Arabic in the City (Routledge Arabic Linguistics)

  • 77 774 1
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

ARABIC IN THE CITY

Filling a gap in the literature currently available on the topic, this edited collection is the first examination of the interplay between urbanization, language variation and language change in fifteen major Arab cities. The Arab world presents very different types and degrees of urbanization, from well established old capital-cities such as Cairo to new emerging capital-cities such as Amman or Nouakchott, these in turn embedded in different types of national construction. It is these urban settings which raise questions concerning the dynamics of homogenization/differentiation and the processes of standardization due to the coexistence of competing linguistic models. Topics investigated include: • • • • • • •

History of settlement The linguistic impact of migration The emergence of new urban vernaculars Dialect convergence and divergence Code-switching, youth language and new urban culture Arabic in the diaspora Arabic among non-Arab groups.

Containing a broad selection of case studies from across the Arab world and featuring contributions from leading urban sociolinguistics and dialectologists, this book presents a fresh approach to our understanding of the interaction between language, society and space. As such, the book will appeal to the linguist as well as to the social scientist in general. Catherine Miller is Senior Research Fellow at the French Council of Research (CNRS), IREMAM, Aix en Provence, France. Enam Al-Wer is Senior Teaching Fellow in Linguistics, University of Essex, UK. Dominique Caubet is Professor of Maghrebi Arabic, Head of the CREAM-LACNAD, National Institute of Languages and Oriental Civilizations (INALCO), France. Janet C. E. Watson is Professor of Arabic and Linguistics, University of Salford, UK. Contributors: Munira Al-Azraqi; Alassane Dia; Mohamed Embarki; Marie-Aimée Germanos; Atiqa Hachimi; Hanadi Ismail; Jonathan Owens; Christophe Pereira; Sherin Rizk; Catherine Taine-Cheikh; Aline Tauzin; Ángeles Vicente and Karima Ziamari.

ROUTLEDGE ARABIC LINGUISTICS SERIES General Editor: Clive Holes (University of Oxford) Editors: El-Sa’id Badawi (American University in Cairo) Adrian Gully (University of Exeter) Yasir Suleiman (University of Edinburgh) Keith Walters (University of Texas) The Routledge Arabic Linguistics Series publishes high quality, academically rigorous research on Arabic linguistics to two main readerships: non-Arabist general linguists with an interest in Arabic, and students and researchers already in the field of Arabic language and linguistics. Both synchronic and diachronic studies of Arabic are welcome which aid our understanding of the historical evolution and the present state of Arabic, whether dialectal or standard. Works written from a sociolinguistic (e.g. language variation), socio-historical (e.g. language history), sociological (e.g. language planning), or psycholinguistic (e.g. language acquisition) perspective are welcome, as are studies of Arabic stylistics, pragmatics, and discourse analysis. Descriptive dialectological works also fall within the scope of the Series, as do works which focus on the evolution of mediaeval Arabic linguistic thought. Proposals or scripts for the Series will be welcomed by the General Editor. PREVIOUSLY PUBLISHED BY CURZON 1. MEDIEVAL ISLAMIC PRAGMATICS Sunni legal theorists’ model of textual communication Mohamed Mohamed Yunis Ali 2. JEWISH AND MUSLIM DIALECTS OF MOROCCAN ARABIC Jeffrey Heath 3. LANGUAGE CONTACT AND LANGUAGE CONFLICT IN ARABIC Variations on a sociolinguistic theme Edited by Aleya Rouchdy PUBLISHED BY ROUTLEDGE 4. STRUCTURE AND FUNCTION OF THE ARABIC VERB Maher Bahloul 5. ARABIC IN THE CITY Issues in dialect contact and language variation Edited by Catherine Miller, Enam Al-Wer, Dominique Caubet and Janet C. E. Watson

ARABIC IN THE CITY Issues in dialect contact and language variation

Edited by Catherine Miller, Enam Al-Wer, Dominique Caubet and Janet C. E. Watson

First published 2007 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2007. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Routledge is an imprint of the Taylor & Francis Group, an informa business © 2007 Editorial selection and matter; Catherine Miller, Enam Al-Wer, Dominique Caubet and Janet Watson; individual chapters the contributors 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. 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 Arabic in the city : issues in dialect contact and language variation / edited by Catherine Miller . . . [et al.] p. cm. — (Routledge Arabic linguistics series) Includes bibliographical references and index. ISBN13: 978-0-415-77311-9 (hardback) ISBN10: 0-415-77311-3 (hardback) ISBN13: 978-0-203-93336-7 (ebook) ISBN10: 0-203-93336-2 (ebook) 1. Arabic language—Dialects. 2. Arabic language—Variation. 3. Urban dialects—Arab countries. 4. Arabic language—Social aspects. 5. Languages in contact—Arab countries. I. Miller, Catherine, 1955– PJ6709.A78 2008 492.7′7—dc22 2007027019 ISBN 0-203-93336-2 Master e-book ISBN

ISBN10: 0-415-77311-3 (hbk) ISBN10: 0-203-93336-2 (ebk) ISBN13: 978-0-415-77311-9 (hbk) ISBN13: 978-0-203-93336-7 (ebk)

CONTENTS

vii viii xi xiii xiv

List of maps List of tables and figures List of contributors Acknowledgements Notes on the transliteration 1

Arabic urban vernaculars: development and change

1

CATHERINE MILLER

PART I

Migration, urbanization and language change 2

The (r)urbanization of Mauritania: historical context and contemporary developments

33

35

CATHERINE TAINE-CHEIKH

3

The formation of the dialect of Amman: from chaos to order

55

ENAM AL-WER

4

Urbanization and dialect change: the Arabic dialect of Tripoli (Libya)

77

CHRISTOPHE PEREIRA

5

Becoming Casablancan: Fessis in Casablanca as a case study

97

ATIQA HACHIMI

6

Two cases of Moroccan Arabic in the diaspora ÁNGELES VICENTE

v

123

CONTENTS

PART II

Urban vernaculars: convergence and divergence 7

Greetings in Beirut: social distribution and attitudes towards different formulae

145

147

MARIE-AIMÉE GERMANOS

8

Linguistic leveling in SanÄani Arabic as reflected in a popular radio serial

166

JANET C. E. WATSON

9

The urban and suburban modes: patterns of linguistic variation and change in Damascus

188

HANADI ISMAIL

10 Segmental and prosodic aspects of Ksar el Kebir’s neo-urban variety

213

MOHAMED EMBARKI

11 The use of kaÌkaÌah/kaskasah and alternative means among educated urban Saudi speakers

230

MUNIRA AL-AZRAQI

PART III

Multilingualism, codeswitching and new urban cultures

247

12 Close encounters of a different kind: two types of insertion in Nigerian Arabic codeswitching

249

JONATHAN OWENS

13 Development and linguistic change in Moroccan Arabic-French codeswitching

275

KARIMA ZIAMARI

14 The language of Cairo’s young university students

291

SHERIN RIZK

15 Rap and rappers in Nouakchott (Mauritania)

309

ALINE TAUZIN

16 Uses and attitudes towards Hassaniyya among Nouakchott’s Negro-Mauritanian population

325

ALASSANE DIA

345

Index vi

LIST OF MAPS

1.1 2.1 3.1 4.1 5.1 6.1 7.1 8.1 8.2 9.1 9.2 11.1

Cities investigated Nouakchott (Mauritania) Jordan Tripoli (Libya) Casablanca (Morocco) Zaragoza and Ceuta Beirut Yemen San¡a Damascus Dialect map of Syria The map of the Kingdom of Saudi Arabia showing locations of the five cities of the study 16.1 Map of Mauritania showing main southern cities

vii

3 36 57 79 99 126 149 168 172 191 200 234 327

LIST OF TABLES, FIGURES AND SPECTOGRAMS

Tables 1.1 1.2 3.1 5.1 5.2 5.3 5.4 5.5 5.6 6.1 6.2 6.3 7.1 7.2 7.3 9.1 9.2 9.3 9.4 9.5 12.1 12.2 12.3 12.4 12.5 12.6

Number of urban inhabitants plus urbanization rate by Arab country from 1900 to 2005 Population growth of 1st, 2nd and 3rd cities of each Arab country from 1860 to 2005 Examples of the patterns found in the second generation Social profile of informants The linguistic variables Frequency of use of Casablancan [r] by Fessi women The variable (qal) by speakers Second person singular in Fessi and Casablancan dialects Frequency of use of gender distinction in 2nd person singular among Fessis Conjugation of kla-verb in prestigious Ceutí sociolect Conjugation of kla-verb in the conservative Ceutí sociolect Conjugation of kla-verb in Zaragoza Arabic Percentage of bonjour in East and West Beirut Percentage of mar}aba by age groups Percentage of hi by age groups Distribution of the speakers by age, gender, and life-mode Percentage use of (h) according to the preceding environment Percentage use of (h) according to age groups Percentage use of (h) according to neighborhood Distribution of (r) by phonetic environment Continuation in same language between speaker’s turns Lexical insertions Marking of definiteness on nouns: definite, indefinite, incorrect hana possessor hana possessor, SA versus monolingual only idAfa possessor viii

24 25 65 105 106 107 108 110 111 133 133 137 156 157 157 191 201 201 202 205 252 252 259 260 261 262

LIST OF TABLES, FIGURES AND SPECTOGRAMS

12.7 12.8 12.9

Adjectives Monolingual norm tracking Monolingual versus English insertions in CS texts, type/token ratios for definite/indefinite nouns 12.10 Correspondence between some SA and NA phonemes 12.11 Total SA words, per CS text 14.1 Use and attitudes towards YL among 40 students 16.1 Classification of the languages of Mauritania, French included, from the most spoken to the least spoken 16.2 Other national language spoken besides native language

262 263 264 266 267 303

329 331

Figures 2.1 4.1 5.1 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 10.1 10.2 10.3 11.1 11.2 11.3 11.4 11.5 11.6

Population of Nouakchott from 1962 to 2005 Lexical evolution of TA Casablanca’s population growth from 1897 to 2003 Percentage use of (h) by gender in Shaghoor Percentage use of (h) in Dummar by gender Use of (h) by gender and age groups in Shaghoor and Dummar Gender distribution of (r) in both communities Distribution of (r) by neighborhood Distribution of (r) by age groups Distribution of (r) by age groups in both communities Distribution of (r) by age groups (Y, M, O) and gender in each neighborhood (S, D) Evolution of the a1/a2 mean ratio according to the schooling level in both areas Word duration among 4 groups Evolution of the mean articulation rate by age among the 4 groups Growth of education sector in Saudi Arabia in number of schools Difference between regions in using /k/, colloquial form and other means Percentage of using /k/, colloquial form and other means by informants of Riyadh Percentage of using /k/, colloquial form and other means by informants of Dammam Percentage of using /k/, colloquial form and other means by informants of Buraidah Percentage of using /k/, colloquial form and other means by informants of Abha ix

44 89 101 203 203 203 206 206 206 207 207 221 223 225 232 236 236 237 237 238

LIST OF TABLES, FIGURES AND SPECTOGRAMS

11.7 Percentage of using /k/, colloquial form and other means by informants of Skaka 14.1 Standard/Non-standard in written and oral forms

238 292

Spectrograms 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9

guYn “cotton” firn “oven” bisbAs “chili” mubargaY “lumpy” za}Awug “spicy dip” dagCg “flour” zabAdC “yoghurt” γa∫∫A∫∫ Cn “cheats” rAgid “sleeping”

178 178 179 180 180 181 182 183 184

x

CONTRIBUTORS

Munira Al-Azraqi, Department of Foreign Languages, King Faisal University, Al-Ahsa, Saudi Arabia. Enam Al-Wer, Department of Language and Linguistics, University of Essex, Colchester, United Kingdom. Alassane Dia, French Department, University of the Manouba, Tunis. Mohamed Embarki, PRAXILING UMR 5267, CNRS, University of Montpellier III, Montpellier, France. Marie-Aimée Germanos, UFR Orient Monde Arabe, University Paris III, Paris, France. Atiqa Hachimi, Department of African and Asian Languages and Literatures, University of Florida, Gainesville, USA. Hanadi Ismail, Department of Language and Linguistics, University of Essex, Colchester, United Kingdom. Catherine Miller, Institut de Recherche sur le Monde Arabe et Musulman (IREMAM-CNRS), University of Aix en Provence, France. Jonathan Owens, University of Maryland CASL, USA and University of Bayreuth, Germany. Christophe Pereira, CREAM-LACNAD, Institut National des Langues et Civilisations Orientales (INALCO), Paris, France. Sherin Rizk, French Department, University of Helwan, Cairo, Egypt. Catherine Taine-Cheikh, CNRS-LACITO UMR 7177, Villejuif, France. Aline Tauzin, FRE 2788 CNRS/University of Paris 7, Paris, France. Ángeles Vicente, Instituto de Estudios Islámicos y del Oriente Próximo (IEIOP), University of Zaragoza, Zaragoza, Spain. xi

CONTRIBUTORS

Janet C. E. Watson, School of Languages, University of Salford, United Kingdom. Karima Ziamari, University of Fez, Morocco/CREAM-LACNAD, Paris, France.

xii

ACKNOWLEDGEMENTS

This book is the result of long-term contacts and extensive collaboration between various researchers through an informal network which aims to foster research on Arabic urban dialectology and sociolinguistics. Funding was provided by the European Scientific Foundation, the French National Council for Research (CNRS), IREMAM, (Institut de Recherche et d’Etudes sur le Monde Arabe et Musulman) and the University of Aix en Provence helped organize a Scientific Preliminary Workshop in October 2004. Arabic urban sociolinguistics is still a relatively new field and our aim in this book is therefore to collect and present different theoretical and methodological trends. Due to constraints of space and time, we could not accommodate all the authors we would have wished, and the chapters presented here have been selected after several readings and reviews. We are aware that much remains to be done, but we hope that this volume lays a foundation for future research. A number of scholars have provided help at various stages of this project. David Britain, Louis-Jean Calvet, Clive Holes, Pierre Larcher, Jerôme Lentin, Leila Messaoudi, Gunvor Mejdell, Jonathan Owens, Samia Sambar, Yasir Suleiman and Kees Versteegh participated actively in the discussions during the workshops and provided us with important scientific insights. In Aix en Provence, thanks are due to Françoise Gyllepsie and Majid Arrif, who were actively involved in the translation and formatting of the papers. We are also extremely grateful to Phil Brew from the University of Essex for translating some papers from French into English; to Peter Behnstedt, who redesigned one of his linquistic maps from Syria, which appears in Chapter 9, his cooperation and prompt response are much appreciated; and, to Florence Troin for drawing the maps.

xiii

NOTES ON THE TRANSLITERATION

The transliteration of the Arabic sounds is the following: emphatics: /x/ (= K), /\/ = (L), /y/ (= M), /z/ (= N), /t/, /l· /, /w/, /â/, /à/ pharyngeals: /]/ (= C) and /¡/ (= O) glottal: /™/ (= 7) velars: /x/ (= D) and /γ/ ( P) interdentals = /θ/ (A) , /o/ (F) , /2/ (N). The sibilant (J) is transcribed as /ʃ/ following the IPA system in most chapters but as /k/ following the Arabic dialectal tradition in a few other chapters. The glide ([) is either transcribed as /j/ following the IPA system in some chapters but as /y/ following the Arabic dialectal tradition in most chapters. Long vowels are phonologically transcribed as /a/, /b/, /c/, /d/, /e/ and phonetically as [a:], [e:], [i:], [o:], [u:], etc.

xiv

ARABIC URBAN VERNACULARS

1 ARABIC URBAN VERNACULARS Development and change Catherine Miller

1.1 Introduction Cities are “par essence” places of contact and heterogeneity; and since the 1960s have been the locus of research on language variation and change. Most of the recent comprehensive publications on dialect contact and language variation in the urban environment focus on the Western world, i.e. on countries where the process of urbanization of the nineteenth and twentieth centuries was closely linked to the process of industrialization (e.g. Auer et al. 2005; Chambers et al. 2002; Kerswill 2005). Urban sociolinguistics, and particularly variationist sociolinguistics, attempts to develop rules, models and typologies; it has turned out to be a rather challenging task due to the number and types of factors and data that need to be investigated (Owens 2005a). The link between social and linguistic processes is particularly complex. The same phenomenon (for example, migration and settlement in a given city) can produce very different linguistic outcomes depending on the historical and social settings. To what extent do the rules observed in industrialized and post-industrialized Western cities, and the socioeconomic categories developed in these countries, apply to other parts of the world? Like any other social science, urban sociolinguistics balances universalism and localism, generalization and particularism. In this respect, investigating non-Western urban settings might help to identify both universal trends and more specific local issues. The Arabic-speaking world covers a wide and heterogeneous geographical area and includes very different types of urban settings, national constructions, social organizations and language situations, in addition to the spread of the Arabic-speaking diaspora in many parts of the world. Historically, Arab countries had experienced very different rates of urbanization. However, urbanization has been one of the important socioeconomic changes of the second part of the twentieth century. Predominantly rural in the mid-twentieth century, the population of most Arab countries is now 1

CATHERINE MILLER

predominantly urban (see Appendix: Table 1.1). This urban expansion cooccurred with a high demographic growth, and arose in a particular political context: the creation or the consolidation of post-colonial states leading to internal as well as external competitions and conflicts in an increasingly globalized world. Little is known about the linguistic outcomes of this massive urbanization process, although hundreds of historical and urban studies have focused on the urbanization of the Middle East (Bonine et al. 1994). For a number of cities, we find relatively old dialect descriptions, which usually fail to account for variation and change. For others, we have more focused variationist studies, restricted to a small number of phonological variables. Although valuable data have been collected for over a century, this data is often not easily accessible.1 The absence of a synthesizing perspective does not facilitate cross-cultural comparison on the correlation between social changes and language changes. The desire to foster dialogue between researchers from different countries and scientific traditions was at the heart of a collective project that led to the organization of an international workshop on Arabic urban vernaculars in Aix en Provence in October 2004.2 From the very beginning, it was clear that the challenge of reaching a more analytic and synthetic perspective based on commonly recognized firm grounds had still a long way to go before its realization. By offering a panorama of Arabic linguistic urbanization, this book represents an initial step in this process.3 It gathers fifteen case studies on eighteen cities from ten countries (see Map 1.1). The selected cities are mainly capital cities, which have undergone different types and degrees of urbanization: old cities (Cairo and Damascus); emerging cities (Amman or Nouakchott); expanding cities (Casablanca, Riyadh, San¡a and Tripoli); cities that went through civil war (Beirut); a few provincial towns (Ksar el Kebir and Meknes in Morocco; Damman, Buraidah, Abha and Skaka in Saudi Arabia); and, cities in which Arabic speakers are a political or demographic minority (Ceuta, Maiduguri, Zaragoza). The book encompasses various methodological and theoretical approaches, some more linked to Arabic dialectology, some more linked to variationist sociolinguistics and some to anthropological linguistics. The chapters present cases of dialect contact and language variation, analyzed at various linguistic levels going from an overall perspective, to phonetics and acoustic analyses. Most chapters discuss the impact of internal migration on both individual speech and on the evolution of urban vernaculars (dialect convergence or divergence); a few chapters focus on other aspects of urbanization, more linked to the spread of education, modernization and globalization. Beyond the diversity of the data, which partly reflects the past and present human and cultural diversity of these cities, some strong trends emerge that will be highlighted in the present chapter. This chapter discusses some key aspects of Arabic urban linguistics.4 It analyzes the status attributed to urban vernaculars by traditional Arab gram2

Map 1.1 Cities investigated

ARABIC URBAN VERNACULARS

3

CATHERINE MILLER

marians and Western dialectologists vis à vis other dialectal categories (see Section 1.2). It then summarizes the main socio-economic characteristics of twentieth-century urbanization trends in the Middle East and the various linguistic impacts of this population renewal (see Section 1.3). It discusses the status of urban vernacular Arabic versus national vernacular and points to the problematic use of the concepts of standardization, prestige and norms in the Arabic setting (see Section 1.4). Finally it presents the issue of multilingualism and new urban cultures in the globalized cities (see Section 1.5). Because we subscribe to the view that language change needs to be explained on the basis of a multiplicity of factors (internal, external and extra-linguistic), particular attention will be given to extra-linguistic phenomena which play an important role in the social construct of the Middle Eastern cities. Urban dynamics cannot be isolated from their wider national or regional political context and from the ideological conflicts that arise in such contexts.

1.2 Classification of Arabic urban vernaculars: stereotypes and facts Before looking at the contemporary settings, it is important to recall how the Arabic linguistic tradition, in both the Arab and Western world, has conceived the linguistic categories, which have shaped, consciously or not, our perception of the language situation. The issue of dialect categorization is closely connected to the theorization of the origin(s) of the Arabic vernaculars, a topic which has been widely discussed over the past fifty years. Only brief reference will be made here to the rural/bedouin/urban distinction, which continues to be a key and controversial classification of Arabic linguistics. 1.2.1 Traditional dialect categorizations Arabic urban vernaculars are considered to have played a crucial role in the history of Arabic. Following the early Arab-Muslim conquest of the seventh-eighth centuries (ad), a number of garrison towns became the first Arabized centers outside the Arabian Peninsula (Donner 1981; Versteegh 1997). Because Arab speakers were in contact with local non-Arab population, the Arabic vernaculars that developed in these cities came progressively to be considered as more “corrupt” than the more “pure” bedouin vernaculars of the Arabic Peninsula. The distinction between the “conservative” Arab bedouin speech and the “corrupted” urban speech is epitomized in fourteenth century Ibn Khaldoun’s Muqaddima but can be traced back to some of the early Arab grammarians such as Ibn Jinni in the tenth century (Larcher 2006; Owens 2005b; Versteegh 1997). According to the tradition, the first Arab grammarians were called upon by the rulers to “protect” the 4

ARABIC URBAN VERNACULARS

pure Arabic language from foreign influence (Versteegh 1997: 3). They started to pinpoint the “faults” (la}n) of the urban speakers and are said to have relied on isolated bedouin speakers to fix the grammatical rules of Classical Arabic. This topos of the purity of the bedouin language and its close relationship to Classical Arabic has survived until now in both Arab societies and the meta-linguistic discourses. The typological division between sedentary (}a|arC) and bedouin (badawC) dialects, and within the sedentary, between urban (madanC) and rural (qarawC or fellA}C) dialects inherited from Ibn Khaldoun was taken over by the early European dialectologists and is still in use today (Palva 2006). The linguistic basis of the urban/rural/bedouin typology led to many controversies, i.e. few features distinguish all bedouin dialects from all sedentary dialects (Holes 1996; Ingham 1982). But the structural similarities recorded between dialects separated sometimes by huge geographical distances indicate that these dialectal classifications are not completely unfounded (Palva 2006; Rosenhouse 2006). An important analytical tool of Arabic dialectology has been the concept of koine/koineization, inherited from the Hellenic linguistic tradition. It was used to explain the origin and the nature of the early urban vernaculars which developed in the garrison towns.5 The concepts of koine and koineization has also been applied to many urban and non-urban Arabic dialects in transitional zones or in areas that had experienced successive waves of settlement (Palva 1982). By using the term koine, linguists accredited the postulate that various Arabic vernaculars share a systemic unity between themselves and with Classical Arabic. They tended to minimize the influence of non-Arabic languages in the historical and contemporary development of Arabic vernaculars. But because the term koine has been dominantly used in reference to Arabic urban vernaculars, it reinforces the idea that urban vernaculars are more mixed than other, particularly bedouin, vernaculars, even if some postulates have since been criticized by a number of linguists (i.e. the possible unity of pre-Islamic Arabic vernaculars, the supposed conservatism of bedouin dialects, the genealogical link between Classical Arabic and modern Arabic vernaculars, etc.).6 In contemporary studies, the term koine refers to a shared variety (see below for Morocco). This traditional dialect classification/representation had important repercussions for contemporary settings, both linguistically and symbolically. 1.2.2 Sociolinguistic implications of the urban/bedouin dichotomy Linguistically, the categories of bedouin, rural and urban dialects are still used by most linguists on the basis of the presence/absence of a set of features rather than by reference to a geographical region or lifestyle. Categorizing a dialect X as a bedouin/bedouinized dialect does not mean that the speakers pursue a nomadic bedouin way of life but that they display in 5

CATHERINE MILLER

their speech a number of features associated with bedouin dialects. This categorization implies that some varieties spoken in urban environments will be nevertheless categorized as “bedouin–bedouinized”, “rural” or, more often, “mixed”. The implication is that “origin” (i.e. genealogy) is considered a more important criterion of categorization than “geographic location”. At this stage, it is interesting to point out that these linguistic representations echo some of the local identity discourses. Reference to “origin”, “lineage” and “family” is a corner-stone of self-affiliation discourses, particularly but not exclusively, among groups claiming an Arab tribal origin. Many urban dwellers categorize themselves by referring to a tribal– regional–family affiliation, rather than a contemporary place of residence.7 The continuing use of the terms rural or bedouin varieties or bedouinized koine in urban contexts can be diversely interpreted. One interpretation is that these terms refer to fixed categories and participate in the orientalist and essentialist vision of Arab societies. Another interpretation is that the use of these terms shows that there is no discontinuity between the city and the bedouin/rural hinterlands, and that urbanization does not necessarily radically transform patterns of affiliation and identification, or language use. Taking into account the theory of ethnic boundaries (Barth 1969), it is evident that the term “bedouin” may refer to very different and changing entities, the important thing being its social meanings and uses in the given society. In this respect, these categories might act as badges of identity and might be just as relevant as more “modern” categories such as social or professional classes. The traditional dialect categorization has proved useful in identifying the various historical linguistic layers found within a city as well as in understanding the origin of communal/religious variants/varieties found in many cities. In North Africa, in particular, historical dialectology has distinguished non-Hilali from Hilali dialects and within non-Hilali, Andalusi from Jbala, etc. (Aguadé et al. 1998). In Iraq, a distinction has been drawn between sedentary q¢ltu dialects and bedouin g¢l¢t dialects (Jastrow 2006). An important correlate is that urban variants/varieties that appear at first sight to have religious or communal or sectarian affiliations (i.e. Jewish or Christian versus Muslim, or Sunni versus Shi¡i) were found to reflect successive patterns of settlements and a division between former sedentary and former bedouin groups (Blanc 1964; Holes 1987, 1995b). To sum up, the population of the old urban centers spoke a sedentary vernacular, irrespective of its religious affiliation (qeltu dialect in Mesopotamia, non-Hilali and often Andalusi dialect in North Africa). With the progressive settlement of former bedouin groups, a process of koineization occurred which led to the emergence of mixed urbanized bedouinized vernaculars spoken mainly by Muslim groups (particularly males), while the old city vernaculars were kept by non-Muslim communities and women.

6

ARABIC URBAN VERNACULARS

Two important sociolinguistic implications can be deduced from these historical processes. First, it is not possible to generalize the linear developmental model proposed by linguists such as Cadora (1992), which postulates a linear evolution from bedouin dialects to rural dialects to urban dialects, given that many Arabic urban vernaculars went through a later bedouinization process, which continued throughout the twentieth century (Abu Haidar 2006). Second, the maintenance over centuries of some “old urban features” attested in pre-Hilali and q¢ltu urban dialects, in spite of huge population movements and koineization processes, indicates that dialect contact induced by migration does not automatically lead to a general process of leveling and koineization and the emergence of a single vernacular that eradicates all previous varieties. Social, communal or spatial segregation can foster the preservation of different varieties (Siegel 1993). This means that the three-generation pattern endorsed by many sociolinguists (such as Calvet (1994) for the French school, Trudgill (1986) and Kerswill (2005) for the British school (see also Chapter 3, this volume) must be investigated very carefully. 1.2.3 The ambivalent image of the city Symbolically, the perception of Arabic urban vernaculars as “mixed” or “corrupted” forms of speech might affect their contemporary status and seems to fit with the ambivalent perception of the city that prevails in many Arab countries. The ambivalent symbolic status of the city is an universal topos, that has come and gone since antiquity according to the historical– political context and the dominant ideological discourse of the time. At some periods, cities are seen as essentially places of cosmopolitanism and corruption as opposed to the rural simplicity and honesty. At other periods, cities are epitomized as places of civilization, refinement, dynamism and modernity as opposed to the backward rural areas. In Arabic, the Arabic root }|r has given ta}a||ur “urbanization” and }a|Ara “civilization”, suggesting that both processes were seen as constitutive to each other. But the modern urban way of life, particularly when associated with Westernization, is also the focus of social and religious criticism. It must be remembered here, that, starting from the fourteenth century, the major Middle Eastern cities developed in a context of political domination, i.e. the urban ruling elite of most countries was of foreign origin (Mamluk, Ottoman, Circassian, Moghol, Spanish, Portuguese, etc.) and the cities attracted merchants, mercenaries, workers, etc. from various places (Dakhlia 2004; Raymond 1993). From the mid-nineteenth century up to the first decades of the twentieth century, under colonial or protectorates rules, many cities had very important non-Muslim European communities and in some cities like Algiers the local Muslim population was a minority (Boucherit 2002).

7

CATHERINE MILLER

When independence took place, urbanization led to an important population renewal in a context of Arab nationalism where the notions of Arab identity, Arab authenticity, purity of origin (§asAla) were some of the cornerstones of a new political discourse, supported by an educational policy in favor of Arabization. The arrival of many provincial migrants and the departure of members of foreign as well as Jewish communities fostered the emergence of new urban practices and cultures. Depending on the city and the national and regional context, the outcome of this population renewal led to different dynamics.

1.3 Urbanization and migration 1.3.1 The growth of the capital cities One of the major characteristics of the urbanization trend in the Arab world in the second part of the twentieth century is the decisive role of internal migration on urban growth, with the notable exception of the Gulf countries, Iraq and Libya, where the oil industries attracted an important regional and international migration. Another exception is Amman (see Chapter 3), which since its creation has accepted many refugees from neighboring countries (Palestinians and more recently Iraqis). This internal migration took place at differing speeds and rates depending on the country. In countries such as Egypt or Syria (see Appendix: Table 1.2) with a pre-1900 high rate of urbanization, internal migration was particularly important up to the 1970s and started to slow down in the 1980s, while it increased considerably since the 1970s in countries with previous low levels of urbanization such as Mauritania, Libya, Jordan and Yemen. Internal migration appeared at first to be directed towards the capital cities or the economically dominant centers. Since the mid 1970s, urbanization has spread to many regional and secondary urban centers as well as former villages (Denis 2007; GEOPOLIS8; Kharoufi 1995). However, every Arab country has one major city which stands far above all the others in terms of demographic growth and economic wealth, and acts as the dominant national pole. This is usually the capital city, except in the case of Morocco where Casablanca surpasses Rabat and acts as the economic capital of the country. This domination has sometimes occurred to the detriment of former important regional cities (cf. Cairo versus Alexandria; Damascus versus Aleppo; Casablanca versus Rabat, Salé, Fez; San¡a versus Aden; Amman versus Irbid, etc., see Appendix: Table 1.2). At the spatial level, the growth in urbanization has witnessed the expansion of urban suburbs, going from upper- and middle-class districts to unplanned or “informal” settlements or “bidonvilles”. While the first type of suburbs has not attracted too much attention, the unplanned suburbs, with sometimes more than 50 percent of the total city population, are often stigma-

8

ARABIC URBAN VERNACULARS

tized in dominant public discourses as places of urban poverty, instability and unstructured rurality that “threaten” the former urban balance and culture.9 Although this negative perception has been qualified by many social studies, it remains a strong urban meta-narrative, which questions the migrants’ ability to integrate within the city. Often expressed in terms of traditional opposition between urbanity and rurality, or even “rurbanity”, it sometimes tends to focus on specific ethno-regional groups such as the Shi¡i living in Eastern Beirut (Harb el Kak 2006), or the Upper Egyptians living in the suburbs of greater Cairo (Miller 2005). At a sociolinguistic level, this growth in urbanization raises the following questions: 1.

2.

3.

What is the influence of internal migration upon the development of urban vernaculars and what kind of dialect contact, dialect accommodation or dialect shift has taken place? To what extent does urbanization lead to the emergence of established (or focused) urban vernaculars which could be considered as urban standards? What are the main factors in the variations and changes that have been recorded? What is the regional–national diffusion of the capital’s vernacular, if it exists? How does it interact with other urban regional vernaculars? 1.3.2 Impact of migration upon the development of Arabic urban vernaculars

The contemporary linguistic impact of migration has been unevenly studied. In the period 1950–70, structuralist and functionalist linguistics dominated Arabic dialect descriptions, giving little place to the study of variation. From the 1960s, more attention has been given to contact between Modern Standard Arabic (MSA) and the dialects than to contact between the various regional/national vernaculars (see Section 1.4). In the early 1980s, a number of variationist studies focused on dialect contact in the urban context, particularly in Jordan–Palestine (Abdel Jawad 1986) and Bahrain (Holes 1987). The development of dialectal studies in the 1980s and the publication of a number of dialect atlases contributed indirectly to the emergence of Arabic urban sociolinguistics, since dialect contact cannot be properly studied without sufficient knowledge of the linguistic features in contact! Different models and approaches have been developed in the Maghreb and the Mashreq. Generally speaking, francophone studies on North African cities have favored a historical–dialectal approach and have focused on the formation of urban koines. The best-studied examples are found in Morocco (see below). Most studies on Middle Eastern cities have followed a variationist approach and have investigated variation based on age, sex, social class and education with a recent shift to ethno-methodology (focus on context of 9

CATHERINE MILLER

interaction and variation seen as a social practice). Finally, political factors (Arab pan-nationalism) have impeded the development of both dialectal and sociolinguistic studies. In most Arab countries, studies on Arabic vernaculars are not yet considered legitimate topics of research and are not supported by the local institutions. Civil wars and political instability have sometimes reinforced this trend. It is therefore not surprising that data are seriously lacking for countries such as Algeria or Lebanon. 1.3.2.1 North Africa In North Africa, urban dialectology has drawn a distinction between cities whose vernaculars are considered to be a bedouin-based koine (Oran, Marrakech, Casablanca, Constantine, etc.) and those characterized as former pre-Hilali or Andalusi dialects (Algiers, Fez, Rabat, Salé, Tunis, Tlemcen, Tangiers, Tetouan, Tripoli, etc.). For the former group, we have very little information concerning the impact of recent migration. The latter groups has attracted most attention, particularly in Morocco, where linguists have developed a further categorization of the urban dialects into vieux parlers citadins (old city vernaculars spoken by the original urban dwellers) and nouveaux parlers urbains (neo-urban vernaculars spoken by the new urban dwellers) for cities such as Rabat, Salé or Fez (Messaoudi 2003). The neourban vernaculars described as koines influenced by the “rural-bedouin features” brought by the immigrants, are said to have spread at the expense of the old city vernaculars. The neo-urban koine is becoming the public inter-group language while the old-city vernacular is increasingly restricted to close family relationships, is mainly spoken by old women and is in a process of attrition following the departure of some old urban communities, particularly the Jewish (Aguadé et al. 1998; Cohen 1973, 1981; Dendane 2002; El-Himer 2001; Jabeur 1996; Messaoudi 2001; Trabelsi 1988). The decline of the old urban vernaculars corresponds to the decline of the aristocratic culture and lifestyle symbolized by the Medina, (often abandoned by rich families who moved to new high-class suburbs) and coincides with the decline of an old urban elite (often from Andalusi origin) and the emergence of a new lifestyle and a new urban middle class of provincial origin (cf. the b¢ldi/barrAni dichotomy of Tunis). In all the big urban centers (Casablanca, Algiers, Tunis, Oran, Constantine, Rabat, Salé, Nouakchott, etc.), the expanding urban koines exhibit various degrees of mixing with the surrounding bedouin/rural dialects. This is certainly the reason why the citadin versus urbain paradigm became so prevalent in the North African urban social sciences (Naciri and Raymond 1997). Each city, however, represents a specific case; below are some examples developed in this book. Nouakchott in Mauritania represents the atypical case of a “new city without a new vernacular” and Taine-Cheikh’s chapter questions the passage from bedouinity to urbanity. Nouakchott was built ex-nihilo in 1957 10

ARABIC URBAN VERNACULARS

and its development coincided with the building of the nation-state. For centuries, the Moorish society was characterized by its bedouin and nomadic way of life. The pre-modern urban centers were mainly caravan cities that, in contrast to the rest of North Africa, did not develop any specific urban dialects. In spite of, or maybe as a result of, the strict social hierarchy of Moorish society, the Hassaniyya bedouin vernacular remained remarkably homogenous and knew very little dialect divergence. Therefore urbanization did not lead to a process of koineization or leveling but rather to increasing contact with MSA or with French, and to a certain amount of lexical renewal and borrowing. Language mixing, which was perceived negatively in the conservative Moorish society, is now spreading either through the use of a classicized colloquial variety or increasing lexical borrowings with French/English and African languages (see Chapter 16). Tripoli (Libya) is another capital that developed in a dominant bedouin environment. Pereira, in Chapter 4, carefully describes the linguistic features of the present-day urban koine. He recalls the long history of the city, which saw numerous population changes and points to the tremendous urbanization that followed the oil revolution in the 1960s. The oldest descriptions of Tripoli Arabic show that at the end of the nineteenth century this vernacular was already a koine, mixing bedouin and pre-Hilali features at all linguistic levels. A pre-Hilali dialect was spoken by the important Jewish community of Tripoli at the end of the nineteenth century. With the departure of this community, the old city dialect has totally disappeared. Compared to these older dialect descriptions, Pereira’s data indicate that certain bedouin features are reinforced (verbal gender distinction, dual nominal forms, synthetic genitive constructions, lexis). In the case of Tripoli, the absence of a strong local Tripolitan elite and the regime’s discourse emphasizing the values of the bedouin ethics accompanied the development of this bedouinized urban koine and the total disappearance of the former urban culture. It seems, however, that very recently a new interest toward the old Tripoli culture has been emerging. In Chapter 5, Hachimi explores contact between an old urban variety (Fessi) and a new urban variety in Casablanca (Morocco). Casablanca developed in the early twentieth century under French colonial rule and attracted a large rural population. Its vernacular is described as a koine that includes many rural/bedouin features taken from the neighboring Chaouia group. Today, Casablanca Arabic is considered to represent the national Moroccan koine, and has attained a certain degree of focusing and stabilization. It has as yet been little studied and we do not know the degree of homogeneization/ differenciation. Hachimi studies the process of accommodation among women of Fessi origin (from the city of Fez) living in Casablanca. Following Eckert’s theory of variation as social practices and not as structures, Hachimi shows that the degree of accommodation varies according to linguistic features, to context of interaction and to the personal attitudes of the 11

CATHERINE MILLER

speakers. Degree of leveling/accommodation is not directly linked to time of migration. An important insight is the symbolic value attributed to each specific feature. A number of Fessi women want to sound “normal”, “modern”, “harsh” and “popular” by using some Casablanca features (and dropping Fessi features viewed as too particular and aristocratic), but refrain from using others considered as too “rural”. Hachimi’s focus on the social meaning of variation highlights the fluidity of individual practices and self-identification. Concerning the old versus neo-urban vernaculars in North Africa, a number of issues deserve more investigation. One is the possible convergence between the different neo-urban koines at the national level. This question remains open for Algeria, due to lack of comparative studies between Oran and Algiers. In the case of Morocco, and due to the attested prestige of Casablanca Arabic, it is not yet clear if what authors such as Messaoudi call the neo-urban vernacular of Rabat, Salé or Fez are local urban koines or if they tend to reproduce the Casablanca koine. Messaoudi (2001, 2003) considers the Rabat neo-urban vernacular to be characterized by the dominance of rural/bedouin features coming from the surrounding tribes, particularly the Zaer, while in Casablanca, the rural/bedouin features are mostly associated with the neighboring Chaouia group. The issue of dialect convergence in the diaspora context is discussed by Vicente (Chapter 6), who is working on the Moroccan communities of two Spanish cities, Ceuta and Zaragoza. Previous studies on Moroccan diaspora in Europe (France, the Netherlands) indicate that Moroccan migrants continue to speak specific local varieties rather than the Moroccan koine. It is well known that migrant groups are often linguistically conservative compared to speakers in their region of origin. Vicente shows that this situation does not apply in Zaragoza, due to the fact that the Moroccan migration is a very recent phenomenon. Migrants (mainly young males) come from different areas and speak the Moroccan koine between themselves. As a Spanish enclave within North Africa, Ceuta constitutes a special case. Due to its proximity to Morocco, the speech of the Arab minority is more influenced by the regional linguistic changes. As in other Moroccan cities, Vicente detects the disappearance of a number of pre-Hilali features (preserved only by old women) and their replacement by koine-like features. However, she notes that some pre-Hilali features associated with the local prestigious Tetouan dialect are maintained. The Moroccan situation appears therefore rather complex with a number of regional urban koines, whose interconnection awaits further investigation. Another question concerns the possible territorialization of the urban variants. Some authors have attempted to map the territorial distribution of the old versus neo-urban varieties within the city and believe there to be a close correlation between types of urban varieties and types of urban districts (cf. El-Himer 2001 for Salé).10 However, this mapping fails to take into 12

ARABIC URBAN VERNACULARS

account social and stylistic variation and has not yet been investigated in a quantitative approach. The territorialization might rather be conceived in a more symbolic perspective, i.e. following what sociologists have labeled “imagined territories” (Tarrius 2000), i.e. the symbolic association between speech and place. Embarki, in Chapter 10, adopts the concept of territorialization and investigates the realization of some phonetic parameters (duration of vowels, syllables and words) among two groups of young Moroccan students born and bred in the town of Ksar el Kebir, in northwest Morocco. One group lives in the old center and the other in the new suburb. Relying on the quantitative analysis of phonetic data, Embarki concludes that there is indeed a difference between the speakers of the new and the old districts. While the girls of the old districts tend to maintain the phonetic parameters of the old city vernacular, the boys of the old districts and the boys and girls of the new districts tend to use the phonetic parameters of the neo-urban vernacular. Embarki’s study seems to indicate that in Ksar el Kebir, as in other Moroccan cities, the neo-urban vernacular is expanding at the expense of the old one, with change spreading from the suburb to the urban center. It might be noted here that realization of duration in the suburbs is closer to that of MSA than to that of the old city vernacular. Here again, more data on Casablanca Arabic are needed in order to compare the various urban koines and the possible influence of MSA. Another issue is the genderization of the distinction between old versus the neo-urban vernacular archetypes. Women are said to be the main users of old city vernaculars (both retainers or acquirers), while men are said to favor new urban features. Therefore, old city vernaculars are typically perceived as sophisticated and more feminine (Boucherit and Lentin 1989; Cohen 1973; Miller 2004), while the new urban koine is perceived as more virile and tough. This “toughness” is often associated with the ruralization/ bedouinization of the city. This genderization is not specific to North Africa and can be found also in the Mashreq, where urban vernaculars tend to be considered more feminine than rural/bedouin vernaculars (Abdel Jawad 1986; Sawaie 1994). This gender distinction typically concerns a limited number of phonological features (such as reflexes of (q), reflexes of (r), +/− emphatization and velarization, backing/fronting vowels) as well as a few morpho-lexical features (use of diminutives) and lexicon. Genderization of speech, however, appears to be evolving. Hachimi indicates that a relative degree of “toughness” is considered a positive sign of modernization by some FessiCasablanca young women, a finding that finds some echoes in studies on code-switching, youth language and new urban music (see Part III). To conclude this brief overview of dialect contact induced by internal migration in North African cities, it appears that the long historical process of ruralization–bedouinization of the former old North African city dialects have accelerated during the second part of the twentieth century, leading to the weakening of the previous Andalusian aristocratic medina culture. 13

CATHERINE MILLER

However, the koineization/homogeneization process has not yet been completed at either city or country levels. Dialectal variety that might lose visibility in the daily life can be maintained through artistic practices, patrimonialization, symbolic memories, etc. As Hachimi shows, dialect variation carries strong but fluid social meanings, and we need to investigate to what extent the neo-urban varieties symbolize a new urban life style. Urbanization in North Africa has other linguistic aspects. One is the increasing contact with other languages (Berber, French, English, Spanish) and the development of new language practices (see Part III). 1.3.2.2 Mashreq In the Mashreq, the linguistic impact of migration varies widely due to the very different types of urbanization that coexist. The best known historical cases of dialect change due to migration have been investigated in areas where Bedouin groups became dominant: Lower Iraq (Baghdad), following the settlements of North Arabian tribal groups between the fourteenth to eighteenth centuries and their political dominance in the twentieth century (Blanc 1964); Bahrain in the twentieth century, following the settlement and political domination in the eighteenth century of the Sunni tribal groups over the local urban Ba]arna Shi¡i group (Holes 1987). In Baghdad, the settlement of numerous Shi¡i southern migrants in Baghdad’s suburb is said to have reinforced the bedouinization of Baghdadi Arabic in the second part of the twentieth century (Abu Haidar 2006). In Bahrain, two varieties coexist, together with a public koine, close to the Sunni bedouin speech (Holes 2006a). Apart from Bahrain and Iraq, Jordan–Palestine is one of the few regions where dialect contact induced by migration/urbanization has been dealt with in some detail (Sawaie 2006). The area had witnessed considerable population movements. Urbanization brought urban (Palestinian) and rural/ bedouin (both Palestinian and Jordanian) vernaculars into contact, leading to a complex pattern of variation related to gender, ethnicity and religion as well as contextual uses. Among the different trends that have been recorded, we note that women tend to dominantly keep /adopt urban (Palestinian) variants, while men are said either to shift to some MSA phonological variants (Amara 2005) or maintain (particularly in the case of Jordanian speakers) their phonological rural/bedouin variants (Abdel Jawad 1986; Sawaie 1994). Christians (both men and women) are also said to use more urban variants (Amara 2005 for Bethlehem) and rural phonological variants appear less resistant than bedouin variants. Al-Wer investigates the recent linguistic history of Amman, the Jordanian capital created in 1923. She recalls the successive waves of settlement which brought first an urban elite from neighboring Palestinian–Syrian cities and then a far larger rural Palestinian population in addition to the local Jordanian population. She shows how progressive “Jordanization” of public 14

ARABIC URBAN VERNACULARS

political life led to a crucial redefinition of the social meanings of localized linguistic features. While urban Palestinian features tended to dominate in the first decades, the rural/bedouin Jordanian features became progressively associated with political influence and a Jordanian identity. Therefore urban Palestinian men also tend now to use some Jordanian dialect features in public. Her study examines the realization of a number of phonological and morpho-phonological features by a group of speakers (from Palestinian and Jordanian origin) living in western Amman and describes the gradual changes that occurred over three generations. She indicates that in the third generation, the mixture and variability recorded in the second generation is considerably reduced and that focusing (i.e. stabilization and regulation) is taking place. Young Ammanis of both origin share numerous new fudged forms with a clear gender distinction with regards to the realization of (q). However, by correlating linguistic variation with social activities rather than preconstructed groups, Al-Wer highlights the fluidity of gender-related variation. The linguistic focusing is corollary to a sense of identification with the locality (Amman). Therefore, a kind of standardized koine Ammani vernacular is expected to stabilize and to represent the local identity, in spite of the continuous influx of foreign migrants. Al-Wer suggests that the formation of the Ammani koine is closely connected to the formation of a Jordanian national entity, which partly contrasts with and partly interacts with the Palestinian identity. The situation of Beirut, an older urban center, appears quite different to that of Amman (Germanos, Chapter 7). The Arabic variety(ies) of Beirut had been little studied (Naim 2006), apart from phonological studies which have focused on gender, religion (Christian/Muslim) and territorial distribution (Eastern/Western Beirut). Between 1975 and 1990, the civil war changed the demographic balance of the city and divided it along clear religiouscommunal lines. Even since the official reunification of the city at the end of the civil war, communal division remains extremely clear-cut. The southern suburb of Beirut (known as Da]yeh) hosts a large and growing number of southern Shi¡i migrants. The presence of a public neutral koine is not clearly attested, although authors such as Srage consider that the young Beiruti generation uses a “constituted urban dialect” very similar to the one used by the inhabitants of Achrafieh, a Beirut Christian area. Germanos describes the uses and attitudes of Beiruti speakers towards greeting formulae and indicates that use varies according to age, sex, place and religious affiliation. While some formulae are considered “neutral” by the speakers, others are closely associated with a specific group. Due to the sectorization of the city, the social distribution is reflected in a territorial division. Some districts, such as Haret Hreik in the southern suburbs, form distinct linguistic territories as far as the use of greetings is concerned. Greetings are a type of linguistic formulae that have important religious significance in the Muslim world (Elzeini 2006); in most Arab countries, 15

CATHERINE MILLER

Christians and Muslims do not use the same greetings. Therefore, greetings alone cannot be used to prove the absence/presence of a common koine shared by all speakers and Germanos concludes with caution that “the hypothesis of the existence of a koine needs to be explored, with further identification of its linguistic features, if it exists”. The present sociolinguistic situation of Beirut and the absence of a well established urban koine might be representative of the political and communal tensions that prevail in this small but very heterogeneous country. San¡a (Yemen) is another capital city of a previously divided country. San¡a has witnessed tremendous population growth over the last few decades, attracting migrants from all provinces. According to Watson (see Chapter 8), this urbanization has led to a limited degree of linguistic leveling, a fact agreed upon by other authors (Freeman 2006). Until 1962, Yemen was a feudal country with very limited centrality; since 1962, the country has witnessed numerous disputes and local wars. Tribal/local affiliation is still very strong and urbanization has not diminished the sense of belonging to an original locality. Male migrants, in particular, are said to keep their own vernaculars and speak a kind of koine in public, but a koine that has not leveled the various vernaculars. Watson analyzes the language of a popular radio series, which arouses contradictory statements among its listeners. Some people consider it to reflect “pure old San¡ani Arabic” while others consider it to be either Yemeni, but not San¡ani, or Classical Arabic. She shows that the actors born in San¡a from non-San¡ani families have a good command of San¡ani Arabic at the morphological and lexical level but that they lack some of the low-level phonological features such as glottalization in pause. The future will see whether glottalization in San¡ani progresses or, whether, as a result of dialect contact, glottalization weakens. Due to its specific history and rich dialect diversity, Yemen deserves far more sociolinguistic studies in order to evaluate the relationship between the different cities and the type of leveling recorded in them (cf. Vanhove 2002). In the old established capital cities of Damascus and Cairo, the stabilization of the urban koine is considered to have occurred at an earlier period, probably the second part of the nineteenth century (Lentin 2006; Woidich 1994). Both countries have experienced a fairly high level of urbanization during the nineteenth century. Since that time, successive waves of migration have not seriously affected the development of the capital dialect and both Damascus Arabic and Cairo Arabic act as national standards. However, although both vernaculars are among the best studied dialects of the Arab world, few studies have focused on dialect contact and the process of accommodation. Miller (2005) studied processes of accommodation among the first generation of Upper Egyptian migrants in Cairo and showed the ambivalence of language uses and attitudes. Lentin (1981) indicated that regional, communal, neighborhood variants are very frequent in Damascus, but also very fluid. Other sociolinguistic studies on Cairo and Damascus 16

ARABIC URBAN VERNACULARS

have highlighted variation relating to education, gender, social class and neighborhood rather than to origin (Daher 1999, Haeri 1996, Royal 1985). Following this trend, Ismail, in Chapter 9, investigates variation and change in two Damascus neighborhoods, Shagoor an inner-city quarter and Dummar, a high-class suburb and asks whether different neighborhoods and lifestyles induce different patterns of change. Ismail compares the realization of two phonological variables in the two neighborhoods. The presence/absence of (h) in suffix position is a long-standing variable of Syrian dialectology. Examining the sources, Ismail postulates that the h-zero form appeared first in the coastal cities before spreading to inland dialects such as Damascus and Aleppo, while the nearby rural areas remained h-full (a process that fits with the “urban hierarchy model” of linguistic diffusion). Her data confirm that the h-zero form is the favored form of both districts. The variable (r) appears to be a change in progress that originated in the suburbs and spread to the inner city and is particularly used by members of the younger generation. Ismail’s data strongly fit with the findings of variationist sociolinguistics that old variants (stable variation) are no longer territorialized and cannot be associated with an age group, while new variants (change in progress) are more closely associated with specific urban territories and age groups (Eckert 2000: 136). The Arabian Peninsula and the Gulf witnessed considerable urbanization following the oil boom of the 1970s. In the Gulf States, this urbanization has reinforced long-standing contact with non-Arabic languages such as Hindi/Urdu, English, as well as with various Arabic dialects spoken by the native population or by the Arab expatriates (Holes 2006b). This has led to the formation of a Gulf Arabic Pidgin used between the Gulf speakers and the Asian migrants and to various degrees of leveling between the local dialects, even if the bedouin/sedentary distinction is still very noticeable. AlAzraqi, in Chapter 11, investigates the leveling of one phonological feature in five Saudi cities. The selected feature is kakkakah/kaskasah, i.e. the realization of 2nd feminine singular pronoun suffix as either [∫] or [s], a well-known feature also attested in a number of other dialects of the Arabian Peninsula. Al-Azraqi compares the realization of this variable among educated speakers of both sexes and various ages. Her study shows that this feature seems to recede, at least in formal context, and particularly among male speakers, and is replaced by a koine form /k/.Women appear to avoid the use of the 2nd feminine pronoun suffix and develop alternative syntactic means. The use of the koine feature varies from one city to another: greatest use of /k/ is recorded in Riyadh, the lowest in Buraidah. Al-Azraqi analyzes the spread of the koine form as a result of both pan-dialectal influence and MSA influence linked to urbanization and education. The results of Al-Azraqi can be contrasted with other studies such as Alessa (2006) on Nejdi speakers in Jeddah, which shows a significantly higher rate of occurrence of affricated /k/ in the suffix form. 17

CATHERINE MILLER

To conclude this preliminary panorama on migration-induced dialect contact in urban environments, we can state that although the sociolinguistic situation of each city does not entirely mirror the state of the social and political relationship between the different urban dwellers, it does provide some cues about it, particularly concerning the rise of new social groups within the city and the country. In this respect, the processes of reallocation, (a variant previously associated with a specific dialect becomes a social variant in the urban context) require additional comparative studies. The current focus on contextual and stylistic variation will certainly help provide a better understanding of the process of social interaction and power negotiation between the various groups. For the time being, it appears that the use of a common public variety which reflects adhesion to a common urban identity is far from being attested in all the cities surveyed. An important question is the potential role of these urban vernaculars as regional/national standards and their relationship with MSA.

1.4 Urbanization and standardization A long-standing issue has been the possible role of urban vernaculars (and particularly that of the capital city) as the supra-local standard at the regional or national levels, and the status of urban vernaculars vis à vis MSA. In the 1950–60s, following the independence and the rise of pan-Arab nationalism, many observers and national actors believed that the local vernaculars, both rural and urban, would converge towards MSA thanks to the mass media and mass education. Most studies were thus dedicated either to descriptions of intermediate varieties emerging from this convergence (arabe median in French, Educated Spoken Arabic in English) or to the variationist analysis of age, gender and social variation induced by the MSAdialect contact, with the postulate that MSA variants represented the “prestigious standard features”.11 A number of voices (Abdel Jawad 1987; Al-Wer 1997; Daher 1999; Ferguson 1997; Gibson 2002; Haeri 1996; Ibrahim 1986; Palva 1982) rejected this unilateral vision of change and discussed the ambivalence of the concept of “prestige” in Arabic. They pointed to the fact that MSA was restricted to certain domains of use (the official public sphere) and was more a male than a female phenomenon, due to division of work and presence in the public sphere. Moreover, convergence towards MSA was mainly recorded at the lexical level (increasing use of MSA words in certain semantic domains), while in all other linguistic levels, variation between MSA and dialect features would be better analyzed as stylistic variation or instances of code-switching rather than an indication of change in progress. Other studies indicated that a number of changes in progress in the various vernaculars had little to do with MSA, including when these changes were initiated among educated speakers, particularly women (AlWer 2002; Haeri 1996). The debate is far from closed. Many authors con18

ARABIC URBAN VERNACULARS

sider that de facto MSA contributes in the processes of koineization and standardization of urban vernaculars, while others stress the opposite and show that current changes in progress have little to do with MSA. Among the arguments of authors who have defended the vitality/prestige of urban vernaculars is the idea that, in a number of countries, the urban vernacular of the capital city is de facto acting as the national standard and has more influence on the other local vernaculars than MSA. This representation is in accordance with the Prague School’s functionalist conception of standard, which considers the presence of a standard language a major correlate of an essentially urban culture. However, it might be noted that the terms urban and national standards are often used in the Arab context without much elaboration concerning the actors and modalities of the process of standardization. Standardization involves a process of more or less conscious, and planned regulation of the language that includes codification of its form in standardized grammars and dictionaries, functional extension, expansion of the range of social uses and acceptance of the norms both officially and on part of the population at large (Haugen 1966; Pedersen 2005). In Europe, the standard variety and standardization are closely associated with the construction of the nation-state, the emergence of a bourgeois culture and the modernization of society. In a number of European countries, for instance, the national standard vernacular developed from one regional urban dialect that became progressively a supra-local vernacular (cf. standard Italian based on the dialect of Tuscany/Florence). But the transformation from a local vernacular to a national standard language was a long process involving the progressive institutionalization of the vernacular and its adoption by the urban bourgeoisie (the Bildungsbürgertum of Germany, England and Scandinavia) as well as the abandonment of the previous literary language (cf. Latin in the case of the Romance languages). The current situation of the Arab world appears rather different. As is well known, the desire for modernization expressed by the nineteenth century na}|a and conducted by Arab intellectuals and nationalists led to the adoption of a “modernized” form of Classical Arabic as the official national standard of most Arab countries rather than a national/local vernacular. Most discussions about the concept of standard/standardization have concerned the status/norms and stabilization of MSA (Mejdell 2006). However, a certain degree of codification and functional extension of the main Arabic vernaculars has occurred, but without official acceptance and institutionalization and often with denial. Focusing (i.e. stabilization and agreement on delineable norms) occurred as a result of speaker practice, i.e. native speakers have clear intuitions about the acceptability of a given utterance and speakers practice converge around certain norms (Walters 2003). Many vernaculars have been described and codified in grammars and dictionaries by foreign as well as local linguists. Urban vernaculars are more and more 19

CATHERINE MILLER

used in official spoken contexts as well as in writing (novels, plays, advertising, internet, SMS, etc.). But the reality of language practice is not recognized through official standardization and institutionalization. The official standard language remains MSA and not the national/urban vernaculars. In this respect, the increasing use of mixed styles (MSA-vernaculars) may appear to be a non-institutional, non-guided tentative move towards the expansion/diffusion and future standardization of the vernaculars. For the time being, the conceptualization of an urban standard as a national standard remains a hypothetical construct for many Arab countries. One issue to be resolved is which social group, life-style, cultural model is associated with the “standard” urban vernacular (when it exists), and how far the urban vernacular is recognized as a symbol of national expression by the overall society. Does the urban standard really correspond to the dialect variety spoken by the professional and educated classes of the capital city as is so often claimed? What is the convergence/similarity between the koineization process and the standardization process? What is the influence of MSA within these processes? Can we consider there to be a nationallybased educated spoken Arabic which de facto functions as the national standard? In some countries, the emergence of a “standard” vernacular goes along with the adoption of a number of non-national pan-dialectal features (cf. the influence of Egyptian/Levantine features in Yemen and the Arabian Peninsula). Another issue concerns the actors in the possible standardization of the urban/national vernacular. Do members of the urban elites play the same role as the urban European bourgeoisie of the nineteenth century, since many members of these urban elite received an education in English or French? Instead of looking for one national vernacular standard, should we postulate the coexistence of numerous varieties and ways of speaking according to context, interlocutors and self-image?

1.5 Multilingualism, globalization and new urban culture Apart from increasing dialect contact through migration and increasing contact with MSA through education and media, urbanization in the Arab world means increasing contact with non-Arabic languages, and opening up to globalization through the use of new technologies, circulation of new cultural models, social mobility and social contestation. Contact with non-Arabic languages implies a wide range of situations that will not be discussed here, including: •

Arabic as a minority language in dominantly non-Arab countries such as the old established Arab communities of Iran, Turkey, sub-Saharan Africa, Cyprus, or the more recent Arab diaspora of Europe and America. 20

ARABIC URBAN VERNACULARS

• •

Non-Arab groups living in contact with a dominant Arab environment including the Berbers in North Africa, the Kurds, Armenians, Nubians, etc. in the Mashreq. Contact with European languages such as French, English or Spanish which have been the official colonial language and still play an important role at the national/official level.

All these types of contact have led to types of language shift, language mixing, borrowing and code-switching. Contrary to the monolingual ideology that prevailed in the Arab countries, language diversity is a historical characteristic of most important Arab cities (Aguadé et al. 1998; Dakhlia 2004; Doss and Miller 1997). Depending on the particular historical and political context, this diversity has been differently accepted and conceptualized. The dominance of Arab pan-nationalism after independence and a context of post-colonial conflicts have contributed to a description in terms of language conflict, language domination, etc. The Arabization policies followed by most Arab states have led to a monolithic/monolingual perception of the relationship between the nation-state and the language and have fueled the linguistic claims of the non-Arab “minorities”. Important is the fact that a shift started to take place during the last decade of the twentieth century in many Arab countries, with progressive opening towards, and recognition of, language diversity, including code-switching and mixing associated with urban cultures. Code-switching is increasingly present in the language practices of the young urban population, in both the written and oral form. It is both a sign of informality and of ability to move across and play within the languages and the culture. Many instances of code-switching are found in new written practices such as chat on the internet, e-mails, SMS, as well as in new urban musical practices such as rap. More often than not, code-switching goes with a trend towards lexical creativity in the dialect. This “fashionable” use of code-switching occurs mainly between Arabic and European languages and functions as a badge of identity for youth belonging to the upper- and middle-class strata. From the available literature, code-switching appears to be socially more widespread in North African cities than in the Mashreq, where it remains an upper-class practice. In North Africa, code-switching is associated with urbanization, modernization, youth lifestyle and “fun”, functioning as a subversive humor device (Caubet 2002). Code-switching between Arabic and local languages (Berber, Kurdish, etc.) has not yet been described in detail, although local languages increasingly play a role in urban musical genres. Studies on code-switching in the Arab world have followed the general theoretical trends of code-switching studies, with many works focusing on linguistic rules and constraints. Owens, in Chapter 12, examines code-switching between Nigerian Arabic (NA), English, Hausa and Classical Arabic (CA) among Nigerian Arabic 21

CATHERINE MILLER

speakers in the city of Maiduguri. Arabic speakers in Nigeria form an old rural community. In Maiduguri, they represent about 10 percent of the total population of the city and are multilingual. Owens’ chapter studies instances of code-switching in five syntactic environments and demonstrates that English and CA insertions in a NA matrix follow various rules. Owens explores structural, psycholinguistic and areal factors to explain the differences in treatment. Code-switching practices among Maiduguri Arab speakers are not confined to youth speakers, but distinguishes urban from rural use. Ziamari, in Chapter 13, studies French-Moroccan Arabic (MA) codeswitching within the speech of young educated students from two professional schools of Meknes. She shows that compared to previous studies on French-MA code-switching, the morpho-syntactic rules of code-switching tend to become increasingly fluid and she points to numerous innovations. Girls seem to be more innovative at the structural levels, while boys appear more innovative at the lexical level. Code-switching is accompanied by lexical creation and semantic shift in both French and Arabic and therefore plays a cryptic and identity function for this age group. Youth urban linguistic and cultural practices have become an active field of urban sociolinguistics in the Western world. Youth, more particularly teenagers, are considered to be the most active initiators of language change (Eckert 2000). In Europe, particularly France, youth language is associated with the suburbs and with a wider process of social contestation (Caubet et al. 2004). In the Western world, the “youth/teenager” category is a well established social fact. Studies on youth language are more developed in North Africa than in the Mashreq. In North Africa, youth languages or youth ways of speaking are correlated to a wider social and cultural phenomenon, and involve criticism of Arabism and a narrow definition of national identity; an opening up towards different cultures considered to be parts of the national heritage; and a claim for more social, political and cultural freedom. The presence of youth language characterized by codeswitching and lexical creation goes together with the valorization of dArija (Moroccan Arabic), the growth of new urban music inspired by rap, rock, fusion, world music, etc. and a more general aspiration toward democratization (Caubet 2006). Rizq, in Chapter 14, describes the emergence of the “youth” category in Cairo (Egypt). She shows how the terms “youth”, “youth language”, “youth songs” became generalized in the mid-1990s in the Egyptian media and analyzes the interaction between public discourses, the cultural scene (movies) and the language practices of young students. Rizq presents a number of lexical and syntactic constructions recorded in the speech of young educated students. She shows that most features associated with youth language are in fact already present in Cairene dialectal Arabic and that “youth” ways of speaking are mainly a systematization of these discursive devices. One of the questions raised by Rizq’s chapter is the relevance of “youth” category in 22

ARABIC URBAN VERNACULARS

Egyptian society. How far is the perception of “youth”, acting as a distinctive group, an importation from western models or a local social reality? This question is important because in most Middle Eastern societies the young age group is demographically dominant and their language use may influence the overall society. But unlike Western societies, “youth” in the Arab world has had little public visibility and freedom. Moreover, family structure and hierarchy is still very strong and the nuclear family does not represent the dominant model even in old urban centers such as Cairo. Within Western sociolinguistic models, the nuclear family is believed to favor the acceleration of change and the influence of the youngsters (Auer et al. 2005). In this respect, Middle Eastern cities and societies may have reached a watershed. Gobalization, access to the internet and television modify the self-image of the younger generation as well as their cultural and linguistic practices. However, this age group is caught between contradictory models and may oscillate between tradition and change. One domain, so far rather neglected in Arabic sociolinguistics (with the exception of Rai music in North Africa) is the social function and linguistic characteristics of new urban music. Here again, Middle Eastern cities are at the junction of numerous trends between Western oriented musical genres, oriental music, the revival of traditional local genres, and African music and types of fusion. Urban music and songs accompany the social transformation of the society. Tauzin’s chapter describes the anthropological and linguistic aspects of rap music in Nouakchott. Rap is a new musical genre mainly performed by young people living in the popular suburbs. Rap singers are members of the low-status class known as Haratins. Rap performances contest the social order and reflect a radical social and cultural rupture. Rap songs mix Hassaniyya, African languages, French and, sometimes, English, each language having a different function. What is fascinating is the recycling of traditional Hassaniyya oral poetry devices as well as the important religious tone, a phenomenon encountered in other countries such as Senegal. Mixing different musical influences, rap is a perfect example of a growing trend towards globalization. The dominant social order is also at the heart of Dia’s chapter investigating the language attitudes and uses of Afro-Mauritanian young people living in Nouakchott. These young people speak Hassaniyya as a lingua franca, together with Wolof, Fulfulde and French. They have an ambivalent relationship with Hassaniyya, due to the existing political and racial tensions between the Moorish population and the African groups. It seems therefore that they refuse to shift exclusively to Hassaniyya and that they favor codeswitching. Dia’s contribution adds to our knowledge of the complex role of Arabic lingua francas in sub-Saharan Africa and multilingual cities such as Ndjamena in Chad and Khartoum in Sudan (Jullien de Pommerol 1997; Miller and Abu Manga 1992). 23

CATHERINE MILLER

References to African roots and cultures, sometimes via Black American models (rap and hip hop) seem to be expanding to other countries such as Morocco or Egypt. In these two countries, acknowledgement of an African heritage is advanced by young musicians and artists who want to free themselves from an “oriental” domination.

1.6 Conclusion The description of Arabic vernaculars has evolved tremendously in the last few decades, moving from the description of homogeneous and fixed diasystems to include attention to social interactions. In this respect, Arabic urban linguistics has followed the general trends of sociolinguistics with a shift from structures to meanings, from developmental perspectives to a contextual and interactionist approach. Many domains remain to be investigated. At the structural level, there is a paucity of studies dealing with syntax. There is also need for an anthropological-oriented approach concerning the role of family structure, the place and role of the individual versus the group, the evolution of gender status and women’s access to the public sphere (Sadiqi 2003), the issue of standardization, the conflicting representation of urbanity and modernity, and the trend toward globalization. Arab cities constitute an extremely important field of investigation not only for urban sociolinguistics but also for general linguistics because of crucial questions such as the conflictual representation of language and modernity and the relation between language, power and identity.

Appendix Table 1.1 Number of urban inhabitants plus urbanization rate by Arab country from 1900 to 2005 1900

1950

1980

2005

Number of urban inhabitants Algeria Morocco Tunisia Mauritania Libya Total Maghreb Egypt Syria Jordan Lebanon Palestine Yemen Bahraïn UAE

425,911 281,500 223,000 55,000 985,411 2,307,720 372,426 155,667 84,800 44,000 47,000

1,472,680 2,217,072 901,893

1900 1950 1980 2005 Rate of urbanization

7 ,157,235 7,350,074 3,121,984 306,081 192,667 1,953,726 4,784,312 19,889,100

19,920,679 9.1 16,322,432 8.0 5,865,969 11.7 915,503 5,010,175 8.0 48,034,758 8.7

8,087,484 24,249,826 1,174,680 4,018,863 149,582 1,407,815 354,000 1,853,756 197,843 729,985 320,768 960,145 65,871 269,307 44,000 784,061

48,226,544 11,194,537 3,850,060 3,553,920 2,617,480 5,500,367 570,255 3,819,000

24

22.8 39.4 26.8 15.3 1.8 69.1

16.5 25.3 26.3 20.5 20.9

39.3 38.1 49.3 20.0 63.9 41.1

59.9 53.8 58.4 32.9 85.7 58.4

39.6 33.2 29.9 37.6 19.2 7.5 59.6 8.8

59.8 46.3 54.9 60.6 56.0 9.5 79.6 30.6

70.0 60.9 74.3 60.8 74.6 27.8 80.2 73.7

ARABIC URBAN VERNACULARS

Table 1.1 (continued) 1900

1950

1980

2005

1900 1950 1980 2005

Number of urban inhabitants

Rate of urbanization

Saudi Arabia 193,000 420,479 4,685,397 18,420,111 12.8 Sudan (North) 139,000 516,692 2,848,543 10,283,321 5.6 Total Middle East 3,343,613 11,331,399 41,807,698 108,035,595 17.5

12.1 7.9 27.4

50.4 19.8 45.0

79.5 36.1 60.3

Total Arab world

25.1

43.7

59.7

4,329,024 16,115,711 61,696,798 156,070,353 14.5

Note: Urbanization is accounted for cities > 10,000 inhabitants Source: GEOPOLIS

Table 1.2 Population growth of 1st, 2nd and 3rd cities of each Arab country from 1860 to 2005 City

Country

1860

1920

Algiers Oran Constantine

Algeria Algeria Algeria

62,174 28,330 38,365

203,927 135,070 61,172

422,100 253,282 88,514

1,646,360 532,763 374,634

3,360,788 754,023 480,278

Casablanca Rabat Fez

Morocco Morocco Morocco

700 40,000 88,000

102,000 55,000 71,000

700,559 200,763 177,578

2,136,088 782,035 420,232

3,569,988 1,670,618 983,751

Tunis Sfax Sousse

Tunisia Tunisia Tunisia

145,000 3,000 8,000

227,837 37,149 37,994

601,695 75,237 80,999

1,062,486 317,328 156,224

1,926,788 485,294 340,731

Nouakchott Nouadhibou Zouerat

Mauritania Mauritania Mauritania

— — —

— — —

4,812 685 —

184,139 29,165 19,834

628,814 77,901 36,385

Tripoli Benghazi Misrata

Libya Libya Libya

40,000 15,000 —

70,000 36,000 14,000

105,741 55,989 29,950

771,928 348,445 87,335

1,595,490 638,050 274,064

Riyadh Jeddah Damman

Saudi Saudi Saudi

— 30,000 —

23,664 25,000 —

97,892 49,898 —

1,054,529 848,290 430,475

4,241,543 2,868,269 1,616,266

Damascus Aleppo Homs

Syria Syria Syria

48,000 125,000 20,000

167,152 154,383 53,360

340,616 379,193 124,232

1,233,389 938,154 328,594

2,799,352 1,688,558 774,567

San¡a Ta¡iz Aden

Yemen Yemen Yemen

— — 20,000

23,000 — 56,500

85,000 25,000 118,768

256,819 86,739 289,040

1,790,275 623,317 601,079

Cairo Alexandria Al-Mahalla Al-Kubra

Egypt Egypt Egypt

277,438 184,447 25,000

961,648 456,672 41,464

2,578,284 1,025,943 126,265

7,732,819 2,409,472 364,475

12,617,647 3,340,353 559,709

25

1950

1980

2005

CATHERINE MILLER

Table 1.2 (continued) City

Country

1860

1920

1950

1980

2005

Amman Irbid Aqaba

Jordan Jordan Jordan

— — —

— — —

89,860 19,947 2,186

747,481 120,093 27,926

257,0596 388,094 81,533

Beirut Tripoli Saida

Lebanon Lebanon Lebanon

— — —

128,529 4,0449 12,000

211,000 100,000 20,000

1,030,474 441,170 120,630

2,153,176 558,519 207,366

Baghdad Mossul Basrah

Iraq Iraq Iraq

250,000 60,000 30,000

627,885 144,925 114,914

2,832,044 490,162 372,019

6,252,189 1,940,672 1,552,537

Source: GEOPOLIS

Notes 1 The publication of the Encyclopedia of Arabic Language and Linguistics (EALL, 4 volumes, 2006–8), fills a very important gap. It includes 14 lemma on specific Arabic urban vernaculars (Amman, Baghdad, Beirut, Bahrain, Cairo, Damascus, Jerusalem, Khartoum, Kuweit, San¡a, Tripoli, Tunis) as well as numerous lemma on national vernaculars. The lemma describe “stable diasystems” and do not account for variation, but they include historical data and valuable bibliographies. 2 The complete report of the conference, including a summary of the contributions can be consulted in the ESF website at http://www.esf.org/fileadmin/be_user/ research_areas/social_sciences/documents/Arabic_Urban_Vernaculars.pdf 3 By focusing on Arabic urban vernaculars, this book complements previous collective publications dealing with contact in Arabic (Owens 2000; Rouchdy 2002), as well as many individual papers published in journals or proceedings. A preliminary “state of the art” study of variation and change in urban vernaculars can be found in Miller (2004). 4 For lack of space, bibliographical references have been kept to the minimum. For Section 1.2, readers looking for more references are advised to consult the EALL. For Section 1.3, the references already mentioned by the contributors of this book have not been repeated and can be found in each relevant chapter. 5 The urban koine hypothesis led to huge controversies and has involved numerous eminent linguists such as Blau, Cohen, Corriente, Ferguson, Fück, Levin, Versteegh, etc. For a brief bibliography see Abboud-Haggar (2006) and Miller (2006). 6 Those issues have been discussed since the beginning of Western Arabic linguistics. For some recent references cf. Holes 1995a, 1996; Owens 2005b; Versteegh 1997. 7 Studies on communal/tribal affiliation in several Middle Eastern cities have shown that tribal affiliation is very much instrumentalized in the urban context (Eickelman 2002; Seurat 1985). In the Gulf, Holes (2005) considers that the Bedouin/sedentary distinction still survives in the collective memory. See also Shryock (1997) for the genealogical imagination of Jordanian nationalism. 8 The GEOPOLIS data-base established by Eric Denis and François Moriconi (CNRS-Paris) records the demographic growth from 1860 up to date of all Arab

26

ARABIC URBAN VERNACULARS

cities that surpass 10,000 inhabitants in 2005. Thanks to Eric Denis for giving me access to these statistics. 9 The stigmatization of the popular suburbs and their identification through derogatory terms which associate spatial and social categorizations (such as “slums”, “ghettos”, etc.) is also a universal paradigm (Depaule 2006). 10 El-Himer distinguishes three varieties: old urban vernacular ( parler citadin) located in the Medina, neo-urban vernacular ( parler urbain) located in the close peripheries of the medina and what he calls parlers à aspect ruraux, i.e. vernaculars with rural features located in the outskirts. Each variety is identified by a set of isoglosses and appears as a kind of prototypical construction. 11 The issue of diglossia, contact and variation between MSA and dialects constitutes the dominant focus of Arabic linguistics and will not be dealt with here. For recent comprehensive reviews of the question see Mejdell (2006), BoussofaraOmar (2006) and numerous issues of the series Perspectives on Arabic Linguistics.

References Abboud-Haggar, S. (2006) “Dialect genesis”, EALL, Vol. 1, 613–622. Abdel Jawad, H. R. (1986) “The emergence of an urban dialect in the Jordanian urban centres”, International Journal of the Sociology of Language 61, 53–63. Abdel Jawad, H. R. (1987) “Cross dialectal variation in Arabic: competing prestigious forms”, Language in Society 16: 3, 359–368. Abu Haidar, F. (2006) “Bedouinization”, in EALL, Vol. 1, 289–274. Aguadé, J., Cressier, P. and Vicente, A. (eds) (1998) Peuplement et arabisation au Maghreb occidental, Madrid-Zaragoza: Casa de Velazquez- Universidad de Zaragoza. Alessa, A. (2006) “When Najd meets Hijaz: dialect contact in Jeddah”, a paper presented at the 7th International Conference of AIDA, Vienna (Austria), 6–9 September 2006. Al-Wer, E. (1997) “Arabic between reality and ideology”, International Journal of Applied Linguistics 7: 2, 51–65. Al-Wer, E. (2002) “Education as a speaker variable”, in A. Rouchdy (ed.), pp. 41– 53. Amara, M. (2005) “Language, migration and urbanization: the case of Bethlehem”, Linguistics 43: 5, 883–902. Auer, P., Hinskens, F. and Kerswill, P. (eds) (2005) Dialect Change, Convergence and Divergence in European Languages, Cambridge: Cambridge University Press. Barth, F. (1969) Ethnic Groups and Boundaries, Boston, MA: Little Brown and Co. Blanc, H. (1964) Communal Dialects in Baghdad, Cambridge, MA: Harvard University Press. Bonine, M., Elhers, E., Krafft, T. and Stober, G. (1994) The Middle Eastern City and Islamic Urbanism: An Annotated Bibliography of Western Literature, Bonn: Ferd. Dümmlers Verlag. Boucherit, A. (2002) L’arabe parlé à Alger: aspects sociolinguistiques et énonciatifs, Paris-Louvain, Peeters. Boucherit, A. and Lentin, J. (1989) “Les dialectes féminins dans le monde arabe: des dialectes minoritaires et leur évolution”, in E. Koskas and D. Leeman (eds) Genre et Langage, Paris: Linx 21, pp. 17–37.

27

CATHERINE MILLER

Boussofara-Omar, N. (2006) “Diglossia”, EALL, Vol. 1, 629–636. Bulot, T., Bauvois, C. and Blanchet, P. (eds) Sociolinguistique urbaine: variations linguistiques, images urbaines et sociales, Rennes: Presses de l’Université de Rennes. Cadora, F. (1992) Bedouin, Village and Urban Arabic: An Ecolinguistic Study, Leiden: Brill. Calvet, L. J. (1994) Les voix de la ville, Paris: Payot. Caubet, D. (2002) “Jeux de langues: humor and codeswitching in the Maghreb”, in A. Rouchdy (ed.), pp. 233–255. Caubet, D. (2006) “The emergence of a new ‘movida’ in Morocco? The use of darija in artistic creation at the beginning of the 3rd Millenium”, a paper presented at the 7th International Conference of AIDA, Vienna (Austria), 6–9 September 2006. Caubet, D., Billiez, J., Bulot, T., Léglise, I. and Miller, C. (eds) (2004) Parlers jeunes, ici et là-bas: pratiques et représentations, Paris: L’Harmattan. Chambers, J. K., Trudgill, P. and Schilling-Estes, N. (eds) (2002) The Handbook of Language Variation and Change, Malden, MA: Blackwell Publishing. Cohen, D. (1973) “Variantes, variétés dialectales et contacts linguistiques en domaine arabe”, Bulletin de la Société de Linguistique (Paris) 68, 215–248. Cohen, D. (1981) “Remarques historiques et sociolinguistiques sur les parlers des juifs maghrébins”, International Journal of the Sociology of Language 30, 91–106. Daher, J. (1999) “Gender in linguistic variation: the variable (q) in Damascus”, Perspectives on Arabic Linguistics XI, 183–206. Dakhlia, J. (ed.) (2004) Trames de langues:usages et métissages linguistiques dans l’histoire du Maghreb, Paris: Maisonneuve et Larose. Dendane, Z. (2002) “Aspects of women’s speech in Tlemcen”, Cahiers de Linguistique et didactique. Nouvelle Série 1/2: 38–46 (Oran). Denis, E. (2007) “Des villes introuvables”, in E. Denis (ed.) Villes et urbanisation des provinces égyptiennes, Paris: Karthala-Cedej, pp. 9–30. Depaule, J. C. (ed.) (2006) Les mots de la stigmatisation urbaine, Paris: Editions UNESCO/MSH. Donner, F. (1981) The Early Islamic Conquests, Princeton, NJ: Princeton University Press. Doss, M. and Miller, C. (eds) (1997) Les langues en Egypte, in Egypte Monde Arabe 27–28, Cairo: Cedej. EALL (2006–8) Encyclopedia of Arabic Languages and Linguistics, 4 volumes, edited by K. Versteegh, M. Eid, A. Elgibali, M. Woidich and A. Zaborski, Leiden: Brill. Eckert, P. (2000) Linguistic Variation as Social Practice: The Linguistic Construction of Identity in Belten High, Malden, MA: Blackwell Publishers. Eickelman, D. F. (2002) The Middle East and Central Asia: An Anthropological Approach, 4th edition, Upper Saddle River, NJ: Prentice Hall. El-Himer, M. (2001) “Identité urbaine de la population de Salé”, in T. Bulot et al. (eds), pp. 99–116. Elzeini, N. (2006) “Greetings”, EALL, Vol. 2, 202–207. Ferguson, C. ([1987] 1997) “Standardization as a form of language spread”, in K. Belnap and N. Haeri (eds) Structural Studies in Arabic Linguistics: Charles A. Ferguson’s Papers 1954–1994, Leiden: Brill, pp. 69–80. Freeman, A. (2006) “Why there is no koine in San’aa, Yemen”, Perspectives on Arabic Linguistics XVI, 68–82.

28

ARABIC URBAN VERNACULARS

Gibson, M. (2002) “Dialect levelling in Tunisian Arabic: towards a new spoken standard”, in A. Rouchdy (ed.), pp. 24–40. Haeri, N. (1996) The Sociolinguistic Market of Cairo: Gender, Class and Education, London and New York: Kegan Paul International. Harb el Kak, M. (2006) “La Dahiye de Beyrouth”, in J. C. Depaule (ed.), pp. 199– 224. Haugen, E. (1966) “Dialect, language, nation”, American Anthropologist 68, 922–935. Holes, C. (1987) Language Variation and Change in a Modernising Arab State: The Case of Bahrain, London: Kegan Paul International. Holes, C. (1995a) Modern Arabic: Structures, Functions and Varieties, London: Longman. Holes, C. (1995b) “Community, dialect and urbanization in the Arabic-speaking Middle East”, BSOAS 58: 2, 270–287. Holes, C. (1996) “The Arabic dialects of south eastern Arabia in a socio-historical perspective”, Zeischrift für Arabische Linguistik, pp. 36–50. Holes, C. (2005) “Dialect and national identity: The cultural politics of selfrepresentation in Bahrain musalsalaat”, in P. Dresch and J. Piscatori (eds) Monarchies and Nations: Globalization and Identity in the Arab States of the Gulf, Reading: I. B. Tauris, pp. 52–72. Holes, C. (2006a) “Bahraini Arabic”, EALL, Vol. 1, 241–255. Holes, C. (2006b) “Gulf States”, EALL, Vol. 2, 210–216. Ibrahim, M. (1986) “Standard and prestige language: a problem in Arabic sociolinguistics”, Anthropological Linguistics 28, 115–126. Ingham, B. (1982) North East Arabian Dialects, London: Kegan Paul International. Jabeur, M. (1996) “Women, social change and linguistic variation in the urban dialect of Tunis”, in C. Holes (ed.) Proceedings of the 2nd International Conference of AIDA held at Cambridge 10–14 September 1995, Cambridge : University Publications Centre, pp. 85–94. Jastrow, O. (2006) “Iraq”, EALL, Vol. 2, 414–424. Jullien de Pommerol, P. (1997) L’arabe tchadien:l’émergence d’une langue véhiculaire, Paris: Karthala. Kerswill, P. (2005) “Migration and language”, in U. Ammon, K. Mattheier and P. Trudgill (eds) Handbook of Sociolinguistics, Vol. 2, Berlin, de Guyter. Kharoufi, M. (1995) “Urbanisation et recherche dans le monde arabe”, MOST Documents no. 11, (http://www.unesco.org/most/kharoufi.htm, accessed January 2007.) Larcher, P. (2006) “Sociolinguistique et histoire de l’arabe selon la muqaddima d’ibn Khaldûn (VIII/XIV siècle)”, in P. G. Borbone, A. Mengozzi and M. Tosco (eds) Loquentes Linguis: Studi linguistici e orientali in onore di F. A. Pennacchietti, Wiesbaden: Harrassowitz, pp. 431–441. Lentin, J. (1981) “Remarques sociolinguistiques sur l’arabe parlé à Damas”, unpublished thesis, Paris: Université Paris III, Sorbonne Nouvelle. Lentin, J. (2006) “Damascus Arabic”, EALL, Vol. 1, 546–555. Mejdell, G. (2006) Mixed Styles in Spoken Arabic in Egypt, Leiden: Brill. Messaoudi, L. (2001) “Urbanisation linguistique et dynamique langagières dans la ville de Rabat”, in T. Bulot et al. (eds), pp. 87–98. Messaoudi, L. (2003) Etudes sociolinguistiques, Kenitra: Faculté des Lettres et Sciences Humaines.

29

CATHERINE MILLER

Miller, C. (2004) “Variation and changes in Arabic urban vernaculars”, in M. Haak, K. Versteegh and R. Dejong (eds) Approaches to Arabic Dialects: Collection of Articles Presented to Manfred Woidich on the Occasion of his Sixtieth Birthday, Amsterdam: Brill, pp. 177–206. Miller, C. (2005) “Between accommodation and resistance: Upper Egyptians migrants in Cairo”, Linguistics 43: 5, 903–956. Miller, C. (2006) “Dialect koine”, EALL, Vol. 1, 593–597. Miller, C. and A.-A. Abu Manga (1992) Language Change and National Integration, Rural Migrants in Khartoum, Reading-Khartoum: Garnett-Khartoum University Press. Naciri, M. and Raymond, A. (eds) (1997) Sciences sociales et phénomènes urbains dans le monde arabe, Casablanca: Fondation du Roi Abdul-Aziz Al Saoud. Naim, S. (2006) “Beirut Arabic”, EALL, Vol. 1, 275–286. Owens, J. (ed.) (2000) Arabic as a Minority Language, Berlin: Mouton de Gruyter. Owens, J. (2005a) “Introduction”, Linguistics 43–5 (Special Issue: Language, Migration and the City: Corpus-based Approaches), 871–882. Owens, J. (2005b) “Pre-diaspora Arabic: dialects, statistics and historical reconstruction”, Diachronica 22: 2, 271–308. Palva, H. (1982) “Patterns of koinezation in modern colloquial Arabic”, Acta Orientalia XLIII, 13–32. Palva, H. (2006) “Dialect classification”, EALL, Vol. 1, 604–613. Pedersen, I. (2005) “Process of standardisation in Scandinavia”, in P. Auer et al., pp. 171–195. Raymond, A. (1993) Le Caire, Paris: Fayard. Rosenhouse, J. (2006) “Bedouin Arabic”, EALL, Vol. 1, 259–269. Rouchdy, A. (ed.) (2002) Language Contact and Language Conflict in Arabic, New York: Routledge-Curzon. Royal, A. M. (1985) “Male/female pharyngealization patterns in Cairo Arabic, a sociolinguistic study of two neighborhoods”, Texas Linguistic Forum 27. Sadiqi, F. (2003) Women, Gender and Language in Morocco, Leiden: E. J. Brill. Sawaie, M. (1994) Linguistic Variation and Speakers Attitudes, Damas: Al Jaffar & Al Jabi Publisher. Sawaie, M. (2006) “Jordan”, EALL, Vol. 2, 498–505. Seurat, M. (1985) “Le quartier de Bâb Tebbâne à Tripoli (Liban): etude d’une ‘assabiyya urbaine”, in Mouvements communautaires et espaces urbains au Mashreq, Beyrouth: Cermoc, pp. 45–86. Shryock, A. (1997) Nationalism and the Genealogical Imagination: Oral History and Textual Authority in Tribal Jordan, Berkeley, CA: University of California Press. Siegel, J. (1993) “Dialect contact and koineization: review article”, International Journal of the Sociology of Language 99, 105–122. Tarrius, A. (2000) Les nouveaux cosmopolitismes: mobilités, identités, territoires, La Tour d’Aigle: Edition de l’Aube. Trudgill, P. (1986) Dialects in Contact, Oxford: Basil Blackwell. Trabelsi, C. (1988) “Les usages linguistiques des femmes de Tunis”, unpublished thesis, Paris, Université de Paris III. Vanhove, M. (2002) “Processus de koïnisation dans le dialecte arabe de Yafi (Yemen)”, in A. Youssi, F. Benjelloun, M. Dahbi and Z. Iraqui-Sinaceur (eds) Aspects of the Dialects of Arabic Today, Rabat: AMAPATRIL, pp. 151–159.

30

ARABIC URBAN VERNACULARS

Versteegh, K. (1997) The Arabic Language, Edinburgh: Edinburgh University Press. Woidich, M. (1994) “Cairo Arabic and the Egyptian dialects”, in D. Caubet and M. Vanhove (eds) Actes des premières journées internationales de dialectologie arabe, Paris: Inalco, pp. 493–510. Walters, K. (2003) “Fergie’s prescience: the changing nature of diglossia in Tunisia”, International Journal of the Sociology of Language 163: 77–110.

31

THE (R)URBANIZATION OF MAURITANIA

Part I MIGRATION, URBANIZATION AND LANGUAGE CHANGE

33

CATHERINE TAINE-CHEIKH

34

THE (R)URBANIZATION OF MAURITANIA

2 THE (R)URBANIZATION OF MAURITANIA Historical context and contemporary developments Catherine Taine-Cheikh

2.1 Introduction The population of Nouakchott numbers around 745,000 inhabitants, and accounts for approximately one quarter of Mauritania’s total population. It is the capital city of one of the multi-ethnic countries of the Sahelo-Saharan region, and was created almost ex-nihilo in 1957, three years before the independence of Mauritania. Its growth is symptomatic of the recent transformation of the Mauritanian way of life, which particularly affected the demographically dominant Moorish ethnic group (the BC1An). This Arabicspeaking population, which was still 95 percent nomadic at the dawn of the twentieth century, is now 80 percent sedentary. Thus the urbanization of Mauritania is almost contemporaneous with the birth of a multi-lingual nation, the establishment of a modern state and the entry of Africa into the global world. To understand the sociolinguistic impact of these recent changes, I shall briefly describe pre-modern Moorish society and examine the historical factors that may account for its impressive linguistic and cultural homogeneity. I shall then discuss the relevance of the notion of urbanity and urban spaces in the pre-modern and modern Mauritanian context and highlight some of the most recent sociolinguistic changes. The (r)urbanization of Mauritania raises the issue of the relationship between bedouinity and urbanity in the Arabic speaking world. It also involves the questions of diversity versus homogeneity, mixing versus boundaries, deviance versus norm, and humor versus state-control.

35

CATHERINE TAINE-CHEIKH

Map 2.1 Nouakchott (Mauritania)

36

THE (R)URBANIZATION OF MAURITANIA

2.2 Bedouinity and linguistic homogeneity in the Moorish context With around three million people (three quarters of whom live within Mauritania’s borders), the Moorish ethnic group represents a small community inside the Arabic-speaking world. Its linguistic homogeneity may therefore seem normal, but it must be noted that the Moorish territory covers almost two million square kilometers. Low population density and nomadism, the traditional Moorish way of life, seem to go along with a certain linguistic conservatism, as attested in other nomadic Arab societies (Ingham 1979: 25). But a number of other historical and social factors might also account for the quite remarkable homogeneity of the Moorish Hassaniyya dialect (}assAniyya). 2.2.1 A slow and gradual process of arabization For millennia, the West African Sahelo-Saharan region has been a contact zone between black populations of sedentary farmers and white populations of nomadic herdsmen. Despite the proximity of the Atlantic, no significant contribution came from the coastal environment and thus there were no important differences between West and Central Sahara. The Berber population was, for a long time, common to both regions. Traditional Arabic sources mention the presence of Zenete groups (Tuareg) in Central Sahara and of Zenaga (also known as oanhaja) in West Sahara. Despite a number of common features, each group seems to have developed independently and had its own specific vernacular. Until the beginning of the fourteenth century, contacts with Arabs remained limited. The islamization that began in the ninth century went further in the eleven century under the influence of the Almoravid movement. But in this case – quite exceptional in the history of the region – the expansionist movement started from the Berber Sahara and spread to the northern regions, some of which were already arabized. The arabization of West Sahara was an indirect consequence of the arrival in North Africa of the Bane Sulaym, Bane Ma¡qil and Bane Hilal Arab nomads. The Bane massan who gave their name to the Arabic dialect of Mauritania (}assAniyya or klAm }assAn), are a branch of the Bane Ma¡qil which was to begin its slow migration to the south at the end of the thirteenth century leaving significant traces in Morocco. Travelers’ reports dating from the fourteenth to sixteenth centuries portray a relatively peaceful coexistence between the Berbers and the Arabs, with some cases of interpenetration. No further population influx followed that of the Bane massan, and the history of the past five centuries is one of internecine fighting, intertribal conflicts and alliances and tribal reconstruction. It resulted in the almost total extinction of the Zenaga language (probably at some time during the nineteenth century, although there is no reliable information on this) 37

CATHERINE TAINE-CHEIKH

and the virtual disappearance of the Berber identity (no one claims to be of oanhaja descent any more). Linguistically speaking, a dialect quite closely related to that spoken by the Bane massan had prevailed. It has of course evolved since the fourteenth century, under the influence of the Berber substrate, but the interesting question is how this evolution has failed to lead to greater internal divergences within such a large geographical area. The “conservative” Hassaniyya dialect shows features common to most Arabic Bedouin dialects (/g/ realization of the qAf; preservation of interdentals; maintenance of synthetic constructions; imperfective without verbal prefix; simple negation, etc.) as well as features peculiar to North African dialects (first person of the imperfective sg. with n- -; a tendency to reorganize the syllabic structure by dropping the short vowel when in non-final open syllable; the opposition /z/-/z/, etc.). But Hassaniyya also has its own specific features such as the use of the particles lAhi to express the future; bChälli “because” or kCv ¢lli “as if”; a passive form in u- and elative forms in a- for the quadriliteral roots and the derived verbs; the creation of a factitive in s- for certain derived verbs in st- (cf. Taine-Cheikh 1991). Whatever the origin of these specific features (whether calques of Zenaga substrates, parallel innovations or original Hassaniyya creations), it is important to note that they are found among all Hassaniyya speakers. This implies a social and cultural setting which favors both the maintenance of the whole linguistic system and the propagation of punctual change. Beyond the geographical isolation factor mentioned above, several points are worth examining in attempting to explain this linguistic uniformity. 2.2.2 Social cohesion and cultural identity The natural environment and the material culture have undergone very few modifications since the arrival of the Bane massan. Many Zenaga words, referring to local cultural practices have been integrated into Hassaniyya, without disrupting the system. The same continuity seems to apply at the social level. The Moorish society had (and still has) a very strong social hierarchy, with the aristocrats at the top and the slaves at the bottom, while the }WAYCn (freed slaves and others) would have a superior status to the dependents, even if the former were generally “black” and the latter “white.” This major opposition was supplemented by two castes: the musicianssingers (the griots iggAw¢n) and the craftsmen-blacksmiths. The oanhaja society may well have been organized along very similar lines, with the exception of the tributary category (known as aZnAgä) which clearly dates from after the arrival of the Bane massan. The distinction between warriors (}assAn or, more prestigious, ∞Wab) and marabouts or learned (zwAyä) establishes a secondary hierarchy within the aristocracy which also follows the arrival of Arab Bedouins. 38

THE (R)URBANIZATION OF MAURITANIA

These hierarchical relationships could have led to a fracturing of the dialect and a limited acquisition of the masters’ idiom, particularly in the case of black slaves (and griots, often but not always of former southern “black” origin). It has to be stated however that the nomadic social unit, the camp, included all classes of society, in particular aristocrats, slaves and blacksmiths (as well as griots in the case of the most prestigious warrior tribes). Even if the }WAYCn were relatively independent (and often more sedentary), their villages fall, both economically and politically, within the tribal sphere of influence. The grouping in sub-groups, groups, and tribes was (and still is) based on (frequently revised) genealogies linked to common ancestors. This could well have brought about linguistic diversification, but uniformity has prevailed, within as well as between tribes.1 Tribal entities, sometimes spread across the whole territory, have continually split and recomposed, but have always maintained numerous neighborhood relationships, including the exchange of women in accordance with the rule of female hypergamy.2 The rule of competitive equality that holds between the “aristocrats” within a group may also have contributed to the limitation of innovations, for surely one gives up a little of one’s own independence if one adopts gratuitously someone else’s speech patterns.3 Culturally speaking, the total religious unity of the Moorish Sahara (based on the trilogy – Malekism, Ash¡arism, (theological theory) and confraternity) is remarkable, and is paralleled by the unified corpus taught by the zwAyä. The klAm }assAn was adopted by everyone at the oral level (as klAm ¢lBC1An) due to its prestige, not only as the language of the victorious warriors – the one in which the griots sang their praises – but because, in the eyes of the literate, it benefited from the great privilege of being a “pathway to classical Arabic” (the formula is from a seventeenth century Moorish sage), i.e. to the language of the Quran. Furthermore, the vitality of popular poetry (γnä and thäydCn), with its quite complex quantitative metric, might have contributed to the development of the expressive capacities of Hassaniyya, incidentally providing a means to spread innovations. It has certainly contributed at least to its recognition as a much appreciated means of oral communication. Finally, one can note that Moorish society in general and the literate in particular – the “ideology professionals” to quote Ould Cheikh (1985) – tend to adopt a negative attitude toward change. This is a fairly common attitude among migrant or peripheral communities as well as among newly converted ones: in the absence of an inherited legitimacy, they often adopt strong normative attitudes. The Moorish people, in their aspiration toward arabness, sought to distance themselves from their southern neighbors (cf. the “Whites/Blacks” opposition of BC1An/SEdAn in Taine-Cheikh 1989). For all these reasons, the zwAyä may have monitored innovations and strongly 39

CATHERINE TAINE-CHEIKH

stigmatized any deviance that might be seen to dilute the “arabness” of the dialect, which was essential to Moorish identity. The linguistic dynamic observed in Hassaniyya seems to correspond broadly to the idea of “acceptable” change.4 The tendency (particularity frequent among the }WAYCn though not solely confined to them) to confuse hamza and ∞ayn or emphatic and non-emphatic consonants (s ~ X, t ~ Y, O ~ 1 ), is often stigmatized by the zwAyä (cf. the realization of YWAb “earth, ground” among the }WaYCn of the southwest).5 All in all, it appears that in spite of its geographical spread and social divisions, Moorish Bedouin society was able to maintain its linguistic homogeneity through time and space. Moreover, no linguistic distinction was prevailing between city-dwellers and Bedouin groups in pre-modern Moorish society.

