Endangered Languages of the Caucasus and Beyond 9789004325647, 9789004328693

Table of contents :
Contents......Page 6
Preface......Page 8
Chapter 1 Consequences of Russian Linguistic Hegemony in (Post-)Soviet Colonial Space......Page 10
Chapter 2 The Contacts between the Ossetians and the Karachay-Balkars, According to V. I. Abaev and Marrian Ideology......Page 26
Chapter 3 Why Caucasian Languages?......Page 48
Chapter 4 International Research Collaboration on Documentation and Revitalization of Endangered Turkic Languages in Ukraine: Crimean Tatar, Gagauz, Karaim, Qrymchak and Urum Experience......Page 60
Chapter 5 Cases-Non-cases: At the Margins of the Tsezic Case System......Page 69
Chapter 6 Language Endangerment in the Balkans with Some Comparisons to the Caucasus......Page 88
Chapter 7 Instilling Pride by Raising a Language’s Prestige......Page 100
Chapter 8 Unwritten Minority Languages of Daghestan: Statusand Conservation Issues......Page 107
Chapter 9 Report on the Fieldwork Studies of the Endangered Turkic Languages......Page 117
Chapter 10 Empire, Lingua Franca, Vernacular: The Rootsof Endangerment......Page 131
Chapter 11 Endangered Turkic Languages from China......Page 144
Chapter 12 The Death of a Language: The Case of Ubykh......Page 160
Chapter 13 Diversity in Dukhan Reindeer Terminology1......Page 175
Chapter 14 How Much Udi is Udi?......Page 196
Chapter 15 Language Contact in Anatolia: The Case of Sason Arabic......Page 218
Chapter 16 Language and Emergent Literacy in Svaneti......Page 235
Chapter 17 The Internet as a Tool for Language Development and Maintenance? The Case of Megrelian......Page 253
Chapter 18 Linguistic Topography and Language Survival......Page 267
Chapter 19 And So Flows History......Page 284
Index......Page 298

Citation preview

Endangered Languages of the Caucasus and Beyond

The Languages of Asia Series Series Editor Alexander Vovin (Ecole des Hautes Etudes en Sciences Sociales (EHESS), France)

Editorial Board José Andrés Alonso de la Fuente (Universitat Autònoma de Barcelona) Wolfgang Behr ( University of Zurich) Uwe Bläsing (Leiden University) Bjarke Frellesvig (University of Oxford) Stefan Georg (University of Bonn) Juwon Kim (Seoul National University) Ross King (University of British Columbia) Dongho Ko ( Jeonbuk National University) Mehmet Ölmez (Istanbul Technical University) Toshiki Osada (Institute of Nature and Humanity, Kyoto) Laurent Sagart (CRLAO, Paris) Claus Schönig (Freie Universität Berlin) Marek Stachowski ( Jagellonian University of Kraków) Yukinori Takubo (Kyoto University) John Whitman (Cornell University)

VOLUME 15

The titles published in this series are listed at brill.com/la

Endangered Languages of the Caucasus and Beyond Edited by

Ramazan Korkmaz and Gürkan Doğan

LEIDEN | BOSTON

Cover illustration: “FLYING ‘WHITE’ SHAMAN IN ‘T’ LETTER SHAPE” by Ahmet Ali Aslan. Reproduced with kind permission. According to Altaian belief, the soul is carried off by the spirits eastward if the youth is destined to become a ‘White’ shaman. The flying shaman form, full of symbols referring to the cycle of life (snake), and the divine (Upper, Middle, and Lower World, Black and White), creates the letter “T” in the original alphabet used by Turkic peoples in their Orhun and Yenisei script. The Library of Congress Cataloging-in-Publication Data is available online at http://catalog.loc.gov LC record available at http://lccn.loc.gov

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface. issn 2452-2961 isbn 978-90-04-32564-7 (hardback) isbn 978-90-04-32869-3 (e-book) Copyright 2017 by Koninklijke Brill nv, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, ma 01923, usa. Fees are subject to change. Brill has made all reasonable efforts to trace all rights holders to any copyrighted material used in this work. In cases where these efforts have not been successful the publisher welcomes communications from copyright holders, so that the appropriate acknowledgements can be made in future editions, and to settle other permission matters. This book is printed on acid-free paper and produced in a sustainable manner.

Contents Preface vii 1 Consequences of Russian Linguistic Hegemony in (Post-)Soviet Colonial Space 1 Gregory D. S. Anderson 2 The Contacts between the Ossetians and the Karachay-Balkars, According to V. I. Abaev and Marrian Ideology 17 Johnny Cheung 3 Why Caucasian Languages? 39 Bernard Comrie 4 International Research Collaboration on Documentation and Revitalization of Endangered Turkic Languages in Ukraine: Crimean Tatar, Gagauz, Karaim, Qrymchak and Urum Experience 51 İryna M. Dryga 5 Cases-Non-cases: At the Margins of the Tsezic Case System 60 Diana Forker 6 Language Endangerment in the Balkans with Some Comparisons to the Caucasus 79 Victor A. Friedman 7 Instilling Pride by Raising a Language’s Prestige 91 George Hewitt 8 Unwritten Minority Languages of Daghestan: Status and Conservation Issues 98 Zaynab Alieva and Madzhid Khalilov 9 Report on the Fieldwork Studies of the Endangered Turkic Languages 108 Yong-Sŏng Li

vi

contents

10

Empire, Lingua Franca, Vernacular: The Roots of Endangerment 122 Nicholas Ostler

11

Endangered Turkic Languages from China 135 Mehmet Ölmez

12

The Death of a Language: The Case of Ubykh 151 A. Sumru Özsoy

13

Diversity in Dukhan Reindeer Terminology 166 Elisabetta Ragagnin

14

How Much Udi is Udi? 187 Wolfgang Schulze

15

Language Contact in Anatolia: The Case of Sason Arabic 209 Eser Erguvanlı Taylan

16

Language and Emergent Literacy in Svaneti 226 Kevin Tuite

17

The Internet as a Tool for Language Development and Maintenance? The Case of Megrelian 244 Karina Vamling

18

Linguistic Topography and Language Survival 258 George van Driem

19

And So Flows History 275 Alexander Vovin Index 289

Preface According to UNESCO, it is believed that at least half of the nearly 7,000 languages spoken around the world will cease to be used within the next 100 years. If this issue is neglected, people will lose not only their cultural heritage but also invaluable understandings about the history of all humankind. In other words, with the disappearance of unwritten and undocumented languages, humanity would lose not only cultural wealth but also important ancestral knowledge embedded, in particular, in indigenous languages. As many of these languages are no longer being handed down from generation to generation, the number of the native speakers of these languages is decreasing, leaving them endangered. For instance, the Ubykh language, one of the languages spoken in the Northwest Caucasus, became extinct when the last speaker Tevfik Esenç died in 1992. For such reasons, if efforts are not made to document the speech and cultural practices of those who use these languages, many speech forms will disappear along with the cultural heritage they embody. Within the context above, the 1st International CUA Conference on Endangered Languages was held by the Caucasus University Association (CUA) on 13-16 October 2014, at Ardahan, Turkey, in collaboration with the Turkish Language Society, Ardahan University and Harvard University. The goal of this “open by invitation conference” was to bring together 35 prominent scholars from all over the world who were actively engaged in research projects and partnerships on different aspects of language endangered, documentation and revitalization to create a forum to be able to discuss the problems of dying languages on both theoretical and practical levels. The motto of the Conference was “before shooting stars vanish” in the sense that the CUA views each language as a star in the sky that could turn into a meteor and believes that as many of these languages should be documented and/or revitalised as possible before they fade away. The regional focus of the Conference was global, with a particular emphasis on the languages and cultures of the Caucasus and the Asia. The aim was to inspire international discussion to a linguistic diverse and unique corner of the globe, through which glocal scholars would be able to benefit in their work. The conference program is available on the official web site (https://www.ardahan.edu.tr/CUAConference2014/). The region in which the 124 participant universities of the CUA are situated is home to a rich linguistic diversity, much of it highly endangered, making it particularly appropriate to situate the Conference at Ardahan. The primary intent of the Conference was to offer the invited scholars the opportunity to address relevant linguistic issues to articulate best practices

viii

preface

in the field. Hence the participants addressed issues such as the state of the field of language documentation, conservation and revitalization; experiences that reflect on establishing research centres and international research collaborations; or topics related to technology, data collection, archiving and preservation. The conference concluded with an action plan for which the ultimate goal was to establish an “Endangered Languages Research Centre” at Ardahan. We are very happy to announce that the Turkish Language Society kindly declared to support this Centre financially. All 35 participants were invited to submit their papers for this publication and 19 scholars submitted their work which represent the papers in this volume. We would like to thank all of the presenters in the panel sessions, who made the conference not only interesting but provocative as well. The Brill’s editors kindly accepted the current editor’s suggestion to publish these papers in the form of a special volume. We believe that the present volume forms a coherent collection to complement the previously published volumes of the Global Oriental. Special thanks go to our conference sponsors for their continued support: Turkish Language Society, Kyrgz – Turkish Manas University and the Turkish National Commission for UNESCO. Last but not least we wish to thank Smithsonion Institution, Foundation for Endangered Languages, and Living Tongues (Institute for Endangered Laguages) for their invaluable academic support. Prof. Ramazan Korkmaz Prof. Gürkan Doğan

Chapter 1

Consequences of Russian Linguistic Hegemony in (Post-)Soviet Colonial Space Gregory D. S. Anderson

Introduction Siberia – the very name can evoke a shudder. Why? True, it is the coldest inhabited part of the earth so a shiver, yes, but why a shudder? This is due to the fact that Siberia is known around the world and in Russia itself, where it constitutes the vast majority of the land mass of that giant nation,1 as a frozen wall-less prison. To be sure, Siberia served as a penal colony for the Tsarist Imperial Russian leaders and this tradition was institutionalized with murderous zeal by their successor Soviet hegemons. But this is only part of the sad story of the consequences of Russian/Soviet colonialism on the diverse native populations of Siberia. For the purposes of the present study I focus only on the linguistic consequences of Russian imperialism and hegemony on the Native Siberian peoples. I start with a brief overview of the diverse Native Siberian groups as they stood at the time of the initial colonialist expansion and exploitation in the 16th century in section 1. I turn in section 2 to an introduction to various preSoviet phases of Russian colonialism and hegemony over Native Siberian populations. In section 3, using post-Soviet census data, I discuss the issues of language shift and ethnic shame that move ever forward among the Native Siberian population groups in the post-Soviet colonial space. Finally in section 4 I present some structural linguistic consequences of Russian linguistic hegemony on the grammatical structures of the dwindling and receding languages of the vast Siberian territory.

1  Indeed we should say the entire Asian portion of Russia in a traditional, non-administrative understanding of the term Siberia.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_00�

2 1

Anderson

Native Siberia at the Time of Russian Contact

Siberia at the time of contact was home to several dozen languages belonging to a range of different language families. Moving west to east we find various northern Samoyedic (Nenets, Enets, Nganasan) and Ob-Ugric (Khanty, Mansi) speaking peoples in the western edge of Siberia between the Urals and the Ob-Irtysh River complex. These people mainly pursued reindeer breeding in the north and hunting and fishing economies in the southern parts of this region. To their east were found a range of Siberian Turkic groups (Tuvan, several Altai and Xakas groups, Shor, Chulym Turks, the Tofa and Siberian Tatars groups) in the southern regions and southern Samoyedic (Selkup, Kamasian, etc.) and Yeniseic groups, today represented only by the Ket and Yugh. The Turkic speakers were largely pastoral nomads, but this mixed with hunting/ fishing or reindeer breeding in the northern mountainous and swampy regions where these economic pursuits were more viable, while the Samoyedic and Yeniseic peoples originally pursued hunter/fishing economies with some limited reindeer-based economies, for example among the northern Selkup. The Tungusic speaking groups occupied a vast territory stretching eastward from central Siberia all the way to Sakhalin in the south, Kamchatka in the east, and the Russian Arctic Far East in the north. These include such languages as Evenki, Even, Negidal, Nanai, Udihe, Ulcha, Oroch and Orok. Tungusic peoples were engaged in either hunting pursuits or reindeer economies, depending on the environmental conditions. In the southern part of the western half of this territory, Tungusic speakers were in contact with the pastoralist Buriat, speakers of a Mongolic language, and their now extinct linguistic cousins the Soyot. On Sakhalin and in the Amur river area, Tungusic speakers were in contact with Nivkh (Gikyak) a riverine fishing-oriented people who speak a language isolate. In the southern and central part of Kamchatka, Itelmenspeaking people were found, and in the north Koryak who speak a ChukotkoKamchatkan language. In the northern part of the Tungus-speaking area were found to the east the reindeer-breeding Yukaghiric-speaking peoples like the Odul, Wadul, Chuvan and Omok, the latter two now extinct linguistically, and the Omok ethnically as well too. To the east of the Yukaghiric peoples were the reindeer-herding Chukchi, also of the Chukotko-Kamchatkan language family (which also includes Chukchi’s sister languages Kerek and Al’utor), while in the coastal parts of Chukotka, Chukchi-speaking people pursued seamammal hunting oriented economies similar to the local Eskimoic-speaking populations, the Sireniki, Naukan and Siberian Yupik. In short, Native Siberia at the time of Russian contact was home to a vast array of different peoples

Consequences of Russian Linguistic Hegemony

3

speaking several dozen languages belonging to a range of unrelated linguistic taxa or genetic units.

2

Phases of Colonialism and Hegemony in Native Siberia

Siberia was subjected to at least two very different patterns of colonialism during the Russian Imperial era. The first stage treated Siberia as revenue resource for the Imperial coffers (Forsyth 1992, Slezkine 1994). At this stage, during roughly three hundred years stretching from the 16th century across the 18th century, Siberia stood as a classic example of an exploitation colony. The expansion to Siberia was pursued to fulfill a Tsarist lust for fur, complete with forts to ensure the tax collection and protect the Crown’s interests. As long as the dreaded yasak or fur tribute tax was paid, often set at debilitatingly high rates, which resulted in a state of virtual enslavement for many families across generations, Native Siberian people were more or less free to continue their traditional livelihoods (Forsyth 1992, Slezkine 1994). When not met, harsh physical punishments (including death) were meted out (Forsyth 1992, Slezkine 1994). The next phase in the colonialist mistreatment of Siberia came in the phase of the penal-cum-settlement colony for which it is known the world over. Exiles with families were to begin an irrevocable shift in the demographics of Siberia. Territorial hegemony was expanded and economic hegemony over the already exploited Native Siberians solidified throughout the 19th century. Ultimately this was followed by attempts at spiritual hegemony over the Native Siberian populations as well, since inevitably with the exiles and newcomers came missionaries determined to bring the heathens of Siberia into the RussoChristian light. While several Native Siberian groups abandoned their languages in this period, the shift was from one indigenous language to another, generally accompanied by a change in traditional economic pursuits. Thus several former hunter-gatherer groups speaking Yeniseic (Arin, Assan, Kott, Pumpokol) or southern Samoyedic languages (Kamasian, Mator, Taigi) adopted a local Siberian Turkic lect when they adopted pastoral nomadism. Russian thus still had relatively little impact on the Native Siberian languages up until the 20th century, except for lexical borrowings which were generally phonologically accommodated to the indigneous sound systems (Anderson 1995). What changed all this was the rise of Soviet Siberia. Ideologically, the Native Siberian populations were problematic to place in the social order of

4

Anderson

Marxism-Leninism, seeing as many were engaged in Stone-Age economies, with little relevance to the urban worker milieu that engendered the Bolshevik Weltanschauung. An early idealism with respect to the proper means of delivering the message of Red Salvation to the exploited masses – i.e., it was seen as best to be delivered in their own tongues – was at best incompetently attempted and quickly abandoned. Native Siberians were definitely, as put by Slezkine (1994), last among equals in the fraternal brotherhood of nations moving to the great Soviet Socialist future. Ideological hegemony over these populations was only part of the goal. The environment of paranoia that spread in the late 1920s and accelerated through the 1930s during Stalin’s reign of terror and power consolidation led to requisitioning vast stretches of Native Siberian territory and putting them into the service of the State supported terror regime against its citizenry, with Native Siberian populations being disproportionately subjected to this. During the Second World War, massive numbers of people including entire ethnic groups from the Caucasus and hundreds of thousands of Ukrainians were deported to Siberia (Grenoble 2003). This further tipped the demographic scales away from Native Siberians. Moreover, officially the only access to education and economic and social advancement was through the medium of Russian, following the implementation of Stalin’s ‘second mother tongue’ policy (Grenoble 2003). To compound the issues even more, traditional lifeways were suppressed (collectivization of herds, shamans executed as kulak-overlords), which further disrupted the populations and eroded the Native Siberian cultural fabric. The result of all of these trends was that there began a massive shift away from Native Siberian ethnic identities and shift to the now supremely dominant Russian language. We can see this manifest itself in both language retention rates, reported rates of ethnic affiliation, discussed in section 3 below, and in the wholescale restructuring of the grammars of the minority languages actually still used by Native Siberians, which I turn to in section 4.

3

Census Data and Indicators of Ethnic Shame & Language Shift

All things being equal, minority population groups should grow steadily unless some catastrophe has occurred like epidemic disease or war that has fractioned the population. Indeed we can see this trend when looking at census returns on ethnic affiliation for some (but not all) Native Siberian groups. If we look at these trends as reported in six successive census returns from 1959-2010, four in the Soviet period and two in the post-Soviet period, for groups like the Nenets,

5

Consequences of Russian Linguistic Hegemony

the Evenki, the Dolgan or the Yukaghir, everything appears normal in terms of demographic growth. Population figures from census records (1) 1959

Nenets 23,007 Evenki 24,151 Dolgan 3,932 Yukaghir 442

1970

1979

1989

2002

2010

28,705 25,471 4,877 615

29,894 27,294 5,053 835

34,665 30,163 6,945 1,142

41,302 35,527 7,261 1,509

44,640 38,396 7,885 1,603

For many other groups however, there are disturbing trends revealed, and the number actually declines in many cases. What the demographic decline numbers actually reveal is not loss of physical representatives/actual population members, but rather the decline in the subjective evaluation of being associated with that identity. This is a process called ethnic shame. Thus, most Native Siberian groups have experienced, or are now undergoing, rapid spread of ethnic shame and are hiding their identity in favor of reporting as Russian. We can divide the Native Siberian groups in roughly four such groups as they reflect somewhat different trends in the spread of ethnic shame. Two groups from western Siberia, the Khanty and Mansi experienced this only during the Soviet era between 1970 and 1979. This period marked a massive upturn in the exploitation of the extractive oil and gas industries on their territories. Since then, the groups that had self-identified as Mansi has grown steadily while Khanty experienced an up-turn in ethnic identity between 1989 and 2002, but then a more plausible increase between 2002 and 2010. (2)

Khanty Mansi

1959

1970

1979

1989

2002

2010

19,410 6,449

21,138 7,710

20,934 7,563

22,521 8,474

28,678 11,432

30,943 12,269

6

Anderson

Four other Native Siberian groups began showing this spread of ethnic shame already in the same period, but this has continued and/or accelerated. These are the Oroch, Nganasan, Ket and Nivkh. The various fluctuating numbers show local trends that distinguish each of these groups. Oroch began this decline in ethinc identity and spread of ethnic shame in the period between 1979 and 1989, the other three between 1970 and 1979. Nivkhs experienced normal growth reports between 1970 and 1979. Nganasans returned a serious upturn in self-reported ethnic identity between 1979 and 1989, but experienced steep decline between 1989 and 2002. Kets showed a reverse trend with continued decline between 1979 and 1989 and an up-turn in the immediate post-Soviet decade (1989-2002). However, all four report declines in the most recent census. (3)

Oroch Nganasan Ket Nivkh

1959

1970

1979

1989

2002

2010

782 748 1,019 3,717

1,089 953 1,182 4,420

1,198 867 1,122 4,397

915 1,278 1,113 4,673

686 862 1,494 5,162

596 834 1,219 4,652

Indeed this is the trend one finds across Siberia: the process of ethnic shame has advanced and spread considerably in the post-Soviet period. Some groups began to show this decline in the period between the end of the USSR and the first post-Soviet census (1989-2002), a group which includes four Tungusic languages, Nanai, Ulcha, Udihe and Negidal, plus Koryak and Aleut. This process has continued in the 2010 census returns. (4)

Nanai Koryak Ulcha Udihe Aleut Negidal

1959

1970

1979

1989

2002

2010

8,026 6,287 2,055 1,444 421 —

10,005 7,487 2,448 1,469 450 537

10,516 7,879 2,552 1,551 5,46 5,04

12,023 9,242 3,233 2,011 702 622

12,160 8,743 2,913 1,657 540 567

12,003 7,953 2,765 1,496 482 513

7

Consequences of Russian Linguistic Hegemony

The last group are just recently revealing these trends in ethnic shame and selfinvisibilization. The 2010 census returns show the first reported demographic decline among the Chukchi, Sel’kup, Itelmen and Tofa. (5)

Chukchi Sel’kup Itelmen Tofa

1959

1970

1979

1989

2002

2010

11,727 3,768 1,109 586

13,597 4,282 1,301 620

14,000 3,565 1,370 763

15,184 3,612 2,481 731

15,767 4,249 3,180 837

15,098 3,649 3,193 762

Unsurprisingly, language retention rates for Native Siberian languages show even further decline. What makes this particularly alarming is that the reported language numbers are generally inflated massively (Grenoble 2003; Anderson 2010, 2011). Thus, while the 2002 census of Tofa reported 378 speakers, I did an actual door to door survey in the villages Tofa is spoken in myself in 2001 and the actual number then was under 40! Indeed Native Siberia is one of the most extreme areas of language endangerment and thus represents one of the hottest of the Language Hotspots (Anderson 2010, 2011, in preparation). The vast majority of languages are either moribund or seriously endangered. Even the groups whose populations do not show advancing ethnic shame are rapidly shifting to Russian and away from their ancestral languages. Compare the reported retention rates for the censuses between 1959 and 2002 for the above mentioned five sets of languages. We see lower percentages across the board as a rule. Some languages only show this decline in reported numbers. Where numbers increase, we are dealing with specific social conditions that reflect more nostalgia for, and identity with, the ancestral language, and not actual linguistic practices. This is in part triggered by the Russian form rodnoj jazyk that census takers use, which is a phrase that connotes something more like the language of your heart and your heritage than the one of your tongue and brain. And all these upwards trends are followed in subsequent returns by more realistic declines.

8

Anderson

(6) 1959

1970

1979

1989

2002

Nenets Evenki Dolgan Yukaghir

84.7% 54.9% 93.9% 52.5%

83.4% 51.3% 89.8% 46.8%

80.4% 42.8% 90.0% 37.5%

77.1% 30.4% 81.7% 32.8%

75.8% 21.3% 67.0% 40.0%

Khanty Mansi

77.0% 59.2%

68.9% 52.4%

67.8% 49.5%

60.5% 37.1%

47.3% 24.0%

Oroch Nganasan Ket Nivkh

68.4% 93.4% 77.1% 76.3%

48.6% 75.4% 74.9% 49.5%

40.7% 90.2% 61.0% 30.6%

18.8% 83.2% 48.3% 23.3%

37.5% 60.6% 32.5% 13.3%

Nanai Koryak Ulcha Udihe Aleut Negidal

86.3% 90.5% 84.9% 73.7% 22.3% —

69.1% 81.1% 60.8% 55.1% 21.8% 53.3%

55.8% 69.0% 38.8% 31.0% 17.8% 44.4%

44.1% 52.4% 30.8% 26.3% 28.3% 28.3%

32.0% 34.5% 25.2% 13.7% 32.4% 25.9%

Chukchi Sel’kup Itelmen Tofa

93.9% 50.6% 36.0% 89.1%

82.6% 51.1% 35.7% 56.3%

78.2% 56.6% 24.4% 62.1%

70.3% 47.6% 19.6% 43.0%

49.1% 38.6% 12.1% 45.2%

Russian linguistic hegemony and language ideology that valorizes Russian as inherently superior to other languages of Russia, and enfranchised through the institutionalization of the Russian language as the only medium to access to transnational culture, education, social and economic advancement spells the doom for most Native Siberian languages. Since it is unlikely these attitudes will change soon, it is unlikely that the rapid decline of Native Siberian languages will be reversed. These processes have accelerated significantly in the post-Soviet period, possibly as a result of the removal of any pretense of state support for indigenous minority identity in Siberia. Indeed it is not

Consequences of Russian Linguistic Hegemony

9

a stretch to say that most of the language families of Siberia will be extinct before 2100.

4

Other Consequences of Russian Linguistic Hegemony: Codeswitching and Restructuring

Since basically all Native Siberian people outside of Tuva (which only joined the USSR in the 1940s) are fluent in Russian, it is no surprise that codemixing or codeswitched utterances are commonplace in the speech of those Native Siberians who continue to use the minority languages, such as the following mixed Russian-Chulym Turkic sentence. (7) Chulym Turkic (Ös) kør-ze-m na nas kakoj=ta ʃybyr moɣalaq [kør-ybyly] see-cond-1 R.at R.us R.some.such poor bear [look-prs] ‘I look and what do I see but some poor bear (looking) at us’ (Anderson and Harrison 2004: 184) Code-switching with Russian is still not very well explored for almost any minority language of the Russian Federation and will likely advance the study of this phenomenon greatly when it can be better investigated. For example, there is complex gender agreement interactions found in Erzya-Russian codeswitching (Janurik 2015) and similar phenomena are found in most languages of Russia today, but the topic has not been explored in most of them. 4.1

Evidence of Contact-Induced Restructuring I: Syntax of Complex Sentences For those languages still extant, Russian structures are penetrating into all domains of grammar in contemporary Native Siberian language usage. This includes restructuring the basic syntactic structures of the languages, e.g., morphosyntactic projections or frames with respect to case assignment to arguments by predicates (4.2) or the whole-scale restructuring of complex sentence structure towards ones where the embedded or dependent clauses has a finite verb and is headed by a complementizer, relative pronoun or subordinator and follows the matrix clause rather than complex sentences with embedded clauses preceding the matrix clause and appearing with a nonfinite (participle or converb) verb form and no clause initial complementizer, relative pronoun or subordinator. I will use mainly data from Siberian Turkic languages (Anderson 2004, 2005) to exemplify the range of changes found

10

Anderson

resulting directly from Russian contact and the adoption of Russian linguistic norms into these originally quite different and distinct language systems. One area of the syntax of complex sentences that has a clear Russian origin and one which was originally entirely alien to Siberian Turkic structure is the use of a clause-initial complementizer followed by a finite verb of the type S COMP S (8) instead of the original structure of S + S (9), often with a genitive marking on the subject of the embedded clause to further show the non-finite or nominalized quality of the embedded predicate. The borrowed Russian complementizer ʃto introduces this new finite complement clause in (8). (8) Abakan Xakas noɣa sɪler saɣɯn-tʃa-zar why you.pl think-prs.i-2pl ‘why do you think I am scared?’ (Anderson 2005: 197)

ʃto min xorɯx-tʃa-m pycck.comp I be.scared-prs.i-1

(9) Abakan Xakas sirer-nɪŋ irtɪ nan-dʒaŋar you.pl-gen early return-modal.nonfin:2pl sal-tɯr-bɯn pfv-evid.pst-1 ‘I forgot that you had to leave early’ (Anderson 2005: 203)

min I

undu-p forget-cv

Another feature of Russian syntax now found in high-contact varieties of Siberian Turkic include relative clauses with finite verbs introduced by relative pronouns that follow the noun relativized on (10). Both such features are alien to Siberian Turkic which rather used non-finite pre-nominal participle forms and no relative pronouns, still found in the speech of some less restructured speakers (11). (10) Abakan Xakas sin pil-bin-tʃe-zɪŋ ol kɪʒɪ-dɪ you know-neg-prs.i-2 that person-acc paz-ɯbɯs-xa-m write-prf-pst-1 ‘you don’t know that person I wrote to’ (Anderson 2005: 206)

xajzɯ-nɣa min which-3.dat I

Consequences of Russian Linguistic Hegemony

11

cf. (11) Abakan Xakas ol kør-gen pyyr-neŋ min xorɯx-pas-tɯɯx-pɯn he see-pst.prtcpl wolf-gen I fear-neg.irr-sbjnctv-1 ‘I wouldn’t have been scared of the wolf he saw’ (Anderson 2005: 205) Another Russian feature that has found its way into Native Siberian languages is the use of a scopeless negative operator and a finite verb together with a borrowed subordinator in a type of temporally subordinate clause generally introduced by ‘until’ or ‘before’ in English translations, and by poka in the original Russian and restructured Siberian Turkic forms. So compare for example (12) and (14) with borrowing/restructuring with the original formations in (13) and (15). (12) Abakan Xakas poka pol-bas-tar subord be-neg.fut-pl ‘until/before it gets cold’ (Anderson 2005: 220)

soox-tar frost-pl

(13) Abakan Xakas soox pol-ɣandʒa frost be-cv ‘until/before it gets cold’ (Anderson 2005: 219) (14) Abakan Xakas poka turu-bas-pɯn subord stand-neg.fut-1 ‘until I stand’

(15)

Abakan Xakas min tur-ɣandʒa I stand-cv ‘until I stand’ (A05: 219-20)

Intermediate stages in this shift can be found in the speech of some speakers too. Thus in (16) we find a form introduced by the borrowed subordinator and with the scopeless negative operator, but with a non-finite verb as well (here in a participle form marked also with the locative case). (16) Abakan Xakas poka pɪs par-ba-an-de ib-zer subord we go-neg-pst.prtcpl-loc house-all ‘before/until we went home’ (Anderson 2005: 219)

12

Anderson

This Russian-originated structure has proven to be particularly easy to borrow. In addition to the Turkic data above, we find similar formations in Yeniseic languages too. So for Yugh Verner (1997: 194) lists poka + bəɲ [neg] + Verb with a fully restructured subordinator + scopeless negative operator + finite verb. In S. Ket on the other hand in the 1970s one finds a similar structure with a calqued and not a borrowed subordinator asjka . . . bən Verb (Kostjakov 1976: 59). In Central Ket there was variation between the original non-finite (‘converb’) structure competing with an intermediate one introduced by the calqued subordinator + scopeless negative operator, i.e., Verb=baŋdiŋa alternating with asjka . . . bən . . . Verb=baŋdiŋa. Central Ket (Grishina 1977: 105). 4.2

Contact-Induced Restructuring II: Case Usage, Morphosyntax and Shift in Verbal Subcategorization Frames Complex sentence structure is not the only domain that shows the tangible effects of contact with Russian. More subtle morphosyntactic patterns also reflect the hegemonic pressure Russian has exerted over Native Siberian languages. One such area is the introduction of dative case for impersonal subjects (17) whereas the original structure still is used by less restructured speakers used the nominative (18) and triggered verb agreement with that referent as the subject. (17) Abakan Xakas maɣaa nan-arɣa I:dat return.home-inf ‘I have to go home’ (Anderson 2005: 170)

kirek nec

(18) Abakan Xakas min ib-zer par-arɣa kirek I house-all go-inf nec ‘I had to go home’ (Anderson 2005: 175)

pol-ɣa-m aux-pst-1

While originally essive constructions were marked by a postposition derived from a converb form of ‘be’ (i.e., ‘being X’ > ‘as X’), and this formation can still be found for example in the speech of the Bel’tir Xakas (20), in urban vernacular varieties one finds the very Russian use of instrumental case in this function (19).

13

Consequences of Russian Linguistic Hegemony

(19) Abakan Xakas ol traktorist-peŋ toɣɯn-tʃa he tractor.driver-ins work-prs.i ‘he works as a tractor-driver’ (Anderson 2005: 170) cf. (20) Bel’tir Xakas dojarka pol-ɯp toɣɯn-tʃa-m milkmaid be-cv work-prs.i-1 ‘I work as a milkmaid’ (Subrakova 1992: 46) Lastly, the introduction of new functions to cases also extends to the genitive in the speech of high-contact varieties of Siberian Turkic, such as Tofa (21) or Abakan Xakas (22). Genitive case previously never was the case subcategorized for by any verb as the form of its complement/object/2nd argument in the Turkic languages, but now it is used in these highly restructured varieties as the case that is projected onto the complement of the verb ‘fear, be afraid of’, a pattern of usage that clearly and obviously reflects Russian norms. Originally and still in less restructured varieties of these Turkic languages the ablative case was/is used (23). (21) Tofa kør-gen-ɪ irezaŋ-nɯŋ see-pst.prtcpl-def bear-gen ‘I was not afraid of the bear I saw’ (Field Notes 2001, SDA-Bear Story) (22) Abakan Xakas ol pyyr-nɪŋ xorɯx-tʃa he wolf-gen fear-prs.i ‘he is afraid of the wolf’ (Anderson 2005: 179)

(23)

men I

kòrt-pa-an men fear-neg-pst 1

Abakan Xakas olar pyyr-deŋ xorɯx-tʃa-lar they wolf-abl fear-prs-3pl ‘they are scared of the wolf’ (Anderson 2005: 180)

Some contact-triggered restructuring towards Russian norms is found throughout the languages of Siberia. For example Gruzdeva (2015) describes a host of changes in Nivkh, such as the importation of forms of ‘give’ into the imperative paradigm as first person imperative forms (24), calqued on the Russian model (25),

14

Anderson

(24) Nivkh t‘ana ñ-aχ give.imp.2sg I-acc ‘let me sing’ (Gruzdeva 2015: 171)

lu-gu-ja sing-caus-imp.2sg

(25) Russian daj spo-ju give.imp sing-fut.1sg ‘let me sing’ (Gruzdeva 2015: 171) or the calqued use of clause-initial conditional marker aif that is redundant, as there is conditional inflection already on the verb, but used exactly as Russian esli is (Gruzdeva 2015: 174). Nor are Siberian languages unique among the subjects of linguistic hegemony in the post-Soviet colonial space, although they are certainly the most severely affected. Similar syntactic restructuring has been reported in such diverse languages under Russian dominion as Udmurt (Kaysina 2015) and Chechen (Guerin 2015).

5

Summary

Five centuries of Russian colonialism and hegemony have left the once rich diversity of languages in Native Siberia in steep decline. Economic exploitation followed by waves of settlement colonization altered the demographics of Siberia forever, and Russian emerged as the only language with a place longterm in the linguistic market. Pressures came from below and above, and language abandonment and ethnic shame have arisen and spread. The future is very bleak for the languages of Siberia. Before they disappear entirely, massive structural borrowing and incorporation of loans from Russian is their fate. Only adequate documentation undertaken now can leave future generations of Native Siberians the legacy they deserve should the social and economic conditions ever once again become favorable to reclaim these identities.

References Anderson, Gregory D. S. 1995. Diachronic Aspects of Russianisms in Siberian Turkic. In Berkeley Linguistics Society 21. Berkeley: Berkeley Linguistics Society, pp. 365-76.

Consequences of Russian Linguistic Hegemony

15

Anderson, Gregory D. S. 2004. The Languages of Central Siberia: Introduction and Overview. In E. Vajda (ed.) Languages and Prehistory of Central Siberia. Amsterdam: John Benjamins, pp. 1-119. Anderson, Gregory D. S. 2005. Language Contact in South Central Siberia. Turcologica 54. Wiesbaden: Harrassowitz Verlag. Anderson, Gregory D. S. 2006. Towards a Typology of the Siberian Linguistic Area. In Yaron Matras, April McMahon, and Nigel Vincent (eds.) Linguistic Areas. Convergence in Historical and Typological Perspective. Basingstoke: Palgrave Macmillan, pp. 266-300. Anderson, Gregory D. S. 2010. Perspectives on the global language extinction crisis: The Oklahoma and Eastern Siberia Language Hotspots. Revue Roumaine de Linguistique XLV: 129-142. Anderson, Gregory D. S. 2011. Language Hotspots: what (applied) linguistics and education should do about language endangerment in the twenty-first century. Language and Education 25 (4): 273-289. Anderson, Gregory D. S. in preparation. Language Extinction. Cambridge University Press. Anderson, Gregory D. S. and K. David Harrison. 2004. Shaman and Bear: Siberian Prehistory in Two Middle Chulym Texts. In E. Vajda (ed.) Languages and Prehistory of Central Siberia. Amsterdam: John Benjamins, pp. 179-197. Forsyth, James. 1992. A History of the Peoples of Siberia. Russia’s North Asian Colony 15811990. Cambridge: Cambridge University Press. Grenoble, L. 2003. Language Policy in the Soviet Union. Dordrecht: Kluwer. Grishina, N. M. 1977. Upotreblenie slova “bang” v slozhnom predlozhenii ketskogo jazyka [Use of the word “bang” in Ket complex sentences]. Jazyki i toponimija 4: 102-107. Gruzdeva, Ekaterina. 1998. Nivkh. Languages of the World Materials LW/M 111. Munich: LINCOM-EUROPA. Gruzdeva, Ekaterina. 2015. Sociolinguistic and linguistic outcomes of Nivkh-Russian language contact. In Stolz (ed.), pp. 153-181. Guerin, F. 2015. The evolution of Chechen in asymmetrical contact with Russian. In Stolz (ed.), pp. 183-198. Janurik, B. 2015. The emergence of gender agreement in code-switched verbal constructions in Erzya-Russian bilingual discourse. In Stolz (ed.), pp. 199-217. Kasyina, I. 2015. Grammatical effects of Russian-Udmurt language contact. In Stolz (ed.), pp. 219-235. Kostjakov, M. M. 1976. Ketskie sootvetstvija russkomu sloznopodchinennomu predlozheniju s pridatochnym vremeni. [Ket correspondences to Russian subordinate sentences with adjectival-subordinate tense] Jazyki i toponimija 1: 56-62. Slezkine, Y. 1994. Arctic Mirrors. Russia and the Small Peoples of the North. Ithaca, New York: Cornell University Press.

16

Anderson

Stolz, Christel (ed.) 2015. Language Empires in Comparative Perspective. Berlin: Mouton de Gruyter. Subrakova, O. P. 1992. Padezhnaja sistema v bel’tirskix govorax xakasskogo jazyka. [The case system of the Bel’tir variety of Xakas] Xakasskaja dialektologija, pp. 32-50. Abakan: XakNIIJALiI. Verner, G. K. 1997. Jugskij jazyk. [The Yugh language] In A. P. Volodin (ed.) Jazyki mira: Paleoaziatskie jazyki. Moscow: Indrik, pp. 187-195.

Internet Sources http://www.perepis2002.ru/ http://www.tooyoo.l.u-tokyo.ac.jp/Russia/bibl/

Chapter 2

The Contacts between the Ossetians and the Karachay-Balkars, According to V. I. Abaev and Marrian Ideology Johnny Cheung

A modest tribute to Uwe Bläsing and his forensic approach to etymology and the origin of words

⸪ 1

Introduction

Ossetic is geographically the most western East Iranian language spoken in the Northern Caucasus. It is spoken in two areas, viz. in North OssetiaAlania, which is part of the Russian Federation, and in South Ossetia, which declared its independence in 2008 in the aftermath of the Georgian-Russian war. According to the Russian census of 2010, around 488,254 persons declare Ossetic as a first language. The dominant dialect is Iron, which also serves as the main or official language for the Ossetians. A minority, less than 100,000 speaks the relatively archaic dialect of Digoron, which is used predominantly by the Sunni Muslim minority in North Ossetia-Alania. In contrast, the Karachays and Balkars speak two very closely related Turkic languages that are usually classified as “West Kıpçak”. They are settled primarily in two Russian republics, viz. in Kabardino-Balkaria (Balkars) and in KarachayCherkessia (Karachays). Both areas are situated in the Northern Caucasus region of the Russian Federation. Although the Karachays and the Balkars share the same standard, literary language, often simply called Karachay-Balkar, it is mostly based on the speech of the numerically superior Karachays, 218,403, vs. 112,924 Balkars, according to the Russian census of 2010. Historically, the language had never acquired literary status, as the speakers would have resorted to writing Arabic and / or Russian instead, until the Soviet government commissioned the introduction of this literary Karachay-Balkar in 1935/6. According

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_003

18

Cheung

to the same census, 212,522 of the combined total of Karachays and Balkars declare to use Karachay-Balkar natively. The Karachay-Balkar settlements are divided in two contiguous political units, but unlike the Ossetians in their home regions, the Karachays and Balkars do not constitute a majority in their respective republics, where there are sizeable, ethnic Russians and Caucasianspeaking Kabardino-Cherkess (also known as Circassians). Historically, the Scythians, Alans, and Sarmatians are considered the linguistic ancestors of the modern Ossetians, although, evidently, the linguistic documentation is rather meagre and often limited to personal names, the occasional quote in a Classical Greek source (such as Herodotus), and grave inscriptions. The same may apply to the attempts to establish a direct linguistic link between the modern day Karachay-Balkars and the (presumably) Kıpçak speaking Cumans and Pechenegs, if not including the other elusive Bolghars with their unclear Turkic affiliation. This is obviously a cause of disagreement. Of course, there are other, mostly lesser-known Turkic languages spoken in the Caucasus, such as Nogay (a South/Central Kıpçak or “Aralo-Kaspian” Turkic language), Kumyk (West Kıpçak), who might also lay claim on these historically attested peoples and tribes. Ossetic and Karachay-Balkar are not in imminent threat of extinction, as these languages have an enshrined position within the political framework of their autonomous republics. Language retention of the native language among the Ossetians and the Karachay-Balkars is high, despite the omnipresence of Russian, which is the language of education and serves as the natural lingua franca among the many Caucasian nationalities. In their respective republics, of the 459,688 North-Ossetians, 402,248 of them indicated that they had a command of Ossetic (87.5%), whereas of the 194,324 Karachays living in the Karachay-Cherkessian Republic 181,740 had a command of Karachay-Balkar (93.5%) and in the case of the 108,577 Balkars in the Kabardino-Balkarian Republic, 96,252 (88.6%), according to the census of 2010.1 Although, in the larger towns and cities, Russian is heard pretty much everywhere, Ossetic and Karachay-Balkar respectively are usually the everyday language of communication in the country side. The level of marginalisation in the former Soviet and contemporary Russian society differs considerably for both communities. Ossetians are well 1  All figures are cited from the documents available at http://www.gks.ru/free_doc/new_site/ perepis2010/croc/perepis_itogi1612.htm.

The Contacts between the Ossetians and the Karachay-Balkars

19

integrated in mainstream Russian society and are therefore treated relatively favourably, especially after the annexation of South Ossetia by Russia in 2008. In contrast, the Karachay-Balkars are generally viewed with some suspicion, because of their religion (Islam) and possible ties to other Turkic groups and communities in Russia and abroad, including possible, political aid and interference from Turkey. Even their full rehabilitation and measures to compensate for the wrongdoings in the past were not in place until Boris Yeltsin signed an official decree on March 3, 1994, which restored their cultural rights in their assigned Republics. Compounding to their rather marginal position in Russian society is the dearth of prominent Karachay-Balkar intellectuals, who could speak and carve out a cultural and political space for their communities in Russian society. The scholarly study of Ossetic and Karachay-Balkar did not start in earnest until the second half of the 19th century, right after the conquest of the Caucasus by the Russians. Ossetic was studied in depth, thanks to the efforts of the Finnish scholar Anders Sjögren and a prominent Russian scholar, Vsevolod Miller. Subsequently, the native Ossetian scholar Vassiliy Abaev would build on their works. Especially the historic relations of Ossetic with the other Iranian languages, several European language groups (Slavic, Celtic), and, of course, with Turkic too, became much better known. Abaev has also published several important synchronic descriptions of Ossetic and textual editions of Ossetic myths and folklore. Karachay-Balkar, on the other hand, lacks similar kinds of wide-ranging research, such as the interpretation of the customs, the historic dimensions of the language, possible contacts with other ethno-linguistic groups, and so on. Actually, it was not until at the turn of the 20th century that the Russian linguist Nikolaj Karaulov recorded the Karachay-Balkar in earnest (Karaulov 1908). Even an in-depth description of the dialects of Karachay-Balkar has yet to appear. A bilingual Karachay-Balkar – Russian dictionary did not appear until 1989 (Tenišev 1989). The older Russian-written literature on KarachayBalkar was meant as an aid to help the Karachay-Balkars to master Russian. The impact of the contributions made by linguists, especially in Russia and the Soviet-Union, on the intellectual formation and appraisal of the “mother” language by its speakers is undeniable though. Ossetic was actively researched by Russian and, later, Soviet scholars, because it was Indo-European and spoken by a largely Christian population, in a sea of largely non-Indo-European languages with sizeable Muslim populations. For this reason, their speakers received a relatively favourable treatment, in comparison to other minorities.

20

Cheung

The situation for Karachay-Balkar could not be more different. The Karachay-Balkars speak a Turkic tongue, and are largely Muslim, and therefore, potentially hostile, overtly or latently, to the official, atheistic Soviet system. The Karachay-Balkars were deliberately broken up in two ethnic designations, despite the clear ethnic communalities and almost identical language. In addition, they were “housed” in separate republics, which they also had to share with unrelated groups, viz. the Cherkessians and Kabardinians respectively. Actually, these Cherkessians and Kabardinians speak very closely related West Caucasian languages. The Soviet authorities actively pursued this kind of ethnic or tribal fragmentation, creating micro-nationalities against the wishes of the local intellectuals, as observed by Alexandr Bennigsen (1983). Obviously, the ulterior motive is “divide and conquer” in order to prevent potentially big challenges to Soviet rule. Russian or Soviet research on Karachay-Balkar was rather limited, other than within the context of its status as a minor Turkic language. According to the School of the prominent Soviet scholar Nikolaj Marr, Karachay-Balkar was a kind of linguistic “mongrel”, the result of “crossbreeding” (скрещения) between Turkic and “Svan” elements, as described in a speech at a 1929 meeting of the Soviet Academy of Sciences: “. . . The timeliness and urgency to study this language [of the Karachay-Balkars, JC] is because of the established relations of the Svan with the Turkic languages of the aforementioned peoples. This makes it possible to identify the process of crossbreeding on the one hand, but on the other hand, to establish the presence of Japhetic elements in those languages, and, therefore in Turkic generally, insomuch Balkar and Karachay appear as the languages of the said system.”2 Even worse, at the end of the Second World War many nationalities were the target of large-scale deportations, such as the Ingush, Chechens, Karachays, Balkars, the Buddhist Kalmyks and the Muslim Digor Ossetians, who were accused of collaboration with the retreating Nazi German troops. One can 2  “. . . . Своевременность и срочность изучения этого языка объясняется установленными связями сванского с тюркскими языками названных народов, что дает возможность с одной стороны выявить процессы скрещения, с другой – установить наличие яфетических элементов в указанных языках, а следовательно и в тюркских вообще, поскольку балкарский и карачаевский являются языками именно этой системы.” In Письмо Н. Я. Марра в Президиум АН СССР с обоснованием необходимости экспедиции в Кабардино-Балкарию и перечнем предполагаемых ее участников [Letter of N. Ya. Marr at the Presidium of the Academy of Sciences of the Soviet-Union on the necessity for an expedition to Kabardino-Balkaria and a list of potential participants], 19 February 1929 (Letter 141, reprinted by Anfert’eva and Kazanskij 2013, 214 f.).

The Contacts between the Ossetians and the Karachay-Balkars

21

notice the huge gap in (pan-)Soviet publications on Karachay-Balkar, nothing was published between 1941 (Bizni zamannı ğigiti, Nalčik) and 1960. The relative academic marginalisation of the Karachay-Balkars can also be seen in Soviet publications on the language contacts between the Karachay-Balkars and the Ossetians. The significant amount of Turkic loanwords in modern Ossetic bears witness to the fact that in the ancient past there were intensive contacts between the ancestors of modern-day Ossetians and the Turkic world. They commenced with the rise of the Judaeo-Turkic Khazar Empire in the 6th century CE and later, the contacts became even more intense during the rule of the Golden Horde. Not to mention, Oğuzic Azerbaijani used to be the lingua franca throughout the Caucasus since the founding of the Safavid Empire in the 16th century until the Russian conquest of the region in the 19th century. In fact, this Turkic influence on the Ossetic language is much more profound than that from the Caucasian languages that are spoken in the region, such as adjacent Ingush-Chechen, Kabardino-Cherkess (= Adyghe) or Georgian. The Turkic groups that have been geographically closest to these Ossetians are these Balkars and Karachays, who themselves have borrowed many Ossetic forms. The modern Ossetians usually call their Balkar neighbours Asy, who, historically speaking referred to the Ossetians themselves. Both names, Balkar and Asy, were already mentioned in the famous early Persian-written geographical work “The Regions of the World” (Ḥudūd al-‘Alam), as ās and balqar respectively. The transfer of the ethnonym ās to the Balkar points to intensive cultural contacts, such as interethnic marriages and strategic alliances between these two peoples. According to the eminent Turkologist Omelyan Pritsak they may have lived together in the Northern Caucasus until the Mongol invasions (Pritsak 1958, 341). Indeed, as asserted recently for instance, by Džurtubaev (2010, 4) in his introduction, the ethnogenesis of both the Ossetians and the Karachay-Balkars is an “interrelated process” (взаимосвязанных процесса). Mediaeval European sources further confirm the co-existence and co-mingling of these groups, when an Ossetic group (known as the Jász) and the Kıpçak speaking Cumans (Kun) settled in Hungary during the 13th century. It is only natural to wonder whether the Karachay-Balkars can be considered the last remnant of the mediaeval Kumans. After all, those Cumans were living very closely, if not in some sort of symbiotic relation, with the Alans, who are the conventionally accepted (immediate) ancestors of the modern Ossetians. In the past, the Mingrelians would apply alani to the Karachays. Even today, the term alan is still employed by the Karachay-Balkars as a self-identification. According to Thordarson (2009, 28 f.), “[w]e are thus justified in regarding the

22

Cheung

Karachay-Balkars as Turkicised Alans”, also on account of the numerous place names of Ossetic / Alanic origin attested in the Karachay-Balkar regions. This is difficult to prove (or disprove) though, as we have no idea how the social circumstances and interactions were between the Iranophone Ossetic- and Turkic-speaking Karachay-Balkars in the modern Karachay-Balkar regions.3

2

Abaev (1933) and Nikolaj Marr

The famous Ossetian linguist and prominent Ossetologist Vasilij I. Abaev seized upon this historic “cohabitation” to confirm Nikolaj Marr’s Japhetic theory, which was proclaimed in the early 1920s.4 According to this pseudo-scientific theory, also known as the New Study of Language, languages rather reflect a continuous merger of previous languages.5 Nikolaj Marr gave a Marxist twist to the European mediaeval idea that, analogous to the legendary origin of the Semitic peoples and their languages from Noah’s son Shem, most European nations and ethnicities descended from Noah’s other son, Japheth. According to Marr’ own interpretation, the languages spoken by Japheth’s children would be the substrate that was later overlaid by Indo-European languages. The different layers (of borrowing) would correspond to the different social classes of ancient societies (in Europe). Language was considered a superstructure on the base of society, concurrent to the creation of a (single) socialist economy. As language mixing was therefore the logical consequence, the notion that the languages of peoples could be traced back and therefore classified according 3  Other explanations are conceivable: a large group of mostly, Iranophone Alans may have vacated these areas, voluntarily or involuntarily (after the well-documented Mongol invasions). Subsequently, the Turkic Karachay-Balkar population occupied or dominated the region, while they would also impose their language on the remaining (Iranophone) sheepherders. Neighbouring groups would still have called them by their older names, etc. It is also conceivable that a previous, centuries-old situation of active bilingualism in these areas had shifted in favour of a predominant monolingual Turkic environment, when an external, Turkic language, i.e. Azeri, became the lingua franca of the Caucasus. This kind of linguistic symbiosis and co-existence between two linguistically unrelated groups is well known elsewhere in the world, e.g. between the Iranophone Balochis and Dravidian-speaking Brahuis in Pakistan. The dominance of one language over another depended on the political constellation and the linguistic preference often alternated with each generation. 4  Most of his articles on the Japhetic Theory can be found in Marr (1933). 5  This has led to a complaint by a Soviet scholar critical of Marr that “the study of the connections between related languages is turned over as a monopoly to bourgeois linguistics” (cited by Pollock 2006, 122).

The Contacts between the Ossetians and the Karachay-Balkars

23

to a common origin (as, notably, proposed by Indo-European linguists), was dismissed as a “false consciousness” (ложное сознание), which was introduced by bourgeois nationalism. New languages were rather the result of crossbreeding, while the ultimate origin of languages derives from the four primordial sounds sung by the ancient people during their chores, viz. ber, yon, roš, sal (Marr 1936, 130). In the article published in the journal, Language and Thought, Abaev (1933) discussed precisely the intermingling and mixing of the Karachay-Balkars and Ossetians, being reflected in the mutual borrowings between these two groups. Later on, he incorporated this study in his collected writings known as Ossetic Language and folklore (Abaev 1949).6 Apparently, shortly after the official endorsement by Marr, a youthful Abaev undertook a field trip to Baksan and its surroundings (located in the KabardinoBalkarian Republic). He then compiled a comprehensive list of putative Ossetic and Karachay-Balkar parallels as published in Abaev (1933), which reflected his faithful adherence to Marr’ Japhetic theory, the prevailing dogma of that period. It comes as no surprise that this publication shows severe, methodological shortcomings. Notably, the article does not give an ultimate origin of the forms, i.e. whether derived from Proto-Iranian, Old / Proto-Turkic or from “Caucasian”. Even if we ignore the ideological bias, Abaev’s paper also contains numerous factual errors, which he did not correct, when, subsequently, these forms were incorporated in his famous Historical-Etymological Dictionary of Ossetic. In this dictionary, he frequently assigned an older origin of these parallels, which was finally permitted after Stalin had denounced the Japhetic theory in an article.7 Unfortunately, Abaev often suggested etymologies that were a priori implausible. Also, it seems that some of the forms cited by Abaev were rather ephemeral, such as the counting system with Digoron sounding names. This system was apparently used in the southern Kabardino-Balkar region of Greater Khulam (near to the Russian – Georgian border). Actually, these numerals do not appear to be attested in other Karachay-Balkar-speaking areas (and therefore, not incorporated in the standard(ized) Karachay-Balkar language). The publication includes other terms too. A good example is Ossetic gæmæx ‘bare, with bare spots’, for which a Karachay-Balkar form gǝmǝx ‘a spot covered by scarce vegetation’ was cited by Abaev as parallel, without giving any source. 6  Lajpanov (1967) more or less repeated this work. 7  This was first published in the newspaper Pravda on June 20, 1950, on which see also the editorial in the linguistic journal Voprosy Jazykoznanija 1952, 3 f.

24

Cheung

So far, I have not found any corroboration for this, only Karachay qımıja ‘bare (footed)’ (?). Forms such as gǝmǝx are perhaps no more than ad hoc borrowings that we may very well encounter in the vocabulary of the few (bilingual?) Balkar speakers who happened to have been in intensive contact with local Digoron speakers (by marriage, trade or otherwise). Although Abaev introduced the region as a kind of melting pot of customs, traditions and languages of the local peoples, he did not explain the exact social or sociolinguistic circumstances (such as code-switching, active bilingualism, and other aspects of interlinguistic and multilingual communications) of this region. It remains, for instance, unclear how competent those informers were in either Digoron Ossetic or Balkar, and how the linguistic skills were acquired, through marriage, upbringing, trade or otherwise. There were arguably no religious objections against intermarriages between Sunni Digoron and Balkar speakers. Abaev asserted that the Ossetic elements in KarachayBalkar were not recent but the result of “the legacy of ancient Alanic-Turkic mingling, which took place in the areas of all the gorges, from the Terek to the Upper Kuban river”8 (Abaev 1949, 18). However, many of the claimed Ossetic loanwords in Karachay-Balkar are not attested elsewhere, which would rather suggest recent or ad hoc borrowing. This could be an indication of (recent) bilingualism. Although Abaev did distinguish elements that Karachay-Balkar had borrowed from Ossetic, from those that Ossetic has borrowed from Karachay-Balkar, the criteria for the distinction were rather erratic. He did invoke certain semantic and morphological criteria though to decide from which direction a term was borrowed, e.g. Ossetic bælas ‘tree’. Its generic sense would have been passed on to Karachay-Balkar balas ‘a wooden hay-dragger’, which is rather specialized (but, theoretically, both bælas and balas could have been independently borrowed from a third source). Elements that Abaev could not have perceived as borrowed from Ossetic to Karachay-Balkar, or vice-versa, were considered to have originated in the postulated Japhetic substrate. At first sight, this kind of categorization was in the spirit of Marr, but Marr’s theory muddles the distinction between original and borrowed forms. Therefore, Abaev relied implicitly on an etymology postulated by previous (non-Marrian) linguists. Without the traditional historical-comparative framework, any etymological attempts in his article were rather ad hoc.

8  “наследие старого алано-тюркского смешения, происходившего на территории всех ущелий, от Терека до верхней Кубани.”

The Contacts between the Ossetians and the Karachay-Balkars

25

Assigning an (ultimate) origin for the forms was, of course, of secondary importance to Abaev (1933), as these “parallels” were rather classified according semantic categories: A. B. C. D. E. F. G. H. I. J. K. L.

terms from the inanimate nature, terms from the animate nature, designations of cultivated plants, designations of domesticated animals, terms from the material culture, anatomical and medical terms, social and ethnic terms, designations of physical and mental properties, varia. counting system (as an indication of economic interactions) religion, mythology and folklore, toponyms.

For this conference, I would like to present a few of my own observations and a personal assessment of Abaev’s treatment of the Ossetic and Karachay-Balkar “parallels”. I will limit myself to the categories A. and B. Abaev’s work is a very instructive example of Soviet linguistics of the interbellum. As for the assessment of the Karachay-Balkar material,9 I have relied principally on the dictionary that Tenišev published in 1989, in order to find a confirmation whether Karachay-Balkar has genuinely assimilated these loanwords mentioned by Abaev in its vocabulary. Another valuable publication is that from Gustav Schmidt, who more explicitly considered the Ossetic borrowings into Karachay (Schmidt 1931). Again, regrettably, he did not always identify the ultimate origin of the Ossetic elements, whether they were inherited from Old Iranian or merely local, Caucasian Wanderwörter, remained unanswered. Recently, Ewa Siemieniec-Gołas has carried out a very valuable lexical study on the Turkic “Erbwortschatz” of Karachay-Balkar (Siemieniec-Gołas 2000). I have considered this work as well for the present talk. 9  The transcription of the Karachay-Balkar form is according to the modern Romanized Turkish alphabet. However, x is used here to denote the voiceless fricative velar, whereas ğ is the voiced correspondence. As it is the case in most Turkic languages, the KarachayBalkar velars k, g have both back and front realizations (the allophones [q]/[k] and [ɢ]/[g]/ [ɣ]), depending on the vocalic environment. Quite often, the complementary distribution of these realizations do not apply to (especially, the most recent) borrowings, e.g. from Arabic or Russian.

26

Cheung

A. In the category of inanimate terms, Abaev cites several “parallels”. Of the 16 forms, 4 forms are of Ossetic / Iranian origin, 3 forms of Karachay-Balkar / Turkic origin, 3 were wrongly analyzed (or simply unclear), perhaps 4 from a third source and 1 was perhaps ephemeral: i. The following forms that have a clear Iranian origin are: – Oss.10 cægat ‘northern side of the mountain’ (PIr. *čakāta-, Middle Persian cagād ‘peak, summit’, Sogdian ck”t ‘peak, forehead’) ~ KB çeget (Balk.) ‘north(ern) direction’ (borrowing from Balk. çeget would have yielded Oss. †cægæt). – Oss. awwon ‘darkness, cover’ (*āwa-wahāna- ‘covering into/down’, cf. Persian bahāneh ‘pretext, cover’) ~ KB awana ‘contour, silhouette, outline’. – KB dorbun (Kar.) ‘(bear) cave’, (Balk.) ‘cave, cavern’ (lit. ‘stone-bottom’, from PIr. *darwa-11 ‘solid, hard (as wood)’, cf. Khotanese dūra- ‘hard’, and *buna‘bottom’, cf. Pers. bon). The corresponding Oss. formation *dorbun is not attested though, only its elements dor ‘stone’ and bun ‘bottom, floor’ are clearly Ossetic in origin. – Oss. swadon / sawædonæ ‘well’ (< pl./f. *syāwā ‘black’ + *dānā ‘river, waters’) ~ KB (Balk.) şawdan ‘springs, well’. Etymologically speaking, Ossetic sawædonæ literally means ‘black water(s)’, which could be calque on an earlier Karachay-Balkar *kara sū for ‘well, or spring’? ii. On the other hand, Ossetic must have borrowed quite substantially from Karachay-Balkar as well. The difficulty is that quite often the KarachayBalkar forms are almost indistinguishable from their Turkic correspondences. The following forms may derive from Karachay-Balkar due to its typical phonological features:

10 

11 

Additional abbreviations include: KB = Karachay-Balkar, Balk. = Balkar, Kar. = Karachay; Oss. = Ossetic, Dig. = Digoron dialect of Ossetic; PIr. = Proto-Iranian, OT= Old-Turkic, PT = Proto-Turkic. Close Iron and Digoron counterparts cited as examples are separated by the sign “/”, with the Iron form on the left and the Digoron form on the right of the slash sign. The reconstruction *dárwa- would explain the vocalism in Ossetic (cf. Cheung 2002, 128 f.), and in Khotanese dūra-, cf. Emmerick (1989, 211). This thematized adjectival formation is a derivative of *dāru (gen. stem dru-) ‘wood’, Persian dār, etc. I have taken the cue for this connection from Maciuszak (2007, 205 f.). The additional cognate forms (Old Persian duruva- ‘secure, firm’, Avestan druua-, Sanskrit dhruvá- ‘healthy’ cited by her are, are unconnected though, as (also) shown by its morphological derivational process (in Sanskrit) and the semantic discrepancies.

The Contacts between the Ossetians and the Karachay-Balkars

27

– KB töppe ‘top, crown (of the head); peak; tuft’ (< PT *töppe, cf. OT töpi, Kumyk töbe, Turkish tepe) ~ Oss. (Dig.) c’opp ‘pluck, wool’, (Iron) c’upp summit, peak’ (c- < *ti). – KB kaya ‘rock, boulder’ (< PT *kaya, Turkish kaya etc.) ~ Oss. k’æj/ k’æjæ ‘slate’. Evidently, the Ossetic forms may also derive from another Turkic language. The Svanetic form k’a ‘slate’ however is rather a direct loanword from Iron Ossetic k’æj. – KB ırxı ‘stream, creek’ (< OT arık ‘irrigation canal’, cf. Chagatay arığ, Turkish ark, etc.) ~ Oss. (Dig.) ærxæ ‘gorge, dry riverbed’. – KB tılpıw ‘vapour; air’ ~ Oss. (Dig.) tulfæ ‘vapour, steam’, see further below. iii. The following forms may be wrongly analyzed or unrelated: – Oss. k’oyldym / k’uldun ‘(mountain) slope, hill’, unrelated to KB küllüm ‘sunkissed spot’ deriv. of kün ‘sun’). – Oss. (Iron) ran, (Digoron) rawæn ‘place’. The cited KB ran is only attested in the expression kaya-ran ‘rock ledge, a certain spot on the rock, rock terrace’, in fact it just reflects a compound with a-elision kaya aran (aran ‘valley, lowland’) > kaya ’ran. KB ran is therefore an accidental form, being unrelated to Oss. ran. – Oss. k’æxæn ‘slope; cliff’. Abaev (1958-1995, 1: 631) no longer incorporated the cited Karachay-Bakar parallel təxən (sic!) ‘flat area on the rock’. iv. The following forms may stem from a third source, perhaps independently: – Oss. (æ)zme(n)sæ (Dig.) ‘sand’ (Iron yzmis) ~ KB (Balk.) üzmez ‘id.’. The Balkar form does not conform to Turkic morphology, hence it might be a borrowing from Ossetic, although it has no further correspondences in Iranian or in the neighbouring Caucasian languages. An Iranian preform *uz-maišā‘mixture, being mixed up’ (*maiz- ‘to mix, mingle’) has often been suggested (cf. Abaev 1958-1995, 4: 282), but this reconstruction is fraught with problems, both semantically and morphologically. – Oss. xuræ ‘gravel’ (Iron xoyr) ~ ? KB (Balk.) xuru ‘stony place, cobblestones’ (no further documentation). – Oss. cuxcur ‘flowing water’ ~ KB çuçxur ‘water fall’ ← South Caucasian / Kartvelian ?, cf. *me-rčx-e ‘shallow (of water)’, *rečx-/rčx- ‘to purl, babble, murmur’ (Klimov 1998, 119, 157). – typpyr / tuppur ‘bloated, fat’; [Dig.] hill’ ~ duppur ‘hill’, with similar forms in Darginian dupur ‘mountain’, Persian topoli ‘fat’, derived from a Turkic formation with *töppe?

28

Cheung

v. A very recent, ephemeral borrowing is: – Oss. gæmæx ‘bare, with bare spots’ ~ KB (Balk.) gǝmǝx ‘a spot covered by scarce vegetation’, see above. B. The 32 terms from the animate natural field are largely neither from Iranian nor Turkic. The botanical terms are usually indigenous (Caucasian). Of the parallels, 7 are Ossetic forms borrowed into Karachay-Balkar, 5 from KarachayBalkar into Ossetic, whereas the remaining 14 may be most likely from a third source (independently). Finally, 6 borrowed forms may be just ephemeral (4) or misinterpreted (2). i. Ossetic forms borrowed into Karachay-Balkar are: – Oss. bærz / bærzæ ‘birch’ ~ KB mırzı ‘id.’ (< PIr. *barzā-, Skt. bhurjá- m. ‘Betula utilis’). – Oss. kærdæg ‘grass’ (< PIr. *karta-ka- ‘cut’) ~ KB kırdık ‘id., greens’ (form contaminated with kırdış?). – Oss. xans (Digoron) ‘long, thick grass; tall weeds’ (< PIr. *kāsa- ‘tall grass’, cf. Pers. kāh ‘straw’ (ºns from fans ‘wool’) ~ KB xans ‘grass’. – Oss. fadawon (Dig.) ‘soft, dry grass (for decking)’ (lit. ‘foot-covering’, with fad ‘foot’ < PIr. *pāda-) ~ KB (Balk.) fadawan ‘straw often used as padding in mountain shoes’. – Oss. mælʒyg / mulʒug ‘ant’ (< Iranian *marwi- + *-čī-ka-) ~ KB (Balk.) gumulcuk ‘id.’ (with gu˚ from gubu). The Balkar form is evidently a borrowing from Dig. mulʒug. This Digoron formation shows an additional u-umlaut in comparison to the Iron correspondence. – Oss. synʒ / sinʒæ ‘thorn; blackthorn; splinter ~ KB (Balk.) şinji ‘spine, (plant) needle’, see below. – Oss. tæk’uzgæ (Dig.) ‘rowan (berry)’ ~ taqüzük (tüqüzgü, Abaev 1958-1995, 3: 255) ‘id.’. The lack of vowel harmony and the velar q in front of ü of the Balkar form all point to borrowing from Dig. tæk’uzgæ, but the ultimate origin is unknown. ii. Several Karachay-Balkar forms from the animate realm have entered Ossetic. We may cite the following forms, which in turn may be borrowings from another language: – KB bittir ‘bat’ ~ Oss. (Dig.) bittir (Iron xælyn-byttyr) ‘id.’, see below. – KB gabu ‘dandruff’, (Karachay) gıbı, (Balkar) gubu ‘spider’ ~ Oss. gæby, gyby / gæbu ‘mite’. The Ossetic forms appear to be borrowings from Karachay Balkar gabu, etc., which again may be an adaptation of a Kartvelian

The Contacts between the Ossetians and the Karachay-Balkars

29

formation, notably from a Georgian dialect form, cf. Gurian ǯɣiba- ‘tick’ (Klimov 1998, 100). – KB gılıw ‘foal; rat’ ~ Oss. (Dig.) gælæw ‘rat’. According to Abaev, gælæw is an “infantile deformation” of k’ælæw ‘foal’, which would be comparable to Kabardian qolow ‘piglet’, Georgian qoqo, Megrelian ɣoɣo ‘calf of buffalo’. Rather than considering “infantile deformation”, gælæw may simply be a loanword from Karachay-Balkar, as gılıw has retained the two meanings ‘foal; rat’. Of course, Karachay-Balkar gılıw may well be Caucasian in origin. – KB mıga ‘quail’ ~ Oss. mæga ‘snipe’. The Balkar form probably directly stems from Kartvelic, notably Georgian mc̣qer- ‘quail’. Balkar would have simplified the consonant cluster of the Kartvelian original formation, and, then, have passed on the term to Ossetic: Oss. mæga clearly shows a semantic shift. – Oss. (Dig.) pursa (Iron pysyra) ‘nettle, Urtica urens’ ~ mursa ~ ‘id.’, see below. iii. The following forms feature borrowings from a third source. They consists mostly of terms from the local flora, which are often Caucasian: – Oss. æxsæli, æxsælæ, (Iron æxsæly) ‘juniper’ ~ KB (Balk.) şkeyli, şkildi ‘id.’ ← South Caucasian?, cf. Georgian ašk’ili ‘wild rose’, Mingrelian šker- ‘rhododendron’. – Oss. sk’eldu ‘cowberry’ ~ KB şkildi ‘juniper’ (‘можжевельник’, Tenišev 1989, 751), kızıl şkildi, (Kar.) kızıl işkildi ‘cowberry, Vaccinium vitis-idaea’, also dial. ışkıldı ?) ← a variant of the Caucasian ‘juniper’ forms (-di: unanalyzable suffix in both Ossetic and Karachay-Balkar). The Karachay-Balkar forms seem to be borrowings from an unattested Iron correspondence *(y)sk’ildy, cf. Dig. sk’eldu. – Oss. cym/cumæ ‘dogwood, Cornus’ ~ KB çum ‘id.’, cf. Lezgian čumal, Tabassaran čemel ‘id.’ (similar forms: Turkish çim ‘grass’). – Oss. ʒedyr, ʒeʒyr, ʒedyræg / ʒæduræ ‘blackberry’ ~ KB züdür ‘id.’ ← a Caucasian language ?, perhaps to be analyzed as *zǝ ‘red; blackberry’ (cf. Adyghe zǝ ‘red’ or Abkhaz Bzyp a-z ‘blackberry shrub, bush’, Chirikba 1996, 87), and *dur ‘fruit’ ? (cf. Lezgian dur ‘dried fruit’, Ossetic dyrǧ ‘fruit’, loanword). Alternatively, it may be a borrowing from Finno-Ugric, according to Tenišev (1989, 807), who apparently follows Abaev (1958-1995, 1: 396). – Oss. ʒæbidyr / ʒæbodur, ʒæbedur ‘mountain goat, Capra caucasica’ ~ cuǧutur ‘id.’ ← a preform *ǯəɣʷətur, undoubtedly of Caucasian origin, probably WestCaucasian, cf. Adyghe šəquɫtər ‘id.’ (Apažev – Kokov 2008, 576). – Oss. mæntæg / mæntæg, mont ‘burdock’ ~ KB mant ‘id.’ ← Wanderwort ?, cf. Svan mant ‘id.’, Greek mínthē ‘mint’.

30

Cheung

– Oss. næzy / næzi ‘pine, Pinus sylvestris’ ~ KB (Balk.) nazı, (Karachay) nızı ‘fir’ ← Kartvelic *naʒw ‘spruce, fir(-tree)’, cf. Georgian naʒv (but also as a regional Wanderwort in other Middle Eastern languages, cf. Persian nāz, nāžu, nājū ?). – Oss. murtgæ, murk’æ ‘Viburnum’ ~ KB (Karachay) murtxu, from Kartvelic, cf. Georgian marc̣qv- ‘strawberry’. – Oss. mæra / mura, pura ‘hollow’ ~ KB pura ‘hollow, rotten (tree)’ ← ?, cf. Chechen, Ingush mur ‘hollow tree’. – Oss. ninæǧ ‘raspberry, Rubus idaeus’ (Iron mænærǧ) ~ KB nanık ‘id.’ ← ? – Oss. tægær ‘maple’ ~ tıgır, (Balk.) tıkır ← Caucasian, cf. Svan tek’er, tek’ra ‘maple’. – Oss. turtu, (Iron) tyrty ‘barbarry, Berberis vulgaris’ ~ KB türtü ← Wanderwort ?, cf. Lezgian turt ‘id.’, similar forms such as Persian tūt. – Oss. ug ‘owl’ (Iron wyg) ~ KB uku. No doubt, these forms are onomatopoetic in origin, cf. Megrelian, Laz ɣu, Svan ɣu, etc. Ossetic and Karachay-Balkar have probably borrowed the forms independently from each other, perhaps from another Caucasian language (if we are not dealing with “spontaneous” expressive forms). – Oss. gælæbo, gæbælo, (Iron) gælæbu ‘butterfly’ ~ KB (Karachay) göbelek ‘id.’ Abaev also cites the Balkar forms gebelo, gelbo (← Digoron?), probably, ultimately, of Turkic origin (cf. Turkish kelebek). Almost all Turkic correspondences of kelebek have retained a final velar (with the exception of geographically distant Uyghur kepilɛ). In addition, the voiced velar g- needs an explanation. – Oss. mæga ‘snipe’ ~ KB (Balk.) mıga ‘quail’ ← a Kartvelic preform * mc̣q̣a-, cf. Georgian mc̣q̣er- ‘quail’. Both Ossetic and Karachay-Balkar would have simplified a difficult to pronounce consonant cluster mc̣q̣°. On final *–r in Proto-Kartvelic, cf. Klimov (1998, 317f.). iv. Ephemeral are probably: – Oss. kældæ ‘dry wood, deadwood’ (Iron kældyn) ~ KB kıldı (not found in Tenišev and other publications): kıldı is rather an ad hoc borrowing from Digoron ? The Digoron form seems to be a lexicalized past participle of the verb kælun (Iron kælyn) ‘to spill, fall down’, which is of PIr. origin.12

12 

Abaev also entertains the possibility of a connection with several European designations for ‘wood; log’, e.g. Greek kládos ‘branch’, Slavic *kòlda ‘block, log’ (Russian kolóda), Germanic (Old Icelandic) holt, German Holz ‘wood’. This may be co-incidental rather than an instance of “Scytho-European” borrowing.

The Contacts between the Ossetians and the Karachay-Balkars

31

– Oss. ʒumarǧ, zumarǧ, (shortened) zum (Iron zym) ‘Caucasian snowcock, Tetraogallus caucasicus’ (lit. ‘winter-bird’ < (thematized) Ir. *zyama- ‘winter’ + *mrga- ‘bird’) ~ KB cumarık ‘id.’ (not in other publications). KarachayBalkar cumarık as a borrowing from Ossetic is not attested elsewhere. – Oss. kændys / kændus a slightly toxic plant ~ KB kündeş ‘id.’ – Oss. qoppæg / qoppæǧ, qobæǧ ‘an edible lily’ ~ KB xömpek, xoppug ‘id.’ v. The following cited forms are fully unclear, also because of the unclear meaning (misinterpretation, misheard?): – Oss. bynʒ / binʒæ ‘fly’ ~ KB didin ‘wasp’ (not confirmed elsewhere, also not included in Abaev 1958, 280). – Oss. ʒægæræg ‘not fully bloomed flower’ ~ KB cıgıra, zıǧıra a kind of edible plant.

3

Some Observations

We can notice several highly interesting forms that the linguistic ancestors of Karachay-Balkar and Ossetic must have borrowed from the period prior to their arrival in the (northern) Caucasus, i.e. before the Mongol invasions in the 13th century. Abaev was the first scholar to label these ancient borrowings as “Scytho-European” isoglosses, which in practice, meant that the ancestors of the Ossetians would have borrowed, mainly, from Germanic and Slavic (also Celtic and Latin), on which see Abaev (1965). However, similar, ancient borrowings from Hungarian were not included in this label, simply because of the fact that Hungarian was not part of the Indo-European language family. A typical example of such a “Scytho-European” isogloss as defined by Abaev is the following “Ossetic ~ Karachay-Balkar parallel”: – Oss. synʒ /sinʒæ ‘thorn; blackthorn; splinter ~ KB (Balk.) şinji ‘spine, (plant) needle’. In this case, sinʒæ may reflect older *spina-13 + dimin. suff. *čī. The preform *spina- would be a loanword, most conceivably from East Slavic, cf. Russian spiná ‘spine’, Old Polish spina ‘id.’ (inherited forms or, loanwords ultimately from Latin ?). The “spine” form appears to be a widespread European cultural term, attested in Latin spina ‘thorn’, Baltic (Latvian) spina ‘rod’, Germanic (e.g. Old High German spinela ‘hairpin’), English spine, etc.

13 

Initial *sp- > *sf- > *s’s’ (palatalization) > modern Oss. s-, cf. sistæ, Iron syst ‘louse’ < PIr. *spiš + *čī (e.g. Avestan nom. sg. spiš, Persian šepeš ‘id.’).

32

Cheung

However, there is an implicit bias towards these ancient borrowings, as Abaev considered mostly (pre-) Ossetic as the first receiver of those so-called “ScythoEuropean isoglosses”, effectively disregarding the possibility that the ancestors of Karachay-Balkars could have contributed to these “Scytho-European isoglosses” as well. After all, we may consider Karachay-Balkars as a modern remnant of the powerful Cumans and Pechenegs, who used to occupy a good chunk of the Eurasian steppes. Thanks to their expansion, Cumans and the Pechenegs certainly came in contact with South Slavic speaking groups, and for a prolonged period. These Southern Slavs may have used just only recently a literary language (which is now known as Old Bulgarian, or alternatively, Old Church Slavonic). The second, politically significant ethnic group the Cumans and the Pechenegs would have met were the Hungarians, who just completed their conquest of Carpathia in the 9-10th century CE. The Hungarian arrival in the Balkans came in the aftermath of the attacks by these Cumans around 895. A written testimony to these contacts is the so-called Codex Cumanicus compiled in Hungary in the 12-13th century. This Codex served as a textbook of Cumanic. We can cite several borrowed forms for which Abaev claims Ossetic as the initial adopter, but actually, they most likely have entered an earlier stage of Karachay-Balkar first, before their adoption into Ossetic: – KB bittir ‘bat’ ~ Oss. (Dig.) bittir (Iron xælyn-byttyr) ‘id.’ ← South Slavic, especially Church Slavic nepŭtyrǐ ‘bat’ (which shows metathesis of t . . . p > p . . . t, cf. Russian netopyr’). The Ossetic and Karachay-Balkar forms appear to be an ancient borrowing from (South) Slavic. The question of course is which language has borrowed first. The apparent loss of ne˚ may give us a clue. Karachay-Balkar (and other Turkic languages) does not have native nouns with initial ne˚, only derivatives of the pronoun ne ‘what’ are attested, cf. Siemieniec-Gołas (2000, 158 f.). A Turkic speaker would have most likely re-analyzed such a foreign formation, South Slavic nepŭtyrǐ, as an expression with the interrogative pronoun ne. In contrast, there would be no apparent reason to resort to such a re-interpretation in Ossetic. There are several inherited formations with initial (Proto-Ossetic) *ne˚, e.g. *nez (= Digoron nez, Iron niz) ‘disease’, *new- ‘to cry’ (= Dig. new-, Iron niw-), *nenegative prefix. Initially, an early predecessor of Karachay-Balkar would thus have borrowed the South Slavic form, after which it was passed on to Ossetic.

The Contacts between the Ossetians and the Karachay-Balkars

33

– KB mursa ~ Oss. (Dig.) pursa (Iron pysyra) ‘nettle, Urtica urens’ ‘id.’. According to Abaev (1949) the Ossetic form has been borrowed into Karachay-Balkar, with the initial labial stop becoming the corresponding nasal m-. This however cannot be correct, as only older voiced b-14 may become m- in Karachay-Balkar, e.g. the indigenous name for the Balkar is Malkar. It is more likely that an earlier Karachay-Balkar form *bursa is the source of the Ossetic form, which looks not very ancient anyway (with atypical p- and final -a, rather than f- and -æ respectively). Therefore, those Cumans with their extensive contacts may have passed on this *bursa to Karachay-Balkar. If so, Karachay-Balkar mursa < *bursa may have been an old borrowing from Hungarian, viz. borsó ‘pea’ (Old Hungarian burso 1254, a place name), with final -ó < *-Vk(V) and de-affricatisation of *č > s).15 The Hungarian form itself reflects a Turkic loanword *burčak (Benkő 1992-1997, 1: 129), which is the term for a legume, pulse(-like) plant, notably pea, vetch (and also ‘hailstone’), cf. Turkish burçak ‘vetch’, Karachay-Balkar burçak ‘hail’, (Balkar) ‘pea’ (Siemieniec-Gołas 2000, 70 f.; Clauson 1972, 357; Sevortjan 1978, 275 f.). The semantic shift from ‘a legume’ to ‘nettle’ in Karachay-Balkar mursa needs an explanation16 though.

14 

15 

16 

Admittedly, the fate of the initial labial stops in Turkic is rather complicated. According to Pritsak (1958, 352), b- becoming m- is a typical Kıpçak development (“echt kiptschakisch!”), e.g. maka ‘frog’ (< PT *bāka, cf. Kumyk baka), (Karachay) miyik ‘big’ (but Balkar biyik, cf. Tatar, Nogay biyik < PT *bädük, cf. Turkish büyük). It is difficult to postulate a watertight phonetic rule though, especially since there are relatively few cases within KarachayBalkar (and borrowings from other Turkic languages may have distorted a possible phonetic distribution). Besides, Chuvash părça ‘pea’ shows loss of the final velar. Assuming that Karachay-Balkar mursa is a loanword from Chuvash is fraught with phonological and historical inconsistencies. The older suggestion that pre-historic Ossetic would have borrowed somehow from Chuvash is equally problematic. According to Gombocz (1912, 52), the Finno-Ugric forms, Mari pursa, pırsa (both modern Chuvash and Mari lack indigenous voiced stops) and Hungarian borsó would all have been borrowed from Old Chuvash *burčaɣ, but this reconstructed Old Chuvash *burčaɣ is simply too close to all the other Turkic forms to corroborate this statement, at least, with regard to Hungarian borsó. Perhaps, the preform *bursa is a blend formation of two similar Hungarian forms: borsó ‘pea, vetch’ and bors ‘pepper’ (bors ← Turkic burç ← ultimately Sanskrit marica, Clauson 1972, 771 f.; Sevortjan 1978, 274 f.).

34

Cheung

As another example of such a labial correspondence / adaptation, we may cite: – KB Abıstol, Amıstol ~ Oss. (Dig.) Amistol Summer month (June-July) ← ultimately Greek apóstolos. Abaev (1949, 283) insinuated that the source of the Karachay-Balkar form is Ossetic Digoron Amistol. The Digoron form is difficult to explain, notably -m- and the vocalism -i-, if it were a direct borrowing from Greek, or more likely via a Slavic intermediary apostolŭ, the expected Ossetic (Digoron) form should have been †ap(’)ostol (and Iron †ap(’)ustul). Rather, the Balkar form may be the source of the Digoron form, a voiced stop *b is normally not found natively17 in intervocalic position, which would therefore have been adapted as -m- in Ossetic. The back-vowel ı represents -i- in Digoron. The Balkar form on the other hand, shows a regular phonetic adaptation of the Slavic outcome apostol of the Greek form. Balkar indigenous vocabulary does not contain an intervocalic, voiceless labial stop, hence Slavic / Greek -p- → Balkar -b-, cf. Proto-Turkic *tāpan (or *tāban?) ‘heel’ > Balkar taban and P-Turkic *tōpık ‘knee’ > Balkar tobuk. In addition, Balkar also shows a regular alternation b ~ m, unlike Ossetic, which does not have an intervocalic -b- in its indigenous phonemic inventory, as all Old Iranian intervocalic *-p-, *-b- have become -v- (except after *u). Finally, the extraneous vowel sequence a . . . o (of apostolŭ) would naturally be adapted as a . . . . ı in Balkar. Finally, from the inanimate sphere (cat. A., see above), we can cite another likely instance of ancient Cumanic borrowing: – KB tılpıw ‘vapour; air’ ~ Oss. tulfæ ‘vapour, steam’. Abaev (1958-1995, 3: 316 f.) cautiously cited a rather far-fetched connection with Sanskrit turīpa‘semen (fluid)’. Both forms, Ossetic tulfæ and Karachay-Balkar tılpıw, have probably been borrowed, perhaps rather from (South) Slavic *toplŭ, Old Church Slavic toplъ ‘warm’ (Derksen 2007, 490), with regular metathesis of *pl > lp. Ossetic does not have a native labial stop p in its phonemic inventory, all forms with p point to either a foreign origin or is the result of a simplifying gemination of a consonant segment (e.g. nk > Iron pp). Considering the more faithful phonetic adaptation of Slavic – ŭ/-ъ in Karachay(-Balkar) as -ıw, Ossetic tulfæ seems to have been borrowed from Cumanic. The Karachay(-Balkar) form tılpıw appears to show umlaut, a feature that can already be noticed in the writing of the Codex Cumanicus, and in 17 

On the development of intervocalic *p > Ossetic v, (after *u) b, cf. Cheung (2002, 18f.).

The Contacts between the Ossetians and the Karachay-Balkars

35

modern Karachay-Balkar also in certain lexicalized phrases, e.g. bu-kün ‘today’ > bügün. However, the exact circumstances of this kind of umlaut are unclear. If the direction of the borrowing were the other way round, Ossetic initial -uwould have been consistently adapted as -u-/-ü- in Karachay-Balkar.

4

Summary and Conclusions

The one-sided concentration on research of the (putative) linguistic ancestors of the Ossetians in the past hundred years by Russian and Soviet scholars has led to the marginalisation and even downplaying of the Turkic linguistic component of this centuries old relation. Ossetic was intensively studied and many aspects of its history, speakers, literature and dialectology became better known, resulting in an appreciation and pride among its modern speakers. This, however, cannot be said of the speakers of Karachay-Balkar. Marked by academic neglect (and deportation of its speakers during the dark days of Stalinism), the Karachay-Balkar language was also considered to be somewhat of a linguistic crossbreed, as fostered by the Japhetic Theory developed by Nikolaj Marr. This has pretty much resulted in an approach in which many, non-Turkic borrowings found in Karachay-Balkar were considered to be taken directly from Ossetic or from a common Japhetic / Caucasian substrate language. A long exposé published by the Ossetian scholar Vassilij Abaev (Abaev 1933) illustrates this situation on the relation between the linguistic ancestors of modern Iranophone Ossetic and Turcophone Karachay-Balkar speakers. He interpreted the Ossetic – Karachay-Balkar parallels found in the local Balkar dialect as the outcome of ancient “Alanic-Turkic” mingling, on top of a Japhetic/Caucasian substrate. Nevertheless, the bias is not only due to the adoption of Marr’s Japhetic Theory. It also had a personal bias, as he ascribed the great majority of these cases to an earlier Ossetic provenance, giving little thought to the possibility that Karachay-Balkar could also have passed on many borrowings to Ossetic as well. We can summarize our conclusions drawn from the assessment of the Ossetic– Karachay-Balkar parallels discussed by Abaev (1933), as follows (based on two semantic categories): – there is no clear tendency in the direction of the borrowings: both Ossetic and Karachay-Balkar have contributed in almost equal measure to each other’s vocabulary.

36

Cheung

– as can be expected, a large group of these “parallels” consists of borrowings from local (Caucasian) languages, and it is often unclear whether they entered Ossetic or Karachay-Balkar first. – in addition, the linguistic ancestors, the Cumans, of the modern KarachayBalkars may have also borrowed from European languages, after which they would have entered Ossetic: (Greek apostólos →) South-Slavic *apostolŭ ‘apostle’ → Cumanic *abïstol (> Karachay-Balkar Abıstol) → Ossetic Amistol ‘Apostle(’s Month)’ (→ dial. Balkar Amystol!). The main criteria that have allowed us to distinguish the direction of borrowing between Ossetic and Karachay-Balkar are: – phonological criteria: e.g. the presence of vowel harmony in Karachay-Balkar forms and its (general) absence in Ossetic, the phonological restrictions and adaptions typical for Karachay-Balkar and Ossetic respectively. In KarachayBalkar, we may notice, for instance, the lack of forms with initial ne-. Ossetic, on the other hand, does not possess (indigenous) p and intervocalic -b-, while it shows the frequent substitution of initial stops (especially from KarachayBalkar and other non-Caucasian languages) with their corresponding ejective consonants. Notable examples are: (Slavic) nepŭtyrǐ → Karachay-Balkar ne bittir → Ossetic byttyr / bittir; Karachay-Balkar töppe, kaya → Ossetic c’upp / c’opp (c < *tj), k’æj / k’æjæ; Ossetic tæk’uzgæ → Karachay-Balkar taqüzük. – semantic shifts: the language that has preserved the meaning of the borrowed form from a donor language most closely, may also have adopted the form first. Examples include: Karachay-Balkar gılıw ‘foal; rat’ → Ossetic gælæw ‘rat’; (Kartvelic) *mc̣q̣a- ‘quail’ → Karachay-Balkar mıga ‘quail’ → Ossetic mæga ‘snipe’. – historical-comparative evidence: forms directly inherited from their linguistic affiliated group, i.e. (Indo-)Iranian or Turkic respectively, as shown by historical-comparative methods, may decisively point to the direction of the borrowing: Karachay-Balkar töppe (< Proto-Turkic *töppe, cf. Turkish tepe, Kumyk töbe, etc.) → Ossetic c’upp / c’opp; Ossetic kærdæg ‘grass’ (< PIr. *karta-ka- ‘cut’) → Karachay-Balkar kırdık ‘id.’. A further (re-)assessment of the Ossetic and Karachay-Balkar material may shed more light on the historical contacts between the Ossetians and Karachay-Balkars, which in turn may assist in the formation of their respective self-image and identity. Precisely, the lack of great, especially, local researchers and scholars has created a cultural and historic void in the national narrative of the Karachay-

The Contacts between the Ossetians and the Karachay-Balkars

37

Balkars. For this reason, many Karachay-Balkars have resorted to “borrow” aspects of their culture and historiography from their Turkic brethren, notably from Turkey. This can have dire consequences in the future. Being ignorant of their unique (language) history, the Karachay-Balkars could well devalue their language to such a degree, that they may decide no longer to pass it on to the next generation, switching to Russian or even (Istanbul) Turkish instead.

References Abaev, Vasilij I. 1933. “Poezdka k verxov’jam Kubani, Baksana i Čereka [Trip to the upper reaches of the Kuban, Baksan and Cherek rivers].” Jazyk i Myšlenie I: 71-89 [= Abaev 1949, 271-290]. Abaev, Vasilij I. 1949. Osetinskij Jazyk i Folk’lor, vypusk I [Ossetian Language and Folklore, part I]. Moskva – Leningrad: Izdatelʹstvo Akademii nauk SSSR. Abaev, Vasilij I. 1958-1995. Istoriko-ètimologičeskij slovar’ osetinskogo jazyka [HistoricalEtymological Dictionary of the Ossetic Language]. 5 vols. Moskva – Leningrad: Institut jazykoznanija RAN. Abaev, Vasilij I. 1965. Skifo-evropejskie izoglossy na styke Vostoka i Zapada [ScythoEuropean isoglosses on the crossroad between East and West]. Moskva: Nauka. Anfert’eva, Antonina N. and Nikolaj N. Kazanskij. 2013. “Materialy k istorii Instituta Lingvističestix issledovanij RAN 1921-1934 gg. (ot Instituta jafetidologičeskix Izyskanij do Instituta jazyka i myšlenija im. N. Ja. Marra) [= Materials on the history of the Institute of Linguistic Research RAN from the years 1921-1934 (from the Institute of Japhethological Investigation to the Institute of Language and Thought, named after N. Ja. Marr)].” Acta Linguistica Petropolitana, Trudy Instituta Lingvističestix issledovanij RAN 9, no. 1: 1-437. Apažev, Muxamed and Džamaldin N. Kokov. 2008. Adygè-urys psal’al’è / Kabardinočerkessko-russkij slovar’ [Kabardino-Cherkess Russian Dictionary]. Nalčik: Gosudarstvennoe izdatelʹstvo inostrannyx i nacionalʹnyx slovarej. Benkő, Lórand. 1992-1997. Etymologisches Wörterbuch des Ungarischen. 3 vols. Budapest: Akadémiai Kiadó. Bennigsen, Alexandre. 1983. The Islamic Threat to the Soviet State. London – Canberra: Croom Helm. Cheung, Johnny. 2002. Studies in the Historical Development of the Ossetic Vocalism. Wiesbaden: Dr. Ludwig Reichert Verlag. Chirikba, Viacheslav. 1996. Common West Caucasian: the reconstruction of its phonological system and parts of its lexicon and morphology. Leiden: Research School CNWS, School of Asian, African, and Amerindian studies.

38

Cheung

Clauson, Gerard. 1972. An Etymological Dictionary of Pre-Thirteenth-Century Turkish. Oxford: Clarendon Press. Derksen, Rick. 2007. Etymological Dictionary of the Slavic Inherited Lexicon. Leiden: Brill Publishers. Gombocz, Zoltán. 1912. Die bulgarisch-türkischen Lehnwörter in der ungarischen Sprache. Helsinki: Société finno-ougrienne. Karaulov, Nikolaj A. 1908. “Balkary na Kavkaze [the Balkars in the Caucasus].” Sbornikʹʹ materialovʹʹ dlja opisanija mestnostej i plemenʹʹ Kavkaza, 38: 131-180. Lajpanov, Kazi T. 1967. “O tjurkskim èlemente v ètnogeneze osetin [On Turkic elements in the ethnogenesis of the Ossetians].” In Proisxoždenie osetinskogo naroda [The Origin of the Ossetian people], edited by Xazbi S. Čerdžiev: 209-214. Ordžonokidze: Severoosetinskoe knižnoe izdat. Maciuszak, Kinga. 2007. Review of Studies in the Historical Development of the Ossetic Vocalism, by Johnny Cheung, Studia Etymologica Cracoviensia, 12: 203-206. Marr, Nikolaj. 1933. Ètapy razvitija jafetičeskoj teorii [The stages of the Japhetic Theory], edited by V. B. Aptekarev. Leningrad: Gosudarstvennaja akademija istorii material’noj kultury. Vol. 1 of Izbrannye raboty [Collected Works]. Marr, Nikolaj. 1936. Voprosy jazykoznanija [Fundamental Questions of Linguistics], edited by A. G. Ioannisjan. Leningrad: Gosudarstvennoe socialno-èkonomičeskoe izdaltel’stvo. Vol. 2 of Izbrannye raboty [Collected Works]. Pollock, Ethan. 2006. Stalin and the Soviet Science Wars. New Jersey: Princeton University Press. Pritsak, Omeljan. 1958. “Das Karatschaische und Balkarische.” In Philologiae Turcicae Fundamenta I, edited by Deny, Jean, Kaare Grønbech, Helmuth Scheel, and Zeki Velidi Togan: 340-368. Wiesbaden: Franz Steiner Verlag. Schmidt, Gustav. 1931. “Über die ossetischen Lehnwörter im Karatschajischen.” In Mélanges de Philologie offerts à M. J. J. Mikkola, professeur de philologie slave à l’Université de Helsinki: 364-395. Helsinki: Suomalainen Tiedeakatemia. Sevortjan, Ervand V. 1978. Ėtimologičeskij slovarʹ tjurkskich jazykov na bukvu “b” [Etymological Dictionary of Turkic Languages on the letter b]. Moskva: Nauka. Siemieniec-Gołas, Ewa. 2000. Karachay-Balkar Vocabulary of Proto-Turkic origin. Kraków: Księgarnia Akademicka. Thordarson, Fridrik. 2009. Ossetic Grammatical Studies. Edited by Sonja Fritz. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Voprosy Jazykoznanija. 1952. “Dva goda dviženija sovetskogo jazykoznanija po novomu puty [Two years of the movement of Soviet Linguistics on a new path].” May-June, 3: 3-18.

Chapter 3

Why Caucasian Languages?1 Bernard Comrie

Introduction The articles in this volume are for the most part concerned with sociolinguistic aspects of languages of the Caucasus, but in the present article I want to draw attention to the fact that these languages also have unusual features of grammatical interest. In addition to the importance of documenting and preserving languages of the Caucasus as part of their communities’ cultural heritage, the languages are also of scientific importance because of their structural properties. I have selected six features for brief discussion, with reference to more extensive treatment in the literature. The examples represent all three of the indigenous language families of the Caucasus, with one each from Kartvelian and West Caucasian (Abkhaz-Adyghe), and four from East Caucasian (NakhDaghestanian), more specifically the Tsezic branch of East Caucasian; the greater concentration on the last mentioned simply reflects my own greater familiarity with these languages.

1

Georgian Verb Indexing (“Agreement”)

In Georgian, the finite verb indexes (or in more traditional terminology: agrees with) its subject and object in person and number. This means that a transitive verb indexes both its subject and its object. The relevant forms can be found in any standard grammar of Georgian, and are to a large extent straightforward. The forms in (1) show a selection of relevant forms in the present tense. 1  A version of this article was presented at the 1st International CUA Conference on Endangered Languages, Ardahan, Turkey, 13-16 October 2014. I am grateful to the conference organizers for making this event possible and to all those who participated in the discussion of my presentation. The following abbreviations are used: abs absolutive, ad ad(essive), adj adjective, all allative, antip antipassive, attr attributive, caus causative, erg ergative, in in(essive), ipfvcvb imperfective converb, loc locative, obl oblique, prf perfect, prs present, pstunw past unwitnessed, rel relative, resptcp resultative participle, sg singular.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_004

40

Comrie

The stem of the verb is invariable, xedav, to which can be attached a prefix and/or a suffix, depending on the person of the arguments to be indexed. (1)

v-

mgmg-

xedav xedav xedav xedav xedav xedav xedav

-s -s -s

‘I see him/her/it’ ‘you see him/her/it’ ‘s/he sees him/her/it’ ‘he/she/it sees me’ ‘he/she/it sees you’ ‘you see me’ ‘I see you’

Some of the cases are morphologically transparent, and likewise transparent in terms of their processing, in particular forms like m-xedav-s and g-xedav-s, where the prefix consistently indexes the object (first or second person) and the suffix consistently indexes the subject (third person). However, the system is somewhat more complex than this, in particular because certain persons are indexed by zero, which means that the hearer2 cannot simply link a given affix to a given combination of grammatical relation (subject or object) and person, but rather has to infer the interpretation from patterns of absence of prefix or suffix. In the simple cases, we can identify subject and object prefixes as in (2). (2) Subject

Object

1sg 2sg 3sg 1sg 2sg 3sg

v∅-s mg∅-

Each of the overt subject prefix, overt subject suffix, and the two overt object prefixes receives a consistent interpretation. Beyond that, inferencing is required, as developed in more detail in Comrie (2013: 24-26). First, absence of an object prefix on a transitive verb always indicates a third person object. Thus, having identified that none of the object prefixes is present, the hearer can make this inference. The case of absence of subject prefixes is more complex, and here the hearer may have to go through several stages. There is no overt second person affix, whether prefix or suffix, so as a 2  Although the inferences are presented as if conscious processes undertaken by the hearer, such processing is of course below the level of consciousness, and combinations may well be routinized by the time a child acquiring Georgian achieves fluency in the system.

Why Caucasian Languages ?

41

general rule the absence of both the first person subject prefix and the third person subject suffix can be taken as an indication of a second person subject. There is, however, one exception to this, which follows from the fact that Georgian does not permit combinations of person prefixes. The form g-xedav has a second person object prefix, so clearly the object is ‘you’. What is the subject? It cannot be third person, since this would require the third person subject suffix -s, as indeed shows up in the form g-xedav-s in (1). Normally, absence of a subject affix indicates a second person subject, so can g-xedav mean ‘you see yourself’, with coreferential subject and object second person arguments? The answer is negative, because another rule of Georgian intervenes, namely one that says that objects coreferential with the subject are not expressed by means of an object affix on the verb, but rather by means of a separate word tavi, literally ‘head’, so that a combination of second person subject and second person object is expressed literally as ‘you hit your head’; since ‘head’ counts as third person, it has no overt person affix in the verb morphology. For g-xedav this leaves only one remaining interpretation, namely ‘I see you’, and this turns out to be correct, since in Georgian when a subject and an object prefix compete for the prefixal person position in the verb morphology, the object wins out and the subject remains unexpressed. In g-xedav, there is no piece of the verb’s structure that we can identify as expressing the first person subject, and a zero subject marker does not in itself indicate a first person subject; rather, it is the interaction of a number of principles of Georgian verb morphology that conspire to determine the unique correct interpretation of g-xedav.

2

Kabardian Pre- and Postnominal Relative Clauses

The discussion of relative clauses in Kabardian in this section is based on Applebaum (2013: 108-118); cf. also the published summary in Applebaum & Berez (2009: 30-33). Kabardian has a fully productive prenominal relative clause construction, as illustrated in (3)-(4). The head noun, ‘man’ in (3) and ‘girl’ in (4), is outside the relative clause and to its right, while the relative clause precedes the head noun, i.e. exactly the opposite order from what one finds in English and most other European languages (although it does parallel the dominant construction in Turkish). (3) [qálɐ-m ∅-kʷ’-á] ɬ’ə́ -r city-obl 3sg-go-prf man-abs ‘The man who went to the city.’

42 (4) [maxʷɐqɐ́ s mektéb-əm jə-ʃə-s-ɬɐɣ-á] every_day school-obl 3sg-there-1sg-see-prf ‘The girl whom I saw at school every day.’

Comrie

həgɐ́ bz-ər girl-abs

In examples (3) and (4), the relative clause consists of a sequence of several words (in fact, two in (3) and three in (4)), and the head is a separate word. This can be seen most obviously from the presence of one stress per word, marked in this transcription by means of an acute accent on the vowel in question. In addition, Kabardian also has a postnominal relative clause, with basically the same internal structure as the prenominal relative clause, except that there are much heavier restrictions and a very different prosodic structure. An example is provided in (5). (5) a žɐlɐ [də-z-ɣɐ-t’əs-á]-m the boy 1pl-rel-caus-sit-prf-obl ‘The boy who made us sit down.’ Perhaps the most important characteristic of (5) is that it is phonologically a single word, despite the transcription with three words following the morphosyntax rather than the phonology. This can be seen most clearly in the fact that (5) has a single accent. But there are also additional prosodic cues that this is a single word. As a separate phonological word, ‘boy’ would have the form ʃálɐ, with an accent on the first vowel, and strengthening of this vowel from [ɐ] to [a]. In (5), the word for ‘boy’ does not show this strengthening, since it is not a separate word and does not have an accent on its first vowel. The initial consonant of the verb form also shows that there is no phonological word boundary between it and the preceding ‘boy’. As a separate phonological word its initial stop would be devoiced, i.e. tə-z-ɣɐ-t’əs-á-m; the absence of devoicing shows that the verb is not at the beginning of a phonological word.3 Given that the postnominal relative clause must form a single phonological word with its head, one might well ask whether this imposes further restrictions, of length or complexity, on the relative clause, given that the whole must be pronounced as a single phonological word. The answer is affirmative: The postnominal relative clause may not be longer than a single grammatical word, a restriction that considerably facilitates the prosodic unification of the construction, although it does mean that ideas requiring expression by means of 3  Though not directly relevant to the structure of relative clauses, the voicing of the initial fricative of ‘boy’ in (5) shows that there is no phonological word boundary between it and the preceding article.

Why Caucasian Languages ?

43

more than a single grammatical word in the relative clause must be expressed by means of prenominal relatives. Kabardian illustrates first a clear example of language that has both prenominal and postnominal relative clauses but where it is the postnominal relative clause that is more restricted in its expressive possibilities. Moreover, it provides excellent illustration of the need to include phonological information in carrying out syntactic analysis.

3

Tsezic Pharyngealization

This section examines two languages from the Tsezic branch of East Caucasian, Tsez (based on the fuller discussion in Maddieson, Rajabov, & Sonnenschein 1996) and Bezhta, with the link between the relevant phenomenon in the two languages following the discussion in Comrie (2003). One of the characteristic phonetic features of Tsez is pharyngealization. In the indigenous vocabulary, phonemic pharyngealization occurs in two sets of circumstances. First, uvulars may occur pharyngealized in any position in the word, including word-finally, cf. the contrast between raq ‘side’ and raqˁ ‘wound’. In this position, phonetic pharyngealization characterizes basically just the consonant, including its release, with only minimal effect on the preceding vowel. Second, and more interestingly for present purposes, phonemic pharyngealization may characterize the initial (C)V of a word, cf. the contrast between -oƛo4 ‘amongst’ and ˁoƛno ‘seven’, and between mo ‘(eye) tear’ and mˁow ‘kind of mushroom’. Here, phonetic pharyngealization characterizes the whole of the vowel, and the last part of the initial consonant if there is one; it does not extend to the rest of the word, i.e. beyond the initial (C)V. Corresponding to a pharyngealized vowel (or CV sequence) in Tsez, in Bezhta one finds what has traditionally been described as an “umlauted” vowel, as seen when comparing Tsez ˁaƛ ‘village’ with Bezhta äƛ. The precise distinctive phonetic value of umlauted vowels in Bezhta remains to be investigated in detail, although impressionistically they are less pharyngealized and

4  The leading hyphen indicates that the postposition is normally preceded by a gender-number prefix. While pharyngealization is frequent in Tsez lexical representations and clearly phonemic, it is not easy to find minimal or even near-minimal pairs.

44

Comrie

more fronted than their Tsez cognates.5 But there is a further, perhaps more interesting difference. In Tsez, as noted, pharyngealization on the initial syllable is restricted phonetically to that syllable. By contrast, in Bezhta umlauting of the vowel of an initial syllable extends to following syllables, though tending to fall off in intensity as one gets towards the end of the word, especially with longer words. Thus, the past tense of the verb ‘cough’ in Bezhta is öhƛö-yö, where not only the second vowel of the stem but also the vowel of the suffix is umlauted, giving rise to vowel alternations in suffixes, cf. xuƛo-yo, past tense of ‘drink’. In other words, Bezhta has developed vowel harmony: In general, the vowels of a word are either all umlauted or all non-umlauted. The comparison of Tsez and Bezhta thus shows us one possible scenario for the origin of vowel harmony, through the extension of what was originally an opposition phonetically characterizing only the first syllable to one that characterizes phonetically the whole of the word.

4

Personification in Tsez

Tsez, like most East Caucasian languages, has a gender system.6 In Tsez, there are four genders, distinguished in the singular for instance by the prefix which they require on vowel-initial verbs and adjectives that agree with that noun phrase. The four classes are set out in (6), with the prefix used for agreeing with a noun of that gender, and a brief semantic characterization of the nouns that constitute each class. (6) Tsez gender (noun class) prefixes in the singular I ∅- all and only male humans II y- all female humans; some inanimates III b- all animals; many inanimates IV r- many and only inanimates 5  While the Bezhta umlauted vowels ä, ö, and ü are often traditionally characterized as front vowels, acoustic investigation of their formant structure by Sven Grawunder (Max Planck Institute for Evolutionary Anthropology) suggests that while there is some fronting, at least for ö and ü, it does not advance even as far as the central position. The Tsez and Bezhta counterparts sound different, but the precise phonetic characterization of the difference remains open. 6  Traditionally, this often referred to as a noun class system in speaking of East Caucasian languages. For present purposes at least, this is purely terminological. In general, the gender of a noun is not reflected in its own form, much as in German one simply has to learn the genders of Löffel ‘spoon’ (masculine), Gabel ‘fork’ (feminine), and Messer ‘knife’ (neuter).

Why Caucasian Languages ?

45

In somewhat more detail: Gender I contains all and only nouns denoting male humans (plus supernatural beings assimilated to male humans). Gender II contains all nouns denoting female humans plus a restricted number of inanimate nouns. Gender III includes all nouns denoting animals, plus a large number of inanimate nouns. Gender IV contains a large number of nouns, all of them inanimate. While in one sense the main problem in learning the gender of nouns in Tsez is the fact that inanimates are assigned across three genders, for present purposes we are only concerned with the semantically determined features of the system: Male humans are in gender I, female humans in gender II, animals in gender III. What, then, happens in a traditional tale where the participants are animals, but playing the role of humans? One traditional Tsez story that I discuss in more detail in Comrie (2005) involves as main participants a rooster and a hen, who play the role of a husband and wife, with traditional family roles, as can be seen in (7), where the rooster goes out to work while the hen stays home to look after the house.7 There is thus a potential conflict between the fact that the protagonists are animals (which would require gender III), but behave as humans (which would require genders I and II). (7) mamalay ɣuddes b-ik’i-x zew-no qaci-x rooster every_day III-go-ipfvcvb be-pstunw firewood-ad ciq-x-ār onoču b-eynoy-xo zew-no idu-z-ā forest-ad-all hen III-work-ipfvcvb be-pstunw home-loc-in ‘Every day the rooster went to the forest for firewood, and the hen did the housework at home.’ A similar problem arises in English with the choice of pronouns in the third person singular, where English distinguishes, roughly speaking, male human he, female human she, non-human (including animals and inanimates) it. If I were to tell the story of the rooster and the hen in English, then I would almost certainly use the pronouns he for the rooster, she for the hen, since what is deemed relevant is the human social role that each illustrates. In Tsez, by contrast, gender III, appropriate for animals, is used, so that in the first line of (b) the verb b-ik’i-x has the gender III prefix agreeing with ‘rooster’, and in the second line the verb b-eynoy-xo again has the gender III prefix, this time agreeing with ‘hen’. 7  The full Tsez text of the story, with a Russian translation, is available in Abdulaev & Abdullaev (2010: 44-47); the version there is slightly different from the one used here, but the differences are not significant.

46

Comrie

This points to an interesting difference in the interaction between grammar and discourse in English and Tsez gender in cases of personification. In English, it is the result of personification—male or female human—that predominates; in Tsez, it is the input (animal).

5

Bezhta Antipassive

In English, voice alternations are in general independent of the expression of tense-aspect-mood (TAM), so that corresponding to the active simple past in (8) one can change the aspect to progressive without changing the voice, as in (9), and likewise one can change the voice without changing the aspect, as in (10). (8) The detective followed the suspect. (9) The detective was following the suspect. (10) The suspect was followed by the detective.

progressive passive

In Bezhta, as generally in Nakh-Daghestanian languages, the alignment of a simple transitive clause is ergative-absolutive, i.e. the subject/agent (A) in (11) stands in the ergative case, while the direct object/patient (P) stands in the absolutive case. (If the verb indexes an argument, this will be the P; in (11), the word ‘bread’ belongs to gender III.) (11) öždi bäbä boy.erg bread(III).abs ‘The boy eats the bread.’

m-üq-čä III-eat-prs

There is an alternative construction as in (12), a so-called antipassive (Comrie, Khalilov & Khalilova 2015: 551-553), in which the P shows up in an oblique case (in Bezhta, usually the instrumental), while the A shows up in the absolutive case; there is a suffix indicating that the verb is in the antipassive voice.8 Just like the English passive, the Bezhta antipassive changes the grammatical relations in the clause; note, for instance, that the verb in (12) now indexes the A (gender I), not the P.

8  The formation of antipassives in Bezhta is, however, rather idiosyncratic, lexicalized, as will be seen in the following examples.

Why Caucasian Languages ?

47

(12) öžö bäbäla-d ∅-üⁿq-dä-š boy(I).abs bread-ins I-eat-antip-prs ‘The boy is busy eating the bread.’ However, unlike the English passive, the Bezhta antipassive not only changes the grammatical relations, it also changes the TAM of the clause, adding the semantic component of durative, i.e. extended in time, which depending on the lexical meaning of the verb might involve simply extending the particular action in time, as in (12), or repeating it so that the resultant complex action takes up more time than a single instance of the action in question, as in (14) and (16) below. The question therefore arises whether the Bezhta antipassive is really to be characterized as a voice, even though in (12) the grammatical relations are different from in (11), or whether it should rather be characterized as an aspectual, durative form, since this is its semantic import. The situation becomes more complex if we include intransitive verbs. First, Bezhta has a small class of intransitive verbs with onomatopoeic semantics that take their single argument in the ergative case, as in (13). Diachronically, these may originally have been transitive constructions (of the general type ‘the boy uttered X’), but synchronically they are one-place predicates. (13) öždi öhƛö-yö boy.erg cough-pst ‘The boy coughed (once).’ In one respect, the antipassive of such an onomatopoetic verb is similar to that of a transitive verb: The argument in the ergative case in (13) shows up in the absolutive case in (14), just like the A of a transitive clause. (14) öžö öhdää-yö boy.abs cough.antip-pst ‘The boy coughed (several times).’ With the general run of intransitive verbs, the antipassive, as in (16), has exactly the same case frame as the basic construction, as in (15); although the form of the verb shifts to indicate the antipassive. (15) öžö ∅-ogic’-iyo boy(I).abs I-jump-pst ‘The boy jumped (once).’

48

Comrie

(16) öžö ∅-ogiyac-ca boy(I).abs cough.antip-prs ‘The boy jumps (several times).’ But is (16) then really an antipassive? There is no change in voice relative to (15), the only change being in the aspectual value. Is the so-called antipassive in Bezhta then purely aspectual, i.e. more accurately described as durative aspect rather than as antipassive voice? The aspectual shift in meaning is constant, but nonetheless there is an undeniable change in grammatical relations with transitive verbs and, though perhaps less spectacular, with onomatopoetic verbs. Bezhta thus illustrates an interesting interaction between voice and TAM. The relevant morphological form is consistently durative. However, if the verb is transitive an obligatory change in grammatical relations is required, paralleling a prototypical antipassive construction. A corresponding change is also required in onomatopoetic verbs. Nothing in the nature of human language predicts this syntactic change. The Bezhta construction does pose a terminological problem (antipassive? durative? both?), although the generalization, though unusual, is easy to state: The resultant form is consistently durative, a change in voice to antipassive is required for transitive and onomatopoetic verbs.

6

Bezhta Adjectives

As the final set of data on Tsezic languages, we consider the word classes represented in the native and borrowed vocabulary of Bezhta, following Comrie & Khalilov (2009). Of particular interest here are adjectives. Bezhta has only a small number of indigenous underived adjectives, such as -uk’o9 ‘big’, k’et’o ‘good’. In addition, within the indigenous vocabulary there are adjectives derived from other word classes, such as participles derived from verbs, e.g. -uɣo-yo ‘dead’, morphologically -die-resptcp, i.e. something like ‘having died’. A large proportion of Bezhta’s adjectives, however, are loans from Avar, the traditional lingua franca of the area, such as bercinab ‘beautiful’. Within the Loanword Typology project, of which Comrie & Khalilov (2009) forms part, for each language a fixed list of lexical meanings was used, translated into the language in question, and then identified as an indigenous formation or a loan, in the latter case with further identification of the immediate 9  A leading hyphen indicates the position where an agreement (indexing) prefix precedes the stem.

49

Why Caucasian Languages ?

source. The data in (17) show, for each word class in Bezhta, the percentage of words on the fixed list that are indigenous versus borrowed from Avar, from Russian, from Georgian, or from some other source. (17) Loanwords in Bezhta by part of speech and source language

Noun Verb Adjective Function word

Indigenous 55.6 94.0 70.4 88.1

Avar 23.1 6.0 29.6 9.9

Russian 12.1 0.0 0.0 1.0

Georgian 8.3 0.0 0.0 1.0

Other 0.8 0.0 0.0 0.0

In the overwhelming majority of languages, from any given source of a substantial number of loans the number of borrowed nouns exceeds that of borrowed adjectives, as can be seen for loans from French into English in (18).10 (18) Loanwords in English by part of speech and source language (Grant 2009)

Noun Verb Adjective Function word

Indigenous 52.0 65.9 66.2 91.1

French 29.5 22.3 19.2 2.2

Latin 9.7 5.9 6.2 2.2

Old Norse 2.7 4.8 6.2 2.2

Other 6.0 1.1 2.3 2.2

Bezhta is unusual in that among its substantial number of loans from Avar, the main source of its borrowings in the LWT wordlist, the percentage of adjectives that is borrowed is greater than the corresponding percentage for any other word class. These adjectives also, incidentally, include many frequent words, such as bercinab.

7

Conclusion

In this article I have drawn attention to some of the properties of indigenous languages of the Caucasus that continue to attract linguists to them. 10 

Likewise from Latin into English, although the numbers are smaller and therefore less probative—a single item can shift the percentages of word classes significantly. From Old Norse into English there are actually more adjectives than nouns or verbs, although the numbers are again small.

50

Comrie

Documenting and preserving these languages is not only an important social and societal task, but also a rewarding scientific endeavor.

References Abdulaev, Arsen K. & Isa K. Abdullaev. 2010. Cezyas folklor [Tsez folklore]. Leipzig– Makhachkala: “Lotos”. Applebaum, Ayla Ayda Bozkurt. 2013. Prosody and grammar in Kabardian. PhD dissertation, University of California, Santa Barbara. Applebaum, Ayla B. & Andrea L. Berez. 2009. A theory is only as good as the data: casting a wide net in Kabardian and Ahtna documentation. In Peter K. Austin, Oliver Bond, Monik Charette, David Nathan, & Peter Sells (eds) Proceedings of Conference on Language Documentation and Linguistic Theory 2, 29-38. London: SOAS. Comrie, Bernard. 2003. A note on pharyngealization and umlaut in two Tsezic languages. In Winfried Boeder (ed.) Kaukasische Sprachprobleme, 105-109. Oldenburg: Bibliotheks- und Informationssystem der Universität Oldenburg. Comrie, Bernard. 2005. Grammatical gender and personification. In Dorit Diskin Ravid & Hava Bat-Zeev Shyldkrot (eds) Perspectives on language and language development: Essays in honor of Ruth A. Berman, 105-114. Dordrecht: Kluwer. Comrie, Bernard. 2013. Ergativity: some recurrent themes. In Edith L. Bavin & Sabine Stoll (eds) The acquisition of ergativity, 15-34. Amsterdam: John Benjamins. Comrie, Bernard & Madzhid Khalilov. 2009. Loanwords in Bezhta, a Nakh-Daghestanian language of the North Caucasus. In Martin Haspelmath & Uri Tadmor (eds) Loanwords in the world’s languages: A comparative handbook, 414-429. Berlin: De Gruyter Mouton. Comrie, Bernard, Madzhid Khalilov, & Zaira Khalilova. 2015. Valency in Bezhta. In Andrej Malchukov & Bernard Comrie (eds) Valency classes in the world’s languages, 541-570. Berlin: De Gruyter Mouton. Grant, Anthony P. 2009. Loanwords in British English. In Martin Haspelmath & Uri Tadmor (eds) Loanwords in the world’s languages: A comparative handbook, 360-383. Berlin: De Gruyter Mouton. Maddieson, Ian, Ramazan Rajabov, & Aaron Sonnenschein. 1996. The main features of Tsez phonetics. UCLA Working Papers in Phonetics 93.

Chapter 4

International Research Collaboration on Documentation and Revitalization of Endangered Turkic Languages in Ukraine: Crimean Tatar, Gagauz, Karaim, Qrymchak and Urum Experience İryna M. Dryga

I would like to begin by thanking Ardahan University and Turk Dil Kurumu for inviting me to speak here today. It’s a real pleasure to have this opportunity to share my views on documentation and revitalization with you. I represent A. Krymsky Institute of Oriental Studies, National Academy of Sciences of Ukraine and at the same time National Network organized for the revitalization of the Turkic minority languages in our country. This initiative of combined efforts of professional linguists, politicians and legislators was demonstrated not at once. It all started years ago with the rare revitalisation attempts and efforts. All of the Turkic languages in Ukraine (Crimean Tatar, Gagauz, Karaim, Krymchak, Urum) are endangered, notably they are at risk of extinction in the short or in the long terms. According to UNESCO’s “Atlas of Endangered Languages” [http://www.unesco.org/languages-atlas/index.php] the degree of their vitality varies from definitely endangered (Gagauz) to extinct (Karaim) [Czató 2010]. Even the most numerous and demographically sizeable Crimean Tatar population is considered linguistically endangered because the language is spoken mainly by the representatives of the older generation. The assessment of this language in the Atlas of Endangered Languages has long appeared to be too optimistic and it is necessary to revise it on the basis of a more profound analysis. First, for the last several years we have spared no effort to collect and preserve the remains of ‘small’ or ‘insular’ Turkic languages in our country [Dryga 2010: 195-200, 220-233, 355-362, 406-419]. In 2006-2007 we conducted a field study in Ukraine and Lithuania in cooperation with the Altaic Society of Korea with a goal to revive, preserve and study the languages and, if possible, cultures of numerically small Turkic language speaking minorities such as Karay, Qrymchaq and Urum, residing mainly in rural localities spread

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_005

52

Dryga

over the multilingual regions of the Crimea, Trakai and Azov and having no native language education [Author’s recordings 2006]. As a result Korean and Ukrainian linguists have collected an invaluable language material given that there was only one speaker of Qrymchak remaining alive, about ten speakers of Qypchaq – Polovets dialects of Urum, eight speakers of the Crimean dialect and two speakers of Halych-Volyn’ dialect of Karaim as of 2008 [Altaic Society of Korea, 2006: 198-203; http://altaireal.snu.ac.kr/askreal_v25/photogroup .html]. The age of all the speakers was over 60, and mostly 80 or more. Second, Lenara Kubedinova and Radovan Garabik from Slovakia began to develop the corpus of Crimean Tatar Wikipedia as a collaborative free content internet encyclopedia [Yavorska, Dryga 2015: 130-136]. Natural language processing often turns to Wikipedia for the source of material, because it exhibits the content, represents contemporary living language and is a valuable source of texts, for example for Linguistic corpus of Crimean Tatar language. For Crimean Tatar it became a point of grassroot prestige, and a major source of accessible online texts in the language. For comparison, no other Turkic language endangered in Ukraine has a Wikipedia corpus. Crimean Tatar Wikipedia (in Latin script) officially started on 12 January 2008, though the first pilot version dates from September 2006. It contains about 4000 articles, ranking 164th by the number of articles. Moreover, Miquel Cabal-Guarro on the basis of the data drawn from the sociolinguistic survey that he conducted in 2011 amongst Crimean Tatars across the peninsula of Crimea, provided an analysis of the language uses and transmission [http://ru.krymr.com/a/25467619.html]. Although the Crimean Tatar language was either seldom or never spoken, especially amongst the individuals of the younger generation (that tend to use only or mostly Russian in their everyday communication, even with their relatives), it is still one of the main identification elements of the Crimean Tatar ethnicity and nearly in all cases claimed to be the identity language of the respondents, almost always even their declared native language. Miquel CabalGuarro also tried to elucidate the degree of endangerment of the Crimean Tatar language. Tudora Arnaut organized 3 international symposiums in Kyiv in 2005-2010 [Arnaut 2005, 2010, 2013] on the problems of the Turkic speaking peoples, and it became our initial experience of public collaborative discussions on the endangered Turkic peoples’ linguistic problems. Quite important decisions taken by the symposiums were quite commendable; they were transferred to the relevant governmental departments and commissions but had no effect [Arnaut 2007]. Moreover, the new language law offered in 2010 by

ENDANGERED TURKIC LANGUAGES IN UKRAINE

53

pro-Russian and Communist Parliament majorities [https://en.wikipedia.org/ wiki/Language_policy_in_Ukraine] ignored these recommendations. Oleksandr Rybalko tried to propose art and cultural events as one of the additional tools for promoting endangered languages in case of Urum language [Yavorska, Dryga 2015: 264-269]. Understanding that in order to be protected, endangered languages require great collective efforts made by various government, academic, and other institutions, he made a field trip with ArtPole Art Agency to Stary Krym (Mariupol), a village founded by Urums speaking the Oghuz dialect, recorded several pieces (fairytales and songs) from a 87-aged native Urum speaker, a storyteller and singer, and one of the fairytales, “Ashyk Garip”, was selected for an art project in a form of video art with subtitles in Ukrainian and English, and accompanied by live music. The project was shown at ArtPole festival, was well received by the audience, had educational value showing to public how rich and diverse Ukraine is linguistically and culturally, and emphasizing the need of urgent support to protect it. Other rare efforts of enthusiasts can be also added. In conclusion: we realized that all the individual efforts were not more than just professional hobbies, and the case of Russian occupation of Crimea and South–Eastern region forced us to reconsider the measures taken before. Russian invasion has set for us new challenges of immediate, target accurate protection of endangered Turkic languages when their situation has fundamentally worsened within less than half a year. OSCE observers having arrived in the country immediately noted the importance of appearence in Ukrainian public discourse on the topic of Turkic language minorities’ rights, against the background of stable deterioration in the Crimean Tatar language’s situation. Add to this targeted physical destruction of non-Russian villages at South– Eastern region of Ukraine. Tragically, settlements of Turks-meskhetians (near Sloviansk) and Urums became places of the most brutal battles. For example, we took as illustration for our conference “Endangered Turkic languages of Ukraine” the photo that became famous among Ukrainian Internet users.1 This is a photo of Urum village Starobeshevo completely razed to the ground where not a single house was left, but just unknown Ukrainian soldiers common grave with homemade cross on it. It became the impetus for the immediate organization of a working group of all interested Ukrainian and Crimean linguistic experts, involving European and Turkish experts that worked on answers to the following questions: how to prevent the threat of language assimilation of Crimean Tatars and other Turkic 1  http://oriental-studies.org.ua/uk/перша-міжнародна-конференція-мови-п/.

54

Dryga

languages represented in Ukraine? What strategies of protection and support should our country choose? The conference analyzed language policies and educational practices towards the endangered languages [Ajniuk 2012] as well as raised awareness of these languages in Ukraine in the context of language diversity in Europe. It also discussed the issues of language planning and language shift including expected cultural and cognitive consequences of language loss from different perspectives. Particularly, when discussing theoretical issues Paul BillbaoSarria mentioned that for successful language recovery we need to meet 4 aspects. These are 1) adequate legislation, 2) proper planning, 3) enough and well spent economic resources, 4) public interest and support, such as backup of a community [Yavorska, Dryga 2015: 69-70]. As a step for creating a civil society, the community must be able to create new speakers of the language, new spaces to use that language and protect means to recover the language as far as create political lobby concerning their language issues, prepare reports on the status of language rights regularly etc. All these tasks were taken on by our National Network on Endangered Turkic languages in Ukraine. Institutionally, we have responded positively to the offer to join projects of European Language Equality Network (ELEN) and hope that this will help in the language rights monitoring in the Crimea, access to which is now unavailable. I would like to draw your attention to the fact that endangered languages in the context of language ideologies are metaphorical notions and security is the main concept here. For many years in Ukraine the situation developed in which not only Turkic languages but Ukrainian and Russian as well were perceived as the languages under threat. As a result of the 2014 conflict perception of the Ukrainian language shifted and it began to be perceived not as a language under threat but as a language-protector [Yavorska, Bogomolov 2010; Yavorska, Dryga 2015: 8]. Crimean Tatar participants of the conference noted that after the events in the Crimea in spring of 2014 Crimean Tatar youth has started to communicate more in native language and it seems to represent a shelter element for the many [Yavorska, Dryga 2015: 119-129; 160-172]. So suddenly endangered languages have become in demand nowadays and we discussed how to explore the shifting mechanism and to use this trend properly. Our Polish colleagues shared their experience of endangered languages’ heritage documenting that is close to us methodologically and technically. In particular, their research group has developed special software for digital processing of various linguistics materials. The researchers have collected data on Kashub, Armenian-Qypchak, Tatar, Karaite and other languages under threat

ENDANGERED TURKIC LANGUAGES IN UKRAINE

55

of extinction and dormant languages. It was also told us about the initiatives of Tatar activists in Poland on revitalization of their native language [Yavorska, Dryga 2015: 10-11]. And the last lesson taught us by the recent events and finally by The language law is that the definition of a particular language endangerment must clearly display the language status in synchrony and diachrony. The fact that this definition is not clearly established created the situation when in certain countries the vigorous languages like German and Russian receive the same protection level as languages on the verge of extinction.2 Conclusions based on reports presented and discussions conducted on the topic: 1.

2.

3.

4.

5.

6.

7.

Dominant culture and the ideologies embedded in it may often negatively affect endangered languages even if unwittingly [Yavorska, Bogomolov 2010]. Split jurisdictions in the aftermath of the collapse of the Soviet Union and wars (occupation of Crimea) complicate sustained and consistent education policies toward the endangered languages. Historical depth that was interrupted from the native perspective and appropriated by cultural others—hence, e.g. the Crimean Tatar texts are perceived as Ottoman. Quite a few of bi-cultural communities still are on the scene as Armenian Qypchaks, Urums, Karaites, Gagauzes who sometimes prefer to be Bulgarians. Facing dilemmas of loyalties and cultural affiliation—affected by the cultural policies and historic narratives of various kin-states and communities. Bilingualism—as both the necessity and the policy issue often unreflected upon properly—both the natives and hosting states tend to side up, explicitly or implicitly, with the preference of monoligualism. The need for a specific alphabet in the case of the endangered Turkic languages—probably phonetic—instead of serving as an instrument to ensure better access to language for those willing to improve their competence, as alphabets often become an ideological minefield. E.g. Crimean Tatars failed to adopt a new Latin based alphabet;

2  About the conference look more at: http://internationaal.pvda.nl/2014/10/01/een-ontmoetingmet-krimtataren/ and http://teraze.org.ua/page.php?id=9&article=4896.

56

8.

Dryga

– the need for digitizing texts including old and new oral ones while native speakers of dialects are still alive; – the need for both bi-lingual and explanatory dictionaries; – the need for oral accounts to be collected while there is still something to collect; – the need for translation as an instrument of enhancing the corpus of modern texts, sustaining and developing language. UNESCO’s Atlas of World’s Endangered Languages still has inadequate characteristics for a number of Turkic languages. The cause is the lack of sources. Consequently, the situation demands a professional re-assessment of the current status of every of Turkic languages.

Our research Institute and our organization are ready to enter as a partner organization into every kind of international research projects, groups, and to join international language corpora and databases with our own language databases and accomplishments, etc. We believe that professional discussion on endangered languages’ preservation issues will contribute to the promotion of cultural diversity as a value, of tolerance towards different ethnic groups and people living in Ukraine. This would ensure the dialog and the understanding between Crimean Tatars, Ukrainian- and Russian-speaking population of the Crimea. It will also create additional opportunities for Crimean Tatars and other Turkic-speaking peoples of Ukraine to take finally, though may be too late, a worthy place in the social and cultural life of the country. It will also help to overcome the political tension around the language issue in Ukraine.

Primary Sources Author’s recordings of Crimean Tatar (Razdol’ne, 1998-1999), Karaim (Simferopol – Evpatoria, 2006, Trakai 2006-2007), Urum (Mariupol – Stary Qrym, 2007, 2008), Krymchak (Simferopol, 2006).

Secondary Sources Ajniuk, Bohdan (ed.), 2012. Ekologia movy i movna polityka v suchasnomu suspilstvi (zbirnyk naukovyh prats).—Кyiv: Vydavnychyj dim Dmytra Burago, 376 p.

ENDANGERED TURKIC LANGUAGES IN UKRAINE

57

Altaic Society of Korea, 2006. Fieldwork Studies of Endangered Altaic Languages. For the Genealogical Study of Korean and the Preservation of Endangered Languages.The Language and Cultural Studies Series 2. Seoul: Altaic Society of Korea. Arnaut, Fedora and Dermenci, Ömer (eds.), 2005. Kruhlyi stil: Tiurkomovni narody Ukrajiny (movy ta kultury tatar, gagauziv, urumiv, karajimiv i krymchakiv). Kyiv: Vydavnychyj dim Dmytra Burago, 112 p. Arnaut, Tudora and Nasrattinoglu Irfan (eds.), 2010. Materialy druhoho Vseukrajinskoho kruhloho stolu “Problemy osvity tiurkomovnyh narodiv Ukrajiny: gagauziv, urumiv, karajimiv, krymchakiv i krymskyh tatar” / II.Ukrayna’daki Türkçe Konuşan Halklar (Gagauz, Urum, Karay, Kırımçak ve Kırım Tatarları’nın eğitim sorunları) Paneli. Kyiv: Vydavnycho—polihrafichnyj tsentr “Kyivskyj universytet”, 257 p. Arnaut, Tudora and Nasrattinoglu Irfan (eds.), 2013. Materialy tretioho Mizhnarodnoho sympoziumu “Tiurkomovni narody Ukrajiny” / III.Uluslararası Ukrayna’da Türkçe Konuşan Halklar Sempozyumu. Кyiv: Vydavnycho—polihrafichnyj tsentr “Kyivskyj universytet”, 335 p. Arnaut, Tudora, 2007. Ukrayna Gagavuzların Arasında Ana Dilinin Gelişmesi, In: Gagavuz Türkçesi Araştırmaları 27-29 Aralık 2007 Bilgi Şöleni. Ankara: Türk Dil Kurumu Yayınları, p. 52. Czató, Èva Á., 2010. ‘Report on an Uppsala workshop on Karaim studies’, In: Johanson, Lars and Csató, Éva Á. (eds.), Turkic languages, Wiesbaden: Routledge. Dryga Iryna, 2010. Pontika: Türkoloji yazıları. Тюркологічні студії. Кyiv: Chetverta hvylia, 530 p. Yavorska, Halyna and Bogomolov, Alexander, 2010. Nepevny object bazhannia: Jevropa v ukrajins’komu politychnomu dyskursi. Кyiv: Vydavnychyj dim Dmytra Burago, 136 p. Yavorska, Halyna, Dryga, Iryna (eds.), 2016 Zahrozheni movy. Krymskotatars’ka ta inshi tiurkski movy v Ukrajini: zbirnyk naukovyh prats’. 1st International Conference Endangered languages: Crimean Tatar and other Turkic languages in Ukraine (26-27th September 2014). Proceedings. Kyiv: Aksioma Medobory, 344 p.

Internet Resources http://altaireal.snu.ac.kr/askreal_v25/photogroup.html http://oriental-studies.org.ua/index.php?news=5379. http://ru.krymr.com/a/25467619.html http://sonseslerduyulmadan.hacettepe.edu.tr http://www.turkiyat.hacettepe.edu.tr/kitap/tehlikedekidillerbildirileri_090413.pdf http://www.unesco.org/languages-atlas/index.php https://en.wikipedia.org/wiki/Language_policy_in_Ukraine

FIGURE 1

Turkic speaking minorities on ethnographic map of Ukraine.

58 Dryga

Appendix

ENDANGERED TURKIC LANGUAGES IN UKRAINE

FIGURE 2

59

Photo of destroyed Urum village Starobeshevo on the banner of the 1st International Conference “Endangered Languages: Crimean Tatar and Other Turkic Languages in Ukraine”.

Chapter 5

Cases-Non-cases: At the Margins of the Tsezic Case System Diana Forker

1

Introduction

The Tsezic languages are a group of closely related languages that form one subbranch within the Nakh-Daghestanian (or East-Caucasian) language family. They can be divided into East Tsezic, comprising Hunzib and Bezhta, and West Tsezic, comprising Khwarshi, Tsez and Hinuq. Tsezic languages are spoken in the Republic of Daghestan, which belongs to the Russian Federation. Daghestan is located in the north-eastern part of the Caucasus. Smaller groups of Tsez and Bezhta speakers also live in Turkey, and some Bezhta speakers live in Georgia. The largest Tsezic language is Tsez with about 12,000 speakers (according to the Russian census of 2010); the smallest language is Hinuq with around 600 speakers. Case assignment in the Tsezic languages is largely semantically motivated, and morphosyntactic features play only a marginal role (cf. Kibrik 1997). Due to the dominant role of semantics in the assignment of case it seems that it is relatively simple to extend the case inventory. That is, suffixes and enclitics with an autonomous distinguished form paired with a clearcut meaning can, in principle, develop into cases. This seems to explain the origin of many spatial cases in Tsezic that most probably go back to spatial postpositions. The aim of this paper is to explore a number of nominal markers that resemble cases and compare them with genuine case markers with respect to functional and formal similarities and differences. I will adopt the canonical approach as exemplified by Corbett (2008) for the feature of case. Corbett (2008) provides ten criteria for canonical case markers and examines the Russian cases in regard to the criteria. I will use these criteria for investigating whether the respective nominal markers from the Tsezic languages could be analyzed as case markers.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_006

61

Cases-Non-cases: At the Margins of the Tsezic Case System

2

Some Notes on the Nominal Inflection in Tsezic Languages

Inflectional categories of nouns in Tsezic languages are number and case. The languages have elaborate gender systems with up to five genders, but nouns are normally not inflected for gender. Inflection is almost exclusively suffixing and largely agglutinative. For the nominal inflection this means that number and case are expressed by different suffixes. Tsezic languages have a rich case inventory with a large number of spatial cases. Non-spatial cases found in all Tsezic languages are absolutive, ergative, genitive, and instrumental. All cases except for the absolutive are expressed by suffixes that are added to the so-called ‘oblique stem’. By contrast, the form that the noun takes in the absolutive case is called ‘direct stem’. The same distinction between direct and oblique stem is found with all sorts of pronouns, and partially also with adjectives and participles. Case formation is straightforward and regular. The main difficulty in the nominal morphology of the Tsezic languages is the formation of the oblique stem from the direct stem. There are several operations that are used in order to form the oblique stem: insertion of vowels, consonants or glides, ablaut, stress shift, conversion, and suffixation of oblique stem markers (Forker 2010a). The last process, the suffixation of an oblique stem marker, is the most frequent and the only productive way of forming oblique stems. The number of oblique markers differs from language to language. In Khwarshi, there are only six markers whereas Hunzib has more than twenty. In all languages, the number of markers is much lower for plural nouns than it is for nouns in the singular. Table 1 illustrates the formation of oblique singular stems to which the genitive suffix has been added.

Table 1

operation

absolutive genitive

Examples of oblique stems for singular nouns Tsez

Hinuq

Khwarshi

Bezhta

Hunzib

ablaut, vowel deletion & oblique marker ‘place’ moči meč-o-s

stress shift

consonant & vowel insertion

deletion of glide

suffixation of oblique marker ‘stable’ bež bež-li-s

‘boy, son’ ‘sibling’ úži ɨs uží-š ɨs-t-ɨ-s

‘eye’ häy hä-l

62

Forker

There is some variation in the stem formation systems. This means that often nouns have more than one way of forming the oblique stem; they follow a rare and unproductive pattern but can also be used with a common oblique marker. Moreover, with spatial cases the case suffix is sometimes attached to the unmodified noun. Depending on the analysis we can either say that in such examples the spatial cases are suffixed to the direct stem or that we deal with conversion.

3

First Case Study: -ɣo / -ɣa

It is plausible to assume that Proto-Tsezic had a case marker *-ɣo that is still found in Bezhta and Khwarshi, but has almost been lost in Hinuq and is absent from Tsez and Hunzib (see Forker 2010b, 2012a for more information on the spatial case systems of Tsezic languages). In Bezhta, the spatial suffix -ɣa has the meaning ‘next to, by, at, near’ (1a) (Kibrik & Testelec 2004: 236). It is also used for the standard of comparison, and occasionally with temporal and metaphorical meaning. The Khwarshi suffix -ɣo translates as ‘in close contact with, nearby’ (1b) (Khalilova 2009: 82). In both languages, further directional cases (lative, ablative, etc.) can be added to the suffix as the Khwarshi example illustrates. (1a) Bezhta buxari-ya-ɣa gäʔä sukʼo=na flue-obl-next be.neg who=add ‘There is nobody next to the flue.’ (1b) Khwarshi (Khalilova 2009: 82) y-õk’-un abaxar yuq’uˁč’eɣo-l=in ii-go-uwpst neighbor(ii) old.woman.apud-lat=add uq’uˁč’eɣo-l=in old.man.apud-lat=add ‘The neighbor went to the grandmother and grandfather.’ In Hinuq, the situation is different because -ɣo is poorly integrated and only occurs in very few spatial and temporal adverbs (Forker 2013: 103). The spatial adverbs č’ek’k’uzaɣo ‘everywhere’ and seda-ɣo ‘in one place’ are formed from quantifiers. The temporal adverbs are sebedoɣo ‘in autumn’ (< sebe ‘autumn’) and aldoɣo ‘formerly, before, in front’. As can be seen in the spatial adverbs the suffix is added to the oblique stem, i.e., seda is the oblique form of the numeral

Cases-Non-cases: At the Margins of the Tsezic Case System

63

hes ‘one’ and -za is an oblique plural stem suffix that has been attached to the quantifier č’ek’k’u ‘all’.

4

Second Case Study: The Vocative

Hinuq and Khwarshi are the only Tsezic Nakh-Daghestanian languages that have a vocative suffix -(i)yu. In Hinuq, it is only used with two common nouns (uži-yu ‘boy-voc’, ked-bi-ža-yu ‘girl-pl-obl.pl-voc’), and it also expresses affectionateness and endearment (Forker 2013: 433-434). In Khwarshi, it seems that it is more widely used because it can occur with various types of common nouns, e.g. cana-yu ‘she.goat.obl-voc’ (Khalilova 2009: 72-73). Neither in Hinuq nor in Khwarshi can the vocative be suffixed to proper names. In both languages the vocative is added to the oblique stem of the noun.

5

Third Case Study: The Functive

5.1 The Functive in Tsezic, Avar and Andic Languages The Tsezic languages as well as Avar and many Andic languages have a suffix that Creissels (2014a, b) calls a ‘functive marker’. I will start my discussion of the functive by describing its functions in Hinuq. The Hinuq functive marker is -ɬun (Forker 2013: 432-433). It is exclusively added to nominals, including case-marked nouns or pronouns and nominalized verbs or adjectives and specifies the property of accomplishing the role or manner or function of whatever the hosting nominal denotes, similar to English ‘as’. More specifically, it occurs in three contexts of use: i.

professions and other social functions (‘as a teacher’, ‘as a friend’, etc.)

(2) Hinuq hago zoq’ʷe-n he be-uwpst ‘He was a doctor.’ ii.

toxtor-ɬun doctor-func

the intellectual activity of regarding a situation in a certain light or aspect (e.g. ‘consider as’)

(3a) sud-mo-y Malla Rasadan bitʼaraw-ɬun ∅-uː-ho court-obl-erg Mullah(i) Nasrudin right-func i-do-prs ‘The court gives Mullah Nasrudin right (lit. makes him the one who is right).’

64

Forker

(3b) deče=gozon debez žo r-eqʼi-yon, me ∅-iči how.much=prt 2sg.dat thing(v) v-know-conc 2sg i-be r-eqʼi-š-me-ɬun v-know-ptcp.res-neg-func ‘How much ever things you know, be as if you would not know them.’ iii.

adverbial clauses expressing causes

(4) аˁši iše y-iq-oru-ƛ’o-ɬun much snow(iv) iv-become-ptcp.pst-spr-func awariya b-iq-iš accident(iii) iii-happen-pst ‘Because of the heavy snowfall he had an accident.’

hayɬoqo he.at

In other Tsezic languages, cognates of the Hinuq suffix are attested (with the exception of Hunzib for which the relevant information is lacking). It seems that in all Tsezic languages the functions i and partially ii are found. The functive suffixes can be added to nouns bearing case suffixes and if they are added to items without overt case suffixes they are attached to the direct stem (not to the oblique stem). In Khwarshi, the suffix is -ɬun / -ɬin (Khalilova 2009: 257-258); in Tsez it is -ɬun (5a). In Bezhta, the functive suffix -ɬun is also used with participles of verbs in a construction with the meaning ‘as if’, e.g. when people are pretending to act in a certain way (5b). Furthermore, all Tsezic languages including Hunzib have a postposition sababɬun consisting of the Avar noun sabab ‘cause, reason’ and the functive suffix. The complex word sababɬun is also used in Avar, but it is not regarded as a postposition in the recent Avar grammar by Alekseev et al. (2014). (5a) Tsez (Abdulaev & Abdullaev 2010) deber di žek’u-ɬun=ä c’ok’i-x-anu=ƛin 2sg.lat 1sg man-func=q consider-ipfv.cvb-neg=quot ‘Don’t you consider me to be a man?’ (5b) Bezhta y-egay-ʔeš-ɬun=na ∅-aq-na giɣa ii-see-ptcp.pst.neg-func=add i-happen-cvb down ẽxe-ɣa=na gowacʼo-na ∅-ẽƛʼe-š Tʼahir river-next=add look.i-cvb i-go-prs Tahir(i) ‘Pretending not to see (her) and looking at the river Tahir is walking by.’

Cases-Non-cases: At the Margins of the Tsezic Case System

65

Avar has the suffix -ɬun that is used in the same or at least in a very similar manner as the functive suffixes of the Tsezic languages (6). The suffix is not explicitly mentioned in the grammars by Charachidzé (1981) and Alekseev et al. (2014). Ebeling (1966: 72), following an older grammar of Avar written in Georgian, writes that it “forms the ‘predicative’, meaning ‘as, in the quality of’ (see Creissels 2014b: 442 for further references). (6) Avar (Axlakov 1976: 13) [Go, but don’t convince him with your hand (i.e. with the force), but] k’ole-b b-at-ani, k’al-zu-l xabar-al-da-ɬun be.able-n n-find-irr.cond mouth-obl-gen talk-obl-loc-func b.aq-e! get.n-imp ‘if you can, get it by talking to him.’ In most of the Andic languages functive suffixes are attested as well, and the meaning corresponds to the meaning of the cognate suffixes in Avar and Tsezic. In Karata, Godoberi, Tindi and Botlikh the suffixes are formally very similar and in one case even identical to the Avar and Tsezic suffixes: – – – –

Karata -ɬe (Creissels 2014b: 444, citing Magomedova & Xalidova 2001) Godoberi -ɬu (Saidova 2006: 226) Tindi -ɬo (Magomedova 2003) Botlikh -ɬun (Creissels 2014b: 444, citing Saidova & Abusov 2012: 109)

Relevant examples are (7a, b) (7a) Godoberi (Saidova 2006: 226) den muʕalim-ɬu=da 1sg teacher-func=cop ‘I am a teacher.’ (7b) Tindi (Magomedova 2003) ihwa-ɬo ħalt’ijaː herdsman-func worked ‘(He) worked as a herdsman.’ In the languages Northern Akhvakh and Bagvalal, which also belong to the Andic branch of Nakh-Daghestanian, the functive suffix is formally slightly

66

Forker

diverse and is followed by a gender agreement marker. The Northern Akhvakh suffix -ɬ+agreement (+-he/-hi) is described in detail by Creissels (2014a, b). It is suffixed to nouns, pronouns, adverbs, etc. The agreement is controlled by the absolutive argument of the clause. The same agreement suffixes are used with the general converb. (8) Northern Akhvakh (Creissels 2014b: 441) di-la hu-be čaka χːirada ʕadati-ɬ-eː 1sg-dat dist-n very dear custom-func-adv.n ‘I considered this a very good custom.’

harigʷ-ari see-pf

The Bagvalal suffix -l(h)i is analyzed by Daniel et al. (2001: 191-193) as a verbforming suffix that attaches to nouns, adjectives, adverbs, and postpositions. Like in Akhvakh it is followed by a gender agreement marker, and in the available examples also by a converbal suffix (9a, b). (9a) Bagvalal gurǯija-j hak’uj-li-j-o j-ah-aː Georgian-f wife-vblz-f-cvb f-take-pot.inf ‘to take a Georgian (woman) as wife’ (Daniel et al. 2001: 193) (9b) milica-lhi-w-o ħalt’idaː-X ida ow police-vblz-m-cvb work.ipfv-cvb be he ‘He works as a police officer.’ (Magomedova 2004) 5.2 Creissels’ Analysis In his analysis of Northern Akhavakh, Creissels (2014a, b) introduces the term ‘functive’ and compares the Akhvakh suffix -ɬ to essive cases in Uralic languages such as Hungarian, Estonian, and Finnish. Creissels (2014b) also makes a proposal concerning the diachronic development of functives in Tsezic, Avar and Andic. Following Alekseev (1988: 35), he suggests that the functive originates from a verb *ɬ- ‘become’ inflected with a converbal suffix. In fact, Avar has а perfective converb suffix -(u)n (Alekseev & Ataev 1998: 62), but synchronically there is no verb ɬ-ize ‘become’.1 Moreover, in Avar-Andic and in Tsezic there are a number of verb-forming suffixes that can reasonably be analyzed as going back to such a verb (Table 2).

1  There is a verb ɬeze ‘put’, but I am not sure about its relation to the functive suffix.

Cases-Non-cases: At the Margins of the Tsezic Case System Table 2

67

Verb-forming suffixes in Avar-Andic and Tsezic

Language

Suffix

Avar Akhvakh Godoberi

xera- ‘old’ > xer-ɬ-ize ‘grow old’ ĩk’a ‘large’ > ĩk’a-ɬ-urula ‘enlarge’ (Creissels 2014b: 441) -eč’uxa ‘big, large’ > -eč’uxa-ɬ-i ‘become large’ (Saidova 2006: 67) -ɬ herk’a ‘big, old’ > herk’ã-ɬ-aɬa ‘become big / old’ (Magomedova & Xalidova 2001: 122) -l(h)i muk’u- ‘small’ > muk’u-li ‘become small, decrease’ (Daniel et al. 2001: 191) -ɬ -oč’č’u ‘cold’ > -oč’-iɬ-a ‘become cold’ (Forker 2013: 329) -ɬ c’odora- ‘clever’ > c’odor-ɬ- ‘become clever’ (Khalilova 2009: 266) -ɬ -uq’o ‘big, old’ > -uq’-ɬ-al ‘grow, increase’ (Khalilov 1995: 136) -ɬ c’uda(ni) ‘red’ > c’uda-ɬ-a ‘become red’ (Khalilov 1999: 279)

Karata Bagvalal Hinuq Khwarshi Bezhta Tsez

Example

-ɬ -ɬ -ɬ

In the Andic languages, not only adjectives and adverbs/postpositions, but also nouns serve as the base for the derivation of verbs. In the Tsezic languages, this seems to be impossible. But Tsez, Hinuq, and Bezhta make use of the same suffix for the derivation of inchoative and potential verbs from other verbs, e.g. Hinuq -ac’- ‘eat’ > -ac’eɬ- ‘be able to eat’ (Forker 2013: 328-330). In Khwarshi, the derivational suffix for potential verbs is -l(l) (Khalilova 2009: 267). When analyzing the Akhvakh functive Creissels (2014b: 441) discusses the question whether it could be considered a derivational suffix employed for the formation of verbs, as has been done by Daniel et al. (2001) for Bagvalal. In fact, such an approach seems plausible since the gender marker following -ɬ in the Akhvakh functive marker also constitutes the suffix of the general converb. Moreover, it is cross-linguistically common to express the functive meaning through a construction with a verb such as ‘be, become’. However, Creissels (2014b) provides four counterarguments and suggests at the same time to treat the functive as a case marker. The counterarguments are: 1. 2.

it can be added to base words from which verbs normally cannot be formed, e.g. interrogative pronouns such as ‘what’ the nouns to which the suffix is added can be marked for case and/or number (6)

68 3. 4.

Forker

nouns bearing the functive suffix can take modifiers (e.g. genitive modifiers or adjectives) (8), (10) nouns with the functive suffix serve not only as adjuncts but also as predicative arguments of verbs such as ‘be’, ‘become’, ‘consider’, etc. (2), (5a), (10)

However, despite the functional similarity to cases such as the Uralic essive, I want to argue that it does not seem the best solution to treat the functive as belonging to the case systems of Tsezic, Avar or Andic. I will concentrate in my argumentation on the functive in Hinuq because this is the language for which I have most data, but my first and my second argument equally apply to the other Tsezic languages. Why should the Hinuq functive not be considered a case marker? How does it diverge from the other case suffixes? First of all, it can be added to other case markers (4). Second, it is suffixed to the direct form of the nominal or nominalized adjective or verb. For instance, if the resultative participle takes case suffixes, it must be in its oblique form -za, but, as (3b) illustrates, this is not the case for the functive. Third, nouns to which the functive is added do not trigger the use of the second genitive (10). These three properties are not attested for the case suffixes. Case suffixes are not added to other case suffixes (though there are complex spatial cases, but they always have as the second suffix a directional marker). Case suffixes are added to the oblique stem (though for some nominals oblique and direct stem are identical). All nominals inflected for any case other than the absolutive trigger the use of the second genitive (10b). The first genitive is only available for nouns in the absolutive case. (10a)

(10b)

Hinuq hago di ∅-egennu essu-ɬun he 1sg.gen1 i-young sibling(i)-func ‘May he be my younger brother!’ dižo essu-y tʼek 1sg.gen2 sibling-erg book ‘My sibling reads a book.’

∅-iči-ƛo! i-be-opt

tʼotʼer-ho read-prs

The similarity of the Hinuq suffix -ɬun with its Avar source suggests that it must be a more recent borrowing and considerably younger than the verb-deriving suffix -ɬ. The derivational suffix is used with a wide range of base words including many native Hinuq words and is rather deeply integrated into the linguistic system. The derived verbs behave just like simple verbs. Therefore,

Cases-Non-cases: At the Margins of the Tsezic Case System

69

Hinuq -ɬun and -ɬ cannot directly be traced back to the same origin, even if this is probably the case for the Avar equivalents.

6

Fourth Case Study: Tsezic -ɬi

The last case study examines the suffix -ɬi and its use in the Tsezic languages. This suffix has again been borrowed from Avar. Its functions in Tsezic largely overlap with its functions in Avar, but there are also some differences between Avar -ɬi and Tsezic -ɬi as well as among the Tsezic languages. The functions are: i.

Derivation of abstract nouns from other nouns, adjectives, non-finite verbs forms, etc. in all Tsezic languages, e.g. Bezhta gäččö ‘be not’ > gäččöɬi ‘absence’, Tsez halmaɣ ‘friend’ > halmaɣɬi ‘friendship’, Hunzib ɨs ‘brother’ > ɨsɬi ‘family’

ii.

Marking complements of verbs of knowledge and cognition in all Tsezic languages except Hunzib (11a, b) (Forker 2016). In Bezhta, this function seems to be broader than in the other languages and also includes other types of subordinate clauses (Kibrik & Testelec 1994: 261).

(11a)

Bezhta [do hoƛoʔ biƛo-ʔ ∅-eče-yo-ɬi] ∅-iqʼe-da . . . 1sg here house-in i-be-ptcp-abst i-know-cond ‘if they knew that I lived here in the house . . .’

(11b)

[biƛo r-oː-ro-ɬi] zuq’o-ro kutakalda zaħmatab house(iv) iv-do-ptcp-abst be-pst strongly difficult ‘Building the house was very difficult.’ (Kibrik & Testelec 1994: 261)

iii.

(12a)

Marking the topic of a conversation or narration in Hinuq and Tsez (12a, b). In Hinuq, the topic of a conversation can also be marked with the first genitive, and in Tsez with the cont-Ablative (or some other spatial cases). Tsez (Abdulaev & Abdullaev 2010) žoyä [babi-ya ɬina-ɬ xizay egäru-ɬi] boy.erg father-erg what.obl-cont after send.ptcp.pst-abst esi-n tell-uwpst ‘The boy told why (lit. after what) the father had sent him.’

70 (12b)

iv.

Forker

Hinuq dižo tʼek-mo-za-ɬi r-egi roži eƛi-š 1sg.gen2 book-obl-obl.pl-abst v-good word(v) say-pst batʼi-batʼiyaw poʔet-za-y=no, ʡalim-za-y=no various poet-obl.pl-erg=add intellectual-obl.pl-erg=add ‘Various poets and intellectuals said good things about my books.’ Marking X in phrases like ‘X turns into Y’ in Hinuq (13), and apparently also in Tsez (in the latter language this is done in combination with the spr-Lative)

(13) Hinuq haɬu ked-i zon-ɬi b-uː-ho arxi, that.obl girl-erg refl.obl-abst iii-make-prs ditch(iii) gulu-za-ɬi r-uː-ho ɬe horse-obl.pl-abst v-make-prs water(v) ‘The girl turns herself into a ditch and the horses into water.’ v.

Marking of nominals that are governed by the postposition ʡolo ‘because of’ in Hinuq (14) (alternatively, the purposive suffix -ƛi can be used in the same function)

(14) Hinuq xexza-ɬi ʡolo eli aƛ-a-do child.obl.pl-abst because.of 1pl village-in-dir ‘Because of the children we came to the village.’ vi.

nox-iš come-pst

Formation of a simultaneous converb and of a few spatial adverbs in Hinuq (Forker 2013: 141, 241, 358-361)

Most of the usages are functionally related because they involve the formation and the use of abstract nominals, either as genuine parts of the nominal lexicon, or in a more abstract sense. For example, complements can be analyzed as nominalized propositions that occur in a position where otherwise nouns occur. The topic of a conversation or narration is an abstract object that can be expressed in the form of a nominalized proposition or other item. Once again we ask the question whether the suffix -ɬi can be analyzed as a case suffix. If we restrict ourselves to Hinuq, which seems to be the language with the broadest range of use, the functions iii-v are functions that are typically

Cases-Non-cases: At the Margins of the Tsezic Case System

71

associated with (spatial) cases.2 What concerns the morphosyntactic side, in functions iii-v -ɬi is added to oblique stems of the respective nominals just like case suffixes (13), (14). Modifiers of the nominal bearing the suffix must appear in their oblique form or as second genitive (12b). Finally, in functions iii-v the suffix can be followed by another suffix -zo that is identical to the second genitive and the second ablative. The second ablative belongs to the directional markers, which are allowed to follow other case suffixes. This might appear to provide us with a further argument in favor of the analysis of -ɬi as a case suffix. However, the directional markers only attach to spatial cases expressing location and not to any other cases, and the functions iii-v do not involve any unambiguously spatial semantics. Yet Hinuq has a couple of spatial adverbs such as hayɬi ‘there’, hibayɬi ‘there’, and haɬi ‘here’ that are diachronically complex consisting of stems of demonstrative pronouns and a suffix -ɬi (Forker 2013: 358-361). It seems that it is precisely the suffix that turns the demonstrative pronouns into spatial adverbs. Under such an approach -ɬi can be viewed as a case suffix that occupies the boundaries of the Hinuq spatial case system.

7

Exploring the Margins of the Hinuq Case System with the Canonical Approach

In order to explore the borderline cases of the Tsezic case systems I adopt the canonical approach as developed by Corbett and his colleagues (see Corbett (2005, 2008) and Brown, Chumakina & Corbett (2013) among others). The basic idea of Canonical Typology is that when investigating a linguistic phenomenon we should start by modelling a canonical item that represents an idealized instance of the examined phenomenon. The properties of the canonical item will help us to classify specific examples with respect to their similarity to the canonical item. In the following, I take Corbett’s (2008) analysis of case and apply it to the four linguistic items that have been treated in Sections 4-7 thereby concentrating on the Hinuq data. In other words, I will clarify to what extend the following suffixes can be analyzed as belonging to the case system of Hinuq: the spatial suffix -ɣo, the vocative -(i)yu, the functive -ɬun, and the abstract suffix -ɬi.

2  Function vi, i.e. the formation of verb forms employed in adverbial clauses, is also attested with some spatial cases in Hinuq, in particular the spr-essive (Forker 2013: 240-241; 255-257).

72

Forker

7.1

The Non-spatial Cases of Hinuq Viewed through the Eyes of Canonical Typology Corbett (2008) lists ten criteria that help to distinguish canonical instances of cases from less canonical ones. Below I briefly present the criteria. In his analysis, Corbett stresses that he treats case as a morphosyntactic feature. He also mentions Nakh-Daghestanian local cases as an example of cases expressing morphosemantic features to which the criteria do not apply. Therefore, I will largely concentrate on two non-spatial cases of Hinuq, the ergative and the dative, and show how these cases behave with respect to the criteria. Criterion 1: Canonical features and their values have a dedicated form. For instance, the Hinuq dative has a suffix that is formally unique within the case system (-z). By contrast, the form of the ergative overlaps with the form of the in-Essive for some nouns (-i) and with the absolutive and partially the first genitive for personal pronouns (see Forker (2013: 130) for the paradigms). Criterion 2: Canonical features and their values are uniquely distinguishable across other logically compatible features and their values. We do not have to select particular combinations: any of them will serve. This criterion is met by all Hinuq cases: they are expressed independently of other categories such as number or gender. Criteria 3 & 4: Canonical features and their values are distinguished consistently across relevant word classes and across lexemes within relevant word classes. Almost all word classes and all lexemes within the different word classes can be inflected for all cases in Hinuq and thus the system is close to being canonical. The ergative represents a minor exception because singular and plural personal pronouns for the first and second person do not distinguish between absolutive and ergative (and with plural personal pronouns the first genitive is also identical to the absolutive/ergative). Criterion 5: The use of canonical morphosyntactic features and their values is obligatory. The use of cases in Hinuq is obligatory and every nominal is treated as bearing a specific case value. In this sense, criterion 5 is met by all Hinuq cases. Choices between cases are treated as non-canonical (see Criteria 6 & 7). Criteria 6 & 7: Canonical use of morphosyntactic features and their values does not admit syntactic or semantic conditions.

Cases-Non-cases: At the Margins of the Tsezic Case System

73

Cases in Hinuq, as well as in the other Nakh-Daghestanian languages, have a clear and specific semantic load which governs their use. For instance, the use of the ergative to mark agents is obligatory; it cannot be replaced by any other case because it is the only case that expresses agentivity. In non-canonical agent constructions, the at-essive instead of the ergative is used, but the case marking turns the argument into a non-canonical agent that lacks a number of agentive properties such as control of his/her actions and volitionality (Forker 2013). Similarly, the dative obligatorily marks experiencers, beneficiaries and purpose. It also marks recipients, but in this function it alternates with the at-essive/at-lative. However, the alternation implies clear differences in the semantics: the dative expresses recipients to whom an object is permanently transferred whereas the at-essive/at-lative expresses temporary recipients (Daniel, Khalilova & Molochieva 2010).3 Therefore, Corbett (2008) excluded the spatial cases right from the beginning. However, it is not clear to me whether the just described alternations between non-spatial and spatial cases are also irrelevant in the view of Corbett or if they can be counted as non-canonical behavior of the non-spatial cases. Corbett points out that differential object marking represents an instance of non-canonical behavior regulated in many languages by the semantic condition of definiteness. Non-canonical agent marking and recipient marking as mentioned above are regulated by similar semantic conditions, but, in contrast to differential object marking in languages such as Turkish, involve an alternation between grammatical and spatial cases and not between two different grammatical cases. Hinuq has a further construction that might count as noncanonical behavior of the case system, the biabsolutive construction, in which the agent occurs in the absolutive instead of the ergative (Forker 2012b, 2013: 522-529). Thus, now we have an alternation between two grammatical cases that is determined by morphosyntactic properties of the verbal predicate and by semanto-pragmatic conditions. In sum, there are no additional syntactic conditions on the use of the Hinuq cases, but there are a few constructions that might count as semantic conditions leading to a deviation from the canonical case ideal.

3  There are even more semantic restrictions with respect to the spatial cases. For example, certain place names and microtoponyms can be only inflected for directional cases but not for the spatial cases expressing locations because of the inherent locational semantics of those nominals. For nouns with animate referents it is difficult if not impossible to find a context of use for spatial cases such as the in-directional. But these restrictions are purely semantic and do not concern the morphology or the syntax.

74

Forker

Criteria 8 & 9: Canonical use of morphosyntactic features and their values does not admit additional lexical conditions from the target (governee) or from the controller (governor). The controller has a single requirement (e.g. it governs the dative). There are (almost) no special lexical conditions on the side of targets, e.g. all nouns take the same case if the same meaning needs to be expressed. There are occasional cases of idiosyncrasy (e.g. the noun zoro ‘barn’ is inflected for the sub-essive that normally means ‘under’ to express the meaning ‘in the barn’), but they perhaps have a semantic explanation. From the side of the controllers there are also (almost) no lexical conditions, e.g. all transitive verbs require the agent to be marked with the ergative, and all affective verbs require the experiencer to be marked with the dative. However, the affective verb -aši‘find’ additionally allows for the at-essive to occur on the experiencer and this might be considered as an additional construction, but again it is a variation between a grammatical case and a spatial case. Criterion 10: The use of canonical morphosyntactic features and their values is sufficient (they are independent). This criterion is fully met in Hinuq because the non-spatial cases do not permit additional markers such as postpositions.4 7.2 The Borderline Cases Table 3 summarizes the behavior of the Hinuq ergative, dative, vocative (Section 5), functive (Section 6) and of the suffix -ɬi in the functions iii to v. The suffix -ɣo (Section 4) has been excluded since it is clearly an old spatial case and spatial cases were regarded by Corbett as expressing morphosemantic and not so much morphosyntactic features. Table 3 displays how well the cases and case-like morphemes fulfill the ten criteria. At the bottom the table indicates if the morphemes require the nominal stem to occur in the oblique form and if modifiers such as demonstrative pronouns also need to occur in the oblique form. These two properties are specific for the Tsezic case system and are therefore also considered.

4  The spatial cases allow them, but the use of postpositions with spatial cases is mostly optional.

75

Cases-Non-cases: At the Margins of the Tsezic Case System Table 3

Comparing non-spatial cases and case-like morphemes

Criteria

Ergative

Dative

Vocative

Functive

-ɬi (functions iii-v)

1 2 3&4 5 6&7 8&9 10 obl stem obl modifier(s)

partially yes partially yes yes* yes yes yes yes

yes yes yes yes* yes* yes* yes yes yes

yes yes no no yes no yes yes ?

yes yes yes partially yes yes yes no no

yes yes yes partially yes yes no yes yes

The ergative behaves relatively canonical except for its partial formal overlap with the absolutive and the in-essive and its alternation with the absolutive and the at-essive in non-canonical agent constructions (indicated by * in Table 3). The dative is even more canonical since it has a unique ending. Again the only instances that might count as non-canonical behavior are alternations with certain spatial cases. The vocative is the least canonical nominal marker. Its use is not obligatory because the absolutive can be used instead, and it occurs only with a very limited number of nouns. However, it is suffixed to the oblique stem. With respect to the criteria 1-10, the functive is as canonical as the ergative or the dative. The only non-canonical feature is found with respect to criterion 5 because the functive can occasionally be replaced with the absolutive, e.g. compare (15) with example (2) above. (15) Hinuq Abdukarim ħaži Buynaksk šahar-mo-s imam Abdukarim Gadzhi Buynaksk town-obl-gen1 imam ‘Abdukarim Gadzhi is the imam of the town of Buynaksk.’

goɬ be

Thus, Creissels’ (2014b: 442-443) critique of excluding the functive from the case inventory as it is done in all grammars of Tsezic languages (as well as in the grammars of Avar and Andic languages) is surely justified. Yet it does not confirm to all other cases when it comes to the specific property of requiring nominals to occur with the oblique stems and modifiers to appear in their

76

Forker

oblique form. Moreover, as mentioned in Section 6, the functive can be added to other case markers, and such a behavior is not typical for genuine case suffixes. Finally, the suffix -ɬi is slightly less canonical than the functive because functions iii and v can also be fulfilled by other markers (criterion 5), and the suffix can optionally be followed another suffix, i.e., it permits additional markers (criterion 10). But in contrast to the functive it behaves well with regard to oblique stems.

8

Conclusion

In this paper, I investigated borderline cases of the Tsezic case systems that resemble core cases, but also deviate from them in various ways. This was done by applying the canonical approach as illustrated by Corbett’s (2008) ten criteria for canonical case markers. The examined nominal markers can be ordered along a scale of canonicity from being more canonical (functive, -ɬi) to less canonical (vocative). The Tsezic case systems are far from being homogenous because of the division into non-spatial and spatial cases with concomitant differences in their formal and functional properties. They lose members (e.g. the spatial suffix -ɣo that was described in Section 4, certain spatial case combinations seem to have fallen out of use in Bezhta, see Forker 2012a, b). But that they also gain new members. The functive and the abstract suffix -ɬi are both borrowings from Avar that are functional equivalents of cases in other languages, and the latter suffix in Hinuq also shares some formal properties with the core case suffixes.

Abbreviations i-v: genders i-v, abst: abstract suffix, add: additive, adv: adverbial, apud: apud-essive, at: location ‘at, by, near’, conc: concessive, cond: conditional, cont: cont-essive, cop: copula, cvb: converb, dat: dative, dir: directional, dist: distal, erg: ergative, f: feminine, func: functive, gen: genitive, imp: imperative, in: in-essive, inf: infinitive, ipfv: imperfective, irr: irrealis, lat: lative, loc: locative, n: neuter, neg: negation, next: location ‘next to’, obl: oblique stem, opt: optative, pf: perfective, pl: plural, pot: potential, prs: present, prt: particle, pst: past, ptcp: participle, q: question marker, quot: quotative, res: resultative, sg: singular, spr: location ‘on’, uwpst: unwitnessed past, vblz: verbalizer

Cases-Non-cases: At the Margins of the Tsezic Case System

77

References Abdulaev, Arsen K. & I. K. Abdullaev. 2010. Didojskij (cezskij) fol’klor. Makhachkala: Lotos. Alekseev, Mixail E, 1988. Sravnitel’no-istoričeskaja morfologija avaro-andijskix jazykov. Moscow: Nauka. Alekseev, Mixail E., S. Z. Alixanov, Boris M. Ataev, M. A. Magomedov, I. I. Magomodov, G. I. Madieva, Patimat A. Saidova & Dzh. S. Samedov. 2014. Sovremennyj avarskij jazyk. Makhachkala: Aleph. Alekseev, Mixail E. & Boris M. Ataev. 1998. Avarskij jazyk. Moscow: Academia. Axlakov, A. A. 1976. Avarskie teksty. In A. A. Axlakov & Kh M. Khalilov (eds.), Satir i jumor narodov Dagestana, 7-42. Makhachkala: Daginogoizdat. Brown, Dunstan, Marina Chumakina & Greville G. Corbett (eds.) 2013. Canonical morphology and syntax. Oxford: Oxford University Press. Charachidzé, Georges. 1981. Grammaire de la langue avar (langue du Caucase NordEst). Paris: Farvard. Corbett, Greville G. 2005. The canonical approach in typology. In Zygmunt Frajzyngier, Adam Hodges & David S. Rood (eds.), Linguistic diversity and language theories, 25-49. Amsterdam: Benjamins. Corbett, Greville G. 2008. Determining morphosyntactic feature values: the case of case. In Greville G. Corbett & Michael Noonan (eds.), Case and grammatical relations, 1-34. Amsterdam: Benjamins. Creissels, Denis. 2014a. Functive phrases in typological and diachronic perspective. Studies in Language 38. 605-647. Creissels, Denis. 2014b. Functive-transformative marking in Akhvakh and other Caucasian languages. In Michael Daniel, Vladimir A. Plungian & Ekatarina A. Ljutikova (eds.), Jazyk. Konstanty. Peremennye: Pamjati Aleksandra Evgen’eviča Kibrika, 430-449. Sankt-Peterburg: Aletejja. Daniel, Michael, Nina Dobrushina, Tat’jana Sosenskaja & Sergei Tatevosov. 2001. Derivacionnaja morfologija. In Aleksandr E. Kibrik (ed.), Bagvalinskij jazyk, 186-197. Moscow: Nasledie. Daniel, Michael, Zaira Khalilova & Zarina Molochieva. 2010. Ditransitive constructions in East Caucasian: A family overview. In Andrej Malchukov, Martin Haspelmath & Bernard Comrie (eds.), Studies in ditransitive constructions, 277-315. Berlin: De Gruyter. Ebeling, Carl L. 1966. The grammar of Literary Avar: Review of Chikobava and Cercvadze’s ‘The Grammar of Literary Avar’. Studia Caucasica 2. 58-100. Forker, Diana. 2010a. Nonlocal uses of local cases in the Tsezic languages. Linguistics 48. 1083-1109.

78

Forker

Forker, Diana. 2010b. Variation in stem formation in Tsezic languages. Suvremena lingvistika 69. 1-19. Forker, Diana. 2012a. Spatial relations in Hinuq and Bezhta. In Luna Filipović & Kasia M. Jaszczolt (eds.), Space and time across languages, disciplines, and cultures: Volume I. Linguistic diversity, 15-34. Amsterdam: John Benjamins. Forker, Diana. 2012b. The bi-absolutive construction in Nakh-Daghestanian. Folia Linguistica 46. 75-108. Forker, Diana. 2013. A grammar of Hinuq. Berlin: De Gruyter. Forker, Diana. 2016. Complementizers in Hinuq. In Kasper Boye & Petar Kehayov (eds.), Semantic functions of complementizers in European languages. Berlin: De Gruyter. Khalilov, Madžid Š. 1995. Bežtinkso-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Khalilov, Madžid Š. 1999. Cezsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Khalilova, Zaira. 2009. A grammar of Khwarshi. Utrecht: LOT. Kibrik, Aleksandr E. 1997. Beyond subject and object: towards a comprehensive relational typology. Linguistic Typology 1. 279-346. Kibrik, Aleksandr E. & Jakov G. Testelec. 2004. Bezhta. In Michael Job (ed.), The indigenous languages of the Caucasus, vol. 3: The North East Caucasian languages, part 1, 217-295. Ann Arbor: Caravan Books. Magomedova, Patimat T. 2003. Tindinsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Magomedova, Patimat T. 2004. Bagvalinsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Magomedova, Patimat T. & Rašidat Š. Xalidova. 2001. Karatinsko-russkij slovar’. SanktPeterburg: Scriptorium. Saidova, Patimat A. 2006. Godoberinsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Saidova, Patimat A. & Magomed G. Abusov. 2012. Botlixsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN.

Chapter 6

Language Endangerment in the Balkans with Some Comparisons to the Caucasus Victor A. Friedman

1

Introduction

In this article, I would like to offer a comparative perspective on language endangerment in the Balkans and the Caucasus. My focus here will be more on the Balkans than the Caucasus, as the former has received less attention in this regard than the latter.1 In a sense, we can view the Balkans and the Caucasus as two prongs of southeastern Europe, both of which have been long-time zones of contact between Christian and Muslim polities. To be sure, the phrase southeastern Europe usually conjures up images of the Balkans but not the Caucasus, and the relationship of the Caucasus to Europe remains ideologically unstable.2 Nonetheless, the Western and Eastern shores of the Black Sea—the west being a peninsula defined by the Adriatic and the East being an isthmus defined by the Caspian—do share aspects of the ecologies of language endangerment. In addition to being zones of contact between Muslim and Christian Empires, the Balkans and the Caucasus also share certain marginalized minorities, viz. Gypsies and Jews, where in each case the linguistic fates and adaptations are also worthy of comparison. In this article, therefore, my goal will be to draw attention to the types of language endangerment occurring in the Balkans and to make appropriate comparisons with the Caucasus.

1  On the situation in the Caucasus, see Friedman (2010). There is no literature to speak of on language endangerment in the Balkans beyond the third edition of the UNESCO atlas (Mosley 2010). 2  A clear example of this instability occurs in Webster’s Geographical Dictionary, where the political border but the northern borders of Iran and Turkey define the southern border of Europe in the entry for Asia, but the Caucasian ridge defines the boundary between Europe and Asia in the entry for Europe (Bethel 1949: 74, 347). © koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_007

80 2

Friedman

Balkans and Caucasus Compared

Among the language ecological commonalities between the Balkans and the Caucasus in the twentieth century we can identify are the effects of forced migration (including genocide), colonization, marginalization, economic transformation (including urbanization, economic migration, and shifts in traditional market patterns), and the rise of new polities.3 One way of combating endangerment is to be found in literacy practices, but this is not in and of itself a simple or sufficient solution. It is now well understood that language preservation cannot be divorced from social context (e.g., Mülhäusler 2002). At the same time, however, in the context of symbolic capital in societies that are already deeply embedded in modern nation-states, literacy becomes important (cf. Friedman 2012a). In order to set this discussion in the present context, a look at history and the concept of indigeneity is useful. For most of the world’s endangered languages, indigeneity is part of both the discourse involved in the exercising of rights including, among others, language rights. 2.1 Indigeneity in the Caucasus In the case of the Caucasus, the indigenous argument is straightforward in so far as the South (Kartvelian), Northwest (Abkhaz-Adyghe) and Northeast (Nakh-Daghestanian) families are concerned. Regardless of the prehistory that accounts for current distributions, the antiquity of the three Caucasian language families in the Caucasus is beyond dispute, just as is the antiquity of Basque in Europe. Nonetheless, such a criterion leaves out many languages that have been spoken in—or achieved their formation in—the Caucasus despite being obvious subsequent but nonetheless ancient arrivals, viz. various IndoEuropean, Turkic, and Semitic languages such as Armenian, Ossetic, KarachayBalkar, Kumyk, Assyrian, etc. 2.2 Indigeneity in the Balkans In the case of the Balkans, all of the languages are either Indo-European or Turkic, and none should therefore claim indigeneity in the sense of the Caucasian language families. Nonetheless, those Indo-European languages descended from languages spoken in the Balkans longer than others use 3  The literature on these phenomena is too vast to cite, but among the relevant works dealing with events that are less well known we can note particularly Ladas (1932), Üngör (2011), and Richmond (2013).

Language Endangerment in the Balkans

81

precisely the “autochthonous” argument as part of their justification for what at times can be viewed as ideologically colonial practices (cf. Friedman 2016). In addition to Albanian and Greek, however, most of the languages of the Balkans took shape precisely in the Balkans, and, like Albanian and Greek, arrived from outside the peninsula and replaced the languages already spoken there. among them Aromanian, Meglenoromanian, Romani, the Balkan dialects of Judezmo, all of the Balkan Slavic languages and dialects, Gagauz, and West Rumelian Turkish, as well as Arvanitika, and Tsakonian in the Albanic and Hellenic branches, respectively (Friedman and Joseph Forthcoming). If we take Bugarski’s (1992: 10) definition of ‘autochthonous’ in the former Yugoslav context, i.e. spoken natively on the territory of the relevant polity for at least a century, and if we extend that measure back by an extra five hundred years or a millennium, we arrive at a concept of indigenous that adequately accounts for the Balkan languages in the sense of the Balkan Sprachbund as well as languages specific to the Caucasus that includes languages descended from later arrivals. We can thus say that in this extended sense, indigeneity in the Balkans and the Caucasus can be defined in terms of the centuries-long language contact that creates a linguistic area (cf. Friedman 2010b, Comrie 2007). As we shall see, competing claims to relative ‘indigeneity’ in the Balkan context are used to justify policies that lead to the extermination of languages that have been spoken on a given territory for centuries and even millennia.

3

Imperial Contexts

Turning for a moment to the competing empires, and beginning with the early modern period, it was the Ottoman, Safavid, and Russian empires that impacted the Caucasus and the Ottoman, Russian, and Austro-Hungarian that impacted the Balkans. On a scale of linguicidal practices, Muslim empires were, at least until the early twentieth century, relatively benign, the AustroHungarian Empire had competing destructive and protective tendencies (cf. Gal 2015), while the Russian Empire has been linguicidal (cf. Wixman 1980). In both the Balkans and the Caucasus, ideologies of totalizing, mono-ethnic nation-states that had their origins in Western Europe worked to destroy traditional multilingualisms, albeit through different mechanisms, and with different results in different locales.

82 4

Friedman

Stateless Languages in the Balkans

Among the Balkan languages, all of those without titular nation-states suffer from some degree of endangerment ranging from moribund to ecological fragility. 4.1 Judezmo Of the moribund languages, the worst affected is Judezmo, the majority of whose speakers were murdered by the Nazis and their collaborators during World War Two. Unlike Yiddish, which has seen a genuine revival in recent years owing in large part to the spread of Lubavitcher Hasidism, now known as Chabad, Judezmo is retreating. In Macedonia, by 2008 there were between five and ten fluent speakers left, all survivors of World War Two. Their numbers have declined since then, although there is still a handful of semi-speakers among their children, prospects for the future are non-existent as of this writing. The situation is similar albeit not as dire, in places such as Sofia, Salonika, Sarajevo, and Bucharest, as well as Istanbul, although there are some younger activists in Istanbul. The communities in the USA, Canada, and Israel are, for the most part, all members of older generations. In the Caucasus, several hundred Juhuro speakers were killed by Nazi occupiers, but for the most part the community survived, although today its dispersal in Israel, the US, and Russia leave the language endangered in Daghestan and Azerbaijan (cf. Altschuler 1990). We can mention here in passing both Judeo-Greek (Romaniote) and Judeo-Georgian (Q’ivruli), which are basically ethnolects of their respective languages. In both cases, the number of speakers is dwindling owing to the Holocaust in the case of the former and migration to Israel in the case of the latter. 4.2 Meglenoromanian The other most highly endangered language in the Balkans is Meglenoromanian, which was originally limited to a cluster of villages on and near Mt. Pajak, in a region that is today divided between the Hellenic Republic and the Republic of Macedonia. By the time this territory was ceded by Turkey to Serbia and Greece in 1913, it was limited to eleven villages—three ended up on the Serbian side, later the Republic of Macedonia, and eight were on the Greek side. Of the eight that were ceded to Greece, however, the largest, Nãte/Nãnte, was Muslim, and so all its inhabitants were sent to Turkey in the 1923 exchange of populations mandated by the Treaty of Lausanne. There they disappeared from view until being rediscovered in Turkish Thrace, independently by Turkish and by Austrian and Turkish ethnologists working in the region (Kahl 2006, Kurtişoğlu and Kurtişoğlu 2012). But the modern Nãntintsi preserve only a few

Language Endangerment in the Balkans

83

Meglenoromanian songs and dances and the memory of where their forebears came from. Meanwhile, the language has gone extinct in two other villages (Atanasov 1990, 2002)—one on each side of the border—leaving a total of seven villages. Prior to the Balkan Wars, Macedonian was the primary contact language for all Meglenoromanian villages, but on the Greek side Macedonian was outlawed in the 1930s and practically all the Macedonian-speaking villages in the Meglen were emptied and/or destroyed at the end of the 1940s. Thus, at present Meglenoromanian is assimilating to Greek in Greece but Macedonian in Macedonia. The future of Meglenoromanian is particularly grim since 1) in Romania it is considered a dialect of Romanian (Rusu 1984), 2) in Greece all citizens of Greece are Greeks and minority languages are actively discouraged (Friedman 2012a), and 3) in Macedonia there is no support for Meglenoromanian maintenance and the speakers are counted with Aromanians (Friedman 2001). 4.3 Aromanian Aromanian is in a distinctly different situation from Meglenoromanian in that there are till hundreds of thousands of speakers. They live in northern Greece, southern Albania, the Republic of Macedonia, and southwestern Bulgaria. Thousands also emigrated to Romania in the wake of the Balkan Wars and World War One. In Romania, Aromanian immigrants were settled mostly around Tulcea in Dobrudja, which at the time was the least Romanian and most ethnically mixed region of Romania and had only recently been assigned to Romania. The situation can be compared to the settlement of Pontic and Cappadocian Greeks in northern Greece, i.e., Greek Macedonia, after the 1923 exchange of populations.4 More or less officially, however, Romania considers Aromanian to be a dialect of Romanian and it thus receives no support aside from the occasional folklore production. In the countries where Aromanian originated, it is only in the Republic of Macedonia, which has one of the smallest number of Aromanian speakers, that the language has any sort of official recognition and support. In Greece, the language is mostly limited to

4  Pontic and Cappadocian—the Hellenic languages spoken by Christians expelled from Turkey in the population exchanges mandated by the 1923 Treaty of Lausanne and, in the case of Pontic, also spoken in the Soviet Caucasus until the dissolution of the USSR, when the speakers left for Greece—could be considered independent Hellenic languages (like Griko in Italy), but owing to Greek language policy, they have been driven to the brink of extinction upon being relocated to Greece. Tskaonian differs historically from these in that it is in many respects descended directly from Doric Greek rather than the Hellenistic Koine that is the source of all other Hellenic languages. Nonetheless, the situation for Tsakonian is likewise precarious.

84

Friedman

speakers over 40, although in some more isolated villages, as elsewhere in the original Aromanian territory, the language is still being passed on to children (cf. Bara 2005). Despite official support for Aromanian in the Republic of Macedonia, however, numbers of speakers are declining, and the language remains vital mostly in isolated villages. There is also a significant Aromanian diaspora in Western Europe and North America, but Constantin Belamaci’s (1888, cited in Balamaci 1987) rage remains a forlorn call, and Hadzhi Daniil’s (1802, reproduced in Ninčev 1977) poem continues as Greek policy.5 4.4 Romani Of the remaining stateless languages, Romani has some speakers in the Caucasus, but, perhaps most significantly, its apparent closest relatives— Lomavren and Domari—are also spoken in the Caucasus (Marushiakova and Popov 2014). This region is thus unique in having—or at least having had— representatives of all three of these diasporic Indic languages, of which Romani took its definite shape in the Byzantine Empire, Domari (also called Karachi, Garbet, Zutt, Nawar, etc.) in the Middle East, and Lomavren (also called Bosha, 5  Constantin Belamaci was the author of an Aromanian poem entitled Parinteascã dimandare ‘Paternal commandment’ that became the Aromanian anthem. The verse cited here gives a sense of the content: Cari-shi alasã limba a lui For whomever leaves his language: S’lu ardã pira focului Let him be burned by flame Si s-dirinã yiu pri loc Let him be destroyed alive where he stands Si lli si frigã limba n-foc. Let his tongue be burned in flame. In contrast, Hadži Daniil of Moschopole (modern Voskopoja, Albania) was a Hellenized Aromanian who published a quadrilingual phrasebook the express purpose of which was the extinction of Balkan languages other than Greek, as seen in the opening lines of his introduction (reproduced in Ninčev 1977:83, given here in original and transliterated Greek and in Wace and Thompson’s (1913:6) English verse translation. Ἀλβανοὶ, Βλάχοι, Βούλγαροι, Ἀλλόγλωσσοι χαρῆτε, Κ᾽ἑτοιμασθῆτε ὅλοι σας Ῥωμαῖοι νὰ γενῆτε. Βαρβαρικὴν ἀφήνοντες γλῶσσαν, φωνὴν καὶ ἢθη, Ὁποῦ στοὺς Ἀπογόνους σας νὰ φαίνωνται σὰν μῦθοι. Albanoi, Vlakhoi, Voulgaroi, Alloglōssoi kharēte, K’etoimasthēte oloi sas Rōmaioi na genēte. Varvarikēn aphēnontes glōssan, fōnēn kai ēthē, Opou stous Apogonous sas na phainōntai san mēthoi. Albanians, Bulgars, Vlachs and all who now do speak An alien tongue rejoice, prepare to make you Greek, Change your barbaric tongue, your customs rude forego, So that as byegone myths your children may them know.

Language Endangerment in the Balkans

85

Posha, etc.) in Armenia. Etymologically, it is clear that all three autonyms—Rom Dom, Lom—are derived from the same Indic case name ḍom, but the languages differ in such fundamental respects that Matras (2012), quite reasonably, doubts that they necessarily ever constituted a single speech community, although it is equally clear that at times their contact was sufficiently intimate for them to have shared innovations. The three languages also occupy three distinct places on the cline of endangerment. Romani is ecologically fragile: it remains vital as long as its speakers suffer a social marginalization that is as effective as the physical isolation of a mountain village, this latter being what protects ecologically fragile languages in, e.g., Daghestan. With advances in social equality, however, strategies for language preservation become necessary. A number of these are being deployed in various contexts, and the need for them continues Domari speakers are in the final stages of shift to majority contact languages. Predication is still possible, but whole sections of the grammar have already shifted. The Lomavren recorded by Patkanov (1887) appears to have been where Domari is now (cf. Matras 2012). At present, Lomavren, like AngloRomani, is only a lexicon (see Friedman 2011). 4.5 Caucasian Languages in the Balkans: Circassian, Armenian, Georgian Aside from Romani, other languages that are or were shared by the Balkans and the Caucasus were Circassian (West Circassian [Adyghe] and East Circassian [Kabardian]), Armenian, and Georgian. According to Kănčov (1900: 116), the majority of Circassians that came to the Balkans did so as refugees from the expanding Russian Empire in 1864. They were settled for the most part in The Ottoman vilayets of Selânik and Üsküp/Kosova. With the Ottoman loss of that territroy in the Balkan Wars 1912-193, followed by World War One, followed by the forced migrations of the 1920 (cf. Ladas 1932), the majority of Balkan Circassians left for or were forced to move to what became the Turkish Republic. By the late twentieth century, there was one Circassian-speaking village left in the Balkans, Čerkes kjoj, about ten kilometers from Prishtina (Kurmel 1994). These last speakers of Balkan Circassian were evacuated to the Adyghe Republic during the NATO bombing of 1999.6 Two other languages of the Caucasus with a presence in the Balkans are Georgian and Armenian. A Georgian monastery (P’et’ric’on) was founded at Bačkovo in the Bulgarian Rhodopes in the eleventh century (construction 6  The only datum we have is Besirov and Tlebzu (1981), who mention the replacement of the ergative by the instrumental. We can speculate that this is a result of Serbian influence. Kănčev (1900: 116, 178, 215) gives locations and statistics for Circassians in Macedonia, but their dialect is completely lost to us.

86

Friedman

completed 1083), when the Georgian province of Tao and Bulgaria were both part of the Byzantine Empire. Two inscriptions and old copies (thirteenth century) of the monastery’s tipikon survive (Šanidze 1971). 4.6 Minority Languages and Dialects Related to Nation-state Languages When we turn to Balkan languages that are related to Balkan nation-state languages but are spoken in polities where the dominant language belongs to a different linguistic group, we see a fairly consistent pattern of endangerment. Arvanitika, which separated from the main body of Albanian and migrated to what is today Greece about a millennium or so ago, has been pushed to the brink of extinction in the course of the past 60 years or so. Of the Slavic languages/dialects spoken in non-Slavic polities, Goran is endangered in Kosovo by Bosnian and in Albania by Albanian, although the relative isolation of the villages is impeding the process. The isolated Macedonian dialect of Boboshtica in the Korcha region of Albania, which speakers themselves called Kajnas ‘like us’ is now a linguistic tourist attraction performed for visiting foreign linguists by a single old woman. Pomak in Bulgaria is Bulgarian, while in Greece, since all its speakers there are Muslim, it is giving way to Turkish. On the one hand, Christian speakers of these Rhodopian dialects consider themselves and their language to be Bulgarian, whereas Muslim speakers do not identify as ethnic Bulgarians. The Greek government has supported this situation by allowing Pomak teaching materials (e.g., Kokkas 2004), which are vigorously opposed by the Turkish elites in Greece (cf. Steinke and Voss 2007).7 In Macedonia, West Rumelian Turkish remained quite vital as a home language despite schooling in standard Turkish, but at present it is in retreat owing to greater freedom of communication. Gagauz has official support in the Republic of Moldova, but not elsewhere, where it is in retreat. In some ways, researching endangered languages in Greece is like researching them in a war zone, because the Greek government and its supporters pretend that Greece’s minority languages constitute a threat to its national security (see Friedman 2012a). The irony is that it is Greece that has contributed to the Republic of Macedonia’s insecurity, and yet, at the same time, endangered language documentation in the Republic of Macedonia is unproblematic.

7  In linguistic terms, Rhodopian dialects are very distinctive, albeit part of the Balkan Slavic dialect continuum that includes Bulgarian, Macedonian, Goran, Torlak, etc. The singling out of Torlak as a separate language in Mosley (2010), while understandable in the context of the former Serbo-Croatian, is more problematic in the context of Balkan Slavic.

Language Endangerment in the Balkans

5

87

Conclusion

I would like to conclude with some thoughts on language endangerment as it relates to the Balkans and the Caucasus. In the case of the Caucasus, the linguistic diversity, typological specificity, and ecological fragility are all obvious arguments for attention. And there are obvious scientific reasons for prioritizing the documentation of typologically unusual or historically relatively unconnected languages. The sense of urgency is profound, and the need for resources—both human and monetary—is undeniable. At the same time, however, the larger intellectual justifications for rescue and revitalization include social issues of equity and human well-being that are equally relevant to minority languages under threat in the Balkans. To this can be added the historical argument that the endangered languages and dialects of the Balkans contain historical and cultural information that will be lost when they are. And owing to the centuries of stable multilingualism, these languages and dialects have much to tell us about language contact as well (see, e.g., Friedman 1997 on the admirative in the Aromanian dialect spoken by the Frasheriote—but not Mbaliote—villagers from Bela di suprã [Gorna Belica, Macedonia]). There is also a certain historical irony in the fact that in many parts of the world, where Europe’s settlement colonies have drastically reduced linguistic diversity, the heirs of these colonists have some recognition of the need for trying to reverse or mitigate the harm that has been done, but in Europe itself, especially in the southeast, there is little to no such recognition, and in some places even resistance and outright harassment even for efforts at documentation, let alone revitalization. This, too, lends a certain sense of urgency to the argument.

Acknowledgments This article draws on more than forty years of my own field work in the Balkans and the Caucasus. I wish to acknowledge support from fellowships from the following US granting agencies: American Council of Learned Societies for a Fellowship in East European Studies (1986) and (2000-01) financed in part by the National Endowment for the Humanities and the Ford Foundation, International Research and Exchanges Board for travel grants to Macedonia in 1991and 1992, National Endowment for the Humanities (2001, Reference FA-36517-01), and Fulbright-Hays Post-Doctoral Fellowship from the U.S. Department of Education as well as a fellowship from the John Simon Guggenheim Memorial Foundation in 2008-2009, American Council of

88

Friedman

Learned Societies for a Fellowship in East European Studies with support from the National Endowment for the Humanities and the Social Science Research Council (2012-2013), an American Councils for International Education (ACTR/ ACCELS) Title VIII Research Fellowship with support from the U.S. Department of State, Title VIII Program for Research and Training on Eastern Europe and Eurasia (Independent States of the former Soviet Union) (2012). Some of the research reported here from other sources was conducted while I was an honorary visitor at the Center for Research on Language Diversity at La Trobe University. The opinions expressed herein are entirely my own.

References Adamou, Evangelia. 2008. “Armenian.” In Evangelia Adamou (ed.), Le Patrimonie plurilingue de la Grèce (Le nom des langues II), 71-76. Leuven: Peeters. Altshuler, M. 1990. Yehudei Mizraħ Kavkaz. Jerusalem: Hebrew University. Atanasov, Petar. 1990. Le mégléno-roumain de nos jours. (Balkan-Archiv Neue Folge— Beiheft 8). Hamburg: Buske. Atanasov, Petar. 2002. Meglenoromâna astăzi. Bucharest: Romanian Academy. Bara, Maria. 2005. Južnoaromunskij govor sela Turia (Pind). Munich: Biblion. Balamaci, Nicolas S. 1987. A New Tone for Our Cultural Discussions. The Newsletter of the Society Farsarotul 1(2)[July]. 11-12. Bethel, John P. (gen. ed.) 1949. Webster’s Geographical Dictionary. Springfield, MA: G. & C. Merriam Co. Bugarski, Ranko. 1992. Language in Yugoslavia: Situation, Policy, Planning. In Ranko Bugarski and Celia Hawkesworth (eds.), Language Planning in Yugoslavia, 9-26. Columbus: Slavica. Comrie, Bernard. 2008. Linguistic Diversity in the Caucasus. Annuasl review of Anthropology 37. 131-143. Friedman, Victor A. 1994. Surprise! Surprise! Arumanian Has Had an Admirative! Indiana Slavic Studies 7. 79-89. Friedman, Victor A. 2001. The Vlah Minority in Macedonia: Language, Identity, Dialectology, and Standardization. In Juhani Nuorluoto, Martii Leiwo, Jussi Halla-aho (eds.), Selected Papers in Slavic, Baltic, and Balkan Studies, (Slavica Helsingiensa 21), 26-50. Helsinki: University of Helsinki. Friedman, Victor 2010a. Sociolinguistics in the Caucasus. In Martin Ball (ed.), Encyclopedia of Sociolinguistics of the World’s Languages, 127-38. London: Routledge. Friedman, Victor A. 2010b. The Balkan Languages and Balkan Linguistics. Annual Review of Anthropology 40. 275-291.

Language Endangerment in the Balkans

89

Friedman, Victor A. 2011. Review of Matras, Yaron. 2010. Romani in Britain: The Afterlife of a Language. Edinburgh: Edinburgh University Press. The Journal of Language Contact, 4. 295-301. Friedman, Victor A. 2012a. A Tantrum from the Cradle of Democracy: On the Dangers of Studying Macedonian. In Victor C. de Munck, and Ljupcho Risteski (eds.), Macedonia: The Political, Social, Economic and Cultural Foundations of a Balkan State, 22-43. London: I.B. Tauris. Friedman, Victor A. 2012b. Copying and Cognates in the Balkan Sprachbund. In Lars Johanson and Martine Robeets (eds.), Copies vs Cognates in Bound Morphology, 323-336. Leiden: Brill. Friedman, Victor. 2016. The Importance of Aromanian for the Study of Balkan Language Contact in the Context of Balkan-Caucasian Parallels. In Thede Kahl and Ioana Nechiti, (eds.), Ethnic and Linguistic Diversity in Southeast Europe and the Caucasus. Vienna: Austrian Academy of Sciences. Friedman, Victor A. and Brian D. Joseph. Forthcoming. The Balkan Languages. Cambridge: Cambridge University. Gal, Suan. 2015. Imperial Linguistics and Polyglot Nationalism in Austria-Hungary: Hunfalvy, Gumplowicz, Schuchardt. Balkanistica 28. 151-173. Kahl, Thede. 2006. The Islamisation of the Meglen Vlachs (Megleno-Romanians): The Village of Nânti (Nótia) and the “Nântinets” in Present-Day Turkey. Nationalities Papers 34. 71-90. Kokkas, N. 2004. Úchem so pomátsko. Ksanthi: Politistiko Anaptyksiako Kentro Thrakēs Kănčov, Vasil. 1900. Makedonija: Etnografija i statistika. Sofia: Bălgarsko knižovno družestvo. Kurmel, Ömer Aytek. 1994. Kosova ve Çerkesler. Yedi Yıldız 4.14-15. Kurtişoğlu Belma and Bülent Kurtişoğlu. 2012. Hidden Latin in Thrace: Notyalılar. Paper delivered at the 3rd Symposium of the Study Group on Music and Dance in Southeast Europe, Berovo, Republic of Macedonia, 4/17/2012-4/23/2012,

accessed 11 December 2015. Ladas, Stephen P. 1932. The Exchange of Minorities : Bulgaria, Greece, and Turkey. New York: Macmillan. Matras, Yaron. 2012. A Grammar of Domari. Berlin/Boston: Walter de Gruyter. Mosley, Christopher (chief ed.) 2010. Atlas of the World’s Languages in Danger. Paris: UNESCO. Marushiakova, Elena and Veselin Popov. 2014. The Gypsies (Dom—Lom—Rom) in Georgia. Proceedings of Annual Meeting of the Gypsy Lore Society and Conference on Romani Studies, Bratislava, September 11-13, 2014. accessed 11 december 2015.

90

Friedman

Mülhäusler, Peter. 2002. How One Cannot Preserve Languages (but can preserve language ecologies). In David Bradley and Maya Bradley (eds.), Language Endangerment and Language Maintenance, 34-38. London: Routledge Curzon. Ninčev, A. 1977. Četiriezičnijat rečnik na Hadži Daniil. Sofia: BAN. Patkanov, K. I. 1887. O narečijah zakavkazskih Cigan: Boša i Karači. St. Petersburg: Imperatorskaja Akademija Nauk. Richmond, Walter. 2013. Circassian Genocide. New Brunswick, NJ: Rutgers University Press. Rusu, Valeriu (chief ed.) 1984. Tratat de dialectologie românească. Craiova: Scrisul românesc. Šanidze, A. 1971. Kartvelta monast’iri bulgaretši da misi t’ip’ik’oni. Tbilisi: Mecniereba. Steinke, Klaus and Christian Voss (eds.) 2007. The Pomaks in Greece and Bulgaria: A Model Case for Borderland Minorities in the Balkans. Munich: Verlag Otto Sagner/ Südoosteuropa Gesellschaft. Üngör, Uğur Ümit. 2011. The Making of Modern Tukey. Oxford: Oxford University. Wace, A. J. B. & M. S. Thompson. 1913. The Nomads of the Balkans. New York: Dutton. Wixman, Ronald. 1980. Language aspects of Ethnic Patterns and Processes in the North Caucasus. Chicago: University of Chicago Department of Geography.

Chapter 7

Instilling Pride by Raising a Language’s Prestige George Hewitt

I have to begin by mentioning Ubykh, given my good fortune in meeting the last speakers in Hacı Osman Köyü and Istanbul in 1974. The demise of the North West Caucasian language Ubykh was played out on Turkish soil following the mass-migration of the Ubykhs (perhaps up to 50,000 souls) at the end of the Great Caucasian War in 1864. The late French scholar, Georges Dumézil, who, of course, made the greatest contribution to documenting and analysing this language, wrote how the Ubykhs even in their homeland (viz. in the environs of modern-day Sochi) may always have been bi-/tri-lingual in one or both of the languages of their neighbours and close linguistic relatives (namely the western Abkhazians and the western Circassians). After resettling in regions of Turkey where the majority of their fellow-Caucasian migrants were either Circassians or Abkhazians, their elders evidently decided that, since they would have to learn the language of their Turkish hosts and also to communicate with their fellow-migrants, they would simply not bother passing on their native Ubykh to the new generations. The decision was regrettable but understandable, given (a) the circumstances in which the Ubykhs found themselves and (b) that at that time nobody worried about the disappearance of languages. What we have here, then, is a clear example, it would seem, of a communitydecision to allow their language to wither away, deeming it to be of less value than those of their neighbours and hosts. And wither it duly did, being described as already moribund by those scholars who first visited the Ubykh settlements (the Dane Å. Benediksten 1898, the German A. Dirr 1913/14, the Hungarian J. Meszaros 1930/31, and Dumézil 1930). When Dumézil returned to Turkey in 1953 after World War II, he expected to find no speakers left, but fate brought him together with Tevfik Esenç, who, much younger than the other speakers, had been raised by his grandparents and thus had had the by-that-time rare opportunity to be exposed largely to Ubykh alone until he started school. Though the language of Tevfik’s wife and the other Ubykh females in the Manyas settlements never seems to have been investigated, Tevfik has become universally described as the last fully competent speaker, taking the language with him to the grave in 1992, though he did pre-decease his wife.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_008

92

Hewitt

Had the Ubykhs reconciled themselves to Russia’s victory in 1864 and accepted the offer to live compactly under Russian domination (albeit perhaps in the Kuban basin away from their traditional hilly terrain), their language would surely be alive today, even if many of its speakers would certainly have perished in the upheavals of the Revolution and Stalin’s Terror. This is a realistic supposition, since both standard Abkhaz (plus its divergent Abaza dialect) and two varieties of Circassian (western Adyghe and eastern Kabardian) were included in the list of so-called ‘Young Written Languages’ by the early Soviets. This gave them literary status and officially approved scripts. It would be interesting to learn if anything has been published about the discussions over which languages were to be included in this classification or even if records of those discussions have survived. The effort that went into supporting languages in the Caucasus in the 1920s until around the mid-1930s is astonishing. For example, although the Daghestanian Lezgic language Udi was spoken in only 3 villages on either side of the Georgian-Azerbaijani border, a 2-dialect primer was published (in Sukhum!) in 1934, and, despite the fact that only a tiny number of South Caucasian Laz speakers lived on Soviet (Georgian) territory, a schoolprimer was published (again in Sukhum) in 1935, two issues of a Laz paper Mch’ita Murutsxi ‘Red Star’ having appeared in 1929 (see Feurstein 1992.320-22 for images of these rarities). Though Abkhaz, Abaza, Adyghe and Kabardian have retained their literary status, literary aspirations for Udi and Laz (perhaps unsurprisingly) quickly faded. But at first glance a far more suprising loss of status was suffered by the S. Caucasian language Mingrelian, given the huge number of ethnic Mingrelians concentrated in the west Georgian province of Mingrelia. I introduce Mingrelian as an example of how a state can act deliberately to undermine the prestige that I argue is necessary for speakers to feel in order to sustain their willingness to pass their language on to their children, thereby ensuring the language’s long-term survival. Back in the 1980s when I was researching Mingrelian the elderly mother of my main Mingrelian informant (in Ochamchira, Abkhazia) asked her son (in his 50s) what use Mingrelian was to a British linguist when it was of no use even to native Mingrelians! It would be foolish, given the numbers involved,1 to suggest that Mingrelian is in any immediate danger, but its fate over the last century well illustrates the negative consequences for languages that government-actions can have. In his paper to the 1990 Caucasian Colloquium at SOAS the German Wolfgang Feurstein, who has long championed the cause of the Laz in Turkey and that of Laz’s 1  The number of ethnic Mingrelians (let alone that of competent speakers) is unknown, but it has been estimated that speakers might number up to half a million.

Instilling Pride by Raising a Language ’ s Prestige

93

close sister Mingrelian (plus that of the more distant sister Svan) in Georgia, compared the treatment of Laz in Turkey with that of Mingrelian and Svan in Georgia (see Feurstein 1992), and I myself have contributed to the discussion (Hewitt 1995). More recent examples could be cited, but simply consider this brief summary: the leading Mingrelian politician of the 1920s, Ishak Zhvania, advocated autonomy for Mingrelia together with the teaching of, and publication in, Mingrelian; books such as Ch’ita Chxoria ‘Red Ray’ and the Zugdidi newspaper Q’azaxishi Gazeti ‘Peasant’s Paper’ appeared from 1930, but around this time, as Stalin secured control of the Kremlin and Lavrent’i Beria assumed Zhvania’s mantle as the leading Mingrelian politician, all talk of autonomy ceased, Mingrelians were officially classified as ‘Georgians’, and publishing in Mingrelian for Mingrelians (as opposed to linguistic or folkloristic works for specialists) ended c.1938. Whilst it is popularly believed in Georgia that Mingrelian, Laz and Svan are mere dialects of Georgian, even local linguists who know this not to be the case but who do not wish to upset the official designation of these peoples as ‘Georgians’ have invented the (to my mind) odd term ‘sociolinguistic dialect’ to describe the status of these tongues in Georgia. This general downplaying of Mingrelian must surely have contributed to the sad remark made by that old lady back in the 1980s. Of course, the clue as to why Tbilisi has wanted as many residents of Georgia as possible to identify themselves as ‘Georgians’ (just as for most of its existence the Turkish Republic has tended to regard all residents of Turkey as ‘Turks’) lies in Zhvania’s desire to establish a Mingrelian autonomy. Separatism was feared and remains feared to the present day.2 Let us recall Stalin’s own words from his famous 1913 article on Marxism and the National Question: ‘What is to be done with the Mingrelians, the Abkhasians, the Adjarians, the Svanetians, the Lesghians [lek’ebi in the Georgian version—BGH], and so on, who speak different languages but do not possess a literature of their own? To what nations are they to be attached? . . . What is to be done with the Ossets, of whom the Transcaucasian Ossets are becoming assimilated (but are as yet by no means wholly assimilated) by the Georgians and the Ciscaucasian Ossets are partly being assimilated by the Russians and partly continuing to develop and are creating their own literature? How are they to be “organised” into a single national union?’ (pp. 48-9 of an undated English translation). I asked an elderly Mingrelian in Upper Gal in the autumn of 2013 about his ethnic self-awareness and how he distinguished between Mingrelians and 2  Interestingly, it seems that the reason for the quick demise of just 2 issues of the Laz newspaper in Georgia in 1929 may have been the result of a request from Turkey to discontinue it (Feurstein 1992.299).

94

Hewitt

Georgians. He replied: ‘We Mingrelians have our own language, but we are only a people [Georgian xalxi], whereas the Georgians are a nation [Georgian eri].’ Asked how he differentiated between peoples and nations, he said that the Georgians formed a nation ‘because they have a literature’! Whilst Stalin’s view might have prevailed amongst (?most/many/all) Mingrelians, recent attempts to introduce this categorisation to the Laz in Turkey by visitors from Georgia have caused what I would judge to be entirely natural resentment. Consider these quotes from a statement released in 2013 by a group of Laz intellectuals: We do not find it ethical and vigorously condemn the appropriation of people who have contributed to the Laz Community and the attribution of those to the Georgians, through distortion and forced interpretations . . . If Georgia and institutions in Georgia seek to raise the Laz’s sympathy and increase their satisfaction it is recommended that they should turn to topics like cultural autonomy and mother-tongue education for Mingrelian and Svan, and support ways to encourage and strengthen these languages. It would be a more democratic manner for Georgia if they intended recognition of these languages as mother-tongues without being pressed by Europe or other political powers but by means of a feeling of responsibility coming from within the country and its institutions. It should be noted that Mingrelian and Svan are a part of the world’s cultural heritage and efforts made to keep these languages alive are nourished by universal principles, which arise out of universal human rights. The extinction of these languages is a problem that concernes not only the speakers themselves but all humanity (To the Turkish and Georgian Public, 2013).3 There is an oblique reference in this statement to the European Charter for Regional or Minority Languages (ECRML), which Georgia has yet to sign (cf. Wheatley 2009). It is true that Turkey too has not signed it either, but, possibly as part of its drive to join the EU, Turkey has relaxed official sanctions against the use of minority-languages on its territory. And the Laz (along with Circassians and Abkhazians) have taken welcome advantage of this positive change in official attitudes.4 It is my firm conviction that linguists wishing 3  Read the statement at: http://abkhazworld.com/news/statements/973-laz-intellectualsexplain-their-view-of-laz-ethnicity.html). 4  The Laz have made tremendous progress as outlined in a recent communication from a member of the community, Eylem Bostancı:

Instilling Pride by Raising a Language ’ s Prestige

95

to preserve endangered languages should, (even at the risk of incurring unpopularity) either themselves or through statist institutions like the ECRML, seek to persuade non-progressive states to recognise that providing proper provision for minority languages (including language-tuition at school upto a modest level perhaps, just to endow the language concerned with some standing)5 is NOT necessarily an encouragement to separatism. If proper linguistic provision is accompanied by respect for ethnic identity and nondiscriminatory treatment across the board, there is absolutely no reason why separatism should result. To turn finally and briefly to Abkhazia, I would note that, possibly as a direct result of the war with Georgia, one has the impression that the need to preserve Abkhaz is more widely felt than previously, language being an obvious and immediate marker of identity—in this case the identity targeted by 1) From September 2013, in the 5th-8th classes of the secondary schools Laz can be taken as an elective lesson. In 2013, there were only five classes in the towns of Fındıklı (Rize) and Arhavi (Artvin). In September 2014, the number of classes went up to 15, with pupils taking Laz as an elective lesson in the towns of Ardeşen, Pazar and Fındıklı in the province of Rize, and in the towns of Borcka and Arhavi in the province of Artvin. We are hoping that the number of classes will go up to at least 30 in the next academic year and that classes will also open in Hopa. 2) The Laz Institute, which was established in 2013, is developing relations with the linguists and scholars around the world. It introduced courses for the teaching of Laz for adults in Istanbul. The Institue has been commissioned by the Turkish Ministry of Education to prepare text books to be used during the elective Laz lessons at the secondary schools. The institute is also having talks with a number of universities in Turkey for Laz courses to be introduced at these universities. Laz elective courses will continue to take place at the University of Bosphorus (Bogazici Universitesi) in the 2014 academic year. 3) Laz intellectuals continue to work towards creating a Laz literature. The Lazika Yayın Kollektifi (The Lazika Publication Collective) has published more than 60 books in the last four years of its establishment, of which only five are in Turkish and all the rest are in Laz. We have published Laz dictionaries, a Laz periodical (Tanura), five Laz novels, books on the Laz history, Laz poem books, a high number of children’s books, and a variety of translations into Laz from famous fairytales/stories from around the world. The fairytales/stories that have been translated into Laz so far are as below and more are to come. We will be attending to the Tüyap Istanbul Book Fair (the largest book fair in the country) in November. The Little Prince (also translated into Mingrelian by the Laz Cultural Association), Romeo and Juliet, The Little Black Fish, Polly-Anna, The Snow White, Don Quixote, Pinocchio. 5  This is not to suggest that the prime responsibility for passing on a language should lie with schools. Native speakers are the ones who can and should most easily fulfil this task by merely opening their mouths and speaking to children. But they might not feel inclined to do so, if they do not recognise the importance of this simple task.

96

Hewitt

Georgian nationalism. With the aim of encouraging universal study and use of the language, a law was introduced on 27 November 2007 under the late Pres. S. Bagapsh requiring all official business to be conducted in Abkhaz from 1 January 2015. Whilst this might have been (and indeed remains as of October 2015) a noble aspiration, passing a law without making provision in terms of teaching, publication of relevant language-materials, etc . . . means that it will be impossible to enforce (the Vice-President elected on 24 August 2014 does not, for example, know Abkhaz!) and renders that law pointless. With Abkhazia’s economy still in a parlous state, any assistance in terms of help with producing suitable manuals and the training of teachers (especially for those classes and subjects that have traditionally been taught in Russian even in Abkhaz-language schools once tuition switches after the first few grades from Abkhaz to Russian) would be welcome. And, of course, the question of Mingrelian is also relevant to Abkhazia. Here I have long advocated that, if the local Mingrelians want their children to be educated through the medium of Georgian (as seems to be the case), this should be allowed on condition that Mingrelian too is taught in local schools up to a certain level of competence in order to raise its profile and prestige amongst its native speakers. This view is not popular, however, with many Abkhazians, I have to admit! But if the cause is noble, the battle is worth fighting, as I hope we can all agree.

Postscript I should add in conclusion that in an e-mail received from Georgia only a week before this paper was read at Ardahan University a Mingrelian correspondent who has translated and published a Mingrelian translation of ‘The Little Prince’ = ch’ich’e mapaskiri6 (organised by the Laz [sic] Cultural Organisation in Turkey, as noted in Footnote 4 above) told me that attitudes in Georgia to publishing Mingrelian materials seem to be changing, a greater readiness than, say, 10-15 years ago to accept such publications being noticeable. This was welcome news, and one can only hope firstly that it is true and secondly that such tolerance widens and deepens there.

6  See http://www.petit-prince-collection.com/lang/show_livre.php?lang=en&id=2615.

Instilling Pride by Raising a Language ’ s Prestige

97

References Feurstein, Wolfgang. 1992. Mingrelisch, Lazisch, Swanisch. Alte Sprachen und Kulturen der Kolchis vor dem baldigen Untergang, in [B.] George Hewitt (ed.) Caucasian Perspectives, 285-328. Unterschleissheim/München: Lincom Europa. Hewitt, [B.] George. 1995. Yet a third consideration of Völker, Sprachen und Kulturen des südlichen Kaukasus., Central Asian Survey, 14.2, 285-310. Wheatley, Jonathan. 2009. Georgia and the European Charter for Regional or Minority Languages, European Centre for Minority Issues (ECMI), Working Paper 42.

Chapter 8

Unwritten Minority Languages of Daghestan: Status and Conservation Issues Zaynab Alieva and Madzhid Khalilov

The modern world is divided into spheres of influence of several economic and cultural (and linguistic) powers, which gradually absorb small nations and ethnic groups. On the contrary, small ethnic groups (with their fragile languages) lose their social significance. The use of native languages becomes limited; they are not spoken any more, and eventually undergo extinction. Among the languages that fall victim of this process are indigenous languages of Daghestan, namely unwritten languages of the minority ethnic groups. In the modern ethno-sociolinguistics, peoples of a country are traditionally classified according to a quantitative characteristic, viz. the number of people speaking a particular language. According to this principle, ‘minority ethnic groups’ appear as part of the peoples of Russia. As shown in the official census, the minority ethnic groups make 63 ethnic units in Russia, which are listed in the ‘Red Book of the Peoples of Russia’, appear to be in the state of ethnic disaster and are to become the object of prioritized protection. All Daghestanian unwritten and newly written languages are listed in this book. The problem of multilingualism in Daghestan and origin of its peoples has always attracted not only linguists’ but also other researchers’ attention. The migration theory of language origin of the Caucasus was popular among early researchers. According to this theory, for many centuries the Caucasus has served as a transit route for many tribes and nations in the era of Great Migrations. The inadequacy of the migration theory of peoples and languages of the Caucasus was partly established by their histrorical comparative study and genetic classification. The Caucasus belongs to one of the few regions in the world where an extraordinary diversity of languages is represented in a relatively small area. More than a half of the languages in the Caucasus are spoken in the republic of Daghestan: there are more than three dozens of languages. Twenty-six of them belong to the indigenous population and are genetically related. These are the following languages:

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_009

99

Unwritten Minority Languages of Daghestan

Avar Agul Andi Archi Akhwakh Bagwalal Bezhta Botlikh Budukh

Hinukh Hunzib Godoberi Dargwa Karata Kryz Lak Lezgi Rutul

Tabasaran Tindi Udi Khwarshi Khinalug Tsakhur Tsez Chamalal Chechen

which form the Nakh-Daghestanian language family. There are also a significant number of languages spoken in Daghestan that belong to other families: Azeri, Kumyk, Nogay, which belong to the Turkic family, and Tat, which belongs to the Iranian branch of the Indo-European family. Currently, there are eleven official languages: Avar Dargwa Lak Lezgi

Tabasaran Chechen Azeri Kumyk

Nogay Russian Tat

three languages that recently became written: Agul

Rutul

Tsakhur

Budukh Hinukh Hunzib Godoberi Karata Kryz

Tindi Udi Khwarshi Khinalug Tsez Chamalal

and 18 unwritten languages: Andi Archi Akhwakh Bagwalal Bezhta Botlikh

in the Republic of Daghestan. It should be noted that all speakers of Budukh, Kryz, Khinalug, and Udi live in the Republic of Azerbaijan. In this list, there are small one-village languages (i.e. languages spoken only in one single village),

100

Alieva and Khalilov

for example, Hinukh spoken by 600 people. The largest of the unwritten ethnic groups is Andi whose language is spoken by about 30 000-40 000. The lack of reliable data on the number of the so-called small ethnic groups falls in the category of curiosities of demographic statistics in Daghestan. However, according to the most precise data it is assumed that the number of speakers of the unwritten languages, as part of the three million people inhabiting the region, exceeds 150 000. More than 100 000 people represent 13 ethnic groups of the Andi-Tsez language group. Agul, Tsakhur, Rutul, and Archi that belong to the Lezgic language group have recently received writing systems and become newly literate, are spoken by ca. 50 000. It is very difficult to determine the number of speakers of other unwritten languages of the Lezgic group such as Budukh, Khinalug, Kryz, Udi, which are located outside Daghestan and almost totally assimilated to Azeri (Gamzatov 1995). Azeri was an inter-ethnic language of communication for Aguls, Rutuls, Tabasarans, and Tsakhurs until their languages got their official status. An interesting situation is within the Avar-Ando-Tsezic group which consist of fourteen languages: Andi, Akhwakh, Bagwalal, Botlikh, Godoberi, Karata, Tindi, Chamalal-Andic subgroup; Bezhta, Hinukh, Hunzib, Khwarshi, TsezTsezic subgroup, among which Avar is the only written and literary language, and the other thirteen languages have no writing tradition and are limited to domestic and family use. Of course, it is expected that multilingualism would be widespread in this region (i.e. knowledge of more than two or three languages). The pattern of multilingualism is the so-called ethnic-ethnicAvar-Russian multilingualism (which is, for example, widespread among Hinukh speakers: Hinukh-Tsez-Avar-Russian or Hinukh-Bezhta-Tsez-AvarRussian multilingualism). Another pattern of multilingualism is ethnic-AvarRussian trilingualism (which is widespread), against which the ethnic identity of the representative of an unwritten language is developed. Trilingualism is a standard way of communication for the Andi-Tsezic speakers. The speakers of the Andi language, for example, identify themselves primarily as Andi and yet call themselves Avar, as Avar language is the lingua franca in this group. Interestingly, outside the republic, the Andi speakers identify themselves as Daghestanians, and outside the country, they are simply Russians. This is the multi-ethnic hierarchy of ethnic and linguistic identity of the Daghestanian highlanders (Gamzatov 2005). The majority of the Andic languages’ speakers live in the most mountainous part of Daghestan, in the Andi Koysu river basin between Andi and Bogos ridges. The western border between Daghestan, Chechnya and Georgia coincides with the ethnic boundaries. Some Andic languages are also represented in the republic of Azerbaijan (for example, village Ahvahdere, where Akhvakh speakers live). The main language of communication there is Avar.

Unwritten Minority Languages of Daghestan

101

Since ancient times, the Tsez people (also known by the Georgian name Dido) inhabit the Western part of Daghestan and partially Georgia, keeping close ties with each other. Interestingly, the languages in the intra-group communication for the Tsezic speakers are Bezhta (which is used with Hunzib and Hinukh speakers) and Tsez (with Hinukh speakers); today Bezhta and Tsez still partially provide these functions. Due to the economic factors and geographical proximity, the Tsezic languages’ speakers were influenced by Georgia, and have been linked more closely with the Georgians, rather than with the Avar. Currently, Avar serves as the language of interethnic communication among the Tsezic peoples. Analysis of the Daghestanian languages clearly confirms that the declarations and decisions of language development in the given circumstances are impossible without taking into account unwritten languages, since there has been a large and prolific branch of linguistics dedicated to these languages. Moreover, unwritten languages of Daghestan have become not just subfield, but an independent field of modern Daghestanian linguistic and ethnographic research. As a result, various projects and special programs for the study of minority peoples, their genesis, historical and cultural past and present have been developed in the historical, philological and sociological departments of the Daghestan Scientific Center of the Russian Academy of Sciences. Serial publications of historical and ethnographic materials titled ‘Small peoples of Daghestan’ are ongoing. As a result, historical and ethnographic descriptions of all Andi-Tsez and Archi languages have been published. Collection of oral and poetic heritage of unwritten peoples and ethnic groups by folklorists have become a very successful endeavor. Much work has been done in the study of folk-art traditions and handicraft culture of these ethnic groups. The idea of preservation of small unwritten languages and protection of ethnic identities assumes a number of plans, projects and programs. The most important current issues for the Daghestanian, particularly small unwritten, languages are collection and documentation, comprehension and interpretation of the lexical data of each language. Therefore, the department of lexicography and lexicology of the Daghestan Scientific Center of the Russian Academy of Sciences carries out a multi-year program of making ethnicRussian bilingual dictionaries of unwritten languages. It is well known, that the only reliable way to gather the lexical material of these minority languages is to work among ethnic groups themselves, i.e. though direct fieldwork with their speakers. Lexical information is being documented during extensive fieldwork in the highland regions of Daghestan, and nowadays, almost all unwritten languages have their dictionary manuscripts, which include about seven to nine thousand of lexical units of up to eight hundred pages. Thirteen of eighteen dictionaries of this type have been published within two decades. This became

102

Alieva and Khalilov

possible due to the sponsorship of the Holland Science Organization, Max Planck Institute for Evolutionary Anthropology (Germany) and the Russian Humanitarian Foundation. The most realistic and effective way of documenting the language of any nation is the preparation, publication and distribution of its lexicon. The following dictionaries have been published within the program of ethnic-Russian dictionaries of unwritten languages: Budukh-Russian dictionary (1984) by Meylanova Bezhta-Russian dictionary (1995) by Khalilov Tsez-Russian dictionary (1999) by Khalilov Chamalal-Russian dictionary (1999) by Magomedova Karata-Russian dictionary (2001) by Magomedova and Khalidova Hunzib-Russian dictionary (2001) by Isakov and Khalilov Khinalug-Russian dictionary (2002) by Ganieva Tindi-Russian dictionary (2003) by Magomedova Bagvalal-Russian dictionary (2004) by Magomedova Hinukh-Russian dictionary (2005) by Khalilov and Isakov Godoberi-Russian dictionary (2006) by Saidova Akhvakh-Russian dictionary (2007) by Magomedova and Abdullaeva Botlikh-Russian dictionary (2012) by Saidova and Abusov Bezhta (2015) by Khalilov Dictionaries of unwritten languages such as Andi, Archi and Khwarshi are in different stages of preparation. Making the Kryz-Russian and Udi-Russian dictionary is in the long-term research plans of the Institute. According to academician Gamzatov (1995: 12), ‘doing a well-supplied dictionary work on unwritten languages of Daghestan, we proceed from the assumption that documentation of native and borrowed vocabulary of a language which has no writing system, its maintenance and preservation for future generations, of the still living experience of the human life and the ethnic way of thinking is certainly a task of huge scientific and humanitarian importance and at the same time in all respects humanitarian and extremely moral event, as well as a noble and rewarding one.’ We believe that, first of all, we need to call attention of the local state authorities and the public to the existing problems in this field in order to preserve the unwritten languages of Daghestan. Therefore, it is necessary to develop appropriate recommendations and take the most serious and urgent action.

Unwritten Minority Languages of Daghestan

103

1. There are no writing systems in the Andi-Tsezic language subgroup. These languages are oral and are used at home as vernaculars. The languages of school education are Russian and Avar, the latter is taught in village schools as the so-called rodnoj jazyk (“native language”). Additionally, one foreign language is taught. This means, that children are exposed to three new languages at school, whereas their native language is absent from the school curriculum. However, in the primary school (especially in the first grade) teachers use native languages as means of communication with the pupils, as children know neither Avar nor Russian yet. Design of writing systems for these languages is a necessary condition in order to preserve and develop ethnic minority groups. Introduction of writing would be valuable for many reasons, particularly given the challenges and difficulties of the primary education, which employs the Avar and Russian languages. 2. It is urgent to develop a project plan, which would specifically focus on the educational aspect of preservation and revival of the Daghestanian oral languages, providing a detailed statement of mother tongue acquisition within the family, at school, institutions; creation and publishing of alphabets, primers and other teaching guides, including radio and television programs; target training and retraining of teaching staff, etc. The implementation of the first stage of the proposed project will serve as a prerequisite for next programs; their humanistic importance and significance cannot be overestimated. 3. In present circumstances, there is no direct threat of disappearance of oral minority languages in Daghestan yet. However, due to the spread of different types of bilingualism, migration and other processes there are certain facts of loss of essential components of language activity like folklore, ‘blurring’ of the structural peculiarities of these languages, and so on. It should be noted also that Russian has a strong impact on these languages: high school students and younger generation communicate in Russian and the knowledge of their native languages is much more restricted. 4. In the last few decades, there has been an increasing interest among minority Ando-Tsezic ethnic groups toward their history, culture, as well as their mother tongue. The language of each nation is a natural and cultural heritage of all humankind. Therefore, our duty is to create necessary conditions for studying ethnic minority groups and their culture and languages (for instance, such study could be detailed historical-ethnographic and grammatical research (Kibrik et al. 1977; van den Berg, Helma, 1995; Radžabov 1999; Kibrik et al. 2001; Khalilova 2009; Forker 2010; Isakov & Khalilov 2012; Magomedova 2012; Comrie, Khalilov, & Khalilova 2015), collection and systematization of folklore and other text genres, different types of dictionaries, etc.).

104

Alieva and Khalilov

Therefore, the main and urgent task is collecting and documentation of the still existing linguistic, historical and cultural, folklore and ethnographic material of the indigenous ethnic groups in Daghestan, which previously had not been studied. Keeping all this wealth is a necessary prerequisite for the study of history, ethnography and language, familiarizing the younger generation to folk wisdom and transmitting the language and the whole bulk of folklore to future generations. A great amount of work in documenting the culture and the unwritten languages of Daghestan was done in the Department of Linguistics in Max Planck Institute for Evolutionary Anthropology (Leipzig, Germany) under the supervision and with the financial support of professor, Ph.D., former director of Department of Linguistics, Bernard Comrie. In particular, fundamental work on folklore of Tsezic people was made within a few years. Today the following books have been prepared and published: Tsez folklore by Abdulaev and Abdulaev and Khwarshi folklore by Karimova. These books include Khwarshi and Tsez tales, legends, stories, anecdotes, folk songs and ritual laments, as well as the sample texts of spoken language (dialogues and monologues). ‘Bezhta-Russian phraseological, folklore-ethnographic dictionary’, of 420 pages, has been prepared by Khalilov and first released in 2014. It includes rich folklore ethnographic material of the Bezhta people. The dictionary is the first work on lexicographical description of phraseology, including paremiology and folklore-ethnographic expressions of Bezhta. The dictionary includes phraseological units, proverbs and a full list of common set expressions. The dictionary lists independent Bezhta good wishes, teachings, swear words, curses, folk beliefs, omens, divinations, spells, riddles, rhymes, tongue twisters, funny expressions, as well as tribal (domestic) labels, signs in order to denote appurtenance. The dictionary can be a source for future theoretical research on phraseology, paremiology and, in particular, Daghestan-Russian phraseography and paremiography. Bezhta phraseological, paremiological and folklore-ethnographic units are diverse in their content and function. The Bezhta folklore emerged from the everyday observations of social and natural phenomena, some units are associated with mythology and real historical events, some of them are borrowed from related (Avar) or unrelated (Georgian and Russian) languages. A certain part of idiomatic, paremiological and folklore-ethnographic material of Bezhta dates back to the oriental languages, namely, Arabic. In general, this folklore work is the most valuable material for studying the life of Bezhta people in its historical aspect. As can be seen from a brief analysis, the study of languages of the Tsezic subgroup has been done relatively well. A great work has been done on compiling

Unwritten Minority Languages of Daghestan

105

and publishing ethnic-Russian dictionaries (except for Khwarshi-Russian dictionary, which is still in progress). Detailed grammars have been prepared and published by different linguists. For almost all Tsezic languages, folklore material (as well as ethnographic material) has been collected and systematized. It is achieved through a serious financial support of Holland Science Organization and Max Planck Institute for Evolutionary Anthropology and active and collaborative work of various scientists (Bernard Comrie, Maria Polinski, Madžid Khalilov, Ramazan Radžabov, Zaira Khalilova, Diana Forker, Arsen Abdulaev, Raisat Karimova, who are fully engaged in the study of the Tsezic languages). Such description of grammatical structure, collection and systematization of folklore and ethnographic material is still to be implemented with oral languages of the Andic subgroup of Avar-Andi-Tsezic languages. It mostly depends on the financial support of local and foreign research centers, and on the researchers involved in the study of the Andic languages. Another important point is worth of notice. Not all unwritten languages of Daghestan are studied and presented at the Institute of Language, Literature and Art, named after G. Tsadasa of the Daghestan Scientific Center RAS, which is the main research center of the study of the unwritten languages of Daghestan. For example, Bezhta, Godoberi and Chamalal are actively investigated: such scientists as Khalilov, Khalilova, Saidova, Magomedova and Alieva work on these minority languages. However, there are no specialists working on other languages of the Andi-Tsezic subgroup, as well as Archi, Budukh, Kryz, Khinalug and Udi. Therefore, the main task of the local research institute is recruitment and training of scientific personnel, who will be representatives of such minority languages and who will be interested and engaged in their study. It is crucial that in October 1991, Russia adopted the ‘Law on Languages of the Peoples of the Russian Federative Republic’, in which languages of all peoples living in the country are declared a national, historical and cultural heritage, and are guaranteed state protection. This law is supported by the Charter for Regional and Minority Languages, adopted a year later by the Council of Europe, in which the linguistic rights of national minorities are declared. However, it should be noted that the Republic of Daghestan is the only region of the Russian Federation, which has not yet adopted its ‘Law on Languages of the Peoples of the Republic of Daghestan’. The draft law, which is under discussion and preparation, should clearly specify certain measures for preservation and development of endangered unwritten languages of Daghestan. However, it is hard to believe that the attitude towards the languages will change radically and there will be a turning point in their study immediately after its adoption. The adoption of this law would guarantee their preservation and strengthen

106

Alieva and Khalilov

their position. The law could clearly specify the place and the role of the native languages in the life of the Daghestanian peoples. The serious social support is still needed for another category of Daghestanian languages, which are literary languages and languages that recently became written (they are about fourteen). These languages already have educational, scientific and artistic material, printed books, and mass media (radio and television). Thus, they can be considered to be “safe” or not endangered. However, nowadays there is a dangerous recurrence of narrowing the quantitative and qualitative areas of native language use, reduction of school programs of the native language and native literature. Many children, who grow up in the urban areas, do not speak their mother tongue. This happens under irreversible process of urbanization, globalization and under the influence of Russian, which is a means of inter-ethnic communication today. Native languages are being moved into the background of public attention. The change in the language situation, a revival of the ethnolinguistic life in the country is only possible with the help of reasoned, comprehensive and radical measures aimed for the future. At present, there is an ongoing mass ethno-linguistic assimilation: out of three million Daghestanian people almost two-third live in ethnic amalgamation. In addition, more than six hundred thousand live outside Daghestan. Such territorial dispersal, as well as ethnic intermarriages (almost every tenth Daghