects such as the ICH project ostensibly protect culture, while at the same time providing employment for certain people. By seeing the problems of Roma in Kosovo purely in terms of ethnicity, culture and rights, international donors have, at a formal level, displaced the discourse of class, but they have also, if inadvertently, provided Roma with jobs. This has led to a divergence between the official and unofficial situations: the priorities of international donors shape official discourses of rights and culture through projects, while the memory of state employment and social rights, and the fluidity of boundaries between governmental and non-governmental bodies, has meant that NGOs are unofficially seen primarily as employers. The social and ethnic categorisations which have shifted in salience over time are reconstituted under the influence of western funders, leading to a semantic dissonance between the way terms are used within the hierarchy of the funders and the funded. This manifests itself in a marked distinction between saying and doing. The understanding and expectation of services and employment which occurs at an informal level can be understood as the legacy of self-management socialism. A post-Cold War lens (cf. Verdery 2002; Chari & Verdery 2009) allows us to capture not just the legacy of socialism, but also the peculiarities of the Yugoslav system as neither fully socialist nor fully capitalist. Equally, Todorova (2015) treats the legacy of socialism as eastern Europe’s position in the capitalist world economy; this perspective reveals the profound effect of privatisation and deindustrialisation in Kosovo on living standards, and also the unevenness of this effect: the previously relatively privileged group of urban workers has lost out in relation to a section of the previously most deprived group, the Albanian peasants. While many of the latter have not profited from the KLA’s victory, those who have profited have acquired the power to represent others. In this sense, hierarchies have been reversed, and class formations have been dramatically reshaped. In the Bosnian context, Baker (2012) discusses the importance of the postsocialist lens to examine the destruction of the socio-economic order, alongside the post-conflict lens; she argues that neither is sufficient on its own to understand how the region has been incorporated into the global economy of development. For Roma in Prizren, the combination of these two lenses can be used to explain the dissonance between the official discourses of minority culture and rights, and

72

Roma, Prizren and language

unofficial economic relations. The self-management system instituted an unequal rural/urban divide, a division achieved primarily through the medium of employment and its associated entitlements. The Milošević era and the conflict meant the increased ethnicisation of this divide, while the NATO intervention and victory of the KLA marked a complete reversal of internal power relations, with those previously experiencing severe oppression taking control of the country. Within this context the translation of policy and funding priorities is reincorporated with different meanings into local discursive practices, which also widens the gap between these practices and actual economic relations between donors, NGOs and employees. This creates a semantic dissonance between the official and the unofficial, shaping diglossia in the language and creating the imperative for an official Romani language.

Notes 1 Members of a Sufi religious group in Prizren. 2 Loose trousers worn by Muslim women in the Balkans. 3 There is an interesting parallel here with ideas of traditional ecological knowledge, which is protected by funders whose understanding of knowledge differs from anthropological understandings of it (Vermonden 2013). 4 Ederlezi, or Đurđevdan, is a festival celebrated by Roma of various religions in the Balkans and elsewhere. 5 The same is true on a larger scale for their funders who, in turn, have to prove to their sources of funding that, on the one hand, they are spending money efficiently but, on the other, that they also continually need more. 6 This is a strategy developed by KFOS to integrate Roma, Ashkali and Egyptians. On a local level it takes the form of an Action Plan which the municipality is supposed to implement. 7 I was not married to my partner at the time but they often referred to him as my husband. 8 I am not sure why they use the term privatism when speaking in BCS and Romani rather than privatizacija. It is possible that it is from the Albanian privatizimi, but more likely that they see it as an –ism, an ideology rather than a process. 9 See Woodward (1995) for a discussion of unemployment at this time. 10 Although some families consider themselves to be ethnically mixed, in most cases where a woman marries into a family of a different ethnicity people do not talk about this as mixed; there is rather an implicit assumption that by joining her husband’s family the bride becomes one of them. 11 This use of gaunani seems to be specific to Prizren. In the Serb enclave of Gračanica, for example, Albanians are called geralo, ‘skinheads,’ probably referring to a traditional Albanian hairstyle. 12 This discussion connects with a wider debate about whether Roma should be defined primarily in terms of class or culture (cf. Stewart 2002; Ladányi & Szelényi 2003; Tremlett 2009). 13 See Chapter 6 for a discussion of the way urban elites position themselves. 14 Milovan Đilas was a leading member of the Communist Party who later became a dissident.

Part 2

Purity, mixture and representation

4

O Romano Teatro and gypsy theatricality

Introduction In her ethnography of a Romani Theatre in Moscow, Lemon (2000) describes how the representation of Roma in Russia has been primarily in the hands of others. Even when the performers are Roma themselves, they perform a stereotype of Roma for a Russian audience, in Russian language. In Prizren, a stereotype is also performed. However, it is performed to a primarily Romani audience, and in Romani language. As such, while these plays in a sense legitimise Romani nationhood by providing a national theatre in the national language, they also reinforce certain stereotypes of ‘gypsiness.’1 Analysis of the meaning of the Romani theatre group to my informants helps show how my informants understand the relationship between their ethnicity and performance, and how their offstage uses of postures and gestures translate into onstage characters. Movements and postures are associated with gypsy stereotypes, which are intersectionally racialised and gendered. In turn, these are located within ideas of Balkanism and east/west dichotomies. The shifting perspective of the tragicomic mode relates to the worldview of those involved in the plays. The theatre group in Prizren has existed since 1989, and plays are presented fairly regularly in Romani, using, and producing, the standard variety. They are funded through small grants either from the Ministry of Culture, Youth and Sport, or from international embassies. Plays are written in Romani by local people, primarily Nexhip, and they usually have performances on April 8 for International Roma Day, and at various other festivals. The actors are staff at Durmish Aslano, and their friends and relatives. I was not able to attend a live performance. Due to suspicions among some of the women in the mahalla about my role in the NGO (discussed in Chapter 1), one of the people involved had told his wife (and by extension other women in the mahalla) that I had left Prizren. When I became aware of this I felt very uncomfortable about the situation, but decided to avoid events where these women were likely to be present. I did, however, attend multiple rehearsals and meetings in the run up to performances, and was involved in ongoing discussions about the dramas while working at the NGO. I was also given recordings of many of the plays to watch later. As such this chapter does not focus on the performances themselves, but rather the discussions, debates and

76

Purity, mixture and representation

contestations surrounding the performances, the meaning of the drama group for those involved, and their audiences, and the image of themselves the group were trying to present. Building on discussions of social hierarchy and language use in previous chapters, this chapter looks at how people think Romani language should be used, this time in performances. A perceived nature/culture divide can be understood in terms of, on the one hand the bodily, and on the other language, as will be discussed in the first section. The second section goes on to discuss this division in terms of representation; people make choices about how they want themselves to be represented, but these choices are always embedded in larger scale hierarchies, in this case discourses of Balkanism. The third section looks at how the nature/culture divide relates to ideas of gender, patriarchy and religion, and the fourth section discusses how these systems of representation become disorienting in times of large-scale shift. The prestige attached to the Romani standard is contrasted to the image of the uncultured gypsy. The disjuncture between these two images shows vacillations in ideas of self and other, as well as how the social hierarchies discussed in previous chapters are structured and reproduced.

Acting like ‘gypsies’ After the main part of my fieldwork was over in December 2014, I went home for Christmas, and then made several shorter trips to conduct interviews and tie up loose ends. In the months before Christmas, so many Roma (and others) had migrated to the west to seek asylum that my departure went unheeded. Upon my return, Nexhip said that they were rehearsing a new play for April 8. His elder son, Fatmir, had written it, with his help. They had decided to write a play about the exodus of Roma. Nexhip explained: It’s called O Sundal isi Amaro Them,2 ‘The world is our land.’ It starts with Roma in a pit, and the Germans digging and burying them, you know like in the Second World War. Then the Roma spring back up. It’s a comedy. The point is whatever they do to us we’ll keep coming back. They can’t get rid of us! The idea of describing a play about migration caused by poverty, and about the Holocaust, as a comedy seemed incongruous. In 2005 they had performed another play about migrating to Germany, called Prizren via Essen. The title refers to the city of Essen, where there is an area jokingly called Terzimahalla 2, housing a large group of Roma from Terzimahalla. Like many of the plays, it is intended as a cautionary tale, to warn people of the dangers of emigrating. Three brothers emigrate to Essen to seek their fortune. One of the brothers becomes a heroin addict. Another, who left a pregnant wife behind, becomes gay.3 His child, without a father, is bullied, and his wife struggles to make ends meet. The brothers’ parents, left on their own, struggle to look after themselves in their old age. When the heroin addict dies of an overdose, the gay one returns to his wife and child,

O Romano Teatro and gypsy theatricality 77 along with the third brother, who has somehow avoided the perils of the west. People described this play to me as being about the problems of migration. When I watched the DVD of the play later I was surprised that they had publicly performed a play with what seemed like such an anti-western message, though this negative association of the west with drugs and homosexuality was widespread off stage. Several other plays deal with a variety of what the theatre group and their funders perceived to be ‘Roma issues.’ They tend to feature girls being ‘stolen’ (marrying against their parents’ wishes), in-laws falling out, divisions among Roma and migration. They feature ‘typical’ older Roma women, dressed in headscarves, dimije and thick armless cardigans, shouting, squabbling, gossiping and plotting. They dote on their sons and persecute their daughters-in-law. In short: the plays represent many of the stereotypes about Roma held both by outsiders and by Roma themselves. This begs the question: why do the Roma who write, organise, direct and act in these plays, and who clearly do not want Roma to be stereotyped in this way, choose to present plays which fit these stereotypes so neatly? In the plays, the stereotype entails dress, personality and the use of gestures. Playful overacting of gypsy women bustling about in headscarves, aprons and dimije, hiking up their dimije, contrasts with a masculine image swaying between drunken affection and aggression. Elbowing and manhandling people around the stage is more common than verbal direction. It is no coincidence that this representation of gypsiness also makes entertaining and engaging drama: the bodily, visual aspects, more than the dialogue, comprise the gypsy stereotype. While my informants, including Nexhip, were keen to separate themselves from this exoticised representation, especially in front of me, they knew how to mobilise this representation for entertainment, and to foreground the flaws in this kind of behaviour to other Roma. Certain forms of language are considered to index gypsiness particularly strongly, such as the phrases ishti devle and zhimi devle, meaning ‘I swear to God.’ When performed on stage, the phrases are evocative, performed with exaggerated body language and emotive intonation. The first part of the phrase is drawn out, with one hand held on the heart, and the stress falls on the second part with one or both hands raised. They are commonly used by the older female gypsy characters. While I was there, the use of expressive language on stage also caused problems off stage. Vera, for example, worried about her daughter swearing on stage, telling me that it would not matter if an older woman, such as Fatka, swore on stage, but that young girls should not. Indeed, parents often worried about girls performing. Onstage persons are reduced to bodies: though acceptable for men or older women, it is problematic for unmarried girls to perform, or even have a photo taken. Even Nexhip’s daughter was no longer allowed to perform once she got engaged. Vera felt that as her daughter’s Romani was limited, she would not understand the swearing, and in addition her father would not allow it. Nexhip gave way to her father’s wishes in this case. But he had complained to me earlier that often audience members did not like the swearing in his plays, especially

78

Purity, mixture and representation

with children present: ‘They don’t like swearing on stage, but when women in the mahalla are brawling and swearing in the street, do they close the door so the children can’t hear, or do they all go out to watch with everyone else?’ The answer was clearly the latter. Nexhip’s representations of daily life on stage are an attempt to represent the daily life he saw off stage. People’s ways of behaving and talking off stage are often theatrical, involving a level of performance. I found that Fatka managed to subvert many of the expectations of Romani women (see Introduction), but this was acceptable as she was divorced, so her reputation was not as fragile as that of unmarried girls, and she did not have to obey a husband and in-laws. She took full advantage of this freedom. Once at the radio, someone had been given some wine, and they were sharing it around. The men drank quickly, Fatka and I had small glasses, while Vera had none. They were speaking in BCS for my benefit. When someone explained something shocking to Fatka, she crossed herself. This gesture is common in Serbia: it may be used in religious contexts, such as entering a church, but is commonly used theatrically as a sign of shock or mock despair at the state of things. Fatka made the gesture to make light of the situation. Saverd reprimanded her: Saverd: Zašto se krstiš be? Ti se Muslimanka. Šta? Misliš da si Srpkinja? Ti si ciganka! Fatka: Da. Ali ja imam pedigre. Saverd: Why are you crossing yourself? You’re a Muslim. What? Do you think you’re a Serb or something? You’re a gypsy! Fatka: Yes, but I’ve got pedigree. In this case a theatrical gesture which may have in the past been acceptable for Roma, was now seen as Serb, and thus became a boundary-marker. For Saverd, the gesture registered difference, non-Romaniness. Fatka managed to co-opt this non-Romaniness to jokingly describe herself as having pedigree, being superior to other Roma. The use of gesture which Saverd took to be Serb and Orthodox brought into question her representation of herself as Roma and Muslim, which she contested by suggesting her self-ascribed social superiority to other Roma. Her unusual position afforded her the opportunity to play with the use of gesture, exploiting the way it can mark both ethnicity and social difference. The theatricality of Roma is more than a stereotype: Roma men and boys are theatrical both on and off stage. With the notable exception of Fatka, I did not come across Romani women engaging in play acting and imitation (this probably would have been seen as disrespectful). But people’s descriptions of Romani women would nonetheless often emphasise this physical theatricality. Roma who otherwise would have avoided negative terms like ciganka (BCS: ‘gypsy woman’) would use it to describe certain types of behaviour. When I asked about the use of the term, they would explain by mimicking or impersonating this ‘gypsy’ behaviour, or by analogy with animals. One woman complained to me about other

O Romano Teatro and gypsy theatricality 79 Roma women shouting, flailing their arms and legs, behaving like chickens. To her this was a particularly ‘gypsy’ way to behave. The postures and behaviours of gypsiness are not only projected onto Roma. Early in my fieldwork an Albanian journalist decided to write an article about me, as an English girl who had come to Kosovo to learn Romani. He took two pictures to go with the article. In one I was standing awkwardly outside the radio, next to Nexhip, trying to smile for the camera. The other I didn’t notice being taken, and I was standing with my hands on my hips, laughing. A Roma friend told me he preferred the latter, saying I looked like ‘one of our people.’ This kind of posture to him made me look Roma. In contrast, on another occasion Fatka explained to me that Roma women were expected to be respectful, saying ‘ciganke are expected to be like this,’ bowing her head and crossing her arms before her waist. The posture suggests subservience. Young Roma women and girls are generally praised for being respectful and peaceful, meaning that they are quiet, don’t answer back and serve their male relatives or in-laws. While these two postures – bowed, subservient, and brazen, noisy, flailing – seem polar opposites, they differ according to the seniority of a woman. Unmarried women and young brides are expected to bow their heads, while senior women whose position is established are more unabashed. These ways of being are both described as physical, and are described (in BCS) as ciganski rather than romski.4 In these examples, gypsiness is something physical, bodily and gestural. The term ‘Roma’ is not used to refer to these kinds of behaviour, or bodily movements, and Romaniness does not index these things. Here, postures and ways of being suggest the physicality of gypsiness. Elements perceived as more ciganski can be seen as lying more on the edge of language. They include posture and gesture, and also bits of language which are more expressive, more connotative than denotative, like swearing or exclamations. Such expressive, connotative elements of language are often ambiguous and peripheral, not quite language, and not quite in any one language. Ishti devle (Ro: ‘I swear to God’) is used off stage, as are the equivalent phrases valla, or vallaj bilaj, accompanied by the same gesture. These latter two expressive phrases can be used when speaking in any language in Prizren. There are a variety of other similar phrases: mashalla is used to compliment children on growing up, to congratulate someone on achieving something or to call women in the street (as with a wolf whistle), and translates as something like ‘Allah has willed it.’ Ishalla is used to express hope, like ‘God willing.’ Allahalla expresses exasperation or contempt, and translates as ‘Oh God.’ These Islamic phrases are used across the Balkans (and beyond), bridging language boundaries. In addition to Islamic phrases, there are exclamations like kuku (usually kuku be or kuku bre), an exclamation used to express dismay or sorrow, and hajde meaning ‘come on.’ Kuku and hajde are usually accompanied by, and often replaced by, gestures. Unlike the Romani ishti devla, these terms can be used in any of the languages spoken, and in a sense are between, or outside, specific languages. While these elements on the edge of language – the less linguistic parts of communication – transgress language boundaries, they are highly contextual and difficult to translate beyond the region. Agar (1991) describes the untranslatable,

80 Purity, mixture and representation culturally specific parts of language as ‘rich points.’ Such contextual, untranslatable elements are expressive, connotative terms that may not have currency in another social setting. Similarly, to be accepted in a new setting, competency in gesture and bodily comportment can be more important than knowledge of formal language rules (Herzfeld 2009). Both gestures themselves and gesture-like bits of language are more likely to be cruxes of contact where codeswitching occurs (Matras 1998a). They cross both the boundary between languages, and that of language itself. There is a tension here: these elements of communication are somehow between languages, less linguistic and therefore less part of a bounded Romani language, but they are also more Romani, in that they are associated with a stereotype of Roma, which is used to represent them on stage. Ishti devla may be used in the plays in place of valla as it is more Romani, in that it is restricted to Romani. But at the same time, it is accompanied by gestures and behaviours which run across languages. This expressive use of gestures, and communication on the boundaries of language, is both more gypsy, in the sense that it fulfils a stereotype of Roma being theatrical, bodily, less like formal language and less Romani, in that it does not work within the boundaries of language and ethnicity. These elements of communication which act between and across languages are not indicative of cosmopolitanism and mixture (to be discussed further in Chapter 6), but rather of informality and bodiliness. This suggests a hierarchy of ethnic characteristics from the formal and cultured, down to informal and bodily, rather than a multiculturalist image of a series of parallel but equal ethnicities. These gestures and postures act simultaneously as ethnic markers and markers of social hierarchy. The very bodiliness of standing and gesturing like a gypsy are also what mark these as socially inferior behaviours. Parts of communication which are more closely associated with nature than culture, those which are more bodily, are seen as the domain of the lower classes (Bourdieu 1991:66–89). Roma are often seen as closer to nature, with a poverty of language (Stewart 1999). People’s ways of talking and bodily comportment evoke hierarchical differences: Bourdieu (2004) describes how male bodily comportment, dress and ways of interacting with potential brides mark the peasant in 1950s southern France as a peasant.5 Here, the physicality of gesticulation, acting, performing – and, in this case, mixing languages without respecting borders – can be seen as part of acting the gypsy, a member of a dominated group represented in terms of a bodily lack of culture. This lower kind of gypsiness cannot be performed through the use of language alone: it requires gestures and movements, alongside exaggerated forms of language. The representation of the gypsy body is thus racialised. Race, understood as a cluster of characteristics projected onto a group and understood as biological (Wade 2012a, 2012b), is in this case indexed by the characteristics of bodiliness itself, projected onto the body and enacted. From the perspective of the Other, the non-racialised subject, the racialised object is more physical, and closer to nature. These features are further enhanced when the body is gendered, with the image of the gypsy woman being represented as the most natural, physical, uncultured.

O Romano Teatro and gypsy theatricality 81 Gender will be further discussed in the third section; suffice it to say here that as the body is used to perform race and gender, it is racialised and gendered. Gypsiness, then, is something which is enacted and performed through both race and gender. The distinction in BCS between romski and ciganski, or in Albanian between gjuha rome and maxhupisht, can therefore be understood as a distinction between language and behaviour, a manner of talking and being respectively. Institutional recognition of Romani language is historically tied to the denomination gjuha rome/romski, while the informal or derogatory terms ciganski/maxhupisht encompass a far less structured understanding of Romani ways of communicating, and a range of behaviours attached to the image of the gypsy. Although those involved in the Romani-language plays speak and diffuse standard Romani, they enact gypsiness, and the two overlap and merge. Behaving like a gypsy includes speaking Romani (in a certain way) and speaking Romani is often accompanied by behaving like a gypsy. By behaving gypsy-like, the characters, women in particular, fulfil many of the stereotypes that non-Roma have about Roma. Language, especially in its institutionalised form, is understood as being on the culture side of the nature/culture divide. Gypsiness is represented as nature, bodiliness, gestures and noise, in contrast to cultured Romani language. The gypsy image, as made in Prizren, is part of broader schemes of representation of the gypsy, and of the Balkans, which are primarily externally constructed. The next section will thus discuss the reactions of Roma in Prizren to the performance of this image, locating them within hierarchical ideas of the Balkans and the west.

Roma in the Balkans While ruminating on my reasons for coming to Prizren, Nexhip announced he didn’t think I was there to learn Romani, but to report on them. Then he turned to me and told me not to make this report a negative one, ‘like Shutka or Kusturica.’ Shutka refers to The Shutka Book of Records, a 2005 documentary film by Serbian director Aleksandar Manić about Šuto Orizeri, a Romani area on the outskirts of Skopje (see Gorski 2005). The film depicts Roma as musical, mystical, exotic and entertaining. Roma I have spoken to about the film complain that the filmmakers chose to only represent eccentrics and stereotypes, and that, while it may be amusing, it doesn’t accurately represent Roma. Kusturica is a renowned Bosnian director who is known for his films about Roma, using magic realism to exoticise Balkan Roma. His films portray Roma as either poor or criminally rich, deceitful, exuberant and mystical. This exoticising tendency makes Kusturica unpopular among Roma I have spoken to. More than one person asked me to help find funding to make a documentary about Roma that showed them as cultured and educated, stating a desire to counterbalance this image. The image of the gypsy presented in Shutka and Kusturica films, and resented by my informants, is deeply embedded in Balkanist hierarchies (Todorova 1994, 1997). In contrast to Said’s (1978) Orientalism, which describes a projected opposition between the west and the Orient, Todorova’s (1997) notion of Balkanism

82

Purity, mixture and representation

focuses on the projection of ambiguity. Orientalist theory cannot be uncritically applied to the Balkans, as the two terms have a very different historical trajectory (Fleming 2000); the Balkans, unlike the Orient, are not projected as the opposite of Europe, but rather are represented as the Other within, an ambiguous and peripheral part of Europe which is never quite European enough. A central focus of work on Balkanism is the way that Balkanist distinctions reproduce themselves. The Balkanist schema of representation is scalar and nesting, whereby the east is compared negatively to the west, the Balkans to eastern Europe, Yugoslavia to the Balkans, the south-east Yugoslav republics to the northwest ones, peasants to urbanites and so on (Bakić-Hayden 1995; Goldsworthy 2002; Razsa & Lindstrom 2004; Volčič 2005).6 Within this schema, the image of the gypsy is often the ‘most’ Balkan. The image of the gypsy is also a form of ‘projective identification’ (Homer 2006), whereby non-Roma project their hopes, fears and fantasises onto the other. This phenomenon is described in the case of Serbia by van der Port (1998) as a way of coping with the wild and the violent in Serbian society. Živković (2000) finds that: Gypsies are exemplary pariahs, outcasts, the very bottom of the internal hierarchies throughout Europe. If the Balkans are perceived as pariah and at the bottom of European hierarchies, one can expect to see the demon of metonymic misrepresentation declare Balkanites to be the Gypsies of Europe. This logic is recursively reproduced on ever smaller scales. (Živković 2000:170) The way the image of the exotic, Balkan gypsy is performed to a non-Roma audience in film and music has been critiqued (Lemon 2000; Silverman 2011; Haliliuc 2015b). Representation is controlled and produced in response to a perceived audience. Dobreva (2007) argues that Gatlif7 focuses on oppression of Roma by mainstream society, in contrast to Kusturica’s portrayal of Roma as detached from the mainstream society. This is a result of the two filmmakers having different projected audiences: Gatlif’s films are aimed solely at a western audience, while Kusturica’s are also aimed at a non-Roma Balkan audience. Petrović’s8 Skupljači Perja is aimed at a similar audience to Kusturica’s but here, rather than the image of the gypsy being surreal, a form of escapism, the film uses ethnographic realism to imply that the film is based in reality (Mladenova 2016). This implication of authenticity breaks down the onstage/offstage distinction by allowing the gypsy image to diffuse into attitudes towards Roma (Iordanova 2003; cf. O’Keeffe 2013). The performance of the gypsy image thus acts to reinforce stereotypes held by non-Roma, leading to a situation where the performance of music (Beissinger 2001) or lifestyle (Iordanova 2008) can be enjoyed and glorified, while actual Roma are still feared or despised. Representations of Roma therefore vary according to author and audience, but they are rarely controlled by Roma themselves. Rather, they are controlled by non-Roma, whose representation in turn affects the way Roma are actually treated.

O Romano Teatro and gypsy theatricality 83 The representation of Roma to the more western Other is part of a broader hierarchical system of representation, whereby the image of the gypsy, and indeed of the Balkans, is presented for the western gaze (Szeman 2009). In its more extreme form this leads to appropriation, with performative forms being fully co-opted by the other (cf. Silverman 2015), and ultimately marketised by western intervention (Ditchev 2002). The representational dichotomy of east and west is thus also structured within the inequality of power to represent. Representation does not just concern impotent images and stereotypes, but is both deeply embedded in power relations, and has effects in the real world.9 The Romani plays in Prizren in many respects perform this racialised Balkan gypsy, epitomised by Kusturica and Shutka, but with one crucial difference: the audience. The plays I am describing are in Romani, and made for a Romani audience. Certainly, the audience often includes a number of internationals, who, ironically, would not understand the image of the west being portrayed as they do not know Romani, as well as local officials and non-Roma NGO elites. But the plays are primarily aimed at local Roma. The observers, those on the other side of the proscenium, are not engaging in an act of differentiating themselves as a western, modern, cultured other, but rather they are local Roma, many of them friends and family of the performers. For them the plays are neither an exotic fantasy, nor cathartic realism. Rather, they are very close to their own experiences. However, this contradicts their insistence on not being seen in these terms: they choose to perform plays which represent Roma in a way they explicitly told me they didn’t want to be represented. In a sense, this is because the image is internalised. The Balkan image has been co-opted and reincorporated in various ways in the region, and is reproduced in the form of ‘nesting orientalisms’ (Bakić-Hayden 1995; cf. Volčič 2005). Gal (1991) describes this as ‘the east/west opposition […] that can be reproduced repeatedly within each side of the dichotomy, fissioning or subdividing ever-smaller social units’ (Gal 1991:446), as modernist hyperdifferentiation, which forms part of a reaction to western homogenising discourse (Kiossev 2002). Green (2005) captures the complexity of this reproduction, or nesting, with the metaphor of fractals, which combine images of premodern ethnic hatreds with postmodern hybridity and resistance to categorisation: ‘essentialist Balkan politicians are being cast as premodern as a result of their apparently promoting an earlier version of modernity’ (Green 2005:155). Fractality, reproduction with slight alteration on ever smaller scales, shows how the Balkans are posited as constantly marginal and inbetween. Since the Balkans are positioned ambiguously, not outside Europe, but not quite European, they are perceived as less far along the trajectory from barbarism to civilisation, and thus in need of development to catch up (Fleming 2000). This ambiguity allows for representational oscillation between the Other within and the absolute Other to be replicated fractally. The structure of differentiation from the absolute Other is reproduced, but the identification with the Other as an internal Other means it is not reproduced fully, but rather splits off as a fractal. This fractal replication of hierarchies can be applied not just to the position of Roma in the Balkans, but more generally to gender and race hierarchies. The

84

Purity, mixture and representation

way gender roles are naturalised, in contrast to ideas about religion, and the way these roles and ideas structure representational hierarchies, will be untangled in the next section.

Natural patriarchy Prizren via Essen, the play mentioned in the first section, warns against the perils of the west, namely homosexuality and drug use, which are portrayed as unnatural and dangerous. While casual homophobia is ever-present, younger people in Prizren also discuss and theorise their reasons for disagreeing with homosexuality more explicitly, in general arguing the following: It has nothing to do with religion, it’s an illness. Parents deserve a daughter-in-law to serve them coffee, a boy can’t be a bride. We can still be European without changing our culture.10 This overt rejection of homosexuality is embedded in a discourse of nature and culture, whereby patriarchal, heterosexual relations are constructed as natural. I do not mean patriarchal in the looser sense of sexist or misogynistic, but rather in the sense of a strict hierarchy of age and gender to be maintained. Both men and women treat this order as natural, or biological. Patriarchal relations are not seen as a religious obligation: in fact, this purported natural order is more often contrasted to religion, which is felt to be part of culture. As part of the cultural sphere, religion is associated with civilisational and educational level, indicative of social hierarchy. The insistence that homosexuality is an incursion on the private, natural domain of the household is overtly kept distinct from religious ideas about homosexuality. Equally, people are aware that the linking of homophobia to Islam is used to denigrate Islam (or religion in general) as backward, barbaric and non-western, and want to avoid this. ‘Religion’ is thus discursively part of the cultured, male sphere. At the same time, gender relations and patriarchal family relations are understood as natural. The naturalisation of gendered differences and the understanding of hierarchically organised patriarchal families also leaves no space for anything other than heterosexuality, and gendered expectations epitomised by virility in men and virginity before marriage in women. Activities and characteristics associated with one gender act as boundary markers. Vera once explained to me, laughing, that if no one else was home one of her daughters would sit in the cold, while the other would light a fire. This was amusing because lighting a fire is a male job, making the daughter who lit the fire ‘like a man.’ Upsetting gender norms slightly is amusing as it is felt to be unnatural. The idea of a woman being involved in other supposedly male activities, especially involving the public sphere, also upsets the ‘natural’ order of things (cf. Helms 2010). Within this hierarchical arrangement, homosexuality also upsets the ‘natural’ order; hence, ‘a boy can’t be a bride.’ Patriarchal family organisation is perceived as natural, especially compared to the unnatural organisation of families in the west, yet it is also legitimised as being ‘our culture,’ as opposed to that of Europe, leading to the argument that ‘we can still be European without changing our culture.’ Here the term culture does

O Romano Teatro and gypsy theatricality 85 not refer to civilisational level, as above, but is rather used to justify and reify difference. In an atmosphere where being openly gay is linked to the west (cf. Mikuš 2011), there is a fear that homosexuality may upset the natural organisation of households: young brides enter the household at the very bottom of the hierarchy and are expected to do the majority of the housework, which men would not ‘naturally’ be able to do; the natural structure of the household would be upset. Young women’s freedom is carefully restricted by their relatives, but this is not, generally, resented. Rather, this is seen as providing women with very necessary safety and protection, and thus creating dependency. Subordinacy within the family is therefore strengthened by the subordinacy of the family to those outside: women’s restriction to the private sphere is inescapable, as it affords them protection in the face of larger scale subordinacies, such as that of Roma to non-Roma. While the freedom of young women is especially restricted, even more senior women are restricted to certain spheres. Unlike many married women, Vera was allowed to work by her husband, but this did not stop other women in the mahalla criticising her, and her work was often seen as inappropriate and a cause for gossip. It is also rare for younger men to question older men. I was thus surprised when Erxhan, young and unmarried, chastised Nexhip for visiting the shrine at Lez (see Chapter 3). This was made slightly more acceptable by the fact that Erxhan was from a well-respected family, but nonetheless it was unusual. I was even more startled when, rather than reprimanding him, Nexhip apologised and asked me to tell Erxhan that he only went as part of the project, and he hadn’t actually entered the shrine but stayed outside. A similar disruption of the ‘natural’ hierarchy occurred when Nexhip challenged the elderly men about the flag and its ability to grant wishes. A considerable shift in ideas of religion within the space of a generation means that religious practices and ideas are one of the few areas where contestation of superiors is possible, and where there is some leeway for younger men to assert their knowledge. Both these instances of contestation were based on new ideas about Islam which are spreading among younger people. Worshipping shrines (or flags), along with a range of other practices, is increasingly seen as backward, rural, Balkan and even pagan, in contrast to modern Islam, which is seen to involve stricter worship, and clearer rules. Many older men for example think it is acceptable for Muslim men to drink alcohol, except maybe not on Fridays, and certainly not during Ramadan, while young men increasingly see it as totally haram. Attitudes towards fasting during Ramadan have also changed, with an increased insistence on an either/or choice, rather than the older in-between practices such as fasting for the first and last days, and maybe one in the middle. Modernity has come to be associated with new religious ideas rather than secular ideas.11 Certain European modernist ideas have refracted in the Balkans within religion, rather than in opposition to it. Nowadays in Prizren, demystification does not mean condemning Islam to the dark ages, abandoning a Balkan past in favour of a European future perceived as modern and secular, but rather condemning Balkan practices such as worshipping shrines to the past, in favour of the educated, enlightened Islamic teachings from the east. These are perceived to be more Islamic, and truer

86

Purity, mixture and representation

Islam. This cultured, modern Islam is treated as distinct both from ‘backward’ practices of the past, and from the ‘natural’ realm of gender and family relations. In contrast to other parts of the Balkans where religiosity, and especially Islam, are often seen in terms of Balkanness and barbarism (Stefansson 2007; cf. Bringa 1995), many of my informants celebrate new forms of Islam as enlightened and civilised, in contrast to certain Balkan Islamic practices. However, as in Bosnia, there is a tension between this idea of ‘true’ Islam, in contrast to local practices, and ‘foreign’ Islam (Maček 2007), epitomised by certain ‘Arab’ styles of dress (beards and shalwar for men, hijab or niqab for women), which many see as extreme or superficial. Women covering their faces is deemed particularly suspicious, and there are rumours that women are paid to do this. Whether religious practices are seen as true (in contrast to Balkan) or foreign (in contrast to local) depends on the position of the people involved, and the discursive context. The idea of religion as being part of a male, cultured sphere, in contrast to the female, natural sphere, relates to notions of public/private, which, while discursively constructed as an opposition, are also ambiguous, being, again, both relative and scalar (Gal & Kligman 2012). The public sphere is the preserve of men, and the more privileged they are, the more public they can be considered to be. Private space is simultaneously part of, and distinct from, public, encompassed by it but also opposed and subordinated to it. In representational terms, fathers are heads of the family, the private sphere, but by virtue of being the patriarch, they also belong to the public sphere, in opposition to the private sphere. The discursive construction of the private as both part of, and opposed to, the public parallels the way the Balkans are both part of, and opposed to, Europe. The Balkans are simultaneously a ‘private’ part of Europe, and distinct from it. The distinction is reproduced fractally. Bourdieu (1990) explains the series of intersecting divisions, such as male/female, inside/outside, upper/lower, that form ideas of public and private in the Kabyle house; he argues that this is not a simple division, but rather a shifting perspective. In a similar manner, dichotomies are reproduced in the form of nesting systems of othering. This again can be seen as fractal: it is replicated, the same but different, in the relationship of mother-in-law to daughter-in-law, and of mother to child. A second degree of separation (such as the relationship between father and child, or grandparent and child) allows for a level of benevolence which the mother, charged with the everyday control of her subordinates, does not have. This phenomenon of benevolence towards those more than one rung down the hierarchy parallels, on a smaller scale, what Spivak calls ‘white men saving brown women from brown men’ (Spivak 1988:297; cf. Abu-Lughod 2002). This attitude of white men to the women of racial and ethnic minorities is an issue addressed by black feminists, and more recently Romani activists, through theories of intersectionality which explore the compounding of subordination through racism and sexism.12 Black feminists have found problems with advocating for their ‘race’ while also trying to advocate for their gender (e.g. Crenshaw 1989; Collins 1998; Swaby 2014). Treating these intersectional identities as two distinct realms of disadvantage is problematic, in that intersectional forms of disadvantage

O Romano Teatro and gypsy theatricality 87 cannot be seen as solely race or gender (Crenshaw 1989). Separating the forms of intersectional disadvantage makes it possible only to help those disadvantaged in one way. A black man with the same profile as a white counterpart may claim racial discrimination if he is not treated equally, but this becomes much harder for someone who is both black and female, let alone working class, gay or disabled. It is not simply that black women experience both racism and sexism, but rather that they often experience a more intense form of sexism than white women, and a more intense form of racism than black men, precisely because of their intersectional position. The same is true for Roma women: Oprea (2005) argues that sexist practices among Roma are either ignored in the interests of cultural autonomy, or portrayed as primitive, and that racism ‘contributes to the oppression of Romani women through the rigidification of Romani practices’ (Oprea 2005:141). Equally, they have less access to the resources necessary to change or escape their situation. Regarding domestic violence, Oprea (2004) argues that authorities turn a blind eye to such practices through fear of looking racist. Such problems are compounded by fear of the state (racism), lack of a place to go (poverty) and their position as women (sexism). They cannot access support networks and services in a way that domestic violence victims from other social and ethnic groups may be able to, making their situation especially intractable. An increasing number of Romani activists have started to look at the possibilities for a specifically Romani feminism, as distinct from a generalised feminism which does not necessarily include this intersectional perspective.13 Racism, like sexism, ‘relies heavily on embodiment, visibility, nature-culture hybridizations and control of reproduction’ (Wade 2010:50). The critiques of the replications of Balkanist dichotomies discussed in the previous section parallel black feminist critiques of the structures of race and gender. Collins argues that ‘hierarchies of gender, age and sexuality that exist within different racial groups (whose alleged family ties lead to a commonality of interest) mirrors the hierarchy characterizing relationships among groups’ (Collins 1998:66). As such, maintaining racial solidarity can replicate other hierarchies. These hierarchical divisions are replicated on different scales, leading to multiple and intersectional subordinations of black women. This subordination is legitimised through a nature/culture divide, whereby men are deemed cultured when compared to women (Ortner 1972), and white people are deemed cultured when compared to black people. The same can be said for Roma women: in Prizren negative stereotypes of the group as a whole are deflected onto women within the group. If this dichotomisation is seen as a fractal, as people are constantly subdivided into unequal oppositions, this goes some way towards understanding how racial and gender subordination of Roma women is more than the sum of its parts. Intersectionality in this context means that Roma women have less agency to mobilise, as they are subordinate to Roma men, who are in turn subordinate to non-Roma men. In addition, I would suggest that the more extreme the contrast is at the largest scale, the more extreme this contrast is likely to be at the smaller scale. An extreme subordination of Roma by non-Roma will fission into an extreme subordination of Roma women by Roma men. The relatively low status

88

Purity, mixture and representation

of the image of Roma, along with their powerlessness to control this image, is replicated in the extreme gendering of Roma women and girls. This leaves little space for gender solidarity: the extreme public/private divide, along with the restriction of women to the private domain, means there is little overlap between women from different groups.14 As the boundaries between public and private have grown stronger in the post-socialist era woman have less scope for participating in the public domain, and, as mentioned, Roma women who worked were often heavily criticised for transgressing this boundary. To say that these relationships replicate fractally on different scales is not of course to say that they are the same at each level: sexism manifests itself very differently from racism. Nor is it to say that subordination within subordinated groups is in any way inevitable or inherent. Rather, subordination engenders further subordination, inhibiting the emancipation and political representation of multiply subordinated persons. Representation has real-world effects, and the power of representation is exercised on subordinates at ever smaller scales. The collision of racialised and gendered images fortifies inequalities, which in turn then reproduce such images. As such the next section goes on to unravel the power of, and motivations behind, the performance of the gypsy image in the plays in Prizren.

Gypsy comedy, Romani tragedy During my fieldwork I attended several rehearsals for the play O Roma thay Ashkaliya, ‘The Roma and the Ashkali.’ The play was intended as ‘a Romani Romeo and Juliet,’ where the young couple are from two opposing families, one Roma and the other Ashkali. The family of the Ashkali girl are keen for her to marry a shaci suitor. Shaci, from the German schatz ‘treasure’ (used as an affectionate term in German), is commonly used ironically in Kosovo to refer to the diaspora. The shacis are the butt of many jokes (not just among Roma), and in the plays their western habits (rap music, fluorescent shell suits, hair gel, male subservience) are contrasted to the normality and tradition of those who stayed behind.15 The shaci character is unappealing, in contrast to the local Roma boy who she really wants to marry. The competition between the two families is entertaining throughout, but while she eventually elopes with the local boy, their marriage does not last, and she then returns to her family. I was asked to translate the following synopsis from Albanian for a funding application: Aktivitetet dhe Objektivave të Projektit Akivitete te drames jane ku në dramë flitet për dy familje Rrome të cilët janë në antagonizëm ndërmjet veti, por të cilët njëherë janë edhe të varura njëra nga tjetra. Njëri paraqitët për ti pajtuar, por pa sukses. Në vazhdim tregohet se si jeta bënë që këto familje edhe më pastaj të janë në kundërshtim, pasi dy të rinj dashurohen dhe martohen, por martesa nuk zgjatë, prandaj kundërshtimi ndërmjet këto dy familje shtohet. Në fund le të thuhet se drama ka karakter komedie.

O Romano Teatro and gypsy theatricality 89 Activities and Objectives of the Project The activities which take place in the drama relate to two Romani families who have an antagonistic relationship, but who are at the same time dependent on one another. One person attempts reconciliation, but without success. As the play continues we see that these two families continue to oppose one another, then two young people fall in love and get married, but the marriage doesn’t last, so the opposition between the families increases. In the end it can be said that the drama has a comic character. As I translated the last sentence, I thought there was a mistake. There was something incongruous in describing a play based on the tragedy Romeo and Juliet as having ‘a comic character,’ but I was assured that the plays are generally tragicomedies, combining the tragic and the comic. Unlike typical comic plots, the plays do not end in a reversion to the start. People die, marriages are ruined, families are destroyed – but they are still funny. Various forms of tragicomedy (or black humour, dark humour) are a common artistic form across south-eastern Europe. A form once criticised for being ‘mongrel’ or hybrid, not fitting classical norms, is also seen to reflect nature, as art reflects life (Guthke 1966). The mixing of genres, while perhaps reflecting life more accurately, unsettles the audience, leaving it unsure how to respond. Accordingly, theorists of tragicomedy often stress the importance of audience response, suggesting that tragicomic authors must have a knowledge of their audience in mind when the plays are written (Styan 1968; Hirst 1984; Foster 2004). Tragicomedy is not just interpreted as a closer, more realistic reflection of the complexities of life, but also as one that requires a different audience response. The author thus needs to be able to gauge the audience response more sensitively than if it were a straightforward comedy or tragedy. As mentioned above, Nexhip and other writers write plays for actors and audiences who are primarily close friends and family. The author shares a common schema of interpretation with the audience, and a level of cultural intimacy with them. Knowing what subjects interest them, and what makes them cry or laugh, is easy, as the author has the same or very similar experiences in life to his audiences. The tragicomic mode can act as a response to both post-socialist and postconflict contexts. In relation to the use of humour in the Romanian film The Death of Mr Lazarescu, Haliliuc (2015a) argues that, unlike benign humour, which is corrective, ‘black humor has no vantage point, it is the humor of disorientation’ (Haliliuc 2015a:157). This dark sort of humour offers comedy, but without the accompanying relief. Irony is commonly incorporated into this type of humour, acting as a way of coping with the ambiguities of shock, trauma and major upheaval (Fernandez & Huber 2001). As such the tragicomic use of irony surrounding grim events such as wars and massacres is particularly salient in the Balkans. This use of humour, which can often be shocking to westerners unattuned to irony in such circumstances, has something uncomfortably inverted about it: ‘rather than people saying the opposite of what is meant, the irony comes

90

Purity, mixture and representation

from reality being opposite from how it should be’ (Kibler 2004:465). Formulaic plots are thus mutated and inverted. Comedy which mocks the self has a precedent in the Balkans, and Nexhip told me that he was inspired by Branislav Nušić, a playwright in the Kingdom of Yugoslavia who wrote satirical comedies. In Nušić’s plays, widely familiar characters fulfil their stereotype hilariously. Nušić used comedy as a form of social criticism, and female characters in particular perform the stereotype of being gossipy, loud, crude and not knowing their place. The Roma plays are in some respects similar. However, Nušić’s work is typical comedy: by the end, everything has returned to its right place, while in the plays I am describing this is not generally the case. Another difference is that Nušić’s satire requires a cold distance from the characters. The satirist needs to be firmly entrenched in a self-constructed superior vantage point in order to engage in this kind of ridicule of his subjects. These Romani plays however, while being funny, show a warmth and tolerance towards the characters, as well as cold ridicule. This warmth is more in line with Herzfeld’s notion of cultural intimacy (2004), which he explains in terms of a humorous, yet regrettable self-knowledge. While Herzfeld is referring to the nation, as an imagined community, here the situation is more complex: while this representation of Roma may in some respects signify the wider Roma nation, here the representation of Roma takes on a very local inflection. People recognise the flaws they see on stage as their own, and this makes them amusing. It is funny to laugh at your own flaws, and Roma performing and mocking gypsiness can therefore be seen as part of a wider trope of people in south-eastern Europe performing and mocking themselves. Audience members often assume that the characters are based on real people from the mahalla, thus locating the stereotype in their own immediate circle. Everyone can relate to the humour in these plays, and the exaggeration and repetition of certain distinctive features makes them entertaining. At the same time, these features are lamentable. The use of this genre means they can combine the tragic and the comic, and shift seamlessly between the humour of self-deprecation, the superior vantage point of satire, and the tragedy of realisation. This shifting of vantage point is a central feature of the tragicomic mode. The tragicomic mode can shift between external and internal vantage points, between detachment and sympathy (Styan 1968), in a similar manner to irony, which displays a tension between inner and outer representations of self, and between public and private (Fernandez & Huber 2001). The ability to ironicise relies on the ambiguity between the formal or official, and the informal or local: irony emerges at the expense of the ambiguous collective (Herzfeld 2001). While the vantage point shifts from self to Other, the scale of self also shifts seamlessly between, on the smallest scale, friends and family in the mahalla, and, on the broadest scale, the Roma nation. At the same time, this perspective allows them to mock, vilify and pity the various scales of self at different points. Ambiguity is inherent in the presentation of gypsiness in these plays: it shifts between humour about the self, as cultural intimacy, and humour about the gypsy

O Romano Teatro and gypsy theatricality 91 as the Other. The public setting of the play is also a private setting of Nexhip, his close friends and family. The gypsy stereotype is viewed both internally and externally, and this vantage point fissions on different scales, fractally reproducing this division. The coexistence of the tragic and the comic, and these shifting vantage points, creates a disorienting and ambivalent response in the audience. Roma performing and mocking gypsiness can therefore be seen as an offshoot of a wider trope of people in south-eastern Europe performing and mocking Balkanness. This mockery wavers between the satirical othering of gypsiness and Balkanness, and self-recognition, laughing at oneself. At the same time the tragic is ever-present. The use of tragicomedy means that they can laugh at ‘gypsies,’ yet cry at their fate in the same breath. Any sense of a stable normality is absent. The ambiguity of people’s own identity is reflected in the way it is performed on stage, and received by the audience. Representing gypsiness through tragicomedy is a way of representing this Balkanist uncomfortable closeness to the constructed Other. The attempt to position oneself at the top end of the east/west progressivist discourses around which Balkanist discourses are structured is clearly present, but it is also contested. Rather than the modern west being seen as the pinnacle of all that as good, and the point that should be aimed for, the west is denigrated as irreligious, unnatural and soulless. The shacis and their western habits are mocked as much as the local gypsy characters. But the east offers no alternative: there are no prospects for migrating east, and while mostly devout Muslims, many people are suspicious of what they see as ‘Arab’ influences in religion, as well as Turkish influence in politics and the economy. Before the war, Prizren Roma participated in the Yugoslav socialist and Serb nationalist state; now in independent Kosovo they participate in the opposing Albanian nationalist state. Influences from the east aim to change Balkan Islamic practices, while from the west there are neoliberal and multiculturalist forces aimed at modernising and stabilising the region. All these ideologies are partially embraced, often less so in private than in public. The plays are thus disorientating in their ambivalence towards an ideal which people should be aiming for: while everyone is aware the image of the Balkan gypsy being presented is not the ideal, no other is posited in its place. There is an additional layer of disorientation: that of the private and public spheres. The private space of family, home and cultural intimacy on the smallest scale is put on display to the public in the form of a play. At the same time, the larger scale imagined community is performed for its own intimates, other Roma. The private language of the mahalla is replaced, to some extent, by attempts to create a public, standard Romani. On the one hand, the possession of a national theatre and national language show Roma, other Prizrenites and funders, that Roma exist in the sphere of culture, rather than nature, while on the other hand the image being portrayed casts the gypsy as nature. Reality for people who have lived through regime change, conflict and extreme social and economic hardship is disorientating: ‘if one’s ironic humor comes from the perversity that we, individually, put upon the world, then the world becomes a very terrifying place in which we have very little surety or control’ (Kibler 2004:465). Shifts in orientation also occurred in Serbian society under

92

Purity, mixture and representation

Milošević, where ‘the warps started piling one atop another at an increasing pace, white being proclaimed black, then white, then black again, and so on, until the situation became dizzyingly, nauseatingly opaque’ (Živković 2000:179). There is a feeling of lack of control, and lack of a clear aim beyond survival, upsetting the possibility of a clear metanarrative. Rather than a postmodern absence of metanarratives, I would argue that this is due to an abundance of compelling yet contradictory metanarratives. As we saw in Chapter 2, repeated shifts in state power have meant that each generation has needed to form a different set of allegiances. Contemporary NGOs, as Chapter 3 showed, provide project work with theoretically neoliberal goals, but without sustained employment and services. People are not disorientated because there is nowhere to turn, but rather because there are too many places to turn, none of which offer a solution to the problems of Roma in present-day Kosovo. Tragicomedy reflects this disorienting shift of metanarratives, through drama which ‘acquits us by laughter at one moment and then convicts us at the next’ (Styan 1968:257). There is no consistent vantage point, aim or teleology in a world in which everything is subject to change beyond our control.

Conclusion The image of Roma presented in these plays is racialised and gendered, presenting the characters as more ‘gypsy,’ closer to nature. This is part of a broader representational schema in which dichotomies of modernity and civilisation versus backwardness and barbarity are reproduced on different scales. These dichotomous representations and the power structures they reinforce descend through different layers, reproducing fractally. While the nature/culture dichotomy may frame differences between non-Roma and Roma, as well as between a Roma husband and Roma wife, this dichotomy has different meanings in each context. At the largest scale, the dichotomy of east/west is represented as total opposition, but the opposition of Europe to the Balkans represents internal othering, an understanding that the Balkans form part of Europe, but a peripheral, abnormal part. In the default, normal image of Europe, the Balkans are absent, only coming into view when they form an opposition, as an internal other. The same can be said of the way non-Roma relate to and represent Roma: whiteness is the default, non-racial category, whereas Roma are racialised. Further down the scale, the next fission of the fractal is the family, where male is considered to be default, in contrast to female. Increased inequality at higher levels of representation is echoed lower down the fractal hierarchies. The power to represent lies with those who have actual power and they represent themselves as neutral in contrast to their subordinates. However, the wielding of power to represent does more than reproduce images: it also reproduces actual inequality. This means that intersectionally disadvantaged people are trapped within these fractals of power relations: Roma women cannot represent themselves as Roma women without undoing the inequalities between Roma women and Roma men, but they cannot do this without first undoing inequalities between Roma and non-Roma.

O Romano Teatro and gypsy theatricality 93 Conservatism is thus more entrenched further down the fractalised hierarchy, offering no avenue of escape. The metaphor of fractals, as I have shown, demonstrates how Balkanism can be reproduced on ever smaller scales. However, the representation of the gypsy as the most Balkan, and least European, and the essentialising of this image, brings it much closer to racial and post-colonial treatments of people, an opposition rather than an ambiguity. European ideas of the Balkans posit racism as a quality of the Other, the irrational hatred of a Balkan savage; the civic nationalism of the European is contrasted to the ethnic nationalism of the Balkans, with the racism directed against Roma being constructed as an essentially Balkan feature. This means the treatment of Roma as a racial Other moves closer to opposition than ambiguity. From the perspective of the west, Roma are completely othered, while being subject to the racism of the ambiguous, peripheral Balkans. However, for those in south-east Europe, Roma are too close. In a sense, there is a continuum between the ambiguity of cultural differentiation present in Balkanist schemes of representation and the oppositions of essentialist racial differentiation. In socialist times, there was a level of stability to this schema of representation and the power structure that controlled it. Socialist ideology lent itself to a teleological narrative of modernisation, experienced in practical terms in socialist Yugoslavia in terms of work, welfare and services, as described in Chapter 3. Notions of normality within the schema were linked to ideas of civilisation and modernity. Economic crisis, war and regime change have destroyed this stable notion of normality (Jansen 2009; Maček 2007; Greenberg 2011). Rather than socialism indexing modernity, from a western perspective, it came to symbolise the temporal and spatial backwardness of the east (Brandstädter 2007). As those with money to fund projects perceive the post-socialist east as backward, socialism’s modernist promise is overturned. This upsetting of large-scale schemata of representation has resulted in the ambivalence and disorientation I described above. The representational victory of capitalism over socialism, of west over east, has reverberated throughout society. At the same time, the major shift in power structures within Kosovo has resulted in reiterations of certain dichotomies on smaller scales. For many outside of Kosovo, post-socialism meant a shift in the public/private distinction, with women again restricted to the private sphere (Pine 2002), and the crisis of war altered, and strengthened, the public/private distinction (Henig 2012). Crises of representation can thus lead to a reaffirmation of hierarchical boundaries, and ‘re-traditionalisation’ (Ströhle 2016; cf. Spivak 1988). Conceptualising these processes as fractalisation allows us to see not just the structures of subordination which control representation, but also the possibilities for change and agency within these structures. In this sense, crisis and disorientation, represented in these dramas as a stereotyping of both the intimate tragic, private self and the comic public Other, can be seen to be part of a wider story of structural inequalities transformed through upheaval. The dramatic representation of the stereotyped backward Balkan gypsy, performed using the modernist ideal of a standard language (to be discussed in the next chapter), and the pervasive ambiguity of the tragicomic mode, reflect this upheaval.

94

Purity, mixture and representation

Notes 1 I use the terms gypsy and gypsiness (in lowercase) to describe an image rather than the ethnic group. 2 In the end the title used was O Sundal si o Romengo Them, ‘The World is the Roma’s Land’ and the scene was somewhat different. 3 I will return to homophobia and understandings of homosexuality later in the chapter. 4 There is no such distinction in Romani, though Roma from Terzimahalla would sometimes use gabel in a similar derogatory fashion. Gabel is derogatory in all languages, but is also used to refer to an older distinction between gabel Roma, who were travellers, and Arli, the settled Roma who were long-term inhabitants of the city. 5 See Gal (1978) for a discussion of how a similar predicament leads to language shift. 6 cf. Jašarević (2007) for a comparison of the image of the peasant and the gypsy. 7 Tony Gatlif is a filmmaker from France, with Romani origins. 8 Sasha Petrović was a French-born Yugoslav filmmaker. 9 Ananiadis (2002) even argues that the NATO bombing of Belgrade was legitimised by a discourse of inequality that posited it as the bombing of the barbaric by the civilised. 10 I have merged several different comments and discussions here in order to avoid identifying specific individuals. 11 This has a precedent in the Balkans in the eighteenth century, where, unlike in western Europe, the diffusion of Enlightenment ideas meant the demystification of organised religion, rather than dispensing with it outright (Đorđević 1975). 12 My own position within these rigid systems of hierarchies was problematic, to say the least. As a highly educated westerner I was in one sense at the top of the hierarchy, an external other with different (and 'unnatural') views and ways of doing things. As an unmarried, young female I was quite the opposite. In fact, in this highly unequal, patriarchally structured setting, my lack of an identifiable intersectional position meant I was out of place. 13 This intersectional perspective also restricts the possibilities for activism: Roma feminists are seen as traitors by both Romani activists (Jovanović & Daroczi 2015; cf. Schultz 2012) and non-Roma feminists (Brooks 2012). cf. Baker (2015) for a discussion of the double alienation of gay Gypsies in the UK. cf. Swaby (2014) for a discussion of divisions and problems of categorisations within a black women’s movement. 14 cf. Stolcke (1994) for the distinct position of indigenous and white women in colonial Latin America. 15 This is related to the image of the gastarbajter more generally in Yugoslav and former Yugoslav cultural production (Daniel 2007; Krstić 2016).

5

Standardisation Learning linguistics in the bath

Introduction Local, regional and international conceptions of Romani language all play a role in local language politics and social hierarchies. Language politics in the Balkans have been shaped by histories of empire, resistance and nationalism, and Romani, though specific, has by no means been immune to these forces. My early encounters with the idea of a Romani standard involved being corrected by intellectuals on the way I spoke and discussions, debates and people’s attitude towards me as a language learner all shed light on the process of standardising Romani. This opened up a range of avenues for me to understand the way people qualify and legitimise certain speech practices at the expense of others, and the way the normative metalinguistic discourse about speech practices diverges from the practices themselves. In Prizren, I witnessed the process of legitimising certain language practices as standard Romani at three main sites: the radio, the theatre group and school classes. All of these are connected to the NGO Durmish Aslano discussed in previous chapters. Radio Romano Avazo is the Romani radio where I spent a substantial part of my fieldwork. One of the central aims of the radio is to hold programmes in what is deemed to be standard Romani, and to present and discuss Romani issues. It is often necessary to hold programmes in BCS or Albanian when guests are nonRoma, or Albanian-speaking Roma. Similarly, many Romani speakers are not familiar with the standard, so speak non-standard (local) Romani. But in general, the staff of the station use what is deemed to be standard during broadcasts. The Romani-language theatre group is also part of the NGO. They usually receive funding from one of the government ministries for a play in the house of culture in Prizren on April 8, International Roma Day, and occasionally go to festivals in Skopje, Macedonia. Many of the plays are written and directed by Nexhip. His teenage children, as well as those of the staff at the radio, act in the plays. The rehearsals were often sites of debates about the use of language. Nexhip was employed by the Ministry of Education, externally to the NGO, to teach optional Romani-language classes in the four primary schools in Prizren

96

Purity, mixture and representation

that have a substantial number of Roma pupils. The classes were once a week after school, and were aimed at second and third years (8–10 years old), though older and younger children also came along. Nexhip often picked children up and dropped them off in his car, which was a novelty for many of the children. I attended too, as something between a teaching assistant, a pupil and a general attraction for the children. The attendance varied widely during my time there. I was told that Osman, the director of Iniciativa, had been involved in designing a curriculum funded by the Ministry and ‘civil society,’ but I never saw it, and the classes varied so widely that it seemed very hard to imagine that there was a set curriculum. Nexhip would sometimes bring two books written by Selahadin Kruezi to the class: an ABC book and a book of mathematical terms. Both were hardback, colourful and full of pictures. Nexhip would sometimes use them for examples of words, and the children were very keen to look through the books. The most consistent attendance was at the school near Terzimahalla, where the children also spoke Romani at home. These classes focused on learning new words, and writing in Romani. Here Nexhip tried, mostly unsuccessfully, to persuade the children to call him profesor (Ro: ‘teacher’) and not ajo (Ro: ‘uncle’). The other three schools had lower attendance, and many of the pupils had little or no knowledge of Romani. Depending on how many pupils attended, and whether they knew how to write, they would repeat or read out words and basic phrases, or play number games. Before the start of the autumn term in 2014, a project to promote the classes to schools and parents ran for a month, and people began to show more interest in the classes. However, this was followed by a mass migration of Roma to the west, and the numbers began to dwindle again. Chapter 1 discussed how my expectations about learning Romani changed over time. The learning process differed according to whether I was learning informally, by imitation in a specific context or didactically, being intentionally taught words out of context (Abercrombie forthcoming). This chapter focuses on decontextualised didactic learning. This more formal aspect of language is embedded in social hierarchies. As such, the chapter also builds on the discussion of the diachronic and synchronic role of Durmish Aslano in Prizren, to look at how the NGO’s prestige shapes language standardisation.

Politics and letters In April 2014, I was walking through town with Moni on the way to the radio station. He was telling me about a meeting where the Ministry announced a plan to institutionalise a Romani alphabet. He told me that many people were annoyed about this, especially The Professor (Ibrahim Elshani), who had not been consulted despite having worked on the Romani language for many years. Moni told me that this new alphabet was based on Croatian, and was more international than the one they used. Those present at the seminar had argued against it on the basis that it was difficult, and that they already had an alphabet, but Moni told me the real reason for their reluctance was that it looked too much like Latin alphabet Serbian.

Standardisation

97

The Ministry for Education, in conjunction with Roma politicians from PREBK (Partia Rome e Bashkuar e Kosovës, Albanian for ‘The United Roma Party of Kosovo’) had decided to implement this new Romani orthography.1 This orthography used diacritics similar to the Latin alphabet used for BCS (hence it being referred to as Croatian), and it had been used to write Romani on a wider scale in Yugoslav times (Friedman 1985a). However, despite this precedence, and its acceptance on a regional scale, this orthography was met with scepticism by most of my informants. The main orthography used at the NGO is also based on the Latin alphabet, but similar to Albanian, except for the use of for /ʤ/ and for /j/, and with the addition of , , and for aspirated consonants (see Note on orthography). The distinction between /ł/ and /ʎ/ (distinguished in Albanian by and , and in BCS by and ) exists in Romani, but in this alphabet both are written as . According to Kruezi (2008), the distinction should be clear from the preceding vowel, though this is not always the case. This orthography is also used at RTK2 (Radio Television Kosovo 2 – the department for minority languages) and The Professor states it has been used in Kosovo (and beyond) for around 40 years (Elshani 2011). It is used in Kruezi’s books for schoolchildren,2 two occasionally produced local magazines,3 various other NGO publications and later the website for the radio. It is used less in the school classes, for fear of confusing the children, who are still learning to write in Albanian. While it is still not so widespread that switching would cause problems, its advantage is that it has much more in common with Albanian. The choices surrounding Romani writing are made at different scales, and writing is used differently at local and international levels. On the whole, most actors choose their own variants, often using the orthography of the national language (Matras 1999; Halwachs 2011). This is problematic in post-conflict Kosovo where the two de jure official languages, with different alphabets, actually exist in almost complete separation. Thus, while the use of , , are ‘evidently borrowed from what is perceived as a more international spelling’ (Matras 1999:489), in Kosovo these graphemes have become associated with Serbian. Roma in Kosovo who have remained after the war are careful to distance themselves from anything seeming to be Serbian because of their insecure position following the war, as discussed in Chapter 2.4 As such an alphabet which may be associated with Serbian language is seen as threatening. Orthography, being both highly symbolic and highly visible, is easier to manipulate than other areas of language. This is not limited to Prizren: Halwachs (2012b) cites the rejection of a more international orthography by Roma in Burgenland, Austria, as an attempt to avoid association with Croats and Hungarians. Nor is it limited to Romani: Jaffe (1999) finds that the use of French spellings in the nascent Corsican standard carries more stigma than the use of Italian spellings. In Kosovo, while the use of Latin alphabet Slavic letters provoked resistance, the use of Cyrillic, emblematic of Serbia, the Orthodox church and, for some, Serbian nationalism, is far more problematic. With the exception of Orthodox churches (which are under armed guard), the only Cyrillic I saw in Prizren was

98

Purity, mixture and representation

on the charred remains of a Serbian building destroyed in 2004. The visibility and permanency of writing makes it easily recognised and stigmatised. While people argued vehemently for the use of their own, Prizren, orthography, in practice, many people mix this Romani orthography with other (primarily Albanian) orthography, showing little consistency. The same word may be spelt differently by the same person, sometimes within the same text or even word. When I first arrived, I asked how to say ‘pleased to meet you,’ and Nexhip wrote avlama shukar so pengjargjum tut, but when I read it out I was told I pronounced it wrongly: in the word pengjargjum (Ro: ‘I met’), the first represents /ʤ/, using the Albanian , which in Gheg (unlike in the Tosk-based standard) is not differentiated from ; the second on the other hand represents /ɟ/ or /gj/. This would not cause problems for native speakers who learn the spelling after they have learnt the word (indeed, no one noticed these inconsistencies), and this inconsistency did not stop them being very vocal in their support of ‘their’ orthography. This phenomenon is not unique to Romani, and many people (both Roma and non-Roma) also write like this in Albanian. Shopfronts often have a name or a sign spelt in two different ways. This is sometimes due to mixing Albanian orthography with Turkish or BCS; sometimes it relates to an approximation of local Gheg forms in Albanian orthography, rather than the use of the Tosk-based standard. Similarly, Nexhip had completed a university degree using Bosnian language as the language of instruction, but this did not prevent him from referring in written communication to the language as boshnjaqki, BCS for ‘Bosniac language,’ written with Albanian orthography. This was not mere carelessness: he, and others, routinely wrote in this unpredictable and spontaneous way in both formal and informal texts. This suggests that at least at the level of orthography, consistency of form is not considered a central feature of defining a standard. Orthographic choices are clearly important, but they are not expected to be consistent. Sporadic and creative use of writing is also a feature of online Romani, which is marked by consistency of dialect features alongside creative pluralism in orthographic representation (Leggio 2011; Leggio & Matras 2013; cf. Halwachs 2012a). In Prizren the sporadic, creative use of orthography is not limited to online communication, nor is it limited to Romani. The internet has opened up a new domain of language use, and while in some languages this domain is in contrast to well-established standards of consistency and propriety, with Romani (and to a lesser extent Albanian), there is no such contrast. The aim when engaging in informal writing practices in Prizren is not the contrast to (and escape from) an older elite idea of written language. Rather, it is part of the contemporary need to differentiate it from alphabets that look Slavic, and to construct a writing system parallel to, yet distinct from, Albanian, as the main dominant language. As such, the regimenting of Romani to match the uses of the dominant standard language (Gal 2006a) lacks an emphasis on orthographic consistency. It is simultaneously necessary to show an allegiance with Albanian, as the dominant ethnic group, by avoiding and purging any visible Serbian elements.

Standardisation

99

Words and purism When I asked Fatmira how to say značiti (BCS: ‘to mean’) in Romani, she responded by saying znaqinela,5 but added that znaqinela is a Serbian word, stating that she didn’t know the ‘real’ Romani word. She added: Mi ne znamo svoj jezik (BCS: ‘we don’t know our own language’). She suggested I asked at the radio station as they would know better. This response of not knowing Romani words, or rather not knowing ‘proper’ words was one I regularly encountered while trying to learn Romani. It was also often accompanied by a deferral to superior knowledge. This was particularly true of words I learnt from women in the mahalla, and in other private settings. Fatmira was aware of something lacking in her own speech, and pointed me in the direction of the NGO. One of the daily tasks at the radio was to translate news from Albanian to Romani, and read it out. Vera and Moni, the two journalists, were responsible for this. When Vera got to an Albanian word she didn’t know in Romani, i papranueshëm (Shq: ‘unacceptable’), she looked it up in the Albanian-Romani dictionary written by Kujtim Paqaku. It wasn’t there, so she asked Nexhip. He didn’t know, so typed pranon (Shq: ‘accept’) into Google Translate to get the English accept. Not knowing English, he struggled to pronounce the double letter, saying something more like ats-tsept, to which he then added the suffix –imi to make ‘acceptable,’ and the prefix bi– to make it negative, giving biacceptimi. He told me that when the journalists didn’t know the ‘real’ Romani word, it was better to make Romani words from ‘international’ words, rather than use words from surrounding languages. ‘International’ words almost always referred to loans from English: the international language, used by the international community. These kinds of spontaneous techniques of borrowing were common. Nexhip and Kujtim often discussed and debated language issues. On one occasion, Kujtim told Nexhip not to use beginava (Terzimahalla Romani: ‘I like’) on the basis that it was korohani (Terzimahalla Romani: ‘Turkish’) – from Turkish beğenmek. Kujtim suggested using man plaqe instead, but Nexhip thought this sounded too much like the Albanian më pëlqen. Presumably the Romani term is from Romanian, and etymologically related to the Albanian term. Terzimahalla Romani had few Albanian borrowings, and Nexhip may have been referring to the fact that Kujtim is from another mahalla and spoke a different dialect. Kujtim on the other hand is likely to have learnt this through his involvement with Romani activists outside of Kosovo, some of whom speak Romani with Romanian borrowings. Then they discussed how to say ‘be careful.’ I had asked Nexhip how to say pazi (BCS: ‘be careful’) a few days before; unsure, he had responded: Mi kažemo ‘javaš’, ama to je turska reč. Nije naša reč (BCS: ‘we say yavaş, but that’s a Turkish word. It’s not our word.’). The primary meaning of yavaş is ‘slowly’ in Turkish, but it can also be used in any language in Prizren to mean be careful, or go slowly, similar to the use of polako in BCS. Nexhip then suggested as a Romani term dikkat. Dikkat in Turkish means ‘attention’ but it also sounds like dikh, the Romani word for ‘look.’ In this sense, it would mean something like ‘look out.’ Nexhip thought it would work, as it is both recognisable, from Turkish,

100

Purity, mixture and representation

and sounds like Romani. Kujtim again responded with akava hitoy korohani lafi (Ro: ‘that’s a Turkish word’). He was clearly unimpressed with Nexhip’s attempts at making standard Romani words, and had more authority on this issue. Kujtim turned to me and said in an instructive tone, and in BCS so I would definitely understand, lingvistika se ne uči u kadi, ‘you can’t study linguistics in the bath.’ Noticing I was somewhat baffled, he explained that you learn about languages by doing research in the field, by going to the forests of Albania and finding isolated groups speaking pure forms of Romani. He had conducted research in the field, not in the bath, and he was therefore more qualified to make new words. Numerous words used in Terzimahalla are deemed by the NGO staff to be ‘foreign’ and therefore incorrect. For example, gindinava (‘I think’) is preferred to mislinava, which is from BCS misliti (‘to think’); shajnava (‘I can’) is preferred to mozhinava, from BCS može (‘can’). Similarly, words which are perceived to be of Turkish origin are talked of as being non-Romani; for example, jeftisar (‘sorry’) was preferred to afikya (from Turkish affi, ‘forgive,’ with the Romani imperative ker, ‘do’). On several occasions I learned a word used in everyday speech, recognisably of BCS or Turkish origin, and was told instead to use a word which seemed to be of Romanian origin. For example, I was told that numay (from Romanian numai, ‘only’) should be used for ‘but’ in place of ama, which is from Turkish but has also spread into other Balkan languages. Though I found the specifics of lexicon and orthography to be contested, the aim of standardisation was not: they all wanted the best, purest and most authentic Romani words from which to construct a standard language. If I ever asked why a certain word or spelling should be used, I was told either that it was gramatika (‘grammar,’ in BCS, Shq and Ro), a term used to refer to an implied standard or proper way of speaking, or simply that this was may laqhi or may shuzhi (Ro, standard: ‘the most beautiful’ or ‘the most pure’).6 Purity and beauty are not clearly delineated categories, and are based on subjective judgements, but subjective judgements made within a particular social setting. When Roma in Prizren talk about what sounds nicest, and what is purest, they tend to imply what sounds least like the other languages known to them. As such, one of the tasks of those involved in standardisation is to identify foreign-sounding words, and to purge them. Because of the need to purify the language, in the examples above, beginava and yavash are identified as Turkish beğenmek and yavaş, and a purer, more Romani word needs to be found. Hence the emphasis on purging Romani of non-Romani words focused on othering words from local languages. Romanian-origin words were acceptable because they were not recognisable as foreign; most people in Prizren do not know Romanian. The rejection not simply of alien words, but specifically of words from the surrounding languages, is apparent with man plaqe, which was acceptable to Kujtim as a Romani word, whereas Nexhip thought it sounded Albanian. Similarly, biacceptimi is acceptable despite its obvious foreign origin. International words can become Romani, and replace impure BCS, Turkish and Albanian words. Words that are less familiar in the Prizren context are interpreted as more authentically Romani.

Standardisation

101

Purism is directed against perceived lexical incursions associated with a dominant neighbouring language. With more than one dominant neighbouring language, the need for differentiation from different neighbours can lead to a lack of solidarity when it comes to standardisation on a wider scale as ‘purists will not agree on the targets for puristic censorship and intervention’ (Thomas 1991:126). In this case however, all local loans are targeted, while words from further afield are accepted and encouraged. Reactions against loanwords are historically situated. Mixed codes are a product of domination, and as such come to symbolise this domination (Jaffe 1999). The foreign-origin words in Romani represent the incursion and domination of foreign elements, and therefore people feel the need to purify them. As such, the commonly used word anglaynava (Terzimahalla Romani: ‘I understand’), from Turkish anlamak, is considered impure. But in Prizren, it has been replaced not with the more international Romani haqarava (or other cognate terms), but instead with lava khan, calqued from the Albanian marr vesh (‘I agree, I understand’); marr, like lava, means ‘take,’ and vesh, like khan, means ‘ear.’ Standardisation often relies on calquing to replace loanwords (Thomas 1991; cf. Hill & Hill 1980), though there are also cases of lexical enrichment relying on loans from the dominant language (cf. Wexler 1989). Here the use of a calque from Albanian, rather than a borrowing from Turkish, or indeed a more international Romani form, shows the importance of correlating with, yet distinguishing from, Albanian. Thomas (1991) argues that separation, solidarity and prestige are central to language standardisation. I suggest that relationships with local languages are more central than solidarity with a wider Romani speech community. Rather than looking to work done on Romani standardisation internationally, they focus on the affinity with and distinction from local languages. Local processes of standardisation have different priorities to larger scale processes. The preference for words of non-BCS and non-Turkish origin is sometimes explained in terms of function, on the basis that they will be understood in neighbouring countries. Moni once explained that I should use hramina in place of pishina, (from BCS pisati) for ‘we write’ not because it was gramatika, but because ‘everyone, even if they don’t know Serbian, will understand this, people in Macedonia, for example, will understand this.’ However, international communication is rarely cited as a priority. In fact, as seen in the discussion on orthography, the need to show allegiance to those in power (the Albanian authorities) and emphasise a difference from previous Serbian authorities, is given far more importance than the need to ally with Roma from other countries. While purity in language is valued, so is richness. If language is understood as a collection of words, richness is understood as the sum of these words, excluding words seen as foreign. The need to use foreign words is thus seen as a sign of the ‘poverty’ of Romani vocabulary, as it is assumed that foreign words compensate for this poverty in Romani. For Nexhip, the fact that the single Romani word Rom could signify three different words in BCS – čovek ‘a man,’ muž ‘a husband’ and narod ‘a nation’ – was a sign of the poverty of Romani. Similarly, he once explained to me that the verb fulela had three meanings: fulava means

102

Purity, mixture and representation

‘I’m going down,’ fulavava means ‘I take off’ (clothes, for example) and fulinava means ‘I throw.’ Despite the different derivational infixes of these verbs, he took the common root as an example of the poverty of the language, a quantitative lack of words. By contrast, Kujtim would extoll the richness of Romani, listing, for example, the word for egg in three different dialects, and stating a high rate of synonymy in Romani. From his perspective, the problem was not a poverty of words in Romani, but rather the ignorance of many Romani speakers. While dialect variation is often seen as an insuperable problem for language standardisation (Friedman 2002a), for Kujtim it was evidence of the strength and richness of Romani. Debates over standard Romani are couched in terms of correctness, rather than consistency or uniformity, and relate primarily to words and orthography rather than to other aspects of language. People were always keen to help me in the process of learning Romani by telling me to use different words, the ones they considered to be standard. While correcting me for using a non-standard word would normally be followed by a comment about the word I had used being Serbian, Turkish or simply non-standard, this was not necessary when my conjugation or declension, for example, was wrong. While they did sometimes correct my misuse of morpho-syntax, this was normally on account of me speaking like a non-native, rather than speaking non-standard, and this did not require legitimation through a metalinguistic discourse in the way that correcting non-standard usage did. In general, as long as people could understand, they did not correct my Romani for sounding non-native, but they did correct the words I used for sounding non-standard. This emphasis on lexical items, rather than other areas of language, suggests an understanding of language as a collection of words. For those without linguistic training, words, especially their roots, are far more tangible than morphosyntax, and therefore more liable to purification. Language contact can lead to structural convergence, while maintaining surface-level separation; this kind of convergence, unlike loanwords, is much less obvious (Matras 2000). Words, on the other hand, are easy to convert into a symbolic resource which can be identified as foreign and purged, or as not foreign and therefore Romani. When people said they did not know their own language, they meant that they could speak fluently but did not know gramatika, did not have command of all the many ‘pure’ words that purportedly exist. ‘Knowing’ the standard means knowing the ‘pure’ variant, rather than (or as well as) their own ‘impure’ vernacular. In this respect, Roma are like others in Prizren, particularly Albanians. An Albanian school caretaker told me: ‘We don’t know our own language, we speak it but we don’t know gramatika. I know Serbian better than Albanian even though I’m Albanian, I always had 5 for Serbian and 3 for Albanian.’7 This feeling of ‘not knowing’ belies an awareness of the difference between the way people speak in Prizren, and the standard variants of the languages they speak. Turkish and BCS speakers express similar disjunctures between their speech and gramatika.8 Many people, then, are aware that their everyday speech practices are considered to be a non-prestigious and impure form of the language. This is what Milroy

Standardisation

103

(2001) describes as a standard language culture: the idea that a standard is correct becomes recognised as common sense, rather than as an ideological stance. This can lead to a reification of the standard, where the standard comes to be something understood as natural, and imbued with truth (Silverstein 1996). Albanian, Turkish, Serbian and Bosnian all have varying levels of codification of their respective standards, which are endorsed and legitimated by state institutions. The source of legitimacy of standard Romani is less obvious, raising the question of how, where and why the standard is being diffused from the intellectuals discussed above to other Roma. The feeling of ‘not knowing one’s own language’ thus relates to the prevalence of an idea of a standard, making the spoken variant seem non-standard, or even sub-standard (cf. Hill and Coombs 1982). Roma in Prizren primarily experience standard language ideology, the idea that some ways of talking are better than others, through the use of surrounding languages. In making Romani into a standard, then, those involved in this project need not just to make a pure standard, but to turn the pure standard, something quintessentially artificial, into a seemingly natural object. It is this object that people are talking about when they say they do not ‘know’ their language.

Intellectuals and barbarisms The use of words from surrounding languages was discussed by my colleagues at the radio. In a group interview with my colleagues, Moni explained: Sakoya qhib si ola barbarismya koya lele lili yaver qhibyata, feri so sakoya qhib, sitoy, er, qhoroli, lela lafya ini yaver qhibyenda. Ake misal, misal lava, eee, e albanyuni qhib, on pana kerena buti vash te kere e standardizaciya … Pandar nane olen yekh ee, adekvatyuno vakeripe ano sa teritoriya kote jivdinena Albayna, a numay e romani qhib sito odoya koya er, sar me dikhava, nane hitoy yekh ee, shuzhi, ee, ee, qhib feri so manglape but phare te kergyol buti. Si amen ano listi akola kerena buti linguistiya phando e romano qhibya, si o Marceli, si o … ini yaver jene may but penjargyole o Marceli ano teritoriya balkanesko. Vash akava sebepi si lafya akola lela kotar yaver qhibya. Every language has barbarisms which are taken from other languages, because every language is, er, poor, it takes words from other languages. Here’s an example, take for example Albanian, they are still working on standardisation …. They still don’t have one, er, adequate way of talking across the territory where Albanians live, but Romani language is such that, as I see it doesn’t have one pure, er, er, language, because it’s necessary to work hard on this. We have a list of linguists who work with Romani language, there’s Marcel, there’s … also other people, Marcel is the most well-known in the Balkans. This is why we need to take words from other languages. In this extract Moni is consciously using the standard: he uses the formal/nonderogatory terms albanyuni and albayna (‘Albanian’ and ‘Albanians’) rather than

104

Purity, mixture and representation

their everyday equivalents gaunani and gaune; the standard feri so rather than soske for ‘because’; the formal terms standardizaciya, adekvatyuno, teritoriya, which are formal concepts with no obvious everyday equivalents. Jivdinena is most likely from BCS, but it is accepted as standard Romani in contrast to zhivinena which sounds closer to the Serbian žive, and is used in the enclaves. However, he uses sebepi, which others avoid because of its Turkish origin (see below). He discusses ‘barbarisms,’ the words from other surrounding languages that have entered Romani, making it impure. Hard work is required to purify the language, in particular by those anointed with the position of linguist, whose role is to pinpoint impure words, and expunge them, replacing them with ‘pure’ forms, or alternatively making new forms without using the resources of other local languages. Most people find pure forms by asking someone in a position of authority. Interestingly however, those such as Kujtim, at the top of the social hierarchy, consider that the purest forms are to be found at the bottom of the hierarchy: among the so-called wild people in the forests of Albania. These Roma are seen as the most excluded, most isolated and least ‘cultured,’ and are therefore treated as untouched specimens, yet to be defiled by barbarisms. From this position, natural language needs to be preserved, and defended against invasion, and while this natural, pure language, is spoken by the natural, pure folk, untainted by contact, it must be defined and legitimated by intellectuals. Here purism is focused not on the Indic origins of Roma, but the contemporary purity of isolation.9 This understanding of language is typical of the Herderian tradition, which started in Germany and spread throughout central, then south-eastern Europe in the nineteenth century (Bauman & Briggs 2000). The mixture of ‘small’ ethnic groups in the Balkans was understood by western scholars as primarily a negative phenomenon, and subsequently local scholars resorted to linguistic purism to find the authentic, untouched, pure speaker (Petrović 2003). This scalar reproduction of hierarchies in language ideology is a form of fractal recursivity, the reproduction of binary differences in different places and at different scales (Irvine & Gal 2000). At the same time, the contrast between cultured and uncultured, urban and rural, mixed and unmixed, depends on Balkanist understandings of the Other within, as discussed in Chapter 4. Here the opposing definitions of self as cosmopolitan, urban and civilised, and the Other as pure, isolated and uncivilised come together to posit the Romani intellectual as a paternalistic researcher and defender of the purity of the ‘wild people.’ In this sense, being multilingual is acceptable (and indeed a sign of being urban and cultured) so long as the languages are clearly bounded and nothing impinges on the purity of one’s own language. As I have explained, when people felt they didn’t know their own language, they would turn to those who they felt did know. At the radio, this meant Nexhip, unless Kujtim happened to be visiting, in which case he generally had the higher authority. But there was an even higher authority than Kujtim: Marcel. Marcel Courthiade, known locally just as O Marceli, is someone talked about in almost

Standardisation

105

legendary terms. Marceleski qhib (Marcel’s language) is considered the ideal language. Marcel Courthiade lives and works in France, was born in Albania and considers himself Roma. He currently works for INALCO in Paris. He visited Prizren once to conduct research on Romani, and I was told that he went to Roma settlements in Albania where even Kosovo Roma were afraid to go, to seek authentic Romani. He has also compiled an orthography which was adopted by the IRU (International Romani Union) (Courthiade 1989).10 Kujtim is the only person I met who uses this orthography; others say they find it too difficult. The orthography was intended to unite the dialects of Romani, a task Romani activists have been trying to achieve for some time. Thus in 1990, the IRU adopted this orthography, aimed at encompassing all Romani dialects. Marcel’s international renown was well-known in Prizren, even if the orthography wasn’t. Kujtim is also seen as an authority on the language as he has been involved in research both at home and abroad, and writes poetry in Romani, as well as having written a dictionary. In spring 2014, when Kujtim, Nexhip and some others had set up a political party (Kosovaki Nevi Romani Partiya [Ro: ‘New Roma Party of Kosovo’]), Kujtim was elected as the Roma deputy of the Kosovo Assembly. His authority, at least in part, stems from his research of obscure dialects among ‘wild people’ or ‘real gypsies’ in the forests of Albania, and he refers to this experience to show he is a real linguist, like Marcel. His other source of authority stems from the widespread knowledge that he had studied at the Sorbonne (more precisely at INALCO) alongside Marcel. Kujtim’s proximity to Marcel clearly lends him an air of authority – but this authority often entails fear or resentment. I often detected a sense of jealousy and competition between Nexhip and Kujtim, which sometimes led to the two men criticising one another. When Bajram suggested Nexhip should go to Paris to learn Romani, Nexhip responded angrily, saying: Me vakarava romani sako dive. Na manglape te nakhav ano Parizi (Ro: ‘I speak Romani every day. I don’t need to go to Paris’). Recognising that Kujtim’s authority flowed from his relationship with Marcel and his stay in Paris, Nexhip sought to bolster his own position, by criticising not Kujtim’s knowledge, but his connection to Marcel and Paris. The debates between Nexhip and Kujtim were no secret; other contestations occurred more privately. While Kujtim, Nexhip and a few others were busy setting up the political party, I had the task of making them endless cups of coffee. This was done on a Yugoslav-era camping stove in the kitchen of the radio station. I asked Kujtim for a lighter, using the phrase situ chakmako (Ro: ‘do you have a lighter’) – the phrase used in Terzimahalla. He corrected me: Chakmako hito korohani lafi, to je od turskog. Vakar ‘yagaripe’ (Ro: ‘Chakmako is a Turkish word. BCS: It is from Turkish. Ro: Say yagaripe’). Yagaripe is derived from the Romani word yag meaning fire. Later, after the people involved with the party had left, the other staff at the radio rolled their eyes and laughed. While this kind of purism was publicly respected, in more private settings it was open to ridicule. The staff also criticise each other on points of language: Saverd once argued I should not be learning from Nexhip, but from him and his older brother, Hysni.

106

Purity, mixture and representation

He said that he grew up speaking proper Romani and knew proper Romani words. From his grandmother, he had learnt ferash for dustpan, and brlyaga for rubbish, which he claimed the other staff didn’t know. Saverd’s older brother Hysni, who I had attempted to conduct a language survey with, was particularly fond of telling me about Romani language, sometimes for hours. In early 2014, one of the first times I met him, he was quite frustrated with the radio station, complaining that the way they spoke Romani was not correct and that I wouldn’t learn properly. He said it sounded like Albanian, that it didn’t have the rhythm and heart of Romani. He suggested I should learn from Kujtim, who he said knew best because he studied in Paris with Marcel. He also found me lots of information on Marcel and showed me some books online. Then he mentioned the visits of various intellectuals, and a trip he had made to Albania to learn proper Romani. He described the Roma in Albania as pravi cigani (BCS: ‘real gypsies’), saying he was scared, as they were wild people. He also mentioned Dragoljub Acković from Belgrade, but complained that Acković used Cyrillic to write Romani, which he said was not international, and also of course would not be acceptable in most parts of independent Kosovo. Those with the social and cultural capital to position themselves as intellectuals, at the top of the hierarchy, are in a different position to journalists, whose task it is to reinforce and diffuse ‘pure’ language directly. This is not always straightforward. In an interview with Avdi, a Roma journalist at RTK2, I asked how journalists dealt with unknown words: Avdi: Dengyum odola misal vash e ‘kondiciya’, er, vay, vash o lafi ‘sebepi’- si Romani lafi, ‘ash’ si. AA: ‘Ash’ si? Avdi: Ash. Va. Vakerna kay ‘ash’. Numay kotar avel akava lafi na jangyol. Misalake mange o Kujtim Paqaku vakerdo kay ‘sebepi’ sito ‘ash’ … Avdi: I gave the example of kondiciya, er, for, for the word sebepi – there is a Romani word, it’s ash. AA: It’s ash? Avdi: ‘Ash.’ Yes. It’s said as ash. But it’s not known where this comes from. Kujtim Paqaku told me this example, that sebepi is said ash … The word normally used for cause/reason in Prizren Romani is sebepi, from Turkish. Avdi clearly accepted Kujtim’s authority about ash being the correct form, although he didn’t know where it came from. However, he questioned whether to use this Romani form, or kondiciya, an international loan (rather than a BCS or Turkish one) that would be more understandable to his listeners. In this case, he decided comprehensibility was more important than using the purest word, ash, but he still wanted to avoid the impure sebepi. Kondiciya was an effective compromise.11 While Kujtim, from his position of authority, was free to insist on more exotic, ‘purer’ forms, those working in media such as the radio station were in a more

Standardisation

107

difficult position, as they had to translate things at short notice, but also had to make them comprehensible to the Romani-speaking public, many of whom are less familiar with the idea of a standard Romani. For Kujtim, it is important to show his knowledge, and to speak the ‘purest’ form possible, while for those working in media it is necessary to negotiate a ‘standard’ form, which would grant the media authority but also be intelligible to the audience. Both these levels of language use (the higher, intellectual level, and the inbetween media level) are purist in their attitude to language, but in slightly different ways. While Kujtim’s ideal is a pure, preserved, isolated form, those working in media focus more on trying to expunge loan vocabulary specifically from surrounding languages. An English word is acceptable where a Turkish-origin word is not. Their construction of the standard is much more negotiated, with the aim of creating an educated, yet widely understood, variant of Romani. In sum, the everyday tactics journalists and intellectuals use for making new terms are creative and spontaneous. Certain people are able to learn ‘pure’ forms, and become Romani intellectuals, by establishing their place in the hierarchy. Kujtim’s authority over Nexhip, however erratic, is not just based on their respective experience in research, or their publications, but also Kujtim’s time with Marcel. What is crucial then is not just going to the periphery to find those isolated people, but also remaining connected to the centre, which is to say to Marcel and the group of intellectuals he favoured. Learning ‘pure’ Romani is not the work of a lone scholar. It cannot be done in the bath. It is necessary for the linguist to brave the wild people of the forests of Albania in order to ‘know’ what is pure, while also using one’s place in a social hierarchy to come closer to the western intellectual. The next section moves on to discuss the way these political wranglings affect speech practices.

Ideology and diffusion Nexhip once asked me about the possibilities of getting funding for a project which would help to increase attendance at Romani-language classes, but he wasn’t sure if the schools would agree to it. I suggested using the office, or a children’s centre. Nexhip was unimpressed with the suggestion and said the classes had to be held in schools – otherwise, they would be pointless. Not all the children who attend Romani classes know Romani, and their backgrounds vary widely. In class, those who know little or no Romani learn some basic phrases and words, while those who are already familiar are taught standard Romani words, and practise writing in Romani. This latter function is crucial: the aim is not to teach them a Romani orthography (Albanian orthography is mostly used), but rather for them to speak and write Romani in a classroom setting, and therefore to associate it with education, with formality, and to use it outside the home or mahalla. Writing Romani in the home, or in a children’s centre, would not be the same. The institutionalisation of Romani through education and media is a means of establishing it in the public sphere, rather than as a limited or domestic language. Language preservation and progression into the

108

Purity, mixture and representation

public sphere (rather than communication) is at the forefront of the standardisation of Romani in Prizren. In terms of diffusing the standard language described in previous sections, neither the classes, nor the drama, nor the radio appear effective. The plays do not reach a particularly wide audience, and despite Nexhip’s insistence that everyone listens to the radio, I was not aware of it being widely listened to. Low attendance at school classes means that they probably don’t have much effect in spreading knowledge of the standard language. They do, however, teach these children that Romani can be used beyond the home or mahalla, and that they can be Roma while in school. Selahadin’s textbooks may lack a clear methodology or curriculum, but their presence in classrooms is important for both Nexhip and the children. They act as a kind of artefact, showing that Romani can be used for formal education. Similarly while attendance at the dramas is limited, the fact of their existence is crucial to showing Romani can be a literary language. What is crucial to those involved is that Romani has a pure, standardised form, appropriate for use in the public sphere. This must be distinct from the mixed, non-standard form used in the home or mahalla. As such, the diffusion of a standard language ideology takes precedence over the diffusion of the content itself. Matras describes how ‘(t)he overwhelming majority of texts produced in Romani do not form part of coordinated language planning efforts. Rather they are the outcome of local and individual text-production initiatives’ (Matras 2002:254). This is certainly the case in Prizren where, despite the role of Marcel as an external reference point, local rather than international efforts are central to the formation of a standard. While with other languages the standards are already more or less set by state institutions, Romani standardisation is an ongoing process. This is not to suggest that other standards are fixed, but that there is at least a point of reference from which to debate standardisation. Correctness is managed institutionally, in Tirana or Belgrade, for example. By contrast, the selection and codification of Romani norms occur sporadically, at a local level, structured by competing authorities on the language. Haugen (1966:933) describes the various aspects of standardisation: ‘selection of norms,’ ‘codification of form,’ ‘elaboration of function’ and ‘acceptance by the community.’ All of these are ongoing in Prizren – though ‘acceptance by the community’ is least advanced. But while Haugen’s model assumes the existence of a codified standard prior to its dissemination, in the case of Romani in Prizren, processes are combined, and occur spontaneously. With the surrounding languages, enregisterment – the ‘processes through which a linguistic repertoire becomes differentiable within a language as a socially recognized register of forms’ (Agha 2003:231) – has occurred through state institutions. In the absence of state structures, the diffusion of Romani is performed by disparate governmental and nongovernmental initiatives and individuals. The classes, media and drama are all sites of the simultaneous selection and diffusion of norms. The vertical diffusion of standard Romani, that is, the attempts of those in institutions to inform people about the standard, brings Romani into the public

Standardisation

109

sphere, and spreads the idea that there is a pure, standardised, literary form of the language. This is intended to show not just Roma, but also non-Roma, that the language is equal to the surrounding languages. As such, it is indicative of the domain change undergone by the surrounding languages at an earlier stage (cf. Paugh 2012; Romaine 1994). The relationship of Romani in Prizren to its contact languages is complex. I would argue that contact primarily affects Romani standardisation in three ways. First, there is structural convergence with Romani’s contact languages (Matras 1998b, 2002:ch.8, 2011; Friedman 1984, 1985b, 1991, 2001). Second, those formulating the standard endeavour to differentiate it from those contact languages. This takes the form of lexical purism, as discussed above. Third, standard Romani itself is conceptualised in relation to other languages. This third point could be described as horizontal diffusion, diffusion of the idea of a standard across languages, as opposed to the diffusion of the idea of a standard vertically, from the intellectuals to the ordinary people. In this view, a standard is needed not just to qualify Roma as a nation with Romani as a national language, in the same way as the Albanians have Albanian language, the Bosniacs have Bosnian language and so on; the existence of the standard also shows that Roma, and their language, can be educated and cultured. While those involved in the media, drama and school classes are attempting to diffuse the standard to other Roma in Prizren, the ideology of a standard is being diffused from speakers of the surrounding languages. There is thus a kind of ideological convergence. The ideology behind BCS standardisation (and to a lesser extent Albanian) in the nineteenth century was a reaction to imperialist domination.12 This ideology has been re-formed and is now mobilised in response to the dominated state of Romani. In situations of contact, purism can be seen as a reaction to the hegemonic purism of the dominant language (cf. Hill & Hill 1980). While standardisers aim, in part, simply to have a standard language, as a cultural artefact, standardisation is also aimed at preserving and promoting Romani in the face of the specific problems faced by Roma in post-conflict Kosovo as described in Chapter 3. In Prizren, then, a metalinguistic discourse of what is correct and incorrect Romani is being spread. Language ideologies do not simply misrepresent actual speech practices: linguistic engineering can actually alter these practices (Golovko 2003). Terzimahalla Roma discuss their language use, and draw a distinction between gramatika, standard or pure Romani, and ‘the way we speak in the mahalla.’ Those familiar with the speech practices defined as standard debate and correct one another and me – as a language learner, who should learn the best variant. However, they only do so in the more institutional contexts of the Romani-language radio station, the Romani-language theatre and school classes. The ‘purest’ forms, promoted by Kujtim, are far from widespread. However, the idea that there is a way people should speak has diffused. There is a developing awareness not that everyone should speak the same, but rather that the way people speak in the mahalla is somehow inferior or inadequate, which acts as a way of distancing the standard from the vernacular (Thomas 1991). A standard is understood as prestigious, not merely uniform, and thus comes to appear neutral and

110

Purity, mixture and representation

common-sensical (Milroy 2001; Gal & Woolard 2014). Why it is correct or pure or standard does not need qualifying: it is legitimised by recognised authorities. This is not to say there is no diffusion of standard forms. In more formal settings, such as my interviews about language use with people such as journalists who had some exposure to the standard, I found it was common for people either to use an everyday word, then correct themselves and use a standard word, or to use the standard word followed by the everyday word, by way of an explanation. Although I often understood ‘international’ terms better, they still felt the need to offer a more usual, less prestigious form. When they wanted both to use the standard and to be understood, they used both. This corrective or explanatory doubling up of terms could be explained as an interim phase, whereby the intellectuals educate ‘the masses’ about the impurity of their language use, gradually bringing them up to the level of pure and educated Romani. While this may be the obvious Herderian explanation, I would suggest that this is not merely a phase, and that neither layer of terms is likely to die out soon. Rather, efforts to make a standard create a diglossia, whereby ‘incorrect’ mixed spoken forms exist in parallel with pure gramatika. That is, the ‘correct’ forms are used in institutions, while the ‘incorrect’ forms are more acceptable in the home. This diglossia, like the ideology behind it, is learnt contextually. Other situations, like a less formal discussion at the radio may allow either: even in institutions, the informal/incorrect forms are used in certain contexts. Standard Romani is discursively constructed not as something which needs to go through the process of selection and codification before diffusion, but rather as something which exists – a codified standard – but needs to be better diffused. When people talk about ‘not knowing our own language,’ they imply that Romani exists in a pure form but needs to be re-learnt and better disseminated, in the same way, for example, as the Tosk-based standard of Albanian does. Informal, mixed variants are contrasted with formal, distinct standards. In the case of the standardisation of Romani, the diglossia apparent in other languages, and the attitudes attached to it, are being replicated. Rather than domain change meaning a shift from the private to the public domain, requiring lexical enrichment, here the use of gramatika in the public domain has resulted in the bifurcation and stratification of the private domain and the Terzimahalla idiom. Making a standard thus also entails producing a non-standard; legitimising one type of speech delegitimises another, such that subalternity is reproduced at different scales. As with Corsican, where domain change resulted in a reproduction of dominant (French) language ideology (Jaffe 1999), a dominant language ideology in which speech practices are understood as hierarchical (and attached to the cultural capital of educational levels) is being reproduced within Romani. The gramatika (the educated, cultured variants) of the surrounding languages is contrasted to the local variants, perceived as mixed and impure. Impurity and peripherality become intertwined, and juxtaposed to the pure, standardised gramatika. While Romani does not have a physical centre in the way that the other languages do, I would argue here that by making a standard, and making

Standardisation

111

Terzimahalla Romani peripheral, that intellectuals are positioning themselves in relation to an imagined centre. The mixed, impure, peripheral variant is contrasted to the pure variant promoted by intellectuals who, by negotiating a position at the top of the social hierarchy, are also putting themselves at the centre of the standardisation process. Figures of authority are attempting to create, and diffuse, a standard. They also often insist that all forms of Romani are mutually intelligible. On one level, they seem to be attempting to create and spread an inclusive standard to people who speak a variety of different, but only slightly different, variants. However, by creating a diglossic situation with two unequal variants, they are in fact forging new forms of inequality in language use. Whereas previously knowledge of another national literary language would create and show social stratification of Roma, by introducing this diglossic situation in Romani they are, albeit unwittingly, creating a layer of inequality within Romani itself. By making the everyday Romani spoken in Prizren more peripheral, they are positioning themselves as gatekeepers of purity and correctness. This is not of course to say that they intend to increase negative perceptions of Roma in lower socio-economic positions, but rather that the generalisation of the language ideology in the area reinforces such perceptions.

Conclusion The standardisation of Romani in Prizren is a response to a particular set of historical circumstances (cf. Jernudd 1989; Annamalai 1989). As previous chapters have shown, there have been substantial changes in ethnic, class and language hierarchies. The Roma response to these shifts is clearest with regards to orthography, where there has been a wholesale rejection of anything that looks Serbian, especially Cyrillic, but also the diacritics used for Latin alphabet Slavic languages in the region (Halwachs 2011). But while Slavic orthographies have been rejected, Albanian orthography has not been fully adopted. Those interested in raising the profile of Romani still want to keep Romani distinct. Likewise, with lexical purism, the focus on the expulsion of foreign words does not extend beyond the linguistic repertoire of Terzimahalla, and words from English or Romanian are in fact actively recruited into the new lexicon. The creation and institutionalisation of this standard opens up a level of choice for Romani speakers: they can further align themselves with Albanian authorities and employers by bringing their children up speaking Albanian, or they can enlist with Romani-language standardisers, and enhance the prestige of Romani. Those involved in standardisation wish to overcome a purported choice between marginalisation, indexed by non-standard Romani, and assimilation, associated with Albanian. This parallels the predicament of many African Americans who often feel they have to ‘talk white’ in order to succeed in prestige fields, especially in education (Young 2004). One response has been the proposal of a written, institutionalised Black Vernacular English, which, like standard Romani, would show that one can be both non-white and educated. The social and linguistic situation

112

Purity, mixture and representation

in the United States is very different, but the choice is similar: the subaltern group can either reinforce a dominant ideology which discriminates against them by using the linguistic resources of the dominant group at the expense of their own, or institutionalise their own variants, but thereby reproduce the division between good and bad language, institutionalising new forms of dominance on the model of those that existed previously. A focus on class, rather than ethnicity, would allow minority groups to improve their situation economically, via the majority language, rather than supporting minority languages at all costs (Hodges 2016). While minority language recognition may be important, the skill of having a good command of the right variant of the dominant language is far more important in negotiating scarce opportunities for employment. Minority languages often suffer when compared with dominant languages for this very reason. Despite attempts to introduce language equality into education, the history of oppression and the residual economic inequality stemming from this mean that minority languages continue to lose out (Tupas 2014). However, the commodification of language does not always imply the hegemony of the dominant language: Heller (2003) describes the case of Canadian French, which she argues has shifted from being primarily a marker of identity into a marketable skill. Making a standard language requires making it marketable, and also a skill which can be learned and can differentiate the possessor of this skill, preferably showing they have invested time or money on acquiring it (Silverstein 1996). For Roma in Kosovo, it may be more worthwhile to invest time and money in acquiring the dominant state language, Albanian, or an additional foreign language. Learning Albanian is a useful skill, beneficial for employment purposes, but raising the prestige of Romani could eventually transform it from a marker of identity into a marketable skill beyond the limited confines of the radio station and RTK. The former is of more immediate and dependable benefit to the household, while the latter holds a more distant potential for Roma more generally. Were Romani to become further institutionalised in Kosovo, standard Romani would be in demand, as a skill for teachers and journalists, for example. Yet poorer Roma are more likely to speak Albanian as a first language, and are less likely to attend school. While they may be able to show allegiance to Albanians by distancing themselves from Roma and Romani language, they are unlikely to climb the rungs of the new Roma social hierarchy by speaking standard Romani and becoming professional NGO workers, teachers or translators. They therefore stand to gain less by supporting this process. Those currently involved in language standardisation have a vested interest in the reproduction of a distinction between correctness and incorrectness, purity and impurity. By reproducing this distinction, they show that they ‘know their own language,’ and are in a position to judge others. Romani has been described as decentralised, pluralistic and polycentric (Matras 2002; Friedman 2005). From the perspective of the international Romani movement, the standard being produced is fragmented, and prioritises local goals over the wider Romani movement (cf. Matras 2013). However, within Prizren it is also extremely centralised, in that the process assumes a perceived centre from which language ideology authority

Standardisation

113

flows. Leggio (2013) has discussed the importance of the internet, with its lack of centralised power, in building a bottom-up rather than top-down approach to writing Romani. In the case of Prizren, I would argue that while this process is decentralised on the international scale, at a local level it is highly centralised and hierarchical. Domination is produced locally, and the standard language both legitimises and is legitimised by this domination.

Notes 1 For more details, see the report by Sveriges Radio (2014), detailing the conference of Partia Rome e Bashkuar e Kosovës (PREBK), the Ministry of Education, Science and Technology and various intellectuals. 2 As well as other works of his such as O Roma & i Romani Qhib (Kruezi 2014), a selfpublished monograph discussing the origins of Roma and Romani language. 3 See Halwachs (2011) for a discussion of the writing used in Yekhipe, a magazine from Prizren. 4 There are still accusations of Roma being closer to Serbs and BCS: see KTV (2014) for a news item that accused Roma children of attending Serbian schools and not speaking Albanian. 5 Interestingly, this is the only time I've heard the word znaqinela used. Prizren Roma generally use the Turkish demek. I can only assume znaqinela was chosen as I asked the question in BCS, so she chose, or possibly improvised, a BCS loan. 6 The use of may to form a comparative is itself a standard form, from Romanian mai. The usual Terzimahalla form would be po shuzhi or po laqhi, from the Slavic. For a discussion of the borrowing of comparatives and superlatives, see Matras (2002); cf. Friedman (1985b). 7 These numbers refer to the grading system at schools, with 5 being the best and 1 the worst. 8 See also Thomas (1978) for a discussion of purism in Croatian, and Thomas (1988) for a broader discussion of purism in Slavic languages. 9 For a discussion of the use of Hindi neologisms and the Devanagari alphabet in Romani on Wikipedia see Leggio (2013). 10 Courthiade’s standardised Romani uses ‘archegraphemic symbols,’ such as and , to be pronounced variously according to the dialect of the speaker (Matras 2002; Leggio 2013). This can be compared to the suggested use of jat to codify the variations within štokavski BCS in the nineteenth century by the Illyrian movement. 11 I’m not sure why kondiciya was used here. He may have mixed up his use of Englishorigin terms or the term may have shifted meaning. 12 For a discussion of this process with regard to Turkish, see Heyd (1954); with regard to Vuk Karadžić’s codification of Serbian, see Wilson (1970); with regard to the Illyrian movement’s codification of BCS, see Despalatović (1975).

6

Before the war A nostalgic speech genre

Introduction Roma from Terzimahalla partake in several spheres of belonging, one of which is that of urban Prizrenites. When Prizrenites talked to me about cosmopolitanism, they linked it to a particular way of speaking Turkish and a local sense of sociality exemplified by the plurality of people and languages in Prizren. This way of speaking Turkish can be interpreted as a speech genre where despite the fact that ‘(e)ach separate utterance is individual […] each sphere in which language is used develops its own relatively stable types of these utterances’ (Bakhtin 1986:60). This genre, and the local dialect of Turkish, is associated with the idea of a harmonious, integrated elite, with the central square Shadervan and with religious festivals. Integration, often used to mean the integration of certain people at the expense of others, is understood as spatial, as well as social. It also carries implications of wholeness and boundedness. For Roma from Terzimahalla, being close to the centre means that they are integrated spatially and have more contact with other ethnic groups. However, it is also a way of differentiating themselves from other people, primarily recent migrants from rural areas. These migrations have been accompanied by a shift in linguistic hierarchies with Albanian now dominant, and a shift in the way speech genres are used. Linguistic anthropologist Heller (2012:31) argues that we need to ‘move away from looking at people as composed of inherent characteristics, or possessing inalienable traits, and look instead at what they are and are not able to mobilise, as well as why they might want to do so.’ Throughout this book I have emphasised that, while people in Prizren have a broad linguistic repertoire, what they are able to mobilise through this repertoire is embedded in the requirements of the broader political economy. Looking at how and why the importance of linguistic resources shifts shows that language mixing and metalinguistic discourses about language mixture nostalgically relate to a lost system of power relations.

Language boundaries and contentment Chatting in English with Galdim, a young Albanian man from Prizren, the topic of languages came up:

Before the war 115 AA: So I guess you don’t speak Serbian … Galdim: Ne [BCS: ‘No’] … (He laughs). AA: And Turkish? Galdim: Yeah … well, some kind of Turkish. We say things like şpiden opştinaya cittim [I went from home to the municipality]. You’ve heard (He laughs). In the sentence şpiden opştinaya cittim, the sentence structure (noun endings and verb) is Turkish, but with the palatalization of the local accent, the verb gittim becomes cittim. The word for home, shpi, is a local variant of the Albanian shtëpi, and the word for municipality is from the BCS opština. This sentence intended to show me how mixed Prizren Turkish is. This is an example of metaphorical mixing (Gumperz & Hymes 1972): the speaker uses his home language, Albanian, for ‘home,’ and the former language of the state, BCS, for a state institution,1 and Turkish, the language of interethnic communication in the city, for the sentence structure. The sentence is indicative of different domains of language use. In this respect it is similar to Poplack’s example ‘Sometimes I’ll start a sentence in Spanish Y TERMINO EN ESPAÑOL’ (Poplack 1980). This is also metaphorical, in that the choice of language (and of switching) describes the content of the language. However, Poplack’s example is self-referential, the languages referring to the sentence itself, while in Galdim’s case the words used relate to the domains of language use. Both examples are not simply unconscious utterances, but part of a conscious and intentional discourse about how language is used. Galdim was not just mixing, but was implicitly showing how he perceived mixing in Prizren. People do mix Turkish like this in everyday speech, and with regards to a wide range of topics, not just language itself, but they are also acutely aware of this as a mixed form.2 When people talk about mixing languages, they often mock it affectionately, mentioning the way languages are mixed in individual conversations, or the way the BCS and Albanian they speak is full of Turkish words, while the Turkish they speak is full of Serbian words, and the Romani has words from all these languages. This is done with a combination of pride and derision. One young Turkish speaker who had gone to study in Istanbul complained to me that Prizrenites use Serbian words in Turkish on purpose. He said they knew proper Turkish but obstinately preferred their own variant. Clearly, knowing what he considered to be ‘proper’ Turkish, this Prizren variant offended his sense of linguistic purity.3 Balkan Turkish, though mutually intelligible with Standard Turkish, has a series of phonological, lexical and morphological differences (Friedman 1982). The structural convergence of Balkan languages that affects Turkish can be described in terms of ‘mutual isomorphism,’ the tendency for sentence planning procedures to synchronise, rather than a one-way ‘borrowing’ of linguistic matter from one language into another (Matras 1998b, 2004). A long history of societal multilingualism in the Ottoman Empire has affected the structure of Balkan Turkish, while more recent Yugoslav history has resulted in a layer of domain-specific lexical borrowing from BCS (and Macedonian) into the variants of Balkan Turkish spoken in former Yugoslavia (Matras & Tufan 2007).

116

Purity, mixture and representation

People regularly comment on the number of languages spoken, and the fact that they are mixed, as a testament to the multiculturalism of the town. While the term multilingualism does not carry with it any of the political intention or essentialist implications of multiculturalism (and is in fact rarely used in the town), the very fact that people enumerate and label languages shows they are perceived as distinct entities which can be mixed. The discursive deployment of the above sentence, its use as a metalinguistic commentary on language use and mixture, suggests an awareness both of the idea of a pure Turkish, and of the foreignness of Albanian and BCS words within the sentence. While mixed practices are commonplace in Prizren, they made things difficult for me when learning, as spontaneous or unexpected uses of words confused me. At the children’s centre, there was a room for children to do homework, as many lacked the space to learn at home. They were helped by Adem, a Roma pensioner, and sometimes me. One girl, a Romani speaker around seven years old, recognised me from the Romani-language class, and wanted me to help her with her Albanian homework. When we finished what she was meant to do, she turned to me smiling and said kenaqisalum. I was confused. I recognised the ending (as a Romani first person past ending), but not the root, and I didn’t understand what she wanted to say. An older girl, around 12 and also a Romani speaker, thought for a moment, shrugged, and said oy si bahtali (Ro: ‘she’s happy’). It was clear she didn’t think this was quite the right word, but an approximation. For them, kenaqisalum didn’t sound unusual in Romani, but expressed something slightly different from the Romani bahtali, while remaining in Romani language, as the language she was most used to talking to me in, and the language which it was easiest for us to communicate in. The Albanian root kënaq translates into English as ‘contented,’ ‘satisfied,’ ‘pleased’ or ‘glad,’ but I didn’t recognise it in a Romani structure. The Romani bahtali would most likely be translated with the Albanian i/e lumtur (‘happy’). This use of mixing may also have been related to her writing task being in Albanian, or the particular dialect of Romani she spoke, but either way, her use of kënaq in Romani was clearly acceptable, and not felt to be transgressing a boundary. While Romani purism is crucial to intellectuals in specific contexts, as we saw in Chapter 5, in other, informal circumstances this kind of mixture is acceptable. This lack of differentiation, or acceptance of difference, is not the case for all languages and all situations. One evening at Vera’s house, her teenage daughter, who speaks primarily Albanian at home, smiled and said something to her mother in Albanian, using the same word kënaq. I didn’t understand, so Vera tried to explain to me, mistakenly saying in BCS ona bi kenaćila da bi ti bila njena sestra (‘she would be kënaq if you were her sister’). They both burst out laughing, and found this so funny that they mentioned it again several times. Vera had mistakenly used the Albanian kënaq in BCS – this was unusual, and entertaining. The latter example was clearly perceived as a switch which crossed a language boundary: that is why it was funny. In the previous example, there was no reaction to the use of kënaq in Romani; the speakers clearly did not see it as transgressing any boundaries. This suggests two possible interpretations: it

Before the war 117 could either be analysed as a borrowing, or as an instance of the bilingual mode. Borrowing entails the content of one language becoming embedded in another language. As described in Chapter 5, loanwords do not always meld into the language imperceptibly, and are often very obviously borrowed. What distinguishes them from switching is their systematic and regular use in the language they are being borrowed into; if I had collected linguistic data to show regular, systematic use of kenaqisalum in Romani, it could be defined as a loanword. As it stands, this remains unknown, and this may be a one-off usage. Regardless of whether kenaqisalum has been fully adopted into Romani, this form (coupled with the knowledge that the girls also spoke Albanian) could also be explained in terms of bilingual mode: a state where both languages known to a bilingual are activated (Grosjean 2008). The metaphor of bilingual mode, and language activation, however, is still based on an assumption that languages can and do exist as distinct entities. This may be the case if her language use is domain-specific, but becomes problematic if this is how she always talks. The concept of bilingual mode makes most sense for those who also have two separate monolingual modes. If the girl always uses mixed forms likes this, then it would be better described as a monolingual mode with borrowings. When people in Prizren say their language is mixed, this tends to relate to the insertion of words from another code, rather than switching either within or between sentences. In linguistics literature ‘codeswitching’ and ‘codemixing’ are used variously (Matras 2009; Crystal 2008). Here I use codeswitching to refer to all instances of language alternation within a conversation, and mixing to refer more generally to the way languages have, and are perceived to have, blurred boundaries. Hence, defining the examples above as codeswitching would imply an acknowledgement of the existence of codes as discrete entities, rather than a more general fluidity of usage. Mixing requires either the elision of a boundary, as with the girl’s use of kenaqisalum, or its transgression (intentional or not), as with Vera’s use of kenaćila. The process of borrowing can be seen as the elision of a boundary between subsets in a linguistic repertoire (Matras 2011), while the elision of a boundary between separate domains can lead to the conventionalisation of certain types of mixing (Heller 1988). Indeed, with no clear demarcation between languages, the distinction between borrowing and codeswitching itself breaks down. Only when languages are treated as discrete entities can borrowing be distinguished from codeswitching. Mixed forms, whether or not they are embraced as mixed forms, can become conventional (cf. Garcia & Mason 2009; Golovko 2003). Where this is the case, it no longer makes sense to distinguish borrowings from switching, nor does it make sense to distinguish monolingual from bilingual modes. Rather, in this instance conventionalised speech practices can be understood by speakers (and analysed by linguists) as mixed. Where a boundary is present, and actively transgressed, the effect is very different. With Vera’s use of kenaćila, an accidental transgression, the effect was humorous. The unexpected juxtaposition of BCS and Albanian made it funny. The more impenetrable the boundary between the two varieties is felt to be, the

118

Purity, mixture and representation

more transgression of this boundary can be considered humorous, or even dangerous. This goes for both languages and variants: Bradley describes how Yanyuwa (an endangered aboriginal language) has a different dialect for men and women, whereby the dialect of the opposite sex may only be used in ‘situations [which] are somewhat ribald and risqué and full of humour’ (Bradley 1988:131). In this case the speech practices and ideology attached to gender difference are strictly policed. The distinction between the two dialects is held to be impenetrable, and crossing it is seen as inappropriate. The use of inappropriate forms can be amusing in some cases, but in other contexts may be considered offensive or dangerous. The example of Vera’s use of kenaćila was unlikely to have elicited any strong reactions when uttered mistakenly among three women, none of them Serb or Albanian, in a private and supportive setting. But given the politicisation of language in Kosovo, it would be easy to imagine a situation where this would not be the case. The various people involved in conversations may not always agree on what constitutes a transgression, and where such a transgression moves from humour into danger. In Romania, using Hungarian in what is perceived as an inappropriate setting is not humorous, but a cause for complaint and policing (Brubaker 2006; cf. McClure & McClure 1988). This varies according to the setting and the audience, and established norms of appropriateness. The humour of unexpected or inappropriate mixes can therefore be problematic, as well as entertaining, and not knowing the rules of appropriateness can risk exposing oneself to ridicule or anger. Conversely, linguistic ambiguity and humour (e.g. through codeswitching) can act as a way of ‘softening’ disputes (Williams 1987). Humorous juxtaposition can also be used intentionally (Garrett 2005). This may involve, for example, the use of a low variety in a conventionally high context, or vice versa (Romaine 1994). The switch itself can also be funny: if a sociolinguistic rule does not call for switching at this point, then the violation of this rule can be humorous (Siegel 1995). Rather than the absence of boundaries being a point of humour, it is the explicit overriding of them that creates humour, and highlights the boundary itself (Woolard 1988; cf. Goldstein 1990). At the children’s centre, the girl’s use of kenaqisalum was not funny, because in that context boundaries were not conventionalised to the extent that they could be transgressed. In situations of societal multilingualism there are clear ideas about the contextual appropriateness of varieties, and of mixing, and breaking these can be funny. My informants conceptualise their linguistic repertoire as comprising distinct languages, each with its own grammar and vocabulary, as discussed in Chapter 5. However, this does not mean that they keep languages separate. Overtly purist ideologies do not always imply separateness in practice, and speakers may not always perceive the separateness of their codes (Minks 2010; Jaffe 1999). There is no simple dichotomy between the pure, unmixed languages envisaged by a monolingualist ideology (Blackledge & Creese 2010) and the fluid or mixed variants in everyday use. Rather, certain contexts are embedded within an ideology of mixture, connected to mixed variants. Acknowledging and even promoting

Before the war 119 language purity does not prevent language mixing, but rather means that in certain contexts mixing is perceived as such. The use of local loanwords discussed in Chapter 5 was understood as a product of not knowing one’s own language; using an Albanian word in a BCS sentence at a family meal was understood as the transgression of a language boundary; the use of an Albanian word by a child speaking Romani was understood as utterly unremarkable to all except me. People’s ideological stance towards impurity in language is not the outcome of an intimate knowledge of etymology or historical linguistics (see Chapter 5). Rather, people are sensitive to where words ‘belong,’ hence the desire to purge recognisable loans. While it is likely that kënaq would be recognised as Albanian even when used in Romani, it is still not perceived as intrusive in certain contexts. Language boundaries are ‘not natural or utterly outside of human will and political forces’ but are socialised systems of distinction (Gal & Irvine 1995:992; see also Gal 1989, 2006b). As such, language boundaries are historically and ideologically constituted; speech practices and attitudes towards hierarchies between speakers and contexts can both reinforce and challenge the way languages become categorised as separate. If Romani children using Albanian words in Romani in the classroom was treated as perfectly natural, while the mixing of an Albanian term into BCS in another context was hilarious, this relates to the historical functions of BCS and Albanian in Prizren. BCS became extremely restricted after the war, because of its connection to the Milošević regime, and because of the exodus of the Serbs. At the same time, Albanian took over from Serbo-Croat as the official language of administration and, to a certain extent, is taking over from Turkish as the language of inter-ethnic communication. The use of Albanian words in BCS, then, was something of an anomaly. Both as Serbo-Croat (in its eastern variant) and as Serbian, BCS has long been highly centralised and standardised, with a concrete idea of correctness. Coupled with the socio-economic segregation of BCS and Albanian speakers before the war, and geographical separation after, the boundary between the languages has been upheld, and mixing has not become usual or conventionalised. This contrasts starkly with the continued acceptability of using Turkish, BCS and Albanian words in Romani (outside the formal or intellectual context), and the use of a mixed variant of Turkish, now in decline. This decline relates to the change in power relations and a demographic shift, and thus evokes comparisons with ‘the good old days before the war,’ which will be discussed in the next section.

Before the war: a nostalgic genre The ‘Horn Knives’ gift shop in the centre of Prizren, just over the stone bridge from Shadervan, has a glass front, displaying an array of horn-handled knives, traditional coffee pots, Turkish delight pots, ancient jugs and so on. When I went in, realising I was a foreigner, a teenage boy spoke to me in limited English until an adult came, presumably his father or uncle. He showed me various things in the shop and the young boy left, at which point I told him we could speak in

120

Purity, mixture and representation

Bosnian if it was easier for him. As I had found older people in Prizren generally spoke BCS well, and were happy to speak it, I thought I would ask. It clearly was easier for him, and after interrogating me about where I was from, he started to tell me about Prizren, and how great it was before the war. ‘We’re very cultured here, we speak all different languages’ he said, a comment that was often used to excuse or explain speaking BCS. ‘It’s not like other places, here in Prizren we are multicultural, harmonious, you can see all different people, hear all different languages, there’s the churches, the mosques ….’ He didn’t mention the attacks on the churches or the Serbian houses (they were never mentioned explicitly). He concluded, ‘but you should have seen what it was like before the war, how much harmony there was here then ….’ This narrative was very familiar to me. Many of those old enough to remember the pre-war period, of all ethnicities, told me how things were better in the past: things were harmonious, people respected one another, and, in short, people were happy and contented. However, people did not always describe contemporary Prizren as unhappy or lacking in harmony. In fact, when comparing Prizren to other places in Kosovo, they did quite the opposite. Prizren was described as the cultural capital, and as being more tolerant and multicultural, evidenced by saying ‘we speak three or four languages, but this isn’t important because we are all Muslim,’ or ‘we speak three or four languages and have both mosques and churches.’ Though it was certainly not the case that one ethnic group spoke one language, enumerating languages was still used to show the number of ethnicities. The different understandings of religious plurality point to two different systems of inclusion and exclusion. The first sees religious (Muslim) unity as more important than ethnolinguistic difference, implying a contrast with people from rural areas, who are seen as the monoethnic, irreligious other. The second points to religious variation in Prizren as positive in contrast to other cities in Kosovo. The emphasis on churches4 and mosques as symbols of a (now mostly lost) religious coexistence speaks of an implicit understanding that tolerance of churches means tolerance of Serbs and hence a lack of conflict. This variant of cosmopolitanism is mobilised in contexts when people compare Prizren to other parts of Kosovo, especially Gjakova. People’s attitudes to ‘places where there had been war,’ more specifically places where Albanians had been massacred by Serb forces, are generally sympathetic yet condescending. They feel for their suffering, but are keen to distance themselves from such areas by emphasising their own tolerance and culturedness. Both during and after the war there have been a series of events in Prizren which undermine this narrative of multicultural harmony. As well as NATO bombing, there were attacks on Albanian heritage sites (most importantly the League of Prizren building), the burning of Serbian houses and churches in 2004 and the flight of the Serbian population (in 1999 and 2004) along with some Roma and Turks (see Chapter 2).5 Despite the violence and the highly visible damage to churches and houses, in comparison with Gjakova, people still feel able to talk about Prizren in terms of harmony: this dominant narrative overwrites other possible accounts of the past.

Before the war 121 The vague period ‘before the war’ is understood as a time that was better, which has passed. As described in Chapter 3, a large part of this relates to the employment that was provided under the previous system. Yet it is more than this. People don’t just talk about jobs and welfare, but also about social relations, morality and contentment. Despite nostalgia for the Yugoslav period, people are not explicitly nostalgic for Yugoslavia: there are no Tito-themed cafes; Yugoslav music is not listened to in the way that it is elsewhere in former Yugoslavia; nor is there the nostalgic celebration and commodification of the symbols of socialism seen in some other post-socialist areas (Petrović 2007, 2010; Jansen 2008). Due to the persecution of Albanians under Milošević (seen as a continuation of Yugoslav socialism), and the subsequent victory of the KLA, Yugoslavia and socialism are topics which are avoided, if not denounced. But people do describe the pre-war period nostalgically. Though commonly viewed in terms of brutality and oppression, it is simultaneously celebrated. In part, this belies the coexistence of a variety of narratives which can be mobilised at different times to stress both violent difference and peaceful coexistence (Megoran 2013; Reeves 2005; Klumbyté 2010). These different narratives are shaped by different power relations, and also interact differently with people’s lived experiences: the glorious liberation from Serb oppression in the Albanian nationalist narratives does not fit with the life-stories of many Prizrenites. Nostalgia is often more complex and fragmented than a simple celebration of all things past, and people can celebrate some elements of the past without others, even where they appear contradictory. In the case of Prizren, the elements of the socialist period which people are nostalgic for are not Yugoslav as such, but rather are a specific version of cosmopolitanism which excluded peasants, and afforded urbanites greater employment opportunities, as we saw in Chapter 3. Unlike Klumbyté’s (2010) argument that ‘Soviet’ brand sausage is popular in Lithuania despite anti-Soviet feeling, Yugoslav symbols are not (and most likely would not be) celebrated in Prizren. The idea of a mixed and harmonious town before the war is no longer related to the Yugoslav idea of ‘brotherhood and unity.’6 Rather, people talk about a time when they were contented, and things were as they should be. Before the war, there were factories, but in addition to having more material privileges than farmers, Prizrenites also enjoyed a specific form of sociality, and their affection for the period goes beyond employment. The narrative of prosperity and peace elides the economic decline, unemployment and political repression of the eighties and nineties. This nostalgic idealisation of the past was regularly evoked by my colleagues and other people I chatted to in Prizren: they described the past as a time of harmony, of sociality, of religion and of love and respect. While the war itself was rarely discussed, for reasons explored in Chapter 2, the periods before and after it were treated as distinct, and compared and contrasted.7 These kind of narrations of the past are of course not limited to Prizren, but take on a specific form here, one which sees mixture and harmony of peoples as good, while seeing the past as qualitatively better in that respect.

122

Purity, mixture and representation

These narratives of the past overlap with discussions of employment in the past, and there is often slippage between the two. But when people move from talking about the material aspects of the past to more typically nostalgic ideas about the sociality and morality of the past, a poetic tone, and a different speech genre is used. A recent poem, Nerde o esçi cünler describes nostalgia for the (undefined) past in Prizren, while complaining about how children today spend too much time on phones and computers. In 2014, it was made into a YouTube video (Üsküp Makedonya 2014) and shared by my colleagues (and presumably many others) on Facebook. The video shows contemporary pictures of old buildings in Prizren through a sepia lens, emphasising not the past itself, but the present view of the past, while one of the authors reads the words of the poem with wistful Balkan music playing in the background. It is in the Prizren variant of Turkish, with borrowings from both BCS and Albanian. The text below is taken from the YouTube page and the orthography represents the local dialect. It can be assumed that the authors were conscious of their choice of this variant to index affection for a bygone era. Otuturduk familyarno, Duyardin ondan bi şaka, bundan bi vitz, Patlardık cülmekten. Kınaç be olurduk, Aaah nerde o esçi cünler nerde! We sat as a family, You could hear a joke from one person, a gag from another, We exploded with laughter, We were contented, Aaah what happened to the old days, what happened! Later in the poem, there is a focus on the way Ramadan and Bayram were celebrated in the past, with Nerde o esçi cünler! (‘What happened to the old days!’) being repeated as Aaahh! Nerde o esçi Ramazanlar! (‘What happened to the old Ramadans!’) and Nerde o esçi bayramlar! (‘What happened to the old Bayrams!’). This relates to a conversation I had, when one of the older Roma NGO workers told me: ‘Before the peasants came down from the hills, Ramadan was celebrated properly, you wouldn’t believe how full Shadervan would be after iftar …8 those people from the villages don’t even know Allah!’ Vera added that before, the food was better, people made their own food, now it was easier to just buy something at the shop. She also told me she had Facebook messages from her Serb friends (who had fled in 1999 or 2004, but stayed in contact) wishing her well for Ramadan, and reminiscing about the Ramadan pitajka9 they used to have when they still lived in Prizren. In these descriptions, the changing in observance of religious festivals, the foods and rituals associated with them, and the harmony of the past are lamented. Despite the spread and partial acceptance of new ideas about Islam, it is still common for people to discuss how it was better before. Unlike the debates

Before the war 123 surrounding religious doctrine described in Chapter 4, here my informants were concerned with a different kind of sociality. This was not a contestation over true Islam, but rather over a kind of contentedness and stability linked to religious festivals, that was felt to be being lost.10 The fourth line, Kınaç be olurduk, ‘we were contented’ uses the Albanian word kënaq discussed in the first section of this chapter. Here the loss of this feeling of contentment is indexed by the use of mixed Turkish associated with pre-war cosmopolitanism. The time before the war is remembered as contented, stable and without upheaval. The loss of this feeling, relating both to the anti-cosmopolitan nature of ethnic war itself, and the perceived invasion of the stable urban space since the war, is comparable to the loss of a feeling of normality described in wartime and immediately post-war Sarajevo (Maček 2007), and in post-Milošević Serbia (Jansen 2009; Greenberg 2011). In Prizren this memory of contentedness and normality also relates to an idea of safety. Speaking to an Albanian taxi driver in BCS again evoked a discussion about before the war, and about harmony in Prizren. While Turkish would be more usual for such reminiscences, the relatively unusual use of BCS seemed to impress a need to show tolerance of other ethnicities and other languages, leading to a discussion of pre-war harmony. It seems unlikely that speaking in Albanian would have had the same response. He described how much better things were before the war, when everyone got on, living harmoniously, and there were no thieves. As Henig (2012) describes in Bosnia, the idea of being able to leave your door open before the war implies a lost sense of communal living, as compared with the privacy and suspicion of the present. This lament was not for the war itself, but for the change in the town, its attitudes and values, which left a feeling of insecurity. Understandings of the past in terms of contentedness, and religious festivities, are tied together by a lost feeling of safety, harmony and community, and these changes are felt more acutely around repeated rituals like Ramadan. As the poem continues: Yok artık o, o adablık, o dostluk, o birbirimiza o saygi, birbirimiza o sevgi It is no more, morality, friendship, respect for each other, love for each other. The poem ends by stressing the past is not returning: Aaaah dostlar! Nerde o esçi cünler? Celecek mi bi daa ceriye? Allah yardımcımiz olsun … Oh friends! What happened to the old days?

124

Purity, mixture and representation Will they return? Allah help us …

People’s nostalgia for the time before the war relates to a kind of sociality specific to Prizren. What is remembered positively is not ‘brotherhood and unity,’ linked to the Yugoslav socialist state, and Serbo-Croat, but rather a time when people were contented, where social relations were based on a common respect for religious ceremony, and where a peculiarly Prizrenite form of Turkish was spoken widely by ‘everyone,’ regardless of their ethnicity and home language. This ‘everyone’ means those present in the public space of Prizren, generally, urban adult males. This distinctly Prizrenite mode of nostalgia is not incompatible with Yugoslav ‘brotherhood and unity,’ nor with older Ottoman forms of cosmopolitanism, but reverberates differently when fitted into these larger scale forms of identification. Social changes are often intricately tied up with linguistic changes, and as such they can become ideologically inseparable. In discussing the nostalgia evoked by language shift, Hill (1998) finds that a shift from Mexicano to Spanish nostalgically evokes an idea of loss of respect, linked to the loss of honorific greetings in Mexicano. Similarly, Hill and Hill (1980) discuss the case of Tlaxcala language in Mexico, finding that urbanisation and a shift in social structures has led to a narrowing of the function of Tlaxcala, with it increasingly becoming a marker of ethnic solidarity. While language use is never neutral, a sudden, dramatic social change can highlight what is at stake with languages. In the case of Prizren, I argue, there has been a change in speech genre (Bakhtin 1986; cf. Garrett 2005): Prizren Turkish was previously a neutral genre among men in public space in Prizren. Neutrality itself can be a genre, and one that requires the unity of the people involved in the conversation (Bakhtin 1986). Once the position of the genre has changed, along with the unity and status of those involved in using the genre, it no longer has such neutrality. For Bakhtin ‘(e)ach utterance is filled with echoes and reverberations of other utterances’ and ‘must be regarded primarily as a response to preceding utterances of the given sphere’ (Bakhtin 1986:91). In the case of this variant of Turkish, formerly neutral among an exclusive group who have since lost the ability to assert exclusivity, the linguistic forms have come to form a new genre, that of nostalgia, filled with the ‘echoes and reverberations’ of the previous genre, and with a sense of its loss. The speech genre of nostalgia thus requires Prizren Turkish. Language use, then, does not just evoke, or represent nostalgia, but in fact can play a key role in its construction.

Sedentarist cosmopolitanism Once as we sat eating lunch at the radio, Nexhip asked if I knew that each Roma household had €15,000 worth of gold. This was given to women when they married, as a kind of security, and it was handed down through the generations, so his wife’s gold would go to his daughters, but in the meantime it was hers. Then he turned to Bajram and said that Bajram’s household had more, maybe €25,000,

Before the war 125 and by way of an explanation added that Bajram speaks Turkish at home. Bajram, usually quiet and dignified, looked embarrassed for a while, but then grinned and added a za srebro, ne pitaj (BCS: ‘don’t even ask about the silver’). Nexhip’s casual explanation suggests an implicit link between Turkish, gold and prestige, which he treated as natural, requiring no further explanation. People are aware that their Turkish is mixed, and differs from the Istanbul standard which they hear regularly on television, but it is still an important local measure of prestige. In fact, I was told that people intentionally used this Turkish to exclude ‘peasants.’ ‘Peasants’ could refer to rural inhabitants visiting the town, but also to post-war migrants. In this sense, living in a city is not enough to make someone ‘urban’; rather, this is a social distinction based on beliefs and way of life (Duijzings 2010). In fact, it could be argued that the incursion of the peasant into the city is what makes them a peasant in this ideological sense, rather than merely an inhabitant of a village or farmer. In town, their otherness becomes apparent. The fact that many people migrated to the town from Albanian villages after the war has meant that this period is seen as an invasion of the city, by peasants, disrupting the contented, civilised, harmonious Turkish speakers. For Roma, this feeling of invasion, a disruption of the stable, integrated mixture of the town, is projected not just onto Albanian peasants, but also rural, Albanian-speaking ‘gypsies,’ many of whom increasingly identify not as Roma, but as Ashkali. While most Terzimahalla Roma do not accept the ‘new’ categories of Ashkali and Egyptian (see Introduction), they do on the whole differentiate themselves from these people, often by co-opting the derogatory terms and stereotypes used by non-Roma to construct a hierarchy among Roma, Ashkali and Egyptians. As mentioned, they differentiate themselves from the ‘peasant’ other (self-ascribed Ashkali), and the ‘intolerant,’ ‘uncultured’ other (self-ascribed Egyptians) through the use of the derogatory terms cigan, maxhup or gabel. In other circumstances, Roma consider these terms offensive, but here they are used to claim exclusive urban membership, while denying ethnic difference to Ashkali and Egyptians. These cultural hierarchies, which I have explored in Chapters 3 and 4, are also embedded in perceptions of mixture. Fatmira once told me the perceived bad behaviour of the children at the children’s centre where I was volunteering was caused by them ‘having no culture,’ rather than poverty. ‘Having no culture,’ contrary to my expectations, did not relate solely to lack of formal education, but rather to the social hierarchies in the town, and the way they were spatialised. Further out, the mahallas are poorer and less connected to infrastructure, and are partly inhabited by post-war migrants. Their recent arrival, and the peripherality of the settlement, is interpreted as lack of integration, lack of cosmopolitanism and lack of culture in the view of longterm residents, which in turn is perceived as a threat to Prizren’s integrity. Here ‘culture’ relates to ideas of religion, as discussed in Chapter 4, Islam being central to urban distinction. When people mobilise the discourse of Islam superseding ethnic difference they are not embracing ethnic differences, but rather treating them as irrelevant. This variant of the image of tolerant Prizren elides differences in attitudes to Islam and ways of practising Islam, as well as

126

Purity, mixture and representation

differences between the various ethnic and social groups in the town. As in other parts of the Balkans, Turkish language is closely associated with Islam. This does not imply that ethnic Turks in Prizren are somehow ‘more’ Muslim, but rather that Turkish language is extended to all Muslims, not just ethnic Turks. However, this is not used to exclude non-Muslims, but rather to stress the urban, cosmopolitan nature of the town in contrast to irreligious, monolingual Albanian-speaking rural areas. A central part of urban identity in former Yugoslavia is opposition to the image of the peasant (Jansen 2005). As discussed, while nationalist ideologies elevate the peasantry to the status of heroes/martyrs, in opposition to the depraved cities, urbanites maintain these distinctions, but reverse their moral evaluation. In the case of Prizren, this does not manifest itself in an opposition between the mixed urbanites and the nationalist religious peasants (see Stefansson 2007), but rather between the urban Muslims and the irreligious peasants. Here the people of Prizren are mixed ethnicities united by a common religion, in contrast to the godless peasants. The discursive construction of Islam as a distinguishing feature of cosmopolitan Prizren does not imply that Prizrenites are more religious than those in the villages. Rather, they perceive themselves as such. In contrast, nationalist discourses in other parts of former Yugoslavia have relied heavily on religion as a way of distinguishing the nation, while Albanian nationalist discourse puts the loyalty of Muslim, Catholic and Orthodox Albanians to the nation above religious distinctions. The division is inverted: while being potentially divisive among Albanians, religion does not delineate ethnic groups in Prizren, and so can be used to reinforce narratives of cosmopolitanism. The perception of people from rural areas as uncultured, irreligious and intolerant was mobilised when they moved to the cities, and the perceived demise of cosmopolitanism is in fact much more commonly ascribed to the influx of rural migrants than to the exodus of the Serbs. Cosmopolitanism legitimises the inclusion of certain types of difference, while excluding others (Spasić 2011). In Prizren this is evident in the use of local Turkish to symbolise mixture, while at the same time excluding peasants. A feeling of lost cosmopolitanism relates to the reversal of power structures described in Chapter 3 which precipitated the language differentiation within Romani described in Chapter 5. A shift away from what was perceived to be a stable, affluent society results in the feeling of being invaded and downgraded, by those whose privilege has declined (Grant 2010). This is analogous to Ballinger’s (2004) notion of authentic hybrids in the Julian Marches. Authentic hybrids are a form of mixture understood as good, longstanding mixtures of people, in contrast to new migrants, in Ballinger’s context particularly refugees. In Prizren too, the pre-war mix of people, and the mixed form of Turkish, could be deemed an authentic hybrid, a traditional form of mixture that is recognised as mixed, and celebrated for being so. Here, rather than a dominant ideology of purism existing in contrast to practices of mixing, there is also an ideological valorisation of mixture. Elitist forms of cosmopolitanism contrast acceptable forms of mixture among those free to travel, and contact with other groups of their own class, to the unacceptable mixture of those with less agency and a

Before the war 127 lower class status (Calhoun 2002; Gupta & Ferguson 1992). This form of cosmopolitanism sees the elite as rootless, with the wealth and power to move freely, in contrast to the rooted elitism of authentic hybrids. The idea of authentic hybrids suggests a different yet equally essentialist and exclusionary way of envisaging the relationship between language, place and community. For the rootless elites, linguistic and cultural differences can be acknowledged but ignored among other people with the same means, while place is transient: mobility itself is exclusive. Authentic hybrids also embrace cultural and religious differences, but only specific emplaced differences. This is then organised around the rural/urban divide (Jansen 2008). In Prizren, speaking a mixed variant of Turkish is valued, as is the ability to speak this in Shadervan, and Romani in Terzimahalla. This positive ideology of a mixed space is embedded in place, rather than in mobility, and as such can be seen as a sedentarist variant of cosmopolitanism. This sedentarist cosmopolitanism exists in contrast to nationalist, purist ideas of space. Nationalist ideas envisage a neat correlation between place, language and culture (Gupta & Ferguson 1992; Olwig & Hastrup 1997). Pratt critiques the essentialism of these ideas with the notion of contact zones, ‘social spaces where cultures meet, clash, and grapple with each other’ (Pratt 1991:34). Rather than languages and cultures belonging to one or other ‘community,’ she argues that groups can co-opt, use, mix and change the resources of others. Similarly, Gutiérrez et al. (1999) argue that Spanish/English mixing in a US classroom can be considered highly constructive, as it creates a zone of proximal development.11 The creation of a third space,12 neither the public English-speaking space outside the classroom, nor the private, Spanish-speaking space of the home, allows for a realm of increased learning and creativity. Canagarajah (2013) goes further, both theorising and actively promoting what he calls ‘codemeshing,’ the fluid mixing of languages that creates a space of synergy and serendipity. He argues that the triad of language, community and place enforces a western monolingualist ideology on parts of the world where this has never fitted, and indeed that it is inappropriate for contemporary globalised and multicultural locales as well. However, for a mixed form to exist, so must a pure form. The hybrid can only be created from, and in opposition to, the pure, so the distinction between pure and hybrid persists (Beltran 2004). Linguistic hybridity, whether viewed positively or negatively, must coexist with the idea of purity. It requires first the construction of a pure, ideal form, and only then the mixed ‘incorrect form,’ whether this is seen as playful or dangerous. To promote mixture and plurality in language use is, then, problematic. Indeed, the promotion of plurality has been criticised by Jaffe (2003) with regards to Corsican. She argues that while on one level attempts to include a variety of language practices is commendable, on another level this process means certain practices are always devalued. In the case of Corsican, regional variation is embraced, while mixing with French is not. Both hybridity and purism are based on patterns of exclusion: promoting purity seeks to see everything as rooted and bounded, fixed in space, while promoting hybridity excludes either those who are not able to travel and therefore to mix, or those who are forced to travel and therefore

128

Purity, mixture and representation

create the wrong kind of mixture (Jansen & Löfving 2008). The same could be said for the notion of translingualism (Canagarajah 2013): arguing for a more fluid, negotiated form of communication overlooks the limited agency of many with regards to linguistic resources. In this case it is the newfound mobility of monolingual Albanians that is perceived as dangerous. In Prizren, sedentarist cosmopolitanism posits an essential link between a specific constellation of languages. The war and subsequent movement of people have upset this essentialised link, evoking a strong sense of nostalgia for particular speech practices. Linguistic practices and socio-economic inequalities have been transformed together. Nostalgia functions as more than a reflection of the past, becoming a critique of the present: Prizrenites lament not just the loss of cosmopolitanism, and the arrival of rural migrants, but also the loss of the socioeconomic conditions which allowed them to demarcate boundaries and represent themselves positively in contrast to the peasants – including through their use of language. Now, they can lament the loss of kënaq attached to a particular structure of exclusion, but there is little they can do about it. Faced with the domination of Albanian nationalism, the ideology and linguistic practices of sedentarist cosmopolitanism have persisted where the power to assert them has not. While people in Prizren talk nostalgically about Prizren’s past, a time of culture, cosmopolitanism and contentedness, they also contextually compare their present to non-Prizrenite others, whether people from former warzones who lack tolerance, people from villages who lack religion, or indeed those perceived to be lower class within the city itself. While in the discussion in the Horn Knives shop described above such narratives were brought up by my use of BCS (presenting people with a need to show their tolerance of other languages in the presence of a foreigner), references to being from Prizren, and being urban, are more often indexed by the use of a specific way of speaking Turkish. This is not the memory of Yugoslav socialism, mobilised against rural nationalists, but urban religious sociality upset by the incursion of those who don’t speak Turkish, and don’t know Allah.

Purism and mixture There is a seeming contradiction between the ideology of purism discussed in Chapter 5, which posits ethnic and linguistic mixture as invasive and in need of purification, and this nostalgic glorification of a mixed, cosmopolitan past. This contradiction can be investigated further by looking at shifts in prestige in relation to speech practices. Shifts in prestige do not always entail the loss of prestige forms. While BCS has been mainly erased from Albanian parts of Kosovo, local Turkish has not: its use has persisted, but its function has been reduced from being an elite language, to being a speech genre used for nostalgic local identity. In the sixties, in answer to the question of why non-prestige variants persisted despite people’s awareness of their being non-prestige, Labov (2006[1966]) proposed the idea of covert prestige, whereby certain contexts actually privilege the non-prestige variant; for

Before the war 129 example, outside the classroom, adolescents may reject standard prestige forms (Rampton 1995). Certain contexts require non-prestige forms not despite, but because of, the existence of prestige, and this correlates with the formation of social structures (Washabaugh 1979). Institutional valorisation of certain speech practices posits them as prestigious within a certain context. At the same time, it demarcates a space for its antithesis and shapes it. By delegitimising certain variants in institutional contexts, they are encouraged in other contexts; rather than people using non-prestige variants through weariness, or apathy towards prestige, there is active pressure to use non-prestige in certain circumstances (Woolard 1985). In this sense, both prestige and non-prestige practices form part of the same social structures, and are informed by fractured and often contradictory ideologies. Ideologies which posit certain forms as correct tend to replicate at different scales of these social structures, reproducing fractally, as described in Chapter 4. In Prizren the post-war regime has made Albanian more prominent in the public sphere, more prestigious, pushing other languages towards the private domain. This is far more the case for BCS than for Turkish, but Turkish has also diminished, while Romani was never a prestige or public form beyond the mahalla. The reaction to this shift has been new forms of covert prestige, or, more precisely, a distinction between prestige and non-prestige being reproduced at a smaller scale. The oppressed group thus converts forms of oppression into weapons of defence (Hill & Coombs 1982), reproducing the distinction between prestige and non-prestige by using the ideological structures of the dominant form. The perseverance of non-prestige, or indeed anti-prestige forms, within a certain restricted space can then be subversive, but can also form part of a circumscribed, private arena of transgression (cf. Hill & Coombs 1982; Rampton 1995). A standard Romani is being made in opposition to, yet based on, the standards of the surrounding languages. Non-dominant languages may adopt dominant ideology in terms of nationalistic purism (Woolard 1998), but at the same time they need to find a niche for this within existing structures. Thus prestige, and its requisite other covert prestige, is always scalar: ideas of prestige and non-prestige are reproduced at different scales through the reproduction of private and public spheres. Many Prizrenites today relate to the current feeling of loss of prestige and lack of opportunities by discursively constructing the past as a time of urban sedentarist cosmopolitanism. The qualities of this sedentarist cosmopolitanism include religious sociality, linked, as we have seen, to a local dialect of Turkish which borrows extensively from surrounding languages. This variant of Turkish, specific to Prizren, is explicitly described as mixed and does not replace ethnic languages, but rather provides a framework for communication beyond this. Indeed, ‘(l)anguage ideologies always include metapragmatics, that is, local suppositions about the relation of speech forms to speakers’ identities and their social situations’ (Gal 2006b:179). This framework of communication, associated with kënaq, contentment, has to a certain extent been lost. Prestige and urban cosmopolitanism have shifted drastically due to the war, the exodus of the Serbs, the influx of rural migrants and the power structures now in place. Nostalgia for the

130

Purity, mixture and representation

social world associated with this way of talking is suggestive of this loss, a loss not only of the language itself, but more the sociality it used to express. In this context, positing oneself as a Prizrenite means showing oneself to be spatially central to the town, to have had a presence there certainly before the war, and preferably longer (recent interpretations of the flag discussed in Chapter 3 testify to this); it also means differentiating oneself from those who do not fulfil these criteria. For Roma this means, in Fatka’s terms described in Chapter 4, having ‘pedigree.’ This group demarcation manifests itself in an ideology of cultured distinction, expressed in certain speech practices. Abu-Lughod and Lutz (1990) argue that ideology ‘is always linked to historically specific social groups assumed from the start to be engaged in struggles of domination and resistance’ (Abu-Lughod & Lutz 1990:9). The dramatic shift in power relations in Prizren during and after the war has highlighted that the ideology of urban cosmopolitanism is precisely that, an ideology, linked to historically specific social groups engaged in struggles of domination and resistance. Here, the ideological link between language, community and place is constructed, rather than essential or natural. However, as it is deeply entrenched in social hierarchies, it still has a very real effect on people’s lives. People’s ideological understanding of language and mixture is both shaped by and shapes what they say and do, but they do this within certain boundaries, as ‘ideologies are not like rarefied, disembodied voices, and are instead given material force in the practice of institutions’ (Rampton 1995:309). Post-war political changes have meant that mixed Turkish is no longer about securing in-group boundaries, but about nostalgia for the days when that in-group was in a position to secure its boundaries. This was previously an unmarked variant, which ‘meets the beliefs and desires of persons in the community who have sufficient power to set norms’ (Myers-Scotton & Bolonyai 2001:10).13 Since the upheavals in power relations, its use has become more marked. The replacement of Turkish with Albanian as a lingua franca does not relate to the replacement of Turks with Albanians as elites. Rather, this shift is metonymic of the victory of Albanian nationalism, the migration of rural Albanian speakers to cities, and the loss of a religious, sedentarist cosmopolitanism. The previous urban elites have lost power, and their elite speech practices are no longer neutral, but have been peripheralised, becoming a nostalgic speech genre. While nostalgia for a pre-war state of contentment is linked to an explicit discourse of harmony, it is also embedded in the unspoken exclusion of certain groups. The ideology attached to the use of Prizren Turkish indexes harmonious mixture both metaphorically, in that a mixed language is understood as implying a mixture of people, and practically, in that it excludes the non-harmonious Other, those who are not from Prizren, or have recently moved there. Here it becomes clear that there is no simple division between language ideology and speech practices. Language ideology is not just that which is explicitly stated, but also the unspoken and implicit structures in which language is embedded. The structures in which ideologies and speech practices are embedded are not static, and the hierarchies that attach ways of speaking to ideologies of prestige in any given context are historically situated. Similarly, while the ideologies connected to particular

Before the war 131 linguistic practices are an essentialised abstraction of these practices, they are also in themselves linguistic. The fact that they are abstracted ideas related to practices means they can be (and often are) made explicit, through the use of language. In this sense, the distinction between ideology and practice could be construed as a distinction between form and content. An utterance can be mixed in form, while also promoting mixture in content. However the relationship between the ideological content of language, and the form of the language, speech practices, is often inconsistent. The contradictions of these explicit ideological discourses become clear at points of dramatic shift, exposing the underlying structures. The inequalities inherent in the Prizren variant of urban cosmopolitanism were punctured in part by the war, but perhaps more fundamentally by the arrival of migrants that followed. Before the war, for Terzimahalla Roma, especially men, being part of this cosmopolitan sociality provided the possibility of obtaining prestige beyond the mahalla. Being Roma and speaking Romani was but one facet of a more complex urban Muslim cosmopolitan identity. Nowadays, power relations are reversed, and this possibility for membership in an urban sociality has shifted to a focus on differentiation from groups not perceived as part of this pre-war elite.

Conclusion In light of this demographic shift and social change, it can be argued that, for Roma, the demise of mixed Prizren Turkish to demarcate social differentiation before the war has left a gap. It has been filled with the construction of a nascent educated, literary form of Romani, to distinguish urban Roma from the incursions of those seen as rural and uncultured. Nostalgia, the residual prestige of a form of sociality, remains, but is no longer a way to access power. The need to ally oneself with those from the town, in contrast to those outside, has been superseded by the need to differentiate from Roma/Ashkali recent arrivals to the town. An elite variant of Romani serves this purpose, while also providing viable access to economic resources through international multiculturalist funding opportunities, which may not have space for mixed Turkish as a non-ethnic language of communication, and certainly would not encourage the religious element of this pre-war configuration. The image of Prizren as multicultural and multilingual, along with the perception of the Turkish spoken there as mixed, has come to be associated with pre-war cosmopolitanism, now felt to be lost. This cosmopolitanism was grounded in an urban elite with the power to exclude rural Albanian speakers. While people have a level of agency in their choice of speech practices, they often have less agency in changing the hierarchies that posit one form as prestige, or unmarked, in any given context (cf. Block 2012). Languages are social processes, and influenced by institutions and power structures; they can be commodified, becoming resources with restricted access (Moyer 2012). The power of the cosmopolitan elite has since been lost, or at least reduced, and, with Albanian taking over as the state language and many Albanian speakers moving to the cities, this mixed Turkish

132

Purity, mixture and representation

has come to symbolise nostalgia for the pre-war status quo. Li Wei (2012) calls for the breakdown of micro/macro distinctions in the study of language, arguing that we need to look at wider social processes to understand language. In this vein, by looking not only at the war and demographic shift, but also at the relationship between ideologies and speech practices, we have been able to see how shifts in power affect people’s understanding and use of language in the everyday.

Notes 1 The word opština for this building is often used across languages despite its official name now being the Albanian komuna. 2 I call this variant of Turkish mixed not in the strict, structural sense of a mixed language (Matras & Bakker 2003), but rather in the sense of a language which is perceived as mixed on the basis of extensive borrowing from surrounding languages. 3 As well as disliking mixing, he may also have disliked the use of Serbian words altogether. 4 The Albanian Catholic Church was rarely mentioned by my informants, except with regards to Caritas, and a nearby pizzeria that was rumoured to serve pork. 5 This list of examples is not meant to be exhaustive, nor is it intended to imply that any one side had a monopoly on suffering. Rather, I intend merely to show that Prizren was not exempt from some of the horrors of the war, despite these narratives of peace and tolerance. 6 In the rare instances when this slogan is mentioned, it is explicitly related to Milošević and Serbian rule. 7 See Boym (2002) for a discussion of nostalgia as a critique of the present. 8 The evening meal during Ramadan where the fast is broken. 9 This is a bakery item eaten only in Ramadan. During the day people bring their own fillings to the bakeries and the baker prepares them ready for iftar. 10 Nostalgia was of course mostly a topic among older people, who remembered the period before the war first-hand, but who were also less involved in promoting new ideas about Islam. 11 A zone of proximal development refers to an appropriate level of learning, where the learner can achieve to the best of their ability (Vygotsky 1978). 12 An intercultural space where the translator and translated are modified (Bhabha & Rutherford 2006). 13 See also Haspelmath (2006) for a critique of ‘markedness’ in linguistics.

Conclusion Between two fires? Dissonance, diglossia, disorientation

Introduction In the first part of the book I examined political and economic changes in Kosovo through the lens of language, arguing that language shifts are metonymic of broader political shifts, and that multilingual repertoires are not simply a representation of a multi-ethnic society, but have an intricate relationship to shifting political, economic and social group formations. Diachronically, shifts in the Yugoslav political system and international relations affect much more mundane everyday speech practices, while synchronically these practices are structured around the transnational funding schemes that shape social relations within the NGO. The second part of the book looked at the contingent, and dissonant, relationship between language ideologies and speech practices with regards to Romani standardisation, Romani representation and the mixed Turkish of a peripheralised speech genre in Prizren. Roma are represented, and represent themselves, within locally produced social hierarchies, and the contestations manifest in language ideologies and speech practices. As such social hierarchy can be studied through language, by bringing the multiple and fractal hierarchies examined in earlier chapters into the broader context of representations of Roma. The contentions involved in the representation of Roma can be illustrated by a very different setting from Prizren: a Serb enclave. The enclave we visited was a small village, over an hour’s drive from Prizren. By looking at an incidence of the Roma I worked with visiting this enclave I seek to shed light on the way Roma are perceived and positioned as ‘in-between’ in different contexts.

Between Prizren and the enclaves As part of an attempt to gain support for the New Roma Party, Osman, Nexhip and others were visiting Roma in a Serb enclave, with whom they usually have very little contact today. They were talking in Romani: Osman Osmani: Tumen, ani enklave, tumen siyan mashkar e duye yaghende. Mashkar e gajyenda thay e gaunenda … X: Na! Na si amen mashkar e duye yaghende. Nane amen problem e gajyencar!

134

Conclusion

Osman Osmani: You, in the enclaves, you’re between two fires. Between the Serbs and the Albanians … X: No! We’re not between two fires. We don’t have any problem with the Serbs! As discussed in the Introduction and Chapter 3, the terms used for Serb and Albanian, gajo and gauno are historically related to the position of Roma in Prizren: gaune referred to peasants, therefore Albanians; korohaya referred to Muslims, therefore Turks; and gaje referred to non-Roma (as it still does in other dialects) which in the context of Prizren meant Serbs. Ethnonyms in the enclaves are often different, but it seems the Roma man from the enclave was familiar enough with this usage, and didn’t confuse matters by switching terms in the middle of a conversation, so accommodated to the Terzimahalla terminology, using gajyencar for ‘with the Serbs.’ At the same time, though it is not clear from the extract above, the Terzimahalla Roma were accommodating to the speech of the enclave Roma. The speech of the enclave Roma is noticeably more BCS, and the linguistic resources of both BCS and Romani are combined almost imperceptibly. Albanian elements which have entered Terzimahalla speech, such as saying po for ‘yes,’ and foshnya for ‘children,’ are not used in the enclave. Later, on the way home to Prizren, they said it was a shame how the Roma in the enclave were ignored by politicians on all sides, which was why they had visited. I commented on their linguistic accommodation, suggesting it was a positive approach. They responded curtly that they did not change their speech: they always spoke ‘their’ proper, Terzimahalla Romani. The aim of their visit was to encourage enclave Roma to vote for the New Roma Party (KNRP), to unite Roma from both sides of the divide in Kosovo. The unity of Romani language should, from this perspective, symbolise the unity of the group. However, the nominal unity of the language masked a range of differences in both language ideology and speech practices. The way language ideology affects Romani speech practices is configured by ideologies of difference related to both Serbs and Albanians as dominant ethnic groups. The man from the enclave, visibly poorer than those who had come from Prizren, objected to being told he was ‘between two fires.’ He did not resent Serbs, and by implication he did resent Albanians; indeed, Roma who live in Serb enclaves are more likely to be seen as collaborators, while many also fled there from Albanian areas in fear. He was therefore positioned on the side of the Serbs in the overarching dichotomous narrative of Serb versus Albanian, in contrast to the Terzimahalla Roma. Although the Terzimahalla Roma altered their speech to communicate better with the enclave, a speech practice of accommodation, this was at odds with the ideology of social difference, and of different allegiances between the groups. They practised accommodation while maintaining an ideology of distinction. This relates to the historical trajectory of these two groups before the war, when there were perhaps less divisions between the two, and the accommodation is therefore ‘automatic,’ while being overtly denied. The message of Roma unity which the KNRP were trying to spread was disrupted by the way they positioned the enclave Roma as in-between. Roma from Prizren may be more powerful than Roma from the enclaves, and as such able to travel there

Conclusion

135

to promote their political party, but their vulnerability to the Albanian authorities outside the enclaves means they must not be seen to take the side of the Serbs, or even be seen as in-between Serbs and Albanians. To take the position that there is no problem between Roma and Serbs is fine, even advisable, in the enclaves, but not in Prizren, or indeed on a wider scale in Kosovo. This brings me back to the question of whether we can study society through language. In the example above, it is clear that the content of what is said tells us what was overtly being contested – the allegiances of the two groups in a postconflict divided space. But a wider ethnography of the speakers involved, their conscious and less conscious attitudes towards different varieties, reveals a far more complex picture of allegiance and historical trajectories of linguistic forms, the ‘social life of the verbal sign’ (Voloshinov 1973:27). An ethnographic study of the variations in use of language, and of the often inconsistent and contradictory relationship between language ideology and speech practices can elicit the complexities of this short dialogue, the unacknowledged accommodation of the Terzimahalla Roma, along with their insistence that they only use ‘their’ Romani. ‘Their’ Romani here means the speech practices which are metonymic of their group, both part of the group and representative of it, a cause and consequence of group distinction. Here they consider their group to be superior, more ‘cultured’ than the enclave Roma, and therefore maintain this distinction. It is not simply that they are not conscious of, or trying to deny, their speech practices; rather, there is a discordance between their speech practices, which are historically constituted, reverberating with previous uses (Bakhtin 1986), and their explicitly stated metalinguistic discourses, which are out of sync with these practices. The ethnographic study of language as metonymic of society shows the unequal relationship between the two groups. Those from Terzimahalla, in a position of relative power, can come to the enclaves and represent the enclave Roma as being ‘between two fires.’ But it also shows the historical relationship between the two groups, a past where linguistic accommodation was common, and natural; despite the post-conflict severance of relations and ideological division of ways of speaking, linguistic accommodation persists.

Variations on a theme: betweenness The notion that Roma are ‘in-between’ is not isolated to the discussion above, but is used far more broadly to describe the predicament of different groups of Roma. The idea that Roma are ‘caught in the middle’ (BBC 1999) or ‘sandwiched’ (Duijzings 1997) between Serbs and Albanians is a common metaphor to describe the situation in Kosovo. It has also been argued that Roma are ‘between’ not two combative ethnic groups, but rather between an EU-sponsored notion of civic nationhood, and the Kosovo Albanian ethnic one (Sigona 2012). Roma in Kosovo thus appear as either caught between two groups, or between different citizenship regimes. Alternatively, in Lemon’s (2000) ethnography Between Two Fires, Roma are posited between assimilation and marginalisation (cf. Gamonte 2013; cf. Kovats

136

Conclusion

2003). This relates more broadly to discussions about whether Roma should be seen in in terms of marginalisation, that is, in socio-economic terms, or in ethnic terms (Ladányi & Szelényi 2003; cf. Stewart 2002; Sardelić 2016; Tiefenbacher 2013).1 This is turn relates to the socio-economic structures which pressure Roma to assimilate or remain marginalised (van Baar 2012; Helbig 2010). Here, not only are Roma as a group seen as uncategorisable, in-between nations, transgressing borders, but Roma individuals are caught between the opposing pressures of marginalisation and assimilation, and Roma as a group again defy categorisation by displaying both class and ethnic characteristics. A further use of the metaphors of betweenness, discussed in Chapter 4, is that voiced by Roma feminists, which looks at the predicament of Roma women, between racism and sexism (Oprea 2004, 2005; cf. Baker 2015). From this perspective, Roma women are trapped by their relationship to Romani traditions (Ilisei 2012; Kyuchukov 2011), which limits their possibilities for organisation and activism (Brooks 2012; Jovanović & Daroczi 2015; Schultz 2012). In addition to the metaphor of Roma being between two groups, or indeed two ideologies, Roma are also described as being between places, nowhere at home. The metaphor of homelessness is sometimes based on a perception of Roma as nomads (Beasley von Burg 2009),2 but is also used to understand migration and precarity of Roma in Europe (Macarie 2014). More than shuttling between two poles, this spatial idea of being nowhere at home, between countries, is symptomatic of a group who transgress modernist categorisations. This ‘in-betweenness,’ rather than being valorised as a positive ideal of free movement, is a state of unsettlement, a problematic, and possibly dangerous, sort of transience. Conversely, proponents of the Romani cause sometimes argue that Roma are the true Europeans: they live in every state and their language is spoken across Europe. This claim is in part a retort to the stigmatising, racialising treatment of Roma vis-à-vis ‘Europeans,’ but it also mobilises elitist cosmopolitanist claims of being rootless to show that Roma, too, can be ‘modern.’ This perspective valorises Roma as rootless cosmopolitans (cf. Jansen & Löfving 2008), rather than a population who migrate in response to rejection and deprivation. Among these varying narratives of in-betweenness, Romani language is often represented as mixed, slippery and hard to define, as will be discussed below. This is metonymic of the way Roma are perceived – too separate from non-Roma, yet also too mixed with them, too marginalised, yet not authentic enough. The array of works which describe Roma and their perceived problems as ‘in-between’ is indicative of the image of both Romani language, and Roma in general.

Mixture, hierarchy and change These varying attributions of betweenness, ascribed to Roma by academics, activists and Roma themselves, both positively and negatively posit Roma as mixed, in-between, and beyond categorisation. Their mixed speech practices (both mixing Romani with other languages, and the lexical borrowings perceived as mixed within Romani) are metonymic of this, while the purism described in Chapter 5 is a reaction against it. Theories which seek to challenge modernist ideas of

Conclusion

137

language as pure and distinct (e.g. Canagarajah 2013) run the risk of celebrating language which is mixed, and boundaries which are blurred. Like other celebrations of rootless multiculturalism, mixture can be seen as an overwhelmingly positive phenomenon in contrast to monolingualist ideologies. While extensive mixing and creative practices in multilingualism may be praised among and by those who are able to move freely in relatively elite circles, the mixed use of nonprestige forms by those not in a position to navigate these linguistic resources is often denigrated as ignorant, or confused. In celebrating mixture there is a risk of losing sight of the social hierarchies which shape linguistic practices. Attempts to show Romani in-betweenness as a quintessentially European cosmopolitanism, and to show all forms of language mixing in a positive, productive light, overlook huge social and economic inequalities. By using language learning as an ethnographic method, I have shown how people draw distinctions between, at one end of the scale, the quotidian multilingualism of some people in Prizren, many of whom speak four native languages but lacked formal education, and, at the other, my own elite multilingualism, which despite my lack of linguistic competence was treated with far higher respect. As someone connected to a university, from one of ‘the Big European Countries,’ my language learning was perceived as prestigious, while theirs was often treated as not knowing any one language properly or purely. Prestigious learning denotes cultural capital, and requires the formal, standard language (Abercrombie forthcoming). The prestige form requires didactic learning, which usually requires some investment of time or money. It cannot simply be picked up, but has to be actively sought out, and policed. As such it is not open to everyone, as not everyone has the same opportunities for education, or the same incentives to learn. While those in control of the standard actively try to diffuse it, access to education, and economic incentives to assimilate, vary widely among Roma; for example, girls, who are more likely to be kept home from school (especially in poorer families), rural migrants who are more vulnerable and thus more pro-Albanian and others, are unable to buy into this exclusive Romani. Teaching and learning standard Romani, without an institutionalised prescriptive grammar, requires making one. This process is founded on a standard language ideology, which juxtaposes the everyday speech of the mahalla to a pure, correct, educated way of speaking appropriate for use in the public sphere. Specifically, lexical purism focuses on erasing loanwords from local languages, creating a parallel lexicon of standard words (Thomas 1991). As such, a hierarchy of practice is created. This hierarchy of speech practices complicates both cosmopolitan ideologies of translingualism (the positive image of mixture), and the opposing negative view of mixture as in-between, impure or improper. If we are to accept that all languages and peoples are inherently mixed, and that true purity does not exist, this begs the question of what mixture is. I would argue that here, mixture of both people and languages is a form of representation, rather than an essential feature. When a group represents themselves (or their equals) as mixed, they tend to do so positively; when differentiating themselves from a dominated group, they tend to use the negative representation. The claim that Roma are true Europeans, in that

138

Conclusion

they live across Europe, as true cosmopolitans, highlights the problematic of cosmopolitanism discussed in Chapter 6. Cosmopolitanism, of the type described by Calhoun (2002), relates to elites who can jet from city to city, encountering only their own class. Roma do not fit this mould: when they do move it is often from necessity, usually to find work. Equally, for Terzimahalla Roma as Prizrenites, the sedentarist cosmopolitanism of Prizren is positive, while for them as elite Roma, the impure Romani of the mahalla is negative. The celebration of multilingualism is metonymic of the celebration of cosmopolitanism, and is equally hierarchical and exclusionary. By ethnographically investigating the ideologies and practices that posit certain forms of mixture as positive, while others as negative, this thesis has shown the ways in which language, and by extension society, is constructed as mixed or pure. The Roma I worked with had a stake in the sedentarist cosmopolitanism of Prizren before the war, but this has been eclipsed. The image of Roma as true Europeans, or the idea that the whole world is the Roma’s land (the title of one of the dramas discussed in Chapter 4) and that they are at home everywhere, displaces the problems of socio-economic inequality that often underlie this need to move around and to subsist everywhere. If cosmopolitanism is about recognising and celebrating difference while constructing new forms of exclusivity (Ballinger 2004), I would argue that so too are ideologies which celebrate multilingualism as they do not acknowledging the huge disadvantage some people face by not knowing the ‘pure’ or ‘educated’ varieties celebrated by monolingualist ideologies. Those in a weaker position are invariably pushed either to embrace the dominant paradigm, or to live outside it and thus be marginalised. In this sense, those in a less powerful position are often stuck between marginalisation and assimilation. In Chapter 5, I discussed the way Roma in Prizren are increasingly choosing to bring their children up speaking Albanian to avoid economic marginalisation, while at the same time they are accused of giving in to pressures of assimilation by those who have a stake in the Romani standard. This is not a simple choice between remaining ostracised or assimilating (cf. Young 2004), or between embracing their own language and taking on that of another (Hodges 2016). For those with sufficient cultural capital, it is precisely through formulating a standard that they can begin to be integrated as equal players in multicultural Kosovo – but they have the luxury of making this choice, while many others do not. Those with a greater stake in Romani language, and the social and cultural capital to make a return on their investment in standardisation, represent this as a choice between economic benefit and loyalty to the ethnic group. From this perspective those with an interest in the standardisation of language are positing their own position as one which values ethnic recognition and solidarity over economic benefits. If language is perceived merely as a skill, then learning the dominant language (in this case Albanian) would result in a greater distribution of economic resources within the group. However, those leading language standardisation prioritise identity recognition. What remains unsaid in this debate is that there is no simple choice between recognition and economic distribution, or loyalty and money; rather, different people have different economic possibilities,

Conclusion

139

as well as ethical and emotional ties to language. Those in a less powerful position not only have less to gain through the recognition of Romani, but also have a greater need to show allegiance to Albanians. This contention relates to the discussion in Chapter 3 regarding the economic benefits of projects ostensibly aimed at ethnic recognition. Here Roma are not just between east and west, as beneficiaries of funds tied to certain discourses, but are also between the de jure rights awarded them and the de facto socioeconomic deprivation they are left with. The semantic dissonance which results from this state of affairs belies a gap, another state of in-betweenness. The social capital that allows some, but not others, to find gainful employment through the NGO means that only people with contacts can access economic redistribution. Both people’s perception of economic benefit and their perception of the identity of themselves and others shift, often in relation to each other. Rather than group identity and economic benefit being seen as two independent choices, the need to invest in one set of speech practices rather than another suggests a fractally divided structure of domination, whereby choosing one language, and one scale of hierarchy, precludes the choice of the other. Therefore, the choice Roma parents face is not solely about the recognition of their group over the economic possibilities for their children, but rather the scale at which they want group recognition to be sought (in this case the mahalla or beyond) and the possibilities for redistribution at that scale. There may be more jobs for Albanian speakers, but there are also many more Albanian speakers requiring jobs, and informal networks of employment may mean Roma are excluded from these jobs. Romanilanguage jobs are far more niche, but there are also far fewer speakers of the standard; given that knowledge of the standard already depends on social capital, it is likely those with this knowledge will also be in a position to be given those jobs. In this sense they are not just between marginalisation and assimilation, but between scales of identification at which they can seek redistribution and recognition. Those with a vested interest in the promotion of Romani language create distinctions between standard and non-standard forms. Purism here is an attempt to centralise Romani, by making it less mixed, less in-between, while at the same time making clear boundaries between Romani and any other encroaching languages. While studies of language standardisation typically look at selection, codification, diffusion and elaboration (Haugen 1966), I have assessed the way the standardisation process can both transform and reinforce social distinctions. An ethnographic approach elucidates not just the metalinguistic discourses about language, but also their relation to everyday speech practices, and the interplay between the two. The standard language ideology which informs these discourses and practices is not innate, but learnt, and I have argued that in this case it is learnt from the language ideologies of BCS, Albanian and Turkish. The belief that there is a correct way of speaking, and therefore that there is also an incorrect way of speaking, is replicated in Romani-language ideology, which then manifests itself in lexical purism. This form of purism, framed by a standard language ideology, is not just diffused from the standard language

140

Conclusion

ideologies of the surrounding language, but is also a reaction to the dominance of these languages (Annamalai 1989; Jaffe 1999). While the overt aim of this process is to ‘purify’ Romani, and diffuse this pure form to all speakers, the actual effect of this parallel lexicon is to give those without privileged access to the standard a feeling of ‘not knowing their own language.’ This ideology thus generates a distinction between formal and informal practices of learning, and to the reproduction of inequality through language. When non-Roma represent Romani as an impure, in-between language, this is metonymic of their attitude to Roma themselves, as impure and out of place, not modern enough yet also not authentic, pure Roma. On a larger scale, the distinction between the modern, pure self and the in-between, semi-peripheral Other is a Balkanist schema of difference (Todorova 1994, 1997; Fleming 2000; Green 2005; Petrović 2013), which is replicated fractally at ever reducing scales (BakićHayden 1995; Irvine & Gal 2000; Petrović 2007). The Roma reaction to this representation of Romani language is to seek to purify it: those in a dominant position distinguish between their own purity, and the impurity and in-betweenness of those below them. The pattern replicates at a smaller scale: the dominant represent the dominated as in-between, and as such assert their own dominance. The same can be said for the intersectional position of women: Roma women are not simply between two ideas or systems, but also further down the hierarchy (see Chapter 4). Being a Roma woman entails more than just double discrimination: it is a case of racism and sexism being intensified, domination reproducing itself fractally, and any possible escape routes being closed. The recurring theme of in-betweenness in studies of Roma, then, is not a result of an inherent in-between position, a need to find a ‘proper’ place somewhere. Rather, it is the outcome of their domination: they are represented as in-between by those with the power to do so. To paraphrase de Beauvoir (2014 [1949]), Roma are not born in-between, they are made in-between. I do not wish to suggest that the scholars and activists cited above are forming a negative representation of Roma. Rather, they are describing the negative representation which structures Roma’s experience. In-betweenness, rather than being a position of a group with regards to others, is the interpretation of a representation projected onto them. This projection of in-betweenness forms part of the projection of semi-peripherality on the Balkans, and is replicated in the representation of Roma by non-Roma, of enclave Roma by Terzimahalla Roma, of Roma women by Roma men, and so on. Those with the power to demarcate boundaries, and therefore to posit the Other as in-between, do so to those without the power to contest this; but the latter react by replicating this gaze, and project this representation onto those who at a smaller scale are powerless to them. The relationship between dominant and dominated means that this representation is one-way, but also that it affects and restricts the position of the dominated: in Green’s (2005:158) words this representation is ‘a fantasy with teeth.’ This replication results in ever smaller fissions, and continuous differentiations on the basis not of ultimate difference, but of inbetweenness. It is futile for dominated groups represented as in-between, in this case Roma, to reflect this image back onto their accuser. Rather, co-opting the

Conclusion

141

terms of the debate from the dominant forces, they attempt to purify Romani, and bring it into the public sphere. Positing the immediate dominated Other as in-between, semi-peripheral, also posits them as dangerously uncategorised. This goes both for the relationship between Europe and the Balkans, and for that between non-Roma and Roma. The Other which is separated by a further degree can be categorised. Rather than being in-between, this complete Other is an inversion which can be pitied and even saved (Spivak 1988). From the perspective of the dominant, this complete Other is not dangerous: they are not ‘matter out of place’ (Douglas 1966). But this complete Other, subjected to multiple forms of domination, cannot resist: to challenge those immediately dominant to them would harm the group, and they are not in a position to resist those further up the hierarchy. The fractalising patterns of representation also reproduce a public/private distinction, meaning that the private at any one scale becomes the public on the next scale down. As such, the appropriate public language depends on the scale of the public/private distinction. Within these fractal dichotomies, it is the public, not the private, which is exclusive. The public sphere is not public in the sense of free and open to all, but rather in the sense that, as described in Chapter 6, those who control the public realm are those who have the power exclude others by peripheralising them. The relativity of this more private domain – private when compared with the dominant, yet public when compared with its own internal, dominated domain – allows for it to be posited as in-between. This creates the nesting distinction of nature/culture, described in Chapter 4. One manifestation of this in Prizren is the orthography debate described in Chapter 5: Roma want an independent, international written language, but this is superseded by the need to align themselves with Albanians through an orthography closer to Albanian, and definitely distinct from Serbian. While the creation of an international Romani language is often thought of as marred by in-fighting, I would suggest that external pressures are far more responsible for disunity and centrifugal forces in Romani activist movements. The immediate dominance of the state of Kosovo takes precedence. As political structures have changed, allegiances have shifted from the central Yugoslav government, to the Kosovo Albanian administration, back to Belgrade, and back again to independent Albanian nationalist authorities. Chapter 2 showed how, without political power, Roma could only navigate their position and nationhood within the changing tides of political change. The KAM Durmish Aslano was set up at a particular historical juncture: Yugoslavia was decentralising, and the International Romani Union was being founded. Because of the ideological linkage between ethnic groups and language, language was often metonymic of such historical changes. Romani language was becoming known at an international level, while Serbo-Croat was in decline in areas where it was not the first language, such as Kosovo. This allowed for Roma with a level of education and cultural capital to assert some, albeit restricted, independence as representatives of the group. The decentralisation and shift in minorities policy did not allow people to challenge Yugoslav socialism itself but rather counterbalanced such challenges, by opening up new spaces for ethnically based organisations within the Yugoslav

142

Conclusion

system (Jović 2003). More than a shift in policy, this can be seen as a shift in what it meant to be Yugoslav: citizens were no longer envisaged as a new Yugoslav man (on the model of the Soviet man), and ‘Yugoslav’ now became adjectival. Roma from Prizren were no longer categorised as just Yugoslav, but as Yugoslav Roma. This category was maintained through the shift to a repressive nationalist regime in the late 1980s, a period of social and ideological upheaval, but as Yugoslavia increasingly came to signify Serbia, and Albanians increasingly withdrew from interaction with the Yugoslav state, Yugoslav Roma ended up on the Serbian side of the chasm between the two groups. In this sense, rather than being between two fires, the problem was more that the binary logic of ethnic conflict meant they found themselves resolutely on one side. As such, after the war they either left, or made a concerted effort to realign themselves. Asserting their position was always reined in by the necessity of asserting allegiance. This, and the accompanying ideological changes, have led to the disorientation described in Chapter 4, but also to a defensive reaction, a tightening of boundaries between the group and the Other, and a solidification of hierarchies within the group. Contrasts at the highest scale of the social hierarchy are increased, which reverberates through the fractals. Realigning themselves has involved disentangling themselves from a Yugoslav orientation, and orientating themselves to the Kosovar Albanian state, but also to the regime of the international community and its liberal multiculturalist goals which promote recognition at the expense of redistribution. As such, the relative lack of power of Roma as an ethnic group means that not only are they subject to dominant groups, but that they have also been subjected to varying power structures and systems of domination. Their ability to navigate their language use has therefore shifted within the bounds of these structures. Jernudd (1989) argues that purism is a reaction to change; here, by ethnographically investigating the dynamics of purism, we can see that Romani purism is part of a broader response to the vacillations and upheavals of political structures in Kosovo. Large-scale political changes have meant a change in the way these boundaries are structured. The status of pre-war cosmopolitan elites described in Chapter 6 has been dissolved. In the immediate aftermath of the war, the renegotiation of boundaries was extreme, which for Roma meant withdrawal to the relative private of the mahalla as a defensive mechanism. At the same time, the dominance of Roma men over Roma women was reasserted as a form of protection. The threat and instability of this period, the fear of not knowing where the boundaries were, led to the increasing withdrawal of Roma women to the private sphere. This is not to say that the treatment of women has got worse in every respect – Roma women today are afforded certain freedoms and opportunities they were not before. Rather, the instability inherent in a period of violent change reinforced gender distinctions. Categories posited as neutral dominate the public space, while others are actively excluded, restricted to the marked private sphere. Before the war this neutrality was owned by Serbs, and urban Muslim males, symbolised by speaking Prizren Turkish, while actively excluding women, and rural Albanians. With the changes in power structures that followed the war the Serb presence was erased,

Conclusion

143

while the other urban elites were decentred, losing privileged access to what was formerly their public space. Displaced in this way, their speech practices lost their neutrality, and the former elites have been pushed to the peripheries of an in-between position, no longer able to assert authority and exclusivity at the level of the town, but able to use these practices as a speech genre for nostalgia. They have not been completely erased from the public space of town in the way that Serbs and Serbian language has been. But they have been marginalised, peripheralised, not quite central nor completely externalised. Language use, and particularly language prestige, is negotiated within these fractalising hierarchical structures, within the public, prestige variants both symbolising and constructing the dominant group. The dominant group constructs the non-dominant group by demarcating difference in forms of expression. The prestige formerly associated with access to and control of this public space has not been totally lost for Prizren Roma, but has been reduced. They are now partially outside. The reduced prestige of Turkish, leading to the lack of a medium to express multi-ethnic urban exclusivity means Roma men need to reassert their dominance and neutrality on a smaller, more private, scale and through a different medium. At this scale, Roma men no longer distinguish themselves from rural Albanians through Turkish, but from the Roma and Ashkali rural migrants, and from women and children. The creation of a local standard language discussed in Chapter 5 is metonymic of this process. Positing oneself as a standard speaker at the level of the mahalla requires positing the dominated Other as impure and uneducated, Romani-speaking but speaking a mixed, in-between Romani, not the pure Romani as defined by those with the power to decide what is standard. Thus prestige is produced and differentiated from non-prestige on the scale of Terzimahalla. Another fission occurs with those who define themselves as Ashkali: they react to, and replicate, the dominance of Roma elites by distinguishing themselves from Roma and replicating hierarchy within the group, while being accused of mimicry by Roma (Petrović 2008). The historical shifts which affect these hierarchies initiate changes in language ideology and speech practices, but not ones which are always in tune with one another, leading to inconsistencies and dissonances between the speech practices people purport to use and those they actually use. In the example given at the beginning of this conclusion, I noted that while Roma from Terzimahalla linguistically accommodated to the speech of the enclave Roma, when I questioned them they denied this. I would suggest that this is because they are familiar with this Serbianised Romani from before the war, and aren’t aware of the change in their practices in this context. At the same time, ideologically, they also want to assert their distinction from, and superiority to, the speech of enclave Roma. They subconsciously change their practices, engaging in practices they are familiar with from the past, while ideologically maintaining this present-day boundary. This fractalisation of structures of linguistic prestige and dominance manifests itself in an unequal diglossia, in response to a historical shift. The replication of diglossic structures from surrounding languages into Romani described in Chapter 5 is indicative of the foreclosure of access to prestige through Turkish, described

144

Conclusion

in Chapter 6. This diglossia is hierarchical, both in terms of the way these forms are learnt, and the level of access people have to them. This hierarchy is structured around a distinction between public and private, where the move to construct a Romani standard accompanies the move to bring Romani into the public sphere, thus reiterating a public/private distinction. This public/private distinction is also reconstituted by the processes described in Chapter 3, whereby the influence of international actors on the explicit discourse of recognition is at odds with the private, implicit systems of redistribution, creating a semantic dissonance between the two levels. This dissonance between the private and public discourses, and between ideology and practice, leads to an overwhelming sense of disorientation, a disillusionment with conflicting metanarratives, coupled with an inability to let go of metanarratives totally. This is represented by the dramatic performance of the gypsy, as bodily and natural, performed primarily not to the purported modern European Other, but to Balkan Roma themselves, not breaking down the boundary between self and Other but severely unbalancing it with a shifting tragicomic vantage point. With semantic duplicity ever-present, and boundaries elusive, grasping for a coherent metanarrative by which to live one’s life becomes fleeting and inconclusive. By looking at the social life of verbal signs (Voloshinov 1973), the ethnographic study of language can be used to expose far deeper synchronic structures, as well as the historical trajectory of these structures, reverberating with utterances of the past (Bakhtin 1986). Equally, as I have shown, linguistic forms and ideological discourses are not just a relation to the past, but are also rooted in a projection of the future: people pre-empt their possible futures by investing in the social and cultural capital that will provide a basis for economic stability for themselves and their families. Speech practices and language ideologies are structured by scalar fractalising hierarchies, within which there are limited possibilities. Language does not just represent the fluctuations in these structures, but is also a part of them, both driven by these changes and driving them. Voloshinov (1973:118) writes that ‘(l)anguage reflects not subjective psychological vacillations, but stable social interrelationships among speakers.’ Speech practices are social processes based on interrelationships, which are often implicit until the stability of these social relationships is interrupted, at which point hierarchies are laid bare. Language ideology changes in tune with the restructuring of social hierarchies, and speech practices in turn indicate these shifts. The position of Roma in Kosovo is not simply a case of being in-between, or out of place. Rather, hierarchies and fluctuations in the stability of relationships, which simultaneously construct language ideologies and are constructed by speech practices, lead to the fractalising of inequality: as boundaries are drawn, the dominated are posited as in-between.

Notes 1 See also the edited volume by Guy (2001) entitled Between past and future. 2 See Sigona (2003) for a discussion of the use of the nomad label for Roma in Italy.

Bibliography

Abercrombie, A. forthcoming. Diglossic learning: Romani language in Prizren. In: Tarsoly, E. & Ćalić, J. (Eds.), Fashioning Language. The Politics of Teaching and Learning Languages. London: UCL Press. Abu-Lughod, L. & Lutz, C. 1990. Introduction: Emotion, discourse, and the politics of everyday life. In: Lutz, C. & Abu-Lughod, L. (Eds.), Language and the Politics of Emotion. Cambridge: Cambridge University Press. Pp. 1–23. Abu-Lughod, L. 2002. Do Muslim women really need saving? Anthropological reflections on cultural relativism and its others. American Anthropologist, 104(3), 783–790. Acković, D. 2009. Romi u Beogradu. Istorija, kultura i tradicija Roma u Beogradu. Beograd: Rominterpres. Adams, L. 2013. Ethnicity and the politics of heritage in Uzbekistan. Central Asian Survey, 32(2), 115–133. Agar, M. 1991. The biculture in bilingual. Language in Society, 20(02), 167–182. Agar, M. 2009. A linguistics for ethnography. Why not second Languaculture learning and translation? Journal of Intercultural Communication, 16. Available at: www.immi.se/ intercultural/nr16/agar.htm. Accessed 11/04/20. Agha, A. 2003. The social life of cultural value. Language and Communication, 23(3–4), 231–273. Allcock, J. B. 2000. Explaining Yugoslavia. London: Hurst & Co. Ananiadis, G. 2002. Carl Schmitt on Kosovo, or, taking war seriously. In: Bjelić, D. I. & Savić, O. (Eds.), Balkan as Metaphor: Between Globalization and Fragmentation. Cambridge, MA: MIT Press. Pp. 117–162. Anderson, B. 1983. Imagined Communities. London: Verso. Anderson, D. 1996. Bringing civil society to and uncivilised place. In: Hann, C. & Dunn, E. (Eds.), Civil Society: Rethinking Western Models. London: Routledge. Pp. 99–120. Annamalai, E. 1989. The linguistic and social dimensions of purism. In: Jernudd, B. H. & Shapiro, M. J. (Eds.), The Politics of Language Purism. Berlin/New York: Walter de Gruyter. Pp. 225–231. Arandarenko, M. & Golicin, P. 2007. Serbia. In: Deacon, B. & Stubbs, P. (Eds.), Social Policy and International Interventions in South East Europe. Cheltenham: Edward Elgar Publishing. Pp. 167–186. Archer, R. 2016. ‘Paid for by the workers, occupied by the bureacrats’: Housing inequalities in 1980s Belgrade. In: Archer, R., Duda, I. & Stubbs, P. (Eds.), Social Inequalities and Discontent in Yugoslav Socialism. London/New York: Routledge. Pp. 58–76.

146

Bibliography

ASK. 2015. Kosovo in Figures 1. General Statistics. Prishtina: Kosovo Agency of Statistics. Available at: http://ask.rks-gov.net/en/kosovo-agency-of-statistics. Accessed 01/02/17. ASK. 2016. Results of the Kosovo 2015. Labour Force Survey. Prishtina: Kosovo Agency of Statistics. Available at: http://ask.rks-gov.net/en/kosovo-agency-of-statistics. Accessed 01/02/17. Baker, C. 2012. Prosperity without security: The precarity of interpreters in postsocialist, postconflict Bosnia-Herzegovina. Slavic Review, 71(4), 849–872. Baker, D. 2015. The queer gypsy. Roma Rights Journal. www.errc.org/roma-rights-journ al/roma-rights-2-2015-nothing-about-us-without-us-roma-participation-in-policy-m aking-and-knowledge-production/4433/14. Accessed 12/04/20 Bakhtin, M.M. 1986. Speech genres and other late essays (Holquist, M. and Emerson, C.[Eds.] McGee, V.W. [Trans]). Austin: University of Texas Press. Bakić-Hayden, M. 1995. Nesting orientalisms: The case of former Yugoslavia. Slavic Review, 54(4), 917–931. Ballinger, P. 2004. ‘Authentic hybrids’ in the Balkan Borderlands. Current Anthropology, 45(1), 31–60. Baučić, I. 1971. Vanjske migracije radne snage kao Faktor socijalnoekonomskih promjena u seoskim sredinama. Sociologija i Prostor, 31–32, 127–140. Bauman, R. & Briggs, C. L. 2000. Language philosophy and language as ideology: John Locke and Johann Gottfried Herder. In: Kroskrity, P. V. (Ed.), Regimes of Language: Ideologies, Polities, and Identities. Santa Fe, NM: School of American Research Press. Pp. 139–204. BBC. 1999. Analysis: Gypsies ‘caught in the middle’. Available at: http://news.bbc.co.uk /1/hi/world/europe/386945.stm. Accessed 12/04/2020. Beasley von Burg, A. 2009. Free to move but nowhere to go: The European Union and the Roma population. Paper Presented at the European Union Studies Association Conference, Los Angeles, CA, April 22, 2009. Available at: http://aei.pitt.edu/33026/. Accessed 12/04/2020. Beha, A. 2014. Minority rights: An opportunity for adjustment of ethnic relations in Kosovo. JEMIE, 13, 85–110. Beissinger, M. H. 2001. Occupation and ethnicity: Constructing identity among professional Romani (Gypsy) musicians in Romania. Slavic Review, 60(1), 24–49. Beltran, C. 2004. Patrolling borders: Hybrids, hierarchies and the challenge of mestizaje. Political Research Quarterly, 57(4), 597–607. Bhabha, H. K. & Rutherford, J. 2006. Third space. Multitudes, 3, 95–107. Blackledge, A. & Creese, A. 2010. Multilingualism: A Critical Perspective. London: Bloomsbury Publishing. Bloch, M. 1991. Language, anthropology and cognitive science. Man, 26(2), 183–198. Block, D. 2012. Unpicking agency in sociolinguistics research with migrants. In: Gardner, S. & Martin-Jones, M. (Eds.), Multilingualism, Discourse, and Ethnography. London: Routledge. Pp. 54–67. Blommeart, J. 2003. Commentary: A sociolinguistics of globalization. Journal of Sociolinguistics, 7(4), 607–623. Blumi, I. 2002. The role of education in the formation of Albanian identity and its myths. In: Schwandner-Sievers, S. & Fischer, B. J. (Eds.), Albanian Identities: Myth and History. London: Hurst. Pp. 49–59. Bonfiglioli, C. 2015. Gendered citizenship in the global European periphery: Textile workers in post-Yugoslav states. Women’s Studies International Forum, 49, 57–65.

Bibliography 147 Bortolotto, C. 2007. From objects to processes: UNESCO’s intangible cultural heritage. Journal of Museum Ethnography, 19, 21–33. Bourdieu, P. 1990. The Logic of Practice. Palo Alto: Stanford University Press. Bourdieu, P. 1991. Language and Symbolic Power. Cambridge: Polity Press. Bourdieu, P. 2004. The peasant and his body. Ethnography, 5(4), 579–599. Boym, S. 2002. The Future of Nostalgia. New York: Basic Books. Bracewell, W. & Drace-Francis, A. 1999. South-Eastern Europe: History, concepts, boundaries. Balkanologie, 3(2), 47–66. Bradley, J. 1988. Yanyuwa: ‘Men speak one way, women speak another’. Aboriginal Linguistics, 1, 126–134. Brandstädter, S. 2007. Transitional spaces: Postsocialism as a cultural process. Critique of Anthropology, 27(2), 131–145. Bringa, T. 1995. Being Muslim the Bosnian Way: Identity and Community in a Central Bosnian Village. Princeton: Princeton University Press. Brooks, E. C. 2012. The possibilities of Romani feminism. Signs, 38(1), 1–11. Brubaker, R. 2006. Nationalist Politics and Everyday Ethnicity in a Transylvanian Town. Princeton: Princeton University Press. Brunnbauer, U. 2004. (Re)Writing History. Historiography in Southeast Europe After Socialism. Münster: Lit-Verlag. Bucholtz, M. 2003. Sociolinguistic nostalgia and the authentication of identity. Journal of Sociolinguistics, 7(3), 398–416. Bugarski, R. 1991. Language in Yugoslavia: Situation, policy, planning. In: Bugarski, R. & Hawkesworth, C. (Eds.), Language Planning in Yugoslavia. Ohio: Slavica Publishers. Pp. 9–28. Burnett, J. H. 1974. On the analog between culture acquisition and ethnographic method. Council on Anthropology and Education Quarterly, 5(1), 25–29. Byram, M. & Esarte-Sarries, V. 1991. Investigating Cultural Studies in Foreign Language Teaching: A Book for Teachers (No. 62). Bristol: Multilingual Matters Limited. Byram, M. 1997. Introduction: Towards a pedagogical framework for visits and exchanges. In: Byram, M. (Ed.), Face to Face: Learning ‘Language-and-Culture’ Through Visits and Exchanges. London: Centre for Information on Language Teaching and Research (CILT). Pp. 3–16. Calhoun, C. J. 2002. The class consciousness of frequent travelers: Toward a critique of actually existing cosmopolitanism. The South Atlantic Quarterly, 101(4), 869–897. Campbell, D. 1998. National Deconstruction: Violence, Identity, and Justice in Bosnia. Minneapolis: University of Minnesota Press. Canagarajah, S. 2013. Translingual Practice: Global Englishes and Cosmopolitan Relations. Oxford: Routledge. Carrier, J. 2015. The concept of class. In: Carrier, J. G. & Kalb, D. (Eds.), Anthropologies of Class: Power, Practice, and Inequality. Cambridge: Cambridge University Press. Pp. 28–40. Chari, S. & Verdery, K. 2009. Thinking between the posts: Postcolonialism, postsocialism, and ethnography after the Cold War. Comparative Studies in Society and History, 51(1), 6–34. Clifford, J. 1988. The Predicament of Culture. Cambridge, MA: Harvard University Press. Cocozzelli, F. P. 2007. Kosovo. In: Deacon, B. & Stubbs, P. (Eds.), Social Policy and International Interventions in South East Europe. Cheltenham: Edward Elgar Publishing. Pp. 203–220.

148

Bibliography

Cocozzelli, F. P. 2009. The Kosovo conflict (1981–1999) and the creation of ethnicized institutions. In: Cocozzelli, F. & Chung, P. S. (Eds.), War and Social Welfare: Reconstruction After Conflict. New York: Palgrave Macmillan. Pp. 33–60. Cohen, A. P. 1985. The Symbolic Construction of Community. London: Ellis Horwood. Collins, P. H. 1998. It’s all in the family: Intersections of gender, race, and nation. Hypatia, 13(3), 62–82. Courthiade, M. 1989. La langue Romani (Tsigane): Évolution, standardisation, unification, réforme. In: Fodor, I. & Hagège, C. (Eds.), Hamburg Language Reform. History and Future, Vol IV. Hamburg: Helmut Buske Verlag. Pp. 87–110. Cowan, J., Dembour, M. & Wilson, R. 2001. Introduction. In: Cowan, J., Dembour, M. & Wilson, R. (Eds.), Culture and Rights: Anthropological Perspectives. Cambridge: Cambridge University Press. Pp. 1–26. Creed, G. 2006. Reconsidering community. In: Creed, G. (Ed.), The Seductions of Community: Emancipations, Oppressions, Quandaries. Santa Fe: School of American Research Press. Pp. 3–22. Crenshaw, K. 1989. Demarginalizing the intersection of race and sex: A black feminist critique of antidiscrimination doctrine, feminist theory and antiracist politics. University of Chicago Legal Forum, 1989, 8. Available at: https://chicagounbound.uchicago.edu/ uclf/vol1989/iss1/8. Accessed 12/04/20 Crystal, D. 2008. A Dictionary of Linguistics and Phonetics. 6th edition. Oxford: Blackwell. Danforth, L. M. 1993. Claims to Macedonian identity: The Macedonian question and the breakup of Yugoslavia. Anthropology Today, 9(4), 3–10. Daniel, E. V. 1996. Crushed glass: A counterpoint to culture. In: Daniel, E. V. (Ed.), Charred Lullabies: Chapters in an Anthropography of Violence. Princeton: Princeton University Press. Pp. 194–212. Daniel, O. 2007. Gastarbajteri: Rethinking Yugoslav economic migrations towards the European North-West through transnationalism and popular culture. In: Ellis, S. G. & Klusáková, L. A. (Eds.), Imagining Frontiers, Contesting Identities. Pisa, Italy: Edizioni Plus, Pisa University Press. Pp. 277–302. de Beauvoir, S. 2014 [1949]. The Second Sex. London: Vintage Books. de Swaan, A. 1995. Widening circles of identification: Emotional concerns in sociogenetic perspective. Theory, Culture, and Society, 12(2), 25–39. Deacon, B. & Stubbs, P. 2007. Transnationalism and the making of social policy in South East Europe. In: Deacon, B. & Stubbs, P. (Eds.), Social Policy and International Interventions in South East Europe. Cheltenham: Edward Elgar Publishing. Pp. 1–21. Despalatović, E. M. 1975. Ljudevit Gaj and the Illyrian Movement. New York: Colombia University Press. Ditchev, I. 2002. The Eros of identity. In: Bjelić, D. I. & Savić, O. (Eds.), Balkan as Metaphor: Between Globalization and Fragmentation. Cambridge, MA: MIT Press. Pp. 235 –250. Dobreva, N. 2007. Constructing the ‘Celluloid Gypsy’: Tony Gatlif and Emir Kusturica’s Gypsy films in the context of New Europe. Romani Studies, 17(2), 141–153. Đorđević, D. 1975. Balkan versus European enlightenment: Parallelisms and dissonances. East European Quarterly, 9(4), 486–497. Douglas, M. 1966. Purity and Danger: An Analysis of Concepts of Pollution and Taboo. London: Routledge. Duijzings, G. 1995. The exodus of Letnica-Croatian refugees from Kosovo in western Slavonia A chronicle. Narodna Umjetnost: Hrvatski Časopis Za Etnologiju i Folkloristiku, 32(2), 152–152.

Bibliography 149 Duijzings, G. 1997. The making of Egyptians in Kosovo and Macedonia. In: Govers, C. & Vermeulen, H. (Eds.), The Politics of Ethnic Consciousness. Basingstoke: Palgrave Macmillan UK. Pp. 194–222. Duijzings, G. 1999. Religion and the Politics of Identity in Kosovo. London: Hurst. Duijzings, G. 2003. Ethnic unmixing under the aegis of the West: A transnational approach to the breakup of Yugoslavia. Bulletin of the Royal Institute for Inter-Faith Studies, 5(2), 1–16. Duijzings, G. 2005. New myths are needed: Reconciliation in Kosovo as an intellectual challenge. Studia Historica Nitriensia, 12, 263–267. Duijzings, G. 2010. From bongo bongo to Boston via the Balkans: Anthropological contributions to the study of urban transformations in southeastern Europe. In: Bohn, T. M. & Calic, M. J. (Eds.), Urbanisierung und Stadtentwicklung in Südosteuropa vom 19. bis zum 21. Jahrhundert. 47. Munich: Otto Sagner (Südosteuropa-Jahrbuch 37). Pp. 93–132. Dunn, C. D. 1999. Toward the study of communicative development as a life-span process. Anthropology and Education Quarterly, 30(4), 451–454. EC Ma Ndryshe. 2015. Erroneous: An Analysis of Numerous and Continuous Faults in Cultural Heritage. Series Analysis “What Went Wrong”? Prishtina: KFOS. Edwards, J., Evans, G. & Smith, K. 2012. Introduction: The middle class-ification of Britain. Focaal, 2012(62), 3–16. Elshani, I. 2011. Romani problematika. Yekhipe, 40, 7–9. Erikson, K. 1995. Notes on trauma and community. In: Caruth, C. (Ed.), Trauma. Explorations in Memory. Baltimore/London: The John Hopkins University Press. Pp. 183–199. Esbenshade, R. S. 1995. Remembering to forget: Memory, history, national identity in postwar East-Central Europe. Representations, 49(1), 72–96. Evans, G. 2006. Learning, violence and the social structure of value. Social Anthropology, 14(2), 247–259. Ferguson, J. & Gupta, A. 2002. Spatializing states: Toward an ethnography of neoliberal governmentality. American Ethnologist, 29(4), 981–1002. Ferguson, J. 2006. Global Shadows: Africa in the Neoliberal World Order. Durham: Duke University Press. Fernandez, J. W. & Taylor Huber, M. 2001. Introduction: The anthropology of irony. In: Fernandez, J. W. & Taylor Huber, M. (Eds.), Irony in Action: Anthropology, Practice and Modern Imagination. Chicago/London: University of Chicago Press. Pp. 1–41. Fleming, K. E. 2000. Orientalism, the Balkans, and Balkan historiography. The American Historical Review, 105(4), 1218–1233. Foster, V. A. 2004. The Name and Nature of Tragicomedy. Hants: Ashgate Publishing Ltd. Fraser, N. 2000. Rethinking recognition. New Left Review, 3, 107–120. Friedman, V. A. 1982. Balkanology and Turcology: West Rumelian Turkish in Yugoslavia as reflected in prescriptive grammar. Studies in Slavic and General Linguistics, 2, 1–77. Friedman, V. A. 1984. Romani te in a Balkan context. Językowe Udia Balkanistyczne, 1, 39–48. Friedman, V. A. 1985a. Problems in the codification of a standard Romani literary language. Papers from the Fourth and Fifth Annual Meetings: Gypsy Lore Society. New York: Gypsy Lore Society. Friedman, V. A. 1985b. Balkan Romani modality and other Balkan languages. Folia Slavica, 7(3), 381–389.

150

Bibliography

Friedman, V. A. 1991. Case in Romani: Old grammar in new affixes. The Journal of the Gypsy Law Society, 1(2), 85–102. Friedman, V. A. 2001. Romani multilingualism in its Balkan context. Sprachtypologie und Universalienforschung, 54(2), 148–161. Friedman, V. A. 2002a. Romani as minority language, as a standard language, and as a contact language: Comparative legal, sociolinguistic, and structural approaches. Proceedings of the Eighth Nordic Conference on Bilingualism, Stockholm University, Stockholm. Friedman, V. A. 2002b. Boundaries and borders in Balkan Slavic. International Journal of Slavic Linguistic and Poetics, 44–45, 161–173. Friedman, V. A. 2005. The Romani language in Macedonia in the third millennium: Progress and problems. In: Matras, Y. (Ed.), Romani in Contact: The History, Structure, and Sociology of a Language. Amsterdam: Benjamins. Pp. 203–217. Gal, S. 1978. Peasant men can’t get wives: Language change and sex roles in a bilingual community. Language in Society, 7(01), 1–16. Gal, S. 1989. Language and political economy. Annual Review of Anthropology, 18(1), 345–367. Gal, S. 1991. Bartok’s funeral: Representations of Europe in Hungarian political rhetoric. American Ethnologist, 18(3), 440–458. Gal, S. & Irvine, J. T. 1995. The boundaries of languages and disciplines: How ideologies construct difference. Social Research, 62(4), 967–1001. Gal, S. 2005. Language ideologies compared. Journal of Linguistic Anthropology, 15(1), 23–37. Gal, S. 2006a. Contradictions of standard language in Europe: Implications for the study of practices and publics. Social Anthropology, 14(2), 163–181. Gal, S. 2006b. Linguistic anthropology. In: Brown, K. (Ed.), Encyclopedia of Language and Linguistics. 2nd edition, volume 7. Amsterdam: Elsevier. Pp. 171–185. Gal, S. & Kligman, G. 2012. The Politics of Gender After Socialism: A ComparativeHistorical Essay. Princeton: Princeton University Press. Gal, S. & Woolard, K. 2014. Languages and Publics: The Making of Authority. London: Routledge. Gamonte, D. S. 2013. Diasporic Roma Students: Performing Ethnic Identity. Sense of Displacement and In-between-ness. Unpublished doctoral thesis. Budapest: Central European University. Garcia, O. & Mason, L. 2009. Where in the world is US Spanish? Creating a space of opportunity for US Latinos. In: Harbert, W. (Eds.), Language and Poverty. Bristol: Multilingual Matters. Pp. 78–101. García, O. & Wei, Li. 2013. Translanguaging: Language, Bilingualism and Education. New York: Springer. Garrett, P. B. 2005. What a language is good for: Language socialization, language shift, and the persistence of code-specific genres in St. Lucia. Language in Society, 34(03), 327–361. Garrett, P. B. 2008. Researching language socialization. In: May, S. & Hornberger, N. H. (Eds.), Encyclopedia of Language and Education: Language Policy and Political Issues in Education. New York: Springer. Pp. 3386–3398. Gay y Blasco, P. 2008. Picturing ‘Gypsies’ interdisciplinary approaches to Roma representation. Third Text, 22(3), 297–303. Gellner, E. 1983. Nations and Nationalism. Oxford: Blackwell.

Bibliography 151 Goldstein, L. 1990. The linguistic interest of verbal humor. Humor-International Journal of Humor Research, 3(1), 37–52. Goldstein, P. 2013. Building Bridges: NGOs, Associations, Movements, Facebook Groups, and the State of Civil Society in Mostar and Novi Sad. Unpublished doctoral thesis. Manchester: University of Manchester. Goldsworthy, V. 2002. Invention and in(ter)vention: The rhetoric of balkanization. In: Bjelić, D. I. & Savić, O. (Eds.), Balkan as Metaphor: Between Globalization and Fragmentation. Cambridge, MA: MIT Press. Pp. 23–38. Golovko, E. 2003. Language contact and group identity: The role of “folk” linguistic engineering. In: Matras, Y. & Bakker, P. (Eds.), The Mixed Language Debate: Theoretical and Empirical Advances. Berlin/New York: Walter de Gruyter. Pp. 177–208. Gorski, G. 2005. The Shutka Book of Records: Sampionite od Sutka (Knjiga rekorda Sutke) CEL FILM. Available at: www.youtube.com/watch?v=RPJnukaw-3Q. Acessed 12/04/20. Gow, J. 1997. After the flood: Literature on the context, causes and course of the Yugoslav War reflections and refractions. Slavonic and East European Review, 75(3), 446–484. Grant, B. 2010. Cosmopolitan Baku. Ethnos, 75(2), 123–147. Grdešić, M. 2015. Workers and unions after socialism. In: Horvat, S. & Štiks, I. (Eds.), Welcome to the Desert of Post-Socialism: Radical Politics After Yugoslavia. London: Verso. Pp. 65–84. Green, S. F. 2005. Notes from the Balkans: Locating Marginality and Ambiguity on the Greek-Albanian Border. Princeton: Princeton University Press. Greenberg, J. 2011. On the road to normal: Negotiating agency and state sovereignty in postsocialist Serbia. American Anthropologist, 113(1), 88–100. Greenberg, R. D. 2004. Language and Identity in the Balkans: Serbo-Croatian and Its Disintegration. Oxford: Oxford University Press. Grosjean, F. 2008. Studying Bilinguals. Oxford: Oxford University Press. Gumperz, J. & Hymes, D. 1972. Directions in Sociolinguistics: The Ethnography of Communication. New York: Holt, Reinhart and Winston. Gupta, A. & Ferguson, J. 1992. Beyond ‘culture’: Space, identity, and the politics of difference. Cultural Anthropology, 7(1), 6–23. [reprinted in Gupta, A. & Ferguson, J. (Eds.) 1997. Culture, power, place. Durham: Duke University Press.] Guthke, K. S. 1966. Modern Tragicomedy. New York: Random House. Gutiérrez, K. D., Baquedano‐López, P. & Tejeda, C. 1999. Rethinking diversity: Hybridity and hybrid language practices in the third space. Mind, Culture, and Activity, 6(4), 286–303. Guy, W. 2001. Between past and Future: The Roma of Central and Eastern Europe. Hatfield: University of Hertfordshire Press. Haliliuc, A. 2015a. What is so funny about the death of Mr. Lăzărescu? The Journal of Popular Culture, 48(1), 154–167. Haliliuc, A. 2015b. Manele music and the discourse of balkanism in Romania. Communication, Culture and Critique, 8(2), 290–308. Halwachs, D. W. 2011. Language planning and media: The case of Romani. Current Issues in Language Planning, 12(3), 381–401. Halwachs, D. W. 2012a. Linguistic diversity, dominated languages and the Internet: The case of Romani. In: Linguistic and Cultural Diversity in Cyberspace. Proceedings of the 2nd International Conference. Paris/Moscow: UNESCO. Pp. 272–281.

152

Bibliography

Halwachs, D. W. 2012b. Romani teaching: Some general considerations based on model cases. European Yearbook of Minority Issues Online, 9(1), 249–269. Hann, C. 1996. Political society and civil anthropology. In: Hann, C. & Dunn, E. (Eds.), Civil Society: Rethinking Western Models. London: Routledge. Pp. 1–26. Haspelmath, M. 2006. Against markedness (and what to replace it with). Journal of Linguistics, 42(01), 25–70. Haugen, E. 1966. Dialect, language, nation. American Anthropologist, 68(4), 922–935. Helbig, A. 2010. The dialogics of development: NGOs, ethnopolitics, and Roma in Ukraine. In: D’Anieri, P. (Ed.), Orange Revolution and Aftermath: Mobilization, Apathy, and the State in Ukraine. Baltimore: Woodrow Wilson International Center for Scholars/John Hopkins University Press. Pp. 254–273. Heller, M. 1988. Introduction. In: Heller, M. (Ed.), Codeswitching: Anthropological and Sociolinguistic Perspectives. Berlin/New York/Amsterdam: Mouton de Gruyter. Pp. 53–76. Heller, M. 2003. Globalization, the new economy, and the commodification of language and identity. Journal of Sociolinguistics. Special Issue: Sociolinguistics and Globalization, 7(4), 473–492. Heller, M. 2012. Rethinking sociolinguistic ethnography: From community and identity to process and practice. In: Gardner, S. & Martin-Jones, M. (Eds.), Multilingualism, Discourse, and Ethnography. London: Routledge. Pp.24–33. Helms, E. 2010. The gender of coffee: Women and reconciliation initiatives in post-war Bosnia and Herzegovina. Focaal, 2010(57), 17–32. Henig, D. 2012. Knocking on my neighbour’s door: On metamorphoses of sociality in rural Bosnia. Critique of Anthropology, 32(1), 3–19. Herzfeld, M. 2001. Irony and power: Toward a politics of mockery in Greece. In: Fernandez, J. W. & Taylor Huber, M. (Eds.), Irony in Action: Anthropology, Practice & Modern Imagination. Chicago/London: University of Chicago Press. Pp. 63–86. Herzfeld, M. 2004. Cultural Intimacy: Social Poetics in the Nation-State. New York: Routledge. Herzfeld, M. 2007. Deskilling, ‘dumbing down,’ and the auditing of knowledge in the practical mastery of artisans and academics: An ethnographer’s response to a global problem. In: Harris, M. (Ed.), Ways of Knowing. New Approaches in the Anthropology of Experience and Learning. New York/Oxford: Berghahn Books. Pp. 91–113. Herzfeld, M. 2009. The cultural politics of gesture: Reflections on the embodiment of ethnographic practice. Ethnography, 10(2), 131–152. Heyd, U. 1954. Language Reform in Modern Turkey. Jerusalem: Israel Oriental Society. Hill, J. 1998. “Today there is no respect”: Nostalgia, “respect,” and oppositional discourse in Mexicano (Nahuatl) language ideology. In: Schieffelin, B. B., Woolard, K. A. & Kroskrity, P. V. (Eds.), Language Ideologies: Practice and Theory. Oxford: Oxford University Press. Pp. 68–86. Hill, J. H. & Hill, K. C. 1980. Mixed grammar, purist grammar, and language attitudes in modern Nahuatl. Language in Society, 9(03), 321–348. Hill, J. H. & Coombs, D. M. 1982. The vernacular remodelling of national and international languages. Applied Linguistics, 3(3), 224–234. Hirst, D. L. 1984. Tragicomedy. London/New York: Methuen. Hodges, A. 2016. Croatian language standardization and the production of nationalized political subjects through language? Perspectives from the social sciences and humanities. Etnološka Tribina: Journal of Croatian Ethnological Society, 46(39), 3–45.

Bibliography 153 Hoepken, W. 1998. War, memory, and education in a fragmented society: The case of Yugoslavia. East European Politics and Societies, 13(1), 190–227. Hoeppe, G. 2007. Conversations on the Beach: Fisherman’s Knowledge, Metaphor and Environmental Change in South India. New York/Oxford: Berghahn Books. Homer, S. 2006. “The Roma do not exist”: The Roma as an object of cinematic representation and the question of authenticity. Gramma, Journal of Theory and Criticism, 14, 183–198. Human Rights Watch. 1999. Federal Republic of Yugoslavia: Abuses against Serbs and Roma in the New Kosovo. Human Rights Watch, 11, 10 (D). Available at: www.hrw .org/reports/1999/kosov2/. Acessed 12/04/20. Hylland Eriksen, T. 1993. Ethnicity and Nationalism: Anthropological Perspectives. London: Pluto Press. Hymes, D. 1972. On communicative competence. In: Pride, J.B. and Holmes, J. (Eds), Sociolinguistics. Selected Readings. Harmondsworth: Penguin. Pp. 269–293. Ilisei, I. 2012. Education of Roma women between feminism and multiculturalism study case: Roma women in Romania. JSSE-Journal of Social Science Education, 12(1), 67–75. Iordanova, D. 2003. Images of Romanies in cinema: A rough sketch? Framework: The Journal of Cinema and Media, 44(2), 5–14. Iordanova, D. 2008. Mimicry and plagiarism: Reconciling actual and metaphoric Gypsies. Third Text, 22(3), 305–310. Irvine, J. & Gal, S. 2000. Language ideology and linguistic differentiation. In: Kroskrity, P. V. (Ed.), Regimes of Language: Ideologies, Polities, and Identities. Santa Fe, NM: School of American Research Press. Pp. 35–84. Jaffe, A. 1999. Ideologies in Action: Language Politics on Corsica. Berlin/New York: Walter de Gruyter. Jaffe, A. 2003. Misrecognition unmasked? ‘Polynomic’ language, expert statuses and orthographic practices in Corsican schools. Pragmatics, 13(4), 515–539. Jansen, S. 2005. Who’s afraid of white socks? Towards a critical understanding of postYugoslav urban self-perceptions. Ethnologia Balkanica, 9, 151–167. Jansen, S. 2008. Cosmopolitan openings and closures in post-Yugoslav antinationalism. In: Nowicka, M. & Rovisco, M. (Eds.), Cosmopolitanism in Practice. Aldershot: Ashgate. Pp. 75–92. Jansen, S. & Löfving, S. 2008. Towards an anthropology of violence, hope, and the movement of people. In: Jansen, S. & Löfving, S. (Eds.), Struggles for Home: Violence, Hope and the Movement of People. Oxford: Berghahn Books. Pp. 1–23. Jansen, S. 2009. After the red passport: Towards an anthropology of the everyday geopolitics of entrapment in the EU’s ‘immediate outside’. Journal of the Royal Anthropological Institute, 15(4), 815–832. Jašarević, L. 2007. Everyday work: Subsistence economy, social belonging and moralities of exchange at a Bosnian (black) market. In: Bougarel, X., Helms, E. & Duijzings, G. (Eds.), The New Bosnian Mosaic: Identities, Memories and Moral Claims in a PostWar Society. Aldershot: Ashgate. Pp. 273–294. Jeffrey, A. 2007. The geopolitical framing of localized struggles: NGOs in Bosnia and Herzegovina. Development and Change, 38(2), 251–274. Jenkins, T. 1994. Fieldwork and the perception of everyday life. Man, 29(2), 433–455. Jernudd, B. H. 1989. The texture of language purism: An introduction. In: Jernudd, B. H. & Shapiro, M. J. (Eds.), The Politics of Language Purism. Berlin/New York: Walter de Gruyter. Pp. 1–20.

154

Bibliography

Jordan, B. 1989. Cosmopolitical obstetrics: Some insights from the training of traditional midwives. Social Science and Medicine, 28(9), 925–937. Jovanović, J. & Daroczi, A. C. 2015. Still missing intersectionality: The relevance of feminist methodologies in the struggle for the rights of Roma. Roma Rights Journal, 2, 79–82. Available at: www.errc.org/roma-rights-journal/roma-rights-2-2015-nothingabout-us-without-us-roma-participation-in-policy-making-and-knowledge-production /4433/12. . Accessed 12/04/20. Jović, D. 2003. Yugoslavism and Yugoslav communism: From Tito to Kardelj. In: Djokić, D. (Ed.), Yugoslavism: Histories of a Failed Idea. Madison: The University of Wisconsin Press. Pp. 157–181. Kalb, D. 2009. Conversations with a Polish populist: Tracing hidden histories of globalization, class, and dispossession in postsocialism (and beyond). American Ethnologist, 36(2), 207–223. Kalb, D. 2014. Class. In: Nonini, D. (Ed.), A Companion to Urban Anthropology. Oxford: Blackwell. Pp. 157–176. Kapralski, S. 1997. Identity building and the Holocaust: Roma political nationalism. Nationalities Papers, 25(2), 269–283. Kent, S. A. 1997. Writing the Yugoslav Wars: English-language books on Bosnia (1992– 1996) and the challenges of analyzing contemporary history. The American Historical Review, 102(4), 1085–1114. Kibler, R. E. 2004. Responses to inhumanity in the Balkans and a preliminary discussion concerning the problem of evil. East European Quarterly, 38(4), 463–472. Kiossev, A. 2002. The dark intimacy: Maps, identities, acts of identifications. In: Bjelić, D. I. & Savić, O. (Eds.), Balkan as Metaphor: Between Globalization and Fragmentation. Cambridge, MA: MIT Press. Pp. 165–190. Kirshenblatt‐Gimblett, B. 2004. Intangible heritage as metacultural production. Museum International, 56(1–2), 52–65. Kitromilides, P. M. 1989. ‘Imagined Communities’ and the origins of the national question in the Balkans. European History Quarterly, 19(2), 149–192. Klumbytė, N. 2010. The Soviet sausage renaissance. American Anthropologist, 112(1), 22–37. Kolstø, P. 2005. Myths & Boundaries in South Eastern Europe. London: Hurst. Kordić, S. 2010. Jezik i nacionalizam. Zagreb: Durieux. Kostovicova, D. 2002. ‘Shkolla shqipe’ and nationhood: Albanians in pursuit of education in the native language in interwar (1918–41) and post-autonomy (1989–98) Kosovo. In: Schwandner-Sievers, S. & Fischer, B. J. (Eds.) Albanian Identities: Myth and History. London: Hurst. Pp. 157–171. Kovats, M. 2003. The politics of Roma identity: Between nationalism and destitution. Open Democracy, 29, 1–8. Kraja, G. 2016. A symbol of repression which should remain just that. Kosovo Two Point Zero. Available at: http://kosovotwopointzero.com/en/a-symbol-of-repression-which -should-remain-just-that/. Accessed 12/04/20. Krasniqi, G. 2015. Equal citizens, uneven communities: Differentiated and hierarchical citizenship in Kosovo. Ethnopolitics, 14(2), 197–217. Kroskrity, P. V. 2000. Regimenting languages: Language ideological perspectives. In: Kroskrity, P. V. (Ed.), Regimes of Language: Ideologies, Polities, and Identities. Santa Fe, NM: School of American Research Press. Pp. 1–34. Krstić, M. 2016. “I’m a poor lonesome cowboy and a long way from home…”: Serbian documentary films about guest workers. Issues in Ethnology and Anthropology, 8(2), 499–518.

Bibliography 155 Kruezi, S. 2008. Standardizumi Romani Qhib. Prizren: Berati. Kruezi, S. 2014. O Roma & i Romani Qhib. Krueziu (Self Published). KTV (Kohavision). 2014. KTV lajme ROMET ne kosove mesojne SERBISHT. Available at: www.youtube.com/watch?v=rwwGc64iE3E. Accessed 12/04/20. Kuutma, K. 2009. Cultural heritage: An introduction to entanglements of knowledge, politics and property. Journal of Ethnology and Folkloristics, 3(2), 5–12. Kyuchukov, H. 2011. Roma girls: Between traditional values and educational aspirations. Intercultural Education, 22(1), 97–104. Labov, W. 2006 [1966]. The Social Stratification of English in New York. Cambridge: Cambridge University Press. Ladányi, J. & Szelényi, I. 2003. Historical variations in inter-ethnic relations: Toward a social history of Roma in Csenyéte, 1857–2000. Romani Studies, 13(1), 1–51. Lampe, J. R. 2000. Yugoslavia as History: Twice There Was a Country. 2nd edition. Cambridge: Cambridge University Press. Lampe, J. R. 2003. The two Yugoslavias as economic unions: Promise and problems. In: Djokić, D. (Ed.), Yugoslavism: Histories of a Failed Idea. Madison: The University of Wisconsin Press. Pp. 182–195. Lave, J. 1982. A comparative approach to educational forms and learning processes. Anthropology and Education Quarterly, 13(2), 181–187. le Normand, B. 2016. The gastarbajteri as a transnational Yugoslav working class. In: Archer, R., Duda, I. & Stubbs, P. (Eds.), Social Inequalities and Discontent in Yugoslav Socialism. London/New York: Routledge. Pp. 38–57. Leggio, D. V. 2011. Cosmopolitan virtual spaces: The Romani case. RomIdent Working Papers. Available at: http://romani.humanities.manchester.ac.uk/virtuallibrary/publicat ions.html. Accessed 12/04/20. Leggio, D. V. 2013. Lace Avilen Ko Radio: Romani Language and Identity on the Internet. Unpublished Doctoral thesis. Manchester: University of Manchester. Leggio, D. V. & Matras, Y. 2013. Social networks as centres of language codification: Romani on YouTube. RomIdent Working Papers. Available at: http://romani.humani ties.manchester.ac.uk/virtuallibrary/publications.html. Lemon, A. 2000. Between Two Fires: Gypsy Performance and Romani Memory from Pushkin to Postsocialism. Durham/London: Duke University Press. Lichnofsky, C. 2013. Ashkali and Egyptians in Kosovo: New ethnic identifications as a result of exclusion during nationalist violence from 1990 till 2010. Romani Studies, 23(1), 29–60. Macarie, I. C. 2014. Half-in, half-out: Roma and non-Roma Romanians with limited rights working and travelling in the European Union. Integrim Online Papers. Available at: www.i ntegrim.eu/wp-content/uploads/2014/02/Macarie-half-in-half-out.pdf. Accessed 12/04/20. Maček, I. 2007. Imitation of life: Negotiating normality in Sarajevo under siege. In: Bougarel, X., Helms, E. & Duijzings, G. (Eds.), The New Bosnian Mosaic: Identities, Memories and Moral Claims in a Post-War Society. Manchester: Ashgate Publishing Ltd. Malcolm, N. 2002. Kosovo: A Short History. 2nd edition. London: Pan Books. Marchand, T. 2007. Crafting knowledge: The role of ‘parsing and production’ in the communication of skill-based knowledge among masons. In: Harris, M. (Ed.), Ways of Knowing. New Approaches in the Anthropology of Experience and Learning. New York/Oxford: Berghahn Books. Pp. 181–202. Marmullaku, R. 1975. Albanian and the Albanians. London: Hurst. Marmullaku, R. 2003. Albanians in Yugoslavia: A personal essay. In: Djokić, D. (Ed.), Yugoslavism: Histories of a Failed Idea. Madison: The University of Wisconsin Press. Pp. 304–316.

156

Bibliography

Marushiakova, E. & Popov, V. 2001. New ethnic identities in the Balkans: The case of the Egyptians. Facta Universitatis-Series: Philosophy and Sociology, 2(8), 465–477. Matras, Y. 1998a. Utterance modifiers and universals of grammatical borrowing. Linguistics, 36–2(2), 281–331. Matras, Y. 1998b. Convergent development, grammaticalization, and the problem of ‘mutual isomorphism’. In: Boeder, W., Schroeder, C., Wagner, K. H. & Wildgen, W. (Eds.), Sprache in Raum und Zeit: Beiträge zur empirischen Sprachwissenschaft. Tübingen: Gunter Narr Verlag. Pp. 89–103. Matras, Y. 1999. Writing Romani: The pragmatics of codification in a stateless language. Applied Linguistics, 20(4), 481–502. Matras, Y. 2000. Fusion and the cognitive basis for bilingual discourse markers. International Journal of Bilingualism, 4(4), 505–528. Matras, Y. & Elšik, V. 2001. Romani Dialectological Questionnaire. Manchester: University of Manchester. Matras, Y. 2002. Romani Linguistics. Cambridge: Cambridge University Press. Matras, Y. & Bakker, P. 2003. The study of mixed languages. In: Matras, Y. & Bakker, P. (Eds.), The Mixed Language Debate: Theoretical and Empirical Advances. Berlin/New York: Walter de Gruyter. Pp. 1–20. Matras, Y. 2004. Layers of convergent syntax in Macedonian Turkish. Mediterranean Language Review, 15, 63–86. Matras, Y. & Tufan, Ş. 2007. Grammatical borrowing in Macedonian Turkish. In: Matras, Y. & Sakel, J. (Eds.), Grammatical Borrowing in Cross-Linguistic Perspective (Vol. 38). Berlin/New York: Walter de Gruyter. Pp. 215–227. Matras, Y. 2009. Language Contact. Cambridge: Cambridge University Press. Matras, Y. 2011. Universals of structural borrowing. In: Siemund, P. (Ed.), Linguistic Universals and Language Variation. Berlin: Walter de Gruyter. Pp. 204–235. Matras, Y. 2013. Scholarship and the politics of Romani identity: Strategic and conceptual issues. European Yearbook of Minority Issues Online, 10(1), 209–247. Matras, Y. 2014. I Met Lucky People: The Story of the Romani Gypsies. London: Allen Lane. Mazower, M. 1995. The cold war and the appropriation of memory: Greece after liberation. East European Politics and Societies, 9(2), 272–294. Mazower, M. 2000. A Short History of the Balkans. New York: Modern Library. McClure, E. & McClure, M. 1988. Macro-and micro-sociolinguistic dimensions of codeswitching in Vingard. In: Heller, M. (Ed.), Codeswitching: Anthropological and Sociolinguistic Perspectives. Berlin/New York/Amsterdam: Mouton de Gruyter. Pp. 25–51. Megoran, N. 2013. Shared space, divided space: Narrating ethnic histories of Osh. Environment and Planning A, 45(4), 892–907. Mehan, H. & Griffin, P. 1980. Socialization: The view from classroom interactions. Sociological Inquiry, 50(3–4), 357–392. Memorandum of the European Roma Rights Centre. 2005. In the aftermath of ethnic cleansing: Continued persecution of Roma, Ashkalis, Egyptians and others perceived as “Gypsies” in Kosovo: June 27, 2005. Available at: www.errc.org/article/in-the-afterma th-of-ethnic-cleansing-continued-persecution-of-roma-ashkalis-egyptians-and-others -perceived-as/2460. Accessed 12/04/20. Menekshe, N. 2004. 35 Bersha e romano K.A.M. “Durmish Aslano” 1969–2004. Self-published. Menekshe, N. 2009. Yubileyi 40 Bersh e KAM “Durmish Aslano” ano Prizren 1969 – 2009. Self-published. Messerschmidt, D. A. 1981. Anthropologists at Home in North America: Methods and Issues in the Study of One's Own Society. Cambridge: Cambridge University Press.

Bibliography 157 Mikuš, M. 2011. State pride Politics of LGBT Rights and Democratisation in European Serbia. East European Politics and Societies, 25(4), 834–851. Mikuš, M. 2013. ‘European Serbia’ and its ‘civil’ discontents: Beyond liberal narratives of modernisation. Working Papers Series (Centre for Southeast European Studies) 7. Graz: Centre for Southeast European Studies. Available at: https://pure.mpg.de/pubma n/faces/ViewItemFullPage.jsp?itemId=item_2260567. Accessed 01/02/17. Mikuš, M. 2015. Indigenizing “civil society” in Serbia what local fund-raising reveals about class and trust. Focaal, 2015(71), 43–56. Milroy, J. 2001. Language ideologies and the consequences of standardization. Journal of Sociolinguistics, 5(4), 530–555. Minks, A. 2010. Socializing heteroglossia among Miskitu children on the Caribbean coast of Nicaragua. Pragmatics, 20(4), 495–522. Misha, P. 2002. Invention of a nationalism: Myth and amnesia. In: Schwandner-Sievers, S. & Fischer, B. J. (Eds.), Albanian Identities: Myth and History. London: Hurst. Pp. 33–48. Misini, A. 2014a. Drama - O Roma thay Ashkalie. Available at: www.youtube.com/watch ?v=nEbMWD1IUFw&t=193s. Accessed 12/04/20. Misini, A. 2014b. Radio Prishtina - KAM Durmish Aslano 1972. Available at: www .youtube.com/watch?v=RYdBKHLsCRc. Accessed 12/04/20. Mladenova, R. 2016. The figure of the imaginary gypsy in film: I even met happy Gypsies (1967). Romani Studies, 26(1), 1–30. Moore, L. C. 2004. Second language acquisition and use in the Mandara Mountains. In: Echu, G. & Gyasi Obeng, S. (Eds.), Africa Meets Europe: Language Contact in West Africa. New York: Nova Science. Pp. 131–148. Moore, L. C. 2009. On communicative competence… in the field. Language and Communication, 29(3), 244–253. Moyer, M. G. 2012. Sociolinguistic perspectives on language and multilingualism in institutions. In: Gardner, S. & Martin-Jones, M. (Eds.), Multilingualism, Discourse, and Ethnography. London: Routledge. Pp.34–46. Mueller, J. 2000. The banality of “ethnic war”. International Security, 25(1), 42–70. Myers-Scotton, C. 1988. Codeswitching as indexical of social negotiations. In: Heller, M. (Ed.), Codeswitching: Anthropological and Sociolinguistic Perspectives. Berlin/New York/Amsterdam: Mouton de Gruyter. Pp. 151–186. Myers-Scotton, C. & Bolonyai, A. 2001. Calculating speakers: Codeswitching in a rational choice model. Language in Society, 30(1), 1–28. Naimark, N. & Case, H. (Eds.). 2003. Yugoslavia and Its Historians: Understanding the Balkan Wars of the 1990s. Palo Alto: Stanford University Press. Niedermüller, P. 1998. History, past, and the post-socialist nation. Ethnologia Europaea, 28(2), 169–182. O’Keeffe, B. 2013. New Soviet Gypsies: Nationality, Performance, and Selfhood in the Early Soviet Union. Toronto: University of Toronto Press. Olwig, K. F. & Hastrup, K. 1997. Introduction. In: Olwig, K. F. & Hastrup, K. (Eds.), Siting Culture: The Shifting Anthropological Object. London: Routledge. Pp. 1–15. Oprea, A. 2004. Re-envisioning social justice from the ground up: Including the experiences of Romani women. Essex Human Rights Review, 1(1), 29–39. Oprea, A. 2005. The arranged marriage of Ana Maria Cioaba, intra-community oppression and Romani feminist ideals transcending the ‘primitive culture’ argument. European Journal of Women’s Studies, 12(2), 133–148. Ortner, S. B. 1972. Is female to male as nature is to culture? Feminist Studies, 1(2), 5–31. OSCE. 2015. Prizren municipal profile. Available at: www.osce.org/kosovo/13128 ?download=true. Accessed 12/04/20.

158

Bibliography

Paugh, A. L. 2012. Playing with Languages: Children and Change in a Caribbean Village. New York/Oxford: Berghahn Books. Pelissier, C. 1991. The anthropology of teaching and learning. Annual Review of Anthropology, 20(1), 75–95. Perić, T. & Demirovski, M. 2000. Unwanted: The exodus of Kosovo Roma (1998–2000). Cambridge Review of International Affairs, 13(2), 83–96. Pétric, B. M. 2005. Post-Soviet Kyrgyzstan or the birth of a globalized protectorate. Central Asian Survey, 24(3), 319–332. Petrova, D. 2003. The Roma: Between a myth and the future. Social Research, 70(1), 111–161. Petrović, T. 2003. Studying the minority groups’ identities in the Balkans from the perspective of language ideology. Balcanica, 34(34), 173–188. Petrović, T. 2007. The territory of the former Yugoslavia in the “mental maps” of former Yugoslavs: Nostalgia for space. Sprawy Narodowościowe, 31, 263–273. Petrović, T. 2008. Serbs, Albanians, and those in between: The gradation of otherness and identity management in the nation-building process. Two Homelands, 27, 67–81. Petrović, T. 2010. “When we were Europe”: Socialist workers in Serbia and their nostalgic narratives. The Case of the Cable Factory Workers in Jagodina. In: Todorova, M. (Ed.), Remembering Communism: Genres of Representation. New York: Social Science Research Council. Pp. 127–153. Petrović, T. 2013. European new colonialism. Belgrade Journal of Media and Communications, 2(4), 111–128. Phipps, A. 2010. Ethnographers as language learners: From oblivion and towards an echo. In: Collins, P. & Gallinat, A. (Eds.), The Ethnographic Self as Resource: Writing Memory and Experience into Ethnography. New York/Oxford: Berghahn Books. Pp. 97–110. Pine, F. 2002. Retreat to the household? In: Hann, C. M. (Ed.) 2002. Postsocialism: Ideals, Ideologies, and Practices in Eurasia. London: Routledge. Pp. 95–113. Poplack, S. 1980. Sometimes I’ll start a sentence in Spanish Y TERMINO en ESPANOL: Towards a typology of codeswitching. Linguistics, 18, 581–618. Poulton, H. 2003. Macedonians and Albanians as Yugoslavs. In: Djokić, D. (Ed.), Yugoslavism: Histories of a Failed Idea. Madison: The University of Wisconsin Press. Pp. 115–134. Pratt, M. L. 1991. Arts of the contact zone. Profession, 1, 33–40. Rampton, B. 1995. Crossings: Language and Ethnicity among Adolescents. New York: Longman Group Limited. Razsa, M. & Lindstrom, N. 2004. Balkan is beautiful: Balkanism in the political discourse of Tudman’s Croatia. East European Politics and Societies, 18(4), 628–650. Reeves, M. 2005. Locating danger: Konfliktologiia and the search for fixity in the Ferghana Valley borderlands. Central Asian Survey, 24(1), 67–81. Roberts, C. 1997. The year abroad as an ethnographic experience. In: Byram, M. (Ed.), Face to Face: Learning ‘Language-and-Culture’ Through Visits and Exchanges. London: Centre for Information on Language Teaching and Research (CILT). Pp. 62–76. Romaine, S. 1994. Hawai’i Creole English as a literary language. Language in Society, 23(04), 527–554. Said, E. 1978. Orientalism. New York: Pantheon Books. Salvadori, E. 1997. Taking language beyond the classroom. In: Byram, M. (Ed.), Face to Face: Learning ‘Language-and-Culture’ Through Visits and Exchanges. London: Centre for Information on Language Teaching and Research (CILT). Pp. 18–35.

Bibliography 159 Sampson, S. 1996. The social life of projects: Importing civil society to Albania. In: Hann, C. & Dunn, E. (Eds.), Civil Society: Rethinking Western Models. London: Routledge. Pp. 120–138. Sardelić, J. 2013. Romani minorities on the margins of post-Yugoslav citizenship regimes. CITSEE Working Paper No. 2013/31. Available at: papers.ssrn.com/sol3/papers.cfm ?abstract_id=2388859. Accessed 12/04/20. Sardelić, J. 2016. Roma between ethnic group and an ‘underclass’ as portrayed through newspaper discourses in socialist Slovenia. In: Archer, R., Duda, I. & Stubbs, P. (Eds.), Social Inequalities and Discontent in Yugoslav Socialism. London/New York: Routledge. Pp. 95–111. Schieffelin, B. & Ochs, E. 1986. Language socialization. Annual Review of Anthropology, 15(1), 163–191. Schultz, D. L. 2012. Translating intersectionality theory into practice: A tale of RomaniGadže feminist alliance. Signs, 38(1), 37–43. Schwartz, T. 1981. The acquisition of culture. Ethos, 9(1), 4–17. Shatri, B. 2010. Arsimi shqip në Kosovë, 1990–1999: Shtëpitë shkolla: Sfidat, arritjet dhe aspiratat. Prishtina: Shtëpia Botuese “Libri Shkollor”. Siegel, J. 1995. How to get a laugh in Fijian: Code-switching and humor. Language in Society, 24(1), 95–95. Sigona, N. 2003. How can a nomad be a refugee? Kosovo Roma and labelling policy in Italy. Sociology, 37(1), 69–79. Sigona, N. 2012. Between competing imaginaries of statehood. Roma: Ashkali and Egyptian (RAE) leadership in newly independent Kosovo. Journal of Ethnic and Migration Studies, 38(8), 1213–1232. Silverman, C. 2011. Gypsy music, hybridity and appropriation: Balkan dilemmas of postmodernity. Ethnologia Balkanica, 15, 15–32. Silverman, C. 2015. DJs and the production of “gypsy” music. Western Folklore, 74(1), 5–29. Silverstein, M. 1996. Monoglot “Standard” in America: Standardization and metaphors of linguistic hegemony. In: Brenneis, D. & Macauley, R. (Eds.), The Matrix of Language: Contemporary Linguistic Anthropology. Boulder: Westview Press. Pp. 284–306. Sorabji, C. 1995. A very modern war: Terror and territory in Bosnia-Hercegovina. In: Hinde, R. A. & Watson, H. (Eds.), War: A Cruel Necessity?: The Bases of Institutionalized Violence. New York: IB Tauris. Pp. 80–98. Sorabji, C. 2006. Managing memories in post‐war Sarajevo: Individuals, bad memories, and new wars. Journal of the Royal Anthropological Institute, 12(1), 1–18. Spasić, I. 2011. Cosmopolitanism as discourse and performance: A view from the semiperiphery. Revija za Sociologiju, 41(3), 269–290. Spivak, G. C. 1988. Can the Subaltern Speak? Reflections on the History of an Idea. New York: Columbia University Press. Pp. 21–78. Stefansson, A. 2007. Urban exile: Locals, newcomers and the cultural transformation of Sarajevo. In: Bougarel, X., Helms, E. & Duijzings, G. (Eds.), The New Bosnian Mosaic: Identities, Memories and Moral Claims in a Post-War Society. Aldershot: Ashgate Publishing Ltd. : Pp. 59–78. Stewart, M. 1999. “Brothers” and “orphans”: Images of equality among Hungarian Rom. In: Day, S., Papataxiarchis, E. & Stewart, M. (Eds.), Lilies of the Field. Boulder: Westview Press. Pp. 27–44.

160

Bibliography

Stewart, M. 2002. Deprivation, the Roma and ‘the Underclass’. In: Hann, C. M. (Ed.) 2002. Postsocialism: Ideals, Ideologies, and Practices in Eurasia. . London: Routledge. Pp. 133–155. Stewart, M. 2004. Remembering without commemoration: The mnemonics and politics of Holocaust memories among European Roma. Journal of the Royal Anthropological Institute, 10(3), 561–582. Stokes, G., Lampe, J., Rusinow, D. & Mostov, J. 1996. Instant history: Understanding the Wars of Yugoslav Succession. Slavic Review, 55(1), 136–160. Stolcke, V. 1994. Invaded women: Sex, race and class in the formation of colonial society. The European Journal of Development Research, 6(2), 7–21. Ströhle, I. 2016. Of social inequalities in a socialist society: The creation of a rural underclass in Yugoslav Kosovo. In: Archer, R., Duda, I. & Stubbs, P. (Eds.), Social Inequalities and Discontent in Yugoslav Socialism. London/New York: Routledge. Pp. 112–131. Stubbs, P. 2002. Globalisation, memory and welfare regimes in transition: Towards an anthropology of transnational policy transfers. International Journal of Social Welfare, 11(4), 321–330. Stubbs, P. 2009. Building capacity to promote social integration and social inclusion in the western Balkans. Paper Produced for Expert Group Meeting on Practical Strategies to Promote Social Integration: Lessons Learnt from Existing Policies and Practices. Accra, Ghana. Available at: www.un.org/esa/socdev/egms/docs/2009/Ghana/Stubbs. pdf. Accessed 12/04.20. Stubbs, P. & Maglajlić, R. A. 2012. Negotiating the transnational politics of social work in post-conflict and transition contexts: Reflections from South-East Europe. British Journal of Social Work, 42(6), 1–18. Styan, J. L. 1968. The Dark Comedy: The Development of Modern Comic Tragedy. Cambridge: Cambridge University Press. Sundar, N. 2004. Toward an anthropology of culpability. American Ethnologist, 31(2), 145–163. Sveriges Radio. 2014. Nevi ABC e romenge ani Kosova. Retrieved from sverigesradio.se/ sida/artikel.aspx?programid=2122&artikel=5821074. Accessed 12/04/20. Swaby, N. A. 2014. ‘Disparate in voice, sympathetic in direction’: Gendered political blackness and the politics of solidarity. Feminist Review, 108(1), 11–25. Szeman, I. 2009. “Gypsy Music” and deejays: Orientalism, balkanism, and Romani musicians. TDR/the Drama Review, 53(3), 98–116. Tanasković, D. 1991. The planning of Turkish as a minority language in Yugoslavia. In: Bugarski, R. & Hawkesworth, C. (Eds.), Language Planning in Yugoslavia. Ohio: Slavica Publishers. Pp. 140–161. Tarone, E. 2007. Sociolinguistic approaches to second language acquisition research— 1997–2007. The Modern Language Journal, 91(1), 837–848. Temple, M. 1996. The politicization of history: Marshal Antonescu and Romania. East European Politics and Societies, 10(3), 457–503. Thaçi, H. 2014. Kosovo is fully behind America’s fight against Isis. The Guardian. Available at: www.theguardian.com/commentisfree/2014/sep/30/kosovo-america-isis -radicals-jihadi-networks. Accessed 12/04/20. Thomas, G. 1978. The origin and nature of lexical purism in the Croatian variant of SerboCroatian. Canadian Slavonic Papers, 20(3), 405–420. Thomas, G. 1988. Towards a typology of lexical purism in the Slavic literary languages. Canadian Slavonic Papers, 30(1), 96–111. Thomas, G. 1991. Linguistic Purism. London: Longman Publishing Group.

Bibliography 161 Tiefenbacher, B. 2013. Identifying “Roma” or constructing “the Other”? Slovak Romani men and women in processes of identification. European Yearbook of Minority Issues Online, 10(1), 249–268. Todorova, M. 1994. The Balkans: From discovery to invention. Slavic Review, 53(2), 453–482. Todorova, M. 1997. Imagining the Balkans. Oxford: Oxford University Press. Todorova, M. 2015. Re-imagining the Balkans. In: Horvat, S. & Štiks, I. (Eds.), Welcome to the Desert of Post-Socialism: Radical Politics After Yugoslavia. London: Verso. Pp. 85–102. Tremlett, A. 2009. Comparing European institutional and Hungarian approaches to Roma (Gypsy) minorities. In: Agarin, T. & Brosig, M. (eds.). Minority Integration: Debating Ethnic Diversity in Eastern Europe. Amsterdam: Rodopi. Pp. 129–150. Trubeta, S. 2005. Balkan Egyptians and gypsy/Roma discourse. Nationalities Papers, 33(1), 71–95. Tupas, R. 2014. Inequalities of Multilingualism: Challenges to mother-tongue based multilingual education. Language and Education, 29(2), 1–13. UNESCO. 2015. What is intangible cultural heritage? Available at: unesco.org/culture/ich /index.php?pg=00002. Accessed 12/04/20. Unkovski-Korica, V. 2015. Self-management, development and debt: The rise and fall of the ‘Yugoslav experiment’. In: Horvat, S. & Štiks, I. (Eds.), Welcome to the Desert of Post-Socialism: Radical Politics After Yugoslavia. London: Verso. Pp. 21–44. Üsküp, Makedonya. 2014. Prizren agziyla…. Ah!-Nerde o esçi cünler…. Available at: www.youtube.com/watch?v=bOz2dMLjCns. Accessed 12/04/20. van Baar, H. 2012. Socio-economic mobility and neo-liberal governmentality in postsocialist Europe: Activation and the dehumanisation of the Roma. Journal of Ethnic and Migration Studies, 38(8), 1289–1304. van Boeschoten, R. 2000. The impossible return: Coping with separation and the reconstruction of memory in the wake of the civil war. In: Mazower, M. (Ed.), After the War Was Over: Reconstructing the Family, Nation and State in Greece, 1943-1960. Princeton: Princeton University Press. Pp. 122–140. van de Port, M. 1998. Gypsies, Wars & Other Instances of the Wild. Amsterdam: Amsterdam University Press. Venkatesan, S. 2010. Learning to weave; weaving to learn… what? Journal of the Royal Anthropological Institute, 16(1), 158–175. Verdery, K. 1991. National Ideology under Socialism: Identity and Cultural Politics in Ceausescu's Romania. Berkeley: University of California Press. Verdery, K. 1993. Nationalism and national sentiment in post-socialist Romania. Slavic Review, 52(2), 179–203. Verdery, K. 1996. Nationalism and national sentiment in postsocialist Romania. In: Verdery, K. (Ed.), What Was Socialism and What Comes Next. Princeton: Princeton University Press. Pp. 83–103. Verdery, K. 2002. Whither postsocialism? In: Hann, C. M. (Ed.), Postsocialism: Ideals, Ideologies, and Practices in Eurasia. London: Routledge. Pp. 15–28. Vermonden, D. 2013. Reproduction and development of expertise within communities of practice: A case study of fishing activities in South Buton. In: Heckler, S. (Ed.), Landscape, Process and Power: Re-Evaluating Traditional Environmental Knowledge. New York/Oxford: Berghahn Books. Pp. 205–229. Vetta, T. 2009. ‘Democracy Building’ in Serbia: The NGO effect. Southeastern Europe, 33(1), 26–47.

162

Bibliography

Volčič, Z. 2005. The notion of ‘the West’ in the Serbian national imaginary. European Journal of Cultural Studies, 8(2), 155–175. Voloshinov, V. N. 1973 [1929]. Marxism and the Philosophy of Language. Cambridge, MA/London: Harvard University Press. Vygotsky, L. S. 1978. Mind in Society: The Development of Higher Psychological Processes. Cambridge, MA: Harvard University Press. Wade, P. 2010. The presence and absence of race. Patterns of Prejudice, 44(1), 43–60. Wade, P. 2012a. Race, kinship and the ambivalence of identity. In: Schramm, K., Skinner, D. & Rottenburg, R. (Eds.), Identity Politics and the New Genetics: Re/Creating Categories of Difference and Belonging. New York/Oxford: Berghahn Books. Pp. 79–96. Wade, P. 2012b. Skin colour and race as analytic concepts. Ethnic and Racial Studies, 35(7), 1169–1173. Washabaugh, W. 1979. Linguistic anti-structure. Journal of Anthropological Research, 35(1), 30–46. Wei, Li. 2012. Introduction to part 1. In: Gardner, S. & Martin-Jones, M. (Eds.), Multilingualism, Discourse, and Ethnography. London: Routledge. Pp. 19–23. Wexler, P. 1989. Hieratic components in soviet dictionaries of Yiddish, Dungan, and Belorussian. In: Jernudd, B. H. & Shapiro, M. J. (Eds.), The Politics of Language Purism. Berlin, New York: Walter de Gruyter. Pp. 141–168. Williams, B. 1987. Humor, linguistic ambiguity, and disputing in a Guyanese community. International Journal of the Sociology of Language, 65(65), 79–94. Wilson, D. 1970. The Life and Times of Vuk St. Karadžić, 1787–1864; Literacy, Literature and National Independence in Serbia. Oxford: Clarendon Press. Winnifrith, T. 1992. The Vlachs of the Balkans: A rural minority which never achieved ethnic identity. In: Howell, D. (Ed.), Roots of Rural Ethnic Mobilisation: Comparative Studies on Governments and Non-Dominant Ethnic Groups in Europe, 1850–1940. New York/Dartmouth: European Science Foundation/New York University Press. Pp. 58–73. Wolcott, H. 1982. The anthropology of learning. Anthropology and Education, 13(2), 83–108. Woodward, S. L. 1995. Socialist Unemployment: The Political Economy of Yugoslavia, 1945–1990. Princeton: Princeton University Press. Woolard, K. A. 1985. Language variation and cultural hegemony: Toward an integration of sociolinguistic and social theory. American Ethnologist, 12(4), 738–748. Woolard, K. A. 1988. Codeswitching and comedy in Catalonia. In: Heller, M. (Ed.), Codeswitching: Anthropological and Sociolinguistic Perspectives. Berlin/New York/ Amsterdam: Mouton de Gruyter. Pp. 53–76. Woolard, K. A. 1998. Introduction: Language ideology as a field of inquiry. In: Schieffelin, B. B., Woolard, K. A. & Kroskrity, P. V. (Eds.), Language Ideologies: Practice and Theory. Oxford: Oxford University Press. Pp. 3–50. Yasamee, F. A. 1995. Nationality in the Balkans: The case of the Macedonians. In: Özdoğan, G. G. & Saybaşılı, K. (Eds.), Balkans: A Mirror of the New International Order. Istanbul: EREN Yayıncılık ve Kitap, cılık. Pp. 121–132. Young, V. A. 2004. Your average nigga. College Composition and Communication, 55(4), 693–715. Živković, A. 2015. From the market … to the market: The debt economy after Yugoslavia. In: Horvat, S. & Štiks, I. (Eds.), Welcome to the Desert of Post-Socialism: Radical Politics After Yugoslavia. London: Verso. Pp. 45–64. Živković, M. 2000. Jelly, slush, and red mists: Poetics of amorphous substances in Serbian jeremiads of the 1990s. Anthropology and Humanism, 25(2), 168–182. Zymberi, I. 1991. Albanian in Yugoslavia. In: Bugarski, R. & Hawkesworth, C. (Eds.), Language Planning in Yugoslavia. Ohio: Slavica Publishers. Pp. 130–139.

Index

Acković, Dragoljub 6, 106 Albania 36–37, 41–42, 44–45, 100, 104–108 Albanian language: in Kosovo 41–44, 114; orthography 97–98, 111; in Prizren 2–5, 21–22, 69, 115, 122–123, 126–131, 138–139; radio 35–36 40, 43, 63, 95; and Romani 99–101, 106, 107, 116–119, 134; television 42; variants and standard vi, 42, 98, 103, 109–112; see also language of education; see also RAE alphabet 96–98, 105–106, 111, 113; see also orthography American embassy 33, 54 Arabic language 36, 56–57 Ashkali see RAE Aslano, Durmish 35–36, 51 Balkan Federation 36 Balkanism 13, 75–76, 81–82, 93 BCS: in Albanian 117; as Bosnian/Bosniac language 3–4, 98, 103, 109,120; as Croatian language 3–4, 41, 96–97; general communication 36, 47, 55, 60, 63, 69; names and variants vi, 2–5; orthography 97–104; radio 95; in Romani 30, 100–105, 134; as Serbian language 20–22, 36, 78, 119–124, 128–129, 141; standardization 109, 139; in Turkish 115–116, 122; see also language of education Belgrade 6, 41–43, 106, 108, 141 Berberski, Slobodan 42–43 betweenness 15, 85, 135–137, 139–140 bilingual mode see mode borrowing 30, 99–102, 106–107, 115–119, 122, 136–137 Bosniac language see BCS Bosnian language see BCS brotherhood and unity 121, 124

churches 1, 2, 6, 97, 120 cigan see gypsy citizen 61–62, 67–68, 135, 142 civil society 55, 61–62, 96 codemixing see codeswitching codeswitching 12, 80, 117–118 Cold War 36, 61, 71 collectivisation 37–38 communicative competence 13, 31 Communist Party 36, 37, 42, 43, 44 community 62–63, 108, 123, 127; and culture 55–59; as ethnic group 10, 64, 67–68 cosmopolitanism: multilingualism 2, 13–15, 114, 120, 123–131, 137–138; sedentarist cosmopolitanism 15, 124, 127–130, 138; Yugoslavia 52, 126 Council of Europe 54–55, 58 Courthiade, Marcel 103–108 Croatian language see BCS cultural intimacy 31, 89–91 decentralisation 33–34, 41–42, 112–113, 141 dictionary 99, 105 diglossia 13, 63, 72, 110–111, 143–144 Đilas, Milovan 70 disorientation 76, 89–93, 142, 144 dissonance 12, 64, 71–72, 139, 143–144 drama: group in Prizren 15, 48, 89, 108–109, 75–77, 138; and the gypsy image 144; tragicomedy 92–93 Drenica 36, 44 Durmish Aslano (KAM/NGO) see Aslano, Durmish (person) Ederlezi 59 Egyptian see RAE enclaves 1, 30, 56–57, 104, 133–135

164

Index

English language: and author 2, 4, 21, 49, 79, 114, 119; as international language 1, 23, 55, 60–64, 99, 107, 111 ethnography of language 25, 27, 30, 135 Europe: and Balkanism 82–86, 89–94, 141, 144; eastern Europe 3, 41; Roma in Europe 10, 136–138; south-eastern Europe 7, 104; western Europe and author positionality 24, 31, 137; western Europe and migration 41, 44, 48 factories 38, 43, 48, 52, 58, 66–70, 121 feminism see gender Flag Day 8, 56–58, 85, 89, 130 gaje 3, 20, 133–134 gaji see gaje gajikani see gaje gajo see gaje gasterbajter see migration and migrant workers Gatlif, Tony 82 gender 20–23, 76, 80–88, 92, 118, 142: and education 44, 137; homosexuality 77, 84–85; marriage 47, 60–63, 77–79, 84, 88–89; patriarchy 24, 31, 63, 76, 84–86; feminism 86–87, 136 genre 89–90: speech genre 13–15, 114, 119–124, 128–130, 133, 143 German language 1, 57, 70, 88 gesture 75, 77–81 Gjakova 9, 50, 57, 120 gramatika 101–102, 109–110 guest workers see migration and migrant workers gypsy: and Balkanism 81–83, 91, 92, 144; gender 77, 80, 88, 92; language 2, 80–81; race 80, 83, 88, 92; stereotypes 10, 15, 75, 78, 81, 91, 70; theatricality 77–81 Herder 104, 110 Holocaust 51, 76 homosexuality see gender Hoxha, Enver 36, 41 humour 89–90, 118 hybridity 15, 83, 87–79, 126–127 imagined community 42, 52, 90–91; and author positionality 27–29; international community 51, 64, 99 India 2, 10–11, 23, 34, 104 Iniciativa 6 8, 20, 21, 60, 96 intangible cultural heritage 14, 54–60, 64, 68, 71

intellectuals 22–23, 44, 103–107, 109–111, 116, 119 International Roma Day 75–76, 95–96 International Romani Union 105, 141 intersectionality 75, 86–87, 92, 140 Islam 7, 84–86, 122–126; Balkan Islamic practices 59, 86, 91; Islamic phrases 79; Islamism 7; Sheikhs 56, 58 Jashari, Adem 45 Jeta e Re 8–9, 21, 49 journalists see media Kardelj, Edvard 41 KLA (Kosovo Liberation Army) 1, 33, 44–46, 50–51, 70–72, 121 Kosovo conflict 11, 33, 45–52, 93, 132; emigration 1–3, 47, 57, 122; in Prizren 47–51, 66, 97, 120–124, 128, 142–145; rural migrants 8–9, 47, 125, 131 Kruezi, Selahadin 23, 96–97 Kusturica, Emir 81–83 language acquisition see language learning language contact 4, 12, 80, 102, 109 language of education 4, 26, 36–43, 112; Romani 4, 20, 40–41, 69, 95–97, 107–109; Albanian 20, 36, 37, 41–44; BCS 22, 36, 41–44, 47, 102; Turkish 36–40, 44 language ideology 13–15; and authority 104, 107, 111–112; and cosmopolitanism 104, 126–131, 137; and multilingualism 103–104, 118, 131, 137; and policy 47, 52; and purism 30, 103–109, 118, 126–129, 137–139; and speech practices 13–15, 29–30, 109–110, 118, 128–131 language learning 14; classes 96–97; formal and informal 26, 96, 107, 116, 137, 140; as participant observation 21, 23, 25–31; as a skill 28, 112, 138; as socialisation 26–29; and standard Romani 102, 105, 107, 137, 140 language politics 4–5, 14, 33, 95–96 language skills 28, 112, 138 LDP (Democratic League of Kosovo) 44 lingua franca 41, 130 linguistic accommodation see Romani language: accommodation linguistic prestige 76, 101, 111–112, 128–131, 137, 143 loanwords see borrowing

Index Macedonian language 115 marriage see gender massacres 1, 45–46, 49, 89, 120 maxhup see gypsy media 26, 33, 106–109; Romani journalists 22, 23, 34, 43, 99, 106–107, 110, 112 memory 14, 51, 71, 123, 128 metalinguistic discourse 95, 102, 109, 114, 135, 139 methodology 14, 25–31, 137 migration 37, 44–45, 76–77, 136; and migrant workers 39, 70, 88, 91; remittances 39, 66; and rural migrants 9, 48, 114, 125–131, 137, 143 Milošević, Slobodan 43–46; disorientation 92; era in Kosovo 33, 52, 62, 72, 119, 121; ethnicity 9, 51 minority languages 26, 97, 112 mode 30, 117 monolingual mode see mode monolingualism 12, 118, 127, 137, 138 monolingualist ideology see monolingualism mosques 1, 6, 7, 120 Mother Teresa Society 44, 54, 62 multiculturalism 91, 137, 138, 142: and NGOs 3, 25, 48, 59, 131; in Prizren 6, 80, 116, 120 multilingualism 5, 12–15, 25–26, 41, 15–18, 137–138 nationalism 33, 37, 41–46, 93, 95, 126–128; Albanian nationalism 7–8, 20, 43–46, 126, 130, 141; Romani nationalism 43; Serb nationalism 3, 7, 43–46, 97 nationality policy 41–43, 47 NATO 1, 45–46, 49, 72, 120 nepotism 66–67, 70 Nevo Jivdipe see Jeta e Re New Roma Party of Kosovo 105, 133–134 non-alignment 36, 39 nostalgia 15, 114, 119, 122–124, 128–130 novice 14, 26–27, 29, 31 Orientalism 81, 83 orthography 23, 41, 97–107, 111, 122, 141 Ottomans 1–2, 6–7, 34, 57, 115, 124 parallel state 1, 44–45, 62 parallel system see parallel state participant observation 19, 25–26, 29–31 partisans 33, 36–37, 52, 69 passive resistance 44–46, 70

165

patriarchy see gender Persian language 57 Petrović, Sasha 82 politics of recognition and redistribution 15, 42–43, 52–54, 68–71, 138–144 post-socialism: gender 88 as a lens 71; modernity 89, 93; NGOs 54, 61–63; and nostalgia 121; privatisation 66; Roma 3, 69–70 Potkalaja 2 precarity 51–52, 65–68, 136 Prishtina 1, 10, 19, 42–43, 55, 58 privatisation 46, 62, 66, 71 protests 41–42, 43, 46 purism 13–15, 99, 109, 116, 126–129, 136–139, 142 race 80–87 racism 86–88, 93, 136, 140 Radio Prishtina 33, 34 Radio Romano Avazo 5, 33, 64, 95 RAE: and drama 88; ethnic identity 1, 7–12, 70, 125, 131, 143; and the NGO 56–57; in Prizren 1, 20, 22, 52, 63, 66; unemployment 65 Rahoveci 57 Ramadan 85, 122–123 Rambouillet ceasefire 45, 51 Ranković, Aleksandar 141 recognition see politics of recognition and redistribution redistribution see politics of recognition and redistribution register 35, 108 repertoire: and language ideology 11, 108, 114; multilingual repertoires 11–14, 117– 118, 133; in Terzimahalla 4–5, 19, 111 representation 13, 15, 27, 60, 75–84, 86–93 Republic of Kosova see parallel state rights 45, 63, 59, 60–64, 71, 139; human rights 33, 62; language rights 41–42; minority rights 41, 59, 67–68, 71; workers’ rights 38, 70 Romani activism 2, 10; internationally 86–87, 99, 105, 136, 140, 141; in Prizren, 22, 25, 99 Romani language: accommodation 133–135; diglossia 13, 63, 72, 110, 143–144; language contact 99–103, 109, 115–117, 119, 134–137; learning 25–28, 79; prestige 76, 101, 111–112, 128–131, 137, 143; Prizren 2–15, 19–24, 30, 47, 60; radio 5, 20–26, 33–35, 63, 95–99, 106–109; school classes 20, 40, 69,

166

Index

95–97, 107–109, 116; standard Romani 29, 72, 91, 95–106, 111–113, 137–144; theatre 75–76, 80–81, 90; see also language of education Romanian language 99–100, 111 RTK (Radio Television Kosovo) 26, 67, 97, 106, 112 Rugova, Ibrahim 45 Second World War 5, 8, 36, 51, 76 self-management 38–39, 42, 45, 61–72 Serbian language see BCS Serbo-Croat see BCS sexism 84, 86–88, 136, 140 shaci see migration and migrant workers Shadervan 114, 119, 122, 127 shrines 57, 85 Shutka 81–83 šiptar (derogatory ethnonym) 2, 3, 7, 42 socialisation 12, 14, 27–29 sociolinguistics 12, 29 Soviet Union 41 speech community 101 speech practices Stalin, Joseph 37, 41 standardisation of Romani 3, 15, 95–96, 111, 133, 138–139; and contact languages 103, 109, 119, 139; and dialect variation 102; and purism 100–101, 108, 110, 112; and translation 62

Tito, Josip Broz 36, 41–43, 52, 121 tragicomedy 15, 75, 89–93, 144 translation 60–64, 72 translingualism 26, 128, 137 Turkey 8, 11, 34, 36, 37, 66 Turkish language: Prizren variant 13, 15, 114–119, 122–131, 142–143; and Romani 19, 30, 98–107; television 20, 125; usage in Prizren 1–7, 22, 57, 69, 139; see also language of education underclass 69, 70 unemployment 44–48, 65–68, 121 UNESCO 55–57, 60, 68 volunteers 64–65 war see Kosovo conflict; Second World War World Romani Congress 42–43 Yugoslavia: break-up of Yugoslavia 1, 44–45, 50, 52; Kingdom of Yugoslavia vii, 36, 90; and language 4, 115; and nationality 34, 41–43, 126, 141–142; and nostalgia 43, 48, 70,121; and selfmanagement socialism 38–39, 69, 93; Socialist Yugoslavia vii, 14, 33, 36–37, 82; Yugoslavism 51