2.2.3 Nomadism and caravan cities in pre-colonial era The mountains (especially the 1haW plateau of Tichit) were the site of an important sedentary settlement during the Neolithic era but, when considering the desert city of the pre-colonial era, it is the image of kXaW (or gXaW) that comes to mind. A certain continuity between the two types of settlement is possible, since the agricultural workers of the oasis (the }WAYCn) may partly be the descendants of an ancient population which inhabited the Sahara even before the arrival of the Berbers. The meaning of qXar/qX¢W as “market town” and “small fortified town” is considered as specifically Saharan according to Maghrebi Arabic dictionaries (Beaussier 1958: 807; Colin 1993: 1578). It differs from the classical meaning of qaXr “palace”. The Moorish region, beyond Tombouctou in Mali, encompassed around twenty kXEW (with not much more than a few hundreds inhabitants each), which can be divided into two groups. The first kXEW were established between the twelfth and fifteenth centuries (Oualata, Tichit, Ouadane and Chinguetti). While Oualata, located at the border of Mauritania and Mali is rather specific, the other three kXEW share many similarities with the cities of the second group, founded between the sixteenth and nineteenth centuries (Atar, Tidjikja, Akreyjit, Rachid and Kasr el-Barka). These were cities of dry stone construction, located in the mountainous regions of Adrar and Tagant, and associated with palm groves, and had little in common with the southern villages (Rosso, Boghe, Kaedi, Selibaby, etc.) inhabited by the ähl l¢-}yEt (the sedentary Hal Pulaaren, Soninkes and Wolofs). They depended on the caravan trade and in most cases were founded by zwAyä tribes. Indeed, marabouts not only had control over the economic activities related to the trans-Saharan trade, but they also undertook the intellectual work associated with the teaching and training provided in the ma}A1¢W.6 40

THE (R)URBANIZATION OF MAURITANIA

The prosperity of the various caravan cities often rose and fell over time, due either to shifting trade routes or to economic and political conflicts. Internal rivalries often forced part of the population to emigrate and several cities were established in the wake of such internecine struggles (cf. Tidjikja founded by a faction of the Idaw¡ali – the “Blacks” – following their expulsion from Chinguetti). In the cities, tribal intermixing was very limited. Generally, no more than two or three tribes lived in the same city and often in separate districts. (cf. the names of districts in Oualata). Even Mosques – the only public meeting places – were subject to a certain form of tribal appropriation, for the designation of the imam was in the power of a particular tribe.7 The kXEW, devoid of public places and markets, appear generally as a juxtaposition of houses, completely closed in on themselves, set along winding sand-covered streets. The main function of the ancient cities was to protect the produce of the oasis and the merchandise of the trans-Saharan trade. The surrounding walls made the surveillance of the houses easier when the whole household resumed their nomadic lifestyle with their herd. For most of the kXEW’s zwAyä were only seasonally sedentary. While Oualata developed an urban culture (mural decorations, a more sophisticated cuisine, pottery, craft objects made of particular types of wood, feminine embroidery, etc.), this is not really true of the other kXEW which seemed to have served essentially the storehouse function. The ancient cities then present a rather chaotic and fragmented picture due to their having been essentially conglomerates of family units – independent and interdependent at the same time.8 From a sociological, economic and cultural point of view, there was no distinction between the city dweller and the nomad. Beyond a certain economic role, the zAwi of the kXAW remained culturally a Bedouin and socially a nomad. The issue is more complex for the agricultural workers of the palm groves. Their integration within the Moorish world was probably more indirect (Bonte 1998) and led, from a linguistic point of view, to two phenomena: 1 2

The presence of specific lexical fields (especially relating to palm cultivation and architecture). The persistence, almost until the twentieth century, of Azayri speakers.

We know very little about Azayr, which was apparently a variety of Soninke mixed with Berber. It may have been either a vernacular Soninke (it was spoken among the Maxna group which seems to have originated in the south) or a survival of an ancient vernacular language spoken at the time of commercial relations with the Ghana empire. In any case, these linguistic particularities did not have much affect on the unity of Hassaniyya. In the ancient cities, the Azayr language may have suffered the fate of Zenaga in the southwest,9 whose role was to resist the Arabic vernacular as best it could 41

CATHERINE TAINE-CHEIKH

and finally to disappear after a transitory period of bilingualism, leaving no more than a few traces in the form of loanwords. Such a phenomenon is certainly not unique but may well be characteristic of a part (often Bedouin?) of the Arab world (Miller 1996).

2.3 (R)urbanization and “modernity” Did the caravan cities contribute to the creation of an urban space? The answer is highly dependent on the definition of “urban”. If we follow the Khaldunian paradigm with regard to the distinction between nomadic and sedentary peoples and if bedouinity is equated with the ∞aXabiyya and the nomadic lifestyle, then it appears that urbanity was not only absent from the ancient cities but is still largely in its infancy even in a city like Nouakchott (Ould Cheikh 2006). If, on the other hand, the city is considered as “a problem and not as an object defined a priori” (Louiset 2001: 232), with the idea that different modes of urbanity co-exist and that not all of them involve a discontinuity between urban and rural (or sedentary and nomadic) peoples, then the answer becomes probably positive both for modern and for ancient cities. If the points of view of anthropologists and geographers tend to diverge, where does the linguist stand? And indeed the dialectologist, who knows the usefulness of certain classifications (the reality of the distinction between Bedouin and sedentary dialects, or even the sub-categorization into rural versus urban dialects) but who also knows that there is no automatic correspondence between the spoken variety and the current lifestyle?10 The prevailing feeling concerning the old Moorish cities is that they lacked the linguistic specificity that characterized many North African cities like Fes, Tlemcen and Djidjelli (the famous North African pre-Hilali city dialects). This of course is due to their completely different history of both arabization and urbanization. Until the middle of the twentieth century then, Mauritania was characterized by a relatively undifferentiated variety of spoken Arabic (Hassaniyya) with few regional, social or communal variants. But the changes that have occurred during the most recent decades of Mauritanian history have had important consequences for the sociolinguistic situation of the country, and particularly for the capital, Nouakchott. This is the very epitome of a modern city, and was originally a relatively controlled space dominated by power and public administration. Soon though its status as a city-state led to its transformation into a sprawling city, with rampant development (a kind of octopus-town sucking in all the resources as well as all the poverty of the country). 2.3.1 Urban spaces and city politics In Mauritania, the first proto-state entities were the four emirates (Trarza in the southwest, Brakna in the center south, Adrar in the north and Tagant 42

THE (R)URBANIZATION OF MAURITANIA

in the center). Offshoots of warlike tribes (and, in three cases out of four, of tribes of Arab origin), the emirates had only an indirect link with the traditional cities described above. It is only in 1902, with the beginning of French colonization, that cities and villages in Mauritania experienced a real blossoming, due to the physical presence of an embryonic administration. 2.3.1.1 Cities with two centers The administrative presence is materialized, in both ancient and modern cities, by the existence of a distinct district, often named ¢l-LädCdä (lit. “the new one”).11 Most of the ancient cities, being located in relatively inaccessible places, did not undergo significant development (with the rare exception of a city like Atar, where the French army settled in the 1950s). The new towns, on the other hand, experienced significant growth, even when their creation was due to a purely administrative decision. The creation of the capital Nouakchott followed the model of Boutilimit, Kiffa or Aleg, each being a city built from scratch in order to satisfy the political or administrative needs of the central power.12 In the case of Nouakchott, the first stone was formally laid, as recently as 1957, by Mokhtar Ould Daddah, already acting as the future president for the nascent Republic during a ceremony in the presence of representatives of the colonial power. The location was chosen for various reasons, including its proximity to the ocean (to benefit from the sea breeze) and its latitude (far enough north of the Senegal River to give Moorish people the impression of being on their own territory). During the colonial period, it had been the site of a small village (or kXaW, similar in style to the ancient cities) but in 1950 this was almost entirely destroyed by an exceptional flood of the Senegal River. It was then rebuilt on slightly higher ground, to a grid plan, but remained for a long time very far from the city center of Nouakchott. After extended negotiations and a succession of development plans (for details, cf. Pitte 1977), Nouakchott eventually took the dual form characteristic of African neo-colonial towns. A wide avenue traversed the city from east to west, from the mosque, on the road to the Ksar, to the hospital, on the road leading to the ocean, each half fulfilling one of the two functions of the town. The different administrative buildings, the presidential palace, the embassies, and the general headquarters of the notables are located in the north of the town: the kapitA/kabbitAl. The popular neighborhoods with the big market occupy the south: the mädCnä (as in Dakar) – or rather the mädCnas because each block, each mädCnä, sports an identifying letter (mädinä R, mädinä L, . . .). Everything contributes to set each half of the town in contrast to the other: the size of the plots and of the houses, the width of the streets and the existence or absence of tarmac surfaces. The contrast appears even in the official names given to the streets. For the up market districts and streets, the most prestigious names from world history were 43

CATHERINE TAINE-CHEIKH

chosen (according to the criteria of the time): Gamel Abdel Nasser, Kennedy, De Gaulle, Bourguiba, Gandhi, Lumumba, etc. For the others, in the mädinä, names were drawn from Mauritania’s own history, or from that of the subregion: Nasser Eddine, Hennoune O. Bouceif, Samori Touré, etc. However, with a few rare exceptions (such as Gamel Abdel Nasser Avenue and, more recently, Kennedy Avenue and De Gaulle Avenue), these names went unused and are now forgotten. 2.3.1.2 Tarmac roads, 4 × 4’s and donkey carts If the notion of a town center exists today in Nouakchott, then it corresponds to the urban space occupied in the 1960s, but the town subsequently underwent very rapid exponential growth that the state could never have anticipated, or even follow. The first unexpected influx of people, at the end of the 1960s, corresponds to the mass settlement of nomads that several years of drought had reduced to extreme poverty. But the end of the drought cycle, several years later, did not reverse the trend: the city state of Nouakchott continues inexorably to attract more and more people, seeking their share of power, employment, medical care or education for their children. The following figures (from Ould Cheikh, 2006, source: ONS) give an idea of the extraordinary demographic boom seen in Nouakchott over the past forty-five years. Year

1962

1972

1975

1977

1988

2005

Population

5,807

55,000

130,000

134,704

590,532

743,511

Figure 2.1 Population of Nouakchott from 1962 to 2005

The expansion of the town (whose population grew fourfold between 1974 and 1990) mainly followed the major road axes spreading out from the capital (the Rosso Road to the south, the road known as the “Hope Road” to the east and the Atar Road to the north). Four new districts (El Mina literally “the seaport”, Arafat, Riyad – originally Wya1 “graves, cemetery” – and Dar Naim) were created in 1989, supplementing the five muqAYa∞At already existing (Teyaret, Ksar, Tevragh Zeina, Toujounine and Sebkha). “Star formation” developments along the tarmac roads have transformed some crossroads into new town centers. Recent plans mention KaWaffEW MadWCd 13 as Nouakchott’s center, at the junction of the North, East, and South Roads: a center displaced from the prosperous neighborhoods of Tevragh Zeina. The official policy has always been to try to curb as much as possible the arrival of internal migrants, but despite this Nouakchott has continued to 44

THE (R)URBANIZATION OF MAURITANIA

grow, whether through the selling of the land (including the theoretically preserved areas), the legalization of informal settlements (gaZWa)14 or frequent resignations in the face of the perpetual renewal of shanty towns or “shanty encampments” (kebbe).15 Thus the relationship between planners and politicians on the one hand and inhabitants on the other is often rather fraught. Popular unofficial place names humorously reflect these conflicts. They either pun on the way the land was appropriated or make fun of the life style of the local inhabitants. Among the many popular place names are: • • • •

• • • •

MgayZCWa “the little squatted”. KäLL ¢l-}Ak¢m “(under) the Prefect’s nose”, a district which was originally a shanty town located just in front of the Prefect’s offices. Säbx¢t ÄwlAd NAX¢r, referring to the tribe of the 1980–84 Finance Minister who is supposed to have shamelessly favored his kinsmen in the distribution of plots of lands. The Aouzou strip (fr. “la bande d’Aouzou”), a district with well situated plots theoretically reserved for green spaces, which after much competition ended up in the hands of the military. It is an allusion in French to the Aouzou territory, a contested area between Libya and Chad. Las Palmas, one the rich new neighborhoods established in the 1990s, where the inhabitants are seen as wheeler-dealers accustomed to the casino of the Canary Islands. Mälla} m¢n }äyt-ak! “make your meal salty [with salt] of your wall!” a popular district where the walls are seriously damaged by the capillary action of the underground water. Bäll w-äL}aW! “wet it and dig!” – a desperate attempt to build house foundations in dune areas. Nättäg Lämbä! “dig up the skins!” – a necessary preliminary operation for any installation in this district, which adjoins the l-BaYwAW “slaughter houses”.

These ironic place names (toponyms), which express the local discontent towards the ruling urban elite, reflect a rupture in social relationships compared to the Bedouin traditions (Taine-Cheikh 1998). 2.3.2 Arabization of the town’s walls The creation of Nouakchott coincided with the withdrawal of the former colonial power. When Mauritania became independent in 1960, Arabic was recognized as the new state’s national language and French (the former colonial language) as its official language. The newly independent Mauritania aimed to give Arabic the same status as French, making it not only an official administrative language but also, more immediately and concretely, an important language of modern education. The proclaimed parity between 45

CATHERINE TAINE-CHEIKH

Arabic and French was quickly abandoned, for as early as 1980 Arabic was made the only official language. However subsequent policy shifted, according to sectors and periods, between an equal balance between the two languages and a reversal of powers in favor of Arabic.16 If there is one domain, however, where the parity between the two languages has become a reality, it is in road signs and markings of all kinds. At the beginning of the 1990s, almost ten years after the adoption of Arabic as Mauritania’s sole official language, shop signs and advertisement hoardings started to become bilingual French–Arabic. It gave a serious boost to the profession of calligrapher.17 Previously, inscriptions on walls had been only in French and were rather rare, limited to official buildings and to certain premises run by foreigners (essentially French and Lebanese–Syrian) like hotels, import-export businesses or grocery shops. All the small trades were in the hands of Moorish people whose shops did not, and still do not exhibit signs. One of the first shops to ignore this rule was the “Bana Bleu” (Blue Bana), a large, well-stocked shop. It painted on its front wall Bana Ble, an approximate transcription of the name given by its French-speaking customers. The term “bana” thus acquired a positive connotation even though originally derived from bana-bana a hawker, and bAnä-bAnä, which in Hassaniyya refers to shoddy goods. Signs became common at the beginning of the 1990s with the development of the liberal professions (doctors, pharmacists, teachers, etc.). In choosing names, at least for administrative purposes, new offices (in particular pharmacies and clinics), tended to use the propitiatory terms of the Arabic lexicon (abstract names such as “Chiva”, “Ennejah”, “Dawa”, “Elbaraka”, “Emel”, “Al-avia”) or names of famous doctors such as Ibn-Rouchd or Ibn-Sina). Like Mauritanian toponyms or ethnic and personal names, those propitiatory terms, borrowed from classical Arabic, were much easier to write in Arabic script than in Latin and contributed greatly to the arabization of written signs. In the 1990s, which saw a boom in mural calligraphy, the spelling was not perfect. Ten years later, signs, mural calligraphy and hoardings have progressed, in quantity as well as in quality.18 They have changed the face of the city, creating a bilingual written environment in French and (literary) Arabic.19 2.3.3 Language contacts and management of boundaries According to geographers such as J. Levy, urbanity is characterized by the geminate concept “density + diversity” (Louiset 2001: 227–8) and one can wonder how this phenomenon is reflected linguistically. In Mauritania, the homogeneity of Hassaniyya has de facto suppressed one of the main sources of linguistic diversity in the Arab world, i.e. dialectal diversity (which has often played a very important role in the history of Arabic urban vernaculars, cf. Miller 2004). Diversity in the Nouakchott context could thus be 46

THE (R)URBANIZATION OF MAURITANIA

thought to be limited to the concurrent and often redundant use of written French and Arabic. However, the situation is not that simple and urban Hassaniyya speakers are often subjected to various and partially new contacts20 to whom they tend to react by adopting attitudes unknown in the past. Previously, the different languages (mainly Hassaniyya, classical Arabic and Zenaga Berber) were rarely mixed and only borrowings and calques that proved essential for inter-communication were accepted.21 Nowadays the dominant trend is toward the blurring of linguistic as well as stylistic boundaries, through lexical borrowings, language mixing and code-switching. The following are some aspects of this development that tend to affect Hassaniya traditional norms. 2.3.3.1 The evolution of Hassaniyya In Nouakchott, a large part of the Hassaniyya traditional lexicon tends to be no longer used. The younger urban born generations ignore the typical vocabulary of the bAdiyya, the traditional location of Bedouin life. Particular terms such as ä∫äylAl “pendulum (from a pendulum well)” or täynnä “pulley” are totally unknown. Similarly, a large part of the vocabulary relating to plants, livestock, farming and agriculture, and which, until the 1960s was still known by the majority of the population, is today scarcely known. Those who do not move outside the urban space have no occasion to use this lexicon and often the gap deepens between younger and older people, the latter being the only ones keeping such vocabulary alive. Poetry is also being slowly forgotten, and given its links with the Bedouin world, one could not expect the γnä to retain the same appeal for city dwellers as it had for nomads. In the 1970s and 1980s, one talked about the “Hassaniyya of the 50’s” when someone was heard speaking in traditional Hassaniyya with obsolete terms, mostly borrowings from European languages, especially French such as boWYmAlä “coin-purse”, g¢rmwäyti “mixed blood”, gläysCs “spinning”, tilbinAW “tribunal” or tAγ¢Y ¢s-s¢kkä “bank note”. Nowadays, the “Hassaniyya of the 70’s” refers to the language used by someone who speaks Hassaniyya without any borrowings, thus showing a good knowledge of the expressive resources of the language. Previously, such knowledge was admired and the proficient speaker was known as udnu bay1a “his ear is white” or äbäy1An m¢n tidinCt w¢ll MAnu “more Moorish than Manu’s lute”. Certain lexical fields, those closely related to the traditional life style and values, disappeared while others were introduced as a result of the profound changes that occurred in all spheres of life – economic, social, political and cultural. But the evolution of language does not reflect communicative needs alone; power and ideology also play a role. The urban lexicon had evolved and integrated dialectal neologisms (cf. käbbä and gaZWa mentioned above), 47

CATHERINE TAINE-CHEIKH

as well as French and Modern Standard Arabic (MSA) borrowings, with a trend toward the increasing use of the latter. 2.3.3.2 Middle Arabic as a “modernization” of the dialect A new usage of Arabic developed during the period 1968–75 under the influence of Arab media and the increasing contacts with other Arab countries (cf. students going abroad and the opening of Arab cultural centers in the capital). This new usage was clearly associated with political movements. Middle Arabic (∞aWabCya l-wuXYa lit. “Arabic of the middle”) was first introduced among the intellectuals who had received a traditional education, in particular those influenced by pan-Arabism. It then spread among a number of young urban students. Indeed, whereas the numerous French speaking educated people of Nouakchott regularly used French in their discussions, the Arabic speakers – often unfamiliar with the oral use of MSA – did not have an adequate means of expression. Unlike Middle-Eastern countries, Mauritania had retained a very traditional Arabic education system, based on the memorization of the Quran and of a great part of the classical Islamic corpus. This left very limited room for free expression, particularly at the oral level, where the scrupulous respect of case and modal markers made the practice of Standard Classical Arabic (SCA) very difficult. The great novelty then was to follow an intermediary path between colloquial and SCA, by dropping those grammatical complexities which did not serve the purposes of communication. Mauritanian Middle Arabic (MMA) corresponds only partially to what English speaking authors call Educated Spoken Arabic (ESA), because, as the dialectal base is fundamentally homogeneous, there is no need to select between the different colloquial realizations. This selection would involve either “purifying” the lowest variety through the removal of its stigmatized features (cf. the affricate realization of a phoneme, the discontinuous variant of the negation and so on, which are associated with the non-prestigious dialect of a local community, often of rural origin), or to abandon the local variety features in favor of koine features, even if the latter are more deviant from the classical norm than the local features (cf. the [™] realization of the qâf instead of [q] felt locally as typically Druze in Palva (1982)). MMA can thus appear as a semi-literary variety, fairly comparable to the literary classicized colloquial variety studied by Palva (1969), even though MMA probably remains closer to the “purely” dialectal base than most of the production considered as ESA, due to the absence of purification tendencies. The transformations observed in MMA are mainly linked to the introduction of words borrowed from MSA. These borrowings respond to new needs of communication and, with a few exceptions, are not inspired by a desire to raise the style level of the discourse through the use of classical Arabic expressions. Borrowings are generally treated according to the morphological 48

THE (R)URBANIZATION OF MAURITANIA

and syntactic rules of Hassaniyya, thus creating a certain systemic discontinuity between Middle Arabic and Standard Arabic. Examples include: the elimination of cases and modes, a very limited use of the TMA (Tense– Modal–Aspect) particles qad and sa-, the absence of dual forms (except for names), a tendency to use the colloquial verbal conjugation such as n- first singular imperfective form. But borrowed terms retain their schematic patterns (some are quite rare in Hassaniyya) and their morphology (especially plural nominal forms). At the phonological level they are only partially integrated into Hassaniyya; MMA largely respects the phonological characteristics of classical Arabic: the |Ad / 1A’ distinction, the voiceless realization [q] of qAf, the regular retention of short vowels in open syllables (even though the quality of the vowel sometimes changes) and partial retention of the hamza. Generally speaking, the phonetic and phonological treatment of the borrowings corresponds to the traditional usage of educated Hassaniyya people, when quoting classical Arabic terms in colloquial speech. However, the use of Standard Arabic is so frequent that the dia-systemic aspects have become the main characteristics of MMA. This tendency is further accentuated by the fact that, in contact with a borrowed noun, the speaker tends to use a more standardized form, even when the dialectal form exists: thus, käbCr instead of the dialectal kbCW (what I have called “contiguous arabization”, Taine-Cheikh 1978). One may note that while “pure” colloquial forms are frequent and belong to all syntactic classes, “pure” literary forms, on the contrary, belong to specific categories (adverbs, particles, conjunctions, etc.), which escape morphological variation but often contribute to a complexification of the syntactic structure. Indeed, verbs and verbal nouns are usually borrowed along with their own particular inflectional regime. In practice, the use of Middle Arabic is rather limited compared to the use of Hassaniyya, but more frequent than Standard Arabic.22 Nevertheless, MMA can be considered a “mixed dialect” in the sense of Trudgill (1986: 59). It clearly tends, in certain circumstances, to constitute a third code alongside Classical Standard Arabic and Hassaniyya. This was the case, for instance, with the popular preacher who spoke all three codes perfectly and knew how to alternate them appropriately during his sermons (Taine-Cheikh 2002). 2.3.3.3 Contacts, mixity and globalization Despite the spread of Middle Arabic, a great number of the Arabic speaking inhabitants of Nouakchott (especially the }WAYCn) speak only one code perfectly: Hassaniyya. However, they are in contact (in town rather than elsewhere) with other languages, either oral or written, such as French (and, more recently, English), or quasi written such as classical Arabic, or (almost exclusively) oral such as Pulaar, Soninke and Wolof, as well as other Arabic dialects heard in the streets or on television. 49

CATHERINE TAINE-CHEIKH

Compared to “bush” Hassaniyya, urban Hassaniyya (particularly in Nouakchott) is essentially characterized by a greater permeability to borrowings, (in addition to the previously mentioned loss of the “rural” lexicon, which had many old Zenaga loanwords). Borrowings are mainly initiated and reinforced in towns, even though they do sometimes penetrate the Hassaniyya of the “bush”. Some old French borrowings are still used, as YAblä “table” or bombotBW “pomme de terre”/“potato” whereas others, such as minyistr “minister”/“minister” or Wadyo “radio”, have more or less disappeared, having been replaced by their classical Arabic equivalents wäzCr and iOA∞a. Other more recent Arabic borrowings betray the influence of Arab media, e.g. bass? “That’s all?” which has penetrated the vocabulary of some traders. Indeed, the frequency of the borrowings in towns is related not only to the presence of the administration and the modern economic (industrial and commercial) sectors, but also to popular markets and to the numerous activities of the informal sector (cf. the activity of michelin which consists in re-inflating tires). Generally speaking, we note that French borrowings are tending to decrease while, as already mentioned, Standard Arabic borrowings are on the increase and borrowings from local African languages remain extremely limited. In parallel, the French spoken in Mauritania (Ould Zein and Queffélec 1997) has been enriched by numerous terms directly borrowed from the national languages: melhafa (women’s veil), ogla (shallow well) and kebbe (shanty town) borrowed from Hassaniyya, gordiguene (gay) and bana-bana (poor quality, itinerant petty trader) borrowed from Wolof, niebe (variety of bean) or walo (fertile lands along the Senegal River) borrowed from Pulaar (Fulani). It is also hospitable to neologisms such as enceinter “to impregnate, i.e. to make someone pregnant” and cabiner “to toilet, i.e. to go to the toilet,” and to semantic shifts such as cow boy “a crafty one, bandit” and écritoire “writer, i.e. writing implement: pencil, pen”. These are fairly common phenomena but it is nonetheless interesting to note that often the same terms, such as gordiguene, niebe and cow boy are also used in Hassaniyya, thus bringing closer the lexicons of the two languages. This is an indicator of a phenomenon which, while difficult to quantify, is representative of an increasing trend for languages to “communicate” with each other. Hitherto, the practice of code-mixing had been restricted to less educated people who had the impression that by resorting to numerous French borrowings they gained in expressivity and modernity. But it seems now to have spread and to have touched even new generations of intellectuals, who do not hesitate to code-switch freely between French and Arabic, even in mid-sentence. Borrowings and code-switching tend to function as signs of claims to identity. While borrowing responds first to the need to name an object, activity or behavior, it very often carries a positive or negative connotation which it acquires by reference to the speakers’ speech community. The need 50

THE (R)URBANIZATION OF MAURITANIA

for modernity and identity is particularly developed within young people’s speech communities, and among them, those most open to the outside world or the most marginal groups. It explains why the relatively less spoken languages in Mauritania such as Wolof and English are the ones which nonetheless supply a number of words to Mauritanian French and to Hassaniyya, cf. tnappi “to rip somebody off ” which was borrowed from Wolof and whose usage is no longer confined to urban delinquents. From this point of view, among the different urban centers which emerged from the period of colonial administration, the little town of Rosso, located on the Senegal River, (and socially very mixed) is often up with the latest fashion – at the crossroad of Mauritanian and Senegalese fashions.23 Increasing processes of borrowing and diversification of the source languages have contributed to the “colorful” aspect of urban Hassaniyya. Moreover, code-switching and code-mixing constitute an even clearer break with former linguistic practices. These developments, together with the emergence of Middle Arabic, are the first linguistic results of the upheaval brought about by the multiplication of contacts with the outside world in an urban environment, despite the maintenance of ethnic and tribal affiliations both within and beyond the city. The effects of globalization have sometimes been described as “glocalization”, in the sense that they produce local from global, and global from local, in a simultaneous process of heterogenization–homogenization. It seems to me that, with regard to linguistic practices, “glocalization” manifests itself first in the increasing diffusion of Arabic and French in society (i.e. the globalization–homogenization facet) and, second, in the development of phenomena of diversification and pluralism – borrowing, code-switching, code-mixing – (i.e. the localization–heterogenization facet). While the phenomenon is not limited to Mauritania, it is characteristic of at least the recent history of the country and contemporaneous with its recent urbanization.

Notes 1 Linguistic differentiation can be observed at the margins of Moorish society, either through the more recent arabization process or through the maintenance of Berber–Arabic bilingualism. 2 The preferred Moorish marriage pattern is with a paternal uncle’s daughter (mint ¢l-∞amm). According to the patrilocality rule, the wife will join her husband’s camp. Even when married to a cousin, she can be taken far away from her usual nomadization area. 3 There is a surprising alliance – which may a priori seem against nature but which relates ultimately to the concept of honor – between the anarchic individualism of nomads and the mutual control (or self-control) of individuals by each other. The control exercised by Moorish society with regard to personal injury gives some idea of the control exerted on the language (cf. Taine-Cheikh 2004b).

51

CATHERINE TAINE-CHEIKH

4 The acceptability of the changes would especially be dependent on the capacity to preserve synthetic constructions and significant semantic oppositions (voice, nominal degrees, comparison, etc.) as well as phonological opposition (where only the dropping of the hamza and the assimilation of the | to 1 are regularly admitted). 5 These tendencies have their origins partially in the phonetics of the substrate language. It may be noted that for certain words, emphatization/de-emphatization occurred among all speakers (X > s in sγ Cr “small” and s > X in XaYla “iron container”). 6 There were moreover itinerant ma}A1¢W, housed under canvas – as well as camp libraries that were transported on camel back. 7 In Ouadane it caused an important tribal conflict (in the 1820s), which was the origin of the construction of a second mosque and the gradual abandonment of the old one. 8 Independent because each one was large enough to accommodate an extended family, understood as a small camp, and interdependent because families, especially those from the same tribe, were supposed to help each other in case of danger. 9 The presence of another vernacular language alongside Berber may have accelerated the disappearance of Zenaga, which was dominant in the ancient towns, probably at least until the sixteenth century. 10 Cf. the case of Baghdad where the Muslim dialect is classified as a former Bedouin dialect, contrary to Christian and Jewish dialects (cf. Blanc 1964). 11 For more details of the names of districts in small towns, cf. Taine-Cheikh 1998: 79–83. 12 Only Zouerate and Akjoujt, other recent cities (and the Cansado district of Nouadhibou) were created because of the implantation of the iron and copper mining industries. 13 The construction of this round-about coincided with the visit of the King of Spain to Mauritania. 14 The term means “squat” and is a neologism derived from the Hassaniyya root GqR, from which is derived gaZZaW “to buck”. 15 Lit. “Rubbish dump(s).” The word, borrowed from Hassaniyya, käbbä, is the nominal derived from käbb “to dump”. 16 On the linguistic policy of Mauritania and its consequences, cf. Taine-Cheikh 2004a. 17 Until then the profession of calligraphist concerned the copying of Arabic manuscripts – a profession clearly in decline at the end of the twentieth century. 18 It is also true for French – the shop sign of “Bana bleu” has indeed added the u of “bleu.” 19 Certain districts however still exhibit the predominant use of Latin script, especially in the district of Sebkha were black Africans are numerous. 20 Spatial mixing is however very limited, first because of ethnicism and the “communalization of space” (Choplin 2003: 71–73), second because of the separation between richer neighborhoods (where inhabitants are in the majority “white” Moorish) and poorer neighborhoods (where the majority of the inhabitants are black or, almost exclusively, “black” Moors). 21 Linguistic mixing has always existed in North Africa (cf. Dakhlia 2004), but the degree of mixing has been far from uniform through time and space. 22 It can be reduced to the use of qaY∞an “certainly” and Yab∞an “of course” (expressions on their way to becoming verbal tics). 23 In spite of a dominant Moorish and Hal Pulaaren population, Wolof is the main lingua franca (Fall 1996–97).

52

THE (R)URBANIZATION OF MAURITANIA

References Beaussier, M. (1958) Dictionnaire pratique arabe-français, Alger: La Maison des Livres. Blanc, H. (1964) Communal Dialects in Baghdad, Middle Eastern Monograph Series X, Cambridge, MA: Harvard University Press. Bonte, P. (1998) L’Emirat de l’Adrar. Histoire et anthropologie d’une société tribale du Sahara Occidental, Thèse de Doctorat d’Etat, Paris, EHESS. Choplin, A. (2003) Rapports ville-Etat. Le cas de Nouakchott (Mauritanie), Paris: Grafigéo. Colin, G. S. (1993) Le Dictionnaire COLIN d’Arabe Dialectal Marocain, Z. IraquiSinaceur (dir.), Rabat–Paris: Al Manahil. Dakhlia, J. (ed.) (2004) Trames de langues: usages et métissages linguistiques dans l’histoire du Maghreb, Paris: Maisonneuve & Larose. Fall, B. (1996–97) “Le langage des jeunes de Rosso”, unpublished MA thesis, University of Nouakchott. Ingham, B. (1979) “Notes on the dialect of the Muyair of Eastern Arabia”, ZAL 2: 23–35. Louiset, O. (2001) “Les villes invisibles”, L’information géographique 65, 3: 219–233. Miller, C. (1996) “Nubien, berbère et beja: notes sur trois langues vernaculaires non arabes de l’Egypte contemporaine”, Egypte/Monde arabe 27–28: 161–70. Miller, C. (2004) “Variation and change in Arabic urban vernaculars”, in M. Haak, R. de Jong and K. Versteegh (eds), Approaches to Arabic Dialects: Collection of Articles Presented to Manfred Woidich on the Occasion of his Sixtieth Birthday, Leiden–Boston: Brill, pp. 177–206. Ould Cheikh, A. W. (1985) “Nomadisme, Islam et pouvoir politique dans la société maure précoloniale (XIe-XIXe siècles). Essai sur quelques aspects du tribalisme”, unpublished thesis, University of Paris V. Ould Cheikh, A. W. (2006) “Nouakchott, capitale nomade?”, in Nouakchott capitale de la Mauritanie: 50 ans de défi, Saint Maur des Fossés: Sépia, pp. 139–148. Ould Zein, B. and Queffélec, A. (1997) Le français en Mauritanie, Vanves: EDICEF/ AUPELF. Palva, H. (1969) “Notes on classicisation in Modern Colloquial Arabic”, Studia Orientalia (Helsinki) XL, 3: 3–41. Palva, H. (1982) “Patterns of koineisation in modern colloquial Arabic”, Acta Orientalia 43: 13–32. Pitte, J.-R. (1977) Nouakchott, capitale de la Mauritanie, Paris: Publications du Département de Géographie de Paris-Sorbonne. Taine-Cheikh, C. (1978) “L’arabe médian parlé par les arabophones de Mauritanie – Etude morpho-syntaxique”, unpublished thesis, University of Paris V – René Descartes. Taine-Cheikh, C. (1989) “La Mauritanie en noir et blanc – petite promenade linguistique en }assAniyya”, Revue des Etudes sur le Monde Méditerranéen et Musulman 54, 4: 90–105. Taine-Cheikh, C. (1991) “L’arabe des BC1An, un dialecte bédouin de Maghreb occidental”, in A. S. Kaye (ed.) Semitic Studies (In Honor of Wolf Leslau: On the Occasion of his Eighty-fifth Birthday November 14th, 1991), vol. 2, Wiesbaden: O. Harrassowitz, pp. 1528–1548.

53

CATHERINE TAINE-CHEIKH

Taine-Cheikh, C. (1998) “Toponymie et urbanization”, Espaces et sociétés en Mauritanie: fascicule de recherches 33, Tours: Urbama, pp. 77–86. Taine-Cheikh, C. (2002) “De la variation linguistique dans le prêche populaire mauritanien”, in A. Rouchdy (ed.), Language Contact and Language Conflict in Arabic, London: RoutledgeCurzon, pp. 177–203. Taine-Cheikh, C. (2004a) “La Mauritanie: vers une nouvelle politique linguistique?”, Aménagement linguistique au Maghreb: revue d’aménagement linguistique, 107: 205–226. Taine-Cheikh, C. (2004b) “L’injure en pays maure ou ‘qui ne loue pas critique’ ”, REMMM 103–104: 83–106. Trudgill, P. (1986) Dialects in Contact, London: Blackwell.

54

THE FORMATION OF THE DIALECT OF AMMAN

3 THE FORMATION OF THE DIALECT OF AMMAN From chaos to order Enam Al-Wer

3.1 Introduction There has been a surge of interest in sociolinguistic research which focuses on the outcome of dialect contact, especially since the publication of Peter Trudgill’s book Dialects in Contact (Trudgill 1986). A wide range of situations have been documented, such as the dialects of new towns (Kerswill and Williams 2000), dialect boundary regions (Britain 1991), immigrant groups in the city (Kerswill 1994), and the two most recent publications on the formation of New Zealand English (Gordon et al. 2004; Trudgill 2004). The Amman project, on which the present chapter is based, is an investigation of the outcome of contact between Jordanian and Palestinian dialects, which are the dialects spoken by the first generations of the vast majority of the city’s population. The analysis of the Amman data is based on the theoretical principles of variation theory, and benefits directly from the accumulation of knowledge in dialect contact research. In contact situations, various linguistic developments can be expected. In some cases, linguistic variants which are characteristic of one of the dialects in the original mix can spread at the expense of other features in the input dialects. In other cases, there can be a fusion of features, resulting in the emergence of intermediate and fudged linguistic forms. In a third category of possibilities, totally new features can emerge, which were neither present in the original mix, nor can be called intermediate features. All of these possible outcomes of dialect contact are exemplified in the data from Amman, as will be shown presently. One of the issues which researchers in the field of dialect contact have had to tackle is how to ascertain that the emergent features are indeed contactinduced, rather than linguistic developments which the dialect in question would have undergone without coming in contact with other dialects. In dealing with the Amman data, this issue is immaterial because in this situation 55

ENAM AL-WER

we are not dealing with a continuation of change in a dialect, but with the formation of a dialect from scratch. Amman had no dialect simply because it did not have a native and stable population. The adult population of Amman, including many who were born in the city, continue to affiliate themselves with the towns and villages of their forefathers. The youngsters, on the other hand, call themselves ∞ammaniyyCn “Ammanis”, by which they mean that they are native to the city. In the region’s dialects the terms shwAm “Damascenes”, ma§Adse “Jerusalemites”, iribdAwiyye “Irbidis, (from the city of Irbid)”, etc. are well integrated terms, but the derivation ¡ammaniyyCn “Ammanis” did not exist. The coinage of the term indicates that the city is acquiring a native population and a regional identity on a par with the established cities in the region, which coincides with the formation of a distinctive dialect. For the “makers” of the Amman dialect, there is no linguistic metropolis to copy within Jordan. Rather, they are constructing a dialect which itself is to become the linguistic metropolis of Jordan. The linguistic and sociolinguistic aspects of this process are compelling. I will start with a general overview of the relevant background information about dialect geography, population growth, and the sociolinguistic situation, focusing on the situation in Amman, and specifically, on the Jordanian–Palestinian relationships as a priori to understanding the sociolinguistic determinants of variation and change in the Jordanian context.

3.2 The dialect geography of Jordan In order to understand the geo-linguistics of contemporary Jordanian dialects, it is helpful to delineate its borders since they were first drafted in 1921. The area which is now Jordan had, until 1921, a geographical (not a political) identity. Ancient travelers referred to it as MashAref ish-ShAm “the approaches to Sham (the northern part of the Peninsula)”. During the Ottoman rule, it was part of the Vilayet of Damascus, and was referred to as sharq il-§urdun “Transjordan, east of the River Jordan”. Shaped like an Arabic coffee pot without a spout (“dalle”), it cuts across the Houran plain and the Badiyat ish-Sham in the north, isolating a sizable part, which includes hundreds of towns and villages, from the Houran heartland in southern Syria. The southern border incorporates fringes from the Hijaz, which include the cities of Aqaba and Ma’an, while the eastern border stretches over some 400 km across the Badiya in a sparsely populated area. The western border ends in the Jordan Valley; it follows the course of the Jordan River from Lake Tiberias to the Dead Sea, and further along the Wadi Araba to the Gulf of Aqaba at the Red Sea. Where the western border was drawn has altered the historical socioeconomic ties between the major urban settlements on the east side of the river and Palestinian cities. It is well known that in this part of Bilad ish-Sham, trade relations had run in an 56

THE FORMATION OF THE DIALECT OF AMMAN

Map 3.1 Jordan

east–west fashion (rather than north–south), as can be witnessed, for instance, by the close familial relations between the tribes of Sult and Nablus, and the tribes of Kerak and al-Khalil (Hebron). Thus, the western border laid the foundations for a new type of integration: a north–south, instead of an east–west. Within these borders, there are dialects which are akin to two major groups: the eastern and far southern dialects are of Arabian type (and have fewer speakers), and the northern and central dialects are of southern Levantine type (see also Cleveland 1963). The Levantine dialects can be subdivided into Hourani in the far north (the dialect of the city of Ajloun is a 57

ENAM AL-WER

good representative), and Balgawi in the northwest and central areas (the dialect of the city of Sult is an example). This leaves a final major urban settlement, that of the city of Kerak and its surroundings (140 km south of Sult), whose traditional dialect diverges from the central and northern dialects in showing affinities with Bedouin dialects further south, and is considered by Palva (1989: 242) as “a dialect of transition between the Syrian area and the Hijaz”. The political re-organization of the Levant after the Great War, and the subsequent political upheavals in the region have favored the emergence of different national norms, as opposed to a pan-Levantine norm based on the dialect of Damascus.

3.3 Population growth, distribution and composition The population of Jordan increased by nearly two million people between 1947 and 1986 (from 500,000 to 2.3 million people), and by more than three million between 1987 and 2005 and it is currently estimated to be 5.7 million people. Sixty-two percent of the population live in the central region (Amman (39 percent), Balga, Zarqa), which comprises only 16 percent of the total area of the country, while the southern region, which comprises 51 percent of the total area, accounts for only 9 percent of the total population. The exceptional increase in the population is due to the war situation in the region, and the displacement of populations. Nearly 1.5 million Palestinians sought refuge in Jordan after 1948 and 1967, and a further 350,000 from Kuwait in 1990. It is obvious from the figures presented above that the vast majority of the displaced population from Palestine settled in the central region. Amman had 5,000 Circassian settlers (originally monolingual in Circassian) in 1906. In 1930, an additional 5,000 people moved into the City, and between 1950 and 1990 the population increased more than fifteen times, to reach 1.6 million in 2003, i.e. nearly one third of the country’s population.

3.4 Internal politics and sociolinguistic ramifications Jordan was the first country in the Levant to have an Arab central government, and, in its formative years, was a refuge for Arab nationalists from Greater Syria in general. It was also the poorest in natural and human resources. The local population looked to Haifa, Jerusalem, Damascus and Beirut as the region’s cultural centers. Until the 1960s, the country had no institutions of higher education. Those who went to universities were educated either outside the Arab world, or in Lebanese, Syrian, and Egyptian universities. In the early decades of the establishment of the state, the governing body and the civil servants were almost exclusively of Syrian or Palestinian origins, which did not go unnoticed by the local population, who from time to 58

THE FORMATION OF THE DIALECT OF AMMAN

time voiced resentments, firstly targeting the Syrians, and later, as the Syrians were numerically overwhelmed by the Palestinians, some nationalists demanded the “Jordanization” of the administration (in Arabic §ardanat iljihAz il-idArC). Political analysts view this demand as the first expression of a Jordanian political identity which was distinct from a pan Syrian identity. However, it would be difficult to understand how the local population were suddenly united and mobilized to express a political identity, or to understand the ramifications of this event on the sociolinguistic situation (see below), without recognizing that a distinctive social identity and social character had already been in existence at least since the mid-nineteenth century when the region was under the rule of the Ottomans (see, for instance, Dawood 1994). Historically, the dominant dialects in the region are the “city” dialects. The reference to “city” in the context of Jordan during the nineteenth and early twentieth centuries meant a settled community, with an economic base that did not only rely on semi-subsistence economy, but also included trade, simple industries, as well as farming. Thus, for instance the dialect of the old city of Sult, the largest in Jordan until the 1930s, was in fact considered madanC “urban” by the local population, despite the fact that all of the (segmental) phonological features of this dialect are “Bedouin” in dialectological terms (see Palva 1994). Where koineization in the central and northern dialects occurs, it is in the direction of features which are characteristic of the traditional dialect of Sult (see Al-Wer 1991). The resurrection of Amman, the ancient site of Rabbath Ammon, and Roman Philadelphia, as the capital of the newly formed Emirate of Transjordan in 1923, deprived Sult of its prominent historical status, and set a new context for future linguistic developments. It is not unreasonable to suggest that had Sult become the capital city, its dialect would have dictated the rubrics of the metropolitan dialect of Jordan. Instead, and over the years, what has happened is that Sult and its dialect were refashioned to become symbols of whatever is “authentic” Jordanian. This role became all the more convincing since Amman had no local population of its own, and since a Jordanian nationalist movement required a territorial base with all the necessary local Jordanian character. The location of Sult at a close proximity to Amman (25 km) enhanced the credibility of its new role.1 After it was chosen as the location of the capital city, Amman became the center of political and economic activity, but without the cultural artefacts. Many of the city’s elite families had come from Damascus, Beirut, Jerusalem, Haifa, and Nablus. These cities already had institutions of higher education when the whole of Jordan had only one secondary school – in Sult. They spoke dialects which were associated with attractive places, and attractive lifestyles. Most of the elite families from the Jordanian side had come from Sult, which, despite enjoying prominent status locally, had no regional significance. However, both sectors of the Amman population stood equal 59

ENAM AL-WER

in relation to their new home, in that neither could claim territorial legitimacy, but while the Syrians and the Palestinians were already trained to take over businesses and public-sector jobs, only a small minority of the Jordanians were equipped in these domains. The Jordanian dialects which were spoken by the early migrants in the city share many linguistic features with Syrian and Palestinian dialects generally, but they are especially more similar phonologically to the village, rather than the city dialects in the rest of the region (see Palva 1970, 1994, 2003 on Balga dialects; Palva 1989 on Kerak dialect). Given the situation outlined here, it is not surprising that it was the Jordanians who initially converged to the linguistic habits of urban Palestinians and Syrians. The Jordanians were not only motivated in this direction, but also had access to the target linguistic features, through direct contact, and frequent exposure, to urban Palestinian and Syrian dialects. The salient urban Levantine linguistic features, therefore, were advanced in Amman in the early stage. The flood of immigrants from Palestinian cities and villages into Amman intensified after 1948 and 1967. The vast majority of the immigrants during this period came from rural locations, which contrasts sharply with the Palestinian presence in Amman in earlier years. The earlier migrants were either of the intellectual and the political elites, or the bourgeoisie, and they, rather than the later Palestinians, were the sector of the Palestinian population who had the most political, economic and social influence on the Jordanian society. It is noticeable, for instance, that despite the fact that the rural Palestinian dialects have a relatively large linguistic representation in Amman (and larger than the urban representation), none of their linguistic features play any role in the formation of the new dialect. In this context, it is worth highlighting a crucial difference between the Jordanian and the Palestinian societies, despite the two societies being very similar culturally. The Palestinian society is sharply stratified in terms of urban (townsfolk) and rural (peasants), which explains the persistence of the sharp linguistic distinctions between town dialect and village dialect, even where the villages are in close proximity to the town. Jordanian society, on the other hand, is tribal, a much more egalitarian system which cuts across the social distinctions between town and village.2 The Palestinian system of social organization was simply transplanted in Amman. Even though the townsfolk and the villagers have become citizens of the same locality, they continue to form two sharply distinctive communities. During the late 1960s and 1970s, the modernization process assumed the dimension of a major sociopolitical transformation, which affected the respective roles of the Jordanians and the Palestinians in general, and their relations in the urban context in particular. This change can be described as a strategy on the part of the state to include the indigenous population in the modernization of Jordan. The September 1970 civil war between the Jordanian army and the Palestinian commando movement, which threat60

THE FORMATION OF THE DIALECT OF AMMAN

ened, or was seen to threaten, the existence of Jordan as a state, made it obvious that the viability of the social and political identity of Jordan depended on the inclusion of the indigenous group in development and economic growth. Four years later, when Jordan recognized the PLO as the sole representative of the Palestinian people, the Jordanian nationalist movement reiterated its demands for a larger share of appointments in the state bureaucracy, and this time, on the grounds that “if the PLO represents the Palestinians, Jordan should be a state for the Jordanians”. During the decade following 1974 the Jordanians increasingly occupied a greater proportion of the posts in the bureaucracy and services. This process enhanced different feelings of identity, and may have exacerbated the perception of a Palestinian threat. An indigenous Jordanian population developed and proceeded to reinstate its own cultural identity in the urban context. Subsequently, this group also assumed a firmer and more secure social and political status, which resulted in a general awareness of a distinct Jordanian identity, and adherence to local norms of social behavior. Above, I made a distinction between the early Palestinian Jordanians and the later Palestinian populations (who also became Jordanian citizens) in terms of their social backgrounds. This distinction is important to understand the analysis of the sociolinguistic developments during the 1970s and 1980s. Members of some of the early Palestinian families, e.g. Touqan and Nabulsi (both originally from Nablus) had in fact been living on the east side of the river (in Sult) for decades prior to the establishment of the state. Others, most notably the Nashashibis of Jerusalem and the Rifa’is of Nablus, had formed a close alliance with the ruling family since the early decades of the twentieth century. If the appointment to high ranking positions during the 1970s and 1980s was a deliberate strategy to give a larger share to indigenous Jordanians, the early Palestinians and the traditional supporters of the state from Palestine were counted among this sector of the population.3 The increase in the number of local Jordanians who came to assume considerable power in politics and in the army, and the fact that Jordan as a country in the region assumed a firmer and securer position led to a crucial redefinition of the social meanings of the use of the localized linguistic features. They became associated with political influence and power, over and above associations with a Jordanian, as opposed to a Palestinian, identity.4 It is noticeable that urban Palestinian men in politically dominant positions also used features of Jordanian dialects in public. Women were largely excluded from the modernization process, which can be confirmed by their very modest contribution to the labor force during that period (roughly 7.7 percent rising to 12.5 percent in the 1980s). No women were appointed to senior management, political, or planning and decision making posts. Women were, thus, also excluded from the domains which called for the use of the local Jordanian linguistic features, or the domains in which the use of the local features legitimized an 61

ENAM AL-WER

individual’s position. The arenas in which women were most involved were largely restricted to the financial sector, education, and the social services. These arenas were (and still are) the domains in which a pan-Levantine urban linguistic norm is required (cf. the notion of the “linguistic market”, Eckert 2000). The onset of the sharp contrast between male preference for a Jordanian accent and female preference for an urban accent, which has been so often reported (see for instance Abdel-Jawad 1981), can be explained with reference to the linguistic requirements (as part of the symbolic necessities) of the domains in which individuals are involved (e.g. their jobs). Put crudely, men were appointed to positions which required the use of the local dialects, and women were excluded from these positions. Over time, the association of these positions with the use of local features became an association between local linguistic features and the individuals who assumed the positions. This line of explanation has a number of benefits. First, it is functional, as it correlates linguistic variation with the social activities of the speakers. It, therefore, provides a simple explanation of the empirical data which show that the speakers’ (male and female) linguistic usages in Amman do not always correlate with their dialectal heritage. Second, it is based on concrete and well-defined criteria, rather than on the abstract and ill-defined notions of “prestige” and “stigma”. Third, it is realistic and flexible since it integrates linguistic findings with sophisticated understandings of the social processes. To sum up, in this section, my aim has been to explain, in broad terms, the social factors which influence the sociolinguistic situation through analyzing the major sociohistorical events. In the course of this chapter, it will become clear that linguistic factors are equally important in the interpretation of the data from Amman.

3.5 The research The data come from ongoing research which I call the Amman project. They were obtained through sociolinguistic interviews, incorporating the latest methodological techniques, and utilizing up to date technological gadgets. So far, the pool of data consists of 25 hours of recorded material. The youngest speaker is 12 years old and the oldest speaker is 78 years old. The first stage of the research provided samples across generation groups of two of the main input dialects: Sult from the Jordanian side, and Nablus from the Palestinian side. The speakers represent three generations (grandparents, parents and their children) of four families (two originally from Sult, and two originally from Nablus). In this sample, there are 8 speakers from the grandparents’ generation, 8 parents, and 14 youngsters (6 male and 8 female). The second stage focuses on the third generation. So far I have 20 speakers in this group, 11 female and 9 male. In order to have comparable data from younger speakers outside Amman, I interviewed an 62

THE FORMATION OF THE DIALECT OF AMMAN

additional 6 speakers (17–20 years), 3 male and 3 female residents of Sult. I also included 2 male speakers from Nablus who were visiting relatives in Amman when the research was being conducted, but who normally live in Nablus. All of the speakers from Amman live in West Amman. The sample of speakers is systematic and, therefore, the data permit comparisons across the speakers; however, the sample is as yet too small to permit statistical testing. It is also too small to warrant generalizations beyond a specific area in the city (the wealthier and more cosmopolitan West Amman), and beyond a specific social group (the upper-middle and upper classes). However, I have good reasons to believe that it is in this part of the city that the dialect is undergoing focusing, and that this particular social group is leading the process. A crucial difference between the more traditional East Amman and West Amman is in the socialization pattern of the youth. In East Amman, the youngsters spend considerably more time with their own families, and extended families often live in the same neighborhoods. On the other hand, in West Amman, the youngsters form intimate peer group relations, and spend most of their leisure time away from their homes and families. In addition, the leisure facilities in West Amman are much more widely available for the youth to spend most of their free time together. In other terms, the familial networks in East Amman are closer, and, therefore, linguistic innovations (divergence from the traditional dialects) would not be expected to permeate such tightly-knit social networks easily (cf. Milroy 1980). In this chapter, I present analysis of the linguistic features which show a fairly advanced stage of focusing, judged by the available corpus of data.

3.6 The data The overall picture across the three generation of speakers in Amman is a textbook case of new dialect formation, which can be summed up in the following way. First generation speakers, who arrived in the city as adults, speak dialects which can be easily identified with the original Jordanian or Palestinian towns from which they migrated. However, the dialects spoken by this generation do show some divergence from the original dialects. The first development involves the leveling out of the most localized, and most marked features, a process which Trudgill calls rudimentary leveling (Trudgill 2004: 89–93). The most obvious feature which undergoes leveling at this stage from the Jordanian side (Sult) is /k/ affrication, which is totally absent in the data from first generation male and female Jordanians. The traditional dialect of Sult affricates [k] conditionally (mostly in the vicinity of front vowels, e.g. [t∫ε:f ] > [kε:f ] “how”) (see Palva 1994). In my corpus of data from Sult (collected in 1987), the oldest and least mobile speakers affricated /k/ consistently in the feminine suffix /ik/ and occasionally in word roots where /k/ preceded front /a/, and /ε/. Some tokens of affrication 63

ENAM AL-WER

also occurred in my supplementary data (collected in 1997) from the male youngsters in Sult, which indicates that although the feature is receding in the original dialect itself, it has not disappeared altogether.5 Judging by the large corpus of data I have from Sult (see Al-Wer 1991), /k/ affrication may have been already undergoing change in the traditional dialect when migration took place. We can, therefore, consider the absence of affrication in the speech of this generation in Amman as a continuation of this process, which was accelerated in the contact situation. Another example of rudimentary leveling from the Jordanian data, but which does not appear to have started in the home town, and was used variably by this generation is gender neutralizations in the verb endings and the pronominal suffixes. From the Palestinian (Nablus) side, the best example of rudimentary leveling is lowering of [ε:] or [e] to [æ:] in some lexical items, e.g. [¡ammε:n] > [¡ammæ:n] “Amman”, [mbe:ri]] > [mbæ:ri]] “yesterday”, [se:¡a] > [sa:¡a] “hour or watch”. Extremely raised variants of /a/ in these examples are generally stereotypical features of Palestinian dialects (most notable in the dialect of Jerusalem), and the tokens with the lowered vowels in the speech of this generation of Palestinian speakers can be considered a convergence or accommodation to Jordanian dialects. As will be explained below, lowering of /a/ and /a:/ becomes almost consistent in the speech of the third Palestinian generation. The second development in the speech of the first generation involves mainly the female Jordanian speakers and the male Palestinian speakers. The Jordanian women in this generation used a few tokens of [t], [d], and [™] instead of the traditional [θ], [o], and [g], respectively. In the speech of Palestinian men some tokens of [g] instead of their traditional [™] were used. These tokens, which represent divergence from the traditional dialects, are really sporadic, and statistically unimportant, but their sociolinguistic patterning is important: it is exactly these two groups of speakers (female Jordanian and male Palestinian) who in subsequent generations diverge most from their heritage dialects. We can, therefore, describe this development as some sort of instatement of sociolinguistic patterning. The second generation show extreme inter-speaker and intra-speaker variability to the extent of appearing chaotic, as a result of using a mixture of the two dialects, although their speech can still be identified as either Jordanian or Palestinian, mostly through vocalic and morphosyntactic features in the speech of some speakers. Additionally, we find a complication in the sociolinguistic correlations, for instance, “gender” emerges as a very important factor, in addition to dialectal or regional origin. Bearing in mind that all speakers showed considerable variability, some patterns are worth noting. The most conservative speakers in this generation, in the sense of diverging least from their traditional dialects, are the Jordanian men and the Palestinian women, and the most innovative speakers are the Jordanian women and the Palestinian men. For example, some Jordanian women in 64

THE FORMATION OF THE DIALECT OF AMMAN

this generation almost consistently used [™] for traditional [g], and the stop variants for the interdental sounds, but they used none of the Palestinian vocalic features; they mixed Jordanian and Palestinian pronominal suffixes (-hum and -hon, -ku and -kon), and they mixed Jordanian pronoun i}na, and Palestinian ni}na “we”. More examples of [g] for [™] occurred in the speech of Palestinian men, the vowels and the plural pronoun were mixed, and the pronominal suffixes were mostly Palestinian (-hon, -kon, rather than Jordanian -hum, -ku). The Jordanian men remained essentially [g] speakers, used a few tokens of stop for interdental, and the vowels, pronominal suffixes and the plural pronoun were consistently Jordanian (-hum, -ku, i}na). The Palestinian women consistently used traditional [™], identifiably Palestinian vowels and pronominal suffixes, and ni}na. These tendencies are tabulated below. Table 3.1 Examples of the patterns found in the second generation

J men P women P men J women

g

§

θ, 6, 1

+



+



+



+



+





+



+



+



+



+

+

+



+

+

+



+

+

+

+

+



+

+



+

+

+

+

t, d, |

J vowels

P vowels

J suffixes

P suffixes

i}na

ni}na

J: Jordanian; P: Palestinian; + denotes “presence”; − denotes “absence”.

Interestingly, the third generation do not adhere to these forms faithfully, but in some cases they focus the patterns which appear in their parents’ generation, while for many features they set totally new patterns, and use new features which were not present in the original mix. It is the speech of the third generation to which the remainder of this chapter is dedicated, and which shows the blueprint of Ammani. 3.6.1 The third generation: Ammani In the speech of this generation, all of whom were born in Amman, the mixture and variability witnessed in the second generation is considerably reduced, and some orderly linguistic behavior is established. We find a high degree of stability of usage of certain linguistic features, use of fudged linguistic forms, and many new patterns and new features. Also in this generation, there is an agreement, a native judgment, concerning what is and what is not Ammani. 65

ENAM AL-WER

3.6.1.1 The consonants The competing features in Ammani are the same sounds that distinguish between the Bedouin and the Sedentary norms in Arabic dialects, namely, [g] versus [™], interdental (plain and emphatic) versus stop (or sibilant), and [d] versus []. The urban Palestinian ingredient in Amman is identical to the system found in all of the other major city dialects in the Levant in having [™], [t], [d], [\] (or the sibilant counterparts [s], [z] [z]) and [].6 All of the Jordanian dialects traditionally have [g], [θ], [o], [2] and [d], but, as I have previously reported, the dialects of Sult, Ajloun and Kerak showed considerable variation, and possibly change, in the interdentals and [d] (AlWer 1991). The data from the youngsters in Amman show that these sounds, [t], [d], [\] (or the sibilants), [] and to a large extent [™] (but see below) have become focused features of Ammani, and, thus, the dialect will be identical to the other Levantine city dialects in its consonantal system. However, the [g] versus [™] alternation has undergone an interesting stylistic and social reallocation. As an outcome of dialect contact, reallocation refers to the phenomenon where “one or more variants in the dialect mix survive the leveling process, but are refunctionalized, evolving new social or linguistic functions in the new dialect” (Britain and Trudgill 2005: 183). The case of [g] in Amman involves social and stylistic reallocation, and here are the data. Recall that in the initial stage of contact the distribution of the usage of [g] and [™] showed a straightforward correlation with the regional origins of the speakers: Jordanians used [g] and urban Palestinians used [™]. Recall also (see Table 3.1) that already in the second generation a gender correlation is apparent, between [™] and the female speakers, and [g] and the male speakers (see Section 3.4 for an explanation of the social motivations and the roots of this development). The third generation inherit this (g)-gender associations from the second generation, but, crucially, the inherited pattern undergoes redefinition of social context, and refunctionalization of the variants in the following ways. The female speakers. Irrespective of their heritage dialects, they use [™] consistently.7 This group of speakers do not really expound the full range of social meanings associated with the use of these variants, but largely advance the same pattern which had been established by the previous generation (see Table 3.1). The male speakers. In the case of [g] versus [™], it is the behavior of the boys which defines the range of meanings associated with this variation. They utilize the variations in a complex and functional manner: • • •

[™] is used among Palestinian boys (talking to each other) [g] is used among Jordanian boys [™] is used by both groups when talking to girls

66

THE FORMATION OF THE DIALECT OF AMMAN



[g] is used by both groups in mixed (Jordanian and Palestinian) boy groups

In the above situations, the context (mainly the interlocutors) is the determining factor. Further, the use of [g] in “fights”, what they call /yuwa∫/, plural of /yo:∫ε/, conflicts of any sort, is highly meaningful, and a highly valued resource. In these situations, [g] is not only valued for what it symbolizes (macho characteristics), but also for the image it averts. All of the boys that I interviewed claimed that a boy who uses [™] in fights can be called tant (plural tantAt) “aunt” (from French tante), by which they meant someone who is not tough.8 Interestingly, tant is not a boy who uses [™], but one who does not, or cannot, make the situational switch to [g], i.e. one who is incapable of appropriating the full range of variations. It is also interesting that the girls too participate in the construction of this meaning through their expectance and acceptance of it, even though they do not implement it themselves (and they would rather be courted in [™]); all of the girls in the sample commented on the fighting situation among boys as one which calls for the use of [g]; one girl claimed that “a boy who uses [™] in a fight is a subject of mockery /mahzale/”. The association of [g] with high ranking officials also survives in the new dialect. Three of the boys in my sample are sons (in one case a grandson) of men who had served as cabinet ministers. These boys used [g] consistently, also when they interacted with girls. The girls, in these boys’ groups of friends, commented on their use of [g], and licensed it by referring to their parents’ occupations. The fact that the association between the use of [g] and political influence is functionalized in this way by this generation is rather remarkable. The association which emerged during their grandparents’ generation on the basis of actual usage survives; it is almost as if they inherit it as a commodity; for others, the association has become a “social value”. The overall envelope of meanings demonstrated by the use of [g] presents a tantalizing paradox: [g] is old fashioned and backward (by association with provinciality), but it is also legitimate and attractive (by association with localism and influential positions). Clearly then, who uses which variant, and the social meanings of these usages is a complex issue. The social meanings as well as the distribution of the variants of (g) have been redefined. It is important to incorporate an understanding of these evolutions in sociolinguistic analysis in order to understand the current situation. 3.6.1.2 The vowels In variation studies on Arabic, especially the Arabic of the Mashreq, it is rare to find analyses of variation and change in vocalic features. There is a widespread impression that consonantal variation is sociolinguistically more 67

ENAM AL-WER

salient. Consonantal variation is certainly much easier to detect since it deals with discrete linguistic differences, whereas vocalic variations are gradient in nature (for elaboration, see Al-Wer 2002). The major urban dialects in the region show considerable vocalic differences, also when the consonantal features and the supra-segmental phonologies are identical. In fact, vocalic differences are the only way to tell the dialects of Damascus and Beirut apart phonologically, and the same goes for distinguishing between the dialects of Haifa and Jerusalem. In the contact situation in Amman, vocalic variations are very important, and have become more so since the consonantal system has become more or less identical to those found in the major cities in the Levant. In the Amman project, short and long vowel features have been analyzed. Below is a summary of some of the findings, and the linguistic processes involved. 3.6.1.2.1 A FUDGED FORM: THE FEMININE ENDING

This variable concerns the raising of the feminine ending /a/ to /ε/. Arabic dialects in general vary with respect to this feature. Some are typically nonraising dialects, and some are typically raising dialects, while others raise the feminine ending conditionally (see Cantineau 1960: 143–166). Urban Palestinian dialects and Jordanian dialects differ in the phonology and phonetics of the feminine ending. Jordanian dialects (central and northern dialects) use /a/ in all environments except after coronal sounds, in which case the ending is raised to a half open [ε] (i.e. /a/ is the default variant), as in the examples below9: /]ilwa/ “pretty”, /maglu:ba/ “Magluba, a traditional dish”, /ya¡ma/ “taste or flavor”, /∫ara:ka/ “partnership”, /bi∫¡a/ “ugly”. But, /madanijjε/ “modern or urban”, /fattε/ “Fatte, a traditional dish”, /sida:dε/ “carpet”, /mdawwazε/ “married”, /sanε/ “a year”, /zγi:rε/ “young”, /¡a:mlε/ “she has done/made” On the other hand, in urban Palestinian the feminine ending is realized as /e/ except after velarized, emphatic and pharyngeal sounds where /a/ is used (i.e. /e/ is the default variant). The raised variant is [e] (cardinal vowel 2 or closer), as in the following examples: /]ilwe/ “pretty”, /nadwe/ “seminar”, /xa¡be/ “difficult”, /ma]kame/ “court”, /mixtilfe/ “different”. But, /mantiqa/ “area/region”, /ruxxa/ “a licence”, /xulya/ “authority”, /a:m¡a/ “university”, /mari:\a/ “ill”. Thus, the word for “pretty” is /]ilwa/ in Sult, but /]ilwe/ in Nablus, and in both dialects /a/ is raised in /∫irse/ “vicious”, but no raising in either dialect in /a:m¡a/ “university”. 68

THE FORMATION OF THE DIALECT OF AMMAN

In Ammani, a fudged form10 is used, which combines Palestinian phonology and Jordanian phonetics, i.e. the younger generation raise everywhere except after velarized, pharyngeal and emphatic sounds, and the raised variant is phonetically half open [ε] (cardinal vowel 3). The Palestinian phonology was used by the third generation speakers consistently, without any exceptions. Regional koineization in this case, as in the case of the consonants, is an influence, considering that the same phonology is found in all urban dialects in the Levant (but not the same phonetics). With respect to the phonetic quality of the raised vowel, none of the youngsters from a Jordanian dialectal background raise further than [ε], and the vast majority of the youngsters from a Palestinian dialectal background lower [e] to [ε]. Interestingly, most of the [e] examples which occur in the data from the Palestinian youngsters come from the speech of the youngest speakers (12 and 13 year olds), which indicates that [e] is the variant that they acquired in the home environment, and that the acquisition of the lowered variant is a developmental process: the more time they spend with the peer group, the more they diverge from their parents’ dialects. With respect to the phonetics of the feminine ending, Ammani and Damascene can be grouped together in opposition to Beirut and Jerusalem, which have very close realizations; the Beirut variant is an /i/ type of vowel. But, all four dialects have an identical phonology. 3.6.1.2.2 LEVELING OF MARKED FEATURES: THE VARIABLE (r a :)

The input dialects differ with respect to the quality of /a:/ in the vicinity of /r/. Although the quality of /a:/ in this environment in Jordanian and Palestinian is back, the Palestinian variant is considerably more back, accompanied with some degree of lip rounding and pharyngeal constriction11 (except where there is an /i/ after /r/, e.g. /mbε:ri]/ “yesterday”). I shall use IPA [ɑ] for the Jordanian variant, and IPA [ɒ] for the Palestinian variant. Thus, the word /ra:]/ “he has gone” in the Jordanian input is [rɑ:]] and in the Palestinian input is [rɒ:]]. The data from Amman’s younger generation show a markedly more front realization than the input varieties, e.g. /dira:se/ “studying”, /isra:™i:l/ “Israel”, /wra:™/ “papers”, /tia:ra/ “trading”, /ra:j]a/ “I am going”, /kta:r/ “many”. Fronting of (ra:) shows a sharp gender distinction. Of the front realization, 90 percent occurred in the speech of the female Ammanis. It is also noticeable that none of the female Palestinian Ammanis used the Palestinian variant (their heritage variant), and the male Ammanis of Jordanian origin did not use the Palestinian variant at all. Clearly, the very back and rounded pronunciation is being leveled out. In the region as a whole, the Palestinian pronunciation is in the minority; it is localized and marked, and, certainly, a prime candidate for leveling in a contact situation. In phonological space, the variant which is being focused in Amman is nearer to the left end of a continuum of front-back /a/, and more similar to the Damascus and Beirut variants. 69

ENAM AL-WER 3.6.1.2.3 MOVEMENT OF VOWELS: THE VARIABLE (a :)

In phonological environments other than around /r/, the quality of /a:/ in the input dialects is as follows: the Palestinian variant is considerably more peripheral, towards the front end, and raised, ranging from raised /æ:/ ~ /ε:/ ~ /e:/, e.g. /ε:/ or /e:/ “chicken”, /mε:xdi:n/ “we/they have taken”; the Jordanian variant is lower and less peripheral, /da:d/ or /dæ:d/ “chicken”, /ma:xoi:n/. There are more than 800 tokens of (a:) in the data from the third generation, and the variant which is most frequently used by Ammanis is [æ:], i.e. slightly higher than the Jordanian input, and lower than the Palestinian input. All Ammanis showed variation in this feature. However, it is noticeable that the Jordanian third generation did not use Palestinian [æ:] or [e:] for /a:/ at all, i.e. the variation in their speech involved [a:] and [æ:]. The few occurrences of the very advanced [e:]-like pronunciations occurred in the speech of the youngest Palestinian boys (see Al-Wer 2002). 3.6.1.2.4 A CONNECTED MOVEMENT: DIPHTHONGIZATION OF /e :/ AND /i /

The movement of /a:/, outlined above, is linear, ending somewhere within the phonological space of /æ:/ or slightly higher. Ammanis also show divergence from their parents dialects in the lexical sets containing /e:/ and /i:/. In the input dialects, these lexical sets contain pure long vowels, e.g. /be:t/ “house”, /ze:t/ “oil”, /ze:d/ “Zaid (a boy’s name)”, /¡e:n/ “eye”, /∫i:l/ “carry!”, /zi:d/ “add!”. The data show that these vowels in the speech of many speakers move in a non-linear fashion, thereby developing an off-glide, as in the following examples: /™ə∫/ “what”, /heək/ “this way”, /zeət/ “oil”, /miən/ “who”, /¡iəb/ “shame”. It is possible that the diphthongization was triggered by the upward movement of /a:/, causing pressure in the phonological space of /ε:/~ /e:/, and /i:/. This suggests that the movements of the long vowels are connected, showing a prototypical pattern of a chain shift (see Labov 1994: Chapter 5). Seen from the perspective of Jordanian dialects, the clockwise movement of the long vowels shows a prototypical pattern of a chain shift, broadly in line with Labov’s Principles that govern the chain shifting of vowels: Principle I, long vowels rise; and, Principle III, back vowels move to the front (Labov 1994: 116). From the perspective of the Palestinian dialects, the movement of long back [ɒ:] to the front conforms to Labov’s Principle III, but the lowering of [e:] and [ε:] to [æ:] is at variance with Principle I.12 3.6.1.3 Markedness and simplification: the pronominal suffix -kum13 One of the most interesting morphosyntactic features which has been focused in Amman is the 2nd person plural pronominal suffix -kum, a clitic form

70

THE FORMATION OF THE DIALECT OF AMMAN

which can attach to nouns, verbs, and prepositions in a variety of NP and VP constructions, e.g. /bju:tkum/ “your homes”, /sa™alna:kum/ “we asked you”, /minkum/ “from you”. As used in Amman, this form is neutralized for gender. The data from the first generation, as well as the data from my previous research in other locations show that this form does not occur in any of these varieties. On the other hand, the younger generation in Amman only use -kum and no other form. I believe that this feature is an innovation, which was locally “made” by the younger generation of both Jordanian and Palestinian origins. Of course, -kum occurs in the Standard variety, but its use in Standard Arabic is restricted to the 2nd person plural masculine in opposition to -kun, the 2nd person plural feminine clitic. -kum can also be found in a number of Arabic dialects, e.g. Cairo and Baghdad. The -kum in Amman could not have come from these dialects, given the limited contact (or any face to face contact) between the speakers who use it consistently and speakers from Cairo or Baghdad. It is unlikely to have been “borrowed” directly from the Standard variety, seeing that the Standard variety does not show any influence in other features. Nonetheless, it is perfectly possible that -kum may have become familiar to Ammanis as a “legitimate” 2nd person plural form through its occurrence in the Standard variety or in other dialects. The interesting question, however, is why it was chosen over the forms which are present in the input dialects. In order to address this question, I shall start by listing the forms found in the major input dialects: Jordanian Old northern dialects New koineized northern forms Old southern dialects New koineized southern forms Palestinian

2nd pl. masc -ku -ku -ku -ku -kun/-kon

2nd pl. fem -kin -ku-ku (gender neutralized form) -kan -ku-ku (gender neutralized form) -kun/-kon (gender neutralized form)

Two developments can be noticed by comparing old with new Jordanian forms: the gender distinction is neutralized, and the neutralization is in favor of the masculine form. The same pattern can be found in the developments which have affected the 2nd plural pronouns, where the old intu/intin distinction has been reduced to intu. The Palestinian dialects in the data from Amman do not show gender distinction, rather an already simplified form, but in this case, the simplification is in favor of the originally feminine form -kun/-kon. The formation of the Amman dialect is, in addition to the input varieties, influenced by regional koineization, as argued above. It would

71

ENAM AL-WER

be sensible, therefore, to look at the forms found in Damascus and Beirut. Fortunately, we have extensive information on this feature in Behnstedt’s Sprachatlas von Syrien (1997), which shows that both cities and their surroundings have -kon (and so do all Syrian city dialects), i.e. similar to the Palestinian input dialects in Amman. In other words, none of the dialects which may have influenced Amman have -kum. To all intents and purposes, the form -kun/kon appears to be the regional urban koine. But, why has this form failed to make it in Amman? To explain the failure of the urban koine -kun/kon and the success of -kum in Amman, I shall refer to two related and widely cited phenomena, namely markedness and simplification. The idea is that in contact situations and new dialect formation the unmarked forms may have a better chance of survival in the new dialect than the marked forms. Trudgill (2004: 117–120) refers to this principle to explain the presence of the weak vowel merger (see Wells 1982, vol. 1) as a feature in Southern Hemisphere English. The principle of markedness has also been referred to in explanations dealing with Creole genesis (Mufwene 2001, cited in Trudgill 2004). Simplification can involve regularization of patterns, reduction, symmetry in paradigms, and generally simpler morphophonemics (see the discussion of simplification in Trudgill 1986: 102–107). With respect to markedness, I argue that -kun/kon in the context of Jordan is marked in two ways. First, it is marked phonetically. All of the Jordanian dialects which have a feminine form to contrast with the masculine form have -kin, with a front vowel, not -kun (or -kon) with a back vowel. Second, -kun/kon is marked paradigmatically in that, as noted above, where gender neutralization occurs in Jordanian dialects, it always generalizes the masculine forms, not the feminine forms. In addition to the neutralizations mentioned earlier, this is also true in the case of neutralizations in the 3rd person plural clitics and pronouns (-hum, humme). Therefore, the choice of -kum may be seen as a way out which avoids locally marked forms and patterns. Simplification in the choice of -kum, specifically morphophonemic regularization, is achieved by way of a more regular phonetic correspondence between the 2nd and the 3rd person clitic forms. Consider the paradigms listed on page 73. As can be seen, the Palestinian and the Ammani forms are more regular, although in different ways, than the Jordanian forms. The choice of -hum for 3rd person plural in Amman is in keeping with the Jordanian pattern, namely, where gender is neutralized, the masculine form is generalized, and the choice of -kum then is perfectly in keeping with this pattern, with the additional bonus of achieving regularity. Notice that the Palestinian forms are no less regular than the Amman forms, but they have not been successful in the formation of the new dialect, which is further evidence that it is the interplay of a number of parameters which determines the outcome of contact. 72

THE FORMATION OF THE DIALECT OF AMMAN

2nd pl. masc 2nd pl. fem 3rd pl. masc 3rd pl. fem Jordanian Old northern dialects New koineized northern forms Old southern dialects New koineized southern forms Palestinian Amman

-ku

-kin

-hum

-hin

-ku

-ku

-hum

-hum

-ku

-kan

-hum

-han

-ku

-ku

-hum

-hum

-kun -kum

-kun -kum

-hun -hum

-hun -hum

3.7 Summary and conclusion The availability of speech samples from all three generations in Amman provides a rather unique opportunity, to observe a language system in the process of formation, and provides a testing ground for the principles of new dialect formation and language change. Although a rounded picture of the dialect will only become clear when a detailed analysis at all linguistic levels has been completed, the data available so far, some of which have been presented in this article outline the essential processes and mechanisms (see also Al-Wer 2002, 2003). The stages of the formation of the dialect of Amman can be summed up in the following way (cf. Trudgill 2004, Chapter 7). In Stage I, first generation speakers arrived in the city as adults. They spoke the dialects which they had acquired as children in their home towns. In the context of a location where no one was native to the place, they came into contact with speakers of different dialects. To the extent that they were linguistically capable of acquiring new habits, or altering old habits, and through contact with and exposure to speakers of other dialects, their native dialects underwent rudimentary leveling, as part of a koineization process. Their children, Stage II, the first native-born generation, were not only exposed to the dialects of their parents, but, in their formative years and unlike their parents, they were also exposed to a wide range of variations. In other words, they would have acquired a variable model. The data from this generation showed that the mixture of features from more than one dialect stock was still present in their speech as adults, and although some of the features which appear in the third generation appear in the second generation, in Stage II there is little stability, which I described as a chaotic situation, a defuse language situation in the terms of Le Page and Tabouret-Keller (1985). 73

ENAM AL-WER

In Stage III, the second native-born generation, some aspects of the koineization process continue (e.g. reallocation). However, there is an evolution of norms, stability of usage, and reduction of the extreme variability found in Stage II. In the final stage, focusing (see Le Page and TabouretKeller 1985, Chapters 5 and 6) starts, and involves regularization in the linguistic system itself, as well as in the behavior of individuals. A corollary to linguistic focusing is a sense of identification with the group and with the locality. In the Amman situation, the derivation Ammaniyyiin may be considered as an emblem of such identification.

Notes 1 It is not clear why Amman, at the time a mere village with less than 10,000 migrant inhabitants, rather than Sult, was chosen as the site of the capital city. Popular views in the country suggest that the founder of the Kingdom, King Abdullah I, until then Emir Abdullah of Hijaz, who himself was an outsider, thought it prudent to avoid well established local towns to set up a central government, as a way to minimize local objections to his project. I am not aware of any scholarly works which may have discussed this issue seriously. 2 Salibi 1993, (Chapter 6), makes a similar observation, which he uses to analyze the repercussions of various political developments. In a similar vein, Shami (1982, 137–139) points to the tribal system in Jordan, at the level of politics and power relations, to explain the weakening of the ethnic corporation among the Circassians in the country. 3 In view of this observation, it is more likely the case that those who were promoted came from the sectors of the population that believed in the tenets of the Jordanian State. 4 A further dimension to this analysis is that the Jordanian bureaucracy, by and large, was immune to the level of corruption and bribery seen in other Arab and Third World bureaucracies. Efficiency and seriousness in applying the law are generally perceived as traits of the traditional bureaucrat. The local Jordanian linguistic features came to connote such characteristics in an official. 5 Research currently in progress on the traditional dialect of Sult by Bruno Herin of Brussels University shows that /k/ affrication continues to be used, mainly by male older speakers (personal communication). 6 A notable exception to this generalization is the dialect of Aleppo which has all of these but [d] not [] (see Behnstedt 1997). 7 Except in specific words or expressions in which [g] is fossilized and not subject to variation, e.g. /mgal¡ay/ “disgusting”, which, when used as a swear word, the speakers claimed “has a stronger effect with [g] than with [™]”). 8 tante in French can be used to describe someone as being homosexual. As used in Amman, the term can imply effeminacy in a boy, without reference to sexuality. It can be also directed at girls to mean “not cool” generally. Elie Kallas has suggested that tante may have entered the speech of young Ammanis via Lebanese, which is quite plausible, considering that French borrowings in Jordanian Arabic are usually borrowed secondhand via Lebanese. The term would have undergone modifications in the range of meanings it denotes in Amman. 9 More examples and further discussion of the vocalic features presented in this chapter can be found in Al-Wer (2002).

74

THE FORMATION OF THE DIALECT OF AMMAN

10 For further analyses and discussion of the emergence of fudged linguistic forms in dialect contact situation consult Trudgill (1986: 60–62). 11 The phonetic quality of Palestinian /r/ is subject to some discussion in the literature; Card (1983) argues that it is not emphatic, whereas Younes (1982) argues that it is emphatic and is de-emphaticized in certain environments. In so far as /ra:/ is concerned, this issue does not really impact on the analysis of the quality of the vowel, especially since /a:/ is changing along the same course in other environments (see Section 3.6.1.2.2). The fact is that long /a:/ in all environments moves to the front and, in some cases, upwards. 12 Labov (1994: 137) cites a counter-example to Principle I from East Lettish, where [ε:] is lowered to [a:]. 13 A more detailed analysis of this variable is in Al-Wer (2003).

References Abdel-Jawad, H. (1981) “Lexical and phonological variation in spoken Arabic in Amman”, unpublished thesis, University of Pennsylvania, USA. Al-Wer, E. (1991) “Phonological variation in the speech of women from three urban areas in Jordan”, unpublished thesis, University of Essex, UK. Al-Wer, E. (2002) “Jordanian and Palestinian dialects in contact: vowel raising in Amman”, in M. Jones and E. Esch (eds) Language Change: The Interplay of Internal, External and Extra-linguistic Factors, Berlin: Mouton de Gruyter, pp. 63–79. Al-Wer, E. (2003) “New dialect formation: the focusing of -kum in Amman”, in D. Britain and J. Cheshire (eds) Social Dialectology: In Honour of Peter Trudgill, Amsterdam: Benjamins, pp. 59–67. Behnstedt, P. (1997) Sprachatlas von Syrien, Wiesbaden: Harrassowitz Verlag. Britain, D. (1991) “Dialect and space: a geolinguistic study of speech variable in the Fens”, unpublished thesis, University of Essex, UK. Britain, D. and Trudgill, P. (2005) “New dialect formation and contact-induced reallocation: three case studies from the Fens”, International Journal of English Studies 5 (1): 183–209. Cantineau, J. (1960) Cours de phonétique arabe, translated by S. AlQarmaadi (1966), Tunis: Centre of Social and Economic Research and Studies. Card, E. (1983) “A phonetic and phonological study of Arabic emphasis”, unpublished thesis, Cornell University, Ithaca, USA. Cleveland, R. (1963) “A classification of the Arabic dialects of Jordan”, Bulletin of the American School of Oriental Research 167: 56–63. Dawood, G. (1994) “As-Sult wa jiwariha khilal al-fatra 1864–1921 [Sult and its neighborhood during 1864–1921]”, unpublished thesis, University of Jordan, Amman. Eckert, P. (2000) Linguistic Variation as Social Practice, Oxford: Blackwell. Gordon, E., Campbell, L., Hay, J., Maclagan, M., Sudbury, A. and Trudgill, P. (2004) New Zealand English: Its Origins and Evolution, Cambridge: Cambridge University Press. Kerswill, P. (1994) Dialects Converging: Rural Speech in Urban Norway, Oxford: Clarendon Press. Kerswill, P. and Williams, A. (2000) “Creating a new town koine: children and language change in Milton Keynes”, Language in Society 29: 65–115.

75

ENAM AL-WER

Labov, W. (1994) Principles of Linguistic Change: Internal Factors, Oxford: Blackwell. Le Page, R. and Tabouret-Keller, A. (1985) Acts of Identity: Creole-based Approaches to Language and Ethnicity, Cambridge: Cambridge University Press. Milroy, Lesley (1980) Language and Social Networks, Oxford: Blackwell. Palva, H. (1970) “Balgawi Arabic: texts from 3afey”, Studia Orientalia XLIII: 1, 3–26. Palva, H. (1989) “Linguistic sketch of the Arabic dialect of El-Karak”, in R. Wexler, A. Borg and S. Somekh (eds) Studia Linguistica et Orientalia Memoriae Haim Blanc Dedicata, Helsinki: Studia Orientalia, pp. 225–251. Palva, H. (1994) “Bedouin and Sedentary elements in the dialect of Es-Salt: diachronic notes on the sociolinguistic development”, in D. Caubet and M. Vanhove (eds) Actes des Premières journées internationales de dialectologie arabe de Paris, Paris: INALCO, pp. 459–469. Palva, H. (2003) “Negation in the dialect of Es-Salt, Jordan”, in M. Haak, R. de Jong and K. Versteegh (eds) Approaches to Arabic Dialects, Leiden: Brill, pp. 221– 226. Salibi, K. (1993) The Modern History of Jordan, London: I. B. Tauris & Co. Shami, S. (1982) “Ethnicity and leadership: the Circassians in Jordan”, unpublished thesis, University of California, Berkeley, USA. Trudgill, P. (1986) Dialects in Contact, Oxford: Blackwell. Trudgill, P. (2004) New Dialect Formation: The Inevitability of Colonial Englishes, Edinburgh: Edinburgh University Press. Wells, J. (1982) Accents of English (Volume 1), Newcastle: Athenaeum Press. Younes, M. (1982) “Problems in the segmental phonology of Palestinian Arabic”, unpublished thesis, University of Texas, Austin, USA.

76

URBANIZATION AND DIALECT CHANGE

4 URBANIZATION AND DIALECT CHANGE The Arabic dialect of Tripoli (Libya) Christophe Pereira

4.1 Introduction In this chapter I shall give an account of the historical and contemporary development of Tripolitan Arabic (TA), describing the historical, sociological and urban contexts which have had an influence. I shall begin by recapping the history of the arabization of Libya from the seventh to eleventh centuries, together with the main historical events of the twentieth century which led to population movements within Libya, and the urbanization of the city of Tripoli itself. I shall then describe some of the characteristics of TA, and compare them with features of pre-Hilalian and Bedouin vernaculars. Finally I shall examine certain questions regarding the nature and origins of TA. The present linguistic description is based on corpora compiled in Tripoli in November 2002 and February 2005, together with a corpus recorded with two 25 year old Tripolitans in Genoa, Italy in September 2003. I also use various materials taken from the published literature on Libyan Arabic.1 Relatively little information is available regarding the linguistic situation in the city of Tripoli. This makes it difficult to provide a sociolinguistic analysis that takes into account variations relating to gender, age, social origin, etc. I have therefore adopted an historical urban dialectology approach, using the categories and dialectal typologies that have been developed to describe the dialectal situation of the Maghreb. I shall attempt to analyze the extent to which TA does or does not match these categories. The categories are based on the distinction between pre-Hilalian and Hilalian dialects (Caubet 2001; P. Marçais 1957; W. Marçais 1950). These terms are used in an historical perspective, based on the classifications set up by William and Philippe Marçais, with the definitions they provided, following Ibn Khaldoun. The two dialect groups can be distinguished in the Maghreb by certain linguistic features (Caubet 2001). However the distinction 77

CHRISTOPHE PEREIRA

between pre-Hilalian and Bedouin dialects is based on an historical demarcation which has evolved significantly with time, with population movements and inter-mingling, often giving rise to hybrid dialects. The description of TA outlined in this chapter uses isogloss lists in an attempt to determine the extent of hybridization.

4.2 The history and arabization of Libya Like all the countries of the Maghreb, both before and since the arrival of the Arabs, Libya has seen a great diversity of populations, especially in its urban centers. Before the arrival of the Arabs in the seventh century the region of present day Libya was populated, successively or simultaneously, by the Libycs,2 Jews, Phoenicians, Greeks, Romans, Vandals and Byzantines. Berber dialects were not completely obliterated by these successive waves of migration. They still form to this day a number of pockets of resistance.3 The Phoenicians established a coastal empire in Tripolitania based on Punic Carthage, which lasted more than six centuries (from the eighth to the second century bc); they spoke Punic, the first Semitic language to become established in North Africa. It was spoken in Tripolitania long after the Roman conquest until the second century ad. The Phoenician trading post of Wy’t, on which Tripoli was built, was founded at this time. The Greeks settled in Eastern Libya as early as the seventh century bc, and in 631 founded Cyrenia, after which the eastern region of present day Libya was to be named (Cyrenaica, or Barqa in Arabic). Eastern Libya remained an active center of Hellenism for more than a thousand years. The Romans destroyed Carthage in 146 bc, and shortly after settled in Libya, where they remained until the fifth century ad. They of course brought Latin, but this remained solely a language of administration, and failed to replace either Berber and Greek in Cyrenaica, or Punic in Tripolitania. The Phoenician trading post of Wy’t was to become the Roman town of Oea. The Vandals, already settled in Spain, invaded North Africa in 430, took Carthage in 439 and went on to occupy Tripolitania. The Byzantines were present in Libya from 535 until the arrival of the Arabs. The Vandal and Byzantine occupations were short lived and did not eliminate the pre-existing linguistic and religious influences. The Phoenicians set up three trading posts in Tripolitania – Sabratha, Oea (Wy’t) and Leptis Magna. Throughout the Punic and Roman periods the coastal towns of Tripolitania were commercially prosperous. The word Tripoli itself derives from these three towns which joined together to form a state (Greek tripolis: tri three, tri-, and polis town). At the time of the Arab invasion Oea was already a large town. The Arabs settled there and the present day city of Tripoli developed on this site (Azema 2000; Lafi 2002; Mantran 1975). The arabization of the Maghreb, resulting from the Muslim conquest which swept in from the East, took place in two stages. The first conquest, 78

Map 4.1 Tripoli (Libya)

URBANIZATION AND DIALECT CHANGE

79

CHRISTOPHE PEREIRA

in the seventh century, did not lead to a comprehensive arabization of Libya, nor of the other North African countries, which for the most part remained Berber-speaking (Mantran 1975; W. Marçais 1961; Rossi 1954, 1968). Tripoli was taken in 645 by ¡Amr ’Ibn al-¡fx, but it was only in 696–697 with the military expedition led by massan al-Ghassani that the Eastern Arabs settled permanently in Libya. This first seventh century conquest was not followed by the systematic settlement of the country by the conquerors. They were relatively few in number and occupied only the towns and points of strategic importance. The rural and mountainous areas were left to the Berber populations until the middle of the eleventh century. The Arabic dialects of the regions that were the first to be invaded and arabized in the seventh century are thought to have been pre-Hilalian dialects. The systematic and widespread arabization of North Africa occurred in the eleventh century, with the second conquest, when Bedouin tribes settled the area. In 1050–51 the Bane Hilal, the Bane Sulaym and the Ma¡qil reached Cyrenaica. While the Bane Sulaym remained there, the Bane Hilal continued their migration westward. In Tripolitania the Bane Zughba tribe, (a branch of the Bane Hilal) occupied the whole region stretching from Tripoli to Gabès, while the Bane Dabbab (who were another branch of the Bane Hilal) settled the eastern region. In the twelfth century, both arabization and islamization made considerable progress across the plains and desert regions. Pre-Hilalian Arabic survived in urban areas, but the Bedouin influence grew progressively stronger. Arab settlement was densest in the coastal areas and part of the Tripolitanian hinterland, a process which intensified during the Hafsidic period, after 1230. Cyrenaica, which was attached to Egypt at this time, was populated mainly by nomadic Arabs. Fezzan, with several prosperous towns, was an independent state, and a transit point for the caravan trade. The Arabic dialects spoken in these regions, occupied and arabized from the eleventh or twelfth century, are thought to have been Bedouin dialects. The arabization of the city of Tripoli thus began in the seventh century (643–645) with the arrival of the first Arab conquerors. It is assumed that from this time a pre-Hilalian Arabic dialect became well implanted and was to develop. It would have undergone a process of koineization following the arrival of migrants of rural/Bedouin origin, but it would appear that from the end of the nineteenth century it was spoken only by the Jewish population (see below).

4.3 Population movements and the urbanization of the Tripolitanian littoral The urban development of the Tripolitanian region was the result of population movements which began during the period of Ottoman rule, and increased throughout the twentieth century. Large scale movements towards 80

URBANIZATION AND DIALECT CHANGE

Tripoli began with the second period of Ottoman domination in 17114 and continued until the arrival of the Italians in 1911. The Ottomans instigated massive displacements of rural and Bedouin populations in the regions under their control (Burgat and Laronde 2000; P. Marçais 1957; Martel 1991). Tripoli and Tripolitania were the regions of Libya most affected by the Ottoman occupation. The Italians and the British also caused population displacements in Tripolitania between 1930 and 1935, and during World War Two (Bessis 1986; Souriau 1975). At the end of World War Two there was a large influx of young people from rural areas into Tripoli, but it was mostly after Libyan independence in 1951 and the discovery and development of oil reserves in 1955 that very large scale population movements and significant demographic changes were seen. The oil revolution very quickly transformed the social situation. It led to the progressive disappearance of the nomadic way of life, a huge rural exodus, inter-urban migrations into Tripoli and Benghazi, the very rapid urbanization of these two cities and the return of many Libyans who had migrated to neighboring countries, and who now chose to settle in Tripoli (Bessis 1986; Djaziri 1996; Martin 1975). Whereas at independence Libya was a country with a rural economy and a large nomadic population, it is now, at the beginning of the twenty-first century, one of the most highly urbanized countries of the Arab world, and the most highly urbanized country in Africa. In 1954, Libya had 288,000 nomads, and 273,400 town dwellers out of a total population of 1,089,000. The town dwellers represented 25.1 percent of the population and were spread across a mere ten towns. Today the population totals six million, of which 85.8 percent are town dwellers, living in a hundred towns. Between 1964 and 1995 the urban population grew four fold and the population of Tripoli as a whole increased by 450 percent (Fontaine 2004). In addition to internal migrations, there has been a significant immigrant inflow. Due to a shortage of local labor, many foreign workers have moved to Libya not only from the neighboring countries of Tunisia and Egypt, but also from the Middle East – Syria, Lebanon and Palestine. Between 1954 and 1973, 187,000 foreigners entered Libya, about half settling in Tripoli and Benghazi. The country from which came the highest numbers of immigrants was Egypt, and it is reasonable to suppose that this would influence the dialects spoken in Libya, especially in the east, where Egyptian immigration was most concentrated. In tandem with these sociohistorical factors which affected the development of TA, we must consider the linguistic and educational policies in force after 1969, which amounted to a strict and uncompromising arabization (Quitout 2004: 5). Classical Arabic was considered the mainstay of ArabMuslim culture and became the sole official language. This involved a rearabization of public life, the work place and education, but more importantly there was a strong literacy policy, for the government of the day wanted not 81

CHRISTOPHE PEREIRA

only to see the arabization of the elite but of the masses as well. It is estimated that in 1973, 32 percent of men and 73 percent of women over 10 years of age were illiterate (Souriau 1986). Libya is currently said to be the country with the lowest illiteracy rate in the Maghreb. In 1995, 76.2 percent of the population (63 percent of women and 87.9 percent of men) was reported to be literate (Ham 2002). At the same time, due to a shortage of teachers, Libya had to recruit foreigners to teach Arabic, many of these coming from other Arab countries, especially Egypt. There are more foreign teachers than Libyan teachers.5 There again it must be supposed that this had an effect on the speech of the Libyan pupils and students taught by these foreign teachers.

4.4 Tripoli from a sociohistorical, demographic and linguistic perspective The profound transformation of the TA dialect caused by these different demographic movements gives rise to a number of important questions. Can it be said that an urban Bedouin dialect emerged in Tripoli as a result of koineization, as happened in other towns of the Maghreb, following the massive rural exodus and the settlement of peasants in the city alongside nomads adopting a sedentary life style (Messaoudi 2003; Miller 2004)? To what extent did Jews and Christians speak different dialects? Did the Jews in Tripoli speak a pre-Hilalian dialect similar to the ancient language of Tripoli, as is indicated in Judeo-Arabic texts?6 And did Muslims also speak a pre-Hilalian dialect, and did they perhaps gradually adopt a koineized Bedouin dialect? Records dating back to the seventeenth century suggest that Tripoli was a cosmopolitan city with a relatively large Jewish community from the Ottoman period onwards, with a territorial division of its different communities which lasted until the beginning of the twentieth century. In the seventeenth century the city had between 15,000 and 20,000 inhabitants, comprising Arabs, Turks, Koulouglis, Jews, Christians (many of them slaves), Janissaries, renegades (Christians converted to Islam) and Arnauts (Albanian Muslims) (cf. Lafi 2002). The Jewish community has been relatively well studied and we know it represented a significant proportion of the population of Tripoli in the eighteenth and nineteenth centuries (Agostini 1954; Bessis 1986; Lafi 2002; Simon 1992). Jews were present in Libya from the end of the third century bc, having come there from Egypt and Palestine. Many Jews expelled from the Iberian Peninsular settled in Libya throughout the fourteenth, fifteenth and sixteenth centuries. When Tripoli was occupied by the Ottomans in 1551 there were very few urban Jews in Libya, but the following years saw a gradual Jewish migration from inland areas of Tripolitania to the coastal towns. Between 1906 and 1954 the majority of the Jewish population in Libya lived in Tripoli.7 82

URBANIZATION AND DIALECT CHANGE

At the beginning of the twentieth century the population of Tripoli was divided into Muslim districts, }Ema, and Jewish districts, }Ara, (Lafi 2002). The Muslims were in the majority and occupied two thirds of the town intra muros, in four districts: }Emat-al-baladCya (1,740 inhabitants), }Emat-γaryAn (3,512 inhabitants), kEKat-as-saffAr (2,277 inhabitants) and bAb-al-b}ar (2,262 inhabitants). The }Emat-γaryAn, which was the most populous district of the Medina, was named after its inhabitants, who had originated in Gharian, a small town at the foot of the Djebel Nefoussa about one hundred kilometers south of Tripoli. It is very probable that the arrival en masse in the mid1960s of these people, with their own particular dialect, would have had an influence on the Arabic spoken in the Tripolitan Medina. No ancient descriptions of TA exist to confirm that there was ever a pre-Hilalian dialect common to the whole population of the city. The oldest published work is that of Hans Stumme (Stumme 1894, 1898), who at the end of the nineteenth century described a Bedouin dialect with some preHilalian features, which indicate that it was a koine. These features included the absence of interdentals, the expression of the possessive relationship analytically through the particle mtA∞, the use of the structure zDz “two” + noun in the plural when counting objects, and various lexical items from pre-Hilalian dialects. Various writers (Cohen 1930; Goldberg 1978; Saada 1984) have provided us with information on the language spoken by the Jewish community of Tripoli, and it seems that this certainly was a pre-Hilalian dialect. Some writers believe that a pre-Hilalian Jewish dialect co-existed in Tripoli with a Muslim Bedouin dialect, but it is difficult to determine at what point this Muslim dialect, the result of a process of koineization, emerged.

4.5 The Tripolitan dialect of Arabic In examining the recent development of the Tripolitan dialect I compared my field data with earlier work by Stumme 1898, Cesàro 1939, Griffini 1913 and Trombetti 1912, as well as my own (Pereira 2003, 2004, 2006, forthcoming). According to these authors, the TA at the end of the nineteenth and beginning of the twentieth centuries was an historically hybrid dialect, combining at all levels Bedouin type features with Hilalian type features. My more recent research confirms that present day TA is more of a Bedouin type dialect. Lexically, the language of the earlier period borrowed much more from Turkish and Italian, the languages of administration, whereas nowadays many such items have been replaced. Borrowings are now made from Standard Arabic, or, depending on the subject matter involved, from English or Egyptian Arabic. Certain Bedouin features seemed to have been reinforced, such as the use of the duel number for nouns and of the synthetic possessive construction. Phonetically, labio-velarization has disappeared from TA (Pereira 2003). Furthermore, with regard to word 83

CHRISTOPHE PEREIRA

internal combinations or sequences of sibilants, certain sequences now occur in modern TA while there are fewer cases of assimilation, dissimilation and metathesis than in the earlier period. In the following sections I shall present a number of features of TA, emphasizing this co-existence of Bedouin and pre-Hilalian features. As a basis for this comparative research I have used Judeo-Arabic texts dealing with a pre-Hilalian dialect taken from the work of Mordekhai Cohen (1930) and the study of Sumikazu Yoda (2005). I shall also compare TA with the Bedouin dialects of Fezzan (Caubet 2004; P. Marçais 2001) and the urban dialects of Benghazi (Owens 1983, 1984; Panetta 1943). 4.5.1 Phonetics As it is in other Arabic dialects in Libya, whether they be urban or rural, the *q is realized as a voiced occlusive [g]: yug∞od “he remains”, ya}r¢g “he burns”, in¢gg¢z “he jumps”, y¢gd¢r “he can”, etc. In the dialect spoken by the Tripolitan Jewish community8 *q was realized as a voiceless occlusive [q]: qmê}a “a grain of wheat”, l-eqrâb “close friends, neighbors”, qamer “moon”, sûq “marché”, fy-ççandûq “in the box”, §aqrab “scorpion”, sûq “market”, qâl “he said”, }álqa “thimble” (Cohen 1930) and q¢ddam “in front of”, q¢ddaK “how much”, áq¢XX “it was cut”, Ü¢||¢yy¢q “severe, rigid” (Yoda 2005: 58). The pronunciation [q] for *q is also found in certain TA words borrowed from Standard Arabic, especially in religious and legal terminology, such as qur§An “Koran”, YarCqa “manner, religious brotherhood”, quYur “diameter”, qas| “goal, aim”, etc. There are moreover certain words which exhibit a frequent alternation between [g] and [q]. All researchers have noted the following examples, from Tripolitania, Fezzan or Cyrenaica: aqAreb and agAreb “parents”, q0|i and g0|i “judge”. The [q] ~ [g] alternation is a stylistic choice, reflecting the register adopted; it is definitely not a remnant of the pre-Hilalian dialect. TA shows no sign of Middle Eastern influence and *q is never pronounced as a laryngeal occlusive [™]. Unlike in many Bedouin type dialects, interdental sibilants do not occur in TA. They fuse with dental occlusives: *ªlaªa > tlAta “three, *ohəb > dh¢b “gold”, *xoa > xde “he took”. There is just one voiced emphatic [\]: *mrc\ > mrC| “sick”, *2-h-r > |+h+r “midday”. The loss of interdental sibilants and their fusion with dental occlusives also occurred in pre-Hilalian dialects. Taine-Cheikh (1998: 30) considers that in Tripoli their disappearance is the sole remaining feature of this urban dialect (Yoda 2005: 70). The absence of interdentals would appear to be an urban feature still found today in western Libya. Caubet (2004: 71), writing about Fezzan, says that it is an example of divergence between sedentary or urban and nomadic vernaculars; interdentals do not occur in 84

URBANIZATION AND DIALECT CHANGE

“sedentary” texts. Bedouin texts on the other hand do contain interdentals, with a t/• – d/O opposition, and a single emphatic 1. With regard to Cyrenaica, Panetta (1943) noted that in the Benghazi dialect the interdental sibilants *ª and *o are replaced by the plosives [t] and [d] and that the emphatic is [\], while Owens, writing in the 1980s noted the presence of interdentals (p. 7): θi’gi:l “heavy”, ik’θir “it increased” . . . ’Oe:l “tail”, ’yikOib “he lies” and (p. 8) ’1ayyig “narrow”, ’yu1rub “he hits”. The voiceless dental occlusive t seems to have been realized as an invariable [t] in TA: t¢s∞a “nine”, rBt “I saw”. However, the palatalization of *t to [n], realized as a voiceless dental africative with hissing offset, was observed at the beginning of the century in the Jewish dialect (Cesaro 1939; Goldberg 1983; P. Marçais 1977; Saada 1984; Yoda 2005). This has since disappeared and is now absent from the dialects spoken in Cyrenaica (Owens 1983; Panetta 1943) and in Fezzan (Caubet 2004). The phonemes /b/ and /d/ are unlike the three other long vowels /a/, /c/ and /e/, as they represent the reduction of the historical diphthongs ay and aw: zBt (< *zayt) “oil”, LBb (< *Layb) “pocket”, bBn (< *bayn) “between”, yDm (< *yawm) “day”, rDKen (< *rawKen) “window”, dDg (< *dawg) “taste”. The reduction of the diphthongs ay and aw to B and D respectively is a feature which is also found in certain nomadic vernaculars of North Africa. Their reduction to C and E, as found in pre-Hilalian dialects, does not occur, such as }EK “house”, biN “room” (Yoda 2005: 92). In Fezzan, (Caubet 2004: 72; P. Marçais 1977: 17), there is said to be free variation between the diphthongs Sw, Qy and complete reduction to D, B; the former often being the choice of female speakers. There again however, the consonantal environment can play a role in favor of the diphthongs: XSwm “fast” (rather than XDm), XQyf “summer” (rather than XBf ). In Benghazi (Owens 1983: 10; Panetta 1943: 17), following the TA model, the diphthongs *aw and *ay are reduced to D and B. To sum up, Bedouin and pre-Hilalian phonetic features alternate in TA, but Bedouin features predominate. Thus the [g] realization of qAf and the behavior of diphthongs which occur in this dialect are peculiar to Bedouin speech patterns, while the only pre-Hilalian feature is the absence of interdentals; the [q] realization of qAf being specific to the ancient Jewish dialect. 4.5.2 Morphology Like Bedouin type dialects, TA makes a gender distinction in the second person singular of the verb inflection for prefixal and suffixal conjugations and the imperative and for the inflections of independent pronouns: • •

klBt (masc.) / klBti (fem.) “you ate” tLCb (masc.) / tLCbi (fem.) “you bring” 85

CHRISTOPHE PEREIRA

• •

dCr (masc.) / dCri (fem.) “do” anta (masc.) / enti (fem.) “you”

Conversely, some pre-Hilalian dialects make this second person singular gender distinction neither for the verb nor for pronouns. Turning to Fezzan and Cyrenaica, there is a gender distinction for both second and third persons, singular and plural, in both prefixal and suffixal verbal inflections, and in the imperative, in independent pronoun inflexions, but also for suffixal pronouns. (For Fezzan see the examples in Caubet 2004: 79–84; for Cyrenaica see Panetta 1943: 185–186 and also Owens 1984: pp. 91–92 (pronouns) and pp. 108 ff for the conjugations). With regard to the Jewish vernacular, there is a gender distinction: nNa and nNi “you” (Yoda 2005: 115), kN¢bN and kN¢bNi “tu as écrit” (Yoda 2005: 146), N¢bda and N¢bday “you begin” (Yoda 2005: 148), N∞¢ll¢m and N∞¢llmi “learn” (Yoda 2005: 151). As to pronominal morphology, the 3rd person masculine singular personal pronoun suffix is usually -a or -¢h with a pausal form and never -u (unlike the pre-Hilalian dialects): e.g. ¢l-xAl fi xBW lBn yukbuW l-a wuld-uxt-¢h “the uncle is fine until his nephew has grown up”; -a is lengthened when a suffix is added, e.g. mA-gAl l-A-K “he did not tell him”. In the Jewish vernacular the pronoun is –u after consonants and –h after vowels, as in pre-Hilalian dialects: qElu l-émmu ta‘mél l-u qofYânu (Cohen 1930: 110) “tell his mother to make his kaftan”, eK sâbb el-,âli râ}, mâ LAbûh (Cohen 1930: 172) “the dear child has got lost; they have not brought him back” and |¢W{ú “they beat him” (Yoda 2005: 115–118). The verb conjugation is not of the pre-Hilalian dialect type. In defective verbs there is no reconstruction of the paradigm, with forms of the type t¢lqAy, y¢bdAw, KrAw, mKAt, in which the long vowel A appears. In the prefixal conjugation the following forms occur: telgi “you find (fem)”; neKru “we buy”; telgu “you find (masc. and fem.)”; yebdu they begin (masc. and fem.)”. In the suffixal conjugation is found Kru “they bought (masc. and fem.)”; in the third person feminine singular, TA has the form C1C2¢t found in Bedouin vernaculars: mK¢t “she left”, lg¢t “she found”, ns¢t “she forgot”. In the pre-Hilalian Jewish dialect however, these forms occur with long vowels: en-nâs isemmCûk “people call you” (Cohen 1930: 36 and Yoda 2005: 174). The same is true for Cyrenaica (cf. Panetta 1943: 227). In Owens and in Fezzan on the other hand, diphthongized forms are found, such as ’nalgaw “we find” or ’algay “find (feminine singular imperative)” (Owens 1984: 114) and t(bdPy “you begin (fem.)”, n(bdPw “we begin”, t(bdPw “you begin (masc.)” and y(bdPw “they begin” (Caubet 2004: 78). In present day TA, as in all the Arabic dialects found in Libya, the diminutive in the triliterals is formed on the pattern C1C2BC3(a): ∞wBla “little family”, }wBK “little house”, rwBs “little head”, l}BmAt “pieces of meat”. The pattern C1C2BC3 peculiar to certain Bedouin dialects stands in opposition to 86

URBANIZATION AND DIALECT CHANGE

the pattern C1C2¢yy¢C3 found in pre-Hilalian dialects: u-mlâ-lu Lwîjibu “and fill his little pocket” (Cohen 1930: 138) and Xif¢r “small birds” (Yoda 2005: 245). Quadriliterals formed on the pattern C1vC2C3AC4 (with a long vowel between the 3rd and 4th root consonants) form their diminutive on the pattern C1C2BC3CC4 (with a long c between C3 and C4): m¢ftA} “key” > mfBtC} “little keys”, s¢rwAl “trousers”> srBwCl “short trousers”, K¢bbAk “window grid” > KbBbCk “little window grid”, s¢ndEg “box” > snBdCg “little box” (Caubet 2004: 86; Panetta 1943: 89; Owens 1983: 76). In this respect TA is closer to Bedouin dialects as argued by William Marçais (1950): “certain features indicate an opposition between the whole Bedouin group and the ‘sedentary’ group; the pre-Hilalian f∞îlel contrasts with a Bedouin paradigm f∞êlîl.” Quadriliterals formed on the same pattern C1vC2C3AC4, form their plural on the pattern C1C2AC3CC4 (with a long c between C3 and C4). Examples: m¢ftA} “key” > mfAtC} “keys”, muXm0r “nail” > mXAmCr “nails”, K¢bbAk “window grid”, KbAbCk “window grids”. However in Griffini’s lexicon published in 19139, certain entries propose plurals and diminutives for quadriliterals formed on the pattern C1vC2C3AC4 and on the pre-Hilalian patterns C1C2AC3¢C4 and C1C2BC3¢C4. Cohen (1930: 86), offers the example kmAm¢n “cumin” and Yoda offers fWAW¢L “coqs”, X{a{¢Y “chaussures” (Yoda 2005: 241). In Fezzan, we find the two patterns CvCACCC and CvCAC¢C, used in both nomadic and sedentary vernaculars, where this had previously been a distinctive difference between them (cf. Caubet 2004: 86). The same is true for Cyrenaica (cf. Panetta 1943: 73–75; Owens 1983: 66–67). Color adjectives formed on the pattern aC1C2vC3 form their diminutives on the pattern iC1CC2vC3. This is restricted to color names and to Bedouin type dialects, with an initial i vowel. Examples: a}maW “red” > i}CmaW, azrag “blue” > izCrag, ax|¢W “green” > ixC|+r, ak}al “black” > ikC}al. For color adjectives in Fezzan, cf. Caubet 2004: 87 and, for Cyrenaica, Panetta 1943: 108–109; Owens 1983: 57–58; though they do not deal with the diminutives of these adjectives. Hans Stumme (1898: 61) and Antonio Cesàro (1939: 106) describe the duel as limited to nouns expressing notions envisaged as one element of a pair. Names of units of measure and of time take the duel, as do those for double body parts. The duel is formed by suffixing –Bn to nouns. Examples: name of the mother-father pair: wAldBn; units of time: dgCgtBn “two minutes”, yDmBn “two days”, KahrBn “two months”, sA∞tBn “two hours”, ∞AmBn “two years”, lBltBn “two nights”; units of measure: LdEltBn (LdEla, surface area measure, plot), kBltBn (volume measure for grain or a liquid), drA∞Bn “two cubits”, mCtBn “two hundred”, AlfBn “two thousand”; double body parts: wudnBn “two ears”, CdBn “(two) hands”, drA∞Bn “two arms”, kur∞Bn “(two) legs”, r¢LlBn “(two) feet”, ketfBn “two shoulders”. For some double body parts, the duel serves (also) as a plural; others have an internal plural. 87

CHRISTOPHE PEREIRA

Examples: wudnBn “two ears” but udAn “ears”, drA∞Bn “two arms” but der∞An “arms”, ketfBn “two shoulders” but ktAf “shoulders”. However, during my last visit to Tripoli, in February 2005, I observed that any noun can take a duel agreement: Krubt gahwutBn “I drank two coffees”, Kba}t sCyAWtBn “I saw two cars”: fta} bAbBn “he opened two doors”, gW¢t ktAbBn “she read two books”. It does not depend on the social class of the speaker, nor the speech register. In Tripoli, when the duel is not used, it is replaced by the numeral zDz “two” followed by a plural noun. Examples: zDz }EtAt “two fishes”, zDz bCbAn “two doors”, zDz knAsa “two brooms” (cf. also, for Cyrenaica, Panetta 1943: 82–85; Owens 1983: 59). In Fezzan, the suffix is -ãn, but it is not often used in texts; it is more frequent with names of units of measure and double body parts (Caubet 2004: 87). Yoda highlights that duel is limited to nouns expressing units of measure and of time and to say 200 and 2000 (Yoda 2005: 204). In TA, morphology alternates Bedouin and pre-Hilalian features, with a preponderance of the former. Features common to Bedouin type dialects include the treatment of the second person singular, defective verbs and diminutives. The frozen use of the duel is more a pre-Hilalian feature but the duel is currently tending to spread in Tripoli.

4.5.3 Syntax In TA the indefinite is marked by the article Ø: bCr “a well”, WAL¢l “a man”, LWAna “a frog”, }ALa “a thing”, muKkila “a problem”, kWAsi “chairs”, m+wA|C∞ “subjects”. An indefinite article wA}d-¢l- has not developed in this dialect, such as occurs in pre-Hilalian dialects in Morocco and Algeria (for dialects of Cyrenaica cf. Panetta 1943: 53; Owens 1983: 52). Sumikazu Yoda does not treat the indefinite article in his study, but we can read }uK kbCra “a big house” (p. 197). The possessive relationship is expressed by coexisting structures of a more Bedouin type (a synthetic construction) and pre-Hilalian type analytic structures. This coexistence is observed in all speakers, whatever their social origin. The synthetic possessive construction occurs quite frequently in TA. Its use does not seem to be restricted to groups of naturally associated notions, to kinship expressions and body parts. Examples: Ax¢r-¢n-nhAW “the end of the day”, taYyCb-¢l-l}am “the cooking of the meat”, l}am-l¢-xrEf “lamb meat”, }DK-L¢dd-i “my grandfather’s house”, YunL ¢rt-¢l-marga and ma∞En-¢l-marga “the sauce boat”, YarCq¢t-mak¢lt-ah “the way of eating”, Low¢l-lamma “the atmosphere of the meeting”. With suffixal pronouns: lDn-ah “its colour”, CdB-ya “my hands”, YbCxt-ah “its sauce”, xAl-i and ∞amm-i “my uncle”, baYn-¢k “your belly”, Low-na “our atmosphere”. 88

URBANIZATION AND DIALECT CHANGE

TA also uses the analytic construction as a possession marker. The particle used in Tripoli is mtAε. Examples: ¢l-marga mtA∞-ah / ¢Y-YbCxa mtA∞-ah “its sauce”, ¢l-Low mtA∞ mak¢lt-¢l-bAzCn “the atmosphere of the meal”, l a∞|Am mtA∞-ah “his bones”, ¢l-wAL ¢b ¢r-ra’isCya mtA∞ }ne “our main meal” – in this latter example it should be noted that the particle mtA∞ is followed not by the first person plural suffixal pronoun –na, but by the first person plural independent personal pronoun }ne (for Fezzan cf. Caubet 2004: 88; and, for Cyrenaica, Panetta 1943: 128–133; Owens 1983: 82). In the speech of the Jews, the particle used is nNa, (Yoda 2005: 261). In TA, syntax, like the morphology and the phonetics, alternates Bedouin and pre-Hilalian features. The use of the article Ø to mark indefiniteness is a Bedouin feature, whereas the use of the analytic construction to express the possessive relationship is pre-Hilalian.

4.5.4 Lexis TA lexis reflects various strata and borrowings. At the beginning of the twentieth century, pre-Hilalian and Bedouin items alternated, as shown in the writings of Stumme (1894, 1898), Cesàro (1939, 1954) and Griffini (1913)10 and in the table below. Changes of lexis are to be noted when comparing the dialect described by the aforementioned writers and that of the present day. Tripolitans used lexical items from Bedouin type vernaculars; other items are not used by my Libyan informants,11 with the exception of the verb Wa “he saw”, used only in the first and second persons of the suffixal conjugation.

Lexical items

Present day TA

TA as described by Stumme, Griffini and Cesàro

To do To enter To want, desire To see, look at To return, come back To put Hand

dAr, idCr xaKK, ixuKK ba, ye{{i Wa ~ Kba}, yeKba} wålla, iwulli dAr, idCr Cd

∞mel, ya∞mel ~ dAr, idCr dxal, yudxul ~ xaKK, ixuKK }abb, i}ubb ~ ba, ye{{i Wa ~ Kba}, yeKba} rLa∞, yerLa∞ ~ wålla, iwulli }åYY, i}+YY ~ dAr, idCr yedd ~ Cd

Figure 4.1 Lexical evolution of TA

Pre-Hilalian items occur in Judeo-Arabic texts (Cohen 1930): n}abb mîat LCnni u Linnîa “I want one hundred geniuses and one hundred fairies” (voglio cento genî e cento fate); n}abb mîat Linn u-Linn câçi “I want one hundred geniuses and one hundred rebellious geniuses” (voglio cento genî u cento genî ribelli); ,êr mektûbi dLCbû-hû-li bîn jiddîja “I only want that you lead me 89

CHRISTOPHE PEREIRA

to my predestined spouse” (voglio solo che mi conduciate dinanzi lo sposo predestinato); where n}abb has the meaning “I want” and jidd means “hand”, which are pre-Hilalian items (cf. Yoda 2005, “Glossary”: 313–347). TA does not appear to have been influenced by Berber but in contemporary TA, we find those two words that are from Berber origin: smCtri and sfCngri with the meaning “very cold weather”, are very commonly used. The Ottoman presence in Libya (1551 to 1911) however, and the Italian occupation (1911 to 1951) did effect Libyan dialects, especially urban ones. In the Arabic vernacular of Tripoli are to be found many Turkish (Türkmen 1988) and Italian (Abdu 1988) borrowings. These borrowings, found particularly in the administrative and military domains, are now used less and less, being replaced by Arabic items. Given the political importance attached to arabization and re-arabization in the public sphere, the work place and in education, Turkish and Italian items are replaced by Standard Arabic and there are also many borrowings from Standard Arabic in the Arabic dialect of Tripoli. Nonetheless, numerous borrowings from Turkish and Italian do still occur in the names of tools and utensils, clothing and the culinary field: kAKCk “spoon” (< kawık), KCKa “bottle” (< wiwe), KanYa “bag, suitcase” (< çanta), kubri “bridge” (< köprü); bAK (< baw) is a mark of respect or superior status, as in bAK-muhandis “senior engineer”, sYa “master, specialist” (< usta), bAmya “okra” (< bamya), are examples of Turkish borrowings; marKabBdi “pavement” (< marciapiede), kaNafCti “screwdriver” (< cacciavite), UibDto “jacket” (< giubotto), kanNBlo “iron gate” (< cancello), simAforo “traffic lights” (< semaforo), Yasa “glass” (< tazza), are Italian borrowings frequently used by my Tripolitan informants. In addition the majority of terms used in the field of motor mechanics are borrowed from Italian: dDmAni “steering wheel” (< due mani), stArsu “steering wheel” (< sterzo), frBno “brake” (< freno), frBnDmAno “hand brake” (< freno a mano) together with many names of fish, such as12: sardCna “sardine” (< sardina), anKEga “anchovy” (< acciuga), lEbbu “sea bass” (< lupo), sAn byCtro “John Dory” (< San Pietro), angwCla “eel” (< anguilla) and mArlEtsu “cod” (< merluzzo). There are also items of Middle Eastern origin (Egyptian Arabic) to be found in TA, but they have been lexicalized and so do not appear as borrowings, for example kwayy¢s “good”, fi “there is/are”, b¢ss “only”, byithayyA l-i “in my opinion”, bardo “also”, which appear in my corpus. These items do not appear in older descriptions. They are very common today in Tripoli. For “there is”, Stumme et al. mention tamma (still found in Tunisia in the form of [ªamma] and which Tripolitan Jews pronounced [namma]), but this item no longer occurs in Tripoli, where only fi occurs. kwayy¢s occurs as frequently as bAhi et mlC} “good”, if not more so; for example when one asks in Tripoli “How are you? Are you well?”, the question is KCn ¢L-Low? KCn }Al-¢k? anta kwayy¢s?

90

URBANIZATION AND DIALECT CHANGE

Young men in present day Tripoli use many English borrowings (my informants are all male, so I cannot say whether the same is true for females). This may be due to the influence of their university curricula (science, technology, IT), new technologies (computers, mobile phones, etc.), music and cinema; other items including swear words are also borrowed from English (Quitout 2004: 10–13): the expression so nice is used by my informants, mo{ayl “mobile phone” (< mobile) is used alongside naqqAl, kombyEt¢r “computer” (< computer), the verb fuck (transcribed and pronouced fœk) is very common, together with the verb nAk, inCk, as in nfœk “I’m stuck with (someone or something)”: nnCku fi g/m-had ¢rza “we’re stuck with this discussion”, nfœk fi g/m-biliyArdi “I’m stuck with having to play billiards”; where g/m is an English borrowing (< game) and biliyArdi is from Italian (< biliardi).

4.6 Conclusion We have seen that from the eighteenth century onward the city of Tripoli has experienced an important period of urban transformation due to population movements that occurred on an ever increasing scale in the second half of the twentieth century. The Arabic dialect of Tripoli has become bedouinized as a result of the huge rural exodus and the settlement in the city of Bedouins and former peasants. These were no doubt the causes of the emergence of a Tripolitan dialectal koine. The Arabic dialect of Tripoli is linguistically hybrid, a koine which has nonetheless preserved certain preHilalian features and which also borrowed numerous items from Turkish and Italian – traces of the significant Ottoman and Italian occupations – though these are now being replaced by terms from Classical Arabic, augmented more recently by borrowings from Egyptian and other Arabic dialects and from Standard Arabic. Tripolitan Arabic is changing, ever more influenced by Standard Arabic but also by borrowings from foreign languages, especially English, resulting on the one hand from the uncompromising arabization policy and on the other from Libyan students in higher education studying science and technology subjects, where English plays a prominent role, as well as involvement with the new technologies – computing, internet, mobile phones, etc. The Arabic dialect of Tripoli does not seem to have achieved the status of a prestigious “national koine” to be imitated throughout the country. It appears that in Libya – and I say this on the basis of field observations and conversations with my informants – every speaker is proud of his or her origins and dialect and does not seek to imitate the Tripolitan vernacular, even when living in the city itself. Some of my informants, from other towns or other regions of Libya (Derna or Benghazi, in Cyrenaica, or from inland Tripolitania, Misurata, Syrte, or Fezzân), living temporarily in Tripoli, were

91

CHRISTOPHE PEREIRA

keen to explain this to me and to point out the most important phonological and lexical differences. One informant for example, who came from Benghazi, told me that hAlba “much” is Tripolitan and that he said wAL¢d, that in Benghazi interdentals are pronounced, that the *t there is pronounced [ty], with slight palatalization, that instead of YAbla “ash tray” they say Y¢ffAya or Y¢ffAy¢t-s ¢bAsi, that there is a gender distinction in the second and third person plural of pronouns and in the verb conjugation. The present government’s rhetoric on “Bedouin legitimacy” must surely have had an influence on people’s use of their own regional dialects, rural or otherwise, when they move to Tripoli. Unlike Rabat, Fez or Salé, in Morocco (Messaoudi 2003), it seems there has never been in Libya a myth of an ancient urban elite. It is indeed doubtful whether there ever existed a prestigious native urban culture in Tripoli. Throughout the nineteenth century the elite consisted of Ottomans, Italians or other Europeans and a few Tripolitan families associated with those in power. Gradually, following the departure of the Ottomans in the nineteenth century, then of the Italians and other Europeans, part of the urban elite disappeared. At independence, King Idriss ruled over a much weakened country, both economically and politically. The capital had been moved to Bayda in Cyrenaica, which certainly caused part of the elite to leave Tripoli. When Gadafi took power in 1969, what remained of the Tripolitan elite left the country. Their ideas were opposed to those of Gadafi’s “people’s socialist republic of the masses”. Conversely, there were large scale population movements and the settling of peasants in Tripoli so that the urban population consisted predominantly of former country dwellers and Bedouins. Does there still exist an ancient declining urban elite in Tripoli? According to my informants, Tripolitans whose families come originally from Tripoli, going back several generations and who used to live in the Medina have now settled in the outer suburbs such as sEg-¢L-Lum∞a, g¢rgAr¢K or tALEra, which used to be outside the town but which are now part of the urban region of Tripoli. Gadafi, known as the “Shepherd of Syrtes”, is of course from Syrtes, of rural Bedouin origin and those who hold political and economic power in Libya are similarly of Bedouin origin. This is an important factor which cuts across the question of Bedouin legitimacy, in contrast to the official rhetoric. The majority of the population is rural and Bedouin by origin and 85 percent of the population is now urbanized.13 Gadafi wanted to demolish the Tripoli Medina in the 1980s which by then was completely devoid of its original population, but blocks of flats were built to house migrants arriving from rural areas. There is now an association for the preservation of Tripoli old town (maKrE∞ tan1Cm wa §idAra al-madCna l-qadCma YarAbulus), which is restoring the Medina. The historical urban inheritance is once more seen as something to be valued. 92

URBANIZATION AND DIALECT CHANGE

With regard to culture, young Tripolitans listen to a lot of foreign music, especially rock, rap and R&B, as well as “oriental” Arabic music from Egypt, Lebanon and the Gulf States, but between 2002 and 2005 I noticed that recordings of ancient Tripolitan music were being re-issued and were much enjoyed by my young informants. Here again we are witnessing the revaluation of the historical urban inheritance. In 2002, I was unable to buy any ancient Tripolitan music; the music shops in Tripoli did not stock it. In 2005, I was able to find it easily. The lyrics are in the Libyan Arabic dialect and it has experienced a big rise in popularity. There is also a large output of reggae sung in Libyan Arabic. The words of these songs need to be analyzed to see what type of Arabic is used and to see whether there are any innovations in the language. The lyrics tend to be nostalgic and romantic, not at all in revolt against the present government. Some allude to a bored generation. Could this be the beginning of a revival of a certain urban city culture?

Notes 1 For a synthesis of the literature on TA see Pereira 2003, 2004 and 2006. 2 The Ancient Libycs were the ancestors of the Berbers. Their vernacular was a Hamito-Semitic language from which modern Berber dialects are derived. The first Lybico-Berber inscriptions appeared around 1500–1000 bc (Larcher 2001: 43). 3 Marçais (1961) states that “in Cyrenaica the old language is still spoken by approximately 1.5 percent of the inhabitants. Tripolitania on the other hand has 23 percent Berber speakers”. Berber is at present the native language of 15–20 percent of the Libyan population. Berber-speakers are concentrated in Tripolitania in the coastal town of Zouara and the inland towns of Djebel Nefoussa and the Ghadames oasis. According to some researchers there are also Berber-speaking groups in the oases of Joufra (Larcher 2001: 46–48; Souriau 1986: 40). 4 The Ottomans occupied Libya without interruption from 1551 to 1911, divided into three periods: 1551 to 1711, 1711 to 1835 (the Karamanli Dynasty) and 1835 to 1911. 5 According to The National Report Presented to the International Conference on Education, held in Geneva in 2001 (The Development of Education in the Great Jamahiria: 73), there are 2,814 Libyan teachers in Libyan universities, compared with 2,714 foreign teachers and 1,403 Libyan teachers in professional training institutions, as against 2,234 foreign teachers. 6 According to Sumikazu Yoda (Yoda 2005) “Jews in North Africa adopted Arabic as their native speech during the first (pre-Hilalian) period and their dialects therefore preserve archaic features no longer present in the dialects of their Muslim neighbors. The Jewish dialects are also distinguished by the use of many words of Hebrew and Aramaic origin. In Tripoli the difference between the Jewish and Muslim vernaculars manifests itself not only in the vocabulary but also in the language type: The Jewish dialect represents the sedentary type while the Muslim dialect belongs to the Bedouin type.” 7 According to Rachel Simon (1992), the number of Jews in Tripoli was: 12,000 in 1906, 17,196 in 1936, 19,330 in 1945, 21,000 in 1948, and 3,580 in 1954. 8 The examples are from Mordekhai Cohen (1930) and are written in this chapter with the same transcription as in the book (pp. 36, 38, 41, 51, 78).

93

CHRISTOPHE PEREIRA

9 L’arabo parlato della Libia by Griffini (1913), despite its title, deals with TA. Cf. the sub-title “Cenni grammaticali e repertorio di oltre 10.000 vocaboli, frasi e modi di dire raccolti in Tripolitania”, p. III and “Gli appunti sono stati presi durante i miei primi cinque mesi di soggiorno a Tripoli (3 gennaio-3 giugno 1912)”, p. VII. 10 Regarding lexis, these writers deal with the TA spoken by Muslims. 11 My Libyan informants consider these words as belonging to the lexis of Tunisian Arabic. They appear in the classifications of William and Philippe Marçais as belonging to pre-Hilalian type dialects, but are in fact at present items that can be considered as pan-dialectal, especially because many of them are close to Standard Arabic, because in Tripoli they are used by the many migrants originating from various different parts of the Arab world, and also because of the influence of the media. 12 Cf. Dalcl muxawwar li ba¡\ al-’asmak al-™iqtixadcya bi myah al-lamahircya al-¡arabcya al-lcbcya ak-ka¡bcya al-™iktirakcya al-¡u2ma, mankerat markaz bu]eth al-™a]ya’ al-ba]rcya, 1990 [in Arabic]. 13 This is somewhat similar to the situation seen in the Gulf States, where some writers have referred to a “Bedouin urban model”.

References Abdu, H. R. (1988) “Italian Loanwords in Colloquial Libyan Arabic as Spoken in the Tripoli Region”, unpublished thesis, University of Arizona. Agostini, E. de (1954) “Sulle popolazioni della Libia”, Libia, II (anno 2): 3–13. Azema, J. (2000) Libya Handbook, Bath: Footprint. Bessis, J. (1986) La Libye contemporaine, Paris: L’Harmattan. Burgat, F. and Laronde, A. (2000) La Libye, Que sais-je? 1634, Paris: PUF. Caubet, D. (2001) “Questionnaire de dialectologie du Maghreb”, EDNA (Estudios de Dialectología Norteafricana y Andalusí) 5: 73–92. —— (2004) “Les parlers arabes nomades et sédentaires du Fezzân, d’après William et Philippe Marçais”, in M. Haak, K. Versteegh and R. Dejong (eds) Approaches to Arabic Dialects: A Collection of Articles Presented to Manfred Woidich on the Occasion of his Sixtieth Birthday, Leiden-Boston: Brill, pp. 67–96. Cesàro, A. (1939) L’arabo parlato a Tripoli, Milan: Mondadori. —— (1954) “Raconti in dialetto tripolino”, Annali di Instituto Universitaro di Napoli 4: 49–59. Cohen, M. (1930) Gli ebrei in Libia: usi e costumi tradotto e annotato da Martino Mario Moreno, Roma: Sindicati Italiano Arti Grafiche. Djaziri, M. (1996) Etat et société en Libye, Paris: L’Harmattan. Fontaine, J. (2004) “La population libyenne, un demi-siècle de mutations”, in La nouvelle Libye: sociétés, espaces et geopolitique au lendemain de l’embargo, Paris: Karthala, pp. 159–175. Goldberg, H. E. (1978) “The Jewish wedding in Tripolitania: a study in cultural sources”, The Maghreb Review 3–9: 1–6. —— (1983) “Language and culture of the Jews of Tripolitania: a Preliminary View”, Mediterranean Language Review I: 85–102. Griffini, E. (1913) L’arabo parlato della Libia, Milan: Ulrico Hoepli. Ham, A. (2002) Libya, Melbourne: Lonely Planet.

94

URBANIZATION AND DIALECT CHANGE

Lafi, N. (2002) Une ville du Maghreb entre ancien régime et réformes ottomanes Genèse des institutions municipales à Tripoli de Barbarie (1795–1911), Paris: L’Harmattan. Larcher, P. (2001) “Les langues de la Libye: passé et présent”, La revue des deux rives 2: 43–51. Mantran, R. (1975) “La Libye des origines à 1912”, in La Libye nouvelle: rupture et continuité, Paris: CNRS, pp. 15–32. Marçais, W. (1950) “Les parlers arabes”, in Initiation à la Tunisie, Paris: Adrien Maisonneuve, pp. 195–219. —— (1961) “Comment l’Afrique du Nord a été arabisée”, in Articles et conférences, Paris: Adrien Maisonneuve, pp. 171–192. Marçais, P. (1957) “Les parlers arabes”, in Initiation à l’Algérie, Paris: Adrien Maisonneuve, pp. 215–237. —— (1977) Esquisse grammaticale de l’arabe maghrébin, Paris: Adrien Maisonneuve. —— (2001) Parlers arabes du Fezzân, textes, traductions et éléments de morphologie rassemblés et présentés par Dominique CAUBET, Aubert MARTIN et Laurence DENOOZ, Université de Liège, Diffusion Librairie Droz: Geneva. Martel, A. (1991) La Libye 1835–1990: essai de géopolitique historique, Paris: PUF. Martin, Y. (1975) “La Libye de 1912 à 1969”, in La Libye nouvelle: rupture et continuité, Paris: CNRS, pp. 33–50. Messaoudi, L. (2003) Etudes sociolinguistiques, Editions OKAD, Kénitra: Université Ibn Tofail. Miller, C. (2004) “Variation and Change in Arabic Urban Vernaculars”, in M. Haak, K. Versteegh and R. Dejong (eds) Approaches to Arabic Dialects: A Collection of Articles Presented to Manfred Woidich on the Occasion of his Sixtieth Birthday, Leiden-Boston: Brill, pp. 177–206. Owens, J. (1983) “Libyan Arabic dialects”, Orbis, XXXII, 1–2: 97–117. —— (1984) A Short Reference Grammar of Eastern Libyan Arabic, Wiesbaden: Otto Harrassowitz. Panetta, E. (1943) L’arabo parlato a Bengasi, I. Testi con traduzione e note, II. Grammatica, Rome: La Libreria dello Stato. Pereira, C. (2003) “Le parler arabe de Tripoli (Libye), etat des lieux (d’après les travaux de Hans Stumme, Antonio Cesàro, Eugenio Griffini . . .)”, EDNA 7: 7–57. —— (2004) “Le parler arabe de Tripoli, un parler mixte”, in I. Ferrando and J. J. Sanchez Sandoral (eds) AIDA Fifth Conference Proceedings, 431–443. —— (2006) “Eléments de description du parler arabe de Tripoli”, in L’arabe dialectal: enquêtes, descriptions, interprétations, Actes d’AIDA 6, Cahier du CERES (Centre d’Etudes et de Recherches Economiques et Sociales), Série Linguistique no. 13, Tunis. —— (forthcoming) “Arabic of Libya”, in EALL, Vol. III, Leiden: Brill. —— (forthcoming) “Tripoli Arabic”, in EALL (Encyclopedia of Arabic Language and Linguistics), Vol. IV, Leiden: Brill. Quitout, M. (2004) “Situation linguistique en Libye (Histoire, statut et usage des langues locales et étrangères)”, Folia Orientalia 40: 313–326. Rossi, E. (1954) “Storia della Libia dalla conquista araba al 1911”, Libia, I (anno 2): 3–43.

95

CHRISTOPHE PEREIRA

—— (1968) Storia di Tripoli e della Tripolitania dalla conquista araba al 1911, Roma: Instituto per l’Oriente. Saada, L. (1984) Eléments de description du parler arabe de Tozeur (Tunisie), Paris: Geuthner. Simon, R. (1992) Change Within Tradition Among Jewish Women in Libya, Seattle and London: University of Washington Press. Souriau, C. (1975) “La Libye moderne”, in La Libye nouvelle: rupture et continuité, Paris: CNRS, pp. 135–160. —— (1986) Libye l’économie des femmes, Paris: L’Harmattan. Stumme, H. (1894) Tripolitanisch-tunisische Beduinenlieder, Leipzig: J. C. Heinrich’sche Buchhandlung. —— (1898) Märchen und Gedichte aus der Stadt Tripolis in Nordafrika, Leipzig: Heinrich’sche Buchhandlung. Taine-Cheikh, C. (1998) “Deux macro-discriminants de la dialectologie arabe (la réalisation du qâf et des interdentales)”, MAS-GELLAS 9: 11–50. Trombetti A. (1912) Manuale dell’arabo parlato a Tripoli, Bologna: L. Beltrami. Türkmen, E. (1988) “Turkish words in the Libyan dialect of Arabic”, Erdem 4–10: 227–243. Yoda, S. (2005) The Arabic Dialect of the Jews in Tripoli (Libya), Wiesbaden: Harrassowitz Verlag.

96

BECOMING CASABLANCAN

5 BECOMING CASABLANCAN Fessis in Casablanca as a case study Atiqa Hachimi

5.1 Introduction Right at the beginning of the twentieth century when Morocco was officially declared a French protectorate, the French decided to shift power to Morocco’s coastal areas (i.e. Rabat and Casablanca) at the expense of its interior areas (i.e. Fez and Marrakech). Rabat was made the administrative capital of the country and Casablanca its economic capital. This decision was a turning point in the history of Morocco in general and that of Casablanca in particular. Being promoted to the leading industrial, commercial and port city in Morocco, Casablanca emerged as a hub for migration attracting large numbers of people from urban and rural areas. As a result, the population of the city mushroomed from less than a million inhabitants in the first half of the twentieth century to more than four million people today, making it by far the largest Moroccan city. This intense human migration has led to important social, cultural and linguistic changes and has made Casablanca a region that offers important opportunities for research on relations between social and linguistic change in a fast changing complex society. One of the immediate consequences of the rapid urbanization of Casablanca has been the disruption of the rural/urban dichotomy that once dominated Moroccan dialects and identities. Mass migration to Casablanca has created a new linguistic urban model in contemporary Morocco quite different from the old urban model of the imperial cities. Thus it is important to recognize that today two competing local urban models co-exist. On one hand, there are the well-established urban dialects of the historical cities, which have defined urbane and prestigious linguistic practice in Morocco for centuries. These dialects have acted as badges of an urban elite and bourgeois identity. On the other hand, there are the newly formed koinés that have emerged as a result of massive internal rural and urban migration to large urban centers in the twentieth century. These koinés are 97

ATIQA HACHIMI

establishing themselves as new urban dialects and are shaping a new brand of urban identity. The best representatives of these types of dialects and identities in contemporary Morocco are, without doubt, Fessi (from the city of Fez) and Casablancan (from the city of Casablanca), respectively. The aim of this chapter is to examine the linguistic and social outcomes of the Fessi-Casablancan contact, considering in particular how Fessi women of Andalusian descent construct linguistic and social identity in Casablanca. The study seeks to determine the extent to which the leveling out, or maintenance, of differences between Fessi and Casablancan linguistic forms plays an important part in constructing both social identities and social meaning among Fessi women in Casablanca. The study of the Fessi-Casablancan linguistic and social contact and change is approached from the framework that views sociolinguistic variation and change as a social practice (Eckert 2000). The appeal of this model rests on its central focus on social meaning in explaining linguistic variation and change. The use of variation by speakers is seen as a symbolic resource in constructing social categories and social meaning rather than merely reflecting membership in predetermined social categories such as “class” and “sex”. Uncovering that a variant is correlated with certain social groups, say middle class or females, and not others, tells us that these groups use the linguistic form more frequently while other social groups use it less. This correlation stops short of revealing the meaning of this variant to the groups who use it more often and those who use it less often. Therefore, understanding the nuanced social meaning of the variants and their local relation to categories and social identity often depends on investigating speakers very closely in their communities of practice. Drawing on the work of Eckert (2000) and a number of other sociolinguistic studies that have shown speakers’ manipulation of socially significant linguistic features in giving meaning to various social identities, the study goes beyond the simple correlation of individual linguistic variables with social membership, and shows how socially meaningful linguistic variables are used as resources in the construction of social identities among Fessi migrants in Casablanca. One of the main goals of this study is to shed light on the interaction between migrants’ perception of their own identity and dialect change and/ or maintenance. It is believed that theories of dialect contact will provide more adequate explanations if social and sociopsychological factors are seriously taken into consideration. Therefore the study has integrated notions of social identity, in-group and out-group identification (Terry and Hogg 1996), and recent work on interspeaker accommodation (Giles and Coupland 1991) and the notion of ethnolinguistic vitality (e.g. Landry and Allard 1994) with linguistic factors from the outset of the investigation. Thus informants’ narratives were elicited about informants’ in-group and out-group identifications and their ideologies and attitudes towards social categorization and its embedded meanings. These narratives reveal a great deal of 98

Map 5.1 Casablanca (Morocco)

BECOMING CASABLANCAN

99

ATIQA HACHIMI

information about what it means to be or to become Casablancan, and what it takes not to be or become one. Latest research on language and identity emphasizes the agency of speakers in dismantling and constructing their own identities (e.g. Gubbins and Holt 2002). Based on informants’ self and group identification, this chapter identifies the categories that have emerged as a result of the Fessi-Casablancan contact, e.g. F wassa-d-Fes “pure Fessis”, F wassa-d-Kasa “Fessi-Casablancans” and Bi|awa/Kasawijjin “Casablancans”, and examines the social practices that shape and give meaning to these categories. In other words, this chapter examines what it means to remain “pure” Fessi, to become Fessi-Casablancan, or Casablancan and shows how linguistic variables are used as symbolic resources in the construction of these identities. I argue that variation in Fessi women’s accommodation to Casablancan linguistic norms plays an important part in creating both social categories and social meaning in Casablanca.

5.2 The urbanization of Casablanca Unlike Morocco’s imperial cities (e.g. Fez and Marrakesh), Casablanca cannot pride itself in being a bourgeois city with a long and rich history of urbanity. Its meager historical contribution has earned it the reputation of a city “without history and without soul” (Dernouny and Léonard 1987). In fact, the story of pre-colonial Casablanca is the story of a marginalized city that is remembered by a succession of destructions and hardships. In the seventh century when Islam reached Morocco, Casablanca was a Berber village known as Anfa “hill”. The inhabitants of Anfa are believed to have rejected the new religion of Islam and as a consequence it remained an independent kingdom for four centuries. In the eleventh century Anfa was occupied and destroyed by the Almoravides. However, under the Merenids in the fourteenth century Anfa stabilized and the town reportedly became important for exporting goods to far away regions but it never reached the rank of a Muslim city (Cohen and Eleb 2002). A century later Anfa became independent again and during this period it became a place for pirates who were attacking incoming ships. In retaliation, the Portuguese invaded and destroyed the town. However, in the sixteenth century the Portuguese decided to come back to the area and settle in it permanently. They built the city and named it Casa Branca “the white house”. In 1755, the Portuguese abandoned the city after an earthquake that destroyed it completely. After the departure of the Portuguese, Casablanca remained deserted until the Alaouite Sultan Sidi Mohammed Ben Abdellah rebuilt it near the end of the eighteenth century. He renamed it Addaru lbay|a§, which is the literal Arabic translation for Casa Branca. The city acquired its Spanish name Casa Blanca when Spanish companies established themselves in the city in 1781. Today the city is known by its Standard Arabic name addaru lbay|a§, and d-dar(l)bi|a in Moroccan Arabic, as well as Casablanca or Casa for short. 100

BECOMING CASABLANCAN

It is undeniable that after the construction of its small port in 1789 Casablanca’s commercial relations with Europe and America started developing, making it an important center in the Atlantic Ocean throughout the nineteenth century. However, its development as a major urban center started in earnest at the beginning of the twentieth century under French colonization. The urbanization of Casablanca is associated de facto with Louis Hubert Lyautey, the French Resident General in the country, who decided to turn the city into the leading industrial, commercial and seaport city in Morocco in 1912. The impact of Lyautey’s decision soon precipitated economic and social changes. As soon as Casablanca started showing signs of economic growth and prosperity, large numbers of people, including both foreigners and Moroccans, started migrating to the city. The migrant population of the first half of the twentieth century included Europeans who came from Europe or from other cities in North Africa, and Moroccans both Muslims and Jews. In 1950, Casablanca contained 414,500 Muslims, 158,000 Europeans and 72,000 Jews. After independence from the French in 1956 until the present, the population of Casablanca has seen dramatic growth. As Figure 5.1 shows, in less than half a century the population of the city went from less than a million to almost four million inhabitants today. Furthermore, the post-independence period in Casablanca is marked by a dramatic decrease in the number of the non-Muslim population and a remarkable increase in the number of the Muslim population. In 1913 for

4,020,000 3,520,000 3,020,000 2,520,000 2,020,000 1,520,000 1,020,000 520,000

18 97 19 07 19 13 19 18 19 21 19 26 19 31 19 36 19 47 19 50 19 51 19 60 19 82 19 94 19 99 20 03

20,000

Figure 5.1 Casablanca’s population growth from 1897 to 2003 Sources: Cohen and Eleb (2002: 286). Adapted from a graph illustrating Casablanca’s population growth between 1897 and 1950, RGPH (1982 and 1994) and CERED (1993).

101

ATIQA HACHIMI

instance the non-Muslims accounted for 49 percent of the population while the Muslims accounted for only 51 percent but in 1998 the Muslim population had risen to 99 percent of the population of the city (CERED 1998). What is remarkable however is that within the Muslim population there was an increase in rural migration to the city at the expense of urban migration. The strong industrial expansion between 1952 and 1960 and the subsequent creation of manufacturing jobs lured migrants from the countryside and small towns to stream to the city in search of jobs in factories. In short, while urban migration to the city characterized the first half of the century, rural migration characterizes the second half. 5.2.1 Major ethnolinguistic migrant groups of Casablanca The population of Casablanca today is made up of a mixture of people born in Casablanca and those born in other areas. We will consider here the three main ethnolinguistic groups that make up the Muslim population of Casablanca today; ∞rubis (the rurals), Soussis (the Berbers) and Fessis. Casablancans of ∞rubi origin make up the bulk of the population in the city, which explains the predominance of rural (historically Bedouin) features in the Casablancan dialect. ∞rubis are believed to be the descendants of the Bedouin Arab tribes that settled back in the twelfth and thirteenth centuries in the neighboring Atlantic plains of Chaouia, Doukkala and Chiadma and Lhaouz, where Greater Casablanca is now located. Originally the term ∞rubi referred to people who originate from these areas but the term today has come to include “unworldly” or “uncultured” behavior or persons. The second largest ethnolinguistic group in Casablanca is the Soussis; Berbers from the Sous region in the southern parts of the High-Atlas in the southwest of Morocco. They are also known as Shleu} in reference to the name of the Berber variety spoken in the Sous region “Tashl}it”. Thus, Soussis bring a different language to the city and most of them are bilingual in Tashl}it and Moroccan Arabic. Soussi-Casablancans are without doubt the economic and political rivals of Fessis in Casablanca. Fessis, known as F wassa or hal Fas “the people of Fez”, are the third largest ethnolinguistic group in Casablanca. They are counted with nonFessi urbanites as urban migrants. According to the 1971 census, urbanites represent 43.3 percent of all immigrants, but their number has been decreasing ever since. The Fessis in Casablanca are not as big in number as the other groups, like the ∞rubis and Soussis but they have big financial capacities and a complex and solid network of alliances. They have enjoyed, since independence, positions of power in the higher administration, politics, economy and international affairs (Adam 1968). In short, today Fessis are well represented in all sectors of power. Fessi identity must be viewed as an ethnolinguistic identity rather than a mere regional identity. Fessis, as Hillili (1987) rightly notes, are a 102

BECOMING CASABLANCAN

homogeneous group not necessarily defined by residence in Fez. In fact, birth in the city of Fez and a lifetime residence there do not bestow urban Fessi identity upon all residents. On the other hand, Fessis born in other cities even if they were there for generations would still be called Fessis. Fessi, therefore, refers to a specific group of aristocratic and bourgeois families from Fez, particularly those of Andalusian descent. Family names, for instance, are always reliable in identifying them. Among the oldest Fessi families in Casablanca are the families of Benzakour, Benkiran, Benjelloun, Bennis, Sqalli, Ya¡qoubi among others (Adam 1968). The first Fessi family, the Benzakour family, established itself in Casablanca beginning in 1830 (Le Tourneau 1965) and the other families have followed since 1880. Earlier accounts of the social role of Fessis in Casablanca hold that they “alone had retained the traditions of an old urban civilization.” According to Adam (1968) it is the Fessis who provided Casablanca with its urban bourgeoisie and its modern elite. Next to family names and several other social practices, Fessi dialect has also been a reliable badge in identifying and distinguishing Fessis from nonFessis. Fessi dialect is a well-established urban (historically sedentary) variety and there are several phonetic, morphological and lexical features that differentiate it from other Moroccan dialects (cf. Hillili 1979, 1987; Zemmama 1975). It is important to mention that Fessi, just like other traditional urban dialects in Morocco and other North African cities, has generally been preserved among old and illiterate speakers, especially among women (Caubet 1998; Iraqui-Sinaceur 1998; Jabeur 1996; Messaoudi 1998, 2002; Trabelsi 1988). The fact that Fessi speech has been maintained and used predominantly among women is largely responsible for associating it with being a feminine variety. It is no coincidence then that in the city of Fez today, young Fessi girls are encouraged to maintain stereotypical Fessi features in their speech while boys tend to adopt a more koiné speech considered more virile for men (Caubet 1993). In addition, Fessi dialect has traditionally been perceived as a “posh” dialect probably because it is the variety spoken by the Fessi bourgeoisie (Hillili 1987). The rich urbane past of the city and the cultural practices of its elite must have contributed to the construction of this prestige. Examining how Fessi women in Casablanca view the Fessi dialect today will shed light on whether this traditional urban prestige lives on or whether it is a thing of the past. 5.2.2 The dialect of Casablanca Casablancan dialect is not as old nor as well established a dialect as Fessi. At the beginning of the twentieth century when Casablanca had a population of less than 24,000 inhabitants, Kampffmeyer (1912) noted that the dialect of Casablanca and that of the rural population of the Chaouia tribes 103

ATIQA HACHIMI

surrounding Casablanca had basically the same rural features. By the 1950s, increased migration to the city from all corners of the country had allowed for contact between different regional dialects, urban and rural, and by this time the Casablancan dialect was described as a “hybrid” and “mixed” dialect (Brunot 1950). Towards the end of the twentieth century, it was described as an “interdialect” that has psychological reality for Moroccans (Moumine 1990). There is no doubt that migration and contact between migrants has greatly complicated the urban/rural split in present-day Casablanca and that a new dialect is forming, however, up to now there is hardly any work documenting the koinéization processes responsible for the emerging dialect in Casablanca. Yet, it is widely accepted that the dominant Casablancan variety has many rural features (Elbiad 1992; Heath 1989, 2002; Lévy 2002; Moumine 1990). In general, Casablancan dialect has received very little attention compared to dialects spoken in other major cities in the Arabic speaking world. Kampffmeyer (1912) is the first linguistic investigation of the dialect spoken in Casablanca. Although the main objective of Kampffmeyer is the study of Moroccan Arabic, the large body of data of spoken Casablancan made available in his work is invaluable in tracing the diachronic changes of the dialect from the beginning of the twentieth century until the present. Some studies have looked at the phonological aspects of the Casablancan dialect (Hamdi and Puech 1987; Khomsi 1975; Nejmi 1995), and there are also some short descriptions of the distinguishing linguistic features of the Casablancan variety (Aguadé 2003; Elbiad 1992). The most comprehensive sociolinguistic study of Casablancan dialect is Moumine (1990), which examines the sociolinguistic patterning of variation in the city following the classical Labovian paradigm. Another important study is Moumine (1995), which looks at attitudes towards the dialect of Casablanca. Moumine found that different regional groups have different attitudes vis-à-vis the dialect spoken in Casablanca. While Marrakshis (people from the city of Marrakesh) and Oujdis (people from the city of Oujda) had positive reactions toward Casablancan on several traits, Fessis in particular did not. Fessis found Casablancan to be “tough”, “masculine”, “uncultured”, “unfamiliar” and “not likeable”. Furthermore, Fessis were the only group of respondents to express greater sensitivity to the rural aspect of the Casablancan dialect (Moumine 1995: 76). To conclude, Casablancan is a recently formed immigrant koiné characterized by the predominance of several rural features, a variety considered particularly by Fessis to be rural, non-prestigious and masculine. In light of this, can we expect any type of accommodation toward Casablancan dialect among Fessi women? A prestige-oriented analysis would predict that Fessi women have no motivation to accommodate to the Casablancan variety because in doing so they would sound ∞rubi and “masculine”. The identity and meaning-oriented approach adopted in this study, on the other hand, 104

BECOMING CASABLANCAN

would not assume a priori that ∞rubi “rural”, “tough” and “masculine” are undesirable traits to aspire to, instead it allows for the local construction of meaning which is at the core of shaping one’s identity. Furthermore, such an approach would also account for speakers who accommodate to Fessi speech when they want or need to.

5.3 Methodology: data collection and informants Data for the present study come from in-depth ethnographic interviews that I carried out with fifteen Fessi women in Casablanca during a fourteenmonth period of fieldwork in 1999–2000. The first generation (i.e. not born in Casablanca) consists of seven Fessi women who have lived in Casablanca for at least seventeen years, which means that we are dealing with a case of long-term rather than short-term contact. The second and third generations, on the other hand, consist of eight Casablanca-born Fessis who were born to Fessi parents and who have lived all their lives in Casablanca. As Table 5.1 shows, the fifteen women not only represent different generations of Fessis but different age groups and educational backgrounds (Hachimi 2005). Table 5.1 Social profile of informants Informants

Age

Place of birth

Level of education

Hajja Fatima Nadia Keltoum Samia Amina Hayat Leila Khadija Ihsan Siham Kawtar Saida Fatma Zakia Lamia

70+ 36 41 39 45 50 61 29 28 24 21 56 26 24 22

Fez France Khenifra Fez Fez Khenifra Oujda Casablanca Casablanca Casablanca Casablanca Casablanca Casablanca Casablanca Casablanca

None University High school University College Junior high school Junior high school University University University High school Junior high school College College College

Being Moroccan myself and having an extensive network of family and friends in Casablanca made it relatively easy to have access to these communities since most of them are friends of friends or family. It is important to mention that I am neither Fessi nor native of Casablanca. Interestingly, however, I was believed to be Casablancan. The reason behind assigning me a Casablancan identity will become clearer when we discuss the social meanings associated with being Casablancan. 105

ATIQA HACHIMI

In order to examine Fessi women’s accommodation, or lack of accommodation, to the Casablancan linguistic forms, the study will assess women’s use of two phonological variables; (r) and (q), and one morphosyntactic variable (-i). These linguistic variables differentiate the Casablanca dialect from the Fessi dialect as shown in Table 5.2. Table 5.2 The linguistic variables Variables

Fessi variants

Casablancan variants

(q) (r)

[ʔ] ~ [q] [ ] Post-alveolar approximant [-Ø] Gender neutralization

[g] ~ [q] [r] Trilled alveolar [-i] Gender distinction

(-i) 2nd fem.sing. clitic on verbs

It is important to note that the phonological variables are highly stereotypical as Fessi while the morphosyntactic variable is not. In fact, it is below the level of social awareness. The Fessi approximant in particular is a minority form, i.e. a form found only among Fessis. Studies of dialect contact have repeatedly shown that minority and stereotypical forms are usually leveled in dialect contact. This would predict that [q] and [ ] are more likely to be left behind while gender neutralization is more likely to be carried over. This study will examine the linguistic processes that have been shown to emerge as a result of accommodation between speakers of different dialects in situations of long-term contact (Trudgill 1986) in order to show evidence for accommodation among Fessis in Casablanca. The chief focus here is on leveling, simplification and mixing in the reflexes of dialect contact in Casablanca.

5.4 Maintenance and leveling of Fessi linguistic forms General findings reveal that maintenance and/or leveling of Fessi linguistic forms are not necessarily constrained by the immigration status of the speakers. That is, some first migrant Fessis have adopted Casablancan variants and use them more frequently than some Fessis born in Casablanca. On the other hand, a native of Casablanca may maintain all Fessi variants. Before turning to the local meaning of maintenance and/or change to our informants general results for each linguistic variable will be discussed first. 5.4.1 The (r) variable The outcome of contact between Fessi post-alveolar approximant [ ] and Casablancan alveolar trill [r] has produced three groups of speakers: (1) 106

BECOMING CASABLANCAN

maintainers of Fessi [ ]; (2) adopters of Casablancan [r]; and (3) users of an intermediate form [m], a form “halfway” between Fessi [ ] and Casablancan [r]. Table 5.3 Frequency of use of Casablancan [r] by Fessi women

Fessi [ó ] users Fatima Amina Samia Ihsan Intermediate [ï] users Kawtar Siham Casablancan [r] users Nadia Hayat Leila Keltoum Khadija Lamia Fatma Zakia

Generation

Fessi [ ]

I I I II

56 172 57 51

0 0 0 0

0 0 0 0

II II

5 0

105 68

5 42

4.34 38.18

I I I I II III III III

0 0 0 0 0 0 0 0

0 0 0 5 1 2 3 6

32 128 95 64 72 92 32 82

100 100 100 92.75 98.63 97.87 91.42 93.18

Intermediate [M]

Casablancan [r]

% of [r] 0 0 0 0

There are no linguistic constraints on the adoption of trilled [r] or the maintenance of Fessi [ ]. As we can see from Table 5.3, speakers who use trilled [r] or maintain Fessi [ ] do so in all their speech without variation between one form and the other. For the intermediate [m] users, however, some linguistic constraints are identified such as the use of trilled [r] next to back vowels and inter-vocalically, but no linguistic generalizations can be made in this regard since this observation is based on few instances and more data are needed to support it. Examination of the social constraints on leveling and/or maintenance reveals an intimate interaction between age of exposure and social network. Overall findings show that trilled [r] is harder to learn and produce and this is evidenced by the fact that Fessis who maintain the approximant were not able to trill the [r] when I asked them to produce it as part of an imitation task. It seems that after a certain age speakers cannot trill the [r] even if they wish to do so. This indicates that a relatively late exposure to trilled [r], which maybe the result of having Fessi-speaking caregivers during formative years, and a tight-knit Fessi network are instrumental in the maintenance of the Fessi approximant. In general, most adopters turn out to have relatively weaker Fessi network ties than the maintainers. For both maintainers and adopters the age and network factors are reinforced by social attitudes and group identity (cf. Hachimi 2005). 107

ATIQA HACHIMI

The fact that some women maintain Fessi [ ], which is both a minority form and a highly stereotypical trait of Fessi speech, does not lend support to theories of dialect leveling, which predict that a form like [ ] would usually be lost in dialect contact (Trudgill 1986). Furthermore, the fact that Ihsan has been born, raised, attended school in Casablanca and has never lived anywhere else, is no guarantee that a Fessi speaker will accommodate to the Casablancan dialect. This indicates that an individual’s network and the attitudes that might develop and be maintained in that network, as well as the varying practices that reinforce one’s Fessiness, may weaken the importance of the nativeness criterion itself. 5.4.2 The (q) variable The qaf variable reveals a different pattern of adoption and maintenance than the raa variable. The main finding is that all speakers maintain the Fessi variant [q] in all lexical items, except for the verb (qal) “to say”. Speakers are thus divided into categorical users of Fessi [™al] or [qal] (the Old and Modern Fessi variants, respectively), categorical users of Casablancan [gal] and one speaker who varies between [qal] and [gal] about 50 percent of the time. Table 5.4 The variable (qal) by speakers Fessi [§al]/[qAl] users

[gAl] users

[gAl]~[qAl] user

[Åal] user Fatima [qÑl] users Samia Amina Ihsan

Leila Keltoum Hayat Saida Khadija Kawtar Fatma Siham Zakia

Nadia

If we compare Tables 5.3 and 5.4, it becomes clear that the same women who maintain Fessi [ ] also maintain the Fessi variants [™al]/[qal] while the women who have leveled Fessi [ ] almost all of them have leveled [qal] and are using [gal] instead. Among these regular maintainers is a Casablancaborn Fessi. Her consistent use of Fessi variants provides further evidence that being native of Casablanca does not necessarily guarantee that a Fessi woman would leave behind stereotypical Fessi features. This is quite different, for instance, from what Miller (2005) found among Upper Egyptian migrants in Cairo where it took only one generation to switch to the Cairene variety. 108

BECOMING CASABLANCAN

Finding that there are lexical constraints on the qaf variable supports findings in other Arabic speaking communities (e.g. Gibson 2002; Haeri 1997; Holes 1986). However, to my knowledge, none of these studies have found variation in one lexical item alone. The question that follows, therefore, is why do Fessi women maintain the Fessi variant [q] in all other lexical items but they vary with respect to the verb (qAl) in particular? In other words, why did they adopt [gal] but not, say, [gbəl] “before”, [trbg] “street”, [srəg] “steal” and several other lexical items where variation between [q] and [g] is possible in the Casablancan variety? The answer to these questions lies in the social meanings encoded in [qal] on the one hand and in [gal] on the other. Based on women’s comments it appears that there is a clear boundary between [qal] users and [gal] users. Most of these women are unwilling to identify with [qal] users mainly because [qal] indexes F wassa-d-Fes or “pure Fessi” identity. They emphasize their use of [gal] to assert their distance from speaking the Fessi dialect and also to assert that [gal] allows them to sound “normal”. For instance, Kawtar likes the fact that she maintains non-trilled Fessi [ ] because it is very feminine and charming. For her [qal] indexes pure Fessiness more so than the variant [ ], in fact, she considers bhal “stupid” a Fessi who uses the Fessi variants of qAf ([ʔ] or [q]). Kawtar: ma∫i fassija fassija, je suis normale / ana ∫wija bi%awija, je parle [r] avec a c’est tout/parce que lfassijat tajgulu ‘qatli qutl¢k’/ ana c’est ‘galli gutl¢k’ dik ‘qalli u-qutl¢k’ c’est bête. “I am not pure Fessi, I am normal . . . I am Casablancan a little. I speak with [Fessi] [ ] that’s all, because Fessi women say ‘qatli qutlək’ . . . Me it’s ‘galli gutlək’ saying ‘qatli qutlək’ is stupid.” If we accept the argument that saying [gal] instead of [qal] casts the speaker as Casablancan, we can therefore suggest that these women employ the most socially significant linguistic resource that readily allows them to sound Casablancan. Furthermore, to adopt [g] in other lexical items beside gAl would mean to sound ∞rubi-Casablancan rather than Casablancan. It seems that these women are distancing themselves not only from pure Fessis, but from ∞rubi-Casablancans as well. They do that by maintaining [q] in all other lexical items while adopting the koiné form [gal]. Both of these linguistic choices allow them to sound “normal”. As we will see later these choices are essential in shaping the hybrid Fessi-Casablancan identity most of these women claim. In addition, it seems that these women are adopting as their target the koiné variety, rather than Casablancan features per se. If this is the case, then this would support results from the Arabic speaking world that women target the local urban vernacular, which is attested in well-established urban 109

ATIQA HACHIMI

dialects such as Cairene Arabic (Haeri 1997) as well as in newly formed dialects such as Amman Arabic (Al-Wer 1999, 2002). 5.4.3 The (-i) variable The morphosyntactic variable (-i) has two variants in the second person singular feminine subject clitic, the Casablancan variant where gender is distinguished and marked by the clitic [-i] and the Fessi variant where gender is neutralized and marked by a zero subject clitic [-Ø]. The results for the morphosyntactic variable (-i) indicate that Fessi women behave differently towards the low-salience morphosyntactic variable than they did towards the highly salient phonological variables (q) and (r). That is, contrary to the high rate of adoption of Casablancan [r] and [gal], most speakers maintain the Fessi morphosyntactic variant, either categorically or partially. Similar to the Moroccan Arabic koiné, Fessi and Casablancan dialects neutralize gender distinction towards the feminine in the perfective but they differ in the imperfective and the imperative. Table 5.5 Second person singular in Fessi and Casablancan dialects Verb k-t-b

Fessi dialect

Casablancan dialect

‘to write’

Masculine/Feminine

Masculine

Feminine

Perfective Imperfective Imperative

kt£bt-i t-kt£b kt£b

kt£bt-i t-kt£b kt£b

kt£bt-i t-kt£b-i k£tb-i

Table 5.5 clearly shows that while feminine and masculine genders are distinguished in suffixed pronouns in imperfective and imperative aspects in Casablancan, this distinction is neutralized in the Fessi dialect. This neutralization is not limited to the Fessi dialect but is spread across old urban dialects in Morocco and North Africa in general. Unlike the (qal) and (r) variables, the majority of informants maintain Fessi gender neutralization since nine women, six of whom are native to Casablanca, have zero usage of the Casablancan variant [-i]. In fact no woman in this study has shown categorical use of gender distinction as they did for example with [r] and [gal], instead the remaining five women vary between gender neutralization and distinction. Interestingly, maintainers include Fessis native to Casablanca and Fessis native to Fez. Among these women are the four women whom we have become accustomed to finding as maintainers (Amina, Fatima, Samia and Ihsan). It is not surprising that these four women would maintain gender neutralization given that they maintain even the highly salient Fessi [ ] and [qal]. 110

BECOMING CASABLANCAN

Table 5.6 Frequency of use of gender distinction in 2nd person singular among Fessis Informants Fessi [-Ø] maintainers Siham Samia Amina Fatima Saida Ihsan Kawtar Fatma Zakia Lamia [-Ø] ~ [-i] users Keltoum Hayat Leila Nadia Khadija

Fessi [-Ø]

Casablancan [-i]

% of [-i]

14 33 45 21 16 27 25 9 12 30

1 0 0 0 0 0 0 0 0 0

6.66 0 0 0 0 0 0 0 0 0

26 28 18 5 13

25 29 16 6 4

49.01 50.87 47.05 54.54 23.52

What is surprising, however, is the maintenance of Fessi gender neutralization by Casablanca-born Fessis who are trilling [r] and using [gal]. One would expect these women to adopt second person singular gender distinction if they are indeed targeting the Moroccan Arabic koiné. However, the fact that almost all Casablanca-born Fessis maintain second person gender neutralization indicates that this feature is the norm among Fessis native to Casablanca. In Hachimi (2005) I suggest that linguistic regularization and simplicity of gender neutralization are potential linguistic factors responsible for the wider retention of the Fessi morphosyntactic variant. The process of regularization has to do with generalization of gender neutralization changes attested in the possessive and object clitic system to subject clitics. In most Moroccan Arabic dialects object and possessive suffixes are neutralized toward the masculine gender in the second person singular. For instance, -¢k is the gender neutral form used in second person singular object and possessive suffixes. The second potential contributing factor in the survival of gender neutralization among Fessi women may be the simplicity of gender neutralization over gender distinction. Following principles of dialect contact and change (e.g. Siegel 1985) simpler forms tend to win out when in competition with more complex forms. This suggests that the simplicity of Fessi gender neutralization allows it to survive, thus inhibiting the adoption of the relatively more “complex” gender distinction. Although language-internal factors may act as forces in inhibiting wider adoption of gender distinction, social and social psychological factors 111

ATIQA HACHIMI

may counteract the linguistic pressures mentioned above. Speakers who have adopted the gender distinction are in fact using a more “complex” form rather than the simple uninflected neutralized form. It seems that the desire to accommodate and be understood by others is important. This suggests that no matter how strong internal-language factors may be, social and social psychological factors do sometimes take precedence. 5.4.4 Local meaning of maintenance and leveling Fessi women in Casablanca vary in reducing differences between the Fessi and Casablancan dialects. In this section we uncover the social meanings behind this variation by looking closely at Fessi women’s attitudes toward the Fessi and Casablancan dialects and identity. The point is to determine the extent to which maintaining salient linguistic forms, or leveling them out, plays an important part in constructing both social categories and social meaning among Fessi women in Casablanca. 5.4.5 Correlations of maintenance, attitudes and identity There are four speakers who are categorized as maintainers of salient Fessi features; non-trilled [ ] and [™al]/[qal]. They include Fez-born women migrants to Casablanca (Amina, Hajja Fatima and Samia) and a native to Casablanca (Ihsan). In this section we examine their attitudes towards the Fessi and Casablancan dialects, their attitudes toward Fessis and non-Fessis, as well as their self and group identifications. A neat correlation between maintenance of Fessi features, positive attitudes towards the Fessi dialect and strong in-group feelings is best seen in Amina and Samia. Both women have expressed great pride in being Fessi and in speaking in a way that indexes their Fessiness. They perceive the Fessi dialect as rYiYba “soft”, m’adba “polite”, }luwa “sweet”, fiha laγna “rhythmic” (lit. sing-song), katliq blamra “suitable for the woman”, in contrast to Casablancan speech, which they label ∞rubi “rural and unsophisticated”, }r¢∫ “tough, coarse” and vulgaire “vulgar”. These traits of softness, politeness and a melodic way of speaking are central to what has been constructed traditionally as desirable femininity. This is also true for other old urban centers in North Africa such as Tlemcen women who hold a comparable position to Fessi women. Similarly, Samia and Amina have a set of criteria by which they distinguish Fessis from non-Fessis in general. They emphasize the historical and sociocultural aspects of this distinction, and they even go a step further than distinguishing between “us” (Fessis) and “them” (non-Fessis), to expressing the superiority of Fessis over the non-Fessi Moroccans. This superiority is attributed to the fact that Fessis have inherited urbane lifestyle and cultural practices from their Andalusian ancestors. Their urbane socio-cultural 112

BECOMING CASABLANCAN

practices and lifestyle, they say, have been passed on §aban ∞an add “from father to son” and have been perfected for centuries, which both women insist cannot be emulated by other groups no matter how urbanized they have become. Both used the following Moroccan saying to show how other regional groups are envious of Fessis and their city: Amina/Samia: Lfassi kajqul ‘ya ¢bbi ∞Yini ¢nna’, u ssusi tajqullu ‘ya ¢bbi ∞Yini Fas’ “The Fessi says ‘God please give me Heaven’ and the Soussi says ‘God please give me Fez’.” Like Amina and Samia, Hajja Fatima agrees that Fez and Fessis are the city and the group of people who define civilization in Morocco. In fact, her comments about Fessis and non-Fessis and her statement iwa kayn Fes ukayn lmaγrib “well, there is Fez and there is Morocco” reminds us of the old Fessi attitude that emphasizes the division of Moroccans into Fessis and non-Fessis. She made another comment that Fes huwa lmaγrib “Fez is Morocco” which also seems to reiterate the old Fessi worldview. It is important to note, however, that Hajja Fatima did not express overt negative or positive attitudes toward the Fessi or the Casablancan dialects. Her unwillingness seems to be in line with attitudes of older people in the 70s and 80s age group who show a certain degree of neutrality towards the dialects (e.g. Moumine 1990). She did acknowledge, however, the difference between the two dialects and brought up the lack of mutual intelligibility between her and speakers of Casablancan dialect. In general, Fessis like Fatima, Samia and Amina comprehend their collective Fessi identity as “primordial”. Under the “primordial model” of ethnic identity, the nature of ethnic ties is considered “‘given’, deeply rooted, not a matter of choice but of circumstantial inheritance” and identity is “rooted in history and tradition” and seen as “stable” and “permanent” (Cornell and Hartmann 1998: 68). This one-to-one relationship between positive attitudes and maintenance of Fessi features does not seem to be as strong for Ihsan; the regular maintainer who is native to Casablanca. Ihsan seems to share the same views and attitudes as the three women from the first generation who consider being Fessi a great asset for a woman but when she ventures to selfidentify with a group, a more nuanced Fessi identity emerges. She clearly discriminates between “pure Fessis”, “Fessi-Casablancans” and “non-Fessis”, in particular the ¡rubis and Soussis. What is striking about Ihsan is her earnest desire to distance herself from “pure” Fessi women who she believes are “naïve, spoiled, snobbish and too delicate”. Instead, Ihsan considers herself }¢r∫a “tough” which, as she observes herself, may not be revealed in her use of typical pure Fessi forms but more so in the way she behaves and sees herself. In her view, speaking with stereotypical Fessi linguistic forms 113

ATIQA HACHIMI

must not be a factor in lumping her with the “pure” Fessis. It seems that for this young Casablanca-born Fessi, self-identifying as a “pure” Fessi has become more of a liability than an asset, as is the case with the Fez-born maintainers discussed earlier. Here the instrumentalist (utilitarian) nature of identity is useful in understanding why individuals (or groups) identify the way they do (Cornell and Hartmann 1998: 56– 60). Ihsan is actively engaged in tailoring her own identity by separating her stereotypical Fessi way of speaking from her identity. Her agency in building a hybrid FessiCasablancan identity and assigning meaning to this category is reflected not only in rejecting any association with “pure” Fessis but in distancing herself from Casablancan women who are of ∞rubi origin. Being perceived as }¢r∫a seems to be positive for Ihsan probably because it carries the meaning of an independent and modern woman. The notion of }¢ru∫ija “roughness and toughness” and how it relates to Fessi-Casablancan identity will be taken up in the next section. 5.4.6 Correlations of leveling, attitudes and identity There are eleven speakers who fall under the category of levelers, i.e. women who leveled salient Fessi features. Four of them (Hayat, Leila, Keltoum and Nadia) are migrants while the others are all native to Casablanca. The goal here is to see whether there is a qualitative difference between the levelers and the maintainers in terms of their attitudes towards the Fessi and Casablancan dialects, towards Fessis and non-Fessis and self and group identifications. Findings show that levelers vary in their attitudes towards Fessi and Casablancan dialects and groups. Thus, while there is correlation between leveling and negative attitudes among some women, this does not hold for all of them as some show ambivalent attitudes. Insofar as in-group and out-group identification are concerned levelers have shown highly nuanced categories and meanings. Similar to the maintainers, levelers who have expressed positive attitudes towards Fessi dialect also emphasize the desirable feminine qualities of Fessi speech particularly its softness, politeness and sing-song prosody. These traits, according to Leila, a 60 year-old migrant, make Fessi speech sound more civilisé “civilized” compared to the ∞rubi dialect, which she considers m∫ufa “savage”. Similar views are reiterated by a 21 year-old second generation Fessi who stresses the aesthetic and the refined aspects of Fessi speech and contrasts them to “the aggressiveness” of Casablancan speech. Lamia: tatini mazjana fl}aqiqa. wllahila fl}aqiqa zwina lfassija, rYiYba/ lh¢|ra djal lkasawijin hija l¢m|arba djal F wassa “I find it really nice actually, honestly. In fact, Fessi speech is beautiful and soft . . . the way Casablancans speak is similar to an argument by Fessis . . .” 114

BECOMING CASABLANCAN

Lamia is a good example of Fessis native to Casablanca who have the greatest delight and pride in belonging to the Fessi group. However, like most levelers in this study, Lamia draws a clear line between F wassa-d-Fes “Fessis from Fez” and F wassa-d-Kasa “Fessis from Casablanca” and gives out several clues on what it means to be one or the other: Lamia: F wassa d Fez, tajkunu ∫wija b¢hlin b¢∞da, uh|¢rthum makat fh¢mha∫ presque ga∞ . . . tajkunu ∫wija m∞¢qdin, f∫i ∫k¢l, ∞¢ndhum l}¢∫ma b¢zzaf, ∞¢ndhum daks∫i t¢qlidi b¢zzaf, alors que lli tajY¢l∞u lkasa, taj}ra∫u . . . “Fessis from Fez are a bit naïve, their dialect is almost unintelligible . . . they are a little behind, they are shy, and they are very traditional. But those who come to Casablanca have become tougher . . .” Similarly, levelers who have negative attitudes towards Fessi agree with Lamia on the near unintelligibility of pure Fessi speech. It is quite revealing to see Nadia who uses qAl about half the time in her speech just as eager to distance herself from pure Fessis as the others: Nadia: kajiwni z¢∞ma lfasiyin pures kajjib li llah b}alli h|art m∞a∫i b¢rrani ma∫i mʁarba. “I find pure Fessis like foreigners, speaking to them is like speaking to non-Moroccans.” In addition, most levelers agree that pure Fessi speech is “slow”, “outdated”, “pompous” and some perceive its features to be “out of the ordinary”. To a 26 year-old third generation Fessi, pure Fessi speech seems to be utterly unbearable. Fatma: Je ne peux pas tenir une discussion m∞a Fessi, pure Fessi, moi ça m’étouffe. “I cannot hold a conversation with a Fessi, a pure Fessi, it suffocates me.” By distancing themselves from the “pure” Fessi stock indicates that to these levelers stereotypical Fessi linguistic forms are a liability rather than an asset. Furthermore, it must be mentioned that these women make a clear distinction also between ∞rubi “rural” and ∞adi “normal” Casablancan speech. All of them find ∞rubi-Casablancan to be “too tough” and “vulgar” except for Keltoum who finds certain ∞rubi expressions to be funny and charming. This sheds light on why these women reject “pure” Fessi and “pure” ∞rubi identities, and reconstruct their identity as Fessi-Casablancan. 115

ATIQA HACHIMI

5.5 Becoming Casablancan Whereas the maintainers, particularly those non-native to Casablanca, stress the difference between “us Fessis” and “them non-Fessis”, levelers stress the difference between “us Fessis of Casablanca” and “them Fessis of Fez”. I argue here that the category Fessi-Casablancan emerges in the intersection between speaking ∞adi “normal” and becoming ∫¢∞bi “like the others” and }r¢∫ “tough”, all of which are important to thinning out pure Fessi identity. 5.5.1 Speaking Çadi Without exception all speakers who leveled stereotypical Fessi linguistic forms have used the adjective “normal” to describe the way they speak. They used the Moroccan Arabic word ∞adi and/or the French normale. But what does “normal” mean and how does it help us understand the process of becoming Casablancan? Several definitions have come up when I asked the women to explain what they mean by “normal”. Leila: ma fasija ma mknasija ma walu normale . . . ∞adija, ma ∞¢ndi la §af wa la aw. ana had∫∫i kullu ma∞¢ndish. ∞¢ndi h¢|ra ∞adija “Not Fessi, Meknesi or anything else. It is normal . . . I do not have the §af or the aw. I do not have all that. I have normal speech.” What is quite fascinating about this description is that it is precisely a definition of leveling; which is the attrition of salient, strongly localized and stigmatized features in favor of forms that are widely used in other dialects (Kerswill and Williams 2002; Milroy et al. 1995). Similarly, other levelers conveyed the idea by defining “normal” in the negative as well. Keltoum, for instance, gives the example of Fessi [ ] and the glottal stop [ʔ], in particular, as being “out of the ordinary”, as she put it “ça sort de l’ordinaire”, whereas Fatma used the phrase h¢|ra mqadda “straight speech” to explain what she means by “normal”. On the other hand, others like Khadija added that “normal” means speech that is “not vulgar” and Nadia says that it is speech that is “not salient” or as she phrased it bhal kikajhdru∞ibad llah “like everybody else”. In their explanation of “normal speech”, almost all levelers contrast “pure Fessi”, “¡rubi Casablancan” to “normal” speech. “Normal” as we can see here is defined in the negative; it is neither this speech nor that. This is akin to the point made by Wolfram and Schilling-Estes (1998) about Standard American English, which is defined more by what it is not rather than by what it is, as they put it “if a person’s speech is free of structures that can be identified as nonstandard, then it is considered standard” (Wolfram and Schilling-Estes 1998: 12). I argue that what these women are referring to as

116

BECOMING CASABLANCAN

“normal speech” is the Moroccan Arabic koiné, which is defined more by what it is not rather than by what it is. The close connection between speaking in a “normal” way and becoming Bi|awija “Casablancan” came up in several discussions with the women. Kawtar, for instance, clearly expresses the interconnectedness between leveling and the thinning out of pure Fessi identity. If we recall from the previous section, Kawtar uses an intermediate [m] and uses [gal]. For Kawtar [qal] variant indexes pure Fessiness more than the variant [ ]. For ease of reference, Kawtar’s quote will be repeated here. Kawtar: ma∫i fassija fassija, je suis normale . . . ana ∫wiya bi|awija, je parle [r] avec [ ] c’est tout, parce que lfassijat tajgulu ‘qatli/qutl¢k’, ana c’est ‘galli gutlak’ dik qalli uqutl¢k c’est bête. “I am not Fessi Fessi. I am normal . . . I am Casablancan a little. I speak with [Fessi] [ ] that’s all, because Fessi women say ‘qatli/ qutlək’. I say ‘galli gutlak’ saying ‘qatli/qutlək’ is stupid.” It is clear from Kawtar’s comments that she puts lots of emphasis on [gal] variant being part of normal Casablancan speech. Her use of [gal] instead of [qal] seems to be an important component in thinning out “pure” Fessi identity and casts her as Fessi-Casablancan. 5.5.2 Becoming CaÇbi The notion of “normal” that came through in the quote above is not simply being convergent linguistically but being “one of the people” as well. I argue that while speaking with Fessi features indexes snobbishness, leveling indexes ta∫ ∞bit, which is a desire to be like the others. Many levelers pride themselves of being ∫¢∞bijat “one of the people, down to earth” as opposed to “snob and pompous”, which is how pure Fessis are described. Keltoum is very keen on distancing herself from Fessis who are “snobbish” and “racist”: dik la haute societé ma∞¢ndi∫ m∞aha “I am not very fond of high society”, emphasizing instead that she likes ordinary people. It is worth mentioning that even among some levelers who have expressed the most positive feelings towards everything Fessi, they are reluctant to identify with “pure” Fessis particularly because of the elitist attitudes they apparently have towards other Moroccan social groups. Leila, for instance, embraces a Fessi identity only as far as certain cultural practices are concerned, such as cooking, dressing, and arranging a bride’s trousseau, but she was proud of the fact that the way she speaks is “normal” and that she is “one of the folks”. For other levelers, like Fatma and Zakia, this notion of ta∫ ∞bit has a lot to do with being able to adapt and blend in both linguistically and socially. It seems that ta∫ ∞bit is important to becoming Fessi-Casablancan or Casablancan. 117

ATIQA HACHIMI

5.5.3 Becoming íërCa So far we have seen that speaking normally and becoming ∫¢∞bi are integral to the category “us Fessis of Casablanca”. Another very important component of this social category is to become }r¢∫. Almost all informants have used the word }¢r∫a, which seems to have become a typical word for describing people and their speech in Casablanca. Almost all Fessi informants, levelers and maintainers, agree that Casablanca-born Fessis are tougher than Fez-born Fessis and it appears that “toughness” distinguishes them from Fessis from Fez. But again what does it mean to become }¢r∫a? Keltoum is aware of the changes that have happened to her dialect and to her personality after she immigrated to Casablanca. She emphasizes the fact that she has become aggressive, independent, street smart, less delicate when she moved to Casablanca. In short, she became }¢r∫a. Keltoum: kunt timide kan∫ufhum kanw¢lli k¢neʁz¢l nw¢lli }¢mra ila ∫af fija fija ∫i wa}ed kulli, it ll¢hna }ra∫it, Bi|awija pure . . . }ra∫it fa∫ it lkasa. hd¢rti wllat }¢r∫a, le r w¢lla }ra∫ . . . ra daba wllat tats¢rreh li ra. “I used to be shy, when I see them (men) I start shaking and I blush. I came here I became tough, pure Casablancan. Everything about me has become tough when I came to Casablanca. My dialect became tough too. My r has become tough. My ra has become straight.” It is clear that leveling is central to becoming }¢r∫a. Keltoum’s comment indicates that leaving behind stereotypical Fessi variants and becoming “tough” are important factors in becoming Casablancan. While many levelers have expressed positive attitudes towards “normal” speech, they agree that in comparison with Fessi speech, this variety is }¢r∫a. Toughness, as studies have shown, is a positive male value usually associated with working class status and with the use of vernacular features (e.g. Trudgill 1974). What is striking here is that what is considered a “tough” dialect, mainly because of [gal] and trilled [r], are precisely the variants some of these Fessi women are adopting. Becoming “tough” is far from carrying negative meanings as one might expect, rather it seems to be an attractive trait that distinguishes Fessi-Casablancans from Fessis from Fez. However, there seems to be general agreement among these women that speaking with Fessi accent makes a man sound effeminate. Fatma, for instance, is unequivocal about the type of identities a man and a woman should embrace; a man should be Bi|awi “Casablancan” whereas the woman should be in-between Fessi and Casablancan. As we have seen so far, Fessi features index traditional femininity. One can argue that if Fessi variants index an outdated, traditional and reserved

118

BECOMING CASABLANCAN

persona, becoming tough and less feminine are desirable probably because they index modernity. This would suggest that Fessi women who leveled Fessi variants do not mind trading off traditional femininity for a modern identity, where the pressures to be “aggressive and tough” are overriding. Thus, sounding “normal”, even if it means sounding “tough” and “less feminine” is important in becoming Fessi-Casablancan.

5.6 Conclusions In this article we have looked at the correlations between dialect leveling and/or maintenance, attitudes and identity. Based on speakers’ overt and covert attitudes and ideologies towards Fessi and Casablancan linguistic and social identity, we have identified the categories that have emerged as a result of the migration of Fessis to Casablanca, such as F wassa-d-Fes “pure Fessis”, F wassa-d-Kasa “Fessi-Casablancans” and Bi|awa “Casablancans”. It turned out that leveling and/or maintenance of stereotypical features is a site and a resource in Fessi identity (re)construction in Casablanca. That is, leveling stereotypical Fessi traits is important in setting the boundary for Fessi-Casablancan identity and giving it meaning, while maintenance of these linguistic forms tends to be, to some extent, crucial in the maintenance of “pure” Fessi identity. Leveling of salient Fessi linguistic forms shapes and defines a hybrid FessiCasablancan identity. In general, those who do not level Fessi linguistic forms have a stronger Fessi identification, while leveling helps in thinning pure Fessi identity. That is, the use of [gal] instead of [qal] is an important resource for a Fessi-Casablancan identity. Informants agree that Casablancaborn Fessis have become “tough” in Casablanca and that it is this “toughness” that distinguishes them from Fessis from Fez. The category F wassa-d-kasa “Fessi-Casablancans” has emerged because of this “toughness”. In other words, part of what defines a Fessi-Casablancan is that they think they are “tougher” than Fessis from Fez, and leveling of stereotypical Fessi variants is central to the existence of this category. The key to understanding what it means for a Fessi to become Casablancan is found in an interaction between linguistic and social practices. To become Casablancan depends on (1) speaking “normally”, i.e. leveling stereotypical linguistic forms, such as leaving behind [™al] or [qal] for Casablancan [gal] and adopting Casablancan trilled [r] instead of the Fessi approximant; (2) to shift from a perceived “soft, delicate and too feminine” persona to becoming }¢r∫a “i.e. tough, street smart, independent, less delicate and over the edge”; and, (3) becoming ∫a∞bija “one of the folks, unassuming and down to earth”. These are important resources in thinning out “pure” Fessi identity, and do in turn index Fessi-Casablancan or Casablancan identity.

119

ATIQA HACHIMI

References Adam, A. (1968) Casablanca: essai sur la transformation de la société marocaine au contact avec l’Occident, Tome I & II, Aix-en-Provence: CREAM. Aguadé, J. (2003) “Notes on the Arabic dialect of Casablanca (Morocco)”, in I. Ferrando and J. J. Sanchez Sandoval (eds) AIDA Fifth Conference Proceedings, Universidad De Cadiz: Servico De Puplicaciones, pp. 301–308. Al-Wer, E. (1999) “Why do different variables behave differently? Data from Arabic”, in Y. Suleiman (ed.) Language and Society in the Middle East and North Africa, Surrey: Curzon Press, pp. 38–58. Al-Wer, E. (2002) “Jordanian and Palestinian dialects in contact: vowel raising in Amman”, in M. Jones and E. Esch (eds) Language Change: The Interplay of Internal, External and Extra-Linguistic Factors, Berlin: Mouton de Gruyter, pp. 63–79. Brunot, L. (1950) Introduction à l’arabe marocain, vol. 1, Paris: Les langues de L’Orient. Caubet, D. (1993) L’arabe marocain: phonologie et morphosyntaxe, Tome I, ParisLouvain: Éditions Peeters. Caubet, D. (1998) “Étude sociolinguistique des traits préhilaliens dans un dialecte en voie d’urbanisation à Fèz”, in J. Aguadé, P. Cressier and A. Vicente (eds) Peuplement et arabisation au Maghreb occidental, Madrid-Zaragoza: Casa de Velazquez–Universidad de Zaragoza, pp. 165–175. CERED (Centre d’Etudes et de Recherches Démographiques) (1993) Migration et urbanisation au Maroc, Rabat: Direction de la Statistique, Ministère chargé de la Prévision Economique et du Plan. CERED (Centre d’Etudes et de Recherches Démographiques) (1998) Population et développement au Maroc, Rabat: Direction de la Statistique, Ministère chargé de la Prévision Economique et du Plan. Cohen, J.-L. and Eleb, M. (2002) Casablanca: Colonial Myths and Architectural Ventures, New York: Monacelli Press. Cornell, S. and Hartmann, D. (1998) Ethnicity and Race: Making Identities in a Changing World, London: Pine Forge Press. Dernouny, M. and Léonard, G. (1987) Casablanca: la parole et la trace, Paris: Afrique Orient. Eckert, P. (2000) Language as Social Practice, Oxford: Blackwell. Elbiad, M. (1992) “The Arab people and the Arabic sounds of greater Casablanca”, in P. H. Nelde (ed.) Plurilingua XIII: It’s Easy to Mingle When you’re a Bilingual, Bilingualism and Contact Linguistics: Dümmler. Gibson, M. (2002) “Dialect leveling in Tunisian Arabic: towards a new spoken standard”, in A. Rouchdi (ed.) Language Contact and Language Conflict in Arabic: Variations on a Sociolinguistic Theme, New York: RoutledgeCurzon, pp. 26–40. Giles, H. and Coupland, N. (1991) Language: Contexts and Consequences, Milton Keynes: Open University Press. Gubbins, P. and Holt, M. (eds) (2002) Beyond Boundaries: Language and Identity in Contemporary Europe, Bristol: Multilingual Matters 122. Hachimi, A. (2005) “Dialect leveling, maintenance, and urban identity in Morocco”, unpublished thesis, University of Hawaii. Haeri, N. (1997) The Sociolinguistic Market in Cairo: Gender, Class and Education, London and New York: Kegan Paul International.

120

BECOMING CASABLANCAN

Hamdi, R. and Puech, G. (1987) “Le domaine prosodique de la pharyngalisation en arabe marocain de Casablanca”, Proceedings of the First International Conference of the Linguistic Society of Morocco, Rabat: Okad Ed. Heath, J. (1989) From Code Switching to Borrowing: A Case Study of Moroccan Arabic, London: Kegan Paul International. Heath, J. (2002) Jewish and Muslim Dialects of Moroccan Arabic, London: RoutledgeCurzon. Hillili, A. (1979) “Phonologie et morphologie de l’ancien Fassi”, unpublished thesis, University Paris III. Hillili, A. (1987) “Contribution à l’étude fonctionnelle de l’arabe classique et de l’arabe fassi”, unpublished thesis, Université Paris III. Holes, C. D. (1986) “The social motivation for phonological convergence in three Arabic dialects”, International Journal of the Sociology of Language 61: 33–51. Iraqui-Sinaceur, Z. (1998) “Le dialecte de Tanger”, in J. Aguadé, P. Cressier and A. Vicente (eds) Peuplement et arabisation au Maghreb occidental, Madrid-Zaragoza: Casa de Velazquez–Universidad de Zaragoza, pp. 131–140. Jabeur, M. (1996) “Women, social change and linguistic variation in the urban dialect of Tunis”, in C. Holes (ed.) Proceedings of the 2nd International Conference of l’Association International pour la Dialectologie Arabe (AIDA), Cambridge, pp. 85–94. Kampffmeyer, G. (1912) Marokkanisch Arabische Gespräche in Dialect von Casablanca, Berlin: Druck and Verlag Georg Reimer. Kerswill, P. and Williams, A. (2002) “ ‘Salience’ as an explanatory factor in language change: evidence from dialect levelling in urban England”, in M. C. Jones and E. Esch (eds) Language Change: The Interplay of Internal, External and ExtraLinguistic Factors, Berlin: Mouton de Gruyter, pp. 81–110. Khomsi, A. (1975) “Etude phonétique et phonologique de l’arabe marocain de Casablanca”, unpublished thesis, Tours: Université Francais Rabelais. Landry, R. and Allard, R. (1994) “Diglossia, ethnolinguistic vitality, and language behavior”, International Journal of the Sociology of Language 108, 15–42. Le Tourneau, R. (1965) La vie quotidienne à Fès, Hachette: Paris. Lévy, S. (2002) “Du nouveau dans le domaine préhilalien”, in A. Youssi et al. (eds) Aspects of the Dialects of Arabic Today, Rabat: AMAPATRIL, pp. 105–113. Messaoudi, L. (1998) “Traits linguistiques du parler ancien de Rabat”, in J. Aguadé, P. Cressier and A. Vicente (eds) Peuplement et arabisation au Maghreb occidental, Madrid-Zaragoza: Casa de Velazquez–Universidad de Zaragoza, pp. 157–163. Messaoudi, L. (2002) “Le parler ancien de Rabat face à l’urbanisation linguistique”, in A. Youssi et al. (eds) Aspects of the Dialects of Arabic Today, Rabat: AMAPATRIL, pp. 223–233. Miller, C. (2005) “Between accommodation and resistance: Upper Egyptian migrants in Cairo”, Linguistics 43: 5, 903–956. Milroy, J., Milroy, L. and Hartley, S. (1995) “Glottal stops and Tyneside glottalization: competing patterns of variation and change in British English”, Language Variation and Change 6: 327–357. Moumine, M. E. (1990) “Sociolinguistic variation in Casablanca Moroccan Arabic”, unpublished thesis, Casablanca, University Mohammed V. Moumine, M. E. (1995) “Mawaqif an-nayiqcn bil-¡arabiyya l-maghribiyya tijjah al-lahja al-bay\awiyya [Attitudes of native speakers of Moroccan Arabic towards

121

ATIQA HACHIMI

Casablancan dialect]”, Recherches en Linguistique Arabe, Publications de la Faculté des Lettres et des Sciences Humaines Ben M’sick: Casablanca, pp. 69–80. Nejmi, H. (1995) “Les variations prosodiques de l’accent de l’arabe marocain parlé a Casablanca: Approche phonétique et phonologique”, Dialectologie et Sciences Humaines au Maroc: Colloques et Séminaires 38: 105–117. RGPH (Recensement Général de la Population et de l’Habitat) (1982) Direction de la Statistique, Rabat: Morocco. RGPH (Reccensement Général de la Population et de l’Habitat 1994) (1996) Direction de la Statistique, Rabat: Morocco. Siegel, J. (1985) “Koines and koineization”, Language in Society 14: 3, 357–378. Terry, D. J. and Hogg, M. A. (1996) “Group norms and the attitude-behavior relationship: a role for group identification”, Personality and Social Psychology Bulletin 22: 776–793. Trabelsi, C. (1988) “Les usages linguistiques des femmes de Tunis”, unpublished thesis, Université Paris III. Trudgill, P. (1974) The Social Differentiation of English in Norwich, Cambridge: Cambridge University Press. Trudgill, P. (1986) Dialects in Contact, Oxford: Blackwell. Wolfram, W. and Schilling-Estes, N. (1998) American English: Dialects and Variation, Oxford: Blackwell. Zemmama, T. (1975) “Phonétique et phonologie de l’arabe fassi”, unpublished thesis, Université Toulouse-Le Mirail.

122

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

6 TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA Ángeles Vicente

6.1 Introduction Migration movements have been one of the most significant social phenomena in the twentieth century, producing considerable changes in both the urban and rural societies of the Arab World. From a sociolinguistic perspective, the greatest effect of these migrations has been an increased contact between various Arabic vernaculars particularly in the cities, due to the rapid urbanization process. But in order to present a complete picture of the impact of urbanization upon Arabic vernaculars, one must take into account not only the internal rural-urban migration trends occurring in present-day Arabicspeaking countries but also the impact of foreign urban migration. In this foreign context, a specific linguistic evolution is taken place, due to the fact that the immigrants’ Arabic mother tongues have a minority status and are evolving in contact with the dominant languages of the host country. Contemporary studies concerning Arabic dialects spoken in migratory situations concern both the Western non-Muslim countries of Europe and the United States, or non-Western countries where vernacular Arabic is spoken in contact with more prestigious or official languages. The most notable cases concerning Arabic-speaking communities in Europe, include: Iraqis in the United Kingdom (Abu-Haidar 1989, 1991), Moroccans in Holland (Boumans 1998; Boumans and de Ruiter 2002), and Maghrebians, particularly Algerians and Moroccans, in France (Caubet 2001a, 2001b; El Minaoui 2001; Melliani 2000). Few studies have been done in the United States, save those of A. Rouchdy (1992, 2002). Numerous works refer to the Arabic dialects spoken by Arabic-speaking minorities in the rest of the world.1 In these countries, Arabic speaking communities often represent ancient diasporas, unlike most of those found in Europe and the United States. Therefore the linguistic evolution of those Arabic vernaculars was/is rather different from what can be observed today in the Western world. Within this scholarship, one can quote studies on Arabic varieties spoken in Africa (Nigeria, Chad, Uganda, or South Sudan), in Asia (Afghanistan, Uzbekistan and Turkey), on Maltese, on Kormakiti 123

ÁNGELES VICENTE

Arabic in Cyprus or the historical studies on Arabic languages that have disappeared, such as Andalusi Arabic and Sicilian Arabic.2 These Arabic dialects have all been influenced by local dominant languages, for example, Italian in Maltese or Sub-Saharan languages in the African group. In any case, these studies highlight the existence of an extremely heterogeneous landscape, so that the Arabic vernaculars in the diaspora cannot be treated as a uniform entity. The particular characteristics of each specific context need to be taken into consideration. The mother tongues of immigrants evolve differently depending upon the neighboring dominant language, but also upon extralinguistic factors peculiar to each community, a point I seek to confirm below. A good example of such a local specificity is the case of France, where colloquial Arabic is considered “a language of France”. The vitality of Maghrebian Arabic in France differs greatly from what has been recorded in other European countries (Caubet 2001a, 2001b, 2002a, 2002b, 2004a, 2004b, 2005). In my opinion, the study of the evolution of Arabic vernaculars in the diaspora, in a diachronic and synchronic perspective, and the comparison with similar processes in the countries of origin is imperative in order to better understand the linguistic processes of urbanization and koineization affecting the Arabic language in general.

6.2 Moroccan Arabic in the diaspora Moroccan emigration has been largely orientated first toward France, subsequently toward Holland, Belgium, and Germany and more recently, toward Spain and Italy. Accordingly, though sharing some common characteristics, the numerous Moroccan communities in Europe lived in different social and linguistic situations. Linguistically, these immigrant groups are quite varied, and several factors have influenced the evolution of their vernaculars. First, the dominant interactive language is distinct in each country, and second the time factor has had an important impact, due to the fact that the linguistic situation of Morocco has evolved since the departure of the early immigrants up to the arrival of the recent ones. With these factors in mind, I have chosen the Spanish–Moroccan Arabic case to demonstrate that the evolution of an Arabic dialect within a minority language context varies from one host country to another, or from one city to another, even if in contact with the same foreign language in the same country.3 With regards to Moroccan Arabic spoken outside of Morocco, the French and Dutch cases are the best known, particularly concerning the processes of interaction between Moroccan Arabic4 and the dominant languages, French and Dutch respectively, through the studies of linguistic practices triggered by contact and bilingualism (Boumans 1998; Boumans and Caubet 2000). According to Boumans and de Ruiter (2002), the language uses of a Moroccan community in the diaspora does not mirror the language situation 124

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

of the country of origin.5 The varieties spoken as mother tongue or lingua franca by these Moroccan immigrants may differ from the local varieties spoken by the monolingual or even bilingual urban speakers in Morocco. A number of European cities are characterized by the presence of a large community of Moroccan migrants coming from a single region, a factor that enhances the maintenance of specific Arabic and Berber vernaculars in these Europeans urban centers and their eventual development as lingua franca among the Moroccan communities. Therefore certain Arabic and Berber vernaculars happen to play a more important role in the diaspora than in the Moroccan urban centers. While this assertion may be correct, it must be qualified, as these diaspora communities are nevertheless different from one another. Accordingly, a number of social factors play an important role in the evolution of their mother tongues, as for example a greater or lesser level of contact with the country of origin, the time of settlement of the community in the host society, whether or not the dominant language is mastered, whether or not a “second” or “third” generation exists,6 the sociolinguistic profile of the community members and its homogeneity, the social networks and so on. All of these factors actively shape the evolution of mother tongues. This chapter seeks to analyze the dynamics of linguistic changes within Moroccan Arabic in different migratory situations taking into account the impact of extra-linguistic factors. To do this, I will study the process of linguistic integration of the population of Moroccan origin within two Spanish cities: Ceuta and Zaragoza. These communities differ on several points, including patterns and timing of their establishment as well as their relationship with the country of origin. Ceuta is a very distinctive city because its present-day Muslim community of Moroccan origin has had a prolonged development, over almost one and a half centuries, and is characterized by its geographical proximity to and continued relations with Morocco, particularly with the population from the northern region. In fact the geographical location of Ceuta and its situation as a Spanish enclave within North Africa makes it rather problematic to talk about a Moroccan diaspora, even if the city is officially Spanish. In this sense, Ceuta can be considered as an intermediate case between a situation of internal ruralurban migration and a situation of foreign migration. Whereas the community of Moroccan-origin Muslims living in Zaragoza settled very recently, and is mainly an “economic migration”, rather similar to those that took place in other European countries in earlier times. Contact with the country of origin in Zaragoza is often quite sporadic, normally limited to vacation periods. The analysis of these two concrete and distinct sociolinguistic contexts will confirm how the linguistic practices of immigrants differ depending upon the internal constituency of each group and upon the degree of their linguistic integration in the host society. These circumstances will have 125

ÁNGELES VICENTE

Map 6.1 Zaragoza and Ceuta

126

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

diverse linguistic impacts on both the evolution of their mother tongues and their relationship with the dominant language.

6.3 The city of Ceuta The history of the city of Ceuta illustrates its distinctiveness in relation to both other Moroccan cities and European cities. Its geographic space has been inhabited since Antiquity by diverse peoples. It was likely a site of Phoenician origins; but its status as an urban center was not confirmed until the Muslim medieval period, when Ceuta enjoyed, from an economic and cultural point of view, a lustrous moment. In the fifteenth century, following a short period under the Portuguese who had control of the Muslim city, Ceuta became a North-African Spanish enclave, a political situation which has endured until modern times, and which has provoked continuous Moroccan land claims on the territory.7 The Muslim population established itself in the city in the modern period starting circa 1860,8 and throughout the twentieth century, the population grew considerably, particularly with the beginning of the Spanish protectorate in the north of Morocco, from 1912–56. This population originated mostly from northern Moroccan regions, like those of Jbala, although others came from the Rcf and the Ses. It is clearly an old-established community existing far before the rise in immigration toward Spain during the last quarter of the twentieth century. This community is mainly made of old families born in Ceuta and completely rooted within the city over several generations, although the newer daily arrivals from Morocco must also be taken into account. This situation triggered the existence of a longstanding social and linguistic space, of particular interest for linguists. The city’s population has been steady for a long time. During the protectorate period it ranged between 70,000 to 80,000 people and is currently around 71,505 persons according to the 2001 census data. According to official statistics, Muslims constitute one third of the city’s population, though this percentage, while often repeated by Spanish authorities, has never been fully justified. In this way, aided by the natural growth of the Muslim community with a higher birth rate than their Christian neighbors, as well as the continued arrival of Moroccans from the other side of the border,9 there has been a notable increase in the Muslim population. Even if the numbers cannot be determined with great accuracy, it can be stated that the number of people within the two communities has started to become balanced, especially amongst the younger generation.10 The city’s population is thus divided between two highly dissimilar groups, who identify themselves as Christians and Muslims (respectively Cristiano and Moro or Musulmán in Spanish, and NaXrAni and Muslim in Arabic).11 Though the situation evolved in the second half of the last century, the Muslim population has generally been compartmentalized into specific 127

ÁNGELES VICENTE

peripheral neighborhoods, particularly concentrated in three of six city districts, also highlighting the segregation between the two groups in all aspects of their daily lives. Besides those two main groups, one finds the Jewish and Indian communities who used to be more numerous until the 1970s and now form very tiny minorities without any prominent role in the daily life of the city (Vicente 2005: 37). With regards to the urbanization process, in the modern and contemporary history of the city, Ceuta did not become an urban center until very recently. Thus, beginning with the fifteenth century Portuguese conquest, Ceuta was a military bastion, and shortly afterwards while under the Spanish, became a fortress and a prison. With a population consisting largely of soldiers and convicts, Ceuta did not become an urban center in the modern period until 1860, with its new status as free port, the considerable growth in its population, and a strong economic expansion; this date coincides also with the rapid arrival of Muslims on the other side of the border. They therefore participated in the city’s urbanization process, a circumstance with consequences in all facets of life for the local population. One specific district of Ceuta, called Benzú, deserves special attention because its linguistic characteristics and specific development help to better understand the evolution of Ceutí vernacular Arabic. The inhabitants of this neighborhood are linguistically distinct from the other Arabic-speakers in the city. Numerous Benzú inhabitants maintain ties with the rural Moroccan world, a linkage with repercussions on the daily lives of its population. These ties are due to its geographical location: it is quite close to the Moroccan territory; it is relatively isolated from the other Muslim city neighborhoods; and, it is very close to the last Moroccan village of Belliounech near the border. Between Belliounech and Benzú there is a border crossing, (distinct from the main official border of Tarajal), where there have been longstanding exchanges of people and various types of influences. Benzú is thus the Ceutí neighborhood with the most marked confrontation between tradition and change; it is a very symbolic space for Ceutí Muslims where ethnic, religious and sexual boundaries are more rigid than in any other of the city’s regions. This situation prompted the existence of a particular Arabic sociolect, which has evolved differently than the other Arabic sociolects of the city.

6.4 Ceutí Arabic At the linguistic level, Muslims in Ceuta form a community wherein the sociolinguistic dynamics are controlled by an asymmetrical power relationship between the status of the Arabic and that of the Spanish. The first functions as the vernacular mother tongue while the second is the dominant language, which acts as a prestigious variety in all public contexts, like in 128

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

education. Spanish is also the lingua franca used among the diverse linguistic communities of the city – i.e. the Moroccan Arabic-speaking community and the Spanish-speaking community – to communicate with one another, as the latter is almost exclusively monolingual. The consequence of this asymmetrical situation is the existence of a Moroccan Arabic and Spanish bilingualism which characterizes exclusively the Ceutí Muslim community. In order to better understand the evolution of the various Arabic sociolects, a brief diachronic survey is necessary. The Moroccan Arabic vernacular, originally brought to Ceuta by the first Moroccans who settled there in 1860, was a Western type, pre-Hilali and rural or mountain dialect.12 It was/ is therefore quite similar to the rural northern Moroccan dialects, like those from the Jbala region, and especially the northern Jebli dialects spoken on the other side of the political border.13 This is due to the regional provenance of the first Muslim inhabitants of the city, who came first from the villages close to the border. However, this panorama has evolved during the twentieth century. Today the linguistic situation of Ceutí Muslims is more diversified due to the language contact situation prevailing in the city and the development of contact-induced linguistic phenomena. But another important factor has played a crucial role in the linguistic evolution of Ceutí Arabic, i.e. the influence of the general linguistic evolution taking place in Morocco itself under the impact of migration and urbanization. In Morocco, like in other North African countries, a similar process has taken place since independence: the expansion of the main urban vernaculars, usually those of the political or economical capital cities like Casablanca in Morocco, Tunis in Tunisia and Algiers in Algeria. After a koineization/homogenization process, these urban vernaculars have become the prestigious varieties in each country, functioning at the national level and at the expense of the old city dialects and the sub-standard colloquial variations. However according to Caubet (1993, I: Introduction): “on assiste à un phénomène de koïnisation certain qui laisse cependant place aux particularismes (plus ou moins accentués) qui caractérisent telle ville, telle communauté ou telle tribu.”14 In this way, the Arabic spoken in Ceuta is influenced by the linguistic changes taking place in Morocco, thereby participating in the koineization process, due to Ceuta’s proximity to Morocco, separated only by a political border, but not by a geographical border. This proximity enhances daily contact between the inhabitants of the city and the Moroccan population, a situation also facilitated by the willingness of the Spanish authorities to permit easy passage to the population to cross the border (this status is granted only to residents of the city of Ceuta and to residents from two northern Moroccan provinces: Tangiers and Tétouan). This daily contact has provoked a sociolinguistic evolution within the Ceutí Moroccan community reproducing that of the country of origin, a fact usually difficult to attain for other Moroccan communities in more distant migration situations. 129

ÁNGELES VICENTE

Ceutí Arabic can be defined as a language which has evolved from a rural Jebli Moroccan dialect and which has been influenced first by the Moroccowide koineization process. In such a way, the actual prestigious Arabic variety of the city is characterized by the presence of certain characteristics of modern Moroccan urban dialects. A second influence is more specific to the northern region and is due to the proximity of the old city of Tétouan. The Tétouani dialect is considered a prestigious variety in this northern Moroccan area. It is a old-city dialect whose local prestige is rather similar to that of Fès or Rabat (Caubet 1998; Messaoudi 1998). Third, the interference of Spanish must not be forgotten with the appearance of several phenomena, which characterize bilingual situations like code-switching, borrowing and interference. These latter linguistic phenomena are most usual among the speech of young people (Vicente 2004) and will not be dealt with detail here. The result of this situation is a heterogeneous panorama where several sociolects can be found and where their existence is conditioned by the context and sociolinguistic profile of their speakers. Thus, the most conservative sociolect, characterized by its pre-Hilali and rural features, is spoken more by the inhabitants of the Benzú neighborhood, and especially by older women, than in any other neighborhood of the city. It has thus become a marked sociolect associated with conservatism, and has even been stigmatized. Meanwhile, the most widespread variety follows the two tendencies indicated above: first, a koineization process which follows the evolution of the national Moroccan koine but keeps a number of local prestigious particularities; and, second a trend toward a greater interaction with the Spanish language. This sociolect is the most prestigious in the city, particularly among the young generation. It deletes the rural features of the old dialect which clearly mark the rural origin of the first Muslims in the city.

6.5 Several features of Ceutí Arabic The data presented here were gathered during a fieldwork carried out over several visits to the city between 2001 and 2004. A systematic survey was undertaken to locate informants neighborhood by neighborhood and to find people of different ages and various educational backgrounds, in order to try to analyze the correlation between linguistic variables and social factors. Data gathering was done through surveys, questionnaires and recordings. The majority of the informants (approximately eighteen) were drawn from a Muslim family who had settled in the city for some time, whose members were numerous and were almost all born in Ceuta, except the eldest who had come from the Jbala region of Morocco. They helped me contact their neighbors and parents to meet informants in other Muslim neighborhoods of the city like, for instance, in Benzú. I also carried out research in primary and high schools with many students and even an Islamic religion instructor. 130

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

I will now present a limited number of features illustrating the two abovementioned trends which seem to play a determining role in the linguistic evolution of Ceutí Arabic: 1. 2.

The appearance of features associated with more prestigious Moroccan dialects, particularly features associated with the national koine. The presence of some features of pre-Hilali origin, characteristic of the northern Moroccan dialects and very often stigmatized by Moroccans from other regions.

The first trend leads to the disappearance of some pre-Hilali and rural features and their replacement by koine-like features. It is the main source of sociolinguistic variation amongst the population of the city, as the presence of koine features varies according to the speakers and the sociolects, the koine-features being more numerous in the prestigious variety while the old women of Benzú keep the pre-Hilali rural features. The second tendency, however, leads to the presence of several features common to the entire language community, in the sense that these features are to be found in both the prestigious sociolect and the more conservative sociolect like that of Benzú. This common presence is due to the fact that these features are of pre-Hilali origin but are associated with prestigious old-city vernaculars and not with rural vernaculars. They also existed in the Tétouani dialect and are maintained due to the prestige of the latter dialect at the regional level. 6.5.1 Phonetic features With respects to phonetics I will cite two examples of the koineization tendency. The first example is the weakening through affrication of several occlusive consonants. This feature has become a sociolinguistic variant, which distinguishes speakers from one another. The fricative pronunciation is characteristic of the more conservative group, i.e. the older women and Benzú residents, whereas occlusive pronunciation is typical of the majority of the population more influenced by the koineization process. One phoneme touched by this phenomenon is, for example, /k/, whose most frequent and dominant pronunciation is the voiceless palatal occlusive like in: kCkBs “cookies”, kAdba “liar”, k¢bb “cube”, k}Pl “black”. But there is also a voiceless palatal fricative allophone [ç], which always appears in inter-vocalic position or after a vowel as in: b¢çri “early”, qTlY l¢ç “I told you”, mAçla “meal”, dCç “that one”, bAbAç “your father”. However, among the conservative members of the Ceutí society, the fricative realization can appear after a consonant or in the initial position like in: çbCr “big”, çTnna “we have been”, dyAlçTm “your”, kYAbçTm “your book”. A second example of the influence of the Moroccan process of koineization on Ceutí Arabic is the evolution of the pronunciation of the phoneme /l/. 131

ÁNGELES VICENTE

This previously voiced pre-palatal sibilant is now realized as a voiced postpalatal occlusive [g] or as a voiced dental occlusive [d] depending on the Moroccan dialect. The first realization, for example gAz “he has passed” (on this phonetic evolution /l/ > [g], see Cantineau 1960: 61), is characteristic of the northern pre-Hilali dialects (Heath 2002: 136–38; Marçais 1977: 22), like those of Anjra and Chaouen (Moscoso 2003a: 44; Vicente 2000: 47) and therefore the old Ceutí Arabic. But a regressive dissimilation took place in other more prestigious Moroccan dialects where /l/ is pronounced [d] in contact with sibilant and fricative phonemes. This last articulation had spread and exists today in southern Moroccan dialects (for example in Skura, see Aguadé and Elyaacoubi 1995: 29) and in some northern dialects of the country (for example in Larache, see Moscoso 2003b). Thus, some speakers in Ceuta articulate this dissimilation and pronounce dAz “he has passed” instead of gAz, the conservative and rural form. The following examples illustrate the presence of northern Moroccan preHilali features in Ceutí Arabic: •







The pronunciation of /q/ as a voiceless velar fricative [x] in certain cases instead of the more common articulation as occlusive [q]. This characteristic feature is an influence from Tarifit (a neighboring Berber dialect), where the fricative articulation of the occlusive consonants is the customary pronunciation, on the northern Moroccan dialects. The articulation wPxt occurs instead of wPqt “time”, the most common form in the other Moroccan dialects, and also the words fEyAx and fCwAx “when?”. Another example, is the voiced velar fricative /γ/ which becomes voiceless in contact with the sibilant /s/ and is pronounced as a voiceless velar fricative [x] like in the verb xs¢l “to wash”, in all forms of its paradigm, instead of γs¢l which is more used throughout Morocco. The xs¢l form is most customary of the northern dialects and must also be included as an influence of the Tarifit dialect, where this devoicing also exists. A third example, is the case of the voiced pre-palatal fricative /l/ pronounced as a voiced pre-palatal affricate [u]. These two types of pronunciation of the same phoneme are characteristic of several sedentary northern Moroccan dialects, yet in the Hilalian dialects, the affricate allophone does not exist. In the northern Moroccan region, it typically appears in the case of gemination because of a phonetic dissimilation of the phonemes in contact, with the result of [du]. Examples: d-UAmA∞ “the mosque”, }AdUa “a pilgrim woman”, d-U¢llAba “jellaba”. In Ceuta, this allophone [u] can appear similarly in other consonant environments when there is not gemination, for example, when /l/ is in contact with /n/ or /r/. Examples: kA-nUPrr¢b “I try”, x¢rUu “They left”. A final example, is the devoicing of the emphatic voiced dental occlusive /\/. In Ceutí Arabic, this phoneme is normally pronounced as an 132

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

emphatic voiceless dental occlusive [y] like in: mrCYa “sick”, bPyYa “egg”, YPhri “my back”. It must be clear that the most common situation in the northern sedentary dialects of Morocco is the devoicing of this phoneme, and thus the [y] and not the /\/ pronunciation occurs quite regularly in Chaouen, in Taza or in Ouargha (Colin 1921: 40; Lévi-Provençal 1922: 20; Moscoso 2003a: 40; Vicente 2000: 42).15 6.5.2 Morphological features At the morphological level, two examples dealing with verbal morphology illustrate the influence of the general Moroccan linguistic shift upon the evolution of Ceutí Arabic. One is the case of the verbal form kAl / kla – yAkTl “to eat”, whose conjugation reflects the sociolinguistic variation occurring among Ceutí speakers. Thus, in the prestigious Ceuti sociolect the perfective conjugation of the verb follows the more current generalized Moroccan pattern, see Table 6.1. This kla-verbal pattern, similar to the defective verbal pattern, is found in the Casablanca dialect, whose features can be considered as part of a koine functioning at a national level. Following the koineization process taking place in Morocco, this pattern also now exists in some northern and preHilali city dialects, such as that of Tetouan, from where it passes to Ceutí Arabic. Nevertheless, the speakers of the Benzú sociolect and the older women of the city (i.e. the groups who keep the rural features present in the old Ceutí vernacular), reproduce the verbal paradigm common to the northern rural dialects, see Table 6.2. Table 6.1 Conjugation of kla-verb in prestigious Ceutí sociolect Singular

Plural

1c. 2c. 3m. 3f.

1c. klCna 2c. klCYu 3c. klAw

klCY klCYi kla klAY

Table 6.2 Conjugation of kla-verb in the conservative Ceutí sociolect Singular

Plural

1c. 2c. 3m. 3f.

1c. kTlna 2c. kTlYu 3c. klAw

kTlY kTlY kAl / kla klAY

133

ÁNGELES VICENTE

In the rural northern Moroccan dialects, this verbal pattern is irregular in the sense that it mainly reproduces the conjugation of hollow verbs, but it also reproduces the paradigm of the defective verbs for certain persons, such as the 3rd feminine singular (klAY) and the 3rd common plural (klAw). The two different kla- patterns reflect a variation linked to age and social factors (i.e. diastratic variation).16 It distinguishes the sociolect of the elderly (especially old women) from the variety considered the prestigious sociolect in the city and adopted by most of the younger generations. It also reflects a territorial–regional variation (i.e. diatopic variation), through which the speaker’s neighborhood of origin can be identified, like that of Benzú. Moreover, there is a third alternative, which appears to be an intermediary solution between the two previous uses. Thus, young people from Benzú say kTlYi “you have eaten”, a form which follows the rural verbal scheme kTlY (Table 6.2), but with the addition of the verbal suffix /-ti/, characteristic of the urban dialects in Morocco (Table 6.1). It is therefore a noteworthy mix or “interstitial” form.17 The second feature which functions as a sociolinguistic marker in Ceutí Arabic is the particle used to express the future, which varies according to the speaker or to the context. Three particles, mAKi, γAdi and ∞Adi, precede the imperfective, and give it the value of a close future. mAKi is common to the entire language community; it exists in both rural and old-city pre-Hilali Moroccan dialects. Being a feature of pre-Hilali dialects of both old-city dialects like Tetouan (Heath 2002: 216) and rural dialects like Anjra (Vicente 2000: 107), it has been kept in Ceutí Arabic: mAKi nKEf¢ç f-¢d-dAr dyAle “I will see you at her home” (old Ceutí sociolect), γPdda mAK ytC} K-Kta “it will rain tomorrow” (most common Ceutí sociolect). The second particle, γAdi and its lesser form γad are the most utilized in many Moroccan dialects. They appear in the Casablanca dialect (Heath 2002: 216) where they probably spread and replaced particles being used elsewhere. In Ceuta, the same process is taking place. Some speakers alternate between mAKi and γAdi while the speakers most influenced by the Moroccan koineization process use mainly γAdi, which does not exist in the more conservative sociolect. Examples of the most common sociolect in the city are:

γAdi n¢mKi m∞Pk “I will go with you”, l-bnA– γAd yLCw mP∞na “the girls will come with us”. Lastly, the third particle ∞Adi, which is very characteristic of Ceutí Arabic, is now used by the speakers of the most conservative sociolect, although it 134

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

is never entirely absent from the speech of other speakers. Examples of this particle are: ∞Ad n¢kY¢b l-bra “I will write you the letter”, ∞Adi YkEn Y¢mma nYCna? “you will be there?”. All these phenomena illustrate the proximity with Morocco and the coinfluences of various trends existing today in northern Morocco. In this respect, the dynamics of dialect change in Ceuta are representative of a wider Moroccan context. What appears more specific is the trend toward Moroccan and Spanish code-switching among the young Ceutí generation, a phenomenon unknown in the rest of Morocco and not yet well rooted among the Moroccan diaspora of continental Spain as we can see in the case of Zaragoza.

6.6 The Moroccan population in Zaragoza The case of Moroccans presently living in the city of Zaragoza, in the North of Spain, is completely different from the Ceutí case. First, because the community of Moroccan origin in Zaragoza settled quite recently, but especially, because it entails an economic migration similar to that of other immigrant communities in several other Spanish and European cities. The evolution, which characterizes this type of migration, is a repeated constant in several countries. At the beginning it consists of a largely masculine phenomenon spurred by economic motives. Then, due to the process of family grouping and to natural population growth, these communities increase and eventually come to include an important number of members of what is known as “the second generation”, i.e. children born in the foreign country from migrant parents but often counted as nationals. The majority of members of the Moroccan community arrived in Zaragoza at the end of the twentieth century. The largest growth of foreigners in the Aragon region (where Zaragoza is situated) was recorded in 1997. Moreover, the number of foreigners of African-descent has increased by 125 percent, in the same year, with 60 percent having settled in the city of Zaragoza, most of them coming from the Maghreb, particularly Morocco. According to the 2004 census, 3,316 Moroccans have officially settled in the city; the number of men (2,215) is double that of women (1,101). They mostly live in three of the city’s twelve districts, neighborhoods that run the risk of being stigmatized by others residents.18 The study of the linguistic integration of Moroccan immigrants in this city is particularly interesting due to the distinctiveness of Zaragoza migratory situation compared to the rest of Spain. Indeed Moroccans represented the largest foreign group living in the city at the end of the 1990s, a situation unknown in any other Spanish city.19 In this way, while Moroccans may not 135

ÁNGELES VICENTE

constitute a significant number, especially when compared to other Spanish regions, and particularly with other European countries, Moroccan Arabic is de facto the second language in the city in terms of the number of speakers. Moreover, several demographic characteristics of these immigrants have important consequences for the linguistic evolution of the community. The population is largely masculine, with 67 percent men and 33 percent women, mostly younger or of middle age, as most fall between the ages of 20 and 44; few people are part of the extremes of the population pyramid, i.e. children and elders.20 They generally work in construction and in commercial sectors.21 These characteristics demonstrate that the settlement of foreign immigrants in the city of Zaragoza is a recent phenomenon, where the family grouping process has recently begun. It entails a work-based migration, as had occurred beforehand in numerous European countries and in other Spanish regions.

6.7 Moroccan Arabic spoken in Zaragoza With regards to Moroccan Arabic, the linguistic diversity of the Zaragoza Moroccan community must be taken into account, due to its diversity of origins in the home country. The diversity of origin was highlighted by most members during surveys carried out on this topic, even if some believe that most of the community stems from the northern regions of the country. Official statistical data do not exist to confirm these assertions, however. In order to better understand the evolution of the mother tongue of these immigrants, I conducted a study in 2003 in Zaragoza. The data presented here are based on several conversations and recordings with five young male students in a Spanish language class for adults.22 From this field-study, it appears that the immigrants use a Moroccan Arabic variety very similar to the variety considered to be the national Moroccan koine, based on the Casablanca dialect. They leave aside regional variants so as not to betray their origins, and in order to be understood by the whole members of the community. Indeed the Moroccan koine is here the most appropriate tool for serving the needs of communication between people who do not share a close regional origin. Therefore, Moroccan immigrants in Zaragoza try to use a prestigious variant which can function as a lingua franca for everyone. Some of the newest migrants might have already spoken this koine before their departure abroad, but it is evident that, in Zaragossa, the homogenization toward the Moroccan koine is also produced in the diaspora, and is an attempt to mitigate the linguistic heterogeneity of the community. The fact that the case studied was undertaken in a language class might have reinforced the impression of a dominant use of the Moroccan koine. But the fact that the migrants are mainly young men living far away from their family, reinforces the need for using this koine instead of more local regional forms. In any case, the possible presence of several local particularities due to the immigrants’ various origins must not 136

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

Table 6.3 Conjugation of kla-verb in Zaragoza Arabic Singular

Plural

1c. 2c. 3m. 3f.

1c. klCna 2c. klCYu 3c. klAw

klCY klCYi kla klAY

be ignored and may even reappear in the near-future, in the context of social networks based upon the same geographical origins. I will thus return to several features already described in relation to the city of Ceuta to study their evolution in the Moroccan Arabic spoken in Zaragoza. For example, in relation to the paradigm for the verb kla-yAkTl “to eat”, the one most employed by the Moroccans living in Zaragoza is generally the one used in the Casablanca dialect (Aguadé 2003: 304), see Table 6.3. This coincides with the pattern belonging to the prestigious variant of Ceutí Arabic, and is due to a common factor: the influence of the Moroccan koineization process on the different variants of the language, even in the diaspora. What is different is the preservation of certain features of the preHilali and old city variant in Ceutí Arabic, which have disappeared or are in the process of disappearing from the Moroccan Arabic spoken in Zaragoza. This disappearance is occurring because the features referred to are more prestigious in the north of Morocco, where they passed over to Ceuta, but outside of this linguistic context, they are very marked and betray the origins of certain members of the community. Beside this, we have in Zaragoza a transitory situation where the dominant function of the language is communication between speakers wishing to erase the heterogeneity of their origins, a notion that does not exist in the city of Ceuta, where the Muslim community is homogenous enough and their linguistic situation is more stable. These features are, for example, the absence of the fricative articulation of the occlusive consonants, as it is considered a feature marking rural origins. In Zaragoza, some speakers therefore say fEqAK “When?” and I also heard fEyAq “When?” a form characteristic of the North, but without fricative articulation. The γs¢l form is generally used in Zaragoza and found throughout Morocco, for instance: γs¢l y¢ddu b-l-ma u X-XAbDn “he washed his hands with water and with soap”, once again without fricative articulation. There is also a sibilant pronunciation of the /l/ phoneme in all the consonant environments, as in L-LAmP∞ “the mosque”, xr¢Lna “we have gone out”, l-}¢LLAm “the barber”, etc. To conclude, the absence of the devoicing of the /\/ phoneme will be cited; it is normally articulated as an emphatic, voiced, dental and occlusive (examples: mrC|a “sick”, bPy|a “egg” and |Phri “my 137

ÁNGELES VICENTE

back”) and not like the voiceless form, characteristic of some of the northern dialects, also employed in the Ceutí dialect. These examples are a brief indication of the linguistic situation of the recently arrived Moroccan immigrants in Zaragoza. It is quite clear that these data are provisional and need to be corroborated by larger data on the evolution of Zaragoza Moroccan Arabic. But it already indicates that a different tendency is taking place in Zaragoza compared to Ceuta.

6.8 Conclusion This study attests that the linguistic evolution of Moroccan Arabic in Morocco is also affecting Moroccan communities in the diaspora: whether due to more sustained and historical migration (as is the case in Ceuta), or to more modern and economic reasons, as in Zaragoza. This corroboration contradicts several previous statements (Boumans 2004), which assert that Arabic dialects in the diaspora have a tendency to become fossilized in the migration community. Consequently, less prestigious or older dialectal features were preserved because of their distance from the center of influence in the home country and its evolutionary process. This might have been true with the old Moroccan diaspora in Europe whose members came before the generalization of the Moroccan koine but it seems less true for the recent migrant generation. Nevertheless, the diaspora sociolinguistic context, within which these Moroccan varieties develop, plays a determinant role in their evolution as a minority language. The two immigrant communities are developing themselves linguistically a priori in the same way, by following the same linguistic changes taking place contemporaneously in the country of origin, that is toward the Moroccan koine. However, the variable constituency of the group described above, is affecting the evolution toward the koine and does not make it similar. In the case of Zaragoza, the community which settled abroad more recently has become linguistically more homogenous than other Moroccan communities in Europe whose members immigrated a long time ago, at a time when the vernacular languages of the home country were more diverse and the koineization process was not that widespread. It must be noted that this linguistic homogenization exemplified by the dominant use of the Moroccan koine is facilitated by the shared social profile of the members of the Moroccan Zaragoza community who consists mainly of young male workers. In Ceuta, on the other hand, the linguistic evolution is more complex and is leading to more diversification. This evolution erases the former rural pre-Hilali features but keeps a number of pre-Hilali city features which no longer exist in the national Moroccan koine. The maintenance of this preHilali city feature is certainly due to the on-going influence of the locally prestigious Tetouan city dialect, facilitated by the geographic proximity 138

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

between the two cities. This evolution indicates a certain northern Moroccan particularism vis à vis the spread of the Moroccan koine. The retained features are typical of pre-Hilali sedentary Moroccan dialects, while as is well-known, the Moroccan koine is particularly shaped by elements of Bedouin-origin.23 Another important difference to highlight between the Moroccan communities abroad is the language spoken as a lingua franca among their members. It had been stated that the dominant language of the host society, generally a European language, might play this role, particularly among immigrants who settled abroad a long time ago and whose mother tongue is in sharp decline, especially among younger generations (Boumans and de Ruiter 2002: 267). However, the study of recently settled communities (like Zaragoza) indicates that the language of daily communication is Moroccan Arabic, despite its minority language status. In this case, in my opinion, the koine of the country of origin is utilized as a homogenizing element and is helpful in mitigating linguistic differences between members of the community. With regard to the dominant language, there is a different panorama in the two cases. In Ceuta, the second and third generations are completely integrated from a linguistic point of view because, especially for the youngest, they have been educated in Spanish schools where the official educational language is Spanish, the dominant language of the society. They are therefore bilingual, but with differing degrees of bilinguism. One notices the presence of linguistic phenomena, which characterize bilingual communities like code-switching, borrowing and interference.24 In Zaragoza however, these linguistic phenomena caused by a bilingual situation do not yet exist. Moroccan immigrants living in the city have not mastered the dominant language of the society where they have settled. They have often learned Spanish in the street and cannot read or write it. Consequently, Moroccan Arabic has remained vital as a minority language in Aragon’s society and there is no trace of the linguistic phenomena associated with bilingual societies.

Notes 1 See for example J. Owens’s edited work dedicated to vernacular Arabic as a minority language (Owens 2000), which includes cases of Afghanistan, Turkey and Israel addressed by Kieffer (2000), Arnold (2000) and Talmon (2000). 2 For lack of space no additional reference will be given to Arabic in non-Western countries, as it does not directly concern our chapter. 3 See Khellil (1996) for a social presentation of the Moroccan diaspora in Spain. 4 It must be noted that many Moroccans living abroad speak Berber as their mother tongue, a point to which I do not refer. 5 See Boumans and de Ruiter (2002: 262), who claim that “in the countries with the largest Moroccan communities, the Berber-speaking element is even more important than in Morocco itself ”.

139

ÁNGELES VICENTE

6 Second or third generation are the terms used in a number of European countries to define the children or grandchildren born in the host country from former migrant parents or grandparents. 7 This rather “colonial” situation is similar to that of another Spanish city situated in the north of the African continent, Melilla. 8 Before this date, Moroccans workers and traders who were assuring the food supply of the city were called moros de paz. They were forbidden to stay in the city and every night were obliged to cross the border towards Morocco (Planet Contreras 1998: 25). 9 The new migrants largely originate from neighboring regions, some arriving clandestinely and staying in the city without Spanish papers, while others have become Spanish, particularly in the case of women arriving in the city following marriage, generally in Morocco, with a Muslim husband of Spanish nationality. 10 The neighborhoods with a large Muslim presence have thus held the highest proportion of youths. It must be noted as well that a greater proportion of Muslim students than of Christian students attend primary public schools in the city; see Vicente (2005: 46–47) for more on this subject and to find the percentages of students in the city’s schools. 11 Christian Spanish speakers tend to use the word Moro, which has a pejorative connotation while Muslim Spanish speakers always use Musulmán. Young people use other collective names, which have sometimes derogatory connotations (Vicente 2004: 35). 12 For a comprehensive view of the dialect classification of Morocco see Aguadé et al. 1998, for the formation of the Ceutí dialect see Vicente (2005: 80–83). 13 See Vicente 2000 for the characteristics of the Jebli dialect spoken on the other side of the border. 14 For an overview of these phenomena in the Arab world see Miller (2004). 15 In other southern dialects it is most often pronounced as a voiced phoneme. 16 See Gadet (2003) for an explanation of the origin of the labels diatopic, diastratic and diaphasic commonly used in Romance studies since the 1950s. 17 With this term, I want to refer to the notion of “interstitial culture”, used by Louis-Jean Calvet in reference to the Chicago’s school of urban sociology (Calvet 1994: 269). 18 The three districts are: district 1 with 28 percent of Moroccan population, district 2 with 15 percent and district 10 with 16 percent (Pinos Quílez 2004: 35, who used information from Zaragoza city hall). 19 They are now the second largest group of foreigners in the city, but given that the first is of Ecuadorian-origin, i.e. Spanish speaking, the linguistic situation has not changed and Moroccan Arabic remains the second language. 20 The number of children born in Spain from Moroccan parents, however, grows every day, as seen in the city’s schools. On this subject, see Elósegui Itxaso (2004), where the author states that there were 571 students of Moroccan-origin educated in Zaragoza during the 2002/3 school year. Most are Spanish and are not counted as foreigners in the official statistics. 21 Outside the urban center, the population of North African origin and especially Moroccan work in agriculture, whatever their former academic background in Morocco. 22 They had all lived in Zaragoza for between three and five years. They came from different zones of Morocco (three from the Northwest, another from Oujda and the fifth from Meknes) and their relationship to one another in Zaragoza was limited to their class attendance.

140

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

23 The majority of the phonological and morphological features of this Moroccan koine stem from the Bedouin dialects, known as ∞rEbi (“rural, Bedouin”), found on the Northern plateau of the Middle Atlas and on a significant portion of the Atlantic coast (Caubet 1993, I: Introduction). 24 For more on these linguistic phenomena stemming from the contact between Moroccan-Arabic and Spanish languages see Vicente (2005: 165–190).

References Abu-Haidar, F. (1989) “Are Iraqi women more prestige conscious than men? Sex differentiation in Baghdad Arabic”, Language in Society 18: 471–481. Abu-Haidar, F. (1991) “Language and sex: the case of expatriate Iraqis”, in A. Kaye (ed.) Semitic Studies in Honor of Wolf Leslau on the Occasion of his Eighty-fifth Birthday, vol. 1, Wiesbaden: Ottro Harrassowitz, pp. 28–38. Aguadé, J. (2003) “Sur le dialecte arabe de Casablanca”, in I. Ferrando and J. J. Sánchez Sandoval (eds) AIDA Fifth Conference Proceedings, Cádiz: Universidad de Cádiz, pp. 301–308. Aguadé, J. and Elyaacoubi, M. (1995) El dialecto árabe de Skûra (Marruecos), Madrid: CSIC. Arnold, W. (2000) “The Arabic dialects in the Turkish province of Hatay and the Aramaic dialects in the Syrian mountains of Qalamûn: two minority languages compared”, in J. Owens (ed.), pp. 347–370. Boumans, L. (1998) The Syntax of Codeswitching: Analysing Moroccan Arabic/Dutch Conversations, Tilburg: Tilburg University Press. Boumans, L. (2004) “L’arabe marocain de la géneration ayant grandi aux PaysBas”, in D. Caubet et al. (eds) Parlers jeunes, ici et là-bas: pratiques et représentations, Paris: L’Harmattan, pp. 49–67. Boumans, L. and Caubet, D. (2000) “Modelling intrasentential codeswitching: a comparative study of Algerian/French in Algeria and Moroccan/Dutch in the Netherlands”, in J. Owens (ed.), pp. 113–180. Boumans, L. and de Ruiter, J. J. (2002) “Moroccan Arabic in European diaspora”, in A. Rouchdy (ed.), pp. 259–285. Calvet, L.-J. (1994) Les voix de la ville: introduction à la sociolinguistique urbaine, Paris: Éditions Payot & Rivages. Cantineau, J. (1960) Cours de phonétique arabe: suivi de notions générales de phonétique et de phonologie, Paris: C. Klincksieck. Caubet, D. (1993) L’arabe marocain. I: Phonologie et morphosyntaxe. II: Syntaxe et catégories grammaticales, textes, Paris-Louvain: Éditions Peeters. Caubet, D. (1998) “Etude sociolinguistique des traits préhilaliens dans un dialecte en voie d’urbanisation à Fès”, in J. Aguadé, P. Cressier and Á. Vicente (eds) Peuplement et arabisation au Maghreb Occidental: dialectologie et histoire, MadridZaragoza: Universidad de Zaragoza-Casa de Velázquez, pp. 165–175. Caubet, D. (2001a) “Maghrebine Arabic in France”, in G. Extra and D. Gorter (eds) The Other Languages of Europe, Clevedon-Buffalo-Toronto-Sydney: Multilingual Matters, pp. 261–277. Caubet, D. (2001b) “L’arabe dialectal en France”, in Arabofrancophonie, Les Cahiers de la Francophonie 10, Paris: L’Harmattan, pp. 199–212.

141

ÁNGELES VICENTE

Caubet, D. (2002a) “Transmission familiale et acquisition non didactique des langues: le cas de l’arabe maghrébin”, Communication à la journée Observer les pratiques linguistiques: pour quelles politiques? de l’Observatoire des pratiques linguistiques, DGLFLF, Ministère de la Culture (accessed 20-02-2002, http://www.dglf.culture. gouv.fr). Caubet, D. (2002b) “Métissages linguistiques ici (en France) et là-bas (au Maghreb)”, VEI-Enjeux (Ville Ecole Intégration) 130: 117–132. Caubet, D. (2004a) “About the transmission of Maghrebi Arabic in France”, in Language and (Im)migration in France, Latin America, and the United States: Sociolinguistic Perspectives, http://www.utexas.edu/cola/france-ut Caubet, D. (2004b) “La darja, langue de culture en France”, Les langues de France, Hommes et Migrations 1252: 34–44. Caubet, D. (2005) “Ce français qui nous (re)vient du Maghreb, mélanges linguistiques en milieux urbains”, Notre Librairie, Revue des Littératures du Sud, Langues, langages, inventions 159: 18–24. Colin, G. S. (1921) “Notes sur le parler arabe du nord de la région de Taza”, BIFAO 18: 33–119. El Minaoui, L. (2001) “Arabe marocain, tamazight et français: l’alternance codique chez un travailleur immigré d’origine marocaine”, in C. Canut and D. Caubet (eds) Comment les langues se mélangent: codeswitching en Francophonie, Paris: L’Harmattan, pp. 159–180. Elósegui Itaxso, M. (2004) “Interculturalismo e Islam en los centros escolares de Zaragoza”, in Á. Vicente (ed.) Musulmanes en el Aragón del siglo XXI, Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, pp. 105–139. Gadet, F. (2003) “Is there a French theory of variation”, International Journal of the Sociology of Language 160: 17– 40. Heath, J. (2002) Jewish and Muslim Dialects of Moroccan Arabic, London and New York: RoutledgeCurzon. Khellil, M. (1996) “La diáspora magrebí”, in Mª-À. Roque (ed.) Las culturas del Magreb: antropología, historia y sociedad, Barcelona: Icaria Antrazyt, pp. 201– 220. Kieffer, Ch. (2000) “The Arabic speech of Bactria (Afghanistan)”, in J. Owens (ed.) pp. 181–198. Lévi-Provençal, E. (1922) Textes arabes de l’Ouargha, dialecte des Jbala (Maroc septentrional), Paris: E. Leroux. Marçais, P. (1977) Esquisse grammaticale de l’arabe maghrébin, Paris: Maisonneuve. Melliani, F. (2000) La langue du quartier: appropriation de l’espace et identités urbaines chez des jeunes issus de l’immigration maghrébine en banlieue rouennaise, Paris: L’Harmattan. Messaoudi, L. (1998) “Traits linguistiques du parler ancien de Rabat”, in J. Aguadé, P. Cressier and Á. Vicente (eds) Peuplement et arabisation au Maghreb Occidental: dialectologie et histoire, Madrid-Zaragoza: Universidad de Zaragoza–Casa de Velázquez, pp. 157–163. Miller, C. (2004) “Variation and Change in Arabic Urban Vernaculars”, in M. Haak et al. (eds) Approaches to Arabic Dialects, Leiden-Boston: Brill, pp. 177–206. Moscoso, F. (2003a) El dialecto árabe de Chaouen (Norte de Marruecos). Estudio lingüístico y textos, Cádiz: Universidad de Cádiz.

142

TWO CASES OF MOROCCAN ARABIC IN THE DIASPORA

Moscoso, F. (2003b) Estudio lingüístico del dialecto árabe de Larache (Marruecos). Basado en los textos recogidos por Maximiliano Alarcón y Santón, Cádiz: Universidad de Cádiz. Owens, J. (ed.) (2000) Arabic as a Minority Language, Berlin-New York: Mouton de Gruyter. Pinos Quílez, M. (2004) “Musulmanes en Zaragoza. Nuestros vecinos desconocidos”, in Á. Vicente (ed.) Musulmanes en el Aragón del siglo XXI, Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, pp. 17–41. Planet Contreras, A. I. (1998) Melilla y Ceuta, espacios-frontera hispano-marroquíes, Melilla-Ceuta: Consejería de Cultura. Rouchdy, A. (1992) The Arabic Language in America, Detroit: Wayne State University Press. Rouchdy, A. (ed.) (2002) Language Contact and Language Conflict in Arabic: Variations on a Sociolinguistic Theme, London and New York: RoutledgeCurzon. Talmon, R. (2000) “Arabic as a minority language in Israel”, in J. Owens (ed.) pp. 199–220. Vicente, Á. (2000) El dialecto árabe de Anjra (Norte de Marruecos). Estudio lingüístico y textos, Zaragoza: Universidad de Zaragoza. Vicente, Á. (2004) ‘La négotiation de langues chez les jeunes de Sebta”, in D. Caubet et al. (eds) Parlers jeunes, ici et là-bas. Pratiques et représentations, Paris: L’Harmattan, pp. 33–47. Vicente, Á. (2005) Ceuta: une ville entre deux langues. Une étude sociolinguistique de sa communauté musulmane, Paris: L’Harmattan.

143

ÁNGELES VICENTE

144

GREETINGS IN BEIRUT

Part II URBAN VERNACULARS: CONVERGENCE AND DIVERGENCE

145

MARIE-AIMÉE GERMANOS

146

GREETINGS IN BEIRUT

7 GREETINGS IN BEIRUT Social distribution and attitudes towards different formulae Marie-Aimée Germanos

7.1 Introduction When various dialectal varieties and different languages meet in urban centers, speakers – both migrants and city dwellers – have to accommodate the contact of linguistic varieties. In multilingual situations, where there is no mutual understanding between them, speakers tend to adopt a common lingua franca in order to be able to communicate, whether it is the language of one of the linguistic groups, or an external lingua franca. The urban lect not only has a communication function, but also an identity one, in the sense that adopting it may become the sign of integration into the city and its values. In cases where various dialectal varieties meet, and mutual understanding is possible between their speakers, the adoption of a common lingua franca is not necessary for communicative purposes. There remains nevertheless, the question of the linguistic integration of groups into the city, and the directions of linguistic changes in urban areas. Urban sociolinguistics proposes to approach urban places as a space that produces new variations. Calvet (1994) thinks that urbanization can cause, along with the linguistic integration to a common vernacular, the appearance of new linguistic forms and languages that functionally replace the languages that disappeared.1 In her remarks about Blom and Gumperz’s 1972 study of code-switching in Norway, Lesley Milroy (1985) also points out that the reason why vernaculars resist pressure to the standard is that they have functions for the groups that use them; and that they might be a sign of the group identity and of the rejection of the “prestige” dialect: Thus, instead of positing a sociolinguistic continuum with a local vernacular at the bottom and a prestige dialect at the top, with linguistic movement of individuals in a generally upward direction, we may view the vernacular as a positive force: it may be in direct 147

MARIE-AIMÉE GERMANOS

conflict with standardized norms, utilized as a symbol by speakers to carry powerful social meanings and so resistant to external pressures. (Milroy 1985: 19) Both tendencies – linguistic integration and differentiation – are strongly linked to linguistic and extra linguistic values. It is thus necessary, before presenting this study on linguistic variables for greetings in Beirut, to highlight some facts about the city itself, and about the demographic changes it has experienced so far.

7.2 Growth, division and reunification: a presentation of Beirut urban area Beirut experienced a quick expansion in the twentieth century. Its population grew from 130,000 inhabitants in 1914 to 581,000 in 1988. The population of the whole Beirut urban area (including the suburbs) grew from 450,000 inhabitants in 1959 to 1,200,000 inhabitants in 1988 (Boudjikanian 1994). The conflicts it experienced later on disrupted its growth, and the number of inhabitants living inside the city’s administrative boundaries has not increased since the beginning of the civil war, mainly because of a steady emigration. In the meantime, the number of residents in the suburbs continued to increase. Until 1975, we could distinguish three distinct areas inside and around Beirut: the city center was essentially commercial, while the city around it was a mixed residential area from the communal point of view – Hamra and Ras Beirut being more mixed and socially prestigious than other areas (Mouzoune 1999: 36). Hamra eventually became an economical center at least as important as the balad (downtown) itself (Bourgey 1979). The suburbs were poorer and more homogeneous regarding the geographical and communal origins of their inhabitants.2 Rural Shiite immigrants settled essentially in southern and eastern suburbs,3 and their Maronites counterparts settled in northern and eastern suburbs. However, this setting changed after the outbreak of the civil war. The city was quickly divided in two separated areas: ∫ ar§iyye and γarbiyye (East and West Beirut). Two concomitant facts accompanied this division: first, the appearance of a demarcation line later known as the “green line” that separated the two sectors and that could be crossed only in some locations, and second, population transfers that occurred inside the city – and across the entire country. During the first two years of the war, 110,000 Muslims left East Beirut and 75,000 Christians moved out of West Beirut; some 50,000 other Christians also left West Beirut after 1983 (Mouzoune 1999: 69). Both division and internal migrations led to a relative communal homogenization of each of the city sectors. This new human geographical 148

GREETINGS IN BEIRUT

Map 7.1 Beirut

149

MARIE-AIMÉE GERMANOS

division was reinforced by the fact that refugees from other parts of the country usually settled in the area that was under the influence of their religious group. The division of the city was also consolidated by the destruction of the city center during the early stages of the conflicts. Banks and shops were forced to move out of downtown Beirut. They opened branches in both sectors of the city whose inhabitants lost the shared public space they had represented (Davie 1993: 2). Territorial, human and economical divisions put a stop to coexistence in the same quarters and suburbs of the capital between different communities, and political and social groups. Even after the end of the civil war, Michael Davie who worked on urban issues noticed in 1993 that the “frontier” was still very alive in the inhabitants’ minds.4 The unprecedented and tremendous growth of Beirut’s southern suburb is also to be partly regarded at as a consequence of the war: thousands of Shiite refugees fled from southern Lebanon and the Beqaa valley to settle there, especially after Israel’s invasion of the country in 1982. This area now known as the Da}yeh has grown enough to be considered as a settlement center at least as important as Beirut itself. In 1990 an estimated 400,000 to 500,000 persons lived in the southern suburbs which cover an area of 16 km2 while Beirut had 510,000 inhabitants with an area of 17.9 km2 (Harb el Kak 1996: 32). Besides, it is often considered as a particular space, relatively autonomous and self-sufficient, poor, and very homogeneous from the communal point of view.5

7.3 Greetings and sociolinguistic variation in Beirut Two major sociolinguistic studies were conducted in the 1980s in two areas of Beirut. Through rigorous phonological description of various idiolects, Naïm-Sanbar (1985), and Srage (1997) respectively described the pattern of phonological variation in Ras-Beyrouth and Moussaytbé. Other levels of linguistic analysis (morphological, syntactical or lexical) were nevertheless left apart in these studies and in the much earlier phonetic study of Mattsson (1910). The aim of this chapter is to focus on another kind of linguistic variable, greetings formulae; and it will deal with both their distribution and with attitudes towards them. These formulae have not been studied yet in the case of Beirut. The three studies mentioned above focused on phonological and not on lexical variables. Recent works conducted by the GRIC (Université de Lyon 2) in an intercultural perspective in small shops in Lebanon (Beirut), Syria, Tunisia, France and Vietnam, have given a “script” of interactions occurring in such situations in these countries (Dimachki and Hmed 2002: 2). But these studies did not focus on the greeting formulae that were used, and did not study their variation. 150

GREETINGS IN BEIRUT

The choice of greetings as a variable imposed itself because of the existence of a variety of terms for the same act – greeting – in Beirut, and because of the interactional and social significance of the greetings in general. Srage referred to the existence of a choice between different formulae in a study dealing with school diglossia in Beirut: Forms of greeting in Beirut illustrate well the great variety of social and linguistic levels. Thus, if I wanted to greet a Cheikh (religious notable) or a practicing Muslim, the best form would be the religious form: assalam alaykum “May my greeting be with you”. If I wanted to speak to a family member, colleagues, or my students, I use the form sabah-il->ayr. It means “good morning” and I find it more appropriate.6 (Srage 1987: 82) The author insists on the importance of social factors over the choice of a precise greeting formula. The example he gives also outlines the religious dimension of this choice. The formulae used to greet are usually interchangeable: they all occur at the same time in a conversation, function in adjacent pairs, and share other characteristics, such as conveying the same meaning (like the recognition of the other as a legitimate interlocutor, cf. Duranti (1997) ). It could thus prove interesting to study the correlations between particular formulae and social variables, especially as the “access rituals” described by Goffman (1973) have mostly been studied essentially from an ethnological or an anthropological point of view.

7.4 Attitudes towards greeting formulae In January 2003 we conducted interviews with 22 informants (see Appendix), during which the topic of greetings was brought up. Mainly five formulae were cited during these interviews: bonjour [bõLEr], mar}aba, hi [hAy], XabA} el->eyr and as-salAmu ∞alaykum. Those formulae can be classified into two categories: the ones that some informants called “neutral formulae” (mainly bonjour and mar}aba), and the ones that were said to be used “depending on the situation”. The first are said to be used “normally” – when the situation does not require a particular formula – while the use of the latter depends on social elements such as his interlocutor’s profile (gender, age, religion, linguistic competence, etc.), the moment of the day, or the formality of the situation. The use of some formulae of this second category was sometimes rejected by informants: they would be able to explain when others use them – or who would use them – but they would totally reject their use for themselves. 151

MARIE-AIMÉE GERMANOS

mar}aba was said to be a neutral greeting by eleven informants. One informant said he uses it “only” when he is not sure if his interlocutor speaks French or English. One informant said she would use it when “you have to greet in Arabic”. Four other informants linked its use to particular situations: they’d greet with a mar}aba if their interlocutor is Muslim, or if they are engaged in an informal conversation, or if they are greeting a lower class interlocutor. Nine informants considered bonjour instead as a neutral formula. One of them insisted on making clear that the use of a French word was not wrong in an Arabic speaking country such as Lebanon: The first thing I will think of (for greeting) is French. We got used to it so it is not wrong. This word, bonjour, is not wrong because we are used to it. I cannot say to someone XabA} el->eyr for example. Other informants said they would only use it when addressing a person that knew French or a Christian person. Three informants said they never used it; they described its use as unusual or pretentious. No one considered XabA} el->eyr as their normal greeting, although twelve informants said it was used frequently but depending on the situation. It was, to them, a term to be used in the morning (contrary to mar}aba that can be used during the entire day), in formal situation, or when addressing an elderly person. Four other informants said it was not a frequent expression. They said it was not easy to utter (“too long and too heavy”) and specialized for “written Arabic” (“If somebody greeted me using XabA} el->eyr, I would feel he is very influenced by Arabic literature”). All of the informants viewed hi as an informal formula. Its main advantage is its brevity according to a majority of informants. This term is said to be used amongst young people, Christians, students, or between friends. When asked, only two informants said they would use as-salAmu ∞alaykum themselves. Christians consider it a “Muslim” greeting, or a “non Lebanese” greeting (used by “Arabs” from Gulf States); and Muslim informants added a nuance to that statement, saying it was used among practicing Muslims. as-salAmu ∞alaykum is also the expression to which the biggest number of informants reacted negatively, and sometimes with violence. A 23-year-old unemployed Shiite woman said: Sometimes I say XabA} el->eyr and people answer me wa ∞alaykum as-salAm as if they wanted to force me to say it. I hate this. I am Lebanese, after all, as-salAmu ∞alaykum I feel it is- well it is a greeting after all. But I am Lebanese. XabA} el->eyr. The most violent reaction against this formula occurred during an interview with a 29-year-old Maronite teacher: 152

GREETINGS IN BEIRUT RESEARCHER : Do you sometimes use as salAmu ∞alaykum? INFORMANT : No way. There is no way. Over my dead body R : Why? I : I don’t like for example ∫ukran or as-salAmu ∞alaykum. Why the

hell am I going to say it? I have a hundred thousand other options. And why the hell should I say something that – R : Something that? I : You want me to be very honest R : Yes I : Well it is a Muslim word. I have nothing against Muslims but I’m proud to be Christian and that’s it. These interviews allowed us to outline the existence of three types of greetings, according to the speaker’s attitude towards them. Some expressions are said to be “neutral”, others are said to be suitable to some kind of situations, and some can be totally rejected. The same formula can belong to each of these three categories, depending on the informant. Sometimes it can belong to the two last types for a single informant. This raises the point of the absence of consensus among all informants over one single greeting to be used in any situation, without it being the sign of a categorization of each interlocutor by the other.

7.5 The use of greetings in Beirut: a distribution according to social, communal, and territorial factors 7.5.1 Presentation of the data collection In order to be able to determine the factors that would influence the choice of a certain greeting formula rather than another in Beirut, we carried out a field study of the greetings usage during January 2004. We had decided to select places where several encounters can happen in a row, to make it possible to have a significant amount of data involving greetings. We thus limited our observations to public places, such as shops (2 newspaper kiosks and 6 grocery stores), doctor’s offices (2), and fast-food outlets (4). We noted that 79.24 percent of the greetings were initiated by the person coming from the outside, and not by the owner of the place. Initially, we had only chosen the kiosks, fast-food outlets, and medical centers, hoping to observe as diverse a range of individuals as possible. One of the kiosks is situated in Concorde, a business area near Hamra. We went there on Monday morning (8.30 am to 11.30 am), when the employees, on their way to work, stopped there to buy their newspapers. The second kiosk is located in Sassine – a more residential area – and we went there on Sunday; it was then mostly frequented by people living in the immediate vicinity. 153

MARIE-AIMÉE GERMANOS

As for the fast-food outlets, two of them were actually two branches of the same brand; one is located in Sodeco, near Saint Joseph University (French curriculum), and the other one is opposite the gate of the American University of Beirut, on Bliss Street. Most of their customers are students, generally from middle or higher social classes. The prices of the other two outlets (in Hamra and Achrafieh) are more affordable, and their menu is also more “classical” and restricted, consisting of one or two types of sandwiches. They are frequented by the residents of the area and by low and middle class workers during the day. Concerning the physician, it was in fact a hearing aid specialist who had two offices, one in Hamra and one in Jal el Dib (northeastern suburb). His patients were either elderly persons, or very young children accompanied by their parents. This allowed us to observe different age brackets and social levels. After this first stage of field work, we noticed that we had not heard some of the expressions that were cited during the interviews (including as-salAmu ∞alaykum that we had only encountered once). This formula had been considered as a “(practicing) Muslim” expression. But all of the areas where we had conducted our researches were central areas, with a certain level of communal and social mixing. Some other areas in Beirut and its suburbs being more marked, socially and communally, so we decided to extend our observations to such areas. In order to be sure that a majority of the observed interlocutors would be inhabitants of the selected areas, we decided to restrict our observations to groceries. The grocer would be able to inform us whether every customer is a habitué – and thus lives in the neighborhood – or not. We selected six areas in this second step. The first grocery was in Borj Hammoud where the majority of the inhabitants are of Armenian origin (the grocer who received us was himself Armenian), the second one was in Haret Hreik, in the southern suburb (inhabited mostly by Shiites), the third one was in Sabra, with lower level social classes formed by Palestinian refugees, displaced Lebanese Shiites and Syrian workers, who clustered in the area according to their nationalities (in three separated sectors). Our grocery owner was Lebanese, with a majority of Lebanese clients. We also went to Nabaa (northern suburb) where there are groups of Armenians, displaced Lebanese Christians, Syrians and Kurdish workers; to Nweyri, close to Basta; and, to Karm el Zeytoun. This enabled us to gather two other formulae: as-salAmu ¡alaykum and barew (Armenian). We noted 480 interactions in all. Of these, 342 began with at least one person greeting the other while the remaining 138 encounters took place without any greeting. For each encounter, we wrote down the gender and approximate age of the interlocutor, whether he was a habitué or not, all of the routine formulae (thanks, greetings, good byes, and so on . . .), and, when necessary, the gestures and the possible relation of kinship between the interlocutors or any other possibly relevant remark. 154

GREETINGS IN BEIRUT

In the analysis of the data that follows, we only take into account the first part of the adjacent pair of the greetings, and only the interlocutor who greeted first, whatever his role was (client, doctor, patient . . .). When speaking of a percentage of use for each formula, or of the number of times it was used, we refer only to the first formula used by the first person to greet, and not to all the greetings exchanged. We do not take into account the answer to the first greetings or anything that follows it. The fact is that the formula used by the first person directly influences the formula used in reply: in a majority of cases, the answer is either the “expected” semantic one (ex: bonjour/bonjour(eyn) or as-salAmu ∞alaykum/ wa ∞alaykum as-salAm), or a “complementary” greeting (for example, the couple “how are you?/fine”, as defined by André-Larochebouvy 1984: 69).7 It was rare that an interlocutor who was greeted with a particular formula produced an unexpected answer (bonjour/masa el->eyr). It sometimes happened that he did not answer at all. 7.5.2 Data analysis The “first” formulae found in the collected data are (number within brackets refers to number of occurrences): bonjour or bonsoir (118), mar}aba (65), hi (40), as-salAmu ∞alaykum (28), kCfak (23), ahla and its variants ahlan and ahlBn (9), XabA} el->eyr (9), ∫u (6), ya∞tCk el ‘afye (5), hello (3), barew – pronounced as parev (8), bari irkown – pronounced payergun (2), sa‘Cde (1), salamAt (1), salAm (1), hi bonsoir (6 occurrences, but with one single person). The most frequent formulae are therefore bonjour, mar}aba, hi and as-salAmu ∞alaykum. Different criteria favor the use of each one of them. The most important criteria are territorial, social, and religious. The other formulae were not frequent enough in the collected data, and it is therefore impossible to propose a relevant interpretation of their use. Nevertheless, we will also present a short analysis on the use of barew and bari irikown, as they were only heard in the two areas where a majority of Lebanese from Armenian origin settled (Nabaa and Borj Hammoud). 7.5.2.1 bonjour This is the most frequently used formula in the collected data: 118 persons used it, i.e. 34.50 percent of the “first greetings” observed. Its use does not vary significantly within different age brackets, but it is used by 46.15 percent of women and only by 28.44 percent of men. It also varies according to the type of place where the encounter takes place. It was more frequent in 155

MARIE-AIMÉE GERMANOS

the doctor’s office (58.02 percent of the “first greetings” in both offices), especially in Jal el Dib (69.39 percent of the “first greetings” are a bonjour). On the other hand, it was rather rare in the grocery stores (17.16 percent of first greetings). The difference between Jal el Dib (69.39 percent of bonjour) and Hamra (40.625 percent) offices reflects the territorial distribution of the use of this formula. Bonjour is much more frequent in those areas that were part of East Beirut (and northern and eastern suburbs of the city; with Jal el Dib being part of these suburbs), than it is in areas anciently part of West Beirut (and southern suburbs). In West Beirut (i.e. Bliss, Concorde, Hamra, Haret Hreïk, Nweyri, Sabra districts) 152 greetings were observed and 183 in East Beirut (i.e. Borj Hammoud, Jal el Dib, Karm el Zaytoun, Nabaa, Sassine and Sodeco districts). Table 7.1 Percentage of bonjour in East and West Beirut

Number of bonjour Total number of greetings Percentage of bonjour

East Beirut

West Beirut

84 183 45.90%

33 152 21.71%

There is another remark about each one of those two groups: bonjour is generally more used in areas inside Beirut’s administrative boundaries, generally mixed, with middle or upper class areas inhabitants or with business centers, than in those areas situated in the suburbs or in popular residential areas inside Beirut. In Borj Hammoud and Nabaa (northern suburbs), bonjour’s percentage is respectively 26.92 percent and 23.08 percent, while it is 51.52 percent in Sassine and 52.32 percent in Sodeco. Not one person used bonjour in Haret Hreik and Sabra (southern suburb). Only one person used it in Nweyri, while in Hamra, Concorde and Bliss, the percentage of bonjour is significant: respectively 34.04 percent, 29.63 percent and 33.33 percent. It appears therefore that bonjour is more used in “Christian” areas, and less used in “Muslim” areas and is also less frequent in popular areas: it is almost never used in popular Muslim areas. Women also tend to use it more than men, and it might be used more frequently in formal situations (such as the doctor’s office), than in informal ones (such as in the grocery store). 7.5.2.2 mar]aba The mar}aba greeting occurred 65 times – 19.05 percent of all first greetings observed. Its use varies with the age of the interlocutor, and the type of situation in which he is involved. Contrary to bonjour, mar}aba is more 156

GREETINGS IN BEIRUT

frequent among men (23.56 percent of them used it) than among women (10.26 percent). Also contrary to bonjour, it is more frequent in groceries (26.12 percent of greetings in groceries are a mar}aba), and less frequent at the hearing aid specialist’s office (6.17 percent of the greetings). In the fast food outlets 10.67 percent of greetings were a mar}aba, but this only occurred in the outlets located in Hamra and Achrafieh – the ones mostly attended by workers and area inhabitants. Only one person said mar}aba in the two branches of the fast food outlets with a younger clientele formed by students mar}aba is apparently more frequent in informal situations (groceries), among men, and in places frequented by workers. Furthermore, it was more frequently used by persons aged between 40/59. Table 7.2 Percentage of mar}aba by age groups Age bracket

−18

19/29

30/39

40/49

50/59

+60

Percentage of use of mar}aba

14.29%

14.04%

17.07%

27.40%

25.53%

8.33%

7.5.2.3 hi The corpus contains 40 hi – as a first greeting – in other words 11.70 percent of all greetings. It is more frequent among women and younger interlocutors, and – not surprisingly – near the American University, in places attended by its students. 62.50 percent of its occurrences were by women – leaving only 37.50 percent by men – whereas the majority of the persons observed were actually men (225 men, and 117 women), i.e. 21.37 percent of the 117 women used it, and only 6.67 percent of the men. Table 7.3 Percentage of hi by age groups Age bracket

−18

19/29

30/39

40/49

50/59

+60

Percentage of use of hi

14.29%

19.30%

12.20%

6.85%

2.13%

0.00%1

Note: Only 8 persons were more then 60 years old.

One last and unsurprising remark: the formula was very frequent in the two fast food outlets frequented by young people and students (Bliss and Sodeco). In Bliss 37.50 percent of greetings were a hi while 19.05 percent of those in Sodeco were a hi. The difference between those two places is more than probably due to the proximity between the Bliss street snackbar and the American University of Beirut. 157

MARIE-AIMÉE GERMANOS

7.5.2.4 as-salamu ¡alaykum as-salAmu ∞alaykum is the fourth most frequent formula in our data: 8.19 percent of all first greetings (28 occurrences). Its distribution varies according to social and territorial factors. A majority (89.29 percent or 25 occurrences) of those using it are men. It is also more widely used in groceries (27 occurrences; the last occurrence happened in the kiosk in Concorde). But both these remarks are not relevant if not related with the territories of the usage of this formula. Contrary to the other formulae, there is a clear division between three areas (Nweyri, Sabra and Haret Hreik) where this formula is massively used, and all the other areas where it hardly ever occurs. It occurred only once in Concorde and once in Nabaa, but it was each time a Syrian supplier – not a client – who used it. We also noticed that two similar formulae – salAm and salAmAt – were also used only by Syrians interlocutors. Of people entering the grocery – and greeting – in Nweyri 25 percent used as-salAmu ‘alaykum, 42.11 percent did in Sabra, and 73.68 percent in Haret Hreik. This is probably why it had a higher rate of usage in groceries than in other situations (the three places visited in these areas were groceries). This also can explain the difference of usage between men and women – only 11 women greeted in those 3 groceries. They greeted as follows: Nweyri (5 women): 3 mar}aba; 1 hi; 1 bonjour Sabra (4 women entered, but only one greeted): 1 mar}aba Haret Hreik (5 women): 3 as salAmu ‘alaykum; 2 mar}aba Not too many women greeted in those areas, but a majority of them preferred another formula (7 out of 11) than as-salAmu ∞alaykum. On the contrary, the majority of men in Nweyri, Sabra and Haret Hreik (23 out of 30 men who greeted) preferred as-salAmu ∞alaykum. Territorial distribution is therefore very important to understand the usage of as-salAmu ∞alaykum. As the inhabitants of the three areas mentioned above are mostly Shiites, we are led to suppose that this formula has a similar function to that of Bavarian “Grüss Gott” which Coulmas (1979: 252)8 refers to in his article about routine formulae: the recognition of group identity. It was also interesting to notice that a rather new formula, salAm al-masC}, has recently appeared on the Télé Lumière, a religious Christian TV channel, during Christian religious broadcasts, while it seems to be absent from daily life encounters. Its construction with salAm reminds us of as-salAmu ∞alaykum, and it might possibly be a Christian religious “identity” response to this formula. When asked about it, informants conversant with salAm al-masC} said it probably has it roots in the message and gesture exchanged in the church during the mass. 158

GREETINGS IN BEIRUT

7.5.2.5 barew and bari irikown Although we heard barew 8 times and bari irikown only 2 times, both their territorial distribution and the remarks about their comparative meanings and uses given by two informants needs to be mentioned. Exactly like as-salAmu ∞alaykum, both formulae were heard in two areas only. Nabaa and Borj Hammoud are places where an important Armenian diaspora lives. One of the two groceries we visited in Nabaa was owned by an Armenian family: we heard 4 barew and 2 bari irokown, which represents 42.85 percent of the 14 greetings that we noted in this store. In Borj Hammoud, only 15.38 percent of the interlocutors used barew. Of course both formulae are very linked to Armenian identity, and they are therefore highly used in sectors were Armenian community lives in Beirut. There is one last hypothesis regarding barew/bari irikown distribution among Armenian interlocutors: it might be possible that older people use bari irikown more frequently than barew – while younger people would prefer using barew. The two persons who used bari irikown in our corpus were over 50 years old, and all those who used barew were under 50. But we cannot of course generalize with such a small quantity of data. We asked each of the grocers about the meaning of both terms. The first grocer was 18 years old. He answered that barew was like mar}aba, and that bari irikown was more sustained, like masa el->eyr. He also said only “old people” use bari irkown. The second grocer was 35 years old. He gave a different explanation, arguing that barew can be used at any time of the day, while bari irkown was only to be used in the evening, and he compared it to bonsoir. It is likely that both sociolinguistic (older/younger usage) and semantic (evening usage) differences overlay. 7.5.3 Remarks concerning Nweyri, Sabra, and Haret Hreik During our study we noticed that the use of as-salAmu ∞alaykum was restricted to the areas of Nweyri, Sabra and Haret Hriek. On the other hand, bonjour and hi – two of the most frequently used greetings in other parts of the city – were almost never used in these three areas. They were not heard at all in Sabra and Haret Hreik. Each of them was used once in Nweyri: hi was used by a friend of the owner (a woman) who came to visit him, and bonjour was also used by a woman (a customer). To that customer, surprisingly – it rarely occurred in the rest of the corpus – the grocer answered using masa el->eyr.9 These three areas differed from the others at different linguistic levels – and not only greetings. It is therefore necessary to present them briefly. Both Sabra and Haret Hreïk are in the suburbs, while Nweyri is an area of Beirut (located near Basta).

159

MARIE-AIMÉE GERMANOS

Sabra is located near the Chatila refugee camp. In the second half of the twentieth century and until 1985, the population living in the area was mostly Palestinian refugees. After the massacres of 1982 and after what was know as the “war of the camps”,10 Sabra was almost totally destroyed and its population fled and took refuge in the Chouf and in southern Lebanon. According to the UNRWA, an estimated 18,000 to 31,000 refugees left the refugee camps situated in Beirut in 1985. They could hardly return to their homes after the end of the war, mostly because of large-scale destruction of the housing (Doraï 2002). The area is therefore now inhabited by few Palestinian refugees, by a significant number of displaced Shiite Lebanese families and by many Syrian workers, who live there in very precarious conditions. Until the middle of the century, Haret Hreïk was still a village located close to Beirut and its population was then mostly Maronites. The number of its residents began to increase – as it did in the neighboring villages – after migrants came to settle there, mostly originating from Shiite villages (from the Baalbek region). In 1950, the Haret Hreïk–Hadeth area was estimated to have 12 percent of Shiite inhabitants, and 55 percent of Maronites inhabitants (Nasr 1979). With the permanent settlement of migrants in it, and with the growth of the Beirut urban area, Haret Hreïk soon became a suburb of the Lebanese capital. During the war, and particularly after 1983, the Maronites fled from Haret Hreïk, while, in the meantime, an increasing number of Shiites displaced from southern Lebanon settled there and in the neighboring areas. The residents of Haret Hreïk are now mainly Lebanese Shiites. The area hosts the headquarters of the Hezbollah political party. Nweyri is located inside the administrative boundaries of Beirut, near Basta, and thus in what was once known as West Beirut. Its population is for the most part constituted of Lebanese Sunni and Shiites inhabitants. The owner of the grocery we stopped in was a Shiite. The second linguistic observation we can make about them – though this has to be verified in a more systematic survey – is that inhabitants of these areas use less borrowings and code-switch less than inhabitants of the other selected areas (we had also noted the occurrence of code-switching and the different borrowings we have heard). The last remark concerns other routine formulae: the ones used in Newyri, Sabra and Haret Hreik are mainly Arabic, while in other areas of the study, they are French or English. The formulae most frequently used to thank in Haret Hreik, Sabra and Nweyri are: yeslamo (and sallemun, yeslamo ha dayyBt etc. . . .), ∫ukran and kellak zD’. Merci was noticed twice in Nweyri, but never in Sabra and Haret Hreik. Thanks/thank you were never heard in these areas. In all other areas of the study, it is merci that is most frequently used – all other formulae including thanks and thank you are also used everywhere else, but merci is by far the most frequent formula, followed by yeslamo. 160

GREETINGS IN BEIRUT

As for leave-taking, bye (and yalla bye) are the most frequent in all areas of the study, except Nweyri, Sabra and Haret Hreik once again. We only heard it once in Nweyri, and once in Sabra. Formulae used instead in these three areas (and that sometimes – but rarely – occur in other places) are: ma∞ es-salBme, ya∞tik el-∞Afye, alla ma∞ak, alla ywaff §ak.

7.6 Conclusion This study demonstrates that the use of greetings in Beirut is linked to social factors, such as the gender or age of the interlocutor. bonjour and hi can then be considered as more “feminine”, and hi as “younger”. Usage is also linked, to a lesser extent, to the formality of the situation – bonjour being more formal than mar}aba. Last but not least, territorial factors can influence the usage of at least three greetings: bonjour, and, to a greater extent, as-salAmu ∞alaykum and the pair barew/bari irikown. Those are greetings more frequently used in some areas, and can therefore be linked to a sectarian identity (Christian (bonjour), Shiite (as-salAmu ∞alaykum), and Armenian (barew/bari irikown). Thus greetings must be considered as sociolinguistic markers, sometimes referring to the identity of a group. Since informants are aware of the existence of such markers and tend to categorize them, they can also sometimes be considered as stereotypes (Labov 1973), especially in the case of as-salAmu ∞alaykum. Some other words can also be treated as stereotypes as tB∞e (my, mine) for example. Some informants mentioned it as the marker that allowed them to identify someone as “Shiite” – the non-marked variable being taba∞e. It is therefore possible to confirm that some linguistic variables are identified with communal groups in the Beirut urban areas. This coincides with phonological remarks made by Naïm-Sanbar (1985) and Srage (1997). A certain number of phonological variables were identified by NaïmSanbar in her study in Ras-Beirut, most of them being relevant from the gender or from the communal point of views. Some of these variables were also outlined in the work of Srage in Moussaytbé. The major variables identified by both researchers in the two areas are the extension of emphasis, the emphasis of /™/ and the following vowel and the distribution of /q/. The realization of emphasis was said to distinguish between the speakers of then East Beirut (de-emphasizing) and those of West Beirut (strong emphasis), the latter variety being thus considered as “heavy” and being identified as a masculine, popular, Muslim, and West Beiruti variety opposed to a more “occidental-like”, feminine, East Beirut and deemphasizing variety. Nevertheless, in later remarks, both researchers refer to what can be considered as a certain kind of koineization in Beirut. In his introduction Srage states: 161

MARIE-AIMÉE GERMANOS

The young Beiruti generation (Sunni and Orthodox Greek in general), uses grosso modo a Lebanese “constituted” urban dialect, distinct from the traditional dialect of Beirut, and from the “Maronite” and “Druze” dialects of the Mountain and of the Metn caza. This urban dialect shares numerous features with the dialect used by the inhabitants of Achrafieh, a Beirut Christian area that groups together Orthodox Greeks and Maronites.11 (Srage 1997: 30) As for Naïm-Sanbar (1985: 217), in the later remarks of her work, she speaks of a “koinic zone”. In the Beirut divided by a demarcation line, these koinic zones had no geographical dimension; they could be found in an informant’s speech and were linked to the presence of a stranger or to the raising of some topics during the conversation for example. They could be defined as the disappearance of distinctive linguistic traits in order to produce a neutral variety. To these koinic zones may correspond, on the other hand, identity zones, where the usage of a distinctive trait is not to be avoided, but might be, on the contrary, socially necessary for a speaker to be fully identified as a member of the group. Now that Beirut has been reunified and that the physical obstacle to the existence of a territorial geographic koinic zone has disappeared, it would be interesting to investigate whether or not a koine is emerging. The idea of such a koine appears to be linked to the disappearance of both extra linguistic (Maronite, Druze, Mountain, Metn), and linguistic particularities (traditional dialect of Beirut), each one “trying to erase distinctive traits from his use, to let it become somehow neutral” (Naïm Sanbar 1985: 217).12 The hypothesis of the existence of a koine needs to be explored, with further identification of its linguistic features, if it exists, and with investigation of its relations to identity zones and distinctive traits.

Appendix: profile of the 22 informants Gender

Age (approximate)

Rite

Occupation

Man Man

24 19

Maronite Orthodox Greek

Man Man Man Man Man Woman

21 42 50 45 50 22

Maronite Maronite Armenian Maronite Catholic Greek Maronite

Unemployed Student (American University of Beirut) Student (Notre Dame University) Small trader Tailor Audio-Visual technician Teacher Nurse

162

GREETINGS IN BEIRUT

Woman Woman Woman Man Man

60 45 50 19 25

Maronite Protestant Protestant Sunni Shiite

Man Woman

54 21

Sunni Shiite

Woman Woman Woman Woman Woman Woman Man

19 23 26 48 60 29 29

Sunni Shiite Sunni Sunni Sunni Maronite Maronite

Retired Teacher Teacher Employee Student (American University of Beirut) Small trader Student (American University of Beirut) Student (University Saint Joseph) Unemployed Employee Small-trader Retired Teacher Employee

Notes 1 (Calvet 1994: 62): “La ville est en effet un creuset dans lequel viennent se fondre les différences – et, au plan linguistique, cette fusion est productrice de langues à fonction véhiculaire –, mais elle les accentue en même temps, comme une centrifugeuse qui sépare divers groupes, séparation qui, au plan linguistique, produit des formes grégaires.” 2 Davie (1996: 38): “Pour schématiser, la période [1960–1975] a vu la perduration de la situation antérieure, à savoir le maintien de la mixité de la ville et une consolidation de banlieues aux identités confessionnelles marquées.” 3 For a detailed study of the transformation of two villages (Ghobeire and Chiyah) of Beirut outskirts into suburbs, with a study of the social consequences of such a transformation, see Khuri (1975). 4 During field researches for this study, in 2003 and 2004, we noticed that ∫ ar§iyye and γarbiyye were still being used by some informants. Furthermore, we heard three significant slips that could only be understood as a negation of the “Beirutness” of each of the two formerly separated areas (West and East Beirut). Once an informant said for example she was “going back to Beirut” while she was going from Mar Elias (area of war time West Beirut) to the Museum sector (close to the ancient “green line”; she was actually going to the eastern part of this sector). 5 Harb el Kak (1996: 14): “Au sein de Beyrouth, la banlieue sud représente plusieurs spécificités. D’abord, sur le plan socio-démographique, c’est un des plus grands bastions communautaires; sur le plan spatial, c’est une entité clairement délimitée par des axes routiers majeurs et caractérisée par des pôles urbains importants; ensuite, sur le plan économique, la banlieue sud regroupe des commerces, des banques et des petites industries qui lui permettent une certaine auto-suffisance; enfin, sur le plan politique, elle est gérée par des partis politiques fortement implantés qui ont longtemps échappé à l’autorité de l’état.” 6 Original text “Les formes de salut à Beyrouth illustrent bien la grande variété de strates sociales et linguistiques: assalam alaykum ‘Que mon salut soit avec vous’ qui est une forme religieuse. Pour m’adresser à ma famille, à mes collègues, ainsi qu’à mes étudiants, j’utilise la forme sabah-il->ayr qui signifie ‘bonjour’, ce qui me paraît le mieux adapté” (Srage 1987: 82).

163

MARIE-AIMÉE GERMANOS

7 André-Larochebouvy (1984: 69): “Les salutations complémentaires sont une amorce à un développement du thème qui servira alors d’exorde avant d’aborder d’autres sujets moins anodins.” 8 Coulmas (1979: 252): “In short, the group identifying or group identity reinforcing effect is another important function of routine formulae (. . .) If a non-native of Bavarian German cannot bring himself to use this formula, he will always be identified as a foreigner upon the first word that he has a chance to utter, even if he has acquired a next to perfect Bavarian accent.” 9 In the entire data, 97 interlocutors in all answered to the bonjour or bonsoir greeting. 70 of them answered using bonjour, bonsoir or bonjoureyn, 13 said ahla (welcome), and 10 answered with a complementary greeting (kCfak or comment ça va?). Only a minority of 4 interlocutors answered with another formula: the grocer in Nweyri said masa el->eyr, and the 3 others said hi. 10 The war of the camps (May 1985–July 1986) opposed the Amal Shiite militia and the Palestinians in the Palestinian refugee camps in Beirut. It began in Sabra, which fell very soon under the control of Amal (May 30, 1985, after 11 days of conflict). 11 Original text: “La jeune génération beyrouthine (sunnite et grecque orthodoxe en général), utilise grosso modo un parler libanais citadin constitué, bien distinct du parler traditionnel de Beyrouth, et des parlers maronite et druze connus dans la Montagne et le caza du Metn. Ce parler citadin comporte de nombreux points et traits communs au parler utilisé par les habitants d’Achrafieh, quartier chrétien de Beyrouth regroupant des grecs orthodoxes en même temps que des maronites” (Srage 1997: 30). 12 Original text “Chacun essayant d’effacer de son usage les traits discriminants, de le rendre en quelque sorte neutre” (Naïm-Sambar 1985: 217).

References André-Larochebouvy, D. (1984) Introduction à l’analyse sémio-linguistique de la conversation, Paris: Crédif. Boudjikanian, A. (1994) “Beyrouth 1920–1991: d’une métropole de croissance au champ de guerre”, in F. Kiwan (ed.) Le Liban d’aujourd’hui, Beirut: Cermoc, pp. 241–273. Bourgey, A. (1979) “L’évolution du centre de Beyrouth de 1960 à 1977”, in D. Chevallier (ed.) L’espace social de la ville arabe, Paris: Maisonneuve et Larose, pp. 245–278. Blom, J. P., and Gumperz, J. J. (1972) “Social meaning in linguistic structures: code-switching in Norway”, in J. Gumperz and D. Hymes (eds) Directions in Sociolinguistics: the Ethnography of Communication, New York and London: Holt. Calvet, L.-J. (1994) Les voix de la ville. Introduction à la sociolinguistique urbaine, Paris: Payot & Rivages. Calvet, L.-J. (2002) “La sociolinguistique et la ville. Hasard ou nécessité?”, Marges Linguistiques. Lieux de ville: langue(s) urbaine(s), identité et territoire. Perspectives en sociolinguistique urbaine. Marges Linguistiques, 3: 46–53. Online resource (accessed April 23, 2005): http://marg.lng.free.fr/documents/01_ml052002_calvet_lj/ 01_ml052002_calvet_lj.pdf Coulmas, F. (1979) “On the sociolinguistic relevance of routine formulae”, Journal of Pragmatics 3 (3/4): 239–266.

164

GREETINGS IN BEIRUT

Davie, M. (1993) “A post-war urban geography of Beirut”, Warwick, EURAMES conference. Online resource (accessed April 23, 2005): http://almashriq.hiof.no/ lebanon/900/902/MICHAEL-Davie/Post-War.html Davie, M. (1996) “ ‘Beyrouth-Est’ et ‘Beyrouth-Ouest’: territories confessionnels ou espaces de guerre?”, in M. Davie (ed.) Beyrouth: regards croisés, Tours: Urbama. 22–49. Dimachki, L. and Hmed, N. (2002) “La comparaison interculturelle: exemple du remerciement dans les interactions de commerce en France, au Liban et en Tunisie”, Journées d’études: analyse des interactions et interculturalité, Lyon, 18/12/2000 and Montpellier, May 2001. Online resource (accessed April 23, 2005): http:// marg.lng6.free.fr/documents/doc0071_loubna_d/doc0071.pdf Doraï, M. K. (2002) “Les réfugiés Palestiniens du Liban en diaspora. Une géographie de l’exil”, PhD thesis, Université de Poitiers. Duranti, A. (1997) “Universal and culture-specific properties of greetings”, Journal of Linguistic Anthropology 7: 63–97. Online resource (April 23, 2005): http:// www.sscnet.ucla.edu/anthro/faculty/duranti/reprints/greetings.pdf Goffman, E. (1973) La mise en scène de la vie quotidienne. 2. Les relations en public, Paris: Les Editions des Minuit. Harb el Kak, M. (1996) Politiques urbaines dans la banlieue-sud de Beyrouth, Beirut: Cermoc. Khuri, F. (1975) From Village to Suburb: Order and Change in Greater Beirut, Chicago: The University of Chicago Press. Labov, W. (1973) Sociolinguistic Patterns, Philadelphia: University of Pennsylvania Press. Mattsson, E. (1910) Etudes phonologiques sur le dialecte arabe vulgaire de Beyrouth, Paris: Upsala. Milroy, L. (1985) Language and Social Networks, London: Blackwell. Mouzoune, A. (1999) Les transformations du paysage spatio-communautaire de Beyrouth (1975–1996), Paris: Publisud. Naïm-Sanbar, S. (1985) Le parler arabe de RAs-Beyrouth, ‘Ayn al Muraysa, Paris: Geuthner. Nasr, S. (1979) “Les formes de regroupement traditionnel (familles, confession, communautés régionales) dans la société de Beyrouth”, in D. Chevallier (ed.) L’espace social de la ville arabe, Paris: Maisonneuve et Larose, pp. 145–198. Srage, N. (1987) “La diglossie et le bilinguisme scolaire à Beyrouth”, in Actes du XIVème colloque international de linguistique fonctionnelle, Louvain-la-Neuve: Cahiers de l’Institut de Louvain, pp. 71–74. Srage, N. (1997) Etude sociolinguistique du parler arabe de Moussaytbé (Beyrouth), Beirut, Publications de l’Université Libanaise.

165

JANET C. E. WATSON

8 LINGUISTIC LEVELING IN SAN¡ANI ARABIC AS REFLECTED IN A POPULAR RADIO SERIAL 1 Janet C. E. Watson

8.1 Introduction Over the past 44 years, Yemen has witnessed momentous social and political changes, which have lead to a 35-fold increase in the population of the capital city, San¡a. As a result of these changes, people from many different parts of Yemen have come into contact with other Yemeni and non-Yemeni Arabic dialects, and, through increasing literacy rates, with Standard Arabic. This contact has in turn led to the acquisition of different registers among some of those whose social circumstances have changed significantly, the acquisition of different dialects, and a limited degree of linguistic leveling. In this chapter, I examine the extent to which the language used in a San¡ani radio serial reflects linguistic leveling in Greater San¡a. I chose to examine the language of episodes in this serial because of the conflicting views it arouses among Yemeni listeners: while the author of the serial claims to use a luγah mubassaYah “simplified language”, listeners tend to say that the language is either pure old San¡ani, or Yemeni but not San¡ani, or “Classical” Arabic. The language of serials is an artificial production, since it is planned, written and performed; it can therefore not be regarded as a reflection of actual speech; however, serials have an enormous audience, they exaggerate local linguistic features for the sake of entertainment, and they reflect local people’s representation of their speech forms. The chapter is organized as follows: Section 8.2 presents a concise sociopolitical history of Yemen and San¡a; Section 8.3 discusses the radio series; Section 8.4 analyses three representative episodes of the radio series in terms of the use of San¡ani Arabic (SA) morphemes and lexemes, pan-Yemeni/ pan-Arabic expressions, and non-SA Yemeni expressions; Section 8.5 discusses pause phenomena in SA. Using the speech analysis program, Praat 166

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

(www.praat.org), I then compare a sample of pause realizations from the episodes with those from the speech of one male and one female SA speaker. I suggest that it is not the systematic linguistic elements that arouse such different evaluations of the language among listeners, but rather the lowlevel phonological features, including, most saliently, pause phenomena.

8.2 Yemen Situated in the southwest corner of the Arabian Peninsula, bordered to the west by the Red Sea, to the south by the Indian Ocean, to the east by Oman, and to the north and northeast by a vast uninhabitable area of desert, Yemen is an oddity within the Arab world. It has never been fully colonized by either western or eastern powers. It has no exportable natural resources. And, importantly, it has existed as a natural unit since times of antiquity, traditions of Yemen before Islam are at the heart of Islamic literature, and deeply embedded in local literature and cultural practices. However, Yemen has never enjoyed unified political power. This lack of unified power is partly due to its geographical makeup – from west to east, the long flat coastal plain by the Red Sea, followed by a series of impenetrable mountains, the central plateau, and finally the inhospitable eastern desert; partly due to doctrinal religious differences between the Zaydi (Shi¡ite) north and the Shafi¡i (Sunni) south; and partly due to the strength of local tribes and the local dominance of the shaykhs (cf. Dresch 2000). Urbanization has indeed taken place, as we will see below; however, migration to the towns has not diminished a sense of belonging to the original locality: all “outsiders” speak of their bilAd, the home village of the family, irrespective of whether they were born there or in the town, and irrespective of whether they have ever visited the village; during the period of the Eids, people who still have family in the bilAd disappear from the town to their home village, and the towns become very quiet. The significance of localized power is seen even today, at a time of democratic government, where large areas of Yemen, particularly to the north and east of the capital, slip in and out of control of the central government. This juxtaposition of lack of centrality besides a strong sense of locality is both reflected and reinforced in a myriad of dialect distinctions, distinctions which are maintained not only in the localities themselves, but also in the towns to which people have migrated. As a result, it will be a long time before local dialects are leveled to identifiable urban vernaculars. 8.2.1 1800s to Republican Yemen The mid- to late-nineteenth century is known by historians of Yemen as “the time of corruption” (Dresch 2000: 4). The coffee trade, which had once made the country prosperous, had decayed; the Red Sea ports had been 167

JANET C. E. WATSON

SAUDI

ARABIA

IRAQ IRAN

Najra¯ n

S ¯I R

TH

cA

EG F UL

I

H.u¯th Shaha¯ rah n y Khamir afa ar Sh alH.ajjah

e Loh. ayyah

SUDAN ERITREA

Y E M E N

C LI UB EN REPYEM OF

INDIAN OCEAN

ETHIOPIA

cAmra¯ n

H

d

J. H . uf a¯ s h

O

A

T

R

U P P E R

U.A.E.

AN

E D S

J. Barat. H.arad.

SAUDI ARABIA

M

S.acdah

RE

J. Ra¯ zih.

Abu¯ cAr¯ısh

EGYPT

J¯ıza¯ n

a k ¯ı l d - B a n d h i ¯ s H. a

Khawla¯ n ibn cA¯mir

S.abya¯

al-Mah. w¯ı t

Sanaa

Ma’rib

Khawla¯ n al-T.iya¯ l



Mana¯ khah

H.odeidah

M

L O W E R Dhama¯ r

S

Rada¯ c

Yar¯ım

Zab¯ıd

Y E M E N

A

e

al-Bayd.a¯ ’

Ibb

a

Tacizz

H Mokha¯

Indian Ocean Aden 0

100 km

2000m 1000m 500m

Map 8.1 Yemen (from Weir 2007)

lost; and the Qasimi state, ruled by descendants of the Zaydi Imam al-Qasim (d. 1620), which had once held the highlands strongly, simply fell apart as a result of warring claimants to the Imamate, famine, and disease. Two foreign powers laid claim to parts of Yemen at this time: Britain in Aden and the Ottomans in the north. In 1839, Captain Haines took Aden for the East India Company’s Bombay Presidency, partly in response to the Ottoman threat from the north. The Ottomans, through Muhammad Ali Pasha, had occupied the southern highland town of Ta¡izz in 1837. They established a presence on the Red Sea coast in 1849, and, following the opening of the Suez Canal in 1869, took San¡a in 1872. Several claimants to 168

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

the Imamate were active at this time, but in 1904, the Imamate passed to Yahya Muhammad Hamid al-Din, whose father had claimed the Imamate since 1890 (Dresch 2000: 5). With Zaydi support, Yahya launched a rising in northern Yemen. In the midst of a general famine, San¡a fell in 1905. Fighting between Yahya’s forces and the Ottomans continued as far south as Qa¡tabah, and many small southern towns remained loyal to the Turks. With the loss of enormous numbers of troops, the Ottomans retook their capital, only to fend off another vast uprising in 1911. Italy’s invasion of Libya, however, forced the Imam and the Ottomans to reach a truce, whereby the Turks retained control of much of the Shafi¡i areas to the south while San¡a became a type of condominium. Ottoman control was finally destroyed by World War I. Between 1918 and 1934, Yahya conquered much of northern Yemen, though there were constant skirmishes with tribes throughout the period of his rule (1904–1948) (for details, see Dresch 2000). Two important treaties were signed in 1934, treaties which gave Imam Yahya’s state northern and southern borders – the Treaty of San¡a, signed with the British in February 1934, which agreed a southern administrative division approximating the old Anglo-Turkish border; and, more disastrously, the Treaty of Ta¡if, signed with Ibn Sa¡ud in May 1934 following the expulsion of Yahya’s troops from Najran, which effectively drew a northern border south of Najran, splitting northern parts of historical Yemen from the modern Yemeni state. In 1948 in a nationalist republican coup, Yahya was murdered. The Crown Prince, Ahmad, sought refuge to the north of San¡a. With help from tribes he advanced on San¡a and crushed the revolution within four weeks, thereafter establishing himself in Ta¡izz as the new Imam Ahmad, al-Nasir li-Din Allah. He ruled for 14 years, until September 1962. 8.2.2 Change in Republican Yemen Shortly after Ahmad’s death, Imami rule in North Yemen was finally overthrown by the revolution of 26 September 1962, when the building in San¡a in which the new Imam, al-Badr, was working was surrounded and shelled. Al-Badr escaped unscathed, but the Imamate did not. This was the birth of the Yemen Arab Republic, often also referred to as North Yemen. The end of the Imamate and the beginning of the Republic was to have wide-reaching political and social implications, particularly in the towns. Although Yemen had had contact with the outside world during the later stages of the Imamate, the last two Imams (1904–1962), in particular, had sought to protect their citizens from outside influences: Yemen had no secular schools; the radio was banned until 1938 (Scott 1942: 131); no consulates or embassies were allowed (Mu™ayyad al-¡Azm 1986: 130); there were no cinemas, restaurants or hotels (ibid.: 185), and banks did not exist (ibid.: 130). Since 26 September 1962, the country, and in particular the towns and 169

JANET C. E. WATSON

cities, have seen tremendous social, demographic, economic and political change. A combination of the September Revolution, unification with the South, which took place on 22 May 1990, and democratization culminating in Yemen’s first local elections in February 2001, has changed Yemen from a self-isolating hereditary autocracy to one of the most dynamic, open and democratic societies in the Arab world. The first secular school was opened in 1963, and the University of San¡a in 1970 (Mission 1980: 46). Television was introduced in 1975, and by 1980 reached 85–90 percent of the total territory (ibid.: 53). Health care improved for all. At first illiteracy rates began to drop gradually, then between 1985 and 1998 they fell by a staggering 27 percent (IMF 2001: 117). Above all, infant mortality fell and the population began to increase significantly, and to become mobile. Between the first demographic census in 1975 and the third in 1986, the population of North Yemen increased by c. 4.1 million – from 5.3 million in 1975 (Steffan 1978, cited in Dresch 1989: 16), to 8.5 million in 1981 (CPO 1983: 25), and to 9.4 million in 1986 (CPO 1989: 37). At the time of unification (1990), the population of the Republic of Yemen was 11.3 million (CSO 1994: 3), rising to 15.8 million in 1994 (CSO 1998: 3), and to over 18 million by 2001 (Banerji and McLeish 2002: 1; CCA 2001: xiii). Most importantly for this study, increasing mobility had a significant effect on the urban–rural population ratio. During the Imamate, North Yemen had an overwhelmingly predominantly rural population with few towns of over 5,000 inhabitants. The 1962 Revolution was followed by a large increase in internal migration to the towns: according to the 1972 socio-demographic survey of the city of San¡a, 68 percent of the workers at the city’s only textile mill had migrated into San¡a between 1960 and 1970 (Mission 1980: 25). In 1975, the population distribution of North Yemen was 11.4 percent urban, 88.6 percent rural (CPO 1983: 6);2 by 1981, the urban population had risen to 12.4 percent of the total (ibid.). It was in the 1990s, however, that a sea change took place in the ratio; by 1997, the percentage of the total population living in urban settlements had increased to 34.4 percent (IMF 2001: 144). These figures do not express the effect of migration on the capital city, however. In 1900, Aden was almost as large as San¡a, with a population of 40,000. During the 1960s, it was not entirely obvious whether San¡a or Ta¡izz would become the major modern city. While San¡a had been the Ottoman capital and the seat of Imam Yahya, it was Ta¡izz that had been the seat of Imam Ahmad. By the 1970s, however, Ta¡izz could no longer keep pace with San¡a. Estimated population figures for San¡a from the middle of the nineteenth century until the late 1930s fluctuated from 40,000 in 1856 (Playfair 1970: 28), 50,000 in 1891 (Bury 1998: 80), 20,000 in 1905 (ibid.),3 to between 40–50,000 in the late 1930s (Scott 1942: 122). By 1975, however, the population of San¡a had grown to 135,000 (CPO 1975: 5); between 1975 and 1986 the city grew by more than 10 percent a year (CSO 1992: 10) with a population of over 211,000 in 1981 (Dresch 1989: 16); and by 2000 a conservative 170

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

estimate of the population was given as 1.7 million. Thus, the population of San¡a is fast approaching 10 percent of the population of the entire unified country, a figure that is all the more significant given that 25 years earlier only 9 percent of the total population of North Yemen lived in settlements larger than 5,000, and 6.4 percent in settlements larger than 50,000 (Steffan 1978, cited in Dresch 1989: 21). In addition to changing the demographic composition of the city such that the pre-Republican San¡anis are now a small minority within Greater San¡a,4 this population increase has completely altered the shape of the city. Until 1962, San¡a occupied a neat figure of eight in the middle of the San¡a plain stretching from west to east: the original walled ancient city in the east, and the once Turkish diplomatic area of BCr al-∞Azab separating the old city from the Jewish quarter, QA∞ al-YuhEd, in the west. It had not changed in shape since Niebuhr visited the city in 1763. After the Revolution, the city of San¡a began to spread beyond its walls. Expansion took off hugely in the 1990s following unification with the South and, as a result of Yemen’s stand in the Gulf War, the forced return of 880,000 Yemeni migrant workers (CCA 2001: 1) from Saudi Arabia and the Gulf. Today Greater San¡a is spilling down the Ta¡izz Road, and has reached the hills that encircle the plain.5 The importance of migration and the role of migrant workers in the socioeconomic history of Yemen cannot be underestimated. A lack of significant natural resources and an increase in oil production in the oilproducing states led to huge numbers of migrant workers traveling to Saudi Arabia and the Gulf. In 1970, it was estimated that 300,000 Yemeni migrants worked in the oil-producing states; by 1975, a census reckoned there were 630,000 migrant workers out of a total population of 5.3 million; and by the end of the decade the number of migrant workers had reached circa 800,000, and they were dwarfing the revenue of the central government with remittances of 1.3 billion US dollars (up from 40 million US dollars in 1969–1970, Dresch 2000: 131). Lack of rain in the early 1980s and the promise of high returns emptied many villages of men-folk for long periods of time. Changes occurring since the Revolution have all had an effect on the language spoken within Greater San¡a. Migration from rural areas to the towns results in exposure to other Yemeni dialects; external migration (and the return of migrants) results in exposure to other Arabic dialects and, to a lesser extent, other languages; education and literacy result in exposure to the written language, and provide an awareness of the world beyond the immediate local environment; television and radio add to the effect of internal and external migration, but also encourage a reduction in traditional leisure activities with a resultant loss of some traditional terminology. Part of the result of increased outside contacts is a certain degree of language leveling. 171

JANET C. E. WATSON

Map 8.2 San¡a

172

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

8.3 The radio series The radio series, Mus¡id wa-Mus¡idih, which has been broadcast by Radio San¡a on a twice daily basis since 1988, takes the form of a dialogue between a man and his wife. The man’s name is Mus¡id, the woman’s name Mus¡idih, names which although known in other parts of the Arab world are, to the best of my knowledge, unused in Yemen. The program lasts between three and five minutes and follows the same basic structure: it is introduced by the theme song, which is based on a well-known proverb; then one of the characters introduces the topic – often in an oblique manner, perhaps through a proverb or song, often criticizing the other for their behavior or attitude; the other character then responds, trying to get to the bottom of the issue; once the issue becomes clear, the two characters usually take opposing sides for the bulk of the episode, each presenting commonly held views on the subject; towards the end of the episode a consensus is reached in favor of the more socially responsible action; finally the winning side of the argument is summarized by one or other of the two characters, usually in a classicizing Arabic register. The episode is concluded by repetition of the theme song. Episodes deal with social issues that are either on-going (such as pollution, or girls’ education), immediate (such as the local elections in February 2002), or seasonal (such as Ramadan, the Eid, end-of-year school exams). Mus¡id wa-Mus¡idih has molded its own Arabic genre drawing on traditional Arabic literary devices. It has the characters of a soap opera, but unlike the soap opera, each episode is complete in itself and marked by resolution. The dialogues examine past events, present issues or future planning, but lack the action of either the soap opera or the serial. Unlike soap operas in the West or the Arab world, Mus¡id wa-Mus¡idih is based in no place more specific than Central Yemen. And unlike in soap operas and serials, the characters have no fixed identity other than being Yemeni, being married to each other, and not being highly educated. From episode to episode everything else can vary: their age, their parentage, the number and age of their children, where they live, their habits. This is deliberate. Mus¡id and Mus¡idih represent everyman and everywoman, and their universality enables them both to raise and solve problems across the whole spectrum of society and to speak for the poor and the not so poor, for the farmer and the townsman, for the old and the child-bearing. The principal aim of the writer, ¡Abd al-Rahman Mutahhar, is didactic – namely, to convey a social message which should be acted upon to a target Central Yemeni audience. The secondary aim is entertainment. This enhances the didactic element because the audience is more receptive to an entertaining program than to one which simply commands or forbids, and because the entertainment engages the listener and allows them to view the issue from different standpoints.

173

JANET C. E. WATSON

Mus¡id wa-Mus¡idih is probably the best-loved radio program in Central Yemen. It is listened to by young and old alike, in houses, in taxis, and in the markets. Linguistically the interest lies not so much in the actual language employed, but rather in the varying conflicting opinions it arouses in listeners: some consider the language to be that of “old” San¡a, others – particularly women from the province of Ibb – consider it to be Classical Arabic. The author himself says that he writes in a luγah mubassaYah – which he defines as a simplified form of San¡ani Arabic – and it is this, the extent to which Mutahhar’s luγah mubassaYah and the actors’ rendering of this language reflects linguistic leveling in San¡a, that I intend to examine in this chapter.

8.4 SanÄani Arabic and the language of the programme In addition to some rather specific lexical items, San¡ani Arabic stands out most clearly from surrounding dialects in its tense voice quality and in specific low-level phonological and phonetic features – the labialization of high vowels following an emphatic consonant within the word, as in manA|ug “regions” versus makAtib “offices”, the labialization of labial and velar consonants immediately following /u/, as in ummwih “his mother” and ∫uggwih “flat”, pre-pausal diphthongization of word-final long vowels and of preconsonantal long vowels, as in ifta}aw “open up m.pl.!” and kabeyr “big”, and glottal closure and glottal constriction in pre-pause position (cf. Jastrow 1984; Naïm-Sanbar 1994). The map in Behnstedt (1985: 58) shows that glottalization in pause is not a feature confined to San¡a. This is, of course, not surprising, since glottalization of syllable-final consonants, including the presence of a glottal stop in association with a final stop or as its replacement, is a common phenomenon in the world’s languages (Michaud 2004: 119). However, glottalization has been used as a cover term for the following three types of phenomena: (1) glottal stop – a gesture of closure with limited coarticulatory effects on the voice quality of surrounding segments; (2) glottal constriction – a tense gesture of adduction of the vocal cords extending over the entire syllable rhyme; and, (3) laryngealization – a lapse into creaky voice with irregular vocal cord vibration, lacking the tenseness of glottal constriction. In San¡ani, glottalization is far more tense and more salient than in surrounding dialects, realized, depending on the phonological environment, principally as glottal closure and glottal constriction. The actors in Mus¡id and Mus¡idih have both lived in San¡a effectively since childhood – ¡Abd al-Rahman Mutahhar in SEg al-Bagar in the center of the Old City and Habibah Muhammad in the old Jewish quarter of al-QA∞. Neither of them have San¡ani parents, though in both cases the parents lived in San¡a and came from villages close to San¡a in the Central

174

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

Plateau. They understand San¡ani and command San¡ani syntax, morphology and phonology. In their everyday speech, however, they lack some of the phonetic features of speakers from well-established San¡ani families. The aims of the program are education and entertainment. In order to attain these aims, issues need to be presented in a way that is simultaneously comprehensible, plausible and humorous. Comprehensibility is achieved partly through the use of a fairly rigid structure within each episode. Plausibility is achieved through discussion of a topic to which the Central Yemeni audience can relate, such as immediate or on-going social issues. And humor is at least partly achieved by exaggerating features of everyday conflict married couples experience in Yemen. Since the target audience comes from Central Yemen, linguistic comprehensibility, however, conflicts at least partially with plausibility and humor. For the episodes to be plausible, the couple must speak naturally, with appropriate use of the discourse devices of interruption and turn taking. If they are to speak naturally, they should preferably speak a single dialect. Humor often exploits the use of wellknown dialect expressions, proverbs, popular songs, play on salient dialect terms, accent and voice quality. However, comprehensibility outside the central dialect area – in this case, San¡a – demands that the more salient dialect terms be avoided and replaced by pan-Yemeni or pan-Arabic terms, unless their sense can be understood from context and tone or is explained through explanatory apposition. In this chapter, I consider the morphemes and lexemes in three episodes, zawAj al-badal “exchange marriage”, al-}alCb al-XinA∞i “artificial milk”, and man zawwaj zallaj “he who marries [his daughter off ], wipes his hands of her”. These three texts are published in Arabic in Mutahhar and Watson (2002) and a small extract is provided here in the Appendix. Using the speech analysis program, Praat, I then examine pausal phenomena from five episodes,6 and through comparing spectrograms from Mus¡id and Mus¡idih with those from native speakers of San¡ani, suggest that it is here, at the sub-conscious level, that we have the clearest indication of linguistic leveling. 8.4.1 Lexemes and morphemes All morphemes and the majority of lexical expressions used in Mus¡id wa-Mus¡idih episodes are San¡ani, even though many are also attested in surrounding dialects. 8.4.1.1 SA morphemes and lexemes SA morphemes used in the episodes include all pronouns and pronominal affixes and other prefixes and suffixes: –i∫ (2 f.s. obj/poss), ∞a- (future 1 pl./

175

JANET C. E. WATSON

2/3 person), ∫A- (future 1 sing.), -ih (3 m.s.) bi- (habitual) bayn- (habitual 1 sg.). Turn-taking and interruption particles are either San¡ani or pan-Yemeni, and include iskah “go on!”, yahChAh, hAh, hayyA, innA, bigCnA, yA s∞am and islam. Presentative particles are used frequently in all three episodes and include fa-hEOA, gaOOA, hC OCk. San¡ani lexical expressions include all the question words – mA “what”, kayf “how”, ayn “where”, lilmA / lilmih “why”, kam “how many/much”, man “who”, mih “isn’t it?” (as tag question); prepositions and prepositional phrases such as sA∞ “like”, bih “there is”, mA bi∫ “there is not”, min jayz “similar to”; particles, adverbials and pronominal forms such as allaOC [relative pronoun], Oal}Cn “now”, tijAh-mA “before [conj.]”, sA∞-mA “like [conj.]”, zid “again”, ∞Ad “still”, law-mA “when [conj.]”; verbs, verb phrases and verbal nouns such as hiniyat li∫ “lucky you f.s.!”, 1uhur li∫ “don’t you f.s. see?”, absar – yibsir “to see”, ista1A – yista1C “to see”, sAr – yisCr “to go”, nafa∞ – yinfa∞ “to go shopping for s.o.”, sayhalih “lack of interest”, daywa∞ah “ditto”, mirA∞ah “waiting”; and nouns such as za∞kamih [name], jahhAl “children”, bahCmih “she-donkey”, ∫igrAn “chickens”, fa∞ilih “problem”. 8.4.1.2 Pan-Yemeni/pan-Arabic expressions Sprinkled within the San¡ani expressions are several pan-Arabic expressions. These include Yufl “child”, which is always used in place of the San¡ani singular jAhil, and whose plural aYfAl is often used rather than either jahhAl or ∞ayyAl. ta}lam “she dreams” is used in place of SA tarAyA§, dajAj “chicken” in place of SA ∫ugrC, filEs “money” always used in place of SA zalaY, miθl “like” alongside SA sA∞, amAm “in front of ” alongside SA tijAh, jamCl “good, beautiful” in place of SA bAhir (in the expression kalAm jamCl ), ∞indamA “when” alongside SA law-mA, a}yAnan “sometimes” in place of SA zArat}Cn, al-fatAh “girl” in summaries in place of either bint or buniyyih, maYbax “kitchen” in place of SA daymih. The demonstrative set hAOA, hAOC and hAOawlA in the episodes is SA, but is considered to be more pan-Yemeni and less SA than the non-ha forms, OayyA, tayyih, Oawlayyih (cf. Watson 1993; Watson and al-¡Amri 2000). 8.4.1.3 Non- San∞ani dialect expressions Expressions from Yemeni dialects other than San¡ani are occasionally also used. In these selected texts, they include sara} – yisra} “to go” in place of sAr – yisCr, and ∫all – yi∫ill “to take” in place of SA bazz – yibuzz. The analysis of these texts suggested that pointers towards a potential emerging urban vernacular were not to be found in the morphology or the vocabulary. Morphemes smaller than a single syntactic word are almost invariably San¡ani, although they may be also found in the surrounding

176

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

dialects. The majority of lexical items are San¡ani, and the exceptions, such as Yufl “child”, ∞indamA “when [conj.]”, a}yAnan “sometimes” and filEs “money”, are more pan-Arabic than pan-Yemeni. I then examined texts from two of my San¡ani speakers – AS, who was born and raised in alBawniyih in Bcr al-¡Azab, the former Ottoman diplomatic quarter which lies next to al-QA∞, and was at the time of recording around 34 years old; W, a girl then in her early teenage years, who was born and raised in BAb ∫u∞Eb on the northern edge of the Old City: her mother comes from a wellestablished San¡ani family and her father was born and raised in San¡a, though his family originates from a village near Kawkaban, c. 90 minutes by car to the northwest of the city. I took one text from each of the speakers, the first which is accessible from the Heidelberg Semitic sound archive (www.semarch.uni-hd.de, Sbahtu 15); the second text is text 6 from Watson and al-¡Amri (2000), extracted pausal elements of which are also accessible from the Semitic sound archive. Analysis shows that although both these San¡ani texts contain non-SA lexical elements, and in the case of W, considerably more non-SA lexical elements than the selected Mus¡id wa-Mus¡idih texts,7 the recordings had a San¡ani element to them which was absent in the latter. This lay principally in the accent, and was particularly noticeable in the nature and degree of glottalization in pre-pause position.

8.5 Glottalization in pause8 Yemeni speakers recognize glottalization as a feature of San¡ani Arabic, despite the fact that it is also attested in many other Yemeni dialects within the Central Plateau (cf. Behnstedt 1985: 58) and, to a far lesser degree, in dialects spoken in the western mountain range (cf. Watson 1989). The difference between SA glottalization and that in surrounding dialects, however, lies in the type of glottalization and the effect glottalization has on pre-pause consonants. Jastrow (1984) in his paper on aspects of the phonetics and phonology of San¡ani notes that final sonorants (l, r, n, m) are often elided in pause in the position CVCSo. Diachronically, this may have begun as simultaneous glottal stop and sonorant pronunciation followed by deletion of the sonorant, as in some of the Sino-Tibetan languages (Michaud 2004). Thus, in the word guYn “cotton”, although closure between the tongue tip and the upper dental area for the final sonorant may take place, the sound is unreleased and visibly absent. W’s pronunciation of firn “oven” is realized as [fir], with typical word-final trill realization of /r/ (four separate taps are visible on the spectrogram) and loss of a visible final sonorant.

177

JANET C. E. WATSON

Spectrogram 8.1 guYn “cotton”

Spectrogram 8.2 firn “oven”

178

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

Final sonorants are pre-glottalized and remain suppressed following long vowels, as in AS’s |CgA(n) “windows”,9 tamA(m) “okay” and iθ nay(n) “two”. In some cases, final voiced fricatives are also visibly absent in pause, as in AS’s ’a}wA(1) “troughs”. Before final consonants, particularly following long vowels, a glottal closure is realized – indicated by glottalized striations followed by complete lack of energy on the spectrogram. This is exemplified in the example of bisbA§s “chili” (8.3) below. Spectrogram 8.3 bisbAs “chili”

Glottal closure occurs not only when the final syllable is stressed, but also when the final syllable follows a stressed syllable, as in W’s marak§ “broth”, za}Awuk§ “spicy dip” and mubargaY§ “lumpy”.

179

JANET C. E. WATSON

Spectrogram 8.4 mubargaY “lumpy”

Spectrogram 8.5 za}Awug “spicy dip”

180

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

Examples of glottal closure in a post-stressed syllable from AS include wsAyit§ “cushions” and barik§ “pool”. And while “weak” consonants – the sonorants and voiced fricatives – are subject to further weakening and, usually, elision in pause, “strong” consonants – the plosives and the affricate /j/ – are released on a glottal airstream mechanism as glottalic pressure consonants or ejectives. This is indicated by the visible spike on the waveform above the spectrogram (Spectrograms (8.4), (8.5) and (8.6) ). Ejective realization is most salient in the case of the velar plosives /g, k/, (see W’s dagC§k§ “flour” and za}Awuk§ “spicy dip”), but also affects the labial plosive /b/ (W’s al-}alC§p§ “the milk”) the affricate /j/ (AS’s sfint§∫ “sponge”) and, to a lesser extent, the coronal plosives /y, t, d/ (AS’s ∞ugEt§ “arches” and musajjilA§t§ “recorders”; W’s mubargaY§ “lumpy” (8.4) ). This is to be predicted, since universally velar ejectives are most favored, followed by bilabials and then coronals (Henton et al. 1992: 91). Compare here the spectrogram of dagC§k§ “flour” (8.6) with that of mubargaY§ “lumpy” (8.4) above. Spectrogram 8.6 dagCg “flour”

Thus, concomitant with glottalization, all voiced consonantal phonemes are devoiced in pause; final sonorants are devoiced and often elided, particularly but not exclusively in post-consonantal position; voiced fricatives may also be elided; final sibilants are preceded by glottal closure, but not suppressed; final plosives and /j/ are preceded by glottal closure and released on 181

JANET C. E. WATSON

a glottalic airstream mechanism as ejectives, particularly saliently when the final syllable is stressed. Glottalization in pause also occurs in the speech of Mus¡id and Mus¡idih. However, a comparison reveals interesting differences. The spectrograms and waveforms for AS and W show a much sharper, less gradual glottalization (represented as a marked increase in distance between the striations on the spectrogram), often accompanied by glottal closure (represented by a complete lack of energy). In the glottalization of final long vowels, Mus¡id10 and Mus¡idih often release the vowel with aspiration, as in: xAlC(h) “empty”, in comparison with the increasingly tense realization of the pre-pausal vowel ending in glottal closure in AS’s madAkC “support cushions”, W’s zabAdC “yoghurt” (8.7) and b-yi}mA “it heats up”. Spectrogram 8.7 zabAdC “yoghurt”

182

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

Final sonorants are clearly articulated in the case of both ARM and HM (as in M’s γa∫∫A∫ Cn “cheats” (8.8) ), frequently voiced (as in Mus¡idih’s11 xuddAm “servants”) and where they are devoiced they are often released with aspiration, as in Ma’s zamAnh “time” and M’s al-’ummh “the mother”. Consider here the spectrogram of M’s γa∫∫A∫ Cn “cheats”. Spectrogram 8.8 γa∫∫A∫Cn “cheats”

In the case of final plosives and the affricate, no complete glottal closure occurs in the speech of M and Ma – energy continues until pause – hence pre-pausal plosives and the affricate are released on a pulmonic and not a glottalic airstream mechanism. In the case of underlyingly voiced plosives, full devoicing rarely occurs in pause, as in Ma’s rAgid “sleeping” (8.9). When fully devoiced in the case of these speakers, release tends to be accompanied by aspiration, as in Ma’s jadCt h ‘new’ and M’s zallaj h ‘sent away’.

183

JANET C. E. WATSON

Spectrogram 8.9 rAgid “sleeping”

Glottalization does take place in the speech of M and Ma, however, it does not take the form of complete glottal closure, nor does it have the same degree of glottal constriction as in the case of the two established SA speakers. Final sonorants are, in contrast to SA, audible and often released with aspiration. Where devoicing of underlyingly voiced obstruents takes place and in the case of voiceless obstruents, pause is often marked by aspiration; due to lack of complete glottal closure, pause is never marked in the case of these speakers by an ejective realization of the pre-pause consonant.

8.6 Conclusion The morphology used in the Mus¡id wa-Mus¡idih episodes is, with the exception of the final didactic summary, San¡ani. Lexically there is a mixture between SA and pan-Yemeni/pan-Arabic with more emphasis on SA. Apart from where register is shifted in order to summarize the argument, there is no evidence of Classical vocabulary. In Mus¡id wa-Mus¡idih, it is, I believe, the mixture between a San¡ani linguistic system – morphology, syntax, lexis, phonemes – and a not-entirely San¡ani accent which arouses such different assessments from native speaker listeners. I suggest that it may not be the deliberate, conscious use of language we should be considering in establishing where and whether a koiné is emerging 184

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

in San¡a, but rather the unconscious or subconscious accent features; features which are not entirely present or absent, but graduated. The future will see whether glottalization in San¡ani progresses, as it appears to have done in some Sino-Tibetan languages, to deletion of final sonorants and non-sibilant fricatives or their replacement by glottal closure, or whether, as a result of dialect contact, glottalization as glottal closure and/or glottal constriction weakens, due to its salient auditory quality, to the type of laryngealization and aspiration attested in incoming dialects.

Appendix Extract from Mus¡id wa-Mus¡idih texts (published in Arabic script with the English translations in Muyahhar and Watson 2002). / indicates pause, including the end of an intonation unit. ZAWfJ AL-BADAL

ma: ma ™ams / amanih innanc 2u]ukt / min ibn axek / law-ma ∫ibi¡t / ya mus¡id / m: hinyatli∫ ]cn ¡ad bi-tu2]ukc / ya mus¡idih / ma: ma bn axek ]aram innahe yibiddc a2-2u]uk / ¡indama gallak tisallifih mcyat alf riyal / yigem yiyallug maratih / inna aw gad ay-yalag bi-files / ya mus¡id / m: antc ma fihimtc∫ al-xabar sawa / ibn axc yi∫tc minnc asallifih mcyat alf riyal / gawlatih / yuwaffc ]agg al-¡iris / ma: inna kayf yuwaffc ]agg al-¡iris / wa-¡ade oa tzawwaj gabla sittih a∫hir / kayf ¡ade haoa l-xabar / m: al-mu∫kilih / ya ]ajjih mus¡idih / hcya zawaj a∫-∫iγar / alla2c bi-nsammc / zawaj al-badal / ma: ma haoa ∫-∫c gade wa2u] / w-ana nabbayt ¡al ibn axek ¡indama yalab minnc ™asra] absir al-bunayyih / gablama yuxyubha / gultalih ebih tizawwuj uxtak badal / wa-titzawwaj badal / zawj al-badal ma fch illa l-mawaji¡ wa-2-2ulm / wa-kullih ¡ala l-fatah al-maskcnih al-maglebih ¡ala ™amraha / m: bigcna / gad tamm ma tamm / ibn axc zawwaj uxtih / bi-walad al-]ajj ]usayn mu]ammad / walad al-]ajj ]sayn mu]ammad zawwaj uxtih / walad axc / ma: ana dariyih / bi-™inn az-zawaj wuga¡ badal ras bi-ras / la ra2aw al-banat wa-la ]atta fi¡ilelahin mihr ¡ala ma ™amarna d-dcn / al-magxed kulla wa]idih / xarajat min bayt ahlaha fc laylat ¡irisha / wa-hc wajc¡at al-galb / makserat al-xayur / la diriyat kayf ¡ade haoa / ]a22aha / wala ma bi-txabbc laha l-iyyam / m: uxburc / gaooa l-iyyam baddit al-mixabba / bint axc / nafarat min walad al-]ajj ]usayn mu]ammad / min ]al ma gurub yistajlcha / 2uhulli∫ ya mus¡idih / 185

JANET C. E. WATSON

ma: gadana dariyih / wa-bigc haoa n-nafer malazim laha yel as-sitti l-a∫hur allaoc ¡a∫atha ma¡a zawjaha / wa-fi l-laxcr i¡lanatih sarc]ah / bi-™innahc la ti∫tch wa-la ti]innih / wa-laflafat adwanha wa-riji¡atlaha bayt ahlaha / /. . ./

Notes 1 I am very grateful to Yahya Asiri for help with producing the spectrograms. Thanks are also due to the British Academy, who awarded me a small research grant 2000–2002, during which time I carried out fieldwork for this paper. Thanks also to Barry Heselwood for very pertinent comments on the phonetics, and to Shelagh Weir, for permission to reproduce the map of Yemen. 2 According to Dresch (2000: 139), 90 percent of the population of the North was rural as late as 1979. 3 This figure is according to a consular report. The drop in population almost inevitably resulted from the effects of an atrocious famine and economic troubles (cf. Dresch 1989: 222). 4 The Old City of San¡a is still populated mainly by original San¡anis. It has an estimated population today of 55,000 (2002 figure, British Council, San¡a source). 5 By local tradition, this latter event presages the onset of the Last Days. 6 In the Semitic Sound Archive identified as social024, social033, social036, social048 and social064. 7 Including classicisms such as al-§An “now” in place of Oal}Cn, natakallam “we talk” in place of nit}AkA, and pan-Arabisms such as ba∞dayn “then; later” in place of ba¡da, ∞a∫An “because” in place of ∞asibb / ∞alAsibb, fCh “there is” in place of bih, and mA fC∫ “there is not” in place of mA bi∫. 8 Broad phonetic transcription is used in the spectrograms below. r indicates both trill and tap, and vowels are transcribed phonemically. 9 Owing to limited space only a small selection of the twenty-five spectrograms analysed have been included here. 10 Henceforth M. 11 Henceforth Ma.

References Banerji, A. and McLeish, C. (2002) Governance and the Investment Climate in Yemen, Middle East and North Africa, Working Papers Series, No. 38, Washington, DC: World Bank. Behnstedt, P. (1985) Die nordjemenitischen Dialekte. Teil I. Atlas, Wiesbaden: Harrassowitz. Bury, G. Wyman (1998) Arabia Infelix: Or the Turks in Yamen, Reading: Garnet. CCA (Common Country Assessment) (2001) Yemen, San¡a: CCA. CPO (Central Planning Organization) (1975) Housing and Population Census, February 1975, San¡a: CPO. CPO (Central Planning Organization) (1983) Statistical Year Book 1983, San¡a: CPO. CPO (Central Planning Organization) (1989) Statistical Year Book 1988, San¡a: CPO. CSO (Central Statistical Organization) (1992) National Population Strategy 1990– 2000 and Population Action Plan, (March), Ministry of Planning and Development and Central Statistical Organization, San¡a: CSO.

186

LINGUISTIC LEVELING IN SAN¡ANI ARABIC

CSO (Central Statistical Organization) (1994) Yemen: Demographic and Maternal and Child Health Survey 1991/1992, (March), Central Statistical Organization, San¡a: CSO. CSO (Central Statistical Organization) (1998) Yemen: Demographic and Maternal and Child Health Survey 1997, (November), Central Statistical Organization, San¡a: CSO. Dresch, P. (1989) Tribes, Government, and History in Yemen, Oxford: Oxford University Press. Dresch, P. (2000) A History of Modern Yemen, Cambridge: Cambridge University Press. Henton, C., Ladefoged, P. and Maddieson, I. (1992) “Stops in the world’s languages”, Phonetica 49: 65–101. IMF (International Monetary Fund) (2001) Republic of Yemen: Selected Issues, IMF Country Report, (April), International Monetary Fund, Washington, DC: IMF. Jastrow, O. (1984) “Zur Phonologie und Phonetik des San¡anischen”, in H. Kopp and G. Schweizer (eds) Enwicklungsprozesse in der Arabischen Republik Jemen, Wiesbaden: Dr Ludwig Reichert. Michaud, A. (2004) “Final consonants and glottalization: new perspectives from Hanoi Vietnamese”, Phonetica 61: 119–146. Mission (1980) Yemen Arab Republic: Report of Mission on Needs Assessment for Population Assistence, Report No. 32, United Nations Fund for Population Activities, (September), New York: Mission. Mu™ayyad al-¡Azm, N. (1986) Ri}lah fi al-∞arabiyyah al-sa∞Cdah, Beirut: Dar al-Tanwcr. Muyahhar, A. R. and Watson, J. C. E. (2002) Social Issues in Popular Yemeni Culture, San¡a: al-Saba]c Press. Naïm-Sanbar, S. (1994) “Contribution à l’étude de l’accent yéménite: le parler des femmes de l’ancienne generation”, ZAS 27: 67–89. Playfair, R. L. (1970) A History of Arabia Felix or Yemen, Farnborough: Gregg International. Scott, H. (1942) In the High Yemen, London: John Murray. Steffan, H. (1978) “Final report of the Swiss Airphoto Interpretation Team”, Airphoto Interpretation Project of the Swiss Technical Co-operation Service, Berne, Department of Geography, University of Zürich, for Central Planning Organization: San¡a. Watson, J. C. E. (1989) “Yemeni vowels in pausal position”, in BRISMES: Proceedings of the 1989 International Conference on Europe and the Middle East, Durham, pp. 351–361. Watson, J. C. E. (1993) A Syntax of San∞AnC Arabic, Wiesbaden: Harrassowitz. Watson, J. C. E. with al-¡Amri, A. S. (2000) WaXf San∞A: Texts in San∞AnC Arabic, Wiesbaden: Harrassowitz. Weir, S. (2007) A Tribal Order: Politics and Law in the Mountains of Yemen, Austin, TX: University of Texas Press.

187

HANADI ISMAIL

9 THE URBAN AND SUBURBAN MODES 1 Patterns of linguistic variation and change in Damascus Hanadi Ismail

9.1 Introduction Damascus is the oldest and largest city in the Levant. Being the main commercial and industrial center in the south of Syria, the city has always attracted workers and migrants from smaller satellite towns as well as other cities in Syria. The metropolis has considerably expanded during the last thirty years. This expansion has proceeded in two directions. The most widely followed trend is the extension and the development of the nearby villages and towns, which has particularly attracted the city’s newcomers and those of less economic ability. The other form of expansion has been the building of new residential areas and suburbs, mainly on the western fringes of the city. The aim of the city planners was to develop less polluted residential areas, and to preserve the Ghuta oasis and the Old City to the east and south of Damascus. The emergence of the metropolitan suburbia is of most interest here since it has brought together communities with a differing lifestyle. The research on which the current chapter is based investigates linguistic variation and change in urban neighborhoods, and applies the concept of life-mode (Hojrup 2003) as a method of analyzing the socio-economic structure in and across neighborhoods. Within this context, neighborhoods are perceived as habitats of certain subcultural and social practices through modes of production, irrespective of any class assumptions. The Philadelphia study (Labov 1989) and Belfast study (Milroy 1980), have substantially improved our understanding of the mechanisms of change in local settings, particularly with respect to the question of whether the “territorial” factor exists independently of other social categories such as social class, age and gender. In this study, I explore the issue of localness in the metropolis 188

THE URBAN AND SUBURBAN MODES

through the mode of production within a socio-economic framework, and how localness, if it does, correlates with linguistic and other social factors. The choice of two Damascene neighborhoods, Shaghoor and Dummar, was based on their socio-geographic character. Shaghoor is an inner-city quarter, and is part of the Old City. It has its distinctive linguistic and social character, which is locally regarded as the archetype of the traditional Damascene dialect and lifestyle. Dummar, on the other hand, is a newlyconstructed suburb, and a residence to the city’s intelligentsia and professionals who originally come from different residential backgrounds in Damascus. The making of these suburbanites is closely integrated with the making of Dummar as a suburban municipality. The issues addressed in this chapter consider the questions of how linguistic change operates in the neighborhood as a subculture and as a mode of production, whether different neighborhoods in the city induce different patterns of change, as well as the profile of the leaders of change in each neighborhood. I begin by giving a brief historical background for Damascus, outlining the language situation in the city since its development as a major urban center (Section 9.2). Section 9.3 gives a description of the speech sample, and Section 9.4 describes the neighborhoods and reviews the historical development of these two urban and suburban centers, and introduces the notion of life-modes as they are applied in the study. The linguistic variables and results are discussed in Sections 9.5 and 9.6.

9.2 Historical and linguistic background Damascus is reported to be one of the oldest continuously inhabited cities in the world (Al-Qayyem 1991). It is not certain when the first urban settlement took place in Damascus; however, evidence from Egyptian records shows that Damascus was conquered in the fifteenth century bc by the Pharaohs. Damascus, or Dar Misiq “the well-watered land” as it was known, was the capital city of the Aramean State or Aram Damascus in 1100 bc. It continued to be a capital city for the Syrian vilayet of the Persians in 538 bc. The city was taken over by Alexander the Great in 333 bc, and then by the Romans, who made it a Roman metropolis and an important center of Roman culture. Its strategic location as a caravan city on trade routes reinforced the city’s status as an important commercial and urban center during the Roman era. During the Umayyad rule (661–750 ad), Damascus gained a significant political status when it became the capital city for a state which extended from Spain to India. In the mid twelfth century, Damascus was also the capital city of the Ayubites (1174–1259 ad). It later became a provincial capital of the Mamluke Empire, before it fell under Ottoman rule in 1517 ad, and remained so for the next 400 years. At the end of the First World War, Damascus was taken by the British and the Allied Arab forces, led by Sharif Hussein bin Ali; Arab rule, under King Faisal I, was soon 189

HANADI ISMAIL

ended by the French mandate over Syria in 1920. In 1946, the French troops fully withdrew and Damascus continued to be the capital city of independent Syria. Due to the heavy influx of economic migrants, the population of the capital has significantly increased. In 1917, the population of the city was around 240,000 people, while in 2002 the city’s population reached 1,516,000 according to the Central Bureau of Statistics (and 3.5 millions including Greater Damascus). There is evidence that Arabic was spoken along with other NorthWestern Semitic languages in some parts of Syria before the Islamic era in the seventh century. During the Hellenistic period, Greek was the language of administration in Syria, before it was replaced by Arabic at the end of the first century of Hiòra (Versteegh 1997: 94). Syriac was the spoken language then, and continued to be spoken by Christians until the eighth century, while it remained as a literary language until the fourteenth century. Versteegh (1997: 152) maintains that the new varieties of Arabic which emerged as a result of Arabicization in Damascus were urban varieties. Two factors lend support to this hypothesis. First, Arabicization spread from the major old Hellenistic cities, which were then Damascus and Aleppo. It was in these cities that the Arab conquerors first settled (Versteegh 1997: 152). Second, Arabic was already spoken in these cities due to the pre-Islamic Bedouin migrations from the Syrian Desert. This situation of dialect contact brought about a faster rate of innovation. The urban dialects spoken in Damascus and Aleppo were linguistically distinct from the dialects of the Syrian Desert and Mesopotamia. The dialect of Damascus falls into the dialect group called Central Syrian dialects, which also includes the dialect of the Druze. Damascus Arabic has been studied for over a century, and is one of the best described Arabic dialects (Lentin 2006).

9.3 The research sample The research sample is taken from two socio-geographically defined communities; inner-city Shaghoor, and suburban Dummar (see Map 9.1). The data include a sample of 56 Damascene speakers (29 male and 27 female). Three independent variables are considered in two neighborhoods: age; gender, and life-mode. The speakers were divided into three age groups: old (51–75), middle (31–50), and young (18–30). The life-mode of speakers is based on the place of residence as a habitat of a particular subculture and social class (Hojrup 2003). Two life-modes are distinguished here, namely the life-mode of “self-employed” represented by the Shaghoor community, and the life-mode of “professionals” represented by Dummar suburbanites. Table 9.1 shows the distribution of speakers by age, gender, and life-mode. Interviews were recorded on Sony MZ-R700 MDs. Individual interviews lasted for an average of one hour, whereas group interviews were recorded for around 200 minutes, and included no more than four speakers in one 190

THE URBAN AND SUBURBAN MODES

Map 9.1 Damascus Table 9.1 Distribution of the speakers by age, gender, and life-mode Self-employed

Young Middle Old

Professionals

Male

Female

Male

Female

5 4 5

5 5 5

5 5 5

4 4 4

191

HANADI ISMAIL

recording session. I was introduced to the Shaghoor community through the local nurse who was a key figure in the area. My network of contacts subsequently expanded through being introduced to new informants as a “friend of a friend” (cf. Milroy 1987). A similar approach was followed in Dummar.

9.4 Neighborhoods and life-modes 9.4.1 Neighborhoods 9.4.1.1 Shaghoor Shaghoor is one of the oldest recorded neighborhoods of Damascus. Part of the neighborhood was known as the Jewish quarter, as it had been a residence for the largest Jewish community in the city. It is also a residence for many of the prominent Shiite families in Damascus. The main road to the neighborhood, Al-Ameen, named after a prominent Shiite figure, accommodates the marketplace and the main shops. During the French mandate of Syria, Shaghoor was one of the important localities of national resistance. Many of the locals were involved in political activities and in the development of Syria’s national political thought during the 1930s. Shaghoor was the home of many prominent intellectual and political figures, including the famous poet Nizar Qabbani and Youssef Al-Azmah, the Syrian defense minister (under King Faisal I) in 1918 and a national figure who led the Syrian army against the invading French troops in 1920 in Maysaloon. Shaghoor, therefore, is associated in the imagination of Damascenes with profound local values of patriotic, social, and linguistic heritage. The architecture of the Old City manifests to a large extent the lifestyle and the social practices of the local dwellers. Known as the archetype of the vintage Damascene style, Shaghoor is widely reputed as the prototype of the old life and building style of Damascus. With the exception of the block buildings along Al-Ameen Street, all of the houses in Shaghoor are of the type known locally as el-bBt el-∞arabi, “the Arabic house”. Each of these traditional houses consists of a courtyard, or bA}t el-bBt, and a fountain. The courtyard, which is overlooked by upstairs balconies and room windows, is the socializing space for the family. The number of rooms in a house is not indicative of the economic status of the family, as the house is designed to accommodate the extended family. Houses are adjacent to each other, and extend along the quarter’s narrow and sinuous alleys, which open up to the neighborhood’s gate. Gates were traditionally used for safety. The speakers of Shaghoor were selected from the Jabri Avenue.2 Some of the avenue families have lived in the neighborhood for over a hundred years, and a good part of them were blood related. Sharing the house with the family, as well as living directly adjacent to neighbors, has traditionally contributed to the maintenance of strong local ties. This pattern is changing, however, as the extended family model has become less favored, according 192

THE URBAN AND SUBURBAN MODES

to field studies. Only 24 percent of women in Damascus who married in the late 1960s have shared the family house with their in-laws (Prothro and Diab 1974, cited in Barakat 2001: 194). The choice of sharing the family house on the part of sons is primarily driven by the economic status of the family. This change has been concomitant with macro socio-economic changes, such as the spread of education, the wider provision of services, and the constant growth of the middle class. Barakat (2001: 178) maintains that the institution’s role, whether private or state-owned, in providing for the individual has replaced that of the extended family. This, in turn, has significantly contributed to reinforcing the concepts of individual freedom and independence among the youth. This urban family structure then has been shaped by the “nuclear model”; however, the family, as a unit, remains the main economic and social resource for the urbanites in Damascus, and probably in most Arab societies (Hopkins 2001). 9.4.1.2 Dummar The emergence of suburbia as a new option of residence is a recent phenomenon in the socio-geographic fabric of Damascus. Dummar suburb was the first project of such a kind. The suburb was built by the Damascus Association of the Professional Guilds. Construction began in the 1970s, but the suburb was only occupied in the mid 1980s. Dummar Project was the name given to the suburb so as to differentiate it from the adjacent town of Dummar. The suburb is locally called “the Project” or el-ma∫rD∞. Dummar lies 5 km to the west of Damascus, under the council of Damascus City administratively. Although planned to house professionals, unlike other emerging suburbs, it has attracted the intelligentsia of Damascene society. In addition to professionals, the residents include TV actors and producers, prominent writers, artists, and technocrats. Most of these prominent figures have their jobs in the city, to which they commute daily. Three types of buildings are distinguished in Dummar: the towers, the three-story buildings, and the villas. The research sample was taken mainly from the towers and the buildings. The character the suburb has developed over the past two decades is not unrelated to the history of its development. Soon after construction was launched on site, a top-level decision was taken to halt the project. A highranking political figure demanded control of the project, which was resisted by the executives of the Guilds as well as by the creators of the project. The Dummar Project eventually triggered a wide political dispute, which entailed a parliamentary as well as presidential intervention. During the process, some of the applicants sold their shares in despair, while for those who carried on, the issue became one of civil rights. This polarized tension has in fact been conducive to the emergence of a class of liberals, and to the fostering of a distinctive character in this particular suburb. 193

HANADI ISMAIL

The events which contributed to defining homogeneity in Dummar are the core elements of what specifies a suburb as a social structure. Greer (1960) explains that it is possible for a metropolitan suburbia to develop without an independent government. This is sustained by the residents’ participation in local formal organizations, which are segmental in their membership and purposes (Greer 1960: 7). This includes all community oriented organizations including improvement associations as well as service clubs and children’s organizations. These organizations, which perform the function of a local political entity, translate the residents’ interdependence into social action. Dummar, indeed, is a case in point. The suburb is maintained by a locally elected committee, which works voluntarily on a highly structured level. Amenities, such as central heating, electricity, and water, are locally administered and locally controlled. In addition, the committee arranges employment opportunities for the residents within these organizations, such as in the central services or through letting premises in the local shopping centers at a nominal cost. The suburb has a school, many kindergartens, sports clubs, and recreational facilities which are mostly run by locals. Today, the suburb has become one of the main leisure attractions for city dwellers. 9.4.2 Life-modes The notion of life-mode is based on Hojrup’s (2003) analysis of class and social differences in Copenhagen. The analysis uses the mode of production as a structure of social relations to describe the class, culture practices, and ideology specifics of a group or subculture. Therefore, each life-mode has its class-defining practices as a mode of production, irrespective of any class category assumptions. The way a life-mode is perceived by the conceptual universe of other life-modes is one of the essential elements related to the understanding of the concept of a life-mode (Hojrup 2003: 29). Two lifemodes are applied in this study, namely the self-employed life-mode and the professional life-mode. 9.4.2.1 The self-employed A distinctive feature of the self-employed life-mode is the concept of “activity”. Individuals in this life-mode do not differentiate between the concepts of “work time” and “free time”. Both concepts are meshed into one activity, namely the daily pursuit. This pursuit is determined by and oriented to the market, where commodities are exchanged. The committed pursuit of daily work is a means and an end, and is perceived as an independent activity rather than a dependent work. The residents of Shaghoor work locally as street vendors or local shop owners. A few others work as taxi drivers and factory workers. Working hours for local street vendors are not restricted to a time schedule. A typical 194

THE URBAN AND SUBURBAN MODES

day for Salem, a 45 year old local fruiterer, is spent by the roadside chatting to other street vendors. Although his house faces his roadside market venue, Salem has his meal sent from home. His day in the market finishes by nine or ten at night, when he joins the family for tea. Salem’s social life is blended within his daily pursuit, through interacting with neighboring sellers. The only additional activities in Salem’s life are the cost-free ones which do not interfere with his daily pursuit, such as evening family gatherings, which may include a card game. Older men, who are in the age of retirement, do taxi driving in the morning and afternoon as their daily pursuit. Ownership is a crucial issue in the community, as all activities are utilized to maintain and extend this concept. Women, on the other hand, are responsible for house issues, and spend their time watching TV and chatting to their neighbors, who, in the majority of cases, are relatives or part of the extended family. Although none of the women I interviewed held a job at the time of the research, the majority of them recounted previous home-based jobs, such as seamstresses, saleswomen, or a home-assistant to a business run by the husband. The job is conceptualized here as part of a family enterprise, rather than a self-development process. As a socio-economic parameter, the life-mode concept can be seen as a more appropriate tool to analyze the urban social structure in Damascus than social class. Initially, the research methodology was designed to factor the socio-economic status as a social variable. I had collected data on house density, education, and occupation of the speakers, the composite of which yielded different measures for social class categories. The drawback of the socio-economic tool was that some speakers had to be pigeonholed, according to the composite socio-economic indicator, in categories which did not reflect their social reality. This is not uncommon in studies which employ composite socio-economic indicators. Some of the Shaghoori families who had dwelt in the neighborhood for over a century were among the leading merchants of Damascus. But their lifestyle, practices, and ideology did not reflect their affluent economic status. For instance, in the case of Sameer and his family, working in the shops they own is a means and an end. The family’s approach to work, including the absence of free time as a leisure activity, is not different from the attitude of the local fruiterer, despite the marked difference in their economic status. Deeb, Sameer’s son, who is 43 years old, told me how he was forced to leave school at the age of 10 in order to work all day in the shop. Work, and, in this case, wealth, are not a means for bettering social status; rather work is an end. 9.4.2.2 The professional Members of the professional life-mode do not acknowledge independence in terms of ownership of an enterprise but through the extent of control and 195

HANADI ISMAIL

creativity at work (Hojrup 2003: 38). Ownership of enterprise does not matter in this mode because it allows no mobility and no development along the career hierarchy. Income is used to develop the status and qualification of the professional. Not only is money used for qualification development but also for other socially promoting activities, including dressing, social gatherings, and leisure or sport activities. These activities are sustained by the wife who assumes the role of backing her husband’s career. Similarly, the house, clothing, and furniture are kept on a par with the social status demands. In this life-mode, the husband–wife relationship is interdependent. On one level, he depends on her to maintain and develop the social supplements necessary to keep up with the status and the lifestyle. She, on the other hand, depends on his income and status as sources of her involvement in fine art, culture issues, clubs, and other social practices which give meaning to her life (Hojrup 2003: 40). Time and free time are valuable resources for the professional life-mode to develop skills and qualifications. Dummar dwellers are mostly professionals, university professors, school teachers, and senior civil servants. Others are TV actors, producers, and broadcasters. The wives of these professionals have either taken an early retirement or are still working as part-time language teachers or translators. The suburb residents identify themselves as part of the §el-musaqqafBn’ “the cultured” community of the city. Although the residents originally come from different areas in Damascus, they share similar and common interests. All of the informants who were interviewed for this study participated in fine arts events, including painting, cinema festivals, sculpture, and literary studies. Activities within the suburb are highly structured and contribute to reinforcing the suburban life style. I was told by men that women are the organizers of the couples’ social activities and networks in the community. These women have founded their own local societies, such as the Reading and Movie Societies, where they meet as a group to discuss a book or a movie over lunch. Other activities concern residential development issues. A 65 year old woman told me how she and her neighbors undertook planting and developing the gardens and roadside trees. Other voluntary activities include the election of local representatives from each building and zone to discuss and resolve issues related to the maintenance of services in the suburb, including the central heating and the water tower of the suburb. This committee is headed by a locally elected chief person. The suburb can, therefore, be described as a spatially-defined microcosm. Broadly speaking, the political orientation in the suburb belongs to what is locally known as “the liberals”, or the civil society advocates, on a par with other intellectuals in Damascus. Adolescents had their own self-founded and run clubs, such as the Drummers and the Chorale clubs.

196

THE URBAN AND SUBURBAN MODES

9.5 The linguistic variables Two phonetic linguistic variables have been chosen to compare the sociolinguistic patterns of these two neighborhoods: (h) and (r). Both occurred frequently in the data. Although variation in the presence of -(h)a/-(h)on has been noted for a long time in the literature, variation in the quality of the trill remains a relatively new phenomenon in the region. Analyzing patterns of variation, and possible change, with regard to these variables will allow us to understand the social meaning of this variation. 9.5.1 Variable (h) 9.5.1.1 Description The variable (h) concerns the presence or absence of [h] in the third person singular feminine suffix /-ha/, and the third person plural, masculine and feminine, suffix /-hon/. Variation in the presence of [h] in the pronominal suffix has been noted in many studies of the dialect, (Bergstraässer 1924; Cowell 1964; Grotzfeld 1965; Lentin 1981). The following lexical items, taken from the data, illustrate this variation: (1) After a consonant /™£b£nha/ “her son”; /]afl£ta/ “her party”; /wa™£tha/ “then or at that time”; /rabbε:ton/ “I brought them up”; /™£lhɔn/ “for them”; /™£lɔn/ “for them”; /j™all£¡hon/ “to kick them out”; /ma¡a/ “with her”. (2) After a vowel (a) High back vowel: /™axadu:ha/ “they took it”; /™abuwa/; “her father”; /™abu:ha/ “her father”; /jiγ£xbuwa/ “to force her”. (b) High front vowel: /ya¡mi:hon/ “feed them”; /da∫∫rijon/ “leave them”; /fi:ha/ “in it”; /fijja/ “in it”. As can be noticed from the examples in (1) above, [h] is variably present after consonants. The variant [h] alternates with the glides [w] and [ j] after vowels. [w] is inserted after the high back vowel [u], and [ j] occurs after the high front vowel [i] (examples under 2). After the long low vowel /a/, [h] is obligatorily present, as in (3) below:3 (3) After long low vowel /a/ /fata]na:ha/ “we opened it”; /¡m£lna:ha/ “we did it”; /hawa:ha/ “its air”; /awwazna:hɔn/ “we arranged for them to marry”.

197

HANADI ISMAIL

9.5.1.2 Previous studies on (h) Previous studies, some dating back to 1924, take note of this variation and offer accounts which shed some light on this variability. On the whole, the presence or absence of [h] in the suffix has been predominantly attributed to the loss of [h] in the suffix as a result of weakening or assimilation by some speakers. Bergsträsser (1924) refers to the absence of [h] in the suffixes as a result of assimilation or weakening, particularly after consonants. However, the transcribed speech sample included in Bergsträsser’s appendix shows a categorical use of h-full pattern, and no background information about the speakers is given. In addition, the status of [h] after a vowel is not discussed. Cowell (1964: 541) points out that the loss of [h] occurs after consonants and sometimes, especially in Lebanon, after long vowels. However, Cowell notes elsewhere that: Commonly in Lebanese pronunciation and to some extent elsewhere, a short vowel a, o, or u (in the suffixes -a “her, it”, and -on or -un “them, their”) . . . may follow a long accented vowel: bt¢ʔra:on (or bt¢ʔra:un) “she reads them” (for bt¢ʔra:hon) . . . (Cowell 1964: 22) The very description of the suffix above as [a] and [on] in Lebanese or elsewhere gives rise to the question of whether the suffix is /-ha, -hon/ or /-a, -on/. In fact, Cowell’s description indicates the legitimacy of /-a, -on/ as pronominal suffixes, besides /-ha, -hon/. Other studies, Blanc (1953), Grotzfeld (1965), Cantineau (1960), and Ambros (1977) note that [h] is reduced to zero in some dialects as a result of assimilation or weakening. Grotzfeld (1965) and Cantineau (1960) base their argument on the analysis of the third person masculine suffix /-o/. According to Cantineau (1960: 120), the underlying form of the third person masculine suffix /-o/, or /-u/ in some dialects, is /-ahu/. Grotzfeld (1965: 19) adds that as a result of the loss of [h] the diphthong /-au/ changed to /-o/, as in /kta:bahu/ → /kta:bau/ → /kta:bo/. This loss of [h] in the masculine suffix, both Cantineau (1960) and Grotzfeld (1965) maintain, applied also to the feminine and plural suffix. Cantineau (1960: 121) mentions that in the dialects of Horan, in the south of Syria, [h] assimilates to [f, t, s, ∫, ©, x], therefore, /xaru:fha/ becomes /xaru:ffa/ “her sheep” and /fra:xha/ /fra:xxa/ “her chicken”.4 Blanc (1953: 29) mentions that in north Palestinian dialects, [h] is reduced to zero and that this deletion “tends to occur in all positions, but is especially frequent word-finally”. In more recent studies, such as Lentin (1981) and Gralla (2003), reference is also made to the variation in the presence of [h] in the suffix. Lentin (1981: 235) points out that the [h] in the suffix seems to be used less frequently; however it appears sometimes in the context of emphasis, particularly in the 198

THE URBAN AND SUBURBAN MODES

speech of the younger generation. Data from Nabk, a small town 80 km north of Damascus show that the third person feminine and plural suffixes are /-a/, /-on/ respectively (Gralla 2003). Unfortunately no statistics are provided in the studies mentioned above, nor some indication of the extent to which (h) is variable. It is noticeable that there is no empirical evidence to confirm that the dialect of Damascus was at some point consistently h-full. The available evidence suggest that the traditional dialect is variable with respect to this feature. Thus, it is theoretically possible to argue that the presence of [h] is an innovation, and that the h-full tokens are cases of h-insertion. In fact, the only conclusion that can be deduced from the data in the literature is that variation in (h) in the dialect of Damascus is a case of inherent variability, which may have been brought about originally through dialect contact. The Sprachatlas von Syrien (Behnstedt 1997) sheds some light on the origins of the variation in (h). Map 265 (530–531) shows Damascus as an [h]-variable zone. Three patterns of variation are shown in the city and in Greater Damascus (Map 9.2). Within the city, (h) is shown to be used interchangeably with zero. In the area of Greater Damascus, [h] is present only after vowels in the western part of Greater Damascus, including Deemaas which is the residence of many commuters to Damascus. In towns located to the east of Damascus (M¡a\\amiyit Sham, öarasta, and Arydz), however, [h] can be retained after a consonant and after a vowel. Towards the south of Damascus, few dispersed towns retain [h] only after vowels (Dbr ¡Ali, 25 km to the south of Damascus, and Salkhad, 34 km to the south, near Suwaida). In the country as a whole, the geographical distribution of (h) in the suffix shows that the h-zero pattern predominantly extends along the western strip of Syria. The farther one moves away from the coast the more the h-full pattern can be observed. The western strip, including Damascus, belongs to the Syro-Lebanese dialect continuum, with the exception of Aleppo, which is considered part of the North Syrian dialect group. According to the map, Damascus is surrounded by h-full dialects on all sides (Qalamun and Horan dialects). Economic migration and daily commuting from these h-full towns into the capital can offer an explanation for the variation found in the city. The city of Aleppo in the north shows a similar pattern to that found in Damascus. The similarity is likely to be due to Aleppo’s location at the border of h-full dialects (to the north, south, and east) and h-less dialects (to the west). Aleppo is the industrial capital of Syria, and has always attracted migrants from surrounding towns as well as from Al-Jazeera (the northeast of Syria). Like Damascus, Aleppo is surrounded by h-full dialects. The hzero form appears to include the coastal cities, whose migrants are mainly from the nearby rural areas and whose dialects are h-zero as well. The map suggests rather strongly that the heartland of h-less dialects are the coastline cities, from where the feature has spread to inland dialects, 199

Map 9.2 Dialect map of Syria

HANADI ISMAIL

200

THE URBAN AND SUBURBAN MODES

Table 9.2 Percentage use of (h) according to the preceding environment [Ø]

[h]

Total

%

Consonant

N %

2,297 96

102 4

2,399

75

Vowel

N %

753 94

47 6

800

25

[h]

Total

%

Table 9.3 Percentage use of (h) according to age groups [Ø] Young

N %

994 99

13 1

1,007

31

Middle

N %

1,148 94

78 6

1,226

38

Old

N %

908 94

58 6

966

30

such as Damascus and Aleppo. It is noticeable that the nearby rural areas of Damascus are h-full, and so are all of the rural locations between Beirut and Damascus. This mode of diffusion seems to fit in the “urban hierarchy model” of linguistic diffusion, which suggests that linguistic innovations can spread from a large city to another large city before they affect rural localities nearer to the center of innovation (Chambers and Trudgill 1980). 9.5.1.3 Results and discussion In this study, the variable (h) is analyzed in terms of presence or absence of [h]. All tokens of (h) after /a:/ were excluded since (h) is not variable in this environment (see examples under 3 above). A total of 3199 tokens of [h] and [Ø] were analyzed using Goldvarb 2001 program. Two environments were coded for, namely, a preceding consonant, and a preceding vowel. Table 9.2 shows the percentage distribution of the variants [h] and [Ø] according to the preceding environment in the sample as a whole. The frequency of [Ø] occurrences is quite high both after a consonant and after a vowel. The results show that the probability of [h] occurring after a vowel is slightly higher than after a consonant: 4 percent and 6 percent, respectively. The values above show that the presence or absence of [h] does not significantly favor a specific environment. Table 9.3 shows that both the middle and old age groups have equal probabilities of [h] use (6 percent), while the young age group’s frequency of [h] use is 1 percent. The difference between the young age group’s use and 201

HANADI ISMAIL

Table 9.4 Percentage use of (h) according to neighborhood [Ø]

[h]

Total

%

Shaghoor

N %

1,676 99

77 4

1,753

55

Dummar

N %

1,374 95

72 5

1,446

45

that of the middle and old age groups does not show statistical significance. Similarly, the use of [h] in Shaghoor and Dummar, as Table 9.4 displays, is almost equal. Figures 9.1 and 9.2 display the use of [h] and [Ø] by gender in both Shaghoor and Dummar. The difference between the female and male patterns of use does not show significance. Females in Shaghoor (Figure 9.1) use [h] in 5 percent of the total, while the males use it in 4 percent. In Dummar, both males and females use [h] in 5 percent. The zero form is clearly favored by both males and females in both communities. It is worth mentioning here that, when asked whether the dialect of Damascus is an h-full dialect or not, speakers could not identify the existing variation, or verify whether they use [h] or [Ø]. In other words, the speakers did not seem to be aware of the variation. The use of [h] by Dummar speakers, except for the young male group, is higher than in Shaghoor. The [h] values in total, however, are relatively low, compared with the values of [Ø] use in both communities which range between 92 and 99 percent. The highest users of the [h] variant are the old male group in Dummar. The leading user of [h] in the professionals’ lifemode group is a 54 year old high-ranking technocrat who was interviewed at his home with his family. His speech style was informal, although some of the lexical items he used were borrowed from the Standard. During the interview, the h-full forms occurred most frequently while the informant was recounting his family history, and in particular while he emphasized his well-established family background. A higher user of [h] in the middle age male group from the professionals’ life-mode is Rami, a 35 year old businessman who presented himself as a soft-spoken modern family man. Rami originally came from a lower socio-economic background than his wife’s. His adopted lifestyle and outlook capitalize on his new mode of production and income, and indeed ratify his membership into the Dummar professional life-mode. On the basis of the performance of these informants, it may be stated that for Damascus speakers the h-full forms are variants which the speakers tend to use in a particular style, e.g. emphasis (cf. Lentin 1981). In Shaghoor, the middle age group use [h] at 6 percent, a figure which slightly exceeds the usage by the old age group (5 percent). The gender difference in each age group is only slight and statistically non-significant. 202

THE URBAN AND SUBURBAN MODES 96

95

100 90 80 70 60

Zero

50 [h]

40 30 20 5

10

4

0 Male

Female

Figure 9.1 Percentage use of (h) by gender in Shaghoor

95

95

100 90 80 70 60 50 40 30 20 10 0

Zero [h]

5

5 Male

Female

Figure 9.2 Percentage use of (h) in Dummar by gender

10 9 8 7 6 5 4 3 2 1 0

Shaghoor Dummar

YF

YM

MF

MM

OF

OM

Figure 9.3 Use of (h) by gender and age groups in Shaghoor and Dummar

203

HANADI ISMAIL

The multivariate analysis conducted on all factors showed a statistical significance of 0.06. The probabilities of the three age groups for the application value [Ø] are: Y = 0.76, M = 0.37, O = 0.39, i.e. the younger age group are the least likely to use [h]. The results overall show that the favored form by both the self-employed and the professionals is the [Ø] variant, with the young age group showing the highest probability of use. The overall pattern of variation in the use of (h) shows no significant differences between age, gender groups, or social groups (here life-mode groups), and the pattern shown in Figure 9.3, a near U-shape pattern, is prototypical of a variable in stable variation (see Labov 1994). Although no statistics with respect to the use of the (h) variants are available in the previous literature (e.g. Blanc, Cantineau, Cowell, Gralla, Grotzfeld, Lentin), it is clear that the [h] variant is the marked and the minority choice in these studies, which is precisely what the figures in my study show. Given the evidence from the previous literature and the statistics and patterns of variation presented above, it is reasonable to conclude that (h) has been in stable variation for a considerable length of time. 9.5.2 Variable (r) Previous studies of sociolinguistic variation in (r) in Arabic cover the context of code switching and borrowing from French in North Africa (Jabeur 1987), and dialect contact in Casablanca (Hachimi 2005). A recent acoustic and auditory analysis of (r) in Jordan was conducted by Khattab and Al-Tamimi (2006 manuscript). The following types of /r/ have been identified in the study: trills [r] [w], taps [ɾ] [ɾ. ], and approximants ['] [ ]. According to Khattab and Al-Tamimi, female speakers lead in the use of the approximant variants in Jordan. In Levantine Arabic and in Egyptian Arabic two realizations of /r/ can be found. These are emphatic /w/, a distinctive feature of Palestinian and Egyptian dialects in the vicinity of low vowels and emphatic consonants, and non-emphatic /r/ (cf. Herzallah 1990; Younes 1994). Uvular fricative [ʁ] has been reported to occur in the speech of Christians in Baghdad (Cantineau 1960: 76) as well as in North Africa (women in Meknes, see Stroomer 2004, and in Morocco generally, Marcais 1977, cited in Stroomer 2004). 9.5.2.1 Description The sound [r] is produced by a single or multiple contacts of the blade or tip of the tongue and the alveolar ridge. It is an apical trill in a geminate and a tap when single (Cowell 1964: 5). Variation in (r) regards weakening of the sound, along a continuum of productions. The attested r-variants in my data are palato-alveolar approximant ['], a retroflex [ ], an alveolar fricative [R] as well as an alveolar trill [r]. The r-variants will collectively be signified 204

THE URBAN AND SUBURBAN MODES

here as ['], and will be treated in the analysis as concerning one change of (r). This is because the data show that the change in the production of /r/ is not unidirectional, and because different speakers adopt different innovative variants. The range of r-variation is broader word-medially and wordfinally, which can be explained by the fact that /r/ is more frequent in these environments than word initially. Examples of the ['] variant are shown in (4) below. (4) Initial /'iwa:™i/ “narrative” /'u:]/ “go”

Medial /di'a:se/ “study” /bt£∫'abi/ “would you drink?” /na∫a't/ “I spread”

Final /xuwa'/ “pictures” /ʔakb£'/ “bigger”

A total of 4,763 tokens of (r) were analyzed using Golvarb 2001 program. Five phonetic environments were coded: word-initial, word-medial preceding a consonant, word-medial preceding a vowel, word-medial intervocalic, and word-final. A few words were excluded from coding: /™ɑwbɑ¡/ “four” and its derived forms, such as /™ɑwb¡i:n/ “forty” and /wwɑ:be¡/ “the fourth”, in addition to /¡ɑwɑbi/ “Arabic”. Emphatic [Ã] was found to obstruct (r) weakening. 9.5.2.2 Results and discussion Table 9.5 displays the frequency of [r] and ['] in five phonetic environments. The distribution of ['] as displayed in Table 9.5 suggests that this variant does not significantly favor a particular phonetic environment, with a value of occurrence which ranges between 12 percent in the environment of preceding consonant to 16 percent in initial position. Figure 9.4 demonstrates that ['] shows minimal gender differentiation in both communities. The variant is adopted almost equally by both sexes. The distribution across both neighborhoods (Figure 9.5) shows that ['] is adopted in 17 percent of the cases by speakers from middle-class Dummar, whereas ['] is used by a frequency of 11 percent in Shaghoor. The cross-tabulation of age groups and neighborhoods, Figures 9.6 and 9.7, show interesting results. The leaders of change in both communities are Table 9.5 Distribution of (r) by phonetic environment

[r] N % [ó ] N %

Initial

Preceding consonant

Preceding vowel

Intervocalic

Final

579 84 110 16

821 88 113 12

844 85 144 15

878 86 138 14

966 85 170 15

205

HANADI ISMAIL 86

100 90 80 70 60 50 40 30 20 10 0

85

r r 14

15

Male

Female

Figure 9.4 Gender distribution of (r) in both communities

89

100 90 80 70 60 50 40 30 20 10 0

83

r r 17

11

Shaghoor

Dummar

Figure 9.5 Distribution of (r) by neighborhood

93

89 76

r r

100 90 80 70 60 50 40 30 20 10 0

24 11

Young

Middle

7

Old

Figure 9.6 Distribution of (r) by age groups

the young age groups with 26 percent of ['] among Shaghoor youth and 23 percent among Dummar youth. The Dummar middle and old age groups use ['] significantly more than their counterparts in Shaghoor (16 percent against 5 percent respectively). The variant ['] occurs only in 2 percent among the old age speakers in Shaghoor, against 13 percent in Dummar. 206

THE URBAN AND SUBURBAN MODES

30 25 20

r Shaghoor

15 r

10

Dummar

5 0 Old

Middle

Young

Figure 9.7 Distribution of (r) by age groups in both communities

50 40 30

29

27

23

Male

21

20 14

20 10

10

Female

10 4 5

13

0 YS

YD

MS

MD

OS

OD

Figure 9.8 Distribution of (r) by age groups (Y, M, O) and gender in each neighborhood (S, D)

The multivariate analysis run on all factors showed that the best groups selected by the stepping up analysis are age groups and neighborhoods. The VarbRul weighting of the best factors are: Y = 0.69, M = 0.45, O = 0.35, S = 0.43, D = 0.57. This indicates that the age factor has the most significant effect on ['] with a range value of 34. The statistics overall suggest that the variation in (r) represents a change in progress. Figure 9.8 shows the distribution of ['] by age groups and gender in each neighborhood. Young and old women in Dummar (YD and OD) use more ['] variants than men in both age groups. The young Dummar women use the new variant at 27 percent, while young Dummar men use it at 10 percent. Similarly, the old Dummar women use ['] at 20 percent, while men of the same age group use it at 14 percent. In Shaghoor the age groups’ pattern shows a sharp age difference, with young age groups using the highest values of ['], nearly 30 percent, followed by the middle age group who use it in approximately 5 percent of the total occurrences, and the old age group whose usage does not exceed 2 percent. While in Dummar the age differences are considerably more gradient than in Shaghoor, the gender differences in Dummar are sharper than in Shaghoor. 207

HANADI ISMAIL

The young age speakers in both communities are the advanced users of [']. Interestingly, Figure 9.8 shows that the most advanced users of the innovative feature are the young men in Shaghoor, followed closely by the young women of Dummar, i.e. both groups can be considered as the leaders of the change in their respective community. The fact that the middle and old age groups in Dummar participate, and to a considerable extent, in this variation, while the same age groups in Shaghoor show a very low rate of participation suggests rather strongly that the change originated in Dummar or at least that it did not originate in Shaghoor. The younger generation in Shaghoor are the early adopters of the variation. The general findings with respect to gender differentiation and sound change in sociolinguistic studies show overwhelmingly that women are most often the leaders of linguistic changes, especially in cases of change from above (Labov 1990). To this end, the behavior of the young men of Shaghoor is rather unusual. In the following section, I attempt an explanation of the gender-related correlations. 9.5.2.3 Men, women and language change: a life-mode account The sex differentiation in the use of ['] in both life-modes, (Figure 9.8), shows interesting patterns. The young men in the self-employed mode have the highest scores of [']. Likewise, in the middle age group of the career/ professional life-mode, men score higher than women in the same group. In order to develop a better understanding of the linguistic behavior of these men, it is important that we first look at the socio-cultural context in which the men and women interact within their modes of production, and the daily social practices in their settings. Within the professionals’ life-mode, old women, whose ages range between 53 and 65, have a relatively high score of [']. The highest users of this group were the wife of a well-known actor, in addition to an ex-wife of a well-known thinker, who works in translation, and is closely linked to the art and acting community. Most of the other women included in this group were of a similar profile and had a similar outlook, namely they capitalize on their husbands’ careers for better achievements. Rather than being “housewives” these women are home-refiners. Following Eckert 1989, we would expect this particular group to be most concerned with accumulating symbolic capital. They are members of the refined society of Dummar, but it is through their husband’s status that their roles in the community are legitimized. The legitimacy of their membership in the group is dependent not on what they do, but on how they appear, i.e. on the symbolic capital. They, their looks, and their skills at keeping a refined home, represent them and their families to the outside world. The change affecting /r/ seems to be associated with the suburb rather than with the inner city, as the statistics show. It would, therefore, be expected that the innovative feature is becom208

THE URBAN AND SUBURBAN MODES

ing a part of the linguistic capital of suburban life, and the groups of speakers for whom the accumulation of such symbolic capital is most important are those whose membership in the suburban community is dependent upon possessing the symbols, namely the unemployed wives of “important” men. By comparison, the younger women in this mode were all single, and were either still in education or have just started their careers. In the selfemployed mode, the young women are the highest female users of ['] in the community. Among the male speakers, the highest users of ['] in the professionals’ life-mode are the middle aged men. Most of these men can be described as progressivists, in terms of social behavior and ideologies. Being in their thirties, single and having finished their education, these men were financially independent and mobile. By comparison, the young men in Dummar were still in education. In this life-mode, education is a tool to better one’s social and professional status. In Shaghoor, none of the women had a job at the time of data collection, with the exception of a young woman who worked as a local hairdresser. A good number of them used to work from home in the past, while one of them worked in an uptown neighborhood of Damascus as a seamstress before starting a family. In fact, this woman was the highest user of [']. She and other female high users of the innovative feature also had a relatively anti-local attitude towards the neighborhood. This attitude was manifested through their constant criticism of local practices, views and residents and their desire to live outside the neighborhood. They had the weakest social links in the community. The behavior of the young men in Shaghoor is of most interest. Informants from this group were in jobs, and their jobs were mostly in the retail business, which put them in direct contact with customers from different backgrounds. The highest users were an 18 year old who ran the family patisserie shop with his father, and his two in-laws, one of whom worked in a local retail store, and the other in a downtown factory. It seems to be the case that in Shaghoor regular contact with outside communities, e.g. through a profession in the retail sector, results in more exposure to the new feature. Bearing in mind that ['] originated elsewhere, the speakers with outside contacts are the carriers of the innovation back into their local community.

9.6 Conclusion In this article, I have presented two cases of variation. The variable (h) typically shows a pattern which is characteristic of variables in stable variation, as evidenced by the fact that there is little generational difference in apparent time. Some real time evidence, albeit without statistics, is also available from the literature, and confirms that in the dialect of Damascus cases of h-full pronunciations can be found although the dialect is essentially 209

HANADI ISMAIL

an h-less dialect. On the basis of the geographical distribution of this feature, as presented in Behnstedt’s Atlas, I have proposed that dropping of [h] in the suffixes may have spread to inland cities, such as Damascus, following the principles of the urban hierarchy model, which essentially predicts that linguistic innovations may hop from one large urban center onto another before affecting rural areas which are nearer to the center of the innovation. In other words, a linguistic innovation can spread from city to city, then to smaller towns and finally to rural areas, regardless of the distances between these regions. This model does not replace the spatial model which simply predicts that innovations which originate in a particular locality will spread to nearby localities before they affect farther away regions; rather it refines it, by incorporating the factors of size of the locality (and hence demography), and contact (Chambers and Trudgill 1980, Trudgill 1986). The second variable presented in this article is (r). In this case, the variation attested seems to be symptomatic of change in progress, as confirmed by the statistical tests. The predicted change to ['] seems to have originated in the suburb, and to have spread to the inner city community of Shaghoor. The case of (r) is rather fascinating, particularly because the male speakers in both communities participate quite actively in advancing the new form. The analysis of the data from Damascus attempts to apply the notion of life-mode, which allowed us to sift through the social practices on account of mode of production, rather than on the basis of the pre-defined category of social class. This approach is still not widely applied in the study of linguistic variation and change in the city (cf. Marshall 2004). In addition, the present study of the mechanisms of linguistic change in suburbia is a novel attempt in the field of Arabic sociolinguistics.

Notes 1 I am grateful to Enam Al-Wer for her comments and suggestions on earlier drafts of this chapter. I also warmly thank Peter Behnstedt for designing a special map to fit in the article. 2 Jabri Avenue is a pseudonym of the local sampling Avenue. 3 Jerome Lentin (personal communication) suggested to me that [h] is not always obligatorily present after the low vowel [a:], and that it can be dropped in this environment. In my data, no examples of deletion after [a:] can be found. 4 Cantineau mentions also that h-zero form is found in the urban dialects of North Africa, particularly in the dialects of the Jews in Algeria, as in /∫£γlum/ “their business”. He also cites /mɔ:nna/ as in /mun huna/ “from here”.

References Al-Qayyem, A. (1991) Dimashq §aqdam madCna ma§hula fi-t-tArCkh (Damascus: the Oldest Inhabited City in the World ), Damascus: Dar Al-Ahali. Ambros, A. (1977) Damascus Arabic, Malibu: Undena Publications.

210

THE URBAN AND SUBURBAN MODES

Barakat, H. (2001) Al-mujtama∞ al-∞arabC al-mu∞AXir (The Modern Arab Society), Beirut: Markaz Dirasaat Al-Wahda Al-Arabiya. Behnstedt, P. (1997) Sprachatlas von Syrien, Weisbaden: Harrassowitz Verlag. Bergsträsser, G. (1924) Zum Arabischen Dialekt von Damaskus, Hannover: OrientBuchhandlung Heinz Lafaire. Blanc, H. (1953) Studies in North Palestinian Arabic: Linguistic Inquiries Among the Druzes of Western Galilee and Mt. Carmel, Jerusalem: Israel Oriental Society. Cantineau, J. (1960) Cours de phonétique arabe, translated by Saleh Al-Qarnaadi (1966), Tunis: Centre of Social and Economic Research and Studies. Chambers, J. K. and Trudgill, P. (1980) Dialectology, Cambridge: Cambridge University Press. Cowell, M. (1964) A Reference Grammar of Syrian Arabic (based on the dialect of Damascus), Washington, D.C.: Georgetown University Press. Eckert, P. (1989) “The whole woman: sex and gender differences in variation”, Language Variation and Change 1: 245–267. Gralla, S. (2003) “Le système vocalique du dialecte Arabe du Nabk (Syrie): changements diachroniques et synchroniques”, in I. Ferrando and J. J. Sánchez Sandoval (eds) AIDA Fifth Conference Proceedings, Cádiz 2003, pp. 501–509. Greer, S. (1960) “The social structure and political process in suburbia’, American Sociological Review 25: 4, 514–526. Grotzfeld, H. (1965) Syrisch-Arabische Grammatik (Dialect von Damaskus), Wiesbaden: Harrassowitz. Hachimi, A. (2005) “Dialect leveling, maintenance, and urban identity in Morocco”, unpublished thesis, University of Hawaii. Herzallah, R. (1990) “Aspects of Palestinian Arabic phonology: a non-linear approach”, unpublished thesis, Cornell University. Hojrup, T. (2003) State, Culture and Life-Modes: the Foundation of Life-mode Analysis, Aldershot: Ashgate. Hopkins, N., (ed.) (2001) The New Arab Family, Cairo Papers in Social Science 24 (1–2), Cairo: American University in Cairo Press. Jabeur, M. (1987) “A sociolinguistic study in Tunisia: Rades”, unpublished thesis, University of Reading. Khattab, G. and Al-Tamimi, F. (2006) “Variation in (r) production by Jordanian speakers”, manuscript. Labov, W. (1989) “The exact description of the speech community: short /a/ in Philadelphia”, in R. Fasold and D. Schiffrin (eds) Language Change and Variation, Washington, D.C.: Georgetown University Press, pp. 1–57. Labov, W. (1990) “The intersection of sex and social class in the course of linguistic change”, Language Variation and Change, 2: 205–254; reprinted in J. Cheshire and P. Trudgill (eds) The Sociolinguistic Reader, Vol. 2: Gender and Discourse, London: Arnold, 1998, 7–52. Labov, William (1994) Principles of Linguistic Change, Vol. I: Internal Factors, Oxford: Basil Blackwell. Lentin, J. (1981) “Remarques sociolinguistiques sur l’arabe parlé a Damas”, unpublished thesis, University of Paris III. Lentin, J. (2006) “Damascus Arabic”, in Encyclopaedia of Arabic Language and Linguistics, Leiden: Brill, pp. 546–555. Marcais, Ph. (1977) Esquisse grammaticale de l’arabe maghrébin, Paris: Maisonneuve.

211

HANADI ISMAIL

Marshall, J. (2004) Language Change and Sociolinguistics: Rethinking Social Networks, Basingstoke: Palgrave Macmillan. Milroy, L. (1980) Language and Social Networks, Oxford: Blackwell. Milroy, L. (1987) Observing and Analysing Natural Language, Oxford: Blackwell. Stroomer, H. (2004) “The Arabic dialect of women in Meknes”, in M. Haak, R. de Jong and K. Versteegh (eds) Approaches to Arabic Dialects, Leiden: Brill, pp. 291–307. Trudgill, P. (1986) Dialects in Contact, Oxford: Blackwell. Versteegh, K. (1997) The Arabic Language, Edinburgh: Edinburgh University Press. Younes, M. A. (1994) “On emphasis and /r/ in Arabic”, in M. Eid, V. Cantarino and K. Walters (eds) Perspectives on Arabic Linguistics VI, Amsterdam: Benjamins, pp. 119–145.

212

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

10 SEGMENTAL AND PROSODIC ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY Mohamed Embarki

10.1 Introduction Our study will focus on the temporal organization of a list of Moroccan Arabic (MA) words produced by native speakers from Ksar el Kebir, an average-sized town in northwestern Morocco, and explore the phonetic, sociolinguistic and dialectal aspects of this parameter. From a phonetic viewpoint, we will examine the cross-speaker variability relating to the duration of the segments (sound, syllable, word) and its phonological status, as well as the temporal setting for the segments. From a sociolinguistic viewpoint, we will compare the temporal organization of those phonetic segments with the social context of their production, based on the theoretical assumption that the age and the gender of the speakers, their academic level in Modern Standard Arabic (MSA), as well as their speech situation influence this temporal organization. From a dialectal viewpoint, we will explore the variety of MA in Ksar el Kebir within a clearly defined space: the city. The temporal organization of the list of words will make it possible to highlight the phonetic clues specific to each type of district, based on the assumption that these clues reflect the past and present relationship between populations and vernaculars of various origins. It might have been possible to divide the phonetic data according to the three viewpoints, but we chose to examine the same phonetic object, vowel, word or articulation rate from phonetic, sociolinguistic and dialectal perspectives. The combination of the three perspectives provides a more detailed account of the diversity of Ksar el Kebir’s modern Arabic. As a result of recent data in Arab dialectology, we decided to use caution in our terminological approach and chose to study the diversity of Ksar el Kebir and not the diversity of MA in general. Indeed, the phonetic data presented here reflects what speakers in this city are likely to produce. Speakers from other cities would generate different elements, with specific phonetic clues. 213

MOHAMED EMBARKI

All things considered, there is not a single modern Arabic, but probably as many varieties of modern Arabic as there are large localities.

10.2 Process of dialectal accommodation in Arabic-speaking urban environments: a brief review of some explanatory factors Following Giles and Coupland (1991), several researchers working within the framework of linguistic accommodation showed how speech adapts according to the partner(s) to mark either a convergence or a divergence, resulting from either social, religious, ethnic or political considerations. In situations of language contact, linguistic accommodation eventually leads to dialectal leveling (Trudgill 1974: 155). Such leveling happens naturally as a consequence of geographic mobility and seems to play a very important part in variation and linguistic changes. From the polarity in the use of standard and vernacular varieties associated with middle, upper and working classes respectively, representations of prestige and covert prestige emerge. The connection, however, seems relatively more complex, and, as Labov (1972) and Trudgill (1974) underlined, pressures are exerted within language communities so that the speakers continue to use less prestigious nonstandard variables as proof of community solidarity and personal identity; in the long run, these acquire their own prestige and are described as covert. From the point of view of gender, men seem more inclined to use varieties with covert prestige (Trudgill 1983). Women are more likely to use standard variables, to follow and adopt the prestigious forms most widespread within their community. Nevertheless, they are the leaders of linguistic change, especially when they occupy a social position associated with higher local prestige (Labov 1994). Early observations of Arab female linguistic practices may have contradicted the tendency outlined above, not that Arabic-speaking women are less attached to prestigious forms than men, but more specifically because of the integration of linguistic variables of MSA to evaluate linguistic behaviors, attitudes and representations in vernacular Arabic, whereas MSA, as Al-Wer underlines, “as a variety is irrelevant in this domain” (2002: 47). Several studies have subsequently corrected these early observations by showing that Arabic-speaking women are receptive to the prestigious character of the current variety within their community (Al-Wer 2002; Benrabah 1992; Vanhove 2002). In the sociolinguistic field, interesting articulations between language and identity were observed, and the language described as a major component and strong marker of identity. As Fishman (1977) underlined, the symbolic function of language is no less important than its communication function in revealing the boundaries of collective identity. Thus the use of marked phonological variables expresses an identity anchoring of a different scale. 214

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

Depending on the circumstances, the same phonological variable (for example /q/) may reflect either a desire to adopt a transnational identity as shown in studies about the Palestinian village of Barta’ (Amara and Spolsky 1996; Amara 1999), or a national identity as shown in the study of Palestinians in Amman (Holes 1995) or by LBCI in Lebanon (Al Batal 2002), or a community identity between the Christians and Moslems of Baghdad (Blanc 1964; Abdel Jawad 1987; Holes 1995), or between Shiites and Sunnis in Manama, Bahrain (Holes 1995) or simply a local identity (Abdel Jawad 1987; Gibson 2002). Geographic mobility is an important factor in the process of transmission of innovative linguistic forms. For reasons of economic, demographic and cultural predominance of cities over rural areas, innovative linguistic forms are initially circulated within urban centers, then from one urban center to another to end up finally in neighboring rural areas (Trudgill 1974). In a large portion of the Arabic-speaking area, the simple process of secondary education is an operation that requires the relocation of the whole family unit. It is no wonder that the schooling level emerges as an extremely important factor in the process of phonological innovation, frequently accelerating the acquisition of urban variables to the detriment of original community variables even if the latter are closer to MSA as shown in, among others, studies by Abdel Jawad (1987), Benrabah (1992), Jabeur (1987), Jassem (1993) and Al-Wer (2002). Thus in the case of Sult (Jordan), Al-Wer (2002) showed that phonological variation of [θ] to [t] was closely related to the level of education of the females surveyed. In the speech analysis, 59 percent of occurrences of the [θ] to [t] variation were observed among people with college education, against 43 percent among people with a high school education, and only 7 percent among illiterate people. The author thus concluded that the degree of education is inversely proportional to the use of phonological elements of MSA when those are different from the locally prestigious variety. Females with a higher education level used the new forms more frequently, whereas those with the lower education level seemed more conservative.

10.3 Urban Arabic varieties in the Maghreb The city is a complex entity where antagonistic forces of homogeneity/ heterogeneity are exerted. This space, which has always been an object of interest to linguists, seems to have activated over the past few years a sociolinguistic revival (Bulot 2004) different from the extensive sociolinguistic investigations initiated by Labov (1978) both from a conceptual and a methodological point of view, and more focused on processes of territorialization. Arabic urban sociolinguistics, if we may call it so, has developed several schools and models which also reflect geographical diversities. Frenchwritten studies relating to old Maghrebian cities, and based essentially on 215

MOHAMED EMBARKI

historical dialectology, show that migrants do not adopt – or no longer adopt – traditional urban standards, but develop new urban standards that are themselves subject to local influences of prestige or covert prestige. These new standards lead to sorts of koines strongly marked by rural phonological features even if they are considered as neo-urban languages because they have become the norm of urban populations in contrast to the traditional urban languages. Researchers such as Messaoudi (2002) thus developed a trichotomy in terms of old urban, new urban and rural vernaculars. One phonological feature, which illustrates among others the coexistence of this North African triple analysis (trichotomie) is the realization (or absence of realization) of the diphthongs. For the sole example of Morocco, several studies have underscored the appearance of the diphthong phenomenon (as in the lexemes [baj\] “eggs”, [sajf ] “sword”, []awli] “sheep” and [mawt] “death”) in Moroccan neo-urban varieties such as those of Fez (Caubet 1998), Rabat (Messaoudi 2002), Salé (El Himer 2004), Casablanca (Ennaji 2004), and have interpreted such phenomenon as a tendency to ruralization/(r)urbanization. The diphthongization process is interpreted by Ennaji (2004) as a mark of ruralization of the Casablanca vernacular, because the MA standard model realizes the first lexeme as [bi\]; El Himer (2004) suggests the [be\] realization for the word “eggs” in “parler urbain slaoui” (PUS) – Slaoui new urban vernacular – and in “parler citadin slaoui” (PCS) – Slaoui old urban vernacular – and the [baj\] realization in “parler à aspects ruraux” (PAR) – semi-rural vernacular. According to Messaoudi (2002) the word “sheep” is realized as []uli] in Rabat’s old urban vernacular, as []awli] in rural vernacular and as an alternation between the simple vowel [u] and the diphthong [aw] in new urban vernacular. Diphthongization, however, is not just a rural feature; it can also be found in a number of “traditional urban” vernaculars in northern Morocco. In Tetouan’s “traditional urban” vernacular, diphthongization still exists in the words [baj\] “eggs”, [sajf ] “sword”, []awli] “sheep” and [mawt] “death” while in Ksar el Kebir’s “traditional urban” vernacular diphthongs alternate with monophthongs [baj\], [sif ], []awli] and [mut].

10.4 Ksar el Kebir’s MA variety Beyond simple diphthong (Tetouan) or monophthong (Rabat, Salé, Casablanca) realizations, the homogeneity of each of these traditional urban vernaculars contrasts sharply with the sometime diphthong, sometime monophtong heterogeneity observed in Ksar el Kebir’s urban variety. This heterogeneity is definitely not accidental; it is the result of historical, ethnographic, geographical, political, and social factors. Historically, Morocco – and to a greater extent all of North Africa – has been the scene of successive waves of population immigration, some coming from the Arabic peninsula, others from Andalusia, still others from the south. If we consider only the 216

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

Arab immigration, the first waves of arrivals began with the conquest of North Africa at the end of the seventh century by the Arabs from Arabia. At that time, Ksar el Kebir, as other cities in Morocco, was Berber. As Ibn Khaldoun described in the third volume of his Universal History, the city was in the territory of the Ghumara, a Berber Masmuda tribe: The Ghumara live in the Rif mountains . . . their country is over five days long, from Ghassasa, north of the Maghreb plains, to Tangiers, and it includes these cities and also Nukur, Badis, Tikisas, Tittawin [Tetouan], Ceuta and Al Qasr. The territory is also five days wide, from the sea to the plains that border Qasr Ketama [Ksar el Kebir] and the Wergha river. (Ibn Khaldoun 2003: 324) The Ghumara, islamicized by Musa Ibn Nusaïr in the middle of the eighth century, stayed in their territory. The town of Ksar el Kebir, on the major thoroughfare between Fez (the Idrissid capital) and the Strait of Gibraltar, had been for two centuries the scene of various successive passages of troops and civilian populations. These waves of immigration had continued between the tenth and the twelfth centuries with the tribes of Banu Hilal and Banu Sulaym. In the late twelfth century, the Almohad khalif Al Mansur established the Riyah tribe, the most powerful and largest branch of the Hilal, in the province of Al Habt, which extends from Qusur Ketama, also called Al Qasr Al Kabir [Ksar el Kebir], to the Azaghar area. The Riyah then settled in the vast plain which extends from there to the Green Sea [the Atlantic Ocean]. (Ibn Khaldoun 2003: 53) Between the twelfth and the fifteenth centuries, various families had moved from Andalusia to Maghreb cities, and in particular to Ksar el Kebir. At the turn of the twentieth century, immigration became primarily internal, with rural migration. During the Protectorate (1912–1956), a rural population coming essentially from the Spanish zone, i.e. from the Masmuda territory, arrived in the city which was then located at the border between the French and Spanish protectorates. The rural migration was more massive in the mid-1970s, because of the construction of the Oued el Makhazine hydraulic dam. These historical factors partly determine other factors. The populations who settled here were not ethnically, linguistically, economically and socioculturally homogeneous. Here the term of layers of immigration interpreted as superimposed is inappropriate: because populations, no matter how diverse, were not isolated from one another. It was in no way a juxtaposition, 217

MOHAMED EMBARKI

but true interaction. In terms of migrations, the construction of the hydraulic dam resulted in the emergence of new peripheral districts (quartiers nouveaux périphériques [QNP] ) in urban areas forming a belt or layer around the old central districts (quartiers anciens du centre [QAC] ). Initially, the QNP accommodated mainly rural populations, who spoke a mountain dialect ( jebli), also described as village or pre-Hilali dialect, and a plains rural dialect (¡rubi), described as Bedouin, whose lands were flooded by dam waters. These districts today include not only families who came with this last wave of immigration and much earlier waves, but also families who have lived in Ksar el Kebir for several generations. Income from salaries is the common denominator to all: these families live on modest or very modest income, where the heads of households are generally day laborers, small craftsmen or unemployed. We will explore the idea that the QNP have probably developed new linguistic standards, different from a segmental perspective from both the original dialects and urban standards. In the absence of officially recognized scientific studies on old urban standards in Ksar el Kebir and based on our knowledge of the subject and of the literature devoted to a number of large Moroccan cities, we can say with confidence that up to the recent past, the language spoken in the town of Ksar el Kebir was different both from northern cities such as Tetouan or Tangiers and from central cities such as Fez, Rabat and Casablanca. We will attempt to determine whether the contact between the town variety and the rural varieties generated, as it did elsewhere, an urban standard influenced by rural phonetic indices. We will focus on the phonetic segments’ duration and on their phonological and prosodic lengthening through the vowel quantity and the articulation rate. The assumption is that speakers from various ethnic origins adopt the local standards for better integration and focus primarily on salient phonetic characteristics such as consonants and vowels, whereas more subtle characteristics such as duration are less used, and therefore, remain controlled by the maternal variety. Whereas the urban classification of the city into QAC/QNP does not reveal a difference among speakers, because they all adopted the same standard by using the same phonological units, the difference may prove more evident in less salient characteristics such as temporal speech organization.

10.5 Corpus and population We used a corpus of 10 lexemes in MA of the C1a1C2a2 type (C = consonant; a = vowel [a] ) as follows: [sa¡a] “hour”, [γaba] “forest”, [¡ada] “habit”, [ba¡a] “he sold it”, [qama] “size”, []aa] “thing”, []ala] “condition”, [ara] “neighbor”, [qa¡a] “place” and [aa] “glass”. The speakers were selected in six schools of Ksar el Kebir, two elementary schools (Tawhid and Sidi Issa Ben Kacem), two middle schools (Mansour Dahbi and Imam Muslim) and two high schools (Mohammadi and Ahmed Rachidi) evenly distributed 218

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

geographically between QAC and QNP. We did not include speakers from nursery schools for three reasons: (1) too much variability in the literature about the time of appearance of socially marked variables among children under 6; (2) the acquisition process by children for the linguistic variables analyzed here is delicate and spread over time; (3) the extreme heterogeneity of the schools with regard to structure, personnel, personnel qualification, and activities offered to the children. We did not include college students either, because there is no higher learning institution in the city. Therefore, integrating speakers who attended college in other cities would have required a control of parameters which we chose not to study here. Our sample includes 48 speakers (24 boys and 24 girls), selected by a questionnaire according to their birthplace and their parents’ birthplace, their place of residence (QAC/QNP), their age, their sex, their schooling level and their family’s socio-economic category. All the speakers are Ksar el Kebir natives, as are their Arabic-speaking parents. The heads-of-household were – at the time of our survey – small craftsmen or daily laborers; the mothers’ occupations were not taken into account. The age made it possible to select speakers who had not failed school, the optimal age for the first schooling level (1st year of the elementary cycle) being 7, the age for the last (final level) being 18. Within the six schools selected, each level was represented by two speakers, a girl and a boy, of the same age, the difference between two adjacent school levels always being one year. The recordings took place within the establishments, during school hours, often in the presence of a teacher and occasionally in the presence of the Director. The speakers did not have any written support for the production of words, each lexeme was guessed beforehand by the speakers before being pronounced. 480 lexemes were collected in all, the various sequences (C1, a1, C2 and a2) were segmented and their duration measured using the PRAAT program. The duration of the word and of the two vowels ( [a1] and [a2] ) were measured, the [a1]/[a2] duration ratios and the articulation rates were calculated.

10.6 Results 10.6.1 [a1] and [a2] vowel duration The mean duration of [a1] among the 48 speakers is 137 milliseconds (ms) with a standard deviation (Δ) equal to 29.15, that of [a2] is 117ms (Δ = 26.67). The difference of 20ms between means is significant in the Student’s t-test (p < 0.0001); the inequality of the variances in the two samples is not significant [F (1, 479) = 1.195; p = 0.052]. When associated with the district variable, the results show among the girls from the QAC a 147ms (Δ = 26.95) and a 131ms (Δ = 30.31) mean duration, respectively for [a1] and [a2]; the difference of 16ms between means is significant in the Student’s t-test (p < 0.0001). Among the girls from the QNP the results 219

MOHAMED EMBARKI

show a 145ms (Δ = 31.76) mean duration for [a1], and a 122ms (Δ = 26.13) for [a2]; the difference of 23ms between means is significant in the same test (p < 0.0001). Cross area, the mean duration of [a1] only differs by 2ms, this difference is obviously not significant in the Student’s t-test (p = 0.569); that of [a2], which is more sizeable (9ms) is significant in the same test (p = 0.013). The boys from the QAC realized [a1] with a 126ms (Δ = 27.16) mean duration and [a2] with a 112ms duration (Δ = 23.20), the difference of 14ms between means is significant in the Student’s t-test (p < 0.0001). The boys from the QNP produce the same differences between the two vowels, [a1] with 129ms (Δ = 24.51), and [a2] with 104ms (Δ = 17.42), the difference between means are significant in the same test (p < 0.0001). Like the girls, the boys from the two types of district produce a significant difference in the Student’s t-test (p = 0.003) in the only means of [a2]; the difference in [a1] is not significant (p = 0.278). The mean duration according to the type of district indicates that the speakers from the QAC realized [a1] in 136ms (Δ = 29.04) and [a2] in 121ms (Δ = 28.62), the difference of 15ms is significant in the Student’s t-test (p < 0.0001); among the speakers from the QNP, the difference of 24ms between the two means is significant in the same test (p < 0.0001), [a1] has a 137ms mean duration (Δ = 29.32) and [a2] 113ms (Δ = 23.89). The same [a1] vowel has a nearly equal mean duration, the difference is not significant in the Student’s t-test (p = 0.782), whereas [a2]’s difference is significant (p < 0.0001) between the QAC and QPN districts. Except for differences in gender, the first vowel [a1] seems reluctant to adopt socially marked variations. The reasons for this limitation are probably related to prosodic constraints. On the other hand, the second vowel [a2] duration, seems to be more flexible and makes it possible to establish social demarcations: girls vs boys and QAC vs QNP. The tendency in the QNP is the shortening of [a2], which creates an easily perceptible mark. 10.6.2 Vowel quantity Embarki and Guilleminot (2003) showed that the vowel duration in Ksar el Kebir’s MA did not make it really possible to oppose short and long phonological vowels. The ratio between vowels in this study is only 1.06 whereas the long/short vowel ratio in languages where the vowel quantity is phonological varies between 1.3 and 1.9 (Embarki 2002: 173). In MSA, the long/ short vowel ratio varies between 2.0 and 3.1; in dialectal varieties, it fluctuates between 1.3 and 2.08 (Embarki 2004: 201). Considering their equivalent in MSA with long [a1] and short [a2], the 10 lexemes in the corpus could have a duration opposition ( [a1] = long, [a2] = short). The mean ratio [a1]/[a2] for the group (n = 480) is 1.21 (D + 0.11), higher than the ratio reported by Embarki and Guilleminot (2003) but is less than the ratio of 1.3 for languages of weak contrast. Among other 220

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

1.5

a1/a2 mean ratio

1.4 1.3 1.2 1.1 1 0.9 0.8 1

2

QAC

3

4

5 6 7 8 Schooling level

9

10

11

12

QNP

Figure 10.1 Evolution of the a1/a2 mean ratio according to the schooling level in both areas

factors, the context in which the interviews were conducted could explain the differences between the two studies. The corpus analyzed here was observed in school circles, whereas that of Embarki and Guilleminot (2003) was observed in out-of school circles. The school institution generally seems to influence the group’s response. The speakers amplified the variations between [a1] and [a2] because they associated them with equivalent lexemes in MSA, which are produced by quantitative opposition, [a1] being long. When we looked at the results of the [a1]/[a2] ratio according to the urban geographical area, we observed that the mean [a1]/[a2] ratio is 1.25 (Δ +0.35) among speakers from the QNP, while it is only 1.16 (Δ = 0.32) among speakers from the QAC, the difference between the means is significant in the Student’s t-test (p = 0.005). As Figure 10.1 shows, the ratios between the two vowels do not evolve in the same way between the first and the final school years. At the beginning of elementary school, the two means are very unequal, 1.16 and 1.35 for QAC and QNP speakers respectively. At the end of the same cycle, the ratios are reversed, 1.35 and 1.11. At the beginning and at the end of middle school and high school, the mean ratio between areas is very close and seems to follow the same development curve. When we compare the evolution within each area, we note that means between QAC speakers, at the beginning of the school cycle (1st year) and at the end of the cycle (12th year) are very similar, whereas among QNP there are many differences between the two extremes. 221

MOHAMED EMBARKI

Boys from the QNP present a higher mean ratio (1.28; Δ = 0.36) than boys from the QAC (1.16; Δ = 0.34), with the difference in means being significant in the Student’s t-test (p = 0.013). The same trend seems to emerge among the girls; the [a1]/[a2] ratio is more noticeable in the QNP (1.22; Δ = 0.34) than in the QAC (1.16; Δ = 0.31), although the difference between the means is not significant in the Student’s t-test (p = 0.166). Within the same area, there is no difference in [a1]/[a2] ratios between girls and boys from the QAC, with both having a same mean of 1.16. Among speakers from the QNP, the girls’ mean is 1.22 (Δ = 0.34) and the boys 1.28 (Δ = 0.36). The difference between the means is not significant for the Student’s t-test (p = 0.230). The data collected in this section leads to the following observations: (1) the vowels of Ksar el Kebir’s MA variety are not opposed by their duration, the [a1]/[a2] mean ratio is not as obvious as that of languages with very weak contrasts; (2) girls and boys seem to handle the duration ratio between the two vocalic segments differently; (3) the differences related to the area variable seem to be confirmed: speakers from the QNP of Ksar el Kebir present more important mean duration ratios than those from the QAC. The differences reveal a salient social identity, the mark from the QNP speakers is an imbalance in the duration to the benefit of the first vowel in the word; (4) the variation seems to reflect a relative awareness of the formalism of the interview location and finally, the influence of MSA’s phonological features: the higher the [a1]/[a2] ratio in MA, the higher the awareness. It appears that awareness of the formalism of the interview location is unequal between the QAC and QNP speakers: the boys and girls from the QNP are more sensitive to the normative influence of the school, while their QAC counterparts seem to minimize the solemnity of the location since they produce a weaker ratio. For this reason, the similarities between younger and elder students are obvious in the QAC, whereas the QNP children do not follow their elders’ example. 10.6.3 Word duration From a developmental point of view, the intrinsic duration of the phonetic segments decreases as the subjects mature and gain better control of their phonatory organs. We should then expect a progressive decrease of the word mean duration. It reached 425ms (Δ = 62.05) among the 48 speakers (n = 480). The boys from the QAC had a 403ms mean duration (Δ = 51.28), whereas it was only 393ms (Δ = 38.19) for the QNP boys, the difference between the means is not significant in the Student’s t-test (p = 0.092). The mean duration was 457ms (Δ = 61.10) for the girls from the QAC versus 446ms (Δ = 68.96) for the QNP girls, the difference of 11ms between means is not significant in the same test (p = 0.183).

222

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY 550

Duration in ms

500 450 400 350 300 1

2

3

4

5

6

7

8

9

10

11

12

Schooling level QAC-B QAC-G

QNP-B QNP-G

Figure 10.2 Word duration among 4 groups (B = boys; G = girls); A = QAC; P = QNP)

The evolution of mean rations according to the schooling level, gender and geographic area highlights some variations (see Figure 10.2). According to area, there are clear differences between young beginners and adults at the end of their school years. From a developmental point of view, boys and girls of the same age should produce segments of practically equal duration. This is not the case across the area. As they begin school, boys and girls from the QAC have the same mean word duration, whereas those from the QNP have distinct profiles, and the girls’ mean word duration is notably longer than the boys’. The results are completely reversed by the end of the school years: girls and boys from the QNP produce similar mean durations, whereas their QAC counterparts produce distinct mean durations, with the girls’ mean duration being longer than the boys’. The difference is not coincidental and is probably the result of very clear social representations of female and male speech. At this time, the quantification of the duration allows us to distinguish two models simply by their spatial location: the first model, located in the QAC and described here as old urban, produces a relatively long word duration; the second model located in the QNP and described here as new urban, produces a shorter duration. Therefore, 15 to 18 year olds from the QAC adopt socially different behaviors, the girls follow the old urban model, the boys the new urban model. In the QNP, there is no difference in gender among 15 to 18 year olds and boys and girls

223

MOHAMED EMBARKI

seem to have adopted the new urban model. Young QAC speakers do not seem receptive to their elders’ models, while their QNP counterparts have apparently followed the example from the QAC 15 to 18 year olds with their socially different behaviors. Our study will have to be expanded not only with an investigation of the representations of the different varieties existing in Ksar el Kebir, but also with a contrastive study including other standards (Casablanca, Fez or Tetouan). 10.6.4 Articulation rate Embarki (1996) showed that in spontaneous speech, the mean articulation rate (the number of syllables/second (Syl/S)) in MA is 5.50 Syl/S (mean observed in a sample of over 5 hours of spontaneous speech produced by 10 adult Moroccan subjects). Given the important age variability in the group, this mean will be used here as an adult model. In our present study, the mean articulation rate of the 48 subjects is 4.79 Syl/S (Δ = 0.68; n = 960). The mean varies with the type of area: the speakers from the QAC have a mean articulation rate of 4.74 Syl/S (Δ = 0.69), those from the QNP a mean rate of 4.85 Syl/S (Δ = 0.66), the difference between the means is not significant in the Student’s t-test (p = 0.067), nor is there significance in the inequality of the variances [F (1, 239) = 1.072, p = 0.590]. Depending on the subject’s gender, the means reveal very small differences: the girls from the QAC have a mean rate of 4.45 Syl/S (Δ = 0.62), the girls from the QNP have a slightly higher mean rate (4.58 Syl/S, Δ = 0.69), the difference between the means is not significant in the Student’s t-test (p = 0.116); the boys from the QAC have a mean articulation rate of 5.03 (Δ = 0.63), their QNP counterparts’ rate is also slightly higher (5.12 Syl/S, Δ = 0.52), the very small difference between the means is non significant (p = 0.209). Whatever the type of area, the mean articulation rate seems to be relatively invariable: the girls seem to have a slower mean articulation rate than that the boys with a faster mean articulation rate for speakers from the QNP compared to their QAC counterparts. When the area and gender variables are combined, differences emerge: the difference between the mean articulation rates of girls and boys from the QAC is significant in the Student’s t-test (p < 0.0001), the difference between the means of boys and girls from the QNP is significant with the same test (p < 0.0001). When the age/schooling level variable is associated with the area and gender variables, different evolution curves are revealed, as shown in Figure 10.3. The progression of the curves shows that at the beginning of school (at age 7), boys and girls from the QAC have an identical mean articulation rate (4.59 Syl/S). At the same age, boys and girls from the QNP produce a difference of over one Syl/S, at 4.28 and 5.31 Syl/S respectively. There are general trends among the four groups: between the ages of 7 and 11 the 224

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

6 5.5

Nbr Sy1/s

5 4.5 4 3.5 3 7

8

9

10

11

12

13

14

15

16

17

18

Age QAC-B

QAC-G

QNP-B

QNP-G

Figure 10.3 Evolution of the mean articulation rate by age among the 4 groups (B = boys; G = girls)

speech flow becomes slower, among girls and boys from both types of area. At the age of 12, the mean articulation rate follows a different evolution curve depending on gender and on the type of area. After reaching a peak of 5.49 Syl/S between the ages of 12 and 13, the mean articulation rate for the girls from the QAC slows down in harmony with their age and education progress. Female speakers in their senior year of high school (18 years old) have a mean articulation rate of 4.25 Syl/S, an even slower rate than the youngest female speakers (7 years old) from the same area. Among girls from the QNP, a significant change seems to occur at age of 12: the articulation rate increases. But contrary to their counterparts from the QAC, the girls from the QNP have a mean articulation rate which grows with age and school progress. The oldest female speakers (18 years old) and in their senior year, have a mean rate of 5.52 Syl/S. Among the boys from the QNP, the mean rate fluctuates between 4.68 and 5.35 Syl/S, with the two extremes of 5.31 Syl/S for the 7 year-old speakers and 5.35 Syl/S for the 18 year old. Although weak, the rate increase is constant beginning at the age of 12 (4.75 Syl/S). The mean rate among boys from the QAC progresses from 4.69 Syl/ S at 11 to 5.90 Syl/S at 18. The articulation rate is thus handled differently by girls and boys from the two types of area. Except for the girls from the QAC, the flow of the other three groups follows an ascending evolution curve that extends over the whole period of secondary studies. This increase is nevertheless in 225

MOHAMED EMBARKI

conformity with the developmental data. If we consider only the data about the 18 year-old speakers, we observe that the speakers from the QAC are at both ends of the evolution curves, with the fastest articulation rate among boys (5.90 Syl/S) and the slowest for the girls (4.25 Syl/S); while speakers from the QNP have a fast rate of articulation, it is very similar, 5.35 and 5.52, respectively for boys and girls. As for the word duration, the articulation rate is marked socially: the old urban model presents a slow articulation rate – a model maintained by the girls from the QAC – while the new urban model presents a rather fast articulation rate – a model adopted by the boys from both types of area and by the girls from the QNP.

10.7 Discussion and conclusion Duration, a non-salient parameter from a sociolinguistic point of view, emerges as a rather helpful variable. The urban organization of the town of Ksar el Kebir into QAC/QNP leads to a different organization from that of the temporal parameter. It goes beyond socio-professional divisions and markings relating to gender. In Ksar el Kebir’s QAC and QNP, girls and boys, primary school children and young adults in their senior year do not handle duration in the same way. The data analyzed here shows that the urban division between QAC and QNP affects even the slightest prosodic parameters which then become marked socially. The “linguistic territories” in the city reflect both a diversity of origins (old urban and new urban) and a symbolic connection with the city. The old urban model competes with the model from the QNP whose impact extends beyond its own space and attracts speakers even from within the QAC, particularly teenagers and young adult males. This attraction of the QNP model is probably emphasized by standards from other cities, such as Casablanca, more readily available on the radio and television, or by trendy speech patterns massively represented in new musical forms (raï, rap, etc.). As mentioned above, the urban standards which, with the help of the media, follow the standard of Casablanca (Ennaji 2004) or Tunis (Gibson 2002), are only moderately influenced and in varying degrees by the characteristics of ruralization. As in the four lexemes presented in Section 10.3, our results show the existence of a true interaction between the varieties of Arabic in Ksar el Kebir. The external layers of immigration that affected the city cannot simply be superimposed, because they introduce various forms of Arabic. Depending on the dialectal origin of the latest newcomers, a number of subtle phonetic indices of the original dialect are still in use. Our data show that linguistic dynamism on the scale of smaller urban structures – neighborhood areas, for instance – exerts influence beyond the area’s boundaries. From a vertical point of view, Ksar el Kebir’s linguistic variety, considered as a product of the various communities who settled side by side, is definitely marked by social dynamics. In fact, the temporal 226

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

organization reveals a plurality of uses based on models ranked as prestigious (old urban) and less prestigious (new urban, rural). From a horizontal point of view, the MA variety in Ksar el Kebir is analyzed in relation to other varieties in Morocco. The phonetic data presented here are part of a better contour definition, compared with the contours of other urban varieties in Morocco. In the same geographical space, either reduced, as a neighborhood area, or more extended, as a national area, the contact between various linguistic varieties produces new standards by mixing old and new forms. If by way of example, populations in urban centers easily adopt diphthongization, it is probably for linguistic as well as historical reasons. The four lexemes (see Section 10.3) were realized in old Arabic and are realized in MSA with diphthongization. It seems like a combined product of persisting old features and the influence exerted by MSA on an increasing number of people. The results concerning vowel duration presented here show a tendency to the lengthening of the first vowel which can be explained on the one hand by the persisting vowel quantity feature in a number of rural varieties, as the pre-Hilali and Bedouin varieties, who continue to live side by side in Ksar el Kebir and whose mark is particularly evident in the QNP production. In addition, this lengthening trend must be seen in relation with the schooling level and the layer borrowed by MA lexemes from their MSA counterparts. Ultimately, the new urban form in Ksar el Kebir is primarily the product of the contact of various popular varieties that belong to a network of relations in a diasystem, but it is also a variety which remains, through its educated subjects, under the permanent influence of MSA.

References Abdel Jawad, H. (1987) “Cross-dialectal variation in Arabic: competing prestigious forms”, Language in Society, 16: 356–368. Al Batal, M. (2002) “Identity and language tension in Lebanon: the Arabic of local news at LBCI”, in A. Rouchdy (ed.), 91–115. Al-Wer, E. (2002) “Education as a speaker variable”, in A. Rouchdy (ed.), pp. 41– 53. Amara, M. H. (1999) Politics and Sociolinguistic Reflexes: Palestinian Border Villages, Amsterdam: John Benjamins. Amara, M. H. and Spolsky, B. (1996) “The construction of identity in a divided Palestinian village: sociolinguistic evidence”, in Y. Suleiman (ed.) Language and Identity in the Middle East and North Africa, Richmond, Surrey: Curzon, pp. 81– 99. Benrabah, M. (1992) “The influence of urbanization on one dialect group”, unpublished conference presentation, Sociolinguistic Symposium 9, University of Reading. Blanc, H. (1964) Communal Dialects in Baghdad, Cambridge, MA: Harvard University Press.

227

MOHAMED EMBARKI

Bulot, T. (2004) Vol. I: Lieux de ville et identité: Perspectives en sociolinguistique urbaine; Vol. II: Lieux de ville et territoires: Perspectives en sociolinguistique urbaine, Paris: L’Harmattan. Caubet, D. (1998) “Etude sociolinguistique des traits préhilaliens dans un dialecte en voie d’urbanisation à Fez”, in J. Aguadé, P. Cressier and A. Vicente (eds) Peuplement et Arabisation au Maghreb Occidental: Dialectologie et Histoire, MadridZaragoza: Casa de Velazquez, pp. 165–175. El Himer, M. (2004) “Ville de Salé: transformations spatiales, sociales et linguistiques”, unpublished conference paper presented at the ESF/SCH exploratory workshop, “The evolution of Arabic urban vernaculars: the effects of migration and social changes”, Aix-en-Provence, 21–23 November. Embarki, M. (1996) “Le discours spontané en Arabe Marocain: mise en évidence de stratégies discursives individuelles dans l’interaction”, unpublished PhD thesis, University of Franche-Comté, Besançon. —— (2002) “L’acquisition de l’allongement vocalique en arabe marocain: production de jeunes enfants marocains en âge préscolaire”, in Actes des XXIVèmes JEP, 24–27 June, Nancy, 173–176. —— (2004) “Quantité vocalique et nature du lexique en arabe marocain”, in Actes des XXVèmes JEP, 19–22 April, Fez, 201–204. Embarki, M. and Guilleminot, C. (2003) “The moving boundaries of the firstacquired variety’s phonological features: evidence from production/perception of Moroccan Arabic’s vowels”, Proceedings of the 15th ICPhS, 3–9 August, Barcelona, 639–642. Ennaji, M. (2004) “Urbanisation in Morocco and changes in Arabic: the case of Casablanca dialect”, Conference at ESF/SCH Exploratory Workshop, “The evolution of Arabic urban vernaculars: the effects of migration and social change”, 21–23 November, Aix-en-Provence. Fishman, J. (1977) “Language and ethnicity”, in H. Giles (ed.) Language, Ethnicity, and Intergroup Relations, London: Academic Press, pp. 15–57. Gibson, M. (2002) “Dialect levelling in Tunisian Arabic: towards a new spoken standard”, in A. Rouchdy (ed.), pp. 24–40. Giles, H. and Coupland, N. (1991) Language: Contexts and Consequences, Pacific Grove, CA: Brooks/Cole Publishing Company. Holes, C. (1995) “Community, dialect, and urbanization in the Arabic-speaking Middle East”, BSOAS, 58: 270–287. Ibn Khaldoun, Abou Zayd, Abderrahman (2003) Kitab al-∞ibar, wa dCwAn al-mubtada§ wa al-khabar, fC ayyAm al-∞arab wa al-∞ajam wa al-barbar (Livre des exemples instructifs et recueil d’origines et de récits, concernant l’histoire des Arabes, des peuples étrangers et des Berbères), Vol. 3, Histoire des Berbères et des dynasties musulmanes de l’Afrique septentrionale, translated from the Arabic by W. MacGuckin de Slane, Algiers: Berti Editions. Jabeur, M. (1987) “A sociolinguistic study in Tunisia: Rades”, PhD dissertation, University of Reading, UK. Jassem, Z. A. (1993) Impact of Arab-Israeli Wars on Language and Social Change in the Arab World: The Case of Syrian Arabic, Kuala Lumpur: Pustaka Antara. Labov, W. (1972) Sociolinguistic Patterns, Philadelphia: University of Pennsylvania Press.

228

ASPECTS OF KSAR EL KEBIR’S NEO-URBAN VARIETY

—— (1978) Le parler ordinaire. La langue dans les ghettos noirs des Etats-Unis, Paris: Les Editions de Minuit. —— (1994) Principles of Linguistic Change: Internal Factors, Oxford and Cambridge, MA: Blackwell. Messaoudi, L. (2002) “Le parler ancien de Rabat face à l’urbanisation linguistique”, in A. Youssi et al. (eds), pp. 223–233. Rouchdy, A. (ed.) (2002) Language Contact and Language Conflict in Arabic: Variations on a Sociolinguistic Theme, London: RoutledgeCurzon. Trudgill, P. (1974) Sociolinguistics: An Introduction to Language and Society, London: Penguin Books (4th edition, 2000). —— (1983) On Dialect: Social and Geographical Perspectives, Oxford: Blackwell. Vanhove, M. (2002) “Processus de koïnisation dans le dialecte arabe de Yafi’ (Yémen)”, in A. Youssi et al. (eds), pp. 151–159. Youssi, A., Benjelloun, F., Dahbi, M. and Iraqui-Sinaceur, F. (eds) (2002) Aspects of the Dialects of Arabic Today, Rabat: Amapatril.

229

MUNIRA AL-AZRAQI

11 ˇ ˇ THE USE OF KASKASAH/ KASKASAH AND ALTERNATIVE MEANS AMONG EDUCATED URBAN SAUDI SPEAKERS Munira Al-Azraqi

11.1 Introduction The Arabic terms kakkakah and kaskasah refer to the use of /k/ and /s/ as reflexes of the 2nd feminine singular dependent object and possessive pronouns, realized in Standard Arabic as /ki/. The Classical Arabic grammarians, Scbawaih (1395/1982: 199–200), Ibn Jinnc (1957: 399–400) and Ibn Ya™ck (n.d.: 48–49) mention two types of kakkakah: replacement of the 2nd feminine singular attached object/possessive pronoun /ki/ by /k/, as in §a∞YaytiK “I gave you (f.)”; and adding /k/ to /ki/, as in §a∞YaytikiK “I gave you (f.)”. Both replacement and addition occur in pre-pause position. Kaskasah is explained by Scbawaih as adding /s/ to /ki/ as in §a∞Yaytikis “I gave you (f.)” in pre-pause position by some speakers. Scbawaih and Ibn Ya™ck do not mention the replacement of /ki/ by /s/; however, according to Ibn Jinnc, /s/ replaces /ki/ among some speakers as in the case of kakkakah. Scbawaih explains the two phenomena as resulting from a need to distinguish between 2nd person singular male and female in pre-pausal position, since pausal short-vowel deletion would result in a collapse of distinction between /ka/ (2 m.s.) and /ki/ (2 f.s.). This study does not aim to examine the history of kakkakah or kaskasah in the Arabian Peninsula, nor does it deal with the purely phonological phenomenon of affrication of /k/, discussed by previous linguists like Johnstone (1963: 212, 1967: 2). Its focus is on the actual social distribution of kakkakah and kaskasah in five modern Saudi Arabian dialects and on alternative variables used by educated speakers in order to avoid these features. The present-day usage of kakkakah and kaskasah is stated and discussed by many linguists like Holes (1984, 1991), Ingham (1982, 1986), Johnstone (1963, 1967), Prochazka (1988) and Watson (1992).

230

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H

11.2 KaÌkaÌah and kaskasah in modern peninsular dialects Kakkakah is attested in the southwestern part of the Arabian Peninsula. Kaskasah is a dialect feature of Najd, from central Saudi Arabia towards the northern region. In the north, we find in addition to /s/ the affricate /Ê/. Another reflex of the 2nd feminine singular object/possessive pronouns, /n/, is used in the eastern and northeastern regions of the Arabian Peninsula. This realization is not mentioned explicitly by Scbawaih or other Classical grammarians; however, Anis (1990: 122) claims that it is referred to, though incorrectly. He supports his conclusion by appealing to the phonological rule in which a velar consonant in the environment of a front vowel is fronted to a palatal affricate (see also Al-Ghunaim 1985: 256). For the last 20 years or more, one observes that this feature is variable and may be undergoing change throughout Saudi Arabia, particularly among educated speakers, resulting in a single koineized reflex of the 2nd feminine singular object/possessive morpheme with two contextually determined allomorphs: /ki/ in suffixation to a vowel-final word, and /ik/ in suffixation to a consonant-final word. Thus, bBtik “your f.s. house” (cp. bBt¢k “your m.s. house”) and §abEki “your f.s. father” (cp. §abEk “your m.s. father”). I shall refer to this new koine form henceforth as /k/. Although we cannot date precisely the beginning of the shift from kakkakah and kaskasah to /k/, it seems to be a rather rapid phenomenon, leading among some educated speakers to the exclusive use of /k/. This shift could have been motivated by a number of modernizing factors, namely urbanization and education. Saudi Arabia has undergone massive development since King Abdulaziz, the founder of Saudi Arabia, initiated the process of urbanization. His aim was to build a country that could achieve the level of wealth and social opportunity of developed countries throughout the world. Since then, many qualitative and quantitative improvements in the standards of education, training, health and social care have been realized in Saudi Arabia (Al-Salem 1416 ah: 21). As part of the urbanization program, education was made a priority, since modern education was thought to contribute to the citizen’s capabilities to cope with domestic and international challenges. The first administration for education established in 1344 ah/1925 ad in Mekkah preceded the unification of Saudi Arabia in 1345/1926 (Mbmish 1419 ah: 33; Al-Salem 1985: 21). Since then, the education sector has expanded hugely, as shown in Figure 11.1 (MEP 2003: 217), including for girls, in spite of early opposition to it from some parts of society. There is no doubt that educational development had a direct effect on language awareness. One of the goals of education in Saudi Arabia has been to enhance language knowledge and to minimize the gap between Standard Arabic and the spoken dialects (Al-Sunbul et al. 1992: 51). Like in many

231

MUNIRA AL-AZRAQI 2500

Thousand

2000

1500

1000

500

0 1969

1974

1979

1984

1989

1994

1999

2000

2001

2002

1969

1974

1979

1984

1989

1994

1999

2000

2001

2002

Other

19

38

48

68

63

148

167

141

150

200

Adult Edu

43

100

142

159

152

116

110

126

106

95

7

19

48

94

122

170

343

348

380

394

Inter & Sec

77

179

338

531

766

1262

1792

1829

1878

1955

Elem & Pre

401

648

886

1297

1831

2238

2362

2380

2403

2409

Higher Edu

Figure 11.1 Growth of education sector in Saudi Arabia in number of schools

other Arab countries, use of Classical Arabic for educational purposes is regularly insisted on. This enhanced awareness of Standard Arabic has had an influence on people’s perception of their spoken dialects and on their spoken uses, particularly in semi-formal contexts. Thus, several features of the spoken dialects in Saudi Arabia have changed, and others are in the process of change. Therefore, due to this educational policy, a related motivation for the change occurring in Saudi dialects, in general, and the shift from kakkakah and kaskasah, in particular, could be the growing influence of Standard Arabic within the diglossia context (Ferguson 1959). Standard Arabic is the medium of instruction in formal settings, most particularly in educational establishments, while local dialects are spoken in informal settings. In school, children learn that /ki/ is the Classical reflex of the 2nd feminine singular object/possessive pronoun, and that it contrasts with the form they use in their own speech. However, the influence of Standard Arabic cannot be the only reason for explaining this change. The broadening of communication, which has inevitably accompanied urbanization, seems also to have encouraged linguistic change toward a process of dialectal koineization. In a large number of Arabian dialects, including that of Hijaz, in western Saudi Arabia, /k/ is the 232

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H

reflex of the 2nd feminine singular suffix. Non-Saudi dialects may also have influenced language use: a significant proportion of Arab expatriates work in Saudi Arabia as professionals: as teachers, university lecturers, engineers, doctors, i.e. in the types of jobs which involve frequent interaction with local people. In addition, many radio and television series feature non-Saudi Arabic dialects, including Egyptian, Syrian, Jordanian and Lebanese. We hypothesize here that the main motivating factors for linguistic change, therefore, are widespread communication and education (cf. Al-Wer 2002: 52), both of which are related and linked to urbanization. On the same lines, the shift from kakkakah and kaskasah to /k/ can be attributed to both increasing influence of Standard Arabic and increasing process of dialect contact and koineization, both phenomena reinforcing each other.

11.3 The study This study aims to examine (1) variability in the rate of use of /k/ among educated people in Saudi Arabia versus use of kakkakah and kaskasah; and, (2) alternative means to refer to the 2nd person feminine singular other than the suffixal pronoun. As will be clear in Section 11.4.1, and contrary to the use of /k/, alternative means cannot be systematically analyzed as covariants of /k/ or /s/ and could be interpreted as stylistic devices. However, the comparison between +/− use of /k/, +/− use of /k/ or /s/ and +/− use of alternative means among each speaker tends to indicate that indeed these alternative means are a way to avoid use of kaskasah or kakkakah. Dialect speakers from the following five major cities were interviewed. Each city represents a geographical region of Saudi Arabia and has its own culture, and its specific language environment but the dialect of each area features either kaskasah or kakkakah. Map 11.1 shows a map of the kingdom with the location of the cities. 1.

2.

Najd is represented by Riyadh (24.65°N, 46.77°E), the capital of Saudi Arabia. With a population of over 4.1 million (CDS Report 2004: 18), Riyadh is regarded as the most modern, developed and educated city in Saudi Arabia. It is one of the regions that use kaskasah – /s/ or /Ê/ – as the 2.f.s. suffix. The eastern region is represented by Dammam (26.43°N, 50.10°E), the capital of the Eastern Province. This city recently absorbed two other cities, Dhahran and Al-Khobar. With a population of about 1.2 million (CDS Report 2004: 23), it is a cosmopolitan city where people of different nationalities live and work, mainly in the oil industry. The Saudi population of Dammam comes mainly from surrounding areas, such as Al-Qatbf, Al-Ahsa or Jubail. The use of /n/ is common among speakers originating from Dammam. This suffix is also attested in dialects of Bahrain, Iraq, Kuwait, Qatar and UAE. 233

MUNIRA AL-AZRAQI

Map 11.1 The map of the Kingdom of Saudi Arabia showing locations of the five cities of the study

3.

4.

5.

Buraidah (26.37°N, 43.97°E) is the capital of Al-Qassim Province. Until recently, agriculture was the primary means of livelihood. Now the province is home to various types of industry, such as pharmaceutical production, as well as agri-business, particularly poultry rearing. The population is 506,000 (CDS Report 2004: 22). The reflexes of the 2nd feminine singular suffix are /s/ and /Ê/. Abha (18.23°N, 42.50°E) is the capital of Asir Province. It is located in the southwest of the kingdom with a population of 350,000 (CDS Report 2004: 24). Although the city is located in Asir, a significant proportion of its population comes from other nearby tribes, such as the Shahran and Qahtan. The reflex of the 2nd feminine singular suffix in Abha is /k/. Skaka (29.97°N, 40.20°E) is a major city in Al-Jawf oasis. It is close to Jordan in the north, and has a population of around 200,000 (CDS 234

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H

Report 2004: 31). The reflexes of the 2nd feminine singular suffix in Skaka are /s/ and /Ê/. 11.3.1 Methodology This study is a precursor to a larger investigation of Saudi speakers’ attitudes towards kakkakah or kaskasah, which involves an analysis of the rate of use of /k/. The data consist of about 12 hours of tape recordings and direct communication between the researcher or field workers and the informants. Tape recording was often difficult with women; in this case, observing and counting was used as an alternative technique. The material deals with general topics. Recordings were made in universities, schools, offices, libraries, and formal social meetings. 11.3.1.1 The informants The data for this study was obtained from 67 informants of both sexes, aged between 24 and 48. All of them are educated and most have a Bachelor’s, Master’s or PhD degree. Most have high-status jobs, such as lecturers, teachers or doctors. A small proportion of the informants are students and housewives. All informants were selected for their tendency to use /k/ instead of their colloquial forms, at least some of the time. Although they cannot be considered as 100 percent representative of the general population, it may be stated that all of the informants are natives of the places in which they live. Some informants were well known to the researcher. The researcher herself was born and had lived for 20 years in Abha and still visits the city from time to time. She spent five years (between 1988 and 1993) in Riyadh and has friends there. She now lives in Al-Ahsa, 87 miles from Dammam, in the Eastern Province. She visited Buraidah twice and Skaka three times in order to carry out this study. 11.3.1.2 Data analysis The data was analyzed using statistical software. Microsoft Excel 2003 was used for column statistics and standard error determination on the informant level. Graph Pad Prism 2.0 was used for contingency table analysis. 11.3.2 Results The chi-square test shows that the difference in using kakkakah or kaskasah among the informants of this study is highly significant (p < 0.0001), as shown in Figure 11.2. Figures 11.3 to 11.7 show the proportional use of /k/ among the informants of Riyadh, Dammam, Buraidah, Abha, and Skaka respectively. 235

MUNIRA AL-AZRAQI 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 SKAKA

ABHA

BURAIDAH

DAMMAM

RIYADH

/k/%

80.1

56.2

51.7

84.9

92.7

Kash or kas %

11.5

29.6

38.2

11.8

6.7

Other means %

8.4

14.1

10.1

3.3

0.6

Figure 11.2 Difference between regions in using /k/, colloquial form and other means

100.00 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 Ahma d

AlMajid

Saleh

AlHuss

Fawa Nayif

z

Mejalli

Afaf

Alya

Hasn

Samir

a

a

Om Hisah

MoDi

Hala

Latifah

ein /k/%

77.78

76.92

95.24

100.00

82.61

91.30

100.00

kash or kas %

22.22

23.08

0.00

0.00

0.00

0.00

0.00

4.76

0.00

8.70

19.23

0.00

14.29

8.70

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

8.70

0.00

0.00

0.00

0.00

0.00

Other means %

100.00 100.00 100.00 100.00 100.00

80.77 100.00 85.71

Figure 11.3 Percentage of using /k/, colloquial form and other means by informants of Riyadh

Informants from Riyadh showed the highest percentage of use of /k/, at 92.7 percent. As illustrated in Figure 11.3, 8 out of 15 informants did not use kaskasah at all. Although the difference was not statistically significant, men showed a greater tendency to use /k/ than women (5 out of 7 men, compared to 3 out of 8 women: Fisher’s exact p > 0.05). 236

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H 100.00 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 Ahmad

Abdull S

Hasan

fadil

Adnan

Arfaj

Nuha

S

Nourah

Gadah

AlJazi

m

100.00

88.89

100.00

90.00

76.92

73.33

kash or kas %

0.00

11.11

0.00

10.00

23.08

Other means %

0.00

0.00

0.00

0.00

0.00

/k/ %

Mariya

91.67

82.61

26.67

8.33

0.00

0.00

Ameer

Fayzah

ah 18 100.00

Fawziy ah

68.42

100.00

100.00

62.50

54.55

17.39

15.79

0.00

0.00

0.00

25.00

27.27

0.00

15.79

0.00

0.00

0.00

12.50

18.18

Figure 11.4 Percentage of using /k/, colloquial form and other means by informants of Dammam 100.00 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 Aljaffali

Saeedi

Rumaih

Sohaim

Abdurhm

Abdullah

Layla

Huda q

Badriyah

Nawal

Hind

Suad

/k/ %

20.00

66.67

64.29

77.78

9.09

71.43

77.78

57.14

69.57

45.83

44.00

17.39

kash or kas %

80.00

33.33

35.71

22.22

90.91

28.57

5.56

31.43

21.74

33.33

32.00

43.48

0.00

0.00

0.00

0.00

0.00

0.00

16.67

11.43

8.70

20.83

24.00

39.13

Other means %

Figure 11.5 Percentage of using /k/, colloquial form and other means by informants of Buraidah

Figure 11.2 illustrates that informants from Dammam showed the second highest rate of use of /k/, at 84.9 percent. However, as Figure 11.4 shows, the difference between men and women in Dammam is relatively small compared to that between the sexes in Riyadh. The informants from both Riyadh and Dammam showed the lowest rate of use of alternative means of avoiding /k/, as shown in Figure 11.2. Only 237

MUNIRA AL-AZRAQI 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 Mousa

Ahmad

Tariq

hussein

Saeed

Ali

Huda

Mona

Dalal

Nouf

Shareef

Khayriy

ah

ah

Salha

/k/ %

80.00

77.78

62.50

73.33

75.00

83.33

12.50

21.05

52.38

60.00

22.22

58.33

kash or kas %

20.00

22.22

25.00

26.67

25.00

16.67

56.25

47.37

14.29

20.00

50.00

25.00

36.84

Other means %

0.00

0.00

12.50

0.00

0.00

0.00

31.25

31.58

33.33

20.00

27.78

16.67

10.53

52.63

Figure 11.6 Percentage of using /k/, colloquial form and other means by informants of Abha 100.00 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 Bader

Jawad

Mahdi

khaleel

Ab.Maje

Haytha

Mansou

ed

m

r

Majdah

Wedad

Najla

Abeer

Ibtisam

Huda J

/k/%

71.43

88.89

100.00

75.00

55.56

80.00

100.00

55.56

75.00

90.91

80.95

84.21

83.33

kash or kas %

28.57

11.11

0.00

0.00

44.44

0.00

0.00

0.00

25.00

9.09

9.52

5.26

16.67

Other means %

0.00

0.00

0.00

25.00

0.00

20.00

0.00

44.44

0.00

0.00

9.52

10.53

0.00

Figure 11.7 Percentage of using /k/, colloquial form and other means by informants of Skaka

one informant from Riyadh and 3 from Dammam used alternatives to the pronoun suffix, all 4 of whom were women, as shown in Figures 11.3 and 11.4. Figures 11.5 to 11.7 show results for informants from Buraidah, Abha and Skaka respectively. These informants showed lower rates of use of /k/, at 51.7 percent, 56.2 percent and 80.1 percent respectively, than informants 238

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H

from Riyadh and Dammam. They showed greater use of their kakkakah or kaskasah colloquial forms. Figures 11.5 and 11.6 show that none of the informants from Buraidah and Abha used /k/ 100 percent of the time. Figure 11.6 shows that a higher percentage of male informants in Abha used /k/ than female informants, while female informants used alternatives to the 2nd feminine singular suffix more than male informants. Figure 11.2 shows that the highest percentage of informants using alternatives to the 2nd feminine singular suffix came from Abha. Significantly, only one male informant from Abha used alternatives to the 2nd feminine singular suffix (see Figure 11.6). Among Skaka informants, 2 out of 5 male informants used alternatives to the 2nd feminine singular suffix (see Figure 11.7), while use of alternatives to the 2nd feminine singular suffix was restricted to females (see Figure 11.5).

11.4 Discussion Patterns of behavior frequently reflect the attitude or perception of the people involved towards a specific social situation (Fasold 1984: 148). This may include language change. A linguistic variant may spread across a population over a matter of generations. The adoption and frequent use of a novel linguistic form gives the impression of its acceptance, whereas refusing or hesitating to use such a form is indicative of speakers’ uncertainty. This study shows that although an increasing number of people are shifting to /k/, some people still use their original dialectal forms or alternatives to the suffixed pronouns. Shifting to /k/ could be interpreted as a sign of urbanization and social development as indicated by the comparison between the different cities. There is a probability that high social status and education must also play a role but this needs to be confirmed by additional research. The analysis of the results shows that the informants from Riyadh and Dammam have a higher rate of use of /k/ than informants from the other three cities. These two cities enjoy a high level of modernization, despite the nomadic backgrounds of the inhabitants. The results for informants from Dammam are very interesting. Dammam is closer to other Gulf cities than Riyadh, as a result of which the general speech patterns have many more features of Gulf dialects than of Riyadh dialect. This phenomenon raises the question of why Dammam speakers tend to shift to /k/, since speakers from cities in the Gulf states do not apparently exhibit such a shift. If the high rate of use of /k/ in Dammam is due to education policy in Saudi Arabia, then Dammam speakers are likely to have the same attitude towards using /k/ as Riyadh speakers, since they share the same educational system. The low rate of use of /k/ among Buraidah, Abha and Skaka speakers could be attributable to the fact that these cities do not enjoy the same level of modernization as Riyadh and Dammam. Although education started 239

MUNIRA AL-AZRAQI

at about the same time (standard education started in Riyadh in 1934 (1354 ah), in Abha in 1935 (1355 ah), in Buraidah and Dammam in 1936 (1356 ah) and in Skaka in 1942 (1362 ah) (cf. Al-Salem 1991: 54–127) ), it did not have the same impact and moreover, these three cities are smaller and less cosmopolitan than Riyadh and Dammam. As the results of this study show, male informants exhibit a higher rate of use of /k/ than females. This is inconsistent with the pattern found in most communities where women were found to be leaders in case of change in progress from above (Labov 1998: 7). The literature about women and change in the Arabic-speaking communities has indicated that Arab women in urban environment use the prestigious urban colloquial variants more than men (Al-Wer 2002; Haeri 1997). However it has also been noted that Arab women use the Modern Standard Arabic (MSA) variants less than men. MSA and urban variants are two competing prestigious forms, MSA being more restricted to formal public contexts, where women were less often present (Abdel Jawad 1986; Ibrahim 1986; Sadiqi 2004). The implication of such findings in our study indicates that /k/ in the Saudi urban society is associated with formality. The use of /k/ by males was greatest when talking to non-family members and decreased significantly when talking with a family member. This is most probably attributable to the degree of formality between speakers. Interaction between men and women in Saudi Arabia is restricted except within families. This kind of relation reflects a different context of interaction between the sexes. When talking to a nonrelative female, a man maintains a high level of formality, whereas within the family, he may shift between /k/ and his dialect form or simply use the dialect form. The lower proportional use of /k/ among women did not necessary result in greater use of the dialectal form. Women, particularly from Abha, Buraidah and Skaka, tended to use alternatives to the 2nd feminine singular suffix (see below). 11.4.1 Alternatives to the 2nd feminine singular suffix There are two categories of alternatives to the 2nd feminine singular suffix: some informants address the receiver directly by using ™inti, the independent feminine singular subject pronoun. Others use alternatives to the 2nd feminine singular suffix including use of the masculine plural pronoun, omission of the pronoun, formulation of a general question, and change of context. 11.4.1.1 Direct means A number of informants from Abha use an alternative means specific to Abha (Al-Azraqi 1998: 156), namely the 2nd feminine unattached subject

240

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H

pronoun §inti, if they want to avoid using dialectal /k/. Consider the following examples: (i) xalAX gid §anti hinah “that is it, you are already here” instead of gidiK hinah (ii) ma∞ad §anti btiCn marrah θAnjah “you would not come any more” instead of ma∞AdiK btiCn (iii) gAlD li in §anti ru}tC lhA “They told me that you went to her” instead of galD li iniK The next usage was identified in the speech of one of the Dammam informants as well, and is, I think, an attempt to use a structure reminiscent of Standard Arabic in the question, rather than a means of avoiding the 2nd person singular pronoun: (iv) wBn §antC? “where are you?” instead of wBniN fCh? 11.4.1.2 Indirect means 11.4.1.2.1 THE PLURAL PRONOUN

The use of the 2nd plural pronoun in place of the 2nd feminine singular pronoun in a question is one of the means that we found to be shared by speakers from the different cities in this study: (v) kBf }Alkum? “how are you (2.f.s.)?” instead of kBf }AliK/Â/N ? (vi) BK ∞lEmkum? “what is your news?” instead of BK ∞lEmiK/Â/N ? (vii) fBnkum? “where are you?” instead of fBniK/Â/N ? This usage is more common among speakers from Buraidah where it is a dialect feature: (viii) ∞ammukum bijii halÍ}Cn “your (2.p.) uncle will come soon” instead of §ammi (ix) bBtkum grajib min hna “your house is nearby” instead of bBti (x) bitru}En la halkum fi X-XBf ? “will you go to your family in the summer?” instead of ‘ahal These examples suggest that speakers play on the fact that the objects “uncle”, “house” and “family” are not the exclusive property of the female addressee. Where the object is the exclusive property of the addressee, as in the case of “dress”, for example, the plural cannot be used.

241

MUNIRA AL-AZRAQI 11.4.1.2.2 GENERAL QUESTIONS

Another feature which is common among speakers from different cities is the use of general questions. Examples from Abha informants include: (xi) man OBh galamah? “whose pen is this?” instead of OBh galamiK ? “is this your pen?” (xii) man tBh γurfatah? “whose room is this?” instead of tBh γurfatiK ? “is this your room?” The same use was identified in the speech of an informant from Riyadh: (xiii) mCn illC kallam? “who called?” instead of mCn illC kallamÂ? “who called you?” 11.4.1.2.3 OMISSION OF THE SUFFIX

Omission of the 2nd feminine singular object pronoun is found widely among speakers from different cities. Consider the following example from a speaker from Skaka: (xiv) §anA §aKKart bass mA KiftCnC “I waved, but you did not see me” instead of §anA §aKKart li bass mA KiftCnC “I waved to you but . . .” Here, the verb or the noun must be followed by a phrase or a clause. A similar example was noticed in the speech of an informant from Dammam: (xv) §ana nabbaht §akθ ar min marrah “I gave a notice more than once” instead of nabbahtiN (xvi) ma fAd it-tanbCh “the notice was not useful” instead of ma fAdiN . . . 11.4.1.2.4 CHANGING THE CONTEXT

There were some instances in which informants appeared to change the context completely in order to avoid directly addressing the receiver. In the following example, a speaker from Buraidah uses the suffix ah “him” in place of the 2nd feminine singular suffix: (xvii) §ana sabag w giltah §ana mA ∞indC xabar billC tgElCnah “I previously said that I have no idea about what you are talking about” instead of gilt liN . . . 242

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H

The same means was used by an informant from Abha. In the following example, the speaker uses the emphatic particle §inn instead of directly addressing the receiver: (xviii) §ana gid gult §inn NAXir ma biyinfa∞ “I have already said that Nasir will not be useful” instead of gult liK . . . By using the definite article before the noun the attached pronoun must be omitted. Consider the following example recorded from one of the Dammam informants advising her female colleague to have a meeting with her students: (xix) §inti lAzim tgablCn iY-YAlbAt illC iddarrsCnhum “You should have a meeting with the students you teach” instead of §inti lAzim tgablCn YalbAtiN . . . In some cases, the speaker generalizes his/her speech by referring to a third party instead of the receiver. This might be partially due to politeness and a wish to avoid giving direct advice to the receiver. The word il-waÍdah “the one f.” was used by one of the female informants when she was advising a woman about her daughter: (xx) lAzim il-wa}dah tis§al bintahA iOA KAfathA ÂiOa “the woman ought to ask her daughter if she sees her like this” instead of lAzim inni tis§alCn binti iOA KiftChA ÂiOa “you ought to ask your daughter if you see her like this”.

11.5 Conclusion This study has examined the change from the old dialectal features of kakkakah and kaskasah to the dialect-neutral suffix /k/ to indicate 2nd feminine object/possessive pronoun. Our data suggest that people from Riyadh and Dammam have adopted the /k/ form more than people from other cities. We hypothesize at this point that urbanization, modernization, education and acceptance of official language policy account for this difference. In order to establish whether this new dialect-neutral suffix /k/ is perceived as the Standard Arabic form received through education or is a dialectal prestige form that emerged through communication with speakers of different dialects within and without the Arabian Peninsula, we would need to investigate other dialect features that are in the process of change. /k/ is not identical to the Classical form since it lacks the final vowel (except in the case of vowel-final stems where the form is /ki/). Thus, it appears to be similar to the dialectal form that is used by many nonpeninsular dialect speakers. However, if we hypothesize that the /k/ form is the result of greater ease of communication, we still have to ask why the 243

MUNIRA AL-AZRAQI

Gulf countries, which became urbanized before Saudi Arabia, did not adopt this form. This chapter suggests that the combined factors of communication and education have led to the use of a dialect-neutral suffix form among highly educated speakers. Further work involving less educated speakers is needed to establish the degree to which non-/k/ reflexes of the 2nd feminine singular suffix are regarded as non-prestigious in the Kingdom of Saudi Arabia.

References Abdel Jawad, H. R. (1986) “The emergence of an urban dialect in the Jordanian urban centres”, International Journal of the Sociology of Language 61: 53–63. Al-Azraqi, M. (1998) “Aspects of the syntax of the dialect of Abha (South-West Saudi Arabia)”, unpublished thesis, University of Durham, UK. Al-Ghunaim, S. (1985) Al-lahajAt fC al-kitAb li SCbawaih, aXwAtan wa bunyah, Mekkah: Umm Al-Qura. Al-Salem, H. (1985) Al-idArah al-ta∞limiyah fC al-mamlakah al-∞arabiyyah alsa∞Ediyyah, Part 1, Vol. 2. (No Publisher). —— (1991) TArCx al-}arakah al-ta∞limiyah fi al-mamlakah al-∞arabiyyah al-sa∞Ediyyah, Vol. II, Washington D.C.: International Graphics Press. —— (1416 ah) Al-tarbiyah wa al-ta∞lCm al-∞Am fC al-mamlakah al-∞arabiyyah al-sa∞Ediyyah bayn al-siyAsah wa al-naZariyyah wa al-taYbCq, Beltsville, MD: International Graphics. Al-Sunbul, A., Mutwalli, M., Al-Khatcb, M. and Abdul-Jawad, N. (1992) NiZAm alta∞lCm fC al-mamlakah al-∞arabiyyah al-sa∞Ediyyah, Riyadh: Al-Kheraiji. Al-Wer, E. (2002) “Education as a speaker variable”, in A. Rouchdy (ed.) Language Contact and Language Conflict in Arabic, London and New York: RoutledgeCurzon, pp. 41–52. Anis, A. (1990) FC al-lahajAt al-∞arabiyyah, Cairo: Maktabat Al-Anglo Al-Masriyyah. CDS Report (2004) Population and Housing Census for the Year 1425 AH (2003/ 2004), Riyadh: Central Department of Statistics, Ministry of Economy and Planning. Fasold, R. (1984) The Sociolinguistics of Society, Oxford, UK and Cambridge, USA: Blackwell. Ferguson, C. A. (1959) “Diglossia”, Word 15: 325–340. Haeri, N. (1997) The Sociolinguistic Market of Cairo: Gender, Class and Education, London and New York: Kegan Paul International. Holes, C. (1984) Colloquial Arabic of the Gulf and Saudi Arabia, London: Routledge. —— (1991) “Kashkasha and the fronting and affrication of the velar stops revisited: a contribution to the historical phonology of the peninsular Arabic dialects”, in A. Kaye (ed.) Semitic Studies in Honor of Wolf Leslau, Vol. 1, Wiesbaden: Harrassowitz, pp. 652–678. Ibn Jinnc (1957) Al-xaXa§iX, Xan∞at Abi al-Fat} ¡UthmAn Ibn JinnC, ed. Mo]ammad al-Nazar, Vol. 2, Lebanon: Dar Al-Kutub al-¡arabiyyah, pp. 11–12. Ibn Ya™ck (no date) Shar} al-mufaXXal, Vol. 2, Part 9, Beirut: ¡alam Al-Kitab, pp. 1–160.

244

T H E U S E O F K A Sˇ K A Sˇ A H / K A S K A S A H

Ibrahim, M. (1986) “Standard and prestige language: a problem in Arabic sociolinguistics”, Anthropological Linguistics 28: 115–126. Ingham, B. (1982) North East Arabian Dialects, London: Kegan Paul International. —— (1986) “Notes on the dialect of the Al-Murra of eastern and southern Arabia”, Bulletin of the School of Oriental and African Studies 49: 271–291. Johnstone, T. M. (1963) “The affrication of ‘kaf’ and ‘gaf’ in the Arabic dialects of the Arabian Peninsula”, Journal of Semitic Studies 8: 210–226. —— (1967) Eastern Arabian Dialect Studies, London: Oxford University Press. Labov, W. (1998) “The intersection of sex and social class in the course of linguistic change”, in J. Cheshire and P. Trudgill (eds) The Sociolinguistics Reader, Vol. 2: Gender and Discourse, London: Arnold, pp. 1–52. MEP (Ministry of Economy and Planning) (2003) Achievements of the Development Plans: Facts and Figures, 21st Issue, Riyadh: Ministry of Education and Planning. Mbmish, M. (1419 ah) WizArat al-ma§Arif fC §ahd wazCriha al-awal al-amCr Fahad bin AbdullazCz (1373–1380). Makkah Al-Mukarramah: Umm Al-Qura University Press. Prochazka, T. (1988) Saudi Arabian Dialects, London: Kegan Paul International. Sadiqi, F. (2004) “Language and gender in Moroccan urban areas”, paper presented in the 1st International Conference on “Arabic urban vernaculars, the impact of migration and social changes, Aix-en-Provence, 22–23 October. Scbawaih, Abu Bishr ™Amr B. ¡Uthman (1395 ah/1982 ad) KitAb SCbawaih, ed. Abdul-Salam Mo]ammad Haren, Vol. 4, Cairo: Al-Xanzc. Watson, J. C. E. (1992) “Kakkakah with reference to modern Yemeni dialects”, Zeitschrift für Arabische Linguistik 24: 60–81.

245

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Part III MULTILINGUALISM, CODESWITCHING AND NEW URBAN CULTURES

247

JONATHAN OWENS

248

CLOSE ENCOUNTERS OF A DIFFERENT KIND

12 CLOSE ENCOUNTERS OF A DIFFERENT KIND Two types of insertion in Nigerian Arabic codeswitching Jonathan Owens

12.1 The linguistic problem This chapter is a study of codeswitching (CS) among Nigerian Arabs in the city of Maiduguri in northeastern Nigeria. As a minority linguistic group in a complex linguistic society, Nigerian Arabs in Maiduguri present an interesting analytical problem. Integrating a number of languages into their repertoire, a basic question is to determine the extent to which the languages are treated equally or differently, and if differently, how? One focus of the chapter is thus descriptive. Five linguistic issues are isolated, in which the integration of English or Standard Arabic (SA) lexical material into Nigerian Arabic (NA) matrices are compared. While the linguistic issues themselves represent standard grammatical constructions, the comparison between English and SA is effected statistically, with a monolingual NA database serving as the basis of comparison. The chapter assumes the relevance of current issues in CS, the borrowing versus CS debate, in particular, and the usefulness of recognizing a matrix language (e.g. Poplack and Meecham 1998; Myers Scotton 1997, 2002). The chapter, however, does not contribute directly to this debate and focuses rather on the role of processing constraints on lexical insertions. The chapter is divided into seven sections. Section 12.2 is a brief historical and ethno-linguistic background summary, Section 12.3 describes the sample used, and Section 12.4 describes the corpus in greater detail, with special emphasis on its relevance to a complex linguistic situation. Section 12.5 presents the five grammatical issues, summarizing their salient points and illustrating them with textual examples. Section 12.6 presents the statistical results of the study, contrasting English and SA insertions in NA matrices. Finally, Section 12.7 highlights five explanatory issues.

249

JONATHAN OWENS

12.2 Historical and ethnic background Arabs have lived in the Lake Chad area since the fourteenth century and in territorial Nigeria since the sixteenth or seventeenth centuries. Their immediate point of dispersal was Upper Egypt (Owens 1993b), a nomadic lifestyle allowing them to spread quickly from Egypt to Lake Chad. Until today they are largely a rural people, practicing cattle nomadism and sedentary farming, or a mixture of the two. In the last 30 years they have increasingly settled in urban areas, however, Maiduguri above all. Today they constitute one of the larger minorities in Bornu as a whole, as well as in Maiduguri, where they make up perhaps 10 percent of the total population. It is in urban areas in particular that they are confronted by a vibrant multilingualism, the ways in which they accommodate to and master it being a main descriptive goal of the chapter. Maiduguri is a city of approximately 500,000 inhabitants in northeastern Nigeria, the capital of Bornu State. Ethnically and linguistically it is very heterogeneous, though it is dominated by a Kanuri minority who historically have been the main political force in the region, and who constitute the largest linguistic group. Hausa has become the main lingua franca of the city over the past 30 years, and English is a second pan-ethnic language widespread in the population.

12.3 The sample The sample consists of two parts, both collections of recorded and transcribed texts. The larger one includes ten separate free conversations in which a good deal of CS occurs. These are called the CS texts and comprise about 84,000 words and have 26 active participants, all native speakers of NA. The topics ranged across local and national politics both past and present, neighborhood matters, customs, education and daily life. A smaller collection of six texts is used as a control. It consists of about 25,000 words with 17 active participants. These are texts recorded for a previous research project (Owens 1993a, 1998), encompassing five texts from Maiduguri and one from a village about 40 miles southeast of Maiduguri. They include interviews and unstructured conversations. They are termed the monolingual corpus, as they have no constituent switching in them, and the number of L2 lexical insertions is small, only 186 or less than 1 percent of the total. This monolingual corpus is used as the norm against which the codeswitched material is measured. That is, material in the CS corpus is not considered to conform to or deviate from NA on the basis of a normative NA grammar, but rather on the degree to which statistical patterns in one corpus correspond to or deviate from statistical patterns in the other. Five participants in the CS texts are also prominent participants in the monolingual corpus, so the two samples are not completely independent.

250

CLOSE ENCOUNTERS OF A DIFFERENT KIND

The two corpora were coded extensively in terms of various parameters including grammatical context, language identity of switched items, pausal position, speaker and length of insertions. However, they were coded to different degrees. In the CS texts, non-switch material was not included in the coding. In the monolingual corpus, about 75–80 percent of the monolingual texts were comprehensively coded, so far as the grammatical parsing categories themselves are comprehensive. The coded material was stored in MS Access, which was the basic program used for building the data bank.1 Statistical tests are carried out with SPSS. The criterion for normative NA behavior used here is different from that used in other work on CS, notably by Sankoff and Poplack (Sankoff et al. 1990). They have compared patterns of inserted items in CS texts with stretches of speech in the CS texts where no CS occurs. To this point, my basic perspective is that ideally one would want to use various definitions of “normative” and work through the entailments of choosing one or the other as the basis of comparison with CS material. Unfortunately, this is not a luxury available to most researchers, as it demands a great deal of time to adequately define even a single corpus type. A problem specific to the current CS corpus is that there are a number of texts where the CS is so dense that it would be difficult to even isolate large stretches of speech to define as non-CS.

12.4 The corpus: an example of a complex linguistic situation The present CS corpus is of special interest to CS studies in that it exemplifies what I term a complex linguistic situation. Linguistically this is characterized by a community of speakers who are fluent in at least two languages, and in addition have a degree of competence in other languages as well. Socially a complex linguistic situation arises where a linguistic minority lives in the presence of widespread lingua francas, the minority besides their own language having to speak the other lingua francas as well. Many parts of Africa are characterized in this way, with minority linguistic groups living in the presence of either complementary or competing lingua francas (Owens 2004). Maiduguri is a good example of this situation. Well over 40 languages are spoken natively, while there are three lingua francas. Up until World War II, Kanuri was the main lingua franca of the city; thereafter Hausa has spread enormously and is now the primary lingua franca, though it has not completely displaced Kanuri (Bross 2002). Besides these two competing lingua francas, English, as the de facto national language and language of instruction, is a further lingua franca, even if competence in it is more confined than that of Hausa (Myers-Scotton 1993). A further language is Standard Arabic (SA). SA has increased in importance since the 1980s, but does not have the status of a lingua franca, being relatively 251

JONATHAN OWENS

Table 12.1 Continuation in same language between speakers’ turns Continue

Number

%

Arabic Hausa English Kanuri Standard Arabic Total

4,076 1,961 32 9 0 6,078

67 32.4 0.5 0.1 0

Table 12.2 Lexical insertions Matrix language

Nigerian Arabic Hausa

Embedded language English

Hausa

SA

1,121 484

139

633 13

confined both in terms of population of speakers, and domains of use (see Sections 12.7.3 and 12.7.6). A crucial aspect of a complex linguistic situation is that groups of speakers use a number of languages. In Maiduguri, all minority linguistic groups will typically have their own language, plus Hausa, English to one degree or another, and perhaps some Kanuri. In addition, segments of the Nigerian Arab population in Maiduguri will have a good knowledge of SA. In the present sample, four languages figure prominently: NA, Hausa, English and SA. While many of the speakers in the sample do know Kanuri, it figures only negligibly in this corpus. With this sociolinguistic background in mind, it becomes a linguistic task to distinguish between discourse and lexical resource languages. Discourse languages are those in which large stretches of discourse are carried out, while lexical source languages are important in providing lexical insertions into matrices formed by the discourse language. In the CS corpus, NA and Hausa are the discourse languages, SA and English the lexical resource languages. Various quantitative indices define this categorization and two are selected here. Table 12.1 presents the statistics for which matrix language is maintained across a speaker switch. That is, Speaker A ends in say NA, while Speaker B continues the matrix language in NA. The only two languages which are consistently maintained as matrix languages across speakers’ turns are NA and Hausa. Turning to the lexical languages, Table 12.2 gives the number of lexical insertions for English, SA and Hausa in matrices formed by Hausa and NA. 252

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Here, English is heavily used as a lexical donor in both Hausa and NA matrices, SA is a heavy lexical provider in NA matrices and Hausa has a dual role. Besides being a discourse language, Hausa is also somewhat a lexical provider in NA matrices. The focus of this study is on the differences between English and SA, as lexical donor languages, in a NA matrix language, since it is only NA these two languages are embedded in. The categories considered are lexical insertions of nouns, adjectives and verbs.

12.5 The data The data for the comparison of English and SA comes from five linguistic domains. At the beginning of each section the normal form or distribution of the category in question will be summarized, after which the “deviant” behavior of the inserted items will be treated. 12.5.1 Definite article alNA, like most varieties of Arabic, formally has a bimodal system for marking definiteness. An indefinite noun is unmarked while a definite noun is marked by the definite article prefix al-, the /l/ assimilated to a following dental or alveolar consonant: bBt “a house” versus al-bBt “the house” The basic function of what is termed an indefinite noun is to introduce non-identified nouns (first mention). The three most important functions of what is termed a definite noun are to mark previously identified nouns, and to mark unique and generic reference. It may happen, however, that sometimes an expected definite article does not appear and what I term a Ø form occurs. Though formally identical to the indefinite form, such forms may be considered to lack an expected definite article. In each of the following two examples, the nouns would be expected to have a definite article on the basis of the unique referent rule. (1) SA insertion mdled fi Ø-∫umal (expected, ∫-∫umAl) born in north “born in the north” (2) English insertion hu bas Ø-kampbn masta hana abAsanjo (expected, al-kampbn) he just campaign manager of Obasanjo “he is (the) campaign manager of Obasanjo” 253

JONATHAN OWENS

12.5.2 Plural morphemes Inserted items may occur in contexts where a plural noun is required. There are only two instances where a SA noun in a plural context does not occur with an expected plural morpheme suffix and therefore SA nouns will be not be treated here. English inserted nouns may appear in three guises. They may be marked by the English plural morpheme -s (with different morphophonological realization), by a suffix from NA, or by Ø. The three cases are illustrated below. (3) English plural -s al-mcn-z hinB-k ART-means hana.PL-your.M “your means” (4) NA plural stbt-At wAd-At bad-o state-PL.F some-PL.F begin-PL.M “some states have begun to pay” (5) Ø stbt-Ø wAd-At qassad-an state- some-PL.F agreed-PL.F “some states agreed”

b-ikaff-an 3-pay-PL.F

Note that the plurality of the noun is confirmed by the F.PL. agreement on the modifier and verb. 12.5.3 Possessive I, idÑfa, Possessive II, hana There are two possessive constructions in NA. In one, termed the idAfa in traditional Arabic grammar, the possessed noun is placed immediately before the possessor. In this construction the possessed noun is severely restricted formally and positionally. It may not bear the definite article prefix and it must occur immediately before the possessor. This means that its definiteness is determined by the definiteness of the possessor noun (if possessor is definite, so is possessed (6b, c), if possessor is indefinite (6a), so is possessed), and any modifiers (e.g. adjectives) must be postposed behind the possessor. A possessor pronoun is suffixed directly to the head noun (6c). (6a) bBt rAjil house man “a house of a man” (6b) bBt ar-rAjil “the house of the man” (6c) bBt-na “our house” 254

CLOSE ENCOUNTERS OF A DIFFERENT KIND

In the other possessive construction, termed possessive II, the possessed is linked to the possessor by an independent morpheme which agrees in terms of number and gender with the head noun, hana (F hCl, PL hine). In this construction, the possessed noun has all the normal properties of an independent noun phrase (NP), e.g. it can be marked by a definite article, modified directly by adjective, etc. A possessive pronoun is attached in this construction to this linker element, not to the head noun. (6d) bBt hana rAjil house of man “a house of a man” (6e) al-bBt hana ar-rAjil “the house of the man” (6f ) al-bBt hanA-na “the house of ours” Insertions in possessor and possessed position are sensitive to three prominent factors. These are whether the insertion occurs at possessor position or at possessed position, and thirdly, whether the possessor is noun or pronoun. These three factors will be considered in each of the two possessive constructions. In these two constructions there is no “incorrect” usage. Rather, as will be seen, the relevant factor is the distribution of the inserted nouns relative to the factors, possessed position, possessor position and noun/pronoun possessor. The following exemplifies SA and English-inserted noun in the various constructions. 12.5.3.1 idafa (7) SA in possessed position tanfcz al-pdlizi implementation ART-policy “implementation of the policy” (8) SA, possessed by pronoun misaham-it-um contribution-F-their.M “their contribution” (9) English, possessed by pronoun yunifdm-ak uniform-your.M “your uniform” (10) English, possessor noun fi Cd al-gaman in hand ART-government “in the hands of the government” 255

JONATHAN OWENS

12.5.3.2 Possessor II, hana (hcl, hine) (11) SA possessed noun yAni unsiriyye hCl abbah-At-um mAfi to say chauvinism hana.F father-PL-their not “which is to say there isn’t the (ethnic) chauvinism/racism of their ancestors” (12) SA possessor min al-fandamental-z hine l-bahas from DEF-fundamental-PL of.PL DEF-research “(this is on the basis) of the fundamentals of research” (13) English, possessed noun, possessed by pronoun saykula hCl-hum di circular hana.F-their this.F “this circular of theirs” 12.5.4 Descriptive adjective In NA the descriptive adjective follows the noun it modifies as in SA. (14) SA da rAjil mujtahid mil§allim this man hard working learned “this is a hard working, learned man” Moreover, in all tokens SA descriptive adjectives basically follow NA agreement rules of definiteness, number and gender agreement. (15) humma nAs ijtimA§-Cn they.M people social-M.PL “they are a social people” In English, the adjective occurs pre-noun, which creates a sequential mismatch with the position of an adjective in NA. In the present corpus there is only one example of an English adjective modifying an NA noun. Although for the statistical basis of this comparison this number can be ignored, it is interesting to see how the mismatch is handled. (16) fi nAs indu-hum, mu∫kila akademik, akademik mu∫kila “there are people who have, a problem academic, an academic problem” The speaker appears to signal his recognition of the problem of using an English adjective with an Arabic noun in the pause after the predicate induhum. Normally there is no pause in this position. He then first tries 256

CLOSE ENCOUNTERS OF A DIFFERENT KIND

the Arabic solution, postposing the adjective to the noun, pauses, then tries the English solution, preposing the adjective to the noun. It is clear that the speaker is well aware of the normative rules of adjective placement in both English and Arabic, the example demonstrating the inherent contradiction of using an English descriptive adjective with an Arabic noun. In two other tokens NA rules of adjective placement are applied to English Adj + N collocations, as in (17) projekt independent t-isow project independent you-do “you do an independent project” For the most part, however, what in English would be identified as descriptive adjectives occur only in English Adj + N collocations, as in (18). (18) fi nomadik edikb∫in da “in the nomadic education” These are not considered independent adjective insertions as the adjectives in this construction have no distribution independent of the collocations they occur in. As with the possessor, the issue here is not whether an SA adjective is morphologically marked correctly according to the rules of NA, but rather the degree to which adjective SA insertions are used. 12.5.5 Verb The final category is that of the verb, or what might be thought of as verbal concepts. SA verbs are usually integrated into NA morphological forms. For instance, in (19) the lexical stem antaxab is of SA provenance. The /t/ is an infix which marks the so-called eighth derived form of verbs in SA. In NA this derived class does not exist. The inflectional elements are provided by NA. The 3M.PL suffix is -o, rather than SA -u and even the so-called preformative element a- at the beginning of the word derives from NA (cf. the derived seventh form, which is found in NA, with the initial -a, as in ankatab “be written”, cf. SA inkatab). (19) SA antaxab-o vote-3M.PL “they voted” English insertions, on the other hand, rarely “arabicize” the English verbs. Rather a construction widely attested in CS situations is used in which a 257

JONATHAN OWENS

dummy verb “do”, sawwa is inserted with the English “verbal” concept represented by a noun (20a) or gerund (20b) inserted as the object of sawwa. (20a) English i-soww-u vdt 3-do-PL vote “they vote” (20b) bisow-u kriybtin hana ndmadik edikb∫in da “they’ll create a nomadic education (program)” In passing it can be noted that four tokens of the “do” + noun strategy are found with SA nouns, as in (21) bisaww-u intixap hana baba ahmat dDl “they’ll elect Baba Ahmad and the like”

da

These represent only a small percentage of all SA verbal concepts, however (4 of 71 tokens).

12.6 Results The five grammatical constructions will be interpreted according either to their degree of incorrectness (1 and 2), or to their degree of occurrences (3, 4 and 5). A baseline of comparison is furnished by the monolingual corpus. Insertions from English and SA are judged to be more or less integrated in the NA matrices according to their statistically-measured difference from the monolingual database. The statistics are presented in tables with observed frequencies in the top line of each cell, expected according to random distribution on the bottom. This mode of presenting data is used, as the reader can immediately ascertain the degree to which a given cell deviates from a randomly given distribution. Significance levels are also given. In some instances, SA is nearly identical to NA, so there is no point in presenting statistics comparing the two (e.g. 99 percent with 100 percent). Otherwise, a full statistical comparison is given. Although Hausa is included in the statistics, it will not be discussed in detail, because, as noted, lexical insertion of Hausa words is considerably less than SA or English. Note that proper nouns (e.g. names of persons, places, institutions) are excluded from the statistics. For each construction I limit myself to what I consider to be the most important aspects of the respective distributions, without discussing all details. Explanatory, generalizing summaries are left until the next section. In general the relations of the three varieties to the monolingual corpus can be summarized as follows: SA mimics the monolingual corpus to the highest degree, while English is the most distinct, with Hausa occupying 258

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Table 12.3 Marking of definiteness on nouns: definite, indefinite, incorrect Language

English Hausa SA Mono Total

Specific

Count Expected Count Expected Count Expected Count Expected Count Expected

count count count count count

Total

Definite

Indefinite

Incorrect

387 497.4 45 52 242 187.1 1,575 1,512.5 2,249 2,249

320 374.2 36 39.1 78 140.8 1,258 1,137.9 1,692 1,692

250 85.4 19 8.9 40 32.1 77 259.6 386 386

957 957 100 100 360 360 2,910 2,910 4,327 4,327

chi sq = 552, df = 6, p < .000

a position between these two languages, though on the whole closer to English than to SA. 12.6.1 Degree to which incorrect Ø definite article marking occurs on inserted SA/English nouns Table 12.3 is to be read as follows. The rows labeled “English, Hausa, SA” represent the number of tokens these three languages are inserted into NA matrices in the CS texts. The row labeled “Mono” represents the number of tokens in the monolingual texts. The column labeled “Definite” gives the number of tokens, for each language category, that a noun is marked by the definite article, that marked “Indefinite”, the number of tokens it correctly occurs in indefinite form, while the column marked “Incorrect” shows how many noun tokens occur in a context where the definite article is expected, but does not occur (as in (1) and (2) above). The comparison of observed and expected allows one to see at a glance whether the “incorrect” marking is over- or underrepresented for a given language. It is apparent that English insertions in the CS texts and the monolingual texts are most distinct. Note that even in the monolingual texts “incorrect” definite marking occurs. However, this represents only a small percentage of the total of all definite/indefinite marking, about 3 percent. By contrast, over 25 percent of English insertions in the CS corpus occur lacking an expected definite article. For English and Hausa the incorrect category is well overrepresented, whereas for SA and the monolingual corpus it is underrepresented, by a small degree in SA, by a large one in the monolingual corpus. A further question is the extent to which SA is distinct from the monolingual corpus. A chi square comparison of these two languages 259

JONATHAN OWENS

Table 12.4 hana possessor Language

English Hausa SA Mono Total

Specific

Count Expected Count Expected Count Expected Count Expected Count Expected

count count count count count

Total

Possessed by pronoun

Possessed by noun

Possessor

82 24.7 8 3.8 0 5 6 62.4 96 96

80 103.6 14 16 26 21.1 282 261.3 402 402

64 97.7 13 15.1 20 19.9 282 246.3 379 379

226 226 35 35 46 46 570 570 877 877

alone returns a significant difference. However, the difference is far smaller than that of the distribution of all the languages (Table 12.3), with a chi square value of 32 (p < .000). 12.6.2 Degree to which incorrect Ø plural marking occurs on inserted SA/English nouns The occurrence of plural inflectional morphology is analogous to the definiteness problem, as nouns in plural contexts can either be marked “correctly” for plurality, or, they may “incorrectly” have a Ø marking. It yields a result similar to that above, except that the contrasts between English and SA, relative to the monolingual corpus, is even more pronounced. In this instance, all but two SA nouns have plural inflection as expected in plural contexts2 and the monolingual corpus is categorically “correct”. Englishinserted nouns, on the other hand, are correctly marked for plural, either with English -s (78 tokens) or with a NA suffix (22 tokens), (like in (3) and (4) above), but incorrectly lack a suffix in 44 tokens (as in (5)). Thus, in this case, English is again maximally differentiated from the monolingual corpus, while SA rather tracks the monolingual norm. 12.6.3 Degree to which SA/English nouns inserted at possessed/ possessor position and bear possessor suffix pronoun The next category involves the two different possessive constructions (I and II). I will start with the hana construction as it yields a result which combines elements of the previous two constructions. In Table 12.4, the columns are divided according to whether the inserted noun is a possessed noun, this 260

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Table 12.5 hana possessor, SA versus monolingual only Type

CS Mono Total

Specific

Count Expected count Count Expected count Count Expected count

Total

Possessed by pronoun

Possessed by noun

Possessor

0 4 6 5.6 6 6

26 23 282 285 308 308

20 22.6 282 279.4 302 302

46 46 570 570 616 616

chi sq = 1.2, df = 2, p < .54

in turn differentiated according to whether it is possessed by a pronoun, a noun, or whether the inserted noun is a possessor. The distribution shows a highly significant association between the two factors. Globally the distributions replicate Table 12.3 for definiteness in that the distributional extremes are found between English and the monolingual texts. Hausa in this case is somewhat closer to SA than is the case for definiteness, while SA is even closer to the monolingual distribution. This can be seen in a comparison between SA and the monolingual texts only, which yields no significant association between the two variables (Table 12.5). In this instance although SA and the monolingual texts are not categorically identical in respect of insertion in the possessed and possessor position, as they are in respect of plural marking, a pairwise statistical comparison shows that the difference between them is not significantly different. The most striking difference between English (and to a lesser degree Hausa) on the one hand, and SA and the monolingual texts on the other, is the distribution of the possessed pronoun. Effectively, a possessed pronoun is categorically absent from the hana morpheme in SA and the monolingual texts, whereas it is the favored method of introducing a pronoun with English insertions (as in (4) and (13) above). The hana possessor has to be read in conjunction with and contrasted to the idAfa possessor (Table 12.6). What is striking is the categorical absence of any inserted English (or Hausa) noun in possessor position, when the possessor is a noun. SA, on the other hand, has a good number of nouns inserted in possessed position (e.g. (7)), even if fewer than expected by chance. A second aspect of the data, compared to the hana possessor, is the relatively lower number of English insertions in possessor position when the possessor is a pronoun. All in all, English nouns are much less likely to appear in possessed position in the idAfa possessive construction than in the hana possessor (Owens 2002). As for the previous constructions, differences between SA and the monolingual distributions are far lower than between 261

JONATHAN OWENS

Table 12.6 idAfa possessor Language

English Hausa SA Mono Total

Specific

Count Expected Count Expected Count Expected Count Expected Count Expected

count count count count count

Total

Possessed by pronoun

Possessed by noun

Possessor

19 10.6 7 3.5 66 31.3 334 380.5 426 426

0 9.6 0 3.2 19 28.2 364 342.1 383 383

20 18.8 6 6.3 30 55.5 698 673.4 754 754

39 39 13 13 115 115 1,396 1,396 1,563 1,563

Table 12.7 Adjectives Type

SA Mono Total

Specific

Count Expected count Count Expected count Count Expected count

Total

Total nouns

Descriptive adjective

427 415 2,865 2,877 3,292 3,292

15 27 199 187 214 214

442 442 3,064 3,064 3,506 3,506

Df = 1, chi sq = 6.5, p < .01

English and monolingual. However, a pairwise comparison between the two still returns a statistically significant difference. 12.6.4 Degree to which SA/English adjectives occur as descriptive adjectives The fourth category is that of descriptive adjectives. As noted in the discussion in Section 12.5.4, English adjectives do not figure here, as they are all but categorically absent from a modifying position when the head noun is NA. It is nonetheless interesting to see the extent to which SA adjectives mimic the occurrence of descriptive adjectives in the monolingual corpus. Table 12.7 gives the distribution based on the following count. The number of descriptive adjectives, i.e. those modifying a noun, are compared to the total count of all nouns in the monolingual texts, or the total of all SA noun 262

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Table 12.8 Monolingual norm tracking

1. 2. 3. 4. 5. 6.

definite article plural idAfa possessor hana possessor adjective verbal concepts

SA

English

+ − + − +/− −

+ + + + + +

+ = significant difference, − = non-significant

insertions in the CS texts.3 The difference between the two is significant, though the degree of difference is smaller than the difference between the SA and monolingual categories in respect of the idAfa possessive. 12.6.5 Degree to which SA/English verbal concepts occur as morphological verbs or in sawwa + N The final category pertains to verbal expressions. SA verbal patterns are nearly always (N = 68) integrated into NA morphological matrices rather than using the periphrastic sawwa “do” + verbal noun construction. This integration occurs in cases where the SA lexeme corresponds to a morphological pattern which is also attested in NA, a form like ba-sabbit-a “I confirm it”, where the SA derived form 2 verb and the NA derived form 2 verb share the imperfect pattern CaCCiC-. It occurs equally where the SA lexeme introduces a morphological pattern non-existent in NA, an example like the form 8 verb antaxab in (19) above, a derived verbal pattern absent from native NA. Among the 68 tokens, 31 occur in morphological verbal forms not attested in native NA (derived forms 5, 6, 8 and 10). There are only four tokens of SA verbal concepts using the “do” construction (see (21) ). Turning to the English, there is only one English-origin verb morphologized according to NA grammar, the verb cenja “to transfer, change”. This is probably originally from Hausa and comes from military terminology, “to transfer”. Otherwise all English verbal concepts use the verb sawwa “do” + NP or gerund, as in (20) above. In respect of verbal concepts, English and the monolingual corpus are maximally different, while SA and the monolingual corpus are virtually identical. In summary, Table 12.8 gives an overview of which categories are distinct from the monolingual norm, for each language. English contrasts in all respects, whereas SA tracks the monolingual distributions in three categories, and is distinct from it in three. Even if distinct, however, it is so to a far lower degree than are the English insertions. 263

JONATHAN OWENS

12.7 Explanations and discussion In this section, I offer various perspectives and interpretations of the data which has been illustrated and summarized. It is not my purpose to reduce the patterns to a single common denominator or to argue that they support one theoretical perspective or another, even if I do emphasize one salient phenomenon in Section 12.7.1. The first five subsections deal with specific linguistic questions, while the last (12.7.6) takes a step back to orientate the data in terms of the macro-sociolinguistic status of the type of CS observed among Nigerian Arabs. 12.7.1 Processing problems, morphology and English The most consistently striking aspect of the data is perhaps the nonconformity of the English insertions to NA grammar. Here I believe there is a straightforward common denominator informing much of this “deviant” behavior. This is the difficulty of inflectional morphological parsing, coupled with slower processing associated with lower-frequency lexical items. This perspective has been spelled out in greater detail in Owens (2005), so I will only summarize it here. A consistent finding of psycholinguistic processing studies is that reaction time latencies are involved in inflectional morphological processing. It takes longer to access word + inflectional affix than it does whole, noninflected words. Most such studies involve single languages (Bybee 1995: 231; Caramazza et al. 1988; Laine et al. 1999; Levelt et al. 1999: 13; Portin and Laine 2001), though I assume here that this result can be extended to bilingual explanations as well. An even more fundamental assumption in psycholinguistics is that speed of lexical access is in general inversely proportional to lexeme frequency (e.g. Bybee 1995: 232; Stemberger 1995: 258). In the present corpus L2s (English, SA, Hausa) have a lower degree of frequency than does native NA, i.e. of the frequency criterion may be applied across languages. For instance, the nouns in the monolingual corpus have a type/token ratio of about 20 percent versus a 50 percent ratio for the English insertions in the CS corpus. These assumptions and observations together can be taken to explain a number of morphological and distributional aspects of the English insertions. The relative lack of the definite article on English-inserted nouns can be taken as prototypical. Table 12.9 Monolingual versus English insertions in CS texts, type/token ratios for definite/indefinite nouns Monolingual corpus

English insertions, CS corpus

529/2,825 = 19%

452/926 = 49%

264

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Access to definite-marked English nouns should be doubly slow: the necessity of accessing the definite article obtains for any definite noun, NA or English, while the lower frequency of the English noun constitutes a further inhibiting factor.4 These two factors would work in the following way. Assuming a standard rate of speech when using NA matrices, speakers may in the milliseconds of time available take more time in accessing an English noun in a context demanding definite marking. Faced between calling up the English lexeme or the definite article, it is obviously the content morpheme, the English lexeme, which would be given precedence. The definite article may be left off, resulting in the Ø form. Most of the other differences between the English insertions on the one hand and the monolingual corpus on the other are explicable, at least in part, along these lines. The lack of the plural morpheme follows the same explanation as the definite article. For the possessive constructions, a number of factors are involved, which I have discussed in detail elsewhere (Owens 2002), including consideration of type/token frequencies, independence of the inserted noun, and morphophonological difficulties associated with possessive pronoun suffixation. For instance, the preference of suffixing a possessive pronoun on hana as opposed to the English possessed noun (see (9) vs. (13) ), would follow the same principle of avoiding morphological complexity on the English noun. Given a choice of suffixing the possessive pronoun on the NA morpheme hana or on the English possessed noun, hana is the faster alternative. The lack of morphologized English verbs, in favor of the sawwa + N strategy, further reflects the relative ease of attaching “complicated” inflectional material to a native NA verb. One distribution for which the current explanation may play a role is the absence of English nouns in possessed position, when the possessor is a noun (see Table 12.4). As seen above, the possessed position of the idAfa construction is grammatically constrained (e.g. cannot be marked by definite article, must immediately precede the possessed). In these constraints it is akin to a N + suffix construction. In fact, a feminine possessed noun takes the same special morphophonological form (tA§ marbEıa) whether the possessor is a pronoun or a noun. Morphology-like constraints may thus be involved here as well. All in all, for every issue except one, considerations of frequency and difficulty of morphological processing go a considerable way in explaining why English insertions in the data diverge so markedly from the monolingual corpus. Key points are as follows: Issue Issue Issue Issue Issue

1: Lack of definite article al2: Lack of plural morpheme 3: hana possessor over idAfa 3:4 Lack of possessed N in idAfa possessor 4: Lack of pronoun on possessed N 265

JONATHAN OWENS

Table 12.10 Correspondence between some SA and NA phonemes SA

NA

] ¡ y 2 θ

h (categorical) ™ (categorical) Ú (categorical) \ (categorical) t s (i.e. t + s categorical) d q (SA) g γ (SA) q

o q γ

Historically O usually was realized as | in the Western Sudanic area, as in |ahab “gold” < Oahab). This correspondence has since been lost.

Issue 5: Lack of inserted English adjectives Issue 6: Lack of verbs, use of periphrastic sawwa + NP The only issue where morphology does not play an obvious role is issue 5. While sequential mismatch, discussed in (16) above, is likely a key factor here, I cannot at this point back this explanation up by reference to general psycholinguistic findings. 12.7.2 Structural congruity: SA and NA SA and NA have much in common, and it is a moot definitional point whether one is dealing with two languages or with varieties of a single language. From the outset NA and SA share a broad core of congruent (Muysken 2000) phonological and morphological structure which facilitates adapting one to the other. For the present data, it is not only what is structurally comparable, but also how the structural correspondences are treated that are important. To limit the discussion to the phonemes of each language, on a structural basis there are a number of differences, as Table 12.10 indicates. What is striking is that words of SA origin are mostly fully integrated into the NA sound system. In the first four correspondences, the NA realization is used exclusively for SA }, ∞, Y and 1. θ also never occurs, only the realization t or s. For the remaining correspondences the most common realization is the NA variant. Note that only two phonemes even appear in their “original” SA form (q, γ), and as noted, even with these the NA realization is the more common. 266

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Table 12.11 Total SA words, per CS text Text

SA

% of total text/corpus

2 3 4 5 6 7 9 10 11 12 Total

0 271 3 150 7 256 65 57 63 36 908

0 3% 0 2% 0 2% 1% 1% 1% 0 1%

In effect, what is imported from SA is a basic lexical form and meaning, with inflectional and much of the phonological material adapted from NA. The high degree of morphological adaptation seen in issues 2, 4 and 6 is simply a reflex of this overall conformity to NA phonological and morphological patterns. The less than perfect integration on the other three issues is a reminder that SA not only structurally, but also in terms of the statistical measures used here, is in some sense an independent system, beyond the scope of this chapter. Issue 1 is perhaps most interesting in this respect, given the processing-based explanation for the absence of the definite article described in Section 12.5.2. SA words are not inherently licensed to full integration in NA, but can be subject to the same frequency-based effects as are words from other languages. 12.7.3 Education and language competence Structural and psycholinguistic factors are, however, not the only ones at work in the present corpus. I will mention two further in this and the next section. As in other CS situations (see e.g. Bentahila and Davies 1998; Finlayson et al. 1998), education and language competence play an important role. This is particularly in evidence in the SA material. Effectively, only those who have studied SA use it to any large degree in the CS corpus. This can be seen in Table 12.11 summarizing the number of SA words in each of the ten texts. The table gives the raw scores per text and the percentage that number represents for all words in the text. It should be noted that the grand total of 908 includes not only lexical insertions, which are treated in this chapter, but longer constituent insertions as well. 267

JONATHAN OWENS

Only five texts, 3, 5, 7, 9, 10 and 11, have an appreciable number of SA tokens. In these texts there are 15 participants, all of whom have studied SA formally in private and/or government schools for at least 11 years. In texts 2 and 4, on the other hand, none of the participants studied SA formally, and the use of SA is all but non-existent. A similar observation pertains to the English insertions in the material, as discussed in greater detail in Owens (2005: 35). 12.7.4 Areal norms In Section 12.5.2 it was suggested that avoidance of morphological processing was an important factor in accounting for data which included Ø forms and use of alternative constructions, like the hana possessive and sawwa + N, allowing one to avoid having to morphologize English words. I do not believe, however, that such an avoidance strategy is the only factor at work. Virtually identical periphrastic constructions occur in other languages of the region, notably for the present corpus, Hausa. The Hausa strategy for incorporating English verbal concepts is the same as NA, namely the verb yi “do, make” + N. Thus, closely equivalent to (20) would be Hausa (22) sun yi vdt “they voted (lit. made a vote)” As with the NA construction, the object noun is either a noun or a gerund, as in (23) nA yi transf£ring nA-shi I made transferring to-him “I transferred it to him” In fact, in the present corpus both NA and Hausa matrices make heavy use of this construction, 150 tokens of sawwa + N (83 N, 67 gerund) and 92 (41/ 51) of yi + N. This same “do” + object complement is equally used in other languages of the area. (24) is from Kanuri, and (25) from Fulfulde of Adamawa (Gottschligg 2003). (24) elek∫in sa-di-n election they-do-PROG “they are holding the election” o waa sumdglin pbtur (25) o he PROG do smuggling petrol “he is smuggling petrol” 268

CLOSE ENCOUNTERS OF A DIFFERENT KIND

A further parallel exists between Hausa and NA possessive marking, discussed in Owens (2002: 195). It may be suggested that regional norms spread throughout co-territorial languages steering the English insertions in NA towards the periphrastic solution. In this context it is worth returning to the treatment of SA verbal concepts in NA. In Section 12.6.5 it was seen that SA verbal concepts rarely use the periphrastic sawwa + N strategy, but rather incorporate SA verbs into NA morphological matrices. Besides the structural factors favoring this treatment (see Section 12.7.3), the relatively isolated sociolinguistic status of SA in northeast Nigeria probably plays a role as well. In contradistinction to its status in the Arabic world, SA is not a widespread variety in any linguistic domain in northeast Nigeria. There are no SA newspapers and no SA radio or television broadcasts, for instance. As a spoken medium it is, as seen above, confined to those who have studied it formally, and it is only Nigerian Arabs who draw extensively on it in a CS mode, as in the present corpus. This means that in contrast to English, which is used universally in northeast Nigeria as a switching resource, SA is confined largely to the community of Arabs. It may be suggested that this facilitates the development of insertion strategies specific to SA, since other linguistic communities will offer no competing models. 12.7.5 CS, borrowing and real issues in Arabic linguistics Both English and SA items are inserted into NA matrices, yet the grammatical outcome of the insertions are quite different. In general, SA insertions are either fully integrated into the NA matrices or closely track integration. English insertions, on the other hand, tend towards defining their own morphology. In a number of cases, this is simply an overuse of the Ø form (plural, possessive, definite article). In others it involves a periphrastic construction which, while fully grammatical in a strict sense, makes use of a construction which in NA itself is rather limited, namely “do” + VN (20). In yet others it simply involves avoidance of a construction, either categorically (e.g. the adjective) or to a marked statistical degree (the idAfa possessive). In both cases the English lexical insertions differ both from the treatment of “native” NA nouns, and the insertion of SA ones. I would suggest that a very cautious lesson can be drawn from the difference between these two types of insertions, SA and English in NA. SA nouns are well-integrated into the NA matrix language, in such a way that they by and large conform to Poplack’s notion of borrowing and to MyerScotton’s classic CS. In this respect NA does not differ essentially from other varieties of modern spoken Arabic. Using one theoretical framework or another, a good number of studies have documented the integration of SA words into Arabic dialectal matrices (Abdel-Jawad 1981; Bani Yasin 269

JONATHAN OWENS

and Owens 1987; Eid 1988; Holes 1987; Sallam 1980). The integration may be said to be perfect in that the grammar either of the dialectal matrix language or the standard embedded language are clearly identifiable. Sallam (1980: 92) notes that Lebanese speakers may use SA plural patterns, with a degree of Lebanese Arabic phonology, as in §aXdikA§ “friends” (Lebanese k for q), Mitchell (1986: 19) speaks of hybrid forms such as ra§-Bt “I saw”, with SA stem and dialect suffix, Meiseles (1980: 133) providing a number of such examples, for instance, bi-yuqAl “it is said” with the dialect prefix biprefixed to a SA passive stem. The production of such forms is not reducible to fixed rule, though the trend is clear, namely the intrusion to a greater or lesser degree of phonological and morphological matrix material from the dialect into lexical entries from SA. NA, a minority Arabic dialect spoken far outside a cultural domain marked by heavy exposure to SA in the media and education, is no less obeisant to this trend than are dialects in core areas of the Arabic-speaking world. On the other hand, SA contrasts quite markedly with insertion of lexical material from other languages, English and Hausa. English insertions are set off grammatically from words of Arabic provenance. At the wider observational level it appears that contact situations with Arabic as the matrix language divide into two basic types. On the one hand is the NA type with L2 lexical material treated in ways which are recognizably different from L1 material. Arabic-Dutch switching in Utrecht also falls broadly into this type (Boumans 1998; Nortier 1990; Owens 2005). On the other are contact situations where L1 Arabic and L2s meld perfectly into each other. A prototypical case here is Arabic-French switching in former French North Africa (e.g. Bentahila and Davies 1983; Nait M’barek and Sankoff 1988), though corpora with other L2s appear to exemplify it as well (e.g. Arabic-English in Jordan or the USA, Mustafa and Al-Khatib 1994). In this perspective, broad issues emerge from the current data, questions relating to the different ways languages are put together in language processing and production and to their sensitivity to the social conditions which support them. 12.7.6 An urban matrix The final issue I would like to take up in this contribution is the urbanness of the phenomenon described here. There are two general perspectives to the question, one relating to local Maiduguri realities, the other to the general theme of this volume, language and the city. From the first angle the answer is clearly that the form of complex lingualism described here is specifically urban. Ipso facto, in Maiduguri it is related to migration. NA is the native language of most Nigerians who identify themselves as Arabs, regardless of where they live. There is thus nothing specifically urban in the fact of speaking Arabic as a native lan270

CLOSE ENCOUNTERS OF A DIFFERENT KIND

guage in Maiduguri. Urbanness, rather, manifests itself in the array of languages of wider communication which Maiduguri Arabs are confronted with. The most striking member of their language repertoire is Hausa, for in using Hausa rather than Kanuri as their main lingua franca they contrast with Arabs living in rural Bornu. Furthermore, the latter group is either bilingual, with Kanuri as a second language (for a very few, perhaps Fulfulde as well), or monolingual, particularly along the Nigerian–Cameroonian border. The other two languages of wider communication, English and SA are structurally urban in that nearly all schools where either English or SA can be learned with any degree of proficiency are located in urban areas. Maiduguri in fact has, compared to any other urban area in Bornu, an impressive number of secondary and post-secondary institutions. The case of SA is interesting. Bornu is widely known throughout Nigeria as being the center for traditional Islamic education (Owens 1995). An important component of this educational system involves koranic memorization conducted under the aegis of itinerant scholars who typically spend a large part of their careers in small villages. Rural Bornu thus has an important tradition of scholarship. This tradition, however, is orientated towards religious education. The SA which appears in the current corpus, however, has a different source. A rough semantic classification was made, for instance, of all short lexical insertions (1, 2, 3 words long), and among those of SA origin none are specifically religious. Rather they describe politics (104 tokens, e.g. intixAbAt “elections”), administration (31 tokens, e.g. wilAya “state”), education (46 tokens, e.g. jAmi§a “university”) and other aspects of contemporary, largely civil society (see Owens 2000). This vocabulary is learned not in a religiously-orientated instruction, but rather via attendance of the privately-funded Islamiyya schools. These schools largely follow the curriculum set by the national Ministry of Education, but emphasize teaching in Arabic, and include religious subjects as one component of instruction only (Pfennig 2005). Like the government school counterparts, they are concentrated in urban areas. The largest such school in Maiduguri has over 2,000 students alone, and there are many others. The SA of the current texts is thus of urban origin. On the basis of field observation, as well as on 54 texts (as in Owens 1998), it may be added that the pattern with SA/NA mixing is hardly attested in rural villages. Maiduguri is established in the area as an educational magnet. Many Arabs reported that they came to Maiduguri for ilm, “knowledge”. By this was often meant traditional Islamic education, though it is noteworthy how many Arabs begin in a traditional system and switch at some point in their careers either to the Islamiyya system or to the national education system. New migrants are thus drawn in part by the promise of a better education, and the languages which they use and master have a specifically urban character. Turning to the second question, it is implausible to imagine that cities and migration to cities should function as a fixed, independent, universal 271

JONATHAN OWENS

variable influencing language in a specific way. The language repertoire of Maiduguri Arabs, as illustrated in the current corpus, has certainly been influenced through contact. However, on a reconstructed basis it can equally be shown that their native language has been historically influenced through contact with other languages in the Lake Chad area (e.g. Owens 1996 on convergent idiomatic structure), and this at a time when urbanization played a far smaller role (probably indeed, a negligible one) in defining the specificities of language contact than it plays today. What can be imagined, however, is that migration and urbanization create conditions for language contact with specifically contemporary affects. Multilinguality in Maiduguri is channeled towards a very small number of languages, English, Hausa, as well as SA in circumscribed circles. Such channeling is repeated in city after African city5 so that cities appear to have the paradoxical role of promoting multilinguality while limiting the choice of languages which will form multilingual repertoires. Migration and urbanization may be more notable for setting constraints on language choice than for determining unique linguistic effects in many African cities.

Notes 1 I would like to acknowledge the invaluable help of Klaus Wolf and the Bayreuth University Computer Center in this work, as well as Bernd Schroeder, who developed the basic search routine for building concordances. 2 The plural category may show gender neutralization (e.g. M plural where F plural is expected) but this is an issue outside the scope of this chapter. 3 It should be noted that post-modifying words like “some”, “certain” are excluded from the category of descriptive adjective in the monolingual corpus. Such words are morphologically inflected in the same way as descriptive adjectives, and like descriptive adjectives occur post-noun. However, semantically they belong to a class of quantifiers or are part of the definiteness system. 4 Should “foreign language” be a variable independent of frequency, this would be yet a third slowing factor. 5 Of course, with different language repertoires, e.g. French instead of English in former French colonies, English and Afrikaans in South Africa, another lingua franca or lingua francas in place of Hausa.

References Abd-el Jawad, H. (1981) “Lexical and phonological variation in spoken Arabic in Amman”, unpublished PhD thesis, University of Pennsylvania. Bani Yasin, R. and Owens, J. (1987) “The lexical basis of variation in Arabic”, Linguistics 25: 705–738. Bentahila, A. and Davies, E. (1983) “The syntax of Arabic-French code switching”, Lingua 59: 301–333. —— (1998) “Codeswitching, an unequal partnership”, in K. Jacobson (ed.) Codeswitching Worldwide, Berlin: Mouton de Gruyter, 25–49.

272

CLOSE ENCOUNTERS OF A DIFFERENT KIND

Boumans, L. (1998) The Syntax of Codeswitching: Analysing Moroccan Arabic/Dutch Conversation, Tilburg: Tilburg University Press. Bross, M. (2002) “State and perspective of languages in Maiduguri”, in R. Kawka (ed.) From Bulamari to Yerwa to Metropolitan Maiduguri, Köln: Köppe, pp. 103–116. Bybee, J. (1995) “Diachronic and typological properties of morphology and their implications for representation”, in Feldman (ed.), pp. 225–246. Caramazza, A., Alessandro, L. and Romani, C. (1988) “Lexical access and inflectional morphology”, Cognition 28: 297–332. Eid, M. (1988) “Principles for code-switching between standard and Egyptian Arabic”, Al-‘Arabiyya 21: 51–80. Feldman, L. (ed.) (1995) Morphological Aspects of Language Processing, Hillsdale, NJ: Lawrence Erlbaum. Finlayson, R., Calteaux, K. and Myers-Scotton, C. (1998) “Orderly mixing and accommodation in South African codeswitching”, Journal of Sociolinguistics 2: 395–420. Gottschligg, P. (2003) “The territorialisation of African language areas by excolonial languages”, paper presented at 12th Megachad, Maiduguri. Holes, C. (1987) Language Variation and Change in a Modernising Arab State, London: Kegan Paul International. Laine, M., Vainio, S. and Hyönä, J. (1999) “Lexical access routes to nouns in a morphologically rich language”, Journal of Memory and Language 40: 109–135. Levelt, W., Roelofs, A. and Meyer, A. (1999) “A theory of lexical access in speech production”, Behavioral and Brain Sciences 22: 1–75. Meiseles, G. (1980) “Educated Spoken Arabic and the Arabic language continuum”, Archivum Linguisticum 11: 118–148. Mitchell, T. (1986) “What is Educated Spoken Arabic?”, International Journal of the Sociology of Language 61: 7–32. Mustafa, Z. and Al-Khatib, M. (1994) “Code-mixing of Arabic and English in teaching science”, World Englishes 13, 215–224. Muysken, P. (2000) Bilingual Speech: A Typology of Code-mixing, Cambridge: Cambridge University Press. Myers-Scotton, C. (1993) “Elite closure as a powerful language strategy: the African case”, IJSL 103: 149–164. —— (1997) Duelling Languages: Grammatical Structure in Codeswitching, Oxford: Clarendon. —— (2002) Contact Linguistics: Bilingual Encounters and Grammatical Outcomes, Oxford: OUP. Nait M’barek, M. and Sankoff, D. (1988) “Le discours mixte arabe/français: emprunts ou alternance de langue?”, Canadian Journal of Linguistics 33: 143–154. Nortier, J. (1990) Dutch and Moroccan Arabic in Contact: Code Switching Among Moroccans in the Netherlands, Dordrecht: Foris. Owens, J. (1993a) A Grammar of Nigerian Arabic, Wiesbaden: Harrassowitz. —— (1993b) “Nigerian Arabic in comparative perspective”, SUGIA 13: 85–176. —— (1995) “Language in the graphics mode: Arabic among the Kanuri of Nigeria”, Language Sciences 17: 181–199. —— (1996) “Idiomatic structure and the theory of genetic relationship”, Diachronica 13: 283–318.

273

JONATHAN OWENS

—— (1998) Neighborhood and Ancestry: Variation in the Spoken Arabic of Maiduguri, Nigeria, Amsterdam: Benjamins. —— (2000) “Loanwords in Nigerian Arabic: a quantitative approach”, in J. Owens (ed.) Arabic as a Minority Language, Berlin: Mouton de Gruyter, pp. 259–346. —— (2002) ‘Processing the world piece by piece: iconicity, lexical insertion and possessives in Nigerian Arabic codeswitching’, Language Variation and Change, 17: 173–209. —— (2004) “Ethnolinguistic Vitality Theory and complex linguistic societies”, in P. Probst and G. Spittler (eds) Between Resistance and Expansion: Explorations of Local Vitality in Africa, Münster: Lit, pp. 35–48. —— (2005) “Ø Forms and lexical insertions in codeswitching: a processing-based account”, Bilingualism: Language and Cognition 8: 23–38. Pfennig, C. (2005) “Arabisch an modernen islamischen Schulen: eine soziolinguistische Untersuchung in Maiduguri, Nigeria”, unpublished PhD thesis, Bayreuth University. Poplack, S. and Meecham, M. (1998) “How languages fit together in codemixing”, International Journal of Bilingualism 2: 127–138. Portin, M. and Laine, M. (2001) “Processing cost associated with inflectional morphology in bilingual speakers”, Bilingualism: Language and Cognition 4, 55–62. Sallam, A. (1980) “Phonological variation in Educated Spoken Arabic”, Bulletin of the School of Oriental and African Studies 43: 77–100. Sankoff, D., Poplack, S. and Vanniarajan, S. (1990) “The use of the nonce-loan in Tamil”, Language Variation and Change 1: 71–101. Stemberger, J. (1995) “Phonological and lexical constraints on morphological processing”, in Feldman (ed.), pp. 247–270.

274

MOROCCAN ARABIC-FRENCH CODESWITCHING

13 DEVELOPMENT AND LINGUISTIC CHANGE IN MOROCCAN ARABIC-FRENCH CODESWITCHING Karima Ziamari

13.1 Codeswitching in students’ speech: an urban phenomenon? In this chapter1 I shall describe various aspects, both linguistic and sociolinguistic, of the ongoing development of Moroccan Arabic-French codeswitching as practiced by young Moroccan students. Codeswitching is a constantly evolving phenomenon, which has been sometimes related to linguistic deficit (Bentahila and Davies 1995).2 The codeswitching evolution can be seen in the creation of new structures. The Moroccan Arabic-French codeswitching of my informants differs from other corpora, whether originating in Morocco or in France, by the absence of certain linguistic phenomena. These differences may be due in part to the profile of the informants, and bring to light certain phenomena of linguistic innovation. The question arises whether these sociolinguistic observations are perhaps related to the specific dynamics of urban youth speech. To answer this question I shall first provide a sociolinguistic characterization of the informants. This will be followed by an analysis of the data within Myers-Scotton’s Matrix Language Frame Theory and an interpretation of the results. The corpus equates to eleven hours of recordings mainly of students of the Ecole Nationale Supérieure d’Arts et Métiers (ENSAM) in Meknes,3 but also of students of the Ecole Nationale d’Agriculture (ENA).4 There were two different groups comprising 33 informants, of which 10 were female. They originated from 12 Moroccan towns,5 with ages ranging from 18 to 30 years. The corpus was established over a three year period, from October 1998 to June 2001, in two different contexts. The first was ENSAM seminar rooms, the second less formal locations, such as rooms in university residences, train compartments, cybercafés etc. 275

KARIMA ZIAMARI

Codeswitching is mainly associated with an urban environment. MyersScotton (1993b: 33) speaks of an “urban syndrome”. It is this urban environment, where different languages come into closest contact and where there is the greatest incidence of bilingualism, which favors the emergence of the practice. Codeswitching in Morocco, a strictly urban phenomenon according to Caubet (2001: 22), while it does occur in various social categories, is essentially a feature of the speech of young bilinguals, such as my informants. Indeed, the ability to switch codes presupposes bi- or multilingualism, which itself is the product of formal education. The question arises, in the case of this group of students from various parts of the country, whether codeswitching is related more to their simply living in close proximity to one another, rather than to an urban context as such. This investigation however will treat the practice as an urban phenomenon, both because it took place in an urban location and because for the most part the informants come from large towns. These informants have Moroccan Arabic as their dominant language but are fully fluent in French.6 While they are capable of functioning in both languages independently, they nonetheless use codeswitching massively and frequently as their means of communication, not only amongst themselves but also with others, both youths and adults, outside their group. Both in Europe and the Maghreb, the frequent use of codeswitching is often associated with “youth speech” (Boumans 2001). But it is also more generally observed in “competent bilinguals who have studied French or who speak French at home” (Caubet 2001: 22). This is confirmed by Lahlou (1991), Abbassi (1977) and many others. The present research adopts the former view, as not only do the informants not separate the two codes, but they use amongst themselves a different, more developed version compared with that generally used in Morocco by other bilinguals. They practice linguistic innovations including lexical creations and prosodic markers which are peculiar to their group, thus excluding adult groups.

13.2 Language dynamics in the data The corpus reflects the contact between four languages: Moroccan Arabic, Standard Arabic, French and English, as shown in the following examples:7 (1) kayen l’ENA ∞end-ha wa}ed association vraiment yuqÑmu la-hÑ wa yuqÄad “There is ENA which has a really fearsome association” (2) je n’ai pas l’esprit de dialogue f X-Xbe} ka neX{e} Yafi switched off turned off “I’m not in the mood for talking in the morning, I wake up feeling tired” 276

MOROCCAN ARABIC-FRENCH CODESWITCHING

However it is Moroccan Arabic-French codeswitching which is most representative of the speech of these students. It is a means of expression used unrestrictedly even in formal situations such as the seminar room. Standard Arabic is brought into play in some situations to fulfill functions such as irony or humor: (3) Ñ mÑ ÅfiwaÄa-k fiahn l-Åi‹ara hada huwwa d-derri lli ka yedwi ∞end-na u ka ygul Ki hALa “And best of all, let me tell you, is the lad who talks well among us and has interesting things to say” (4) hada huwwa n-naqÑbi dyal-kum “He’s your trade unionist (i.e. a good talker)” (3) and (4) are two successive turns in which the first student (from ENSAM) uses Standard Arabic to express irony. The implication is quickly picked up by the second (from ENA), who replies in the same tone. The humor is expressed through the use of the Standard Arabic equivalent of a word, which gives it a noble connotation, as shown in this example where the use of the word îiflan “horse” contrasts with }ÜaW “ass” which in Moroccan society carries negative connotations: (5) ma qaWi∫ ga∞ ma qaWi∫ ma qari∫ }ÜaW “He never went to school at all, he is an ass” [Laughter from the group. The same speaker continues] (6) ma∫i îiflan l-muhimm . . . “He is not a horse, in other words . . .” In general, Standard Arabic for these young people is reserved for formal situations and religious discussions: (7) l-Åislam ma ∞eÜÜeW ma kan ta‹addud parce que kun kan ta‹addud ma γadi∫ ykun flaleî li kulli zaman wa makan “Islam has never been rigid. If it were it would never be applicable to all circumstances” Structurally when Standard Arabic is involved as one of the codes, codeswitching is rarely insertional. It exhibits alternating fragments of speech, unlike other languages such as Moroccan Arabic-French. This relates to the degree of familiarity and competence in the practice of the language. A number of researchers agree that there is a close link between a speaker’s competence and the directional or insertional character of codeswitching (Backus 1996; Myers-Scotton 1993a, 2002). 277

KARIMA ZIAMARI

Most of the time the informants use a vernacular variety of Moroccan Arabic which corresponds to their origins. In certain contexts however a contrary phenomenon occurs: females of rural origin display a leveling of differences, though females from prestigious towns such as Fez tend to maintain certain specific features. For example, a female informant from Fez uses the Fassi (i.e. from Fez) preposition lil-, rather than the more common l- or li-: (8) C’est une bonne affaire lil-u “It’s a good deal for him” Male informants in general display identifiable or even very marked regional intonations and pronunciations. We noticed the alternation e/a in the verb h|eW “he spoke” used by a speaker from Ksar El Kebir: ka nha|Wu “we speak” rather than ka nhe|Wu. The use of ba “he wanted” rather than bγa has also been recorded among speakers from Marrakech. Where female speakers are concerned however, it is often difficult to identify their origin. Consider the following examples of speakers of different genders from the same town of Oujda: Female: (9) fa∫ ka yem∫i ka ne∞Wef dert-ha b yeddi-yya / }etta ana ma sahla∫ / u ka nebqa nregreti nar-i / merde / je tiens à lui / u ka nebqa nÔeÆÆeW fi-h “When he leaves, I know it’s my fault / I’m not easy to live with either / and I start to feel sorry / god damn it / I love him / and I start thinking about him” Males: (10) xYeW lli ma yeddi∫ weLdiyya ma ykun∫ ∞ayi∫ ∫i ∞i∫a mli}a “Because if you don’t marry an Oujda girl you won’t be happy” (11) Ketti }n¢aya ngul l¢k ∫i }wayeL matalan yLiwna nwertu-hum ngulu }na ila ma derna-hum∫ par contre ila ma derna-hum∫ bon t’es pas civilisé “I’ll tell you something, we care about traditional things and if you don’t observe them, well you’re not civilized” Note in (9) the use by the female speaker of the suffix pronoun –yya “my” in the word yeddi-yya “my hands” from the standard Moroccan plural form yeddin. The male informants on the other hand use region specific words such as xYeW “because” in (10) and use verbs in the imperfect without the preverb ka-, which tends to be omitted in Oujda, as shown in example (11) with the verbs yLiwna and ngulu. 278

MOROCCAN ARABIC-FRENCH CODESWITCHING

13.3 Language change within the Matrix Language Frame: Matrix Language Turnover This research is based on Myers-Scotton’s Matrix Language Frame theory, an insertional model, as distinct from linear models (Poplack 1980). The Matrix Language Frame (MLF) is based on the notion of a hierarchical relation between codes. A Matrix Language (ML) is seen as the dominant code over the Embedded Language (EL). The ML generates the morphosyntactic frame of the utterance and then provides the grammatical link between constituents by imposing its order and supplying active morphemes8 known as system morphemes. The EL is represented by morphemes known as content morphemes. The MLF’s specificity lies in the value attributed to the ML, which is seen as extremely dynamic. It changes both diachronically and synchronically. Diachronically, a change in the sociopolitical balance within a community can modify the status and power of a language. Synchronically, any change in extralinguistic factors, such as subject, speakers etc., could reduce the status which would become an EL. The change of ML, or Matrix Language Turnover (MLT) does not occur in all corpora nor in all codeswitching contexts. Certain corpora largely reflect the stability of the ML (MyersScotton 2002: 64).9 In other contexts though MLT is characteristic, as shown by Backus (1996) in relation to Turkish-Dutch codeswitching, by Boumans (1998) for Moroccan Arabic in contact with Dutch, and by Caubet (1998) for Algerian Arabic-French. In the sociolinguistic context of this research, MLT constitutes a key linguistic characteristic. Moroccan Arabic is not necessarily the only language which can serve as ML, as Caubet (2002) points out.10 French frequently challenges this position. MLT reflects the sociolinguistic choice of the informants. With ENSAM11 students there is a significant divergence of status between the languages. Moroccan Arabic, though used, is strongly stigmatized compared with French, especially by female informants. This gives French the social and linguistic power to function as ML. MLT occurs in the corpus not only triggered by external factors but in particular in the same context, the same conversation and, contrary to MLF “rules”, in the same turn, as seen in the following examples: (12) l-muhimm c’est l’occasion pour qu’on en parle // je me suis jamais ze∞ma vue sortir avec une personne de mon âge // il faut que je sorte avec une personne kbeW men-ni // une personne plus mûre tu vois // mais une personne m∞a-ya // ka teqWa m∞a-ya “In other words it’s time to talk about it. In fact I have never been out with someone of my age. I have to go out with someone older, someone more mature, you see. But someone who is with me, who studies with me” 279

KARIMA ZIAMARI

(13) ka t∫uf wa}ed riche ka yetzewweL we}da pauvre // intellectuelle peut-être elle a des problèmes avec la famille // walakin kat kat // il y a une entente binat-hum “You see a rich guy marrying a poor girl. She may be intellectual and have problems with her family, but there is an understanding between them” Synchronically, MLT in the corpus occurs in both directions, from French to Moroccan Arabic (12) and from Moroccan Arabic to French (13), though the former is much more frequent. This provides great dynamism in the language structures produced by the informants. Examples of such dynamism can be seen in a number of structures such as verb and nouns insertions, lexical creations and phonological realizations.

13.4 Verb insertion The insertion of French verbs into the morphological and syntactic structure of Moroccan Arabic consists in keeping the French verbal root and adding the ML inflection: (14) kõpawa 3rd sg. masc. perfect “comparer” “He compared” The French verb adopts the morphology of the ML, viz. that of quadriliteral verbs in the perfect and imperfect ( ykõpaWi “he compares”). It can also be prefixed t-/tt- to indicate the passive (middle voice) and to express reflexiveness and reciprocity. (15) ka netkõpowya 1st sg. imperfect “comporter” “I behave” Many studies have explored this type of insertion and shown that French verbs adopt Moroccan Arabic morphology.12 But the verbs which occur are generally of the -er (first) group. In the codeswitching of Moroccan students in Morocco verbal insertion is very productive, and involves verbs of all three French groups, all conjugated in the same way: (16) yetabli 3rd sg. masc. imperfect “établir” “He will establish” (17) γa tatendi-h going 3rd sg. fem. imperfect “atteindre” “She will attain it” 280

MOROCCAN ARABIC-FRENCH CODESWITCHING

It must be noticed that English verbs are beginning to be inserted in the same way: (18) understand it 1st sg. perfect “to understand” “I understood” (19) laykiti-na 2nd sg. perfect “to like” us “You liked us” The insertion of an English verb into a mixed structure is a rapidly growing feature of my informants’ codeswitching unlike what has been recorded in other English-Arabic codeswitching. According to Myers-Scotton and Jake (1997, 2001), this type of verbal insertion shows a lack of congruence in Palestinian Arabic-English codeswitching. When the two languages come into contact, the English verb tends to be inserted as an inflectional phrase (Myers-Scotton and Jake 1997: 30): (20) ana I don’t like it “Myself, I don’t like it” So verbal insertion, though rare in other corpora, occurs frequently in my informants’ conversation, which seems to indicate that their codeswitching is evolving. Witness also certain linguistic creations in which the French adjective “méplat”13 is inserted as if it were a verb: (21) ma ka tmeplati-ha∫ f l-weXY “You aren’t making it flat in the middle” Besides its morphological integration, the evolution of verbal insertion is also seen in cases of semantic slippage. Certain verbs are used with a particular meaning by both male and female informants: Female: (22) regarde comment il est ∫uf ka yetfakxa “See how he is, he can be sent by fax”, (because he is so thin) Male: (23) xeXX¢k tkun franc Wa-k ka t}awel tformini u Xafi “You must be frank; you are just trying to spare my feelings and nothing else” 281

KARIMA ZIAMARI

As has been demonstrated (Caubet 2003; Vicente 2003), the semantic slippage characteristic of youth speech in Morocco is expressed here through French verbs. Furthermore, changes in linguistic practice can be seen in the similarity between certain insertions, like the insertion of Arabic verbs, featuring youth speech in migrant context: (24) je vais nelqa sûrement une solution “I’m bound to find a solution” This type of insertion occurs in the “mixed speech” of young immigrants of Maghreb origin living in France, as analyzed by Melliani (2001): (25) je vais les }aKi “I’m going to swindle them” (Melliani 2001: 68) The two Arabic verbs (24, 25) function differently. In Melliani’s corpus verbs are inserted as bare forms, whereas in mine they retain their fully inflected form. In the immigrant corpus the verb is normally inserted into structures without the auxiliary: (26) tu xdem toi Said? “Do you have a job, Said? (Melliani 2001: 66) However in my corpus this structure is rare and has an ironic function: (27) tu heWWef beaucoup “You’re working too hard” Whatever function is served by this type of insertion, the linguistic creativity involved is clear, showing how much the students’ codeswitching has evolved. It offers modes of integration which go beyond those seen in other corpora compiled in Morocco, and is closer to in some respects to the mixed speech of second generation immigrants in France.

13.5 The insertion of the noun phrase In general, when the ML is Arabic (Moroccan or Algerian), as has been shown in various studies, the inserted noun is accompanied by a definite article (le, l’, la, les) following the determiners had and wa}ed known as complex determiners: (28) f wa}ed la publicité “In an advert” (Lahlou 1991: 135) 282

MOROCCAN ARABIC-FRENCH CODESWITCHING

My corpus includes other types of determiners, such as indefinite articles and possessives: (29) γada l wa}ed un mois “It goes back about a month” (30) had mon genre “Mon genre” (i.e. my kind) This use of the indefinite article or the possessive in place of the definite article l- is new, and is not recorded in previous studies. It is a structure produced by female informants, which thus shows they are more daring in innovation. What is involved is not only the insertion of definite articles, masculine and feminine, singular and plural, or even of indefinite articles, but more especially the use of possessives and even of quantifiers: (31) wa}ed huit représentants “About eight representatives” In my corpus the structure [possessive + noun] occurs inserted into a prepositional phrase, which is an infrequent combination in other corpora: (32) m∞a tes enfants “With your children” Normally, when French is the EL, possessives occur only rarely. Instead we find the use of the dyal structure: (33) la retraite dyal-ha “Her retirement” (Lahlou 1991: 135) In Moroccan Arabic, the possessive relationship is expressed in two ways: an analytic structure with the preposition dyal and its variants (d/nta∞), and a synthetic construction involving the juxtaposition of two noun phrases (Caubet 1993, vol. I: 20): (34) bnat la fac “The girls from the university” Unlike the synthetic structure, the construction with dyal is productive in Moroccan Arabic-French codeswitching. The corpus exhibits particular uses of dyal which facilitate the bridging between a number of terms: (35) le rapport ma bin le volume dyal la molécule dyalt la structure u l’oxyde dyal-ha “The relationship between the volume of the molecule of the structure and its oxide” 283

KARIMA ZIAMARI

The frequent unexplained use of the preposition dyal is indeed one of the prime examples of linguistic structure evolution. Boumans (2001, 2006) likewise links the appearance of the analytic construction with the linguistic change of Moroccan Arabic-Dutch codeswitching among migrants. The preposition dyal14 appears in a variety of syntactic environments and also replaces the genitive in Boumans’s work. But for my informants it tends, by a remarkable semantico-syntactic slippage, to replace the French preposition “de” in a number of constructions such as: la plupart dyal / sous forme dyal / faire partie dyal / à partir dyal / à cause dyal: (36) ça dépend dyal la moyenne “It depends on the average”

13.6 Codeswitching: linguistic innovations and crypto-ludic functions Through their codeswitching my informants display many linguistic innovations. Various examples are to be found in the corpus. The word l’amitié “friendship” is used, among others, to address someone or to express an affinity, similar to the use in Moroccan Arabic of the term l-xawa “fraternity”. (37) kamlin inférieur γi h|eW A l’amitié “We are all inferior, go ahead, speak, my friend” The students also use other words to express their relationships within the group. The ENSAM students use binôm-i (i.e. mon binôme “my partner”), while those from ENA use binox, formed from the same base, binôme. The word binox is formed by inverting the morphemes ex-binôme and eliminating the phoneme m. The ENA binox are directed by a leader known as KexXuX. This is another creation based on the word KexXiyya “personality” and modeled on the schema for quadriliteral nouns (CeCCuC) in Moroccan Arabic (Caubet 1993: 94). Linguistic innovation and semantic slippage serve the functions of humor and disguise. Word play reveals a linguistic appropriation of the language as a source of fun for youngsters “en perpétuelle recherche d’humour” (Caubet 2003: 267). This function is displayed in some surprising creations: (38) Y}en Pampers15 “Let’s work seriously at the humidity course” Y}en, literally “to grind”, means “to work seriously” while Pampers “nappies” is a cryptic reference to a scientific course on humidity. The cryptic feature marks adherence to a particular social group, as any group will seek to ensure it is not understood by others. The phenomenon 284

MOROCCAN ARABIC-FRENCH CODESWITCHING

was noted (Doran 2000) in the speech of young people in Paris suburbs. Thus my informants compete to create lexical items with the same base, such as binôme. If both groups use the same word as a base and claim credit for one or other creation, it is in order to distinguish themselves from other university students, who may for example use the term comrade to indicate their own political allegiance.

13.7 The phonological dimension The phonological and prosodic dimension is a very important marker of identification. Our corpus reveals evolution in this area too. Indeed it occurs to such a degree that the ML may be identified from nothing more than a pause: (39) l’internat # problème “Boarding school is indeed a problem” Although the utterance is apparently in French, the structure is from Moroccan Arabic which allows the subject–predicate relationship to be expressed through a zero copula. Such utterances in French are in fact syntactic replicas of Moroccan Arabic as is shown by the following example taken from the corpus: (40) l-meγrib # meLhed “Morocco is well advanced” Phonological and prosodic clues distinguish different varieties of French in the Maghreb (Abbassi 1977; Caubet 1998). Abbassi for example specifies three varieties: High Standard French, Standard Moroccan French and Substandard Moroccan French. This classification presupposes an educational differential in favor of the first variety (1977: 28–32). The influence of Moroccan Arabic, with variations according to regional origins, on the French spoken by informants, especially the males, is very strong, especially on their intonation and pronunciation. The female informants on the other hand, as already noted, make an effort to achieve a good “French” pronunciation, as close as possible to High Standard French. They see this as a way of imposing an urban character on their intonation. Females moreover stigmatize Moroccan Arabic more strongly and make less use of Standard Arabic. Male informants on the other hand have different priorities. They remain closer to their original pronunciation and intonation as an affirmation of identity. They express their masculinity by rolling their /r/ in French, changing the point of articulation for certain phonemes and especially by lengthening certain vowels. The rolling of the /r/ is particularly noticeable when 285

KARIMA ZIAMARI

the two languages are in contact. The French dorso-uvular /r/ becomes the apico-alveolar [r] or [w]. Two alternatives occur: a rolled /r/ when there is assimilation with the article l- in accordance with the rules of Arabic as ML (41), and the normal phonetic realization of standard French (42). (41) r-resto “The restaurant” (42) l-rayon “The radius” This rarely occurs with female informants. The gender difference manifested in the treatment of /r/ is neutralized in certain cases. Sometimes females roll the /r/ to conform to the speech style of others with the intention of establishing an identity which goes beyond the feminine–masculine distinction. They seek thus to portray themselves rather as ‘AM’ (arts et métiers) students. Within codeswitching, variations in the articulation of the French /t/ are very significant for certain speakers. The French apico-alveolar /t/ is pronounced with a more dental articulation in words such as: but “goal”, cité “residence” and animateur “instructor”. The vowel system of the EL, French, likewise undergoes modifications when in contact with Moroccan Arabic, the most important change being the lengthening of the vowel /a/ in open syllables and of the vowel /i/ under the stress: (43) comunication “What a communication!” (44) s-synoncme “Oh, the synonym” These male informants are certainly capable of speaking French with a normal French pronunciation and intonation, but they adopt their specific speech style to mark the gender difference with the female students.

13.8 Changes and developments in Moroccan Arabic-French codeswitching Changes in Arabic-French codeswitching are seen in the specificity of the linguistic forms which occur in the corpus. Comparing the features found in my corpus with those of others reveals certain divergent characteristics. The synchronic instability of the ML is the most salient feature of the linguistic landscape of this research. The ongoing social power struggle between the two languages leads to MLT, with sometimes one and sometimes the other language functioning as ML. This in turn gives rise to new combinations.

286

MOROCCAN ARABIC-FRENCH CODESWITCHING

As stated by Caubet (2002: 124), “le mélange ne date pas d’hier au Maroc (mixing is nothing new in Morocco)” and the two languages have long coexisted in Moroccan society, codeswitching continues to change and develop. The linguistic proficiency in both French and Arabic of our informants can be added to this old contact between the two languages. Therefore, the dominant use of codeswitching does not reflect a linguistic deficiency in French and cannot be attributed to the Arabization policy as stated by many people (Benthila and Davies 1995). This change is not due to any linguistic deficiency, but is the result of innovation (Backus 1996: 28–29)16 and creativity in a ‘ “default-mode” of conversation’ (Lahlou 1991). Often a particular group of students will create new cryptic and ludic linguistic forms, which are subsequently taken up more generally: (45) Lab-ha f-l’axe “He spoke well” Literally, “he brought it on the same line / direct to the target”, an expression used by young people in Casablanca. (46) γadi nem∫i l à deux mètres “I’m off to see the dealer” Literally “I am going for two meters”, a cryptic expression re-used in certain rap songs by ENSAM students. Male informants exhibit a different codeswitching compared to females. They display prosodic markers, verbal insertion and a greater number of lexical innovations. Females innovate more in terms of structures such as the insertion of possessive and definite articles with nouns. Like any form of youth speech, codeswitching presents an important field for research. The spread of this means of communication in society is leading these students to create their own version, as a distinctive sociolect. The urban location is the focus of the special and identity differentiation which motivates the linguistic dynamic. ENSAM and ENA students constitute two distinct and competing groups. For many years ENA was the only grande école d’ingénieurs in Meknes. ENSAM was founded only in 1997. The rivalry is not based on different diplomas, but it is important to the students in identity terms. Unlike ENSAM,17 ENA is located on the outskirts of the town. The expressions used by each group to refer to the other reveal that the town center location of ENSAM is seen as significant. ENSAM students use the expression kilomètre six (6K) to refer to ENA students, alluding to the distance of the site from the town center. They in turn are known by their ENA rivals as zuL ptipanat (“two ‘pains au chocolat’ ”) implying that they are spoiled by their favorable location.

287

KARIMA ZIAMARI

We see then that youth speech styles are characterized not only by abundant lexical and structural creativity, but are also motivated by a need to mark their spatial identity. Therefore “Our young people have only to open their big mouths (Il suffit que nos jeunes ouvrent leur gueule)”, as the Moroccan weekly Telquel claimed in its June 2004 headline, to show linguistic innovation and creativity in codeswitching, as indeed do young people in general and our ENSAM students in particular.

Notes 1 I would like to thank Ad Backus for his useful and constructive comments. 2 The Moroccan educational system had imposed the Arabicization of technical and scientific subjects for primary and secondary education. In this context, one can link the codeswitching of this generation to the lack of proficiency in French due to this Arabicization policy. In Bentahila and Davies (1995: 84) some structures (the insertion of verbs) are linked to the fluency of speakers: “The speakers are using French building bricks to erect what is clearly marked as an Arabic structure.” 3 ENSAM (Ecole Nationale Supérieure des Arts et Métiers) was created in 1997. Its main mission is the formation of engineers. The course of study lasts for five years after the scientific and technical baccalaureates. 4 ENA (Ecole Nationale d’Agriculture) in Meknes is a high school of engineers in agronomy. 5 The Moroccan towns are: Casablanca (8 informants), Fez (7), Oujda (4), Meknes (3), Marrakech (3), Salé (2), Rabat (1), Khouribga (1), Ahfir (1), Ksar El Kebir (1), Sidi Kacem (1), Sidi Yahya (1). 6 These students are admitted in ENSAM after a pre-selection requiring a good level in French, an exam and an interview which aim to verify their level in French. 7 Moroccan Arabic is transcribed in italic, Standard Arabic in bold, French in regular and English is underlined. 8 Myers-Scotton (2002: 73) distinguishes four types of morphemes: (1) early system morphemes (determiners); (2) bridge late system morphemes (prepositions: dyal “of ”); (3) outsider late system morphemes (flexion, tense, aspect); and, (4) content morphemes (nouns, adjectives). 9 The ML is stable when it does not change across sentences. The following example shows that Arabic as the ML is maintained in all bilingual utterances: l-biochimie kamla qWit-ha f nhar hadi matière hadi bnadem ka ysme∞ l-biochimie w ana qari-ha f nhar walakin une fois ka tmitrizi s-système ka tquleb-hum ∞la∫ mes parents ma weldu-ni∫ ∞la beWWa ana ga∞ “All the biochemistry, you see, I have learned it in one day. Is it a discipline? We heard about the biochemistry, then I learned it in one day, but when you master the system, you fool them. Why didn’t my parents give birth to me aboard?” 10 Caubet (2002: 124) reports this regarding the Maghreb: “Au Maghreb, la langue ‘matrice’ est généralement l’arabe maghrébin et la langue ‘encastrée’ le français”. 11 Informants were given a questionnaire that aimed at describing their sociolinguistic profile and their attitude towards the languages they speak. The majority favors French while Standard Arabic appears in the second position competing with English. Moroccan Arabic is stigmatized and considered less important than the other languages.

288

MOROCCAN ARABIC-FRENCH CODESWITCHING

12 Bentahila and Davies (1983, 1995), Lahlou (1991), Caubet (1998, 2002), BarillotFadil (2002). 13 “Méplat” is not a common word. It is a technical word in some scientific disciplines. The informants use it in this context. 14 The constructions with dyal are among those which reveal some of the MLF’s limits. Though this preposition is a “bridge late system morpheme”, it is pervasive when French is the ML. So, the MLF (and the other models) cannot explain why the preposition is realized in Moroccan Arabic in a French syntactical environment. The preposition dyal is sensitive to the activation of the embedded language (Moroccan Arabic) and constitutes a limit to the MLF as a model able to explain all structures (Ziamari 2006). 15 Pampers is a trade-name of babies’ nappies. Here, it is used as a nickname for a course on “humidity”. 16 “Individuals employ new elements all the time and by doing that they introduce innovations into their idiolect. If an innovation becomes conventionalized, their idiolect has undergone change” (Backus 1996: 28). 17 ENSAM’s students would like their school to be located in Casablanca.

References Abbassi, A. (1977) “A sociolinguistic analysis of multilingualism in Morocco”, unpublished PhD thesis, University of Texas, Austin. Backus, A. (1996) Two in One: Bilingual Speech of Turkish Immigrants in the Netherlands, Tilburg: Tilburg University Press. Barillot-Fadil, N. (2002) “Etude sociolinguistique et intonative de l’alternance codique arabe marocain/français”, unpublished thesis, University of Paris 7. Bentahila, A. and Davies, E. (1983) “The syntax of Arabic-French code switching”, Lingua 59: 301–330. —— (1992) “Code-switching and language dominance”, in R. J. Harris (ed.) Cognitive Processing in Bilinguals, Amsterdam: Elsevier, pp. 443–458. —— (1995) “Patterns of code-switching and patterns of language contact”, Lingua 96: 75–93. Boumans, L. (1998) The Syntax of Codeswitching: Analyzing Moroccan Arabic/ Dutch Conversations, Studies in Multilingualism 12, Tilburg: Tilburg University Press. —— (2001) “Moroccan Arabic and Dutch: language of Moroccan youth in the Netherlands”, in J. J. de Ruiter (ed.) Plurilinguisme au Maroc et dans les communautés marocaines en Europe, Langue et linguistique 8, 97–120, Fez. —— (2006) “The attributive possessive in Moroccan Arabic spoken by young bilinguals in the Netherlands and their peers in Morocco”, Bilingualism: Language and Cognition 9, 3: 213–231. Caubet, D. (1993) L’arabe marocain, vols I–II, Paris-Louvain: Peeters. —— (1998) “Alternance de codes au Maghreb: pourquoi le français est-il arabisé?”, in Plurilinguismes 14 (Alternance des langues et apprentissage en contextes plurilingues): 121–142. —— (2001) “Comment appréhender le codeswitching”, in C. Canut and D. Caubet (eds) Comment les langues se mélangent: codeswitching en Francophonie, Paris: L’Harmattan, pp. 21–32.

289

KARIMA ZIAMARI

—— (2002) “Métissages linguistiques ici (en France) et là-bas (au Maghreb)”, in Pratiques langagières urbaines, Enjeux identitaires, enjeux cognitifs, Ville-EcoleIntégration, Enjeux 130: 117–132. —— (2003) “L’Intrusion des téléphones portables et des ‘SMS’ dans l’arabe marocain en 2002–2003”, in D. Caubet et al. (eds) Parlers jeunes, ici et là-bas: pratiques et repésentations, Paris: L’Harmattan, pp. 247–270. —— (2004) Les mots du Bled, Paris: L’Harmattan. Doran, M. (2000) “Speaking verlan: performing hybrid identity in suburban Paris”, text of AAAL presentation, Vancouver. Lahlou, M. (1991) “A morpho-syntactic study of codeswitching between Moroccan Arabic and French”, unpublished PhD thesis, University of Texas, Austin. Melliani, F. (2001) “Le métissage langagier en questions: de quelques aspects morphosyntaxiques”, in C. Canut and D. Caubet (eds) Comment les langues se mélangent: codeswitching en Francophonie, Paris: L’Harmattan, pp. 59–72. Myers-Scotton, C. (1993a) Duelling Languages: Grammatical Structure in Codeswitching, Oxford: Clarendon Press. Paperback edition 1997. —— (1993b) Social Motivations for Codeswitching: Evidence from Africa, Oxford: Clarendon Press. —— (1998) “A way to dusty death: the Matrix Language Turnover Hypothesis”, in L. Grenoble and L. Whaley (eds) Endangered Languages: Language Loss and Community Response, Cambridge: Cambridge University Press, pp. 289–316. —— (2002) Contact Linguistics: Bilingual Encounters and Grammatical Outcomes, Oxford: Oxford University Press. Myers-Scotton, C. and Jake, J. (1997) “Codeswitching and compromise strategies: implications for lexical structure”, International Journal of Bilingualism 1: 25–39. —— (2001) “Explaining aspects of codeswitching and their implications”, in J. Nicol (ed.) One Mind, Two Languages: Bilingual Language Processing, Oxford: Blackwell, pp. 84–116. Poplack, S. (1980) “Sometimes I’ll start a sentence in English y termino en Espagnol: toward a typology of code-switching”, Linguistics 18: 561–618. Sankoff, D. and Nait M’barak, M. (1988) “Le discours mixte arabe/français: emprunts ou alternances de langue?”, Revue Canadienne de Linguistique 33: 143–154. Telquel 130, June 2004 “Nos jeunes ouvrent leur gueule”, C. Bensalmia and A. R. Benchemsi, 22–32. Vicente, A. (2003) “La négociation des langues chez les jeunes de Sebta”, in D. Caubet et al. (eds) Parlers jeunes, ici et là-bas: pratiques et repésentations, Paris: L’Harmattan, pp. 33–47. Ziamari, K. (2003) “Le codeswitching intra-phrastique dans les conversations des étudiants marocains de l’ENSAM: approche linguistique du duel entre l’arabe marocain et le français”, unpublished thesis, INALCO-Fès. —— (2006) “La construction SN dyal SN en codeswitching arabe marocain/ français: les failles du Matrix Language Frame”, in S. Mejri (ed.) L’arabe dialectal: enquêtes, descriptions, interprétations, Actes Aïda 6, Tunis: Cahiers du CERES 13, pp. 499–507.

290

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

14 THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS Sherin Rizk

14.1 Introduction This chapter is the result of a study undertaken between 1999 and 2000 on what is labeled as “youth language” (YL) in Egypt (Rizk 2000), a very recent social phenomenon that has caught the attention of press columnists and the media. Our interest in these linguistic practices was aroused by press articles that criticized the way young people in Cairo spoke, and raised concerns about its negative influence on “our Arabic language”, a problem of special note given the almost complete absence of any scholarly study of these practices in the Egyptian context.1 The role played by the media in the representations and tokenization of such phenomenon can be significant and emphasizes the complex interaction developing between language users and the media establishment. We will begin with a study of the various attitudes and representations transmitted by the media on the subject of this “language” and then present the results of an exploratory survey of 40 young Cairo students in the faculties of Arts, Engineering, Medicine and Commerce (5 girls and 5 boys from each section). This survey sheds light on students’ linguistic practices and their attitudes towards this YL. In order to apprehend the phenomenon of YL in a city like Cairo, we must first question the relevance of such a category, the scope of these practices, (who speaks it, when and why) and the possible role this “language” could play in the evolution of Cairene Arabic. How does YL express – or not – a certain form of urbanity? How does it represent – or not – a linguistic innovation compared to a more ordinary colloquial practice? Is it a “transitional” form of language reserved for a specific age group or is it a greater phenomenon encompassing the emergence of a new social category on a globalized urban scene: the youth? It is not easy to answer these questions in a city like Cairo because the phenomenon is relatively new and has hardly ever been addressed in Egypt; furthermore we lack in-depth sociolinguistic studies on the actual speaking practices of the Cairene population. The few linguistic studies on the speech 291

SHERIN RIZK

of young Egyptians (Allam 2000; Peterson 2002, and our present study) remain confined to a specific social group (the students of Cairo), which limits its sociological impact. These studies have been conducted in foreign languages (English, French). We did find one study in Arabic (Lotfi 2002); however, this is only a glossary headed by a few introductory pages that reveals the current stereotypes about the vulgarity of this language and the danger it represents. As Niloofar Haeri (1997: 18) says, “Lack of attention and interest in non-classical Arabic . . . is difficult to dissociate from an ideology that considers non-classical varieties of Arabic as less than ‘real language’.”

14.2 Attitudes and representations of the written press and audio-visual media 14.2.1 Is the linguistic situation in Egypt a double diglossia? For a long time, linguists and Arabic-speaking intellectuals have described the language situation of Egypt within the theoretical frame of diglossia, assigning to each linguistic pole a distinctive function: literary Arabic – al-fuX}A 2 – was supposed to be used mainly for writing and in all formal and official oral situations, while colloquial Egyptian Arabic – al-∞Ammiya – was supposed to be used essentially for oral and non-official communications. In Egypt, the colloquial form is often considered sub-standard with respect to al-fuX}A, the academic standard. Its written use is always stigmatized by the “defenders of the Arabic language”, even though ∞Ammiya has massively infiltrated written domains. To quote a few examples, colloquial Arabic is to be found in Beyram Eltonessy’s poems dating from the early twentieth century, in Ahmad Fouad Negm’s writings of the 1960s and 1970s, in plays, and more recently in caricatures and advertisements published in the daily press and therefore reaching a mass audience, in addition to its increasingly frequent use in newspaper articles and novels. We observe somehow a parallel relationship at the oral level between the YL and the colloquial language (Figure 14.1).

Written form

Oral form

Standard/Non-standard

Standard/Non-standard

Literary Arabic/Colloquial Arabic

Colloquial Arabic/Youth language

Figure 14.1 Standard/Non-standard in written and oral forms

292

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

YL is therefore considered non-standard compared to colloquial Arabic and is generally stigmatized, particularly by those who represent the symbolic linguistic authorities (linguists, professors, intellectuals, etc.). But this has not slowed its remarkable development in recent times. Moreover, the press, and particularly large daily newspapers (such as al-AhrAm and al-AxbAr) frequently use lexical terms of YL, especially in advertisements, while publishing simultaneously – next to those advertisements – articles expressing concern about diminishing standards, and dwelling on the strangeness of YL and its threatening influence on the national language (see, for example, Elghazaly 2000; Elshaer 1999; Esmaïl 2002). The deprecatory lexemes used to describe YL are very similar to those used against colloquial Arabic (Farag Allah 1996). Thus YL is described as “a deformed monster of bastard expressions” (masx mu∫awwah min alfAz laqCYa), full of “faulty” (alfAz maγlEYa), “intruding” (daxCla), “abbreviated and easy expressions” (muxtaXara wa sahla) that do not belong in “our beautiful language” (luγatuna al-∞arabiya al-jamCla). 14.2.2 From young “deviant” to “hip” youth – cross-influences among young people, media and the artistic scene One can distinguish two main phases in the recent attitudes of broadcast media. In the first period, between 1990 and 1995, the phenomenon described as “youth language” started to emerge. The media, especially television media, presented a stereotypical image of YL speakers. The latter were essentially “deviants”, “bad kids” that television audiences should not imitate; they included characters such as a student failing at school, taking drugs and devoid of morals – played by Ahmad Abdel Aziz in the series el-bBt el-kebCW “the large house”. The negative image changed completely after 1995, and more precisely with the release of “youth movies”, the first of which is generally recognized as Isma∞Cleya Waye} gAy (Ismaïlia Back and Forth 1997) followed by Se∞CdC f-el-gAm∞a el-amWCkeya (An Upper Egyptian at AUC 1998) – both were huge hits. In addition, the advertising industry has appropriated many other forms of YL. The stereotype of the young delinquent has therefore disappeared and been replaced by the image of the cool kid who speaks a funny language but whose life is successful. When one of the Cairo students surveyed referred to YL as loγat el-kAmannannA “language of Kamannanna”, he emphasized, perhaps unconsciously, the significant role played by artistic creations (films, songs, etc.) in the propagation and creation of youth expressions. KAmannannA is a made-up word in a song performed by the two young actors in the movie Isma∞Cleya Waye} gAy. The meaning of the word remains unclear, but it is understood in the video clip of kAmannannA as everything a young man may wish for: money, girl, car, etc. (the boys act out these “wishes” every time they pronounce the word). 293

SHERIN RIZK

Students recognize the role played by the media, not only in broadcasting youth expressions but also sometimes in creating new ones. Thus, an expression such as ka}Wata – a neologism meaning “transgression of the moral and social taboos” – probably known only to a small group, was made highly popular by the movie Se∞CdC f-el-gAm∞a el-amWCkeya. But the too frequent use of a word can lead to an adverse reaction: young people lose interest in that word and try to create new ones. For example some of the students we surveyed used the words ∞A∫, literally “Hail” (as in “Hail to the king”) between themselves to express agreement or consent rather than ∞e∫Ya literally “cream”3 which, to them, had become “commonplace” and “stereotyped” as a result of the movies. Thus, if a student says: fCk men senCma bokWa? “Would you like to go to the movies tomorrow?”, the answer may be ∞e∫Ya “okay, good idea” or ∞A∫ yallA “okay, let’s go”. The development of YL is intimately associated with the evolution of Egyptian musical production. In 2002/2003, the popular Egyptian rappers MTM enjoyed a huge success with the hit song ommC msAfWa “My Mother is Going on a Trip”, which opened the way for the rap movement in Egypt (see Appendix, pp. 305–306). At this rather advanced stage in the expansion of YL, the song strewn with youth expressions had no trouble being accepted. From then on, rap songs proliferated and included many youth expressions and even youth titles such as the song performed by MTM, mAlak∫ fChA, meaning literally “you have nothing there”, and used by young people to say “you are hopeless, you don’t understand anything”, or the song by Aly, naffa| which means “sweep” but which, when labeled as a youth term, becomes “drop it” (see Section 14.3.3). The relationship between YL and rap, or more generally the hip-hop movement, was identified in many countries, such as the USA, Europe and several African countries like Gabon, Senegal or Algeria (Auzanneau et al. 2002; Miliani 2002; Moulard-Kouka 2004). In these countries, rap appeared much earlier than in Egypt where it is still a rather marginal phenomenon whose performers or audiences are upper- or middle-class young people, in contrast to the “Street culture” of American hip-hop. Compared with other African rap forms, Egyptian rap uses different methods of expression. If Senegalese rap tends to put emphasis on religious values (Moulard-Kouka 2004), Egyptian rap uses an “entertaining”, “lighter” tone similar to the values and themes of the American hip-hop movement (Alim et al. forthcoming). The role played by broadcast media is often criticized by the language control authorities. Television cultural programs respond to these criticisms by blaming teachers for their lack of effort in encouraging young people to use classical Arabic. Interestingly, YL is viewed by adults through the normative prism of the diglossic conflict. Indeed, YL is opposed not on the grounds of a colloquial standard norm (as, for example, adult use of ∞Ammiya that would represent the dominant social standard), but for the non-use of classical Arabic, which is rarely, if ever used in everyday conversation, but 294

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

remains the only recognized standard. The public emergence of YL is not in itself an enhancement of the colloquial form. The influence of the classical norm might be the reason why some practices which already occurred in colloquial Arabic are considered as deviant youth expressions. 14.2.3 What language for which young people? The impact of labels and categorizations “lu›at a‹-‹abÑb” and “lu›at as-s®m” A study of the YL of Cairo brings up the issue of the relevance of such a category in the society being considered. Which youth (young people) are we dealing with? Why call it YL if it is also used by not so young people? In fact, who are the “youth” in Egypt? The term YL is the translation of the Arabic expression luγat a∫-∫abAb. This label, developed in the broadcast and print media, is part of a youth category that includes youth movies, youth songs, youth fashion. These labels portray young people as members of a distinct group and opposed to adults. The expression luγat a∫-∫abAb is used by young people themselves to identify their way of speaking, and may confer authenticity to its existence. As Louis Jean Calvet said: “labeling the language, giving it a name is an acknowledgment, a statement for both the speakers and the linguist. The linguist’s view however, does not necessarily match the speakers’ statements” (Calvet and Moreau 1998: 25). In this study, we have chosen to reproduce the labels/categorizations of the actors without expressing a judgment on their validity from a linguistic point of view. One of the issues concerns the specificity and origin of the students’ YL as compared to other popular types of language, particularly slang and professional jargon. Thus, we find terms labeled as “youth” by Y. Hemaya in his dictionary (Hemaya 2002–2003)4 also mentioned in earlier studies about secret languages – luγat as-sCm – in Egypt (Issa 1988; Rowson 1983) such as the term fAW mewalla∞, literally “burned rat”, an expression now used by a number of students to speak of a “cigarette”, which originally appeared in thieves’ slang with the same meaning. Other expressions take different meanings, such as a∫fEW, which meant “empty shop” in thieves’ cant and “something falsified” in jewelers’ jargon; the expression is now used to speak of “someone without money”. In the same way, yAfet used by jewelers to designate “a rich customer”, is used by young people today to refer to any well-made or good quality object (for example, a tee-shirt may be yAfet). These examples emphasize the relativity of the youth label as given to a number of expressions, and highlights possible linkages with the secret languages of some social groups (dancers, prisoners, etc.) or the professional jargon (jewelers, mechanics, electricians, etc.). It should be stressed that links between YL and popular or corporate argot exists in many countries where slang expressions, sometimes obsolete, are adopted by young people as their own (Goudailler 2001). 295

SHERIN RIZK

14.2.4 A Cairene student practice? The links between YL and cant/secrete languages in Egypt would require further analysis to ascertain whether young people actually imitate the linguistic behavior of these groups generally identified as the bottom of the social scale.5 The point we wish to emphasize, however, is that our preliminary observations suggest that YL in Egypt seems to be largely a student phenomenon, whereas in large European cities the phenomenon is usually attributed to teenagers living in lower-class housing districts (Caubet et al. 2004). Studies of the speech patterns of young Egyptians have until now focused on the educated urban young, living in Cairo (Allam 2000; Peterson 2002; Rizk 2000, and our own research in progress). This does not necessarily mean either that this language is used exclusively by young students or that it is specific to those who live in Cairo. An in-depth comparison between the capital city and the provinces would be necessary to find out whether a YL is spoken in cities such as Assiout or Ismaïlia, including measurement of the role played by commercial media in the propagation of this phenomenon and the potential homogenization of linguistic behaviors. For example, we found that medical students at the University of Sohag may download Y. Hemaya’s dictionary from a website called Dactoor.cjb.net.

14.3 Dynamic linguistic processes The speech of Cairene students is characterized mainly by lexical and semantic processes combined with phonetic and morphosyntactic alterations. Our analysis reveals that the majority of the processes used by young students previously existed in Cairene Colloquial Arabic and cannot be considered as radical or specific innovations. In many cases, young people exploit previously productive processes in Egyptian Arabic (cf. metaphors, metonymy, rhymes, use of derivation, etc.). These processes are frequently and almost systematically used by young people and have therefore become markers of YL. Interestingly, the representations (YL as a dangerous and imported innovation threatening the Arabic language) are not in tune with the reality of the processes observed (a lexical register using the expressive possibilities of colloquial Arabic). 14.3.1 Phonological and morphosyntactic characteristics We observed two phonological processes: intonation and palatalization. Intonation is a major factor particularly for terms that have acquired new connotations. Thus, when asked to do something, a person may answer §en ∫A§ §allah “God willing” in a flat tone to signify “all right, I’ll do it”, as is generally the case in everyday speech. But when the same answer is pronounced with a high pitch with emphasis on the last syllable of §allah it has the opposite meaning of “I will not do it”. The intonation implies: “God 296

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

willing, but He will not”. The ironic intonations that characterize many youth expressions are used to create such antiphrasis. Cairene Arabic also has room for irony. What is new, however, is the high pitch melody that sometimes completely changes the meaning as in the expression kol sana wenta Yayeb, used in Cairo’s dialect to express wishes for the New Year, but also used by young people with a higher pitch to say: “drop it, let’s forget it”, as in el-moWatab? – kol sana w-enta Yayeb “my paycheck? – that’s it, I spent it”. The YL in Cairo is also characterized by a fast speech flow that leaves out quite a few vowels resulting occasionally in syllabic patterns that do not exist in Cairene dialect, such as initial consonant clusters: klot “breeches” (CA kolot) or msAfWa “is traveling” (CA mesAfra). In Cairene, a number of middle- or lower-class female social groups tend to produce the t and d consonants and the more emphatic Y and | with strong palatalization.6 Discredited by upper-class social groups, this speech pattern is taken up and joked about by young people. For example, one of the students we surveyed displayed a bumper sticker on his car with the , reflecting the palatalized pronunciation γalYyet ∞omWC of words the contemporary hit γalYet ∞omWC or “The Mistake of My Life”. Among morphosyntactic characteristics, we should mention the grammatical transposition of some words, i.e. their change from one syntactic category to another. For instance, the noun Ya}n (literally “grinding”) is used as an adjective as in el-emte}An kAn Ya}n “the examination was difficult”. Also the term Ya}n changes from the adjective “difficult, terrible” with negative connotations, to the degree adverb “very” as in: el-emte}An kAn sahl Ya}n “the examination was very easy”. Ya}n is the most popular expression in a series which includes expressions such as xan§ “smothering”, ∫an§ “strangling”, fa}t “drilling”, Wadm “obstructing” and others: da felm xan§. “it is a bad movie”; da felm helw ∫an§ “it is a very beautiful movie”; wAd ∞abCY fa}t “a very idiotic guy”. mDt “death” is another noun used as a degree adverb, but not as an adjective: felm }elw mDt “very beautiful movie” and felm we}e∫ mDt “very bad movie”. The use of mDt as a degree adverb is neither recent nor specific to young people and can be found in many Arabic dialects. We noticed that the girls surveyed used mDt more frequently and avoided words such as xan§ “smothering” and ∫ an§ “strangling” loaded with connotations of violence, or others, more obscene, such as nBk “screw”.7 These expressions were considered by the students as “male expressions” not to be used by “wellmannered girls”. Even those who tolerate the use of these words (only 9 boys and 3 girls) insisted that girls should use them only among friends, so that they would not undermine their reputation. Another morphosyntactic characteristic is the use of the preposition fC/fe “in” before a noun to turn it into an adverb, as in fe-l-xabAsa “clandestinely, in secret”. This construction is found in literary and colloquial Arabic as in 297

SHERIN RIZK

fC al-§alan “in the open, openly”; fC al-§aXl “in the beginning”, etc. What is new is the repeated use of this pattern, and its occasional use as an adjective. For instance, el-yDm kAn fe-l-lazCz, literally “the day was in the good”, but meaning “it is a good day”; anA fe-|-|ayA’, literally “I am in a loss” but meaning “I am depressed and not able to concentrate”; enta fe-Y-YaWAwa, literally “you are in fresh air” meaning “you are disconnected from the world”. 14.3.2 Form substitution: truncation and initialing Cairene students seldom use formal processes such as word encryption and truncation or suffixation, contrary to other YL where these practices are widespread (Goudailler 2001; Miller 2004). In our study group, we found only one example of truncation: Bh en-neZ for Bh en-neZAm, meaning “What is the system” used to signify “How is it going?” or “What are we going to do?” Moreover, the expression Bh el-neZAm is part of the so-called communication rituals such as Bh el-axbAr, meaning “What news?” The regularity of such expressions has led to word and rhyme games in question and answer series: Bh el-neZAm? – lCn ∞eZAm “a disease affecting bones”. Bh el-axbAW? – el-axbAW zayy el-ahWAm “el-akhbar is like el-ahram”, a word game on the two major daily Egyptian newspapers al-AkhbAr and al-AhrAm. These puns based on rhymes are frequently found in colloquial communication rituals as well as in poetry and popular songs. We also found examples of initialing. These acronymic patterns are either used with a humorous connotation as for instance ∫Cn kAf ∞Bn (∫K¡) the initials of ∫aklak keda ∞abCY, “you seem to be idiotic” or used euphemistically to soften the effect of words socially proscribed and considered vulgar, such as e dabel et∫e (a}}) the English acronym of the Arabic exclamation a}}a an expression of protest and disapproval.8 We found yet another example among girls at the Department of French Studies at the Faculty of Arts and Letters. These girls take advantage of their knowledge of French as a foreign language less familiar than English. In the expression ∫aklC C U, they tried to reproduce the youth expression ∫aklC klDt, literally “I look like culotte (breeches)” meaning “it is embarrassing, I feel awkward”. To ensure the encoding of the word for “culotte”, they pronounce the abbreviation in French. The spelling of the abbreviation also enables them to avoid /ky/ and the vulgar connotations associated with its meaning in French, “ass”.

298

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

14.3.3 Substituting meanings: metaphors and metonymies Metaphors are the leading semantic mechanisms and de rigueur among the young people of Cairo. Here are a few examples: the expression naffa|telu literally means “to sweep with/for somebody”. However, it becomes “to get rid of”, so when a young person speaking of a subject matter says naffa| he actually means “drop the subject, forget it”. The “advice” often comes with a gesture of a hand shaking a part of the ear. Sometimes the gesture is enough to convey the message. Many of the metaphors used by young people emphasize the fun functions of language; they make up the most provoking and sarcastic expressions such as a series of phrases used to describe an unattractive girl. She “looks like a torn up sandal” ∫abah es-sandal el-ma§YE∞, or she looks like “a bus backing up” otDbCs WAge∞ be |ahWu, or she’s been “run over by a train” dAyes ∞alBha §aYr, or even, “she’s been shot with a firearm” ma|WEba b-en-nAr (i.e. “she is very ugly as though she has been bent out of shape in an accident”). We should point out that metaphors and similes are particularly well developed in popular colloquial practices in Cairo, especially when it comes to physical descriptions of men or women, sexual intercourse, etc. Metonymies are less common than metaphors. One example is the use of the term el-la}lE} meaning “Egyptian pound”, by extension “money”. The word comes from the verb la}la}-yela}la} “to move”. Money is la}lE} because it makes people move to help you. The meaning is explained by the proverbial expression popularized by the famous humorous writer Ahmad Ragab and taken up by young people: el-kol yeWE} maYWa} mA §aWE} mAdAm ba∫AweW b-el-la}lE}, “everyone follows me as long as I show them money”. While metonymies are not as widespread as metaphors in YL, the expressions created by this process offer the possibility of creating new ones based on the same semantic matrix. Therefore when young people refer to someone’s head as baYCxa “water melon” or ∫ammAma “melon” they are using a universal semantic matrix, comparing someone’s head with a fruit (or a vegetable) as is done in French with the words “lemon, pear, potato”.9 Shifts in meaning occur mainly through metaphors and metonymies although they frequently loose the original image or association; zobEn, literally “customer”, is used by young people to speak of “someone who can be easily deceived, naïve”. In the same way, the original meaning of the adjective Wewe∫ “insane” becomes “friendly, cool”. Starting from Wewe∫ and based on resources of the Arabic derivation system, young people have created the quadriliteral verb Wawwe∫ which means “to entertain, make more pleasant” and two nouns WawA∫a, taWwC∫, meaning “recreation, entertainment”.10 Such derivations are consistent with existing morphological patterns in the Arabic system. Indeed this is a very widespread and ancient process present throughout the Arabic-speaking world, particularly when

299

SHERIN RIZK

words borrowed from foreign languages are integrated into the language and generate derivatives. Many youth expressions are phrases taken from colloquial and/or classical Arabic and given a new meaning. The noun bC§a, which means literally “environment, atmosphere, surroundings, milieu” takes on the meaning of “popular” or even “vulgar” when spoken by young people.11 The verb YabbaxyeYabbax derived from the verb Yabax-yeYbox “to cook” means “to falsify, distort”: el-entexAbAt eYYabbaxet ∞a∫An }a|Wetu yeksab “the elections were falsified to benefit his majesty”. The new verb spawns a noun taYbCx “falsification” different from Yabx “cooking” derived from Yabax. Yet a few young people sometimes say Yabax with that same meaning, as in YabaxEha ma∞a ba∞| we akelEhAlC literally “they cooked it and made me eat it”, which means “they plotted against me; they set a trap for me and I was caught”.

14.3.4 Loans and neologisms One reason why the language of young people is regarded as a threat to Arabic is the significant proportion of borrowed words.12 Borrowed words rank third in Cairo as a source for YL behind morpho-phonological and semantic processes. English being the first foreign language in Egypt, borrowed words come primarily from Anglo-American. These loans may either appear in their original forms as: pCs (peace) ya mAn, a phrase used to greet someone or to agree with what he says kEl, from the English word “cool” besbes, distorted pronunciation of “business” xalCk on mateb∞A∫ of, / let you be on, don’t be off/ “be flexible”, “understand me” stAyel, from “style”, used as an adjective “elegant” me∫ fe-l-mEd, “not in the mood”, i.e. “not feeling well, disappointed” They can also be adapted to the morphosyntactic Arabic system. They then form principally quadrilateral verbs such as: tan∫en-yetan∫en “be tense” from English “tension” as in ma∞lB∫ anA mtan∫en ∫ewaya “sorry I am a little tense” kansel-yekansel, from English “cancel” haypar-yehaypar, from English “hyper”, i.e. “to hallucinate” or “to take drugs” santar demAγak, /center your mind/ i.e. “concentrate” 300

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

sakWet-yesakWet, “secret”, i.e. “to keep secret, to keep to oneself”, cf. xod el-felEs di solfa bas ya∞nC matsakWet∫ ∞alBhA “hold this money but only as a loan” estmDWning, portmanteau word which imitates the colloquial phrase estebA}a meaning everything that someone drinks (or eats) early in the day (first cigarette, first coffee) sBstemak Bh, from English “system”, i.e. “What are your plans?” Young people also use the verb hanneg-yehanneg derived from “hanging”. This word reveals the influence of new technologies on the linguistic practices of young people, the primary users of new technologies. The English word hanging, used to speak about a computer that has crashed, is used by young people to talk of someone who refuses something and does not want to change his mind. Therefore he seems to be “crashed as a computer”. We were also able to gather examples of French calques among the students of the French department. These calques were probably only known to and used by them since they were “youth” Arabic words and expressions translated literally into French as jokes such as ramasse le tour “collect the turn”, word for word translation of lamm ed-dDW which means “stop the silliness”, or assied toi dans un endroit sec “sit in a dry place”, literal translation of the youth Arabic expression o§∞od fC }etta nA∫ fa used to make fun of someone and ask him to keep to himself and mind his own business. The expression implies that the person is as stupid and immature as a child who wets his pants and the spot where he was sitting. As for huit kilos d’oreilles “eight kilos of ears”, it is a calque of the youth expression tamAnya kClo wedAn which refers to someone who is curious and tries to listen to everything that people say. In the same way, as a joke or as a code, French-speaking students speak about a foul-mouthed or graceless person as environnement from the youth Arabic word bC§a. Others use the word atmosphère, which leaves the door open to further calques using synonyms of the same word. Connivance is a significant factor and these calques can only be understood and deciphered by a small group. Only the young French-speaking people of Cairo can grasp their meanings. The comic effect created by speaking French takes the tension out of the situation and diminishes the power of expressions which might otherwise be taken as insults. Thus, the use of French has a euphemistic effect and may also help lessen the stigma brought on by the use of a “youth” expression. “Neologisms ex-nihilo” rank fourth on the list of lexical innovation processes. Presumably created out of nothing, these neologisms can neither be classified as procedures of semantic and/or formal intervention over existing words nor as easily identifiable loans. The role played by the media in the diffusion or creation of these neologisms is undeniable. As we mentioned earlier, the words ka}Wata and kAmannannA were made popular by a successful movie. The ∫ankEtC word comes from the title of an Egyptian song. In the song, the word refers to 301

SHERIN RIZK

everything that can be bad in a man, i.e. a “guy”. The use of the word ∫ankEtC makes sense only in a given context where it is possible to determine its nature. Therefore the singer describes a liar, a woman-chaser, a careless guy etc., as a ∫ankEtC. In the same way, the adjective moftakes, created exnihilo, means both “sad” and “happy”. Based on this adjective, a new noun is created – eftekAsAt – as well as the verb eftakas-jeftekess with the meaning of “to invent, to create”. The importance given to the context seems to be one of the characteristics of the language of young students in Cairo, and includes a few ambivalent or ambiguous terms. For instance, terms that existed in colloquial forms have acquired opposite meanings to that of the original, such as the adjectives fazC∞ “terrible”, ba∫e∞ “horrible”, moXCba “catastrophic”, etc., used by young people with both positive and negative connotations. Artistic productions play a part in defining new terms that are frequently ambiguous. Hence, the group MTM’s song Wawwa} badrC “return early” begins with a definition of the term eftekAsa as the invention and creation of stories or pretexts to do or to avoid doing something. But later in the same song, the singer admits that he may not really understand the term. What matters here is to point out the ambivalence of many youth Arabic terms. The use of ambiguous terms is very widespread in other non-Arabic YL such as the adjectives “terrible” or “grave” in French. The phenomenon also exists in (classical and colloquial) Arabic where it is known by the name of a|dAd (Cohen 1961). But, here again, the young speech patterns systematize and ritualize the process. Most processes used by young people are not entirely new and can be traced back to Cairo’s colloquial forms. We may therefore wonder why the language is stigmatized, since the processes used by young people do not really transgress the structure of the Arabic language. Several theories are offered. First, young people systematize and valorize linguistic processes that remain less spread in everyday speech, and the accumulation of these processes characterizes the lexical specificity of YL. The main reasons however, are undoubtedly of an extra-linguistic nature. By pushing the fun side of the lexical and semantic innovation to the extreme, young people also express a craving for freedom. The controversy is about the new “subversive and provocative” values and ideals claimed by these young people and expressed in such terms as nAm w-eWtA} ya∞tCk el-nagA} “sleep and get some rest, success will come to you”, or zAkeW tenga} γe∫ tegCb magmE∞ “if you study you’ll succeed, if you cheat you’ll get a good grade”. Finally it should be stressed that while distancing themselves from “high speech patterns” and favoring the expressions and pronunciations associated with popular classes, young students transgress their society’s barriers and social taboos. The fact that their language is perceived as “daring” and “vulgar” is therefore not surprising, knowing that this type of value judgment about younger generations is an unvarying feature throughout history. 302

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

14.4 YL as seen by the young students of Cairo But what is the attitude of the students toward this new language practice? Based on semi-directive interviews of forty young students we considered three following criteria: 1. 2. 3.

Using versus not using this language Positive versus negative attitude towards this language Academic department of enrollment

Although our sample was limited, the investigation made it possible to postulate a number of hypotheses which further investigation may or may not confirm. We summarize our results based on two main hypotheses. First, whereas all our students belonged to relatively homogeneous milieus from a social point of view, it is obvious that different educational curricula lead to behavioral differences, depending on the prestige and hierarchical rank of the various academic departments. The majority of those who said they use this language and feel positive about it are Engineering and Medical students. These students are perfectly aware that when they speak “youth”, they deviate from the standard but they are proud to do so. Some even claim credit for initiating some lexical innovations. This positive attitude can be explained in terms of “linguistic security” by virtue of “social security”. These students attend highly hierarchical schools and either have a prestigious social identity or at least do not have a stigmatized social identity. They do not feel that the way they speak is an issue. What seems to confirm this observation is that, at the opposite end, the majority of those who use this young language while feeling negative about it, are students in Arts and Letters and Commercial Studies departments whose reputations are much less prestigious. The socially ranked educational system engenders among the latter a sort of linguistic insecurity which results

Table 14.1 Use and attitudes towards YL among 40 students Faculties

Use and positive attitude

Use and negative attitude

Non-use and negative attitude

Medicine

4 males 5 females 5 males 3 females 2 males 2 females 3 males 2 females 26

− 1 male − − 2 males 1 females 2 males 3 females 9

− − − 2 females 1 male 2 females − − 5

Engineering Arts Commercial Total

303

Total

10 students 10 students 10 students 10 students 40

SHERIN RIZK

in the stigmatization of this speech pattern that they do not (or claim that they do not) wish to use. For instance, a female student from the Faculty of Arts and Letters showed us a paper where she had listed a number of lexemes of this YL under the title al-alfAz as-sEqeyya “popular/vulgar expressions”. She told us that as she went over the list again she found these words mo§Wefa §awC “very disgusting”. There may be another explanation of the differences in attitude between the students. In Egyptian culture, as in other cultures, being a good student means being a serious person. One’s self-image is then built around the concept of serious person, based on a much debated universal notion of the serious person as lacking a sense of humor. Thus the humorous and nonserious characteristic usually associated with this way of speaking serves to diminish the negative image of the studiousness of engineering and medical students. Second, we noticed that negative attitudes and normative behaviors (nonuse) are found mostly among female students (four out of five surveyed). This seems to agree with the theory of W. Labov (1998) and many other linguists, among them R. Lakoff (1975), about the more normative, more polite and more conservative aspect of women’s language in so-called traditional societies. This would imply, again from a point of view based on Labov’s work, that YL was not yet perceived at the time of our survey as a prestigious variant leading to social advancement. But the nonrepresentativity of our sample and the fact that the forms of this language have significantly evolved these past few years rules out any hasty conclusion. Indeed it would be necessary to undertake a more complete study on the linguistic behaviors of girls in Cairo and on their feelings about the use of this language in order to validate the hypothesis that the use of this language by males is better tolerated than by females.

14.5 Conclusion Our survey touched on a linguistic phenomenon hardly ever explored in Egypt in spite of its recent impact on the Egyptian media scene. We focused mainly on the description of those features considered most characteristic of this language and shared by all the students we surveyed. The media promotion of youth speech seems to be part of a globalization process and for this reason, we may wonder to what extent YL is a trend imported and transposed into an Egyptian context. At the same time, it is interesting to note that the principal linguistic processes that characterize this language are relatively (but not completely) similar to those observed in other languages (metaphors, metonymies, calques) while at the same time, they contribute to discursive and humorous strategies, fairly common and appreciated in colloquial Arabic in Cairo, a city where humor plays a very important social function. 304

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

Many paths remain unexplored, including the difference between male and female practices or the influence of co-education on these social networks. Can the impact of this language lead to more significant changes from a linguistic point of view as well as in social behaviors, particularly for girls with privileged social backgrounds? What makes some girls in Cairo use a YL and run the risk of being deemed bad-mannered by members of the society? What about youth phrases considered as too obscene or vulgar to be used by girls? Are there changes in standards and self-image? Can we establish a link between the use of this language by females and its use by students of highly competitive schools? We have suggested that the latter do so to escape the stereotype: “only those deprived of humor are successful students”. Do females try to escape from the stereotype mentioned by Marina Yaguello (1992: 60), that “men have a sense of humor, women do not?” Our survey revealed that this YL seems to be used and recognized primarily by the young students of Cairo in Engineering and Medicine schools and that its practice is fashionable and essentially fun. The students from less prestigious schools (such as the Faculty of Commerce and Trade or the Faculty of Arts and Letters) seem to have a more normative and conservative attitude even though they also use the language. Here we have considered users who are students from the middle or upper classes. What about young people from lower classes who are not part of the academic world? How do they identify or not identify with this type of language or with the stereotypes associated with the category of “fashionable young people?” No survey has been conducted among those more underprivileged groups and it is therefore impossible to answer such questions. However, it should be noted that novels or movies (see the movie el-lembC or the novel abbAs el-∞abd), portray characters from lower-class backgrounds who use a “special language” close to the YL, where we find similar expressions alongside others more violent, and numerous puns.

Appendix Omni msAfWa From MTM www.kalmet.co.il en-nahawda f-ydm el-¡agaza ommc msafwa way]a ganaza waxaltaha l-]ad el-bab xabayetnc fekwa ostaza ¡andc ¡eda we ¡andc sc-dcz we en-nahawda ydm el-xamcs ya¡nc kolo fe-l-lazcz

mafa\el∫c ella et-tanfcz kalemt kol ex]abc ¡ala yel ma kan∫ fchom ]ad ma∫γel kolo kan beywod ye™el ya ™e∫ya ya pcs ya emma kel keda ma fa\el∫ ella ]aga bas afa\\c el-bbt men kol el-¡af∫ axl fc nas teb™a γawiya kasw we me∫ yalba te™leb be-™af∫ ]aybt fe kol ]eta sama¡a we ]aybt scdchatc fe bta¡a 305

SHERIN RIZK

kolo gahez l-el-eza¡a we lesa badwc fa\el sa¡a ™a¡at afakaw sa™alt nafsc ana lesa haf\al mestanc wo]t gebt el-mekwefdn waxxalto we ™a¡at aγanc

law el-kalam da me∫ ma¡™el ma¡le∫ ana el-mas™el ™olt a∫ef ba™a nafsc ∫waya ™a¡at aw™os we xhabc ]awalaya fag™a wamelc mayek fe-edaya we ™alelc γannc oγneya

Refrain: ommc msafwa we ha¡mel ]afla bas yawbt matgc∫ ¡ala γafla walla gcwanna tbalaγ ¡anna walla el-kahwaba te¡mel ™afla

el-]afla dc ana mestancha xat el-¡awabeya we gbt bcha men kotw ez-za]ma ]awalbha mal™bt∫ ]eta awken fcha wo]t wamcha ¡ala ew-waxcf we-l-manzaw kan bawa zawcf w-en-nas ketcw ∫aklaha n\cf banat lazcza damaha xafcf we gowa dah kan ¡alam tanc es-sama betmayaw aγanc w-ana wa™ef baw™ox fe makanc ∫oft ∫ellet banat ¡agbanc kan fchom wa]da a¡wafha bas makonte∫ bastazwafha ™olt ma¡le∫ azabayha ¡ala∫an ar™ox ma¡a x]abha w-ex-xob] yele™ ¡albna w-en-nas me∫yet men ]awalbna ™a¡adna watebna we waxbna ba¡dbn yabelna we γanbna

waxalt letakc es-sa¡a tamanya labes tc∫ewt fawc™ almanya baxbt la™bt bbte ba™a ]aga tanya we-l-]afla etmalet fe sanya mazcka gamda we kolaha gaz wap hawes . . . we kol di™c™a tzcd en-nas wana wa™ef ¡al-bab me]tas we fch nas bawa ¡amlcn seya] gaycn we m¡ahom sela] xalenc ba™bt me∫ mewta] wo]t ™afel bel-mofta] we katabt wawa™a ¡a-l-bab te™el mamne¡ doxel ay masyel

Notes 1 Our research started in 1999 and the first publications date back to 2000. We identified six studies: Allam (2000); Ezzat (2000); Hemaya (2002, 2003); Peterson (2002) and Lotfi (2002). Hemaya (2002, 2003) and Lotfi (2002) are glossaries of the YL, Ezzat’s psychosociological study (2000) includes a collection of youth expressions. Allam (2000) and Peterson (2002) are the only linguistic studies. 2 In Egypt the term fuX}A refers to literary as well as classical Arabic and does not reflect the difference underscored by linguists between classical Arabic (the language of the Koran), literary Arabic and standard modern Arabic – see Kamal Beshr (1997: 175–177). Here we refer to the Egyptian categories and keep aside the debate on diglossia which has been extensively studied elsewhere. 3 The term §e∫ta means “cream”. It is frequently used in Cairo’s dialect to mean good/well: yomak §e∫ta “good day” or bent zay el-§e∫ta “a very beautiful girl”. Young people however, use it simply to say “okay”. It is one of the three key terms observed at first and revealed by the media. The other two are rewe∫ and Ya}n.

306

THE LANGUAGE OF CAIRO’S YOUNG UNIVERSITY STUDENTS

4 Y. Hemaya is an author of paperbacks for young people. He was the first and only one so far to publish in Arabic what he entitled “a dictionary” of the youth lexicon. 5 Many writers speak of a reversal in the social pyramid which propels these professions to the top, and not the bottom, of the social scale. On the social mobility in Egypt see Ezzat (2000) and Amin (2001). 6 Haeri (1997: 78) speaks of a strong palatalization as characteristic of the Low Middle Class and of a simple (weak) palatalization as characteristic of the Upper Middle Class. In an analysis of the sociolinguistic representations of this phenomenon, Haeri noticed that the people she surveyed did not comment on the weak but only on the strong palatalization which, for them, was characteristic of “lower class” women. As Haeri herself says, this undermines her theory about palatalization as a “change from above”. 7 Expressions such as we}e∫ nBk and }elw nBk are very close to the English “fucking bad” and “fucking good”. 8 The sound a}} in interjections such as a}}a adds a coarse or vulgar touch because it is considered as erotic. These representations are seldom rational; as proof the interjection a}}e, meaning “shucks” also labeled coarse by the people of Cairo, though less vulgar than a}}a, is frequently used in Alexandria without an associated negative. However, Alexandrines, especially females, who arrive in Cairo are usually advised to avoid the interjection. 9 On the semantic matrix and how “it makes it possible to produce new words ad infinitum”, see Calvet (1994: 35–42). 10 Gemination ensures the difference between rawwe∫ “to entertain” and rawa∫ “to drive crazy”. 11 In the 1990s, the brand BMW, considered as a “rich person’s car”, was interpreted by some young people as the initials of BC§a MaXreya WAYya, meaning “vulgar (or corrupted?) Egyptian milieu”. After it spread among young people, the expression was reduced to bC§a waYya “vulgar, low class milieu” and later to bC§a, “vulgar”. 12 Peterson (2002) however, notes that the proportion of loans is relatively weak in her data (although it was collected among students of the American University).

References Alim, H. S., Meghelli, S. and Spady, J. G. (forthcoming) Enter the Cipha: Hip Hop Culture and Consciousness in Global Context, vol. 1, Philadelphia: Black History Museum Press. Allam, J. (2000) “Youngsters and new vocabulary”, Al-luγa 2: 111–135; Cairo. Amin, G. (1998; 2nd edn 2001) Madha }adath lil-maXriyCn, Cairo: Dâr El-Hilâl. Auzanneau, M., Bento, M. and Fayolle, V. (2002) “De la diversité lexicale dans le rap au Gabon et au Sénégal”, La Linguistique 38/1: 69–98; Paris. Beshr, K. (1997) ∞ilm al-luγa al-ijtimA∞C, Cairo: Dâr Gharcb. Calvet, L. J. (1994) L’argot, Paris: Presses Universitaires de France. Calvet, L. J. and Moreau, M. L. (eds) (1998) Une ou des normes? Insécurité linguistique et normes endogènes en Afrique francophone, Paris: Diffusion Didier Erudition. Caubet, D., Billiez, J., Bulot, T., Léglise, I. and Miller, C. (eds) (2004) Parlers jeunes, ici et là-bas. Pratiques et représentations, Paris: L’Harmattan. Cohen, D. (1961) “A\dad et ambiguïté linguistique en arabe”, Arabica VIII: 1–29.

307

SHERIN RIZK

Elghazaly, S. (2000) “]adcth al-re]”, al-AhrAm, January 11, Cairo. Elshaer, H. (1999) “Alfath maγleta’, al-AhrAm, November 2, Cairo. Esmaïl, S. (2002) “Al-faris alladhi qatal jiwaduh’, al-AxbAr, September 25, Cairo. Ezzat, A. (2000) Al-ta}awwElAt fi a∫-∫axXCya al-maXrCya, Cairo: Dâr al-Hilâl. Farag Allah, A. A. R. (1996) “La question de la langue dans la presse égyptienne”, Egypte/Monde Arabe 27–28: 435–449; Cairo: Cedej. Goudailler, J.-P. (1997; 2nd edn 2001) Comment tu tchatches. Dictionnaire du français contemporain des cités, Paris: Maisonneuve et Larose. Haeri, N. (1997) The Sociolinguistic Market of Cairo: Gender, Class and Education, New York: Kegan Paul International. Hemaya, Y. (2002–2003) QamEs riwi∫ Ya}n, vols 1 and 2, Cairo: Dâr al-Hossam. Issa, A. (1988) Al-luγAt as-sirrCya, Cairo: Intissar Press. Labov, W. (1998) “Vers une réévaluation de l’insécurité linguistique des femmes”, in P. Singy (ed.), Les femmes et la langue, l’insécurité linguistique en question, Paris– Lausanne: Delachaux et Niestlé, pp. 25–35. Lakoff, R. (1975) Language and Woman’s Place, New York: Harper and Row. Lotfi, S. (2002) QamEs al-gihbAz al-riwi∫, unpublished MA thesis, University of Cairo. Miliani, H. (2002) “Cultures planétaires et identités frontalières. A propos du rap en Algérie”, Cahiers d’études africaines 168: 763–776. Miller, C. (2004) “Un parler ‘argotique’ à Juba, Sud Soudan”, in D. Caubet et al. (eds), pp. 69–89. Moulard-Kouka, S. (2004) “Le rap à Dakar. Approche sociolinguistique du langage hip-hop au Sénégal”, in D. Caubet et al. (eds), pp. 111–126. Peterson, J. L. (2002) “Contemporary Cairene youth terminology: linguistic deviation or social art?”, in Youssi et al. (eds), pp. 422–429. Rizk, S. (2000) “Le parler des jeunes en Egypte”, unpublished MA thesis, University of Cairo. Rowson, E. (1983) “Cant and argot in Cairo Colloquial Arabic”, Newsletter of the American Research Center in Cairo 122: 13–24. Yaguello, M. (1978; 2nd edn 1992) Les mots et les femmes. Essai d’approche sociolinguistique de la condition féminine, Paris: Payot. Youssi, A., Benjelloun, F., Dahbi, M. and Iraqui-Sinaceur, Z. (eds) (2002) Aspects of the Dialects of Arabic Today, Rabat: Amapatril.

308

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

15 RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA) Aline Tauzin

15.1 The emergence of Mauritanian rap Our study will focus on rap music composed and sung in the Hassaniyya Arabic dialect currently spoken in Mauritania among the Moors.1 Mauritania is a multiethnic country where several languages co-exist; rap sung in Hassaniyya appeared in the early 2000s. Lyrics in the other languages spoken in the country such as Wolof, Pulaar or Bambara, had emerged a few years before, but Soninké is still missing from the local scene. Rap is the favorite form of expression of the young people who live in large city working-class neighborhoods and surrounding shanty towns. Its very existence and contents signal the beginning of significant changes, both in the status of music and the traditional organization of society. In Hassaniyya, the singers are known as Wap¢W (pl. Wap¢WAt), a borrowing from French. Their musical practices are at the cross-roads of external and local influences. Rap music, primarily American and Senegalese and, to a lesser extent, French, was widely broadcast at a time when Mauritanian society was undergoing major changes that can be summarized in one word: urbanization. But the appeal and eclectic appropriation of musical forms are long-established practices in this part of the world. American music has been performed for several decades in West Africa, after it was introduced by soldiers who fought in support of the French colonial power during the Second World War. Jazz bands soon came to life in Dakar, Saint-Louis or Bamako. Soon after, Afro-Cuban music was a huge success in the same cities.2 As a result of various migrations, bonds ultimately remained strong with the United States primarily, and also with France. Several singers said that, in the 1990s, they listened mostly to American rap, then to French groups, among them NTM and IAM. Most ins (instrumentals)3 come from North America, some come from France. Rappers also emphasize the legacy of their “elders”, the generation known as “old warriors” who paved the way for them. Of the latter, they say that, 309

ALINE TAUZIN

they were very bright and well-educated young people. Some stayed in France or in America, others came back to Mauritania and tried to keep up the same way of life. They sang and danced like Americans. We learned from them. Later on, when we heard that the French did rap, we sang like them. The names of the groups or singers are also influenced by the American model: B.O.B. (combined initials of the three members’ names, Bocara Ba, Ousmane and Boundaw); Positive Attié Black Possi;4 M.D. Max; Overdose; D.J.H. (abbreviation of Hachim); Dental Men Possi (dental means “we meet”, in Pulaar); M.C. Mandela; W.P. or Black Snakes Possi. The geographical proximity of Senegal and the extraordinary development of rap music there also played a part in the emergence of Mauritanian rap initially expressed in Wolof and Pulaar, two languages also spoken in the neighboring country. The conflict that opposed the two nations in 1989 also left a mark because it resulted in the return of emigrants to their respective countries. In Mauritania, the “haratines” (see Section 15.2) were among those who came back and played an important part in the emergence of rap. Many of them had been born in Senegal and all had been clearly influenced by their adoptive society. So much so that in Mauritania, they were referred to in terms of transformation or changes such as mestekwar “turned kwAr” – kwAr means Black in Hassaniyya – or metgelleb, from the verb gelleb “to change”. At the same time, Moor society was experiencing a great deal of deep changes. The drought in the 1970s precipitated a process that had already begun on a large scale since the colonial period. It led to a massive settlement of nomads, followed by an increased concentration of population in cities, primarily in Nouakchott, the capital city, and in Nouadhibou, the economic hub, just as the farmers in the Senegal river valley were affected by a widespread rural exodus. Growing city neighborhoods were quick to set themselves apart from one another, as individuals who belonged to the same social groups gathered in increasingly homogeneous areas. Whereas Moor society remained traditionally hierarchical, life was quite different from that of yesterday’s camps; population density bred protest, where rap music found its place.

15.2 Urban protest and the break with the past: rappers and Griots Rap’s main theme is the condemnation of corruption. Within Moor ethnic groups, it is produced only by the Haratines. In the past, the name Haratines meant dependent, specializing in agriculture and rather attached to the Marabout noble tribes. In contemporary language, the Haratine group includes former slaves. Rap music however, is not the exclusive mode of 310

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

expression of such groups at the bottom of the traditional hierarchy. It is the language of the poor, of the “galériens” (lit. galley-slaves from the French slang word “galérer”), of those who face the daily challenge of survival in “hell”. Rap is not differentiated among the various ethnic groups either. On the contrary, it claims to be a unifying and patriotic element within the Mauritanian nation. It is expressed in several languages and each rapper frequently uses two, even three languages: one or two are local, the third is French. Groups often include singers who master all languages spoken on Mauritanian soil – all that is except Soninké, as mentioned above – in an attempt to reach and rally all young people. There is no claim of membership in specific ethnic groups, and in fact, such issues are rejected. Actually, a number of rappers are the products of mixed-raced unions, another sign of their modernity.5 Since most have had little formal education, they often acquire their multilingual skills at home, but also in the street or, as far as French is concerned, by viewing series on French television channels. The use of multiple languages is widely promoted, as one of the rappers said: “You will often hear me say pentagon. Pentagon, it is because of the five languages.” Another break with the past is the fact that those who sing do not belong to the old-established group of musicians. Among ethnic Moors – as among all other groups – only the Griots had this established right. Indeed, part of the musical repertoire was the prerogative of dominated groups, such as the women slaves’ “crushing songs” for example, or songs of praise to the Prophet performed by the same group. If they were performed only in specific circumstances, their existence related back to the very elements of social stratification. Therefore, “crushing songs” were a background to the work performed exclusively by maids and, although they criticized masters, they were – through their flat and repetitive nature – a means of escape rather than an instrument of protest (Tauzin 1989). The praises to the Prophet intended to prove to the masters that the performers were knowledgeable about Islam, that they were true Muslims, and that, as a result, they should be emancipated. Rap breaks up that musical production hierarchy and defines a different hierarchy. No one can be kept from performing rap, at least in theory. The themes associated with rap music also confirm the break. The Griots’ music referred to specific social relations, and in part, continues to do so, even if attempts are made to bring the themes and forms up to date; the Griots’ music was intended to praise the aristocratic warrior groups to which the official musicians were bound and to disparage the honor of enemy groups. The repertoire was usually anonymous and transmitted from generation to generation and, when the Griots introduced new songs, they did so within the fixed framework of the same social relations and their duplication. It served as one of the privileged tools of such duplication. Rap music itself is not an inherited form. It is in fact the spontaneous expression of new social conditions. It arose within a dominated class and 311

ALINE TAUZIN

claims to be heard not only by its social peers but among all elements of society. A number of rappers even feel the need to sing in Hassaniyya in order to be understood by the politically and demographically dominant Moors. Moreover, the protest expressed through rap is quite different from a simple re-evaluation of social stratification and embodies instead a shift in meaning and an attempt at revising social stereotypes. Indeed, definitions applying to the various groups that form society have always implied elements of an essentialist nature. Thus, a warrior was by nature, courageous, a Marabout, self-confident, a Griot, shameless, a blacksmith, greedy and talkative, a slave, irresponsible and so forth. Oral literature played a major part in transmitting these definitions and perpetuating the social differences which it helped to justify (Tauzin 1993). The list included as many stereotypes as there were groups of affiliation. By exposing misery and condemning corruption, rap music, for its part, initiates a process of definition for the various social groups based only on their living conditions and, at the very end, a project of modernity, quite reminiscent of Marxist thought, even though rappers do not in any way identify themselves with such currents. Rappers, in reality, are not alone in the re-evaluation process. In the field of music, a number of Griots have also become involved in creating new sounds, developing new topics, with some even eventually distancing their work from their past masters, even if they had received payment. This goes up to the refusal of using the Hassaniyya word, iggCw (pl. iggAwen) “Griot” in favor of the title fennAn, or artist. This is because the word “griot” implies an attitude of submission toward the noble class and limits musical expression within the narrow frame of this relationship. This tendency however, has reached a limit as members of dominant groups among the Moors reject non-traditional musical expression based on a culture of personality. Public displays and the singing of lyrics to the beat of musical instruments are incompatible with traditional, upper-class values.

15.3 Break between generations Beyond disruption to the social order – at least as a perspective, through condemning its abuses – rap induces a reversal of the generational order. The “normal”, usual, course of interaction in traditional society is reversed, since criticism now comes from young people and is directed at their elders. It is interesting to note that there was a preliminary stage to the birth of rap that paved the way to its development. In the 1980s, a short-lived style named ∫abAt or t∫ebAt came into being in the neighborhoods surrounding Nouakchott. The name comes from the French verb “saboter” (to sabotage), and means locally “to make fun of, to ridicule someone”. Young boys would exchange rhythmic insults on the alleged faults of their adversaries’ parents, taken separately or together: a father might be scorned for his lack of virility, a mother for her sexual impurity, together they might be repre312

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

sented in a very crude caricature of the sex act. These insults reproduced forms already existing in the nomadic society and therefore follow a certain continuity. However, they also innovated by introducing the theme, as yet unknown, of family relations and alleged dishonor. We have suggested elsewhere (Tauzin forthcoming) the idea that what was at stake was not a matter of permanence, through the fundamental value of honor, but of suspension. It was no longer about an age group, as for traditional society, but about an emerging category of adolescents confronted with a range of instinctual drives, free from the constraint of legacy and consequently in a position to create. The break is also revealed in the clothes worn by rappers. Whereas the poorest classes in Mauritania have long given up the traditional boubou for worn out trousers and shirts bought at low prices in the fukudyey,6 rappers flaunt sweat shirts featuring names of American states or universities, as well as bandanas, caps and headbands. Only on the day of Tabaski, Aïd el-Kebir, some do wear boubous out of respect for tradition, to “please (their) mothers”. Those who can afford it may purchase, in the market, the perfect rapper’s uniform from overseas: narrow hat, clear plastic glasses, hooded sweat shirt, baggy trousers worn below the hips and sneakers. As a rapper commented: We dress like Americans. The only difference is the language. They speak English, we speak Hassaniyya. Clothes arrive in Dakar and are immediately available to us. We pay for that, even if it is very expensive. Selling used clothes to friends is also a common practice, which makes it possible to amortize high costs. The issue of clothing is far from trivial among the Moors. Keeping up with traditional clothing is essential within dominant groups, who manage to impose it on younger generations, even if these are in rebellion against tradition. Issues regarding male dress and the use of boubous, or even more so, of women wearing mela}fa – large cloths in which they wrap themselves – are inextricably related to issues of ethnic identity, social status and religious belonging; any attempt at revising these values is perceived as a betrayal of one or the other or all of these values (Tauzin 2006). The rappers’ choice of what looks like a uniform is an effort to express their will to break with traditional codes and deny the extreme poverty which affects their social group. Their choice of dress is evidence of a refusal to let poverty be perceived through their physical appearance. But new trends are emerging among some who appear set to reaffirm their Africanness, as much in terms of dress as in the use of musical sounds. At this time, it is difficult to distinguish among attitudes, between those who wish to distinguish themselves in a local context, and those whose dress 313

ALINE TAUZIN

codes express a longing for autonomy and a certain maturity with respect to the American model.

15.4 Writing and staging The writing of texts may be the work of an individual or of a group. In any case, it is an intense collective moment in community life. The young rappers and their friends usually rent a room where they gather during free time. They spend long hours there talking, laughing, even eating and rehearsing. Like clothing, the room is an important investment and is often furnished with care. Thus the task of decorating the room occupied by Papis Kimy and his “quartel” – i.e. his group, the Hassaniyya clan – belongs to one of his members. As in any Moorish house, mattresses are aligned along the walls. The walls are covered with large lace curtains and drapes, all bought at the fukudyey. Glasses and cups are lined up on small racks painted in white. Photographs of Western women, cut out of magazines, are glued to one of the walls. Lætitia Casta wearing a small camisole with one strap down holds the spotlight next to Britney Spears, Shakira or Mariah Carey. In the center of the room, a large round coffee table is also covered with lace. A number of pots with artificial flowers are scattered here and there. A smallish American flag is hung on one of the walls. We are told by the tenants that it is a reminder of the designer’s passion for an overseas rapper (Tupac). Then of course, there is a large blaster, stacks of cassettes and a CD player. Many people come in and out, some stay only briefly while others settle down for longer. We are not dealing with traditional society but with groups of young people, in the modern sense of the term; the prevailing rule of co-mingling is proof enough. Venues where rap groups can perform are limited and include some nightclubs, occasional stages set up in low class neighborhoods and the French Cultural Center where a stage is available every two weeks and where they can use computers to create sounds. Recording opportunities are even more sporadic. Control over broadcasting is impossible, a situation also condemned by Griots for many years. Finally, we must emphasize the near total lack of female rap performers in any language. There is one exception however: Sister Feuz, originally from Senegal, sings in Pulaar and is involved in clothes creation. The absence of female performers is undoubtedly due, for the most part, to the extreme difficulty for a woman who does not belong to the Griot caste, to display herself publicly on stage.

15.5 Islam and social criticism Let us now examine the texts themselves. The first texts, chronologically speaking, are considered by rappers as “recreation lyrics” and are generally 314

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

referred to as free style. Some feel they hold no special interest and are only good for their rhymes. The performer sings his own praises and the lyrics have no social content: this is reason enough to reject them as nAz, a slang word borrowed from the French for out of date. Others however enjoy their rhythm and play on words. Texts created after the initial period and developed by the rapper community are dedicated for the most part to condemning social conditions considered disastrous. Mauritanian rap is positioned within society’s value system. It does not promote the destruction of society, but rather, its perfection. Therefore, it holds major moral implications, much like African rap in general. As one of the performers said, their rap is a “message to everyone”, their fight is “against xenophobia, racism, tribalism, tobacco, AIDS”. Rap music is the only means for young people to protest: “to be a revolutionary, to unveil all that takes place in our society, it is necessary to be a rapper”. This role underscores the significance of the message of selfrespect, about which many singers are adamant. “As teachers we must respect ourselves, we must set our limits.” “All that is antisocial is not unveiled brutally. We are not trying to divide people, but unite them.” It is a matter of being taken seriously by society as a whole and not just by their peers. Coherence of ideas is a key issue in the search for recognition that motivates these rappers, whom society disparages as hooligans and delinquents. Their message is based on respect for the rules and proscriptions of Islam and their attack is aimed at those who would transgress it in spite of what they may say publicly, and without regard to their social standing: A self-proclaimed Muslim cannot be the thesis’ antithesis. A real Muslim is one hundred percent Muslim. You’d have to explain to me what a Muslim in the wrong is, a Muslim who does not respect what Islam prescribes . . . corruption and Islam can’t go together . . . You are a Muslim, you don’t touch alcohol, everyone knows that, you can’t misbehave sexually, you can’t be corrupted, you are incorruptible, you can do no harm, you must share your feelings. What kind of religion do Muslims represent when they use alcohol, when they prostitute themselves, when they are corrupted? That is no longer Islam. In spite of such positions they do not support fundamentalist movements, and, in fact denounce their hegemonic and totalitarian ways. “Appearance is misleading. Some people call themselves Muslims yet they are terrorists. Religion is something you keep in your heart.” Plurality, in its many forms, coexistence of differences, reconciliation, they sing their creed: “We rappers are here to reconcile people, to give future generations something to look forward to.” They wish “to reconcile all ethnic groups” which, according to contemporary usage includes black 315

ALINE TAUZIN

Moors (dependents and slaves of the traditional ruling classes) and white Moors, and they oppose the enforcement of Muslim rules on non-Muslims: If you consider yourself a Muslim, you must be a Muslim. But nationwide proscriptions cannot be enforced. The guiding principles of Islam are for Muslims, and these cannot be imposed on everyone, on foreigners for example. At the same time, the world is conceived only through absolute belief and everyone is expected to abide by religious principles, following the discourse produced by the Mauritanian society as a whole: “We live in a world of many races, many religions, many ideas. Each one of us must abide by the principles of his own faith”. “I am not against Christians, I am not against Jews, each one may practice his own religion. If you claim to be of such or such confession, you must follow through.”

15.6 The language standard and figures of speech What is striking in the style of the lyrics is not so much linguistic creativity as the use of expressions borrowed from a now obsolete Hassaniyya, reminiscent of nomadic culture. Old expressions, sentences and proverbs are frequently used, even though they are no longer used in urban language, much less among the younger generation. At the same time, a youth slang is emerging and is spreading among the elder generation. Rappers however rarely acknowledge the “recycling” of forms from the past and insist that such lexical choices are the fruit of their own creativity. The use of archaic forms was frequent in the performance of t∫ebAt “sabotage”, mentioned above. One might thus identify bits of wedding songs, quite defiant and occasionally slightly modified, although sometimes simply because of the age of the protagonists. Frequently, rappers were not aware of such borrowings and some are barely even aware they had heard such phrasing elsewhere. The search for words wherever they are to be found – in the streets, in the markets, among lost sources close to contemporary usage – enable them, at little cost, to come out ahead in the competition. We should note that the interest sparked by older forms of language tinged with nostalgia, is shared by many adults who themselves were confronted with the brutal change from nomadic to urban sedentary lifestyles. When arguments take place between city people and “bushmen” it is not uncommon to see witnesses stop to savor, in silence, and with a smile on their face, the subtleties of this language of the past, considered richer than the language of the urban space. Are rappers then like “saboteurs”, unaware or perhaps unwilling to admit as much? This is perhaps a reasonable possibility given that the moralizing contents of rap lyrics are greatly enriched by borrowed forms earlier used to 316

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

the same effect. The use of slang or the invention of a new language would only further discredit rap music as a voice for mainstream youth. In this sense it is important to use a language as neutral as possible and accessible to all, and which does not interfere with the rap message. On the other hand, it appears to us that creativity is infinitely greater when the texts are in French. Paronomasia (punning) is used frequently in rap. Rhymes are sometimes created using local pronunciations without consideration for spelling, sometimes even words are used with a purely Mauritanian meaning. Which does not prevent the authors’ intensive use of word and rhyme dictionaries.

15.7 Samples of rap texts Most of the excerpts quoted below were written by Papis Kimy,7 considered as the best rapper in the Hassaniyya language. The first excerpt, entitled Conflit (Conflict), includes a number of themes in various languages (the example quoted below is from the Hassaniyya). The first topic is about a competition with another, either real or imaginary. The text then affirms a commitment to both homeland and Islam, and exhorts the other to reject corruption and to preserve the power of his speech and purity of his condemnation. (Another example, which we are not able to reproduce for lack of space, expresses the difficulties of living for those intent on singing.) The drive to succeed and to win over an audience is an important factor in the choice of language as French makes it possible to reach an audience well beyond the national borders. The desire for success is tied to a love for poetry and music, an awareness of others’ success thanks to media play, and their need to find the means to survive in a very difficult environment. dLey u em ay si representi crazy Kimy biyye elli bC1Ani Wappi be l-kalAm el-}assAni Bye wa}de tneyn reprezen kimy man klAmi zeyn klAmi zeyn ixallaX ed-deyn u isowwi el-ma∫Akel beyn θneyn ente OAk el gelt ∞elen eW-Wap rïm lïlak Ane ba∞d kïmy mA vett YWabni biyye elli mA inehwenni mAhu klAm waYani Ane mDrïtAni miyye ve l-miyye met∞allem ellA be l-ugiyye na∞rav ∞lA∫ en-nAs 317

ALINE TAUZIN

ta∞Yi el-hediyye w ed-diyye hemmak ente tXDbi ∞liyye ∫weyy le-klAm el-}assAniyye elli mAhu metwAzen même pas m∞a lEqa el-aXliyye }Ez hDk ∞anni ∞essak }atte menni eW-Wap huwwe venni ba∞∞ad lA tgerreb menni Translation D. J. and MIC represent Crazy Kimy because I am a Moor My rap is in Hassaniyya Eye one two represents Kimy man My words are beautiful, my words are beautiful, They pay back the debt8 They solve problems between people9 You who say you’re master of Mauritanian rap I, Kimy, was not pleased I was not transported These are not nationalistic words I am one hundred percent Mauritanian I only know the ouguiya10 I know why people give the offering and price of blood11 You are trying to steal from me Some of my Hassaniyya Which does not even rhyme12 With the national language Keep away from me Beware of me Rap is my art Keep off do not get close to me. The following text draws on a number of sentences and proverbs from the traditional corpus. le-∞du mA yurabbe ye owleyd el-∞amm ∞essak men ∞dûk elli yaxetlak lAhi yeketlak lA tevli∫ mAhu WAXak waÙÙa elli idAve∞ ∞annak 318

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

bïh elli e∫beh le-mdAve∞ lA men elli yetkellem u e∫beh er-raLLAle ellA elli menhum m∞Ak er-raLLAle elli mA yenv∞Ene ilA mAtu mA yowL∞Ene elli lAhi yaxed∞ak ta}kem ∞annu WAXak mA YYï} e∫beh tdïr WAXak beyn eW-WEX u ted∞i be s-slAme axeyr men l-ihAne eywe a1baY Xa}ab le-m}Ali men lAhi yelgï lak ∞essak lA yedxel sebïlak ∞rav gA∞ elli mθïlak izïd Onûbak w iga11ak ∞an XA}ab el-ma∞lEm w i∫ïlak lA taX}ab yekEn el-∞angaW u lA tetkel yekEn ∞lïh bïh elli xaYY eY-Yab∞a idegdeg YEbak Translation Enemies, they are not raised up with you13 O my cousin be wary of your enemies He who chases after you will kill you14 Only trust yourself Or he who defends you For he who defends you is worth more that the smooth talker And the best man is the one who is with you15 Men who do not give us anything When they die we do not suffer16 He who may betray you Do not trust him So that you don’t fail Better stick to worthy people17 and ask God for peace Than be mistreated You must know how to tell Evil’s friend From he who will help you18 Make sure he does not walk next to you Know who looks like you 319

ALINE TAUZIN

He increases your sins He belittles you whereas your worthy friend Lifts you up Be friends only with the best And rely only on him19 Because the mark of a good education Is that it can be fashioned20 The theme of love is only approached and, indeed, acknowledged by rappers insofar as it carries a political message and makes it possible to denounce one of the society’s rules of marriage. That rule is female hypergamy, which requires that a woman marry only a man with a higher social status. Any other union is prohibited, as well as any kind of attraction between two people who do not fit into the patrilineal scheme imposed by society. Here is an example:21 Ane ∫enhu blA }obbi ve ram∫et ∞ayn xa∫∫et Oi eY-Yovle gaÙbi xaWWast el-∞arbiyye w el-¡asriyye mA Wayt mAhu entiyye dA§imen nzeyyen vïk en-niyye }obbi mA itgeddem ∫Dr ehlek biyye elli eswed kDri walla }arYAni men gedd ehlek lAhi ntemm ellA ∞abd ment xayme kbïre lA txammem vïhe kell sba} mAreg ed-dAr v el-lebne mAhi miyye Oi elli txammem bïhe }obbet gaÙbak mAhi hiyye Translation What am I without love With one glance this girl entered into my heart I looked to the right and to the left and saw only you My intentions towards you are pure But my love cannot be22 because I am black Black or vassal, to your parents I am but a slave, A girl from a well-to-do family, do not think about her Every morning when I go out I don’t have a penny in my pocket She who is in your thoughts, the love of your heart, it won’t be her. The following final example is French rap, also written by Papis Kimy: Dis la réalité même si elle t’en coûte la vie Papis Kimy représente Médina tree23 320

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

Les lœufs24 galèrent ici 90% ne sont pas instruits Exprime imprime mes rimes viennent de RIM25 Ma vie est fiasco le marasme c’est mon ghetto Y a trop de marmots manchots Notre magot c’est nos scenario Nos micros sac à dos et stylos Même pour fumer on est là à ramasser des mégots Amigo here we go y a des go26 prostituées à gogo Qui se font comme boulot jouer des films porno à cause du pogno Chez moi c’est le cago27 Le chaos règne le sang saigne Qu’est-ce que tu veux que je t’enseigne Quand on est sale on se baigne Y a trop de dégâts rien n’est méga Méga mes immédiats28 je suis la mass media J’ai grandi dans la lerga Entre la colère et la joie Près des vieillards lascars soûlards Qui ont comme fonction fortune et pouvoir Cambrioler, agresser, voler, violer Je suis simple dans mes lyrics On est à la proximité de l’apocalypse Les jeunes se perdent en conjectures de plus 30% atteints déjà par le virus La police et la justice Les parents ne peuvent plus rien contre leurs filles et fils Qui ont comme vice Etre taspé29 ou drogué Trop de grasses matinées Les lœufs n’ont plus le temps de prier A plus forte raison se confesser O mon Dieu de ce monde hypocrite Je vous conseille d’en avoir des limites Car y a des gens qui zimittent30 Et les couples en foyer concrétisés par des enfants illicites Ils en ont rien à cure de la Bible qu’ils parcœurisent Ou le Coran qu’ils récitent Y a trop de rivalités de divorces et de mariages forcés

15.8 Conclusion Similar to what we may observe in other societies (Hadj Miliani 2002: 767), Mauritanian rap is both product of a recent urbanization, specifically that 321

ALINE TAUZIN

of the Moor ethnic group, and at the core, expression of former subordinate groups. It is evident that rap music can only really develop within population concentrations and that such concentration triggers the re-evaluation of long-established relationships and the emergence of a more universal message, based on issues specific to the lower social classes. In a comparison with other rap forms performed on the African continent, Mauritanian rap unquestionably shares many common characteristics with Senegalese rap – probably due to the similar way in which the two countries developed rather than to their geographical proximity. Particularly noteworthy is the fact that, until the emergence of rappers, both in Mauritania and Senegal, Griots had a monopoly on the production of songs. Any person who did not belong to the established group of musicians was excluded from it – which reveals the enduring significance of that traditional form of social organization. In such a context, rap music gives meaning to a break which is both sudden and easily identifiable through the challenges posed to established order and group specialization, and which inevitably affects social and generational hierarchies. Two identical characteristics are evident on either side of the river separating the two countries. The first is the idealization of an earlier form of language, which Sophie Moulard-Kouka (2004: 121–122) says – speaking of the Wolof language in Senegal – meets the approval of “elders, who see in it an attempt to return to traditional values”. In Mauritania, rappers hunt for forgotten proverbs whereas linguistic innovation is only possible in a foreign language, i.e. in French. What is the actual purpose of this archaeological work on language? What can it tell us, for example, on the activity of rappers in a modernizing context? The second characteristic is the religious message of rap, which defends the ideals preached by Islam and condemns the behavior of public officials who belittle their religion while claiming to support it. These two characteristics, specific use of language and reference to religious ideals, would definitely need further study. A comparison with Afro-American rap, including its Islamic component (Béthune 2003), would undoubtedly be enlightening. We feel, however, that it would be complete only if associated to a diachronic definition of the culture specific to the rap producing dominated groups – in Mauritania, former slaves, may we recall – using the same approach as analyses relating to the United States of America (Béthune 2003; Martin 1998).

Notes 1 This study is based on research completed in Nouakchott, in 2001 and 2002– 2003. At the time of the second stay, there were only two “official” cassettes of rap in Hassaniyya, one by Papis Kimy (released in March 2002 in Senegal) and one by D.J.H. They had been recorded in studios in Dakar and were sold in a number of standarts, shops specializing in the sale of recorded music.

322

RAP AND RAPPERS IN NOUAKCHOTT (MAURITANIA)

2 During the entire period of colonization, Mauritania was part of French West Africa whose capital was Saint-Louis of Senegal at first and Dakar later on. Nouakchott was founded as capital of the future state in the very last months of French presence in the area. 3 It is a pre-recorded beat to which the singers add their lyrics. It makes up for the lack of real music, due to the extreme poverty of the rappers, but it is very restrictive. 4 Possi is an American term ( posse) that means a group of friends, for which Olivier Cachin (1996) suggested the translation, crew or set. 5 The statistics on the number of mixed-race unions in Mauritanian society is impossible to calculate. Their very existence is hardly mentioned. 6 The word, borrowed from Wolof, means both second-hand goods and the shops where they are sold. Most clothes come from the United States. 7 This rapper’s choice of name is atypical: indeed, it is a childhood name – not an imitation of American models. Papis is the name his relatives call him, though his actual first name is Harouna. Since there were several Papis, to avoid confusion he added to his first name the name of the dog he had adopted and with whom he walked all the time: Kimy. 8 Here the singer uses a Moorish proverb which reads: le-klAm ez-zeyn ixallas eddeyn, “fine words pay back the debt”. 9 Another adaptation of a Moorish proverb: le-klAm ez-zeyn isowwi ma∫Akel bweyn θ neyn. 10 Name of the Mauritanian currency. 11 With this sentence, the author claims to be a good Muslim, aware of his religious obligation. In Hassaniyya, hediyye means the gift one gives to a Marabout or to the descendants of the Prophet. The price of blood, diyye, refers to the code of honor. 12 The first – unspoken – part of the verse should be “to do something”. 13 The expression means that one should stay away from his enemies. 14 A proverb saying that one should beware of bad people. A riddle uses the same words yaxtlak lAhi yeketlak, for which the answer is, sleep. 15 “The one who is with you”, traditionally implies tribe or relatives. 16 Here the proverb is used in the plural form and rendered: elli mA yenva∞ni ilA mAt mA yowLa∞ni. 17 A proverb that translates literally as: “Better put your head between heads”. 18 The proverb reads a∞rav Oe lli lAhi yelgï lak. The verb lge, whose first meaning is “to meet”, sometimes also means, as it does here, “hold the load on a camel while someone else fastens to it the luggage rack”. 19 This verse and the one before are part of a proverb which reads: lA taX}ab yekEn el-∞angWa “Be friends only with the top of the head (= the best people) u lA tetkel yekEn ∞lïhe And rely only on them lA ∞elle tedxel zangWa For fear of having to face a big problem ta∞geb be n-nedem ∞lïhe Later you would regret”. 20 The expression idegdeg eY-YEb, “clay can be fashioned”, means a quality person (obedient etc.), and is used in western Mauritania. The equivalent expression in eastern Mauritania is “his iron is malleable”. 21 This text was written by Memme, a young man who first wrote in French and later in Hassaniyya. Memme neither sings nor performs publicly his compositions. As many others, he has a mixed ethnic background and is a university student. 22 Literally this Hassaniyya text should be translated as “I cannot walk towards your parents”, which means, “ask in marriage”.

323

ALINE TAUZIN

23 Medina Three, one of Nouakchott’s oldest neighborhoods, is both working-class and multiethnic. It is here that rappers gather on stages for concerts. 24 Loeufs = youngsters. 25 Initials of the Islamic Republic of Mauritania, République Islamique de Mauritanie. 26 go = girls. 27 cago = bordello. 28 Neighbors. 29 Pétasse in French slang (inverting syllables), i.e. a slut. 30 With this verb the author means, “Those who are swept away by other religions, those who are atheist, or Muslims who no longer pray, imitate other people and lose their roots”.

References Béthune, C. (2003) Le rap. Une esthétique hors la loi, Paris: Editions Autrement (coll. Mutations). Cachin, O. (1996) L’offensive rap, Paris: Gallimard. Lapassade, G. and Rousselot, P. (1990) Le rap ou la fureur de dire, Paris: Editions Loris Talmart. Martin, D.-C. (1998) Le gospel afro-américain. Des spirituals au rap religieux, Paris/ Arles: Cité de la Musique/Actes Sud. Miliani, H. (2002) “Culture planétaire et identités frontalières. A propos du rap en Algérie”, Cahiers d’Etudes Africaines 168, XL II–4: 763–775. Moulard-Kouka, S. (2004) “Le rap à Dakar. Approche sociolinguistique du langage hip-hop au Sénégal”, in D. Caubet, J. Billiez, T. Bulot, I. Léglise and C. Miller (eds) Parlers jeunes, ici et là-bas, Paris: L’Harmattan. Tauzin, A. (1989) “Le gigot et l’encrier. Maîtres et esclaves en Mauritanie à travers la littérature orale”, Revue du Monde Musulman et de la Méditerranée 51: 74–90. Tauzin, A. (1993) “Le forgeron, la hyène et les bonnes manières. Contes mauritaniens”, Littérature Orale Arabo-Berbère 21: 181–196; Contes arabes de Mauritanie, Paris: Karthala. Tauzin, A. (2006) “Le corps féminin et ses transformations dans la société maure de Mauritanie: influences exogènes et évolutions internes”, in J.-F. Werner (ed.) Médias visuels et femmes en Afrique de l’Ouest, Paris: L’Harmattan. Tauzin, A. (forthcoming) “Insultes rituelles chez les jeunes en Mauritanie: l’art du ‘sabotage’ ”, in A. Tauzin (ed.) Insultes, injures et vannes, en France et au Maghreb, Paris: Karthala.

324

USES AND ATTITUDES TOWARDS HASSANIYYA

16 USES AND ATTITUDES TOWARDS HASSANIYYA AMONG NOUAKCHOTT’S NEGRO-MAURITANIAN POPULATION Alassane Dia

16.1 Introduction Hassaniyya, a Mauritanian Arabic dialect, tends to become the main lingua franca, if not of the whole country, at least in the large urban centers such as Nouakchott, the capital, where all ethnolinguistic components of the Mauritanian population live side by side. The prevalence of Hassaniyya is due to several factors. The first factor is demographic. The native speakers of the language, the Bidhan (white Moors, Bidhani in the local singular form) and the Haratin (descendants of black slaves, Hartani in the local singular form), actually make up the majority of the population. In addition, the Bidhan hold most of the political and economic power. Another significant factor has been the government’s determination to implement an Arabization policy at all levels since the early years of independence. Such partisan policy however, viewed by black Africans, correctly or incorrectly, as an obvious assimilation policy, also hampered the greater expansion of Hassaniyya. In fact, it created a form of resistance among black Africans which implies that Hassaniyya, despite its many resources, could never become the only true and indisputable vernacular, or even lingua franca. TaineCheikh (1997: 74) describes it well when she writes: rare are those among them [black Africans] who recognize Arabic as their own language and consider it a fundamental element of their cultural identity. In actual fact, the pragmatic acceptance of Arabic has improved following the repression of the Liberation Forces of Africans in Mauritania (F.L.A.M.) and the events of 1989. 325

ALASSANE DIA

These so-called “events”1 changed the order of things and upset the fragile linguistic balance which had prevailed until then. Hassaniyya has since spread significantly among black African communities, but its development has not really been explored and raises many issues. To what extent has Hassaniyya become the first lingua franca in the country and in its capital in particular? How is this functional expansion expressed at the linguistic level? Are we observing the development of a lingua franca Hassaniyya characterized by a phonological and morphological simplification as many other African lingua francas (Lafage 1979; Manessy 1992; Thiam 1992)? Does the development of Hassaniyya among black African communities result in multilingualism and the coexistence of various languages or in the progressive Arabization of the communities who increasingly use Hassaniyya as their first language? In short, are there new lingua franca or communal Hassaniyya varieties developing in Nouakchott, together with the Hassaniyya spoken by the Bidhan population?2 Considering the lack of research on these issues, we feel it is necessary to (a) introduce Nouakchott’s Negro-Mauritanian community; (b) describe briefly the speakers’ conceptualization of the various languages’ status, based on a small survey we conducted in a number of schools in Nouakchott; (c) illustrate some phonological and syntactic characteristics of the Hassaniyya variety spoken among Negro-Mauritanian communities, which we observed or recorded on site; and, (d) analyze the attitudes of the speakers with respect to Hassaniyya.

16.2 Introducing Nouakchott’s Negro-Mauritanian community The Mauritanian population, estimated today at 2,800,000 (2,508,159 according to the 2000 census as reported by sources at the National Office of Statistics), consists of two large subcategories or communities: the Moors mentioned above, and the Negro-Mauritanians. The latter designation includes the four black African ethnic groups in the country, namely the Pulaar or Haalpulaaren (Fulani), the Soninké or Sarakholle, the Wolof and the Bambara. According to Balta (1980: 27), the demographic distribution among the three main Negro-Mauritanian ethnic groups is the following: Pulaar (66 percent), Soninké (22 percent) and Wolof (5 percent). NegroMauritanians are generally considered a minority within the country even though official statistics do not mention ethnic origins. In the absence of official statistics, most speculations about the respective proportions of the different population components are unreliable. Each author comes up with his own figures which may fluctuate depending on the publishers. For instance, the estimated numbers introduced by Ould-Zein and Queffélec go from 80 percent Moors and 20 percent Negro-Mauritanians in 1998 (1998: 7 note 2)3 to “between 60 and 80 percent Moors and 20 to 40 percent NegroMauritanians” in 2001 (2001: 10 note 1). Daure-Serfaty (1993: 85) reports 326

USES AND ATTITUDES TOWARDS HASSANIYYA

that Negro-Mauritanians generally reject their status of minority group and claim to account for at least half the population.4 It is therefore rather difficult to have a clear notion of the representativeness of each of the communities. We think however, that the most reasonable figures are 60 percent Moorish and 40 percent Negro-Mauritanian. It is even more difficult to establish statistics for a city like Nouakchott where nearly one third of the country’s diverse population mixes and blends together. Some districts of the city however are known for hosting such and such community. Thus, certain districts of Old Nouakchott such as Medina R or Medina 3 are mostly black African while others such as the Ksar are mostly Moorish. An administrative city born after independence (the foundation stone was officially laid in 1957), Nouakchott was at first populated mainly by Negro-

Map 16.1 Map of Mauritania showing main southern cities

327

ALASSANE DIA

Mauritanians, more sedentary than their Moorish fellow-citizens and much more receptive to the colonial schooling and, consequently, the first servants of the early administration (Stewart 1989: 162). In the mid 1970s however, the young capital “was literally taken by storm” as a result of an unrelenting drought. According to figures of the National Office of Statistics (June 2001), the population of Nouakchott increased from 134,704 inhabitants in 1977 to 303,328 in 1988, and to 558,195 in 2000. Today, semi-official sources suggest a figure of 800,000 inhabitants, i.e. a third of the total population of the country. The demographic explosion resulted in “uncontrolled” development of the city and emergence of kebbas (“dumps”, in Hassaniyya) on the outskirts of the city inhabited mostly by Haratin people, now freed from their former masters. It is impossible to determine which community is the largest in the city, but one could guess that the Moors are dominant because they include the Haratin who are also considered the majority individual component in the country. The fact remains that even if Negro-Mauritanians were to be considered a minority group in Nouakchott, they would remain, at the very least, a very strong minority group. In the same way, the absence of official statistics makes it impossible to determine the percentage of those “born and bred”5 in Nouakchott versus more recent arrivals. However, in spite of the massive arrival of rural populations fleeing the drought, the Negro-Mauritanian community is not made up only of poor people who came to join the ranks of the city’s unemployed. Members of the community work in all socio-professional activities (civil servants, tradesmen, workers, etc.) although they tend to specialize in the informal sector such as tailoring, or in semi-skilled work (plumbers, mechanics, electricians, etc.).

16.3 The status of languages according to the representation of Negro-Mauritanian students Before we describe the linguistic characteristics of the Hassaniyya spoken by Negro-Mauritanians of Nouakchott, we need to reflect upon the conceptualizations which Negro-Mauritanians may have of the socio-linguistic reality and on the status of each language considered. 16.3.1 Hierarchic classification of the languages in use There is generally a large gap in situations of multilingualism between public management – linguistic planning – projected by the government, and the actual or in-vivo management, to use a term borrowed from Calvet (1987), by those who speak the languages. We will begin with a very preliminary study of the Negro-Mauritanian representations of their linguistic environment. We selected a sample of 117 students from three senior classes and

328

USES AND ATTITUDES TOWARDS HASSANIYYA

Table 16.1 Classification of the languages of Mauritania, French included, from the most spoken to the least spoken Classification

Hassaniyya

Pulaar

Soninké

Wolof

French

Arabic

1st position 2nd position 3rd position Total

31 13 10 54

7 17 18 42



3 6 9 18

16 17 4 37

2 5 4 11

2 15 17

Source: Survey conducted between April and May 2002 among 63 students

one junior from bilingual sections of both public and private schools and gave them a questionnaire based on the status evaluation of all languages spoken in the country. The selected sample included 73 students claiming Pulaar as their native language, 13 speaking Soninké, 20 speaking Wolof, 6 speaking Hassaniyya, 2 speaking Bambara and 3 foreigners. The resulting classification of the most spoken national languages (Table 16.1) according to 63 students calls for several observations. Among African languages, we note that Bambara is not mentioned at all, undoubtedly because of its minority status. Pulaar is mentioned 42 times, which can be explained by the high proportion of Pulaar speakers in our sample. The share of Wolof is relatively modest here (18 times in all) when compared to its use as a second language (see Table 16.2, p. 331). As for the two major academic languages, French and Arabic, we note that whereas French is quoted 16 times as the first spoken language in the country and 17 times as the second, Arabic, (i.e. literary Arabic) is mentioned only 11 times (less than Wolof ), and only twice as the first spoken language in the country. The low ranking of Arabic, the official language of the country, can be explained by several factors. One is that – because of their French academic background – the students do not recognize literary Arabic as a spoken national language. Another is that the majority of the people we surveyed consider Hassaniyya and literary Arabic as two related varieties of the same language, Hassaniyya being considered the spoken language. At any rate, Hassaniyya is recognized by most as the first language spoken in the country. It is the language most frequently mentioned (54 times) and ranks as the first language in the country for nearly half of the people surveyed (31). The significance of this first place is boosted by the fact that the overwhelming majority of the people surveyed belong to the NegroAfrican community. The majority of the students surveyed admit using Hassaniyya in order to communicate with native speakers of the language. In situations involving only Negro-Mauritanians, albeit of different lan-

329

ALASSANE DIA

guages, French is generally spoken among educated speakers; otherwise, one of the Negro-African languages is used (frequently Pulaar or Wolof ). Whereas Hassaniyya definitely ranks first, it is not obvious which language ranks second. French seems to emerge as the second language since it is mentioned as such by 17 students against 16 mentions of Pulaar. We can also note that 25 percent of the students surveyed consider French the first language in the country whereas Pulaar ranks first only for 7 students. The criteria for classification may not have been the same for all languages. Indeed it seems that languages were classified according to their social prestige. The fact that Arabic (actually Hassaniyya) and French rank first and second respectively seems to reinforce a distinction between culturally and socially dominant languages versus dominated languages (Pulaar, Soninké, Wolof); an opposition which could also be interpreted in terms of languages with a written tradition (the traditionally oral Hassaniyya language is assimilated to written Arabic and benefits from its prestige) versus languages with an oral tradition. In such a case, the classification shows that the Negro-African speakers, while they continue to speak their own languages, have partly internalized the representations and stereotypes of the Moorish society who, short of assimilating literary Arabic to Hassaniyya, consider the latter as a superior language (compared to other African vernaculars) because of its relationship to the former. It is further significant that those who speak Hassaniyya do not distinguish between the various African languages and classify them all under the name of “klaam likwar” (language of the kwars, in other words language of the blacks). 16.3.2 Bilingualism and the functional distribution of languages Beyond the information provided by the questionnaire about the status and the lingua franca function of the languages used, we examined the degree of bi- or multilingualism of the students surveyed as well as the contexts or areas considered suitable for the use of such and such other language. Therefore we asked them whether they spoke national languages other than their native languages and in which circumstances they used these languages.6 Out of a total of 114, 105 speak at least two national languages. Hassaniyya is mentioned 74 times and thus ranks first as the most frequently spoken second language among Negro-Africans, followed by Wolof (65), Pulaar (29) and Soninké (11). Whereas Negro-African speakers are typically multilingual, the three Moorish students in our survey said they did not speak any national language other than Hassaniyya. This contrast shows that those speaking other languages are the ones making the effort to speak Hassaniyya whereas the Moors are not learning African languages: as a result Hassaniyya is one of the great inter-ethnic communication languages in the country.

330

USES AND ATTITUDES TOWARDS HASSANIYYA

Table 16.2 Other national language spoken besides native language Student’s native language

Other national languages spoken Pulaar Soninké

Pulaar Soninké Wolof Hassaniyya Other Total non-native speakers

9

Wolof Hassaniyya

10 18 – 1

53 7

2 – –

– 5

47 9 15 – 3

29

11

65

74

Total native speakers surveyed 73 13 20 3 5

speakers speakers speakers speakers speakers

Source: Survey conducted among 117 students; 114 valid responses

The above figures however do not mean that Hassaniyya is the only and single lingua franca in the city and country. A number of students said they use French primarily to communicate with other ethnic groups and that they use Hassaniyya or other African languages only if their interlocutors understand neither their language nor French. But in everyday life, the situation is much more complex than our students expressed. We know, for instance, that French is learned primarily through school education; considering the relatively low rate of Mauritanians attending schools (the rate of literacy among Mauritanian young people between 15 and 24 years of age is 61.3 percent according to the UNESCO Institute for Statistics),7 we can easily assume that conversations between speakers of different languages are more likely to resort to one of the national languages, in particular Hassaniyya or Wolof. Wolof ranks second among the national languages used in addition to one’s native language (65 responses). The fact that Wolof ranks so high, in spite of the marginal position of its native speakers (Wolofs account for only 5 percent of the Negro-African community) is due to the strength of the language in nearby Senegal where it stands as a lingua franca. The proximity of the two populations and the influence of Senegalese television – the only one to reach Mauritania for a long time – also helped. Wolof thus comes before Pulaar (27 responses) and Soninké (11). The relatively poor showing for Pulaar is undoubtedly explained by the fact that it is the native language of more than two-thirds of the students surveyed. But, in spite of the relatively low figure, the percentage is rather significant since more than half of the students who are not Pulaar native speakers claim to speak the language. Finally, Soninké is hardly ever spoken by non-natives (with the exception of the Guidimakha area where it is the first language). It is interesting to consider the link between the languages spoken by each of the students surveyed and their backgrounds. The majority of the people

331

ALASSANE DIA

who claim to use Hassaniyya as their first language are either born and bred in Nouakchott, come from other large urban areas in the country such as Nouadhibou or Zouerate, or live in areas of the country where Hassaniyya prevails. Among the people who grew up in the south of the country, few speak Hassaniyya. The Pulaar speakers who claim to speak Soninké are all, except two, from the Guidimakha area. The Soninké natives who speak Wolof or Pulaar come either from the mostly Negro-African neighborhoods of Nouakchott or from the Gorgol area (especially those who speak Pulaar). Hence, it seems that for non-native speakers, the choice of national language spoken is determined essentially by the speakers’ backgrounds. This shows the relativity of the concept of lingua franca in Mauritania. Even if Hassaniyya appears as the language best suited to achieve this function, reality is more complex since each language prevails in a specific geographical area. As observed by Diagana (2000: 244): Everywhere in the country, one or more languages or varieties of languages are used as means of communication between populations with different native languages. We identified three different situations: Southern Mauritania, Northern Mauritania and Nouakchott, the capital, with its own specifics. The surveys carried out by the author in the Gorgol area, in the southern part of the country, showed that 72 percent of the Pulaar speakers did not use any national language other than their own; they did not consider it useful to speak another language since “everyone in Kaedi spoke Pulaar”. Therefore Pulaar seems to come first in the South and more precisely in the Gorgol and Brakna areas whereas Soninké seems to prevail in Guidimakha and Wolof in Trarza. Hassaniyya, for its part, seems to be first everywhere else except in the capital where it may be competing with Wolof. Chartrand (1977: 118) mentions a survey conducted by the ENA ([French] National School of Administration) which gives a slightly different distribution of languages: Hassaniyya dominates clearly [. . .] in all areas of Mauritania, except for the 4th (Gorgol) and the 5th (Brakna) areas where Pulaar prevails and for the 10th (Guidimakha) area where Soninké comes first; Hassaniyya prevails in the district of Nouakchott. Such a distribution does not seem to reflect truly the real life situation even if the context was different at the time from what it is today. For instance, one could at least wonder why no mention is made of Wolof considered as the common language par excellence in the Trarza area (border with

332

USES AND ATTITUDES TOWARDS HASSANIYYA

Senegal). The lingua franca status of Hassaniyya is also relative in the city of Nouakchott and in the other large metropolitan areas in the country where the mixing of populations creates very complex situations. To conclude this preliminary survey about speakers’ representations, we will keep in mind that the people surveyed tend to internalize a diglossic representation of the functional distribution of the languages, which does not reflect the reality of everyday practices. Thus, national languages (except French and Literary Arabic) are allegedly said to be used in the streets or with friends or with parents speaking different languages, in short in informal situations, whereas Arabic and French would be reserved for more formal contexts. However, it has been observed that Hassaniyya (TaineCheikh 2002) as well as Negro-African languages are also used with various degrees in sermons, meetings or political speeches and that the boundary between formal and informal contexts is fluid.

16.4 Which type of Hassaniyya is spoken among Negro-Africans? In order to answer this question, we used two types of data. We studied a group of 12 young people from the Medina R and Socogim neighborhoods, with whom we had worked previously and who, in some cases, had filled out our questionnaire. These young people attend school and are natives of Nouakchott or grew up in the city for the most part. Most come from the various ethnic groups who make up the Negro-African community but there are also a few Haratin among them. We observed and recorded the group’s language practices and we asked them about their feelings in relation to these practices and to the country’s linguistic situation in general. The second approach consisted in taking notes as we moved about markets and grocery stores and also a number of government offices in Nouakchott, which allowed us to make contact with people from various backgrounds. In this way we completed about 20 recordings totaling one hour which we added to the previous two-hour long recording of young people. The data collected confirm that we cannot speak of one practice but of the several practices of Hassaniyya among Negro-Mauritanians. Indeed, those practices vary according to each individual and to each individual’s background, particularly to the town/area where s/he grew up. We can however identify in these practices a number of significant permanent features generally common to most speakers, which implement processes of phonological, morphological and syntactic reorganization characteristic of those observed in other lingua francas whether they derive from African languages or Arabic and which lead, in particular, to a reduction of the phonological inventory and a simplification of gender and number agreements (Billiez 1992; Diouf 1992; Manessy 1992; Miller 1987, 2004).

333

ALASSANE DIA

16.4.1 Phonological characteristics One of the principal difficulties facing Negro-Mauritanians who resort to Hassaniyya resides in the phonological system of the language, quite different from their own. Indeed, whereas all Negro-African languages spoken in Mauritania belong to the Niger Congo phylum, Hassaniyya, on the other hand, belongs to the Afroasiatic phylum. Several phonemic distinctions of Hassaniyya, such as the interdentals /θ/ as in θ umma “then”, /2/ as in 1il “shade” and /o/ as in haOa “that” or the pharyngeals /¡/ and /]/, as well as the emphatic consonants do not occur in the Niger Congo languages. Hence we frequently observe that Negro-Mauritanian speakers who were not born or did not grow up in Nouakchott neutralize the contrasts existing between these sounds and reduce them to a single phoneme. For instance, the three interdentals /o/, /2/ and /θ / are frequently reduced to the single phoneme [d] and less frequently to [θ] as in the following example : (1) Lady customer (pointing to a bag): he haθ e kem ? Hey this how-much “How much does this cost?” Shopkeeper: geb1ih we tawv bi deux milles cinq cents Take it and only at two thousand five hundred “You can have it for just two thousand five hundred.” Lady customer: ana kay maw gabaθ i haθ e deux milles cinq cents I, kay-NEG take it this two thousand five hundred. “I will not buy it for two thousand five hundred.” In the above exchange, overheard at the Nouakchott central market, also known as the capital market, between a Moorish shopkeeper and a black African woman customer, we noted that the customer assimilated three interdentals into a single phoneme, the voiceless interdental phoneme /θ/. This woman’s example is somewhat atypical, because in instances of neutralization, archiphonemes are generally borrowed from the speaker’s native language. As for interdentals, they are generally replaced by the phoneme [d] as in this phrase by a Negro-Mauritanian speaking about the neighborhood grocer to another customer: (2) hade razil mAlEm This man good “He is a good man.” In the same manner, the emphatic consonants /x/ as in Xawt “voice” or /y/ as in Yawle “table” are also difficult to enunciate for some Negro-Mauritanian 334

USES AND ATTITUDES TOWARDS HASSANIYYA

speakers. The former is generally assimilated to [s] while the latter is reduced to [t] as in the following example where Yfoyl becomes tufoyl and Xalli changes to salli: (3) huwe tufoyl mezal ysalli He young man still pray-IMPERF3M.SG “The young man (teenager) is still praying” (has not finished praying) Other Hassaniyyan phonemes which Negro-Mauritanian speakers find hard to articulate are the pharyngeals /¡/ and /]/ and the palatal [l]. The pharyngeal /¡/ sometimes disappears to be replaced by lengthening of the following vowel: (4) ma∞lEm > mAlEm “nice, generous” (4′) Ana anndu muskil kebCr I have-3M.SG problem big “I have a big problem.” In this example the /¡/ of ∞andC (here changed to anndu) disappears and is replaced by the nasal vowel /p/ which does not exist in Hassaniyya. The palatal /l/, is frequently replaced by the dental /z/: (5) raLil > razil “man” razzAle ya-Edu hok (= er-raLLAle ye-∞Edu hok) men stand-IMPERF3M.PL over there “The men stand over there.” Other phonemes such as the palato-alveolar fricative /∫/ or the labial /v/ are also the object of reorganizations. The former is generally changed to /s/ as in the examples mentioned: (mu∫kile > muskil in example (4), im∫i > imsi in example (7), ∫rabt > sarabtu in example (8)). The /v/, for its part, almost systematically changes to /f/ when spoken by Negro-Mauritanians not familiar with Hassaniyya and particularly by those whose introduction to Arabic was through the Koran, like in the case of mele}ve > melehfe “veil”: (6) hiyye lAbis melhefe zein she wear-3SG.FEM veil-FEM pretty-MASC “She wears a pretty veil.” We observe that the neutralization of interdentals, emphatics and pharyngeals is a phenomenon fairly characteristic of people who lived in areas where Hassaniyya is not as popular as in Nouakchott. Thus in example (2), we must specify that the speaker did not grow up in Nouakchott and that he 335

ALASSANE DIA

has lived there only three years. Therefore all our examples are characteristic of the Hassaniyya spoken by Negro-Mauritanians who did not grow up in an urban environment. On the other hand, the use of these phonemes does not represent any particular difficulty for speakers who grew up in urban environments and, more generally, to those with a good academic background (educated either in public or traditional Koranic schools). City people, and particularly the “true Nouakchott citizens”, as some of the people surveyed called themselves, make it a point of honor to realize these phonemes correctly. For them, the correct usage is a way of showing that they belong to the urban community, as opposed to Nouakchott’s communities of recent rural immigrants. In my data I found 76 percent of correct realization of interdentals among the young Negro-Africans born and bred in Nouakchott (132 in 173 occurrences) while only 21 percent of correct realization of interdentals among people who came recently to Nouakchott (25 occurrences in 117). Manessy (1990: 8) notes, among the modes of structure of urban speech, a tendency to use “a ‘correct’ pronunciation [. . .] in as much as this way of speaking is considered characteristic of city speech, and is therefore a sign of integration into the urban community”. These phonemes, particularly interdentals, remain comprehensible even when neutralized. What’s more, the treatment of these phonemes by Hassaniyya native speakers is evolving. A short while ago, Taine-Cheikh pointed out (1979: 168) the conservative nature of Hassaniyya and gave as an example “in phonology, the maintenance of the three interdentals of classical Arabic”. Today, we can say unmistakably that the contrast between interdentals is not as well-defined as one might expect since we find, even among Moors with whom we spoke, a tendency to neutralize the phonemes /θ / and /o/. Perhaps this is part of a restructuring process of urban as opposed to rural Hassaniyya (see Taine-Cheikh, this volume). 16.4.2 Morphosyntactic characteristics Negro-Mauritanian speakers tend to restructure the morphology and syntax of Hassaniyya as much as they do the phonology. These reorganizations affect primarily the verbal system as well as the gender and number agreements. 16.4.2.1 Verbal system The rules of conjugation (marks of person, aspect, tense) are not always respected, as in: (7) Ana yamis imsi sDr polis huwa gali zi hDn I yesterday go-IMP3M.SG to police he tell-PERF3M.SG-PR1 come here “I went to the police yesterday, they told me to come here.” 336

USES AND ATTITUDES TOWARDS HASSANIYYA

In this statement, the verb “to leave” me∫a is in the imperative mood (imperfective) instead of perfective 1st person singular (m∫eit). Similarly, a speaker tells a young Moor who gives him a glass of tea: (8) anha Ana mAw sarabtu ataye No I-NEG drink-PERF1.SG ART-tea “No I do not drink tea.” The perfective 1st person singular sarabtu used here is confused with the imperfective which would be more appropriate since the message expresses a permanent attitude (e.g. present general) cf. anhe Ane mA na∫rub etAy. We noticed many occurrences (48 percent of total verbs) of such specific uses among our study groups, particularly in the recordings we made of chance encounters, more than among the group of young people in the older neighborhoods of Nouakchott. These uses cannot be attributed to phenomena of interference (influence of native languages) because these languages have, like Hassaniyya, a verbal system with morphological aspecttense values. In Pulaar, for example, the marks of persons and aspect-tense are expressed through verbal flexion8 as in Arabic. It is therefore more likely to be a process of formal simplification, very widespread in lingua francas, which tends to reduce the aspect-tense morphological marks. We noted one occurrence of an invariable verbal form: (9) ewa tey dengay Ylaå ? then today dengay-IMPERF2M.SG go-up-PERF3M.SG “Are you going up today?” This occurrence occurs in a case of codeswitching: only the verb Ylaå “to go up” belongs to Hassaniyya, the rest of the phrase is in Wolof. dengay marks both the imperfective and the person and therefore replaces the Hassaniyyan morpheme -ta (2nd person imperfective singular). It is important to point out that the phrase was spoken by a bilingual speaker who has a perfect knowledge of both languages (Wolof and Hassaniyya) and that the practice does not result from limited language skills but more from the limits imposed by codeswitching. This explanation seems all the more likely as it is the only example of this type found among our collected recordings and as it is revealed in a hybrid statement. 16.4.2.2 Invariability of possessive marks Another morphosyntactic characteristic is the use of invariable possessives such as anndu, disregarding the person agreement as in this occurrence where the speaker/client tells the grocer: 337

ALASSANE DIA

(10) Ana aandu miyé mAw aandu degdeg I have3.SG hundred NEG have3.SG coins “I have a one hundred [bill], I have no change.” instead of the standard form Ana ∞andi. The absence of agreement for the possessive object is so frequent among neo-urban Negro-Mauritanians that Hassaniyya native speakers imitate it derisively, not unlike “negro pidgin” in colonial times or what we now call foreign-talk. To the question ∞and-ak keθ a /have-PR2M.SG/“do you have such and such item?”, the grocer, a native Hassaniyya speaker, generally answers ∞and-ak baå “you do have” instead of ∞and-C keθ a /have-PR1SG/ “I do have that item”. Other possessive forms are also influenced by these transformations. Thus uxtC /sister-PR1SG/ “my sister” will be changed by some speakers into uxtu Ana /sister-PR3M.SG I/. (11) he xalli hAze uxt -u Ana he let+PR3M.SG DEM sister-PR3M.SG I “He let her, she’s my sister.” instead of the regular form he xallChe hAθ i uxtC 16.4.2.3 Gender neutralization Another syntactic characteristic is the lack of distinction between genders, a distinction about which Hassaniyya native speakers are adamant. We noticed that Negro-Africans speaking Hassaniyya tend to use almost exclusively the masculine gender (cf. (11) for example), even those who speak almost perfect Hassaniyya. On one occasion we listened to a half-hour conversation in Hassaniyya between cousins (a girl, Fayaye, and a boy, Moudiri), both native Pulaar speakers, born and bred in Nouakchott. The boy, in spite of his perfect command of Hassaniyya, systematically ignored gender distinction and spoke to his cousin as if he were addressing a boy, as the following example illustrates: (12) n’ta dak l’accident el bare} vi IlDt C Kinhu ? You.M.SG that ART-accident last night in block C what “You, what was that accident last night in block C?” In this sentence the speaker chose to use the masculine, 2nd person singular pronoun, n’ta, instead of the feminine 2nd person singular, n’ti. This might reflect a voluntary attempt to simplify the language. Such linguistic behavior is not an isolated case and seemed quite normal to the girl, though she herself had a perfect command of Hassaniyya. Other group conversations 338

USES AND ATTITUDES TOWARDS HASSANIYYA

validated that point. And one wonders if it could be a way of resisting Hassaniyya and its encroaching dominance. Such gender neutralization however, is not always the result of a conscious effort at simplification. Sometimes it results from a speaker’s limited proficiency. We did observe, in some cases, variations in the treatment of agreements depending on the speakers’ more or less perfect command of Hassaniyya. For example, these young people teasingly said to each other: (13) N’ta }ayAtak xAmre (You.MASC life.FEM-your awful.FEM) or }ayAtak xAmir (life.FEM-your awful.MASC) “you your life awful”, “Your life is awful/you live a shitty life.” The following examples show various cases of incorrect gender agreement in the speech of both native and non-native young people of Nouakchott. In example (14) (from a native of Nouakchott), the person uses the fem. personal pronoun hiyye but the masc. form of the verb (marrag instead of fem. marrgit). In example (15) the demonstrative and the adjective lack gender agreement while in example (16) the adjective “cold” takes the feminine agreement while the verb remains in the masculine form (xAlig) in agreement with the word ma§ “water”. (14) hiyye (et)tele mA marrag l’match (instead of hiyye (et)tele mA marrgit l’match) she ART-TV NEG come out-PERF3M.SG ART-match “The TV did not retransmit the match.” (15) dAk leylChe mab1Eg hatte (instead of dCk l’leylChe mab1Ege) DEM.M monkey-F.SG puffy-M.SG well “This type of she-monkey is really puffy.” (16) ma§ bArde xAlig water.M cold.F exist.M “There is cold water.” 16.4.3 Lexico-syntactic characteristics It is not easy to identify Hassaniyyan lexical characteristics used by NegroMauritanians since codeswitching is so frequent. We must note however that sentences are generally punctuated with non-Arabic grammatical words. Therefore emphatic terms such as kay or kam derived from Pulaar, or rek derived from Wolof (?) are often used. There is also the frequent recourse to French words or words borrowed from local languages as illustrated below: (17) Ana kay maw gabaθ i haθ e deux mille cinq cents I kay NEG buyPERF1SG this deux mille cenq cents “Me, I will not buy that for two thousand five hundred.” 339

ALASSANE DIA

(18) mohammed kam hrab Mohamed kam run-away-PERF3M.SG “Mohamed, he ran away.” (19) γAsit toqrAt he “She went to her namesake’s.” (tokara, “namesake”, is a Pulaar word) (20) huwwa m’murti “He’s playing smart.” (murti, “smart”, is a Pulaar word). As for the treatment of French words, we observed a slight difference between Hassaniyya native speakers and others. The former tend to use a French qualifier before the borrowed word, when the word is not considered to be integrated: (21) hEme les jeunes m∫Aw They les jeunes left-IMPERF3M.PL “Have the young people left?” The substantive, “jeunes”, is preceded by the French plural qualifier, “les”. It should be stressed that this type of insertion is very frequent in occurrences of codeswitching in North Africa (Boumans and Caubet 2000). Negro-Mauritanians often use the Hassaniyya determiner as in this sentence where Fayaye asks Moudiri: (22) n’ta hade e∫ chanteur minhu You this ART-chanteur who “Tell me who this singer is.” Another example of the way Negro-African Hassaniyya speakers treat words borrowed from French as Hassaniyya words is the use of the verb “galérer” (to have a hard time) in this statement by Oumar: (23) les gens m’galerCn hun ART.PL people.PL, galérer-PART.3M.PL here “Here, people have a hard time.” In this sentence, the verb “galérer” agrees in number and gender with its subject (the people) using the Hassaniyya morphological markers. We observed that the same verb is treated the same way in Pulaar, and also takes the Pulaar verbal marks. 16.4.4 Summary We will conclude this brief description by mentioning that restructuring affects all levels of language, even if at varying degrees. Distinctive phono340

USES AND ATTITUDES TOWARDS HASSANIYYA

logical variables are mainly the prerogative of new immigrants and are almost non-existent among young speakers who grew up in Nouakchott. The morphosyntactic restructuring of Hassaniyya was observed among both categories of speakers. Gender neutralization, for example, was observed even among those who speak Hassaniyya rather well. In light of the information we gathered, we may presume that the practices observed are, for the most part, the result of a process of acquisition rather than of a desire of “appropriation” or “divergence” as some of the interviewees claimed. However, one cannot fully deny the existence of such a desire and for some speakers (those with a perfect command of Hassaniyya) this restructuring is used for fun more than anything else. Finally, it is interesting to note that young people between the ages of 15 and 25 generally have a better command of Hassaniyya than their elders – no matter how long either of them have lived in Nouakchott.

16.5 Attitudes towards Hassaniyya As is their practice, the attitudes of Negro-Mauritanians towards Hassaniyya are many and varied. In most cases, in urban environments, the relationship to the language is rather paradoxical, and reflects both attraction and repulsion, love and hatred. Hassaniyya is spoken because it is impossible to avoid in a Mauritanian context, as Fayaye explains. But, as Fayaye continues, it is also used to display one’s identity, primarily their Mauritanian identity but also their urban citizenship: Every time we get together we speak Hassaniyya. This irritates our other cousins who come from the village. But we are Mauritanian and these languages belong to us all. And then, we are also Nouakchott residents. As if embarrassed by the pride they feel in knowing the Hassaniyya language, they speak what they consider to be a specific Hassaniyya, devoid of a number of grammatical constraints. Such feelings of ambivalence with respect to the language are rather widespread. The young people we met all like the language but they resent the fact that it tends to be imposed on them. Amadou Tidiane says he has nothing against Hassaniyya, and enjoys speaking the language with his friends, “but asking me to speak about my day in Hassaniyya during an exam to enter public office, as I had to do recently, is pushing it too far.” As long as it is imposed, they all definitively refuse to adopt Hassaniyya as the sole lingua franca. Indeed, many of them enjoy speaking French or switching codes when they are with Moorish friends. Following the same logic, young Negro-Africans prefer to use between themselves composite forms of language which borrow vocabulary and structures from all languages around and even beyond (Dia 2004). 341

ALASSANE DIA

There is however, an undeniable reality upon which everyone agrees: the knowledge of Hassaniyya, however rudimentary, is absolutely essential in Nouakchott, if only to satisfy the most elementary daily needs. For many Negro-Mauritanians, Hassaniyya only serves this function.

16.6 Conclusion To the Mauritanian people generally, Hassaniyya emerges as the first language of the country. At first glance, the development of Hassaniyya seems to fit with the model developed by Calvet (1994),9 who postulates that urban multilingual situations (resulting from migration) lead in the long run (three generations) almost ineluctably to monolinguism either through the emergence of a common pidgin/lingua franca or through the acquisition of the urban vernacular by the second or third generation of the newcomers. But the urban Mauritanian multilingual setting does not fit with this mechanical model, particularly due to identity and political factors. Despite its predominance, Hassaniyya is not considered the sole lingua franca by all Mauritanians for two main reasons. The first one is due to the governmental ideological stand to privilege Arabic, and through Arabic, Hassaniyya. The feeling, created by this choice of submitting to a policy that tends to destroy their own languages, pushed Negro-Mauritanians into a corner by making them reject any thought of definitely adopting Hassaniyya as their first language. The second reason is related to the dynamics of the language’s internal evolution. Hassaniyya is still rather conservative grammatically as well as phonologically and may need to adopt some of the simplification processes as practiced by Negro-Mauritanians, in order to achieve a standard level that will facilitate its gaining popularity. It is very unlikely that a simplified form of Hassaniyya will one day become a stable urban variety and it is more accurately described as a transitory step. Negro-Mauritanians however, will certainly have their say about possible urban and common forms of Hassaniyya since a number of their practices have been taken up by Hassaniyya native speakers, particularly by Haratin.

Notes 1 The period modestly referred to as “events” is that dark page in the history of independent Mauritania which lasted several years (1986–1993). The April 1989 conflict between Mauritania and Senegal served as a pretext for the Mauritanian government to carry out a vast operation of “denegrification” in the country, which had started in 1986 with the earliest arrests of black intellectuals and repressive measures against the community. Tens of thousands of Negro-Mauritanians were then deported to Senegal and Mali. The Senegal valley, occupied by NegroMauritanians, was the scene of all sorts of exactions (extra-judicial executions, spoliations, etc.) by the Mauritanian army and the Haratin militia set up by the government. In September 1990, the Negro-African elements were, in turn, defini-

342

USES AND ATTITUDES TOWARDS HASSANIYYA

2 3 4

5 6 7 8 9

tively purged from the army (cf. Mauritania 1986–1989: Context of a Crisis, Amnesty International, November 1989, and other annual reports of this organization). See Cohen 1963 and the research work by C. Taine-Cheikh on the Hassaniyya language. The authors humbly acknowledge that their “figures are personal and thus inevitably approximate” (1998: 7 note 2). Indeed, demographic growth is quite high among mostly polygamous NegroMauritanians. The new 2001 Civil Code prohibiting polygamy, although never enforced, is considered by many as an attempt by the government to slow down this growth. Another argument in favor of Negro-Africans is the population density. The department of M’Bagne in the Negro-Mauritanian area has the highest density with 35 inhabitants per km2 whereas, in contrast, there are 2 inhabitants per km2 in the northern Moorish area. Nouakchott, as mentioned earlier, was only created in 1957. “Born and bred in Nouakchott” refers to the first inhabitants in the city and of course, to those who were born or grew up there. The answers relating to practices were not presented in the form of statistics and were often collected orally and not systematically by way of questionnaires. Until the reform of 1999, education in Mauritania was divided into two sections, one Arabic and the other known as bilingual (see Sonkalo 1995), yet more than two-thirds of the students are taught in Arabic. The aspect-tense marks are identified through the form of the verb. Calvet 1994: 5: “telle une pompe, la ville aspire du plurilinguisme et recrache du monolinguisme et elle joue ainsi un rôle fondamental dans l’avenir linguistique de la région ou de l’Etat.”

References Balta, P. (1980) “Une réforme linguistique courageuse mais complexe”, Le Monde Diplomatique (July). Billiez, J. (1992) “Le parler véhiculaire interethnique de groupes d’adolescents en milieu urbain”, in R. Chaudençon (ed.) Des langues et des villes. Actes du Colloque International, Dakar 15–17 Decembre 1990, Paris: Didier Erudition, pp. 117–126. Boumans, L. and Caubet, D. (2000) “Modelling intrasentential codeswitching: a comparative study of Algerian/French in Algeria and Moroccan/Dutch in the Netherlands”, in J. Owens (ed.) Arabic as a Minority Language, Berlin: Mouton de Gruyter. Calvet, L. J. (1987) La guerre des langues et les politiques linguistique, Paris: Payot. Calvet, L. J. (1994) Les voix de la ville, Paris: Payot. Chartrand, P. (ed.) (1977) Situation linguistique et politique de la langue en Mauritanie: essai de description, Nouakchott: ENA. Cohen, D. (1963) Le dialecte arabe Hassaniyya de Mauritanie, Paris: Klincksieck. Daure-Serfaty, C. (1993) La Mauritanie, Paris: L’Harmattan. Dia, A. (2004) “Politique et réalités linguistiques en Mauritanie : parlers jeunes en milieu négro-africain de Nouakchott”, in D. Caubet et al. (eds) Parlers jeunes ici et là-bas, Paris: L’Harmattan, pp. 17–31. Diagana, S. O. (1992) “Contacts de langues: approche sociolinguistique des emprunts du soninké au français, à l’arabe et au pulaar”, unpublished thesis, Université Paris V.

343

ALASSANE DIA

Diagana, S. O. (2000) “Concurrence linguistique déloyale en Mauritanie”, in P. Dumont (ed.) La coexistence des langues dans l’espace francophone, approche macrosociolinguistique, Paris: AUPELF-UREF, pp. 243–251. Diouf, J.-L. (1992) “Quelle langue parlent-ils à Dakar”, in R. Chaudençon (ed.) Des langues et des villes. Actes du Colloque International, Dakar 15–17 Decembre 1990, Paris: Didier Erudition, pp. 227–236. Lafage, S. (1979) “Esquisse d’un cadre de référence pragmatique pour une analyse sociolinguistique en contexte africain”, in G. Manessy and P. Wald (eds) Plurilinguisme, normes, situations, stratégies, Paris: l’Harmattan, pp. 41–60. Manessy, G. (1992) “Modes de structuration des parlers urbains”, in R. Chaudençon (ed.) Des langues et des villes. Actes du Colloque International, Dakar 15–17 Decembre 1990, Paris: Didier Erudition, pp. 7–28. Miller, C. (1987) “De la campagne à la ville, évolution fonctionnelle de l’arabe véhiculaire en Equatoria (Sud Soudan)”, Bulletin du Centre d’Etude des Plurilinguismes (Nice) 9, 1–26. Miller, C. (2004) “Un parler argotique à Juba, sud Soudan”, in D. Caubet et al. (eds) Parlers jeunes ici et là-bas, Paris: L’Harmattan, pp. 69–89. Ould Zein, B. and Queffelec, A. (1998) Le français en Mauritanie, Paris: AUPELF, Hachette. Ould Zein, B. and Queffelec, A. (2001) “La longue marche de l’arabisation en Mauritanie”, La Revue du français en Afrique 15 (electronic version). Sounkalo, J. (1995) “La situation linguistique en Mauritanie”, Littérature mauritanienne, Notre Librairie 120–121: 36–39. Stewart, C. C. (1989) “Une interprétation du conflit Sénégalo-Mauritanien”, Revue des Mondes Musulmans et de la Méditerranée 54, 4: 161–170. Taine-Cheikh, C. (1979) “Aperçu sur la situation sociolinguistique en Mauritanie”, in Introduction à la Mauritanie, Paris: CNRS, pp. 167–173. Taine-Cheikh, C. (1997) “Pratiques de l’arabe et processus identitaires en Mauritanie”, in F. Laroussi (ed.) Plurilinguisme et identité au Maghreb, Rouen: Publications de l’Université de Rouen, pp. 67–79. Taine-Cheikh, C. (2002) “De la variation linguistique dans le prêche populaire mauritanien”, in A. Rouchdy (ed.) Language Contact and Language Conflict in Arabic, London: RoutledgeCurzon, pp. 177–202. Thiam, N’D. (1992) “Nouveaux modèles de parlers et processus identitaires en milieu urbain: le cas de Dakar”, in R. Chaudençon (ed.) Des langues et des villes. Actes du Colloque International, Dakar 15–17 Decembre 1990, Paris: Didier Erudition, 495–510.

344

INDEX

INDEX

References to tables, figures and notes are indexed as, e.g., 116t, 116f or 116n; references to maps are in italics, e.g. 3 Abbassi, A. 285 Abdel Jawad, H. R. 9 Abha 234, 238–239, 239–240 Abu Haidar, F. 14 accent 184–185 accommodation 11–12, 16–17, 98, 109–110, 214–215 Adam, A. 102 adjectives 87, 256–257, 262–263 affiliation 6, 16, 26n affixes 85–86 age 205–209, 223–224, 224–226, see also youth language agricultural workers 41–42 Aguadé, J. 6, 137, 140n Al-Azraqi, Munira 17 Al-Gunaim, S. 231 Al-Salbm, H. 231 Al-Sunbul, A. 231–232 Al-Tamimi, F. 204 Al-Wer, Enam 14–15, 18, 59, 214, 215, 240 Aleppo 199 Alessa, A. 17 Allam, J. 292 Allard, R. 98 ambiguity 302 Amman 8, 14–15, 59–62 Ammani dialect 55–56; development 59–60; phonology 66–70; variation 15, 63–73 Amman project 62–74 André-Larouchebouvy, D. 155, 164n Anis, A. 243–244 Arabic, Classical, see Classical Arabic

Arabic, Mauritanian Middle, see Mauritanian Middle Arabic Arabic, Modern Standard, see Modern Standard Arabic Arabic urban vernaculars, see urban vernaculars arabization: Middle East 190; North Africa 37–38, 78–80, 81–82, 90, 288n, 325 Arabo parlato della Libia 93n archaic language 316–317, 322 areal norms 268–269 Armenians 159 articles 282–283; definite 259–260, 263, 263t, 264–265; indefinite 88, 89 articulation rate 224–226 Auer, P. 1 avoidance strategies 240–243 Azayr language 41–42 Backus, A. 277, 287, 289n Baghdad 14, 204 Bahrain 9, 14 Barakat, H. 193 Barth, F. 6 bedouin dialects 4–7; morphology 85, 86–87; phonetics 83–84; syntax 88 Bedouin people 91–92 Behnstedt, P. 72, 199 Beirut 148–150, 149 Beirut dialect: attitudes to 151–153; greetings 15–16, 150–151, 153–162; koineization 161–162; variation 15–16, 150–151, 161–162

345

INDEX

Benghazi dialect 85 Bentahila, A. 275 Benzú 127, 133, 134 Berbers 37–38, 93n, 102 Bergsträsser, G. 198 Bessis, J. 81 Béthune, C. 322 bilingualism 128–129; and codeswitching 251–253, 270–271; and education 139; and language domains 330–333; and signage 46 Black Mauritanians 325, 326–327, 328–342 Blanc, H. 6, 14, 198 Bonine, M. 2 borrowing: Cairene Arabic 300–302; Hassaniyya 10–11, 47–48; Mauritanian Middle Arabic 48–49; Tripolitan Arabic 89–91 Boumans, L. 124–125, 139n Buraidah 234, 237f, 238–239, 239–240, 241 Burgat, F. 81 CA, see Classical Arabic Cadora, F. 7 Cairene Arabic: attitudes to 291, 292–295, 302; code-switching 298; dialect variation 296–297; lexical borrowing 300–302; morphology 300–301; youth language 22–23, 291–292, 296–306 Cairo 16–17 calques 301 Calvet, L.-J. 147, 163n, 295, 328 camps, and social organization 39, 52n Cantineau, J. 68, 198, 210n capital cities: Amman 59–60; growth of 8–9 caravan cities 40–41 Card, E. 75n Casablanca 8, 11–12, 99; ethnolinguistic groups 102–103; history 97, 100–102 Casablanca dialect 8, 11–12, 97–98, 134, 137, see also Moroccan Arabic; attitudes to 104, 112–115, 116–119; descriptions 103–105; variation 98, 106–112 Caubet, D. 21, 77, 84–85, 87, 124, 282, 284, 287, 288n Cesàro, Antonio 87–88 Ceuta 12, 125, 126, 127–128, 140n

Ceutí Arabic 128–135, 138–139; definition 130; descriptions 131–135 chain shifts 70 Chambers, J. K. 1, 201 Change, see language change Chartrand, P. 332 Christians 152–153, 158 civil war 15, 60–61, 148–150 Classical Arabic 230; attitudes to 291–292; domains of use 292, 294–295; as official language 81–82; standardization 4–5 code-switching 21–22, 23–24, see also lexical insertions; Ceuti Arabic 135; coding of 250–251; Hassaniyya 47, 49–51, 337; and humor 21, 277, 282, 287–288; and identity 51–52, 284–285; and linguistic deficit 275; and multilingualism 251–253, 270–271; types 270; and youth language 51–52, 275, 276–278, 280–285, 287–288 Cohen, D. 302 colonial period 7–8, 101 color adjectives 87 communications technologies 232–233, see also radio series competence, linguistic 267–268 complex linguistic situations 251–253, see also code-switching; contact situations conjugation 86–87 contact situations: and code-switching 251–253, 276–278; expected outcomes 55–56; and koineization 129–130; and leveling 171; and migration 7, 9–18, 124–125 context, effect on results 220–221 convergence 12, 18–20 Cornell, S. 114 corpus analysis 250–251 Coulmas, F. 164n Coupland, M. 98, 214 Cowell, M. 198, 204 cryptic language 284–285 Cyranaica dialects 85, 86, 93n Damascus 16–17, 191, see also Dummar; Shaghoor; history 189–190; suburbs 188, 189, 192–196 Dammam 233, 237–238, 239 data, accessibility of 2

346

INDEX

data analysis 155–159, 201, 205, 219–226, 235–239, 258–270, 328–333, 333–341 Davie, M. 163n Davies, E. 275 definite article 253, 259–260, 263t, 264–265, 282 Denis, E. 8 Dia, Alassane 23 Diagana, S. O. 332 dialect geography 56–58 dialect maintenance, see maintenance dialect variation: and gender 6, 7, 13, 14–15, 64–65, 66–67, 98–100, 103, 236–239; and identity 15, 66–67, 98–100, 214–215, 226–227; and migration 199, 215 dialects, see also contact situations; names of individual dialects, e.g. Ceuti Arabic: classification 4–8, 42, 59, 77–78, 103–104; descriptions 2; formation of 73–74, 147, see also Ammani dialect; Jewish dialects 84, 85; linear development model 7 diaspora 12, 123–124, 124–127 Dimachki, L. 150 diminutives 86–87 diphthongs 70, 85, 216 discourse languages 252, see also matrix languages Djaziri, M. 81 “do” + verb construction 257–258, 263, 265, 268 domains of use 18; French 45–46, 329–330, 331; Hassaniyya 330–331t, 331–332; Modern Standard Arabic 45–46, 214, 269, 329; restriction 61–62, 214 dominant languages, effect on dialects 124, 128–129, 139 Doran, M. 284 duel construction (plurals) 87–88 Dummar 189, 191, 193–194; dialect variation 201–202, 203f, 205–209; life-modes 195–196 duration 219–226, 227 Eckert, P. 17, 98, 208 economic migrants 135–136, 138 education: effects on language use 215, 267–268, 271; languages of instruction 45–46, 81–82, 251, 329

educational policies: Middle East 231–232, 239–240; North Africa 81–82, 139 Egyptian Arabic, see also Cairene Arabic: loan words from 90; phonological variation 204 Egyptian music 294 El-Himer, M. 12, 27n Elzeini, N. 15 Embarki, Mohamed 13, 220–221, 224 Encyclopedia of Arabic language and linguistics 26n English: loan words from 90–91, 252–253, 253–254, 255–258, 258–266, 269, 270, 281, 300–301, 340; as national language 251 ethno-methodology 9–10 ethnolinguistic groups 102–103 euphemisms 301 extended family 192–193 family: changes in 192–193; and identity 6; and youth 23, 63 feminine endings, and vowel raising 67–68 Ferguson, C. A. 232 Fessi dialect: contact situations 98; leveling 114–117, 119–120; maintenance 103, 106–112, 112–114; morphology 86, 87, 88; phonetics 84–85 Fessi women 10, 11–12, 13; accommodation strategies 109–110; attitudes to Casablanca dialect 104–105, 112–115, 116–119; leveling 114–117, 119–120; maintenance of Fessi forms 106–112, 112–114, 278; research subjects 105–106 first generation speakers 63–64, 105 Fishman, J. 214 focusing 65, 70–73, 74; definition 19 forced migrations 342n–343n form substitution 298 formality, greetings 151–152 French 329; code-switching 275, 276–278, 298; domains of use 45–46, 329–330; influence on Hassaniyya 49–50; loan words from 280–284, 301, 339, 340; North African dialects 285; and rap music 317, 320–321 Fulfulde of Adamawa 268

347

INDEX

fun, see humor future particles 134–135 Gadafi, Muammar 92 gemination 307n gender, see also identity: and dialect variation 6, 7, 13, 14–15, 64–65, 66–67, 98–100, 103, 236–239, 297, 304; distinction of 85–86; neutralization 71–72, 110–112, 338–339 Germanos, Marie-Aimée 15–16 Giles, H. 98, 214 glides 70 glottalization 177–184 Gottschligg, P. 268 Goudailler, J.-P. 295 Gralla, S. 199 Greer, S. 194 greetings 15–16, 151–153, 161; usage 150–151, 153–161 Griffini, E. 93n Griots 310–312, 322 Grotzfeld, H. 198 Guilleminot, C. 220–221 Gulf Arabic Pidgin 17 Gulf countries 8, 17, 26n (h) variable 197–204, 209–210 Hachimi, Atiqa 11–12 Haeri, N. 18, 240, 292, 307n Haratines, see Moorish ethnic group Harb el Kak, M. 163n Haret Hreik 159–161 Hartmann, D. 114 Hassaniyya 23, 325–326; archaic forms 316–317, 322; attitudes to 39–40, 52n, 329–330, 341–342; codeswitching 47, 49–51, 337; distinctive features 38, 52n, 333–341; domains of use 330–331t, 331–332; influence of French 49–50; as lingua franca 330, 331–332, 342; poetry 39, 47, 312; rap music 309–322; vocabulary 41–42, 47–48, 50–51 Haughen, E. 19 Hausa 251, 268, 269; as lexical source language 253, 258, 259; as matrix language 252 Heath, J. 134 Hemaya, Y. 295, 307n

hierarchical societies: Moorish ethnic group 38–40, 310–311, 312–313; Palestinians 60 Hilali dialects 77–78 Hillili, A. 102–103 hip-hop music 294 history, see names of individual places, e.g. Morocco Hmed, N. 150 Hogg, M. A. 98 Hojrup, T. 188, 190, 194, 195–196 Holes, C. 6, 9, 14, 17, 26n Hopkins, N. 193 humor 284–285, 298, 299, 301, 302, 304; and code-switching 21, 277, 282, 287–288; and place names 45 hypergamy 51n, 320 (-i) variable 110–112 Ibn Khaldoun 4, 5 identity 6, see also gender; Armenians 159; and code-switching 51–52, 284–285, 287–288; construction of 100, 116–119; and dialect maintenance 147–148; and dialect variation 15, 66–67, 98–100, 214–215, 226–227, 286, 336; Fessis 102–103, 112–120; and macho language 66–67; Moorish ethnic group 38–40; and politics 59, 60–61; religious groups 152–153, 158; and youth language 303–304 ideology, and dialect development 47–50 in-group/out-group identification 98 indefinite article 88, 89, 283 informants 65, 105, 130, 162–163, 190, 191t, 218–219, 233, 235, 328–329, 333 initialing 298 innovations 284–285, 286–288, 289n integration, migrants 9, 44–45 interspeaker accommodation 98 interviews 62–63, 105–106, 151, 190–192, 218–219 intonation 296 irony 282, 297 Islam, and rap music 314–316, 322 Islamic education (Islamiyya) 271 islamization, West Sahara region 37–38 Ismail, Hanadi 17

348

INDEX

Issa, A. 295 Italian, loan words from 90 Jastrow, O. 6, 177 Jeddah 17 Jewish dialects 84, 85, 86, 89–90 Jews 82–83, 93n Jordan 9, 14–15, 57; dialects 57–58, 204, see also Ammani dialect; history 58–62, 74n; nationalism 26n; population 58 Kampffmeyer, G. 103–104 Kanuri 251, 268 kakkakah/kaskasah 231, 232–233, 235–240, 243–244, see also pronominal suffixes; avoidance strategies 240–243; definition 230 Kerswill, P. 1, 116 Kharoufi, M. 8 Khattab, G. 204 Khelil, M. 139n Kimy, Papis 317–321, 323n koineization 26n, 83, see also urban vernaculars; Beirut dialect 161–162; definition 5; Moroccan Arabic 131–135, 136–138; regional 69, 71–72; Saudi Arabian dialects 232–233; and segregation 6, 7 Ksar el Kebir 13; history 216–218 Ksar el Kebir dialect: classification 218; phonological variation 216, 219–226, 227 Labov, W. 70, 75n, 161, 188, 214, 304 Lahlou, M. 287 Lakoff, R. 304 Landry, R. 98 language attitudes 302 language change: attitudes to 39–40, 52n, 303–304; and educational policies 231–232; leaders/early adopters 205–206, 208, 209, 214; and life-modes 188–189, 208–209; and modernity 239 language competence 267–268 language contact, see contact situations language diversity 20–24 language domains 18 language policies 81–82, 139 language status: classification 328–330; prestige dialects 214, 239, 240, 330

language variation research 1, 2, 10–14, 23 Laronde, A. 81 Le Page, R. 73, 74 leave-taking formulae 161 Lentin, J. 16, 198–199, 210n Levantine Arabic 204 leveling 7, 11–12, 16, 17, 137; Ammani dialect 63–64, 69–70; definition 116; Fessi dialect 114–117, 119–120; Saudi Arabian dialects 235–240, 243–244 Levy, J. 46 lexical borrowing: Cairene Arabic 300–302; Hassaniyya 10–11, 47–48; Mauritanian Middle Arabic 48–49; SanÉani Arabic 176–177; Tripolitan Arabic 89–91 lexical insertions, see also codeswitching: Hassaniyya 339–340; Moroccan Arabic 280–284; Nigerian Arabic 253–267, 269–270 lexical source languages 252–253, see also lexical borrowing; definition 252 Libya 8, 78–83, 92, see also Tripoli life-modes 194–196; definition 190; and language change 188–189, 208–209 lifestyle, and dialect classifications 5–6 linguistic deficit 275, 287 linguistic diffusion, urban hierarchy model 201 linguistic innovations 284–285, 286–288, 289n literacy 81–82 loan words, see lexical borrowing Lotfi, S. 292 Louiset, O. 42 lowering, vowels 70 macho language 66–67, 74n, see also toughness Maghreb, see North Africa Maiduguri 21–22, 250, 270–272, see also Nigerian Arabic maintenance 7, 103; Fessi dialect forms 106–112; and identity 112–114, 147–148 Marçais, P. 77, 85, 93n, 94n Marçais, W. 77, 87, 94n markedness 72, 204 marriage customs 51n, 320

349

INDEX

Marshall, J. 210 Martin, D.-C. 322 Martin, Y. 81 Mashreq, see Middle East Matrix Language Frame/Turnover 279–280 matrix languages 252, 279–280, 288n; turnover 286–287 Mattson, E. 150 Mauritania 10–11, see also Nouakchott; history 37–42, 42–44, 45–46, 309; population 35, 343n; rap music 309–322 Mauritanian Middle Arabic 48–49 meaning, changes in 299–300 media, attitudes to youth language 293–294 Medina culture 10, 13 Meiseles, G. 270 Meknes 22 Melliani, F. 282 men: and code-switching 285–286, 287; community activities 195, 196; and dialect variation 223–224, 224–226, 236–239, 278, 297; and language change 208, 209, 240; macho language 66–67, 74n Messaoudi, L. 12, 216 metaphors 299 methodology 218–219, see also informants; variables; corpus analysis 250–251; data analysis 155–159, 201, 205, 219–226, 235–239, 258–270, 328–333, 333–341; ethno-methodology 9–10, 98–100, 105–106; historical-dialectal 9, 77–78; interviews 62–63, 105–106, 151, 190–192, 218–219, 233, 235; lifemodes 190, 194–196; observations 153–155, 235, 333; variationist 1, 2, 9, 62–63, 98–100, 130–131, 190–192 metonyms 299–300 Middle East 9–10; educational policies 231–232, 239–240; research studies 14–18 migrants: communities 124–125, 135–136, 138; integration 9, 44–45 migration: and dialect variation 199, 215; and urbanization 8–18, 80–81, 101, 171; and war 58, 81, 148–150, 342n–343n Miller, C. 16, 26n, 108

Milroy, J. 116 Milroy, L. 147–148, 188 Mitchell, T. 270 mixed dialects 48–49 mixing, see code-switching MMA, see Mauritanian Middle Arabic Modern Standard Arabic 251–252, 329; and code-switching 276–278; and dialect variation 231–232; domains of use 45–46, 214, 269, 292, 294–295, 329; loan words from 252–253, 255–256, 257, 258–263, 266–268, 269–270; phonology 220; and urban vernaculars 9, 18–20, 240, 269–270 modernity: and code-switching 51–52; and language change 239; and toughness 13, 119–120 modernization 19, 60–62 Moorish ethnic group 37–42, 51n; music 309–312; social hierarchies 310–311, 312–313 Moroccan Arabic: classification of dialects 97–98, 140n; koineization 131–135, 136–138; lexical insertions 280–284; as matrix language 279; phonological/prosodic variation 285–286; youth language 22, 275, 276–278, 280–285, 287–288; in Zaragoza 136–138 Morocco 8, 10, 11–12, 13, 14; diaspora 12, 124–127, 135–136 morphological simplification 48–49, 110–112 morphology: areal norms 268–269; Cairene Arabic 300–301; Ceuti Arabic 133–135; Moroccan Arabic 280–282; Nigerian Arabic 254–258, 263, 264–266, 267; processing constraints 264–266; San¡ani Arabic 175–176; Tripolitan Arabic 85–88 morphosyntactic variation 70–73; Cairene Arabic 297–298; Casablanca dialect 106, 110–112; Hassaniyya 336–339 Moulard-Kouka, Sophie 322 Moumine, M. E. 104 MSA, see Modern Standard Arabic multilingualism 20–24, 328–333; and code-switching 251–253, 270–271; and rap music 311

350

INDEX

nouns 87–88, 282–284 nuclear family 193 Nweyri 159–161

music 22, 23, 24; traditional styles 92–93; and youth language 294, 301–302 Mus¡id wa Mus¡idih 173–177, 182–184, 185–186 Muslims 127–129, 158 Mutahhar, A. R. 173–174 Myers-Scotton, C. 276, 277, 279, 288n Naciri, M. 10 Naïm-Sanbar, S. 150, 161, 162, 164n names, and humor 45 national languages 45–46, 251, 329 Negro-Mauritanians 325, 326–327, 328–342 neighborhoods 188, 189, 192–194, 194–196 neo-urban vernaculars 12, 13, 27n, see also Ammani dialect; Casablanca dialect; Hassaniyya; Tripolitan Arabic neologisms 50, 294, 301–302 new city vernaculars, see neo-urban vernaculars new technologies, see communications technologies Niger Congo languages 334 Nigeria 21–22, 250 Nigerian Arabic 249; lexical insertions 250–251, 253–267, 269–270, 271; as matrix language 252; morphology 254–258, 263, 264–266, 267; phonological variation 266–267 nomads 39, 44–45, 81 non-Arabic languages, see also names of individual languages, e.g. French: influence of 5, 17, 20–22, 23, 123–124; loan words from 90–91 “normal” speech 116–117, see also leveling; standardization North Africa 9, see also individual places, e.g. Morocco; arabization 37–38, 78–80, 81–82, 288n, 325; dialect variation 204, 215–216; French dialects 285; research studies 10–14, 23; youth language 22 Nouakchott 10–11, 23, 36, 327, see also Hassaniyya; history 43–44, 310, 327–328; multilingualism 328–333; population 35, 44–45, 325, 326–328; rap music 309–322; urban identity 336; urban vernaculars 42

Oea 78, see also Tripoli offficial languages 20, 45–46, 81–82, see also standardization old city vernaculars, see also Fessi dialect: definition 27n; prestige 97, 240; and women 10, 11–12, 14, 112, 240 Oualata 41 Ould Cheikh, A. W. 39, 42, 44 Owens, J. 1, 21–22, 85, 86, 87, 139n palatalization 297, 307n Palestine 9, 14–15 Palestinian dialect 75n, 198 Palestinians 60 palm groves, cultivation of 41 Palva, H. 5, 48, 59, 60, 63 Panetta, E. 85, 86, 87 parler citadin, see old city vernaculars parler urbain, see neo-urban vernaculars parlers à aspect ruraux 27n particles 89, 134–135 pause, and glottalization 177–184 Pedersen, I. 19 Pereira, Christophe 11 performing, rap music 314 Peterson, J. L. 292 phonetics 52n, 75n, 84–85 phonological variation 13, 17, 214–215; Ammani dialect 63–70; Beirut dialect 150, 161–162; Cairene Arabic 296–297; Casablanca dialect 106–110; Ceuti Arabic 131–133; Damascus Arabic 197–210; Hassaniyya 334–336; and identity 214–215; Ksar el Kebir dialect 219–226, 227; Mauritanian Middle Arabic 48–49; Moroccan Arabic 285–286; Nigerian Arabic 266–267; Palestinian dialect 75n, 198; San¡ani Arabic 177–184 place names, and humor 45 plurals 87–88, 197, 254, 260, 263t, 265; as avoidance strategies 241 poetry 39, 47 politics: effect on research studies 10; and identity 59, 60–61, 67; and language policies 21

351

INDEX

population 1–2; ethnolinguistic groups 102–103; growth of 25–26, 44–45, 81, 166, 170–171, 343n; religious groups 101–102 possessives 88–89, 254–256, 260–262, 263t, 265, 283–284, 289n, 337–338 post-colonial states 2 power relationships 47–50, 67 pre-Hilali dialects 11, 77–78, 80, 83, 134; morphology 86–87; phonetics 84, 132–133; syntax 89 prestige dialects 214, 239, 240, 330 processing constraints, morphology 264–266 professional life-mode 190, 195–196, 202, 204 pronominal suffixes 70–73, 86, 197, see also kashkasha/kaskasa pronouns 70–73, 85–86, 88 prosodic variation 285–286 Pulaar 329, 331, 332, 337, 340 (q) variable 107–110 Quitout, M. 81, 91 (r) variable 106–107, 204–208, 210 Rabat 12 radio series 173–177, 182–184, 185–186 raising, vowels 67–68, 70 rap music 23, 24, 294; and Islam 314–316, 322; Mauritania 309–322; themes 310–312, 320; writing/ performing 314 Raymond, A. 10 reallocation 18, 66–67 reggae 93 relexification 89–91 religious education 271 religious groups 101–102; greetings 151, 152–153; identity 152–153, 158; segregation 83, 127–128, 148–150 research studies, see also data; informants; methodology; variables: Middle East 14–18; North Africa 10–14 restructuring 340–341, see also dialect variation rhyme games 298 Riyadh 233, 236, 237–238, 239 Rizk, Sherin 22–23 road signs, bilingual 46 Rowson, E. 295

rudimentary leveling 63–64 Ruiter, J. J. de 124–125, 139n rural dialects 5 rurals, in Casablanca 102, 115 SA, see Modern Standard Arabic Sabra 159–161 Salé 12 Salibi, K. 74n Sallam, A. 270 San¡a 16, 172; population 166, 170–171, 186n San¡ani Arabic 174–184 Saudi Arabia 17, 231–233, 239–240 Saudi Arabian dialects 231–244 “sawwa” + verb construction 257–258, 263, 265 Schilling-Estes, N. 116 second generation speakers 64–65, 105, 140n secret languages 295 sedentary dialects 5 segregation: economic 52n; ethnic 52n; and koineization 6, 7; religious 83, 127–128, 148–150 self-employed life-mode 190, 194–195, 204 Senegal music 310, 322 Shaghoor 189, 191, 192–193; dialect variation 202–203f, 205–208, 209; life-modes 194–195 Shami, S. 74n Shryock, A. 26n Siegel, J. 7, 111 signs, bilingual 46 Simon, Rachel 93n simplification 72, 110–112, 337–339 Skaka 235–236, 238–239, 239–240 slang 295 social hierarchies: Moorish ethnic group 38–40, 310–311, 312–313; Palestinians 60 sociolinguistic markers 161 Soninké 331, 332 Souriau, C. 81, 82 Soussis 102 Spanish 128–129, 135, 139 speakers, categories of 63–65 Sprachatlas von Syrien 72, 199 Srage, N. 150, 151, 161–162, 163n, 164n Standard Arabic, see Modern Standard Arabic

352

INDEX

standardization: Classical Arabic 4–5; definition 19, 116; and urban vernaculars 18–20, 215–216 stereotypes 161 students: Cairo 296–305; and codeswitching 275, 276–278, 280–285, 287–288; language attitudes 303–304; Nouakchott 326–327, 328–342 Stumme, Hans 83, 87–88, 90 substitution 298–300 suburbs 8–9, 188, 189, 192–194, 194–196 suffixes 70–73, 86, 197 Sult 59, 63–64, 74n; Sult dialect, phonological variation 215 syntax 88–89 Syrian Arabic 198–201 TA, see Tripolitan Arabic Tabouret-Keller, A. 73, 74 Taine-Cheikh, Catherine 10–11, 84, 325 tant (tante) 67, 74n Tarrius, A. 13 Tauzin, Aline 23, 311, 312 television, and youth language 293–295 territorialization 12–13, 219–226, 227, 330–333 Terry, D. J. 98 thanks 160 third generation speakers 65, 105, 140n toponyms, see place names toughness 13, 104, 113, 115, 118–119, 119–120, see also macho language towns 40–41 tribal affiliations 6, 16, 26n, 39, 60, 74n Tripoli 11, 78, 79, 80–83 Tripolitan Arabic: attitudes to 91–92; classification 83–84, 91; descriptions 83–91, 93n; history 82– 83; lexis 89–91; morphology 85–88; phonetics 84–85; syntax 88–89 Trudgill, P. 63, 73, 108, 201, 214 truncation 298 Tunisian Arabic 94n Turkish loan words 90 Türkmen, E. 90 urban urban urban urban

cultures 41 dialects, see urban vernaculars hierarchy model 201 music, see music

urban vernaculars, see also names of individual dialects, e.g. Hassaniyya: attitudes to 4–5, 7, 91–92, 97–98; classification 4–8; descriptions 26n; and migration 9–18; and Modern Standard Arabic 9, 18–20, 240, 269–270; and standardization 18–20, 215–216 urbanization 1–4, 24–25, 128; definition 42, 46; and dialect formation 147; and migration 8–18, 80–81, 101, 171; and music 309 variables 106, 150–151, 190, 191t, 197, 220–221 variationist sociolinguistics 98–100; research studies 1, 2, 9, 188 verbs 86–87, 133–135, 257–258, 263, 265, 268, 280–282, 336–337 Versteegh, K. 190 Vicente, Ángeles 12, 282 vowels 67–70, 85, 219–222 war 15, 60–61, 342n–343n; and migration 58, 81, 148–150 Watson, Janet 16 West Sahara region 37–38 Westernization, attitudes to 7–8 Williams, A. 116 Woidich, M. 16 Wolfram, W. 116 Wolof 329, 331, 332 women: and code-switching 286, 287; community activities 195, 196; and dialect variation 98–100, 223–224, 224–226, 236–239, 285, 297; and language change 208–209, 214, 304; and old city vernaculars 10, 11–12, 14, 112, 240, 278; and rap music 314; restriction of domains 61–62, 214 words, duration 222–224 work, attitudes to 194–196 writing music 314 Yemen 16, 167–171 YL, see youth language Yoda, S. 84, 85, 86, 88, 93n Younes, M. 75n youth culture 92–93 youth language 22–24; attitudes to 291, 292–295, 302; Cairene Arabic 291–292, 296–306; and change

353

INDEX

205–206, 207–208; and codeswitching 51–52, 275, 276–278, 280–285, 287–288; greetings 157; and identity 303–304; and media 293–294; and rap music 23, 24, 294, 309–322

Zaragoza 12, 125, 126; migrants 135–136, 138, 139, 140n zero marker: (h) variable 198, 199–201, 209–210; indefinite article 88, 89 Ziamari, Karima 21

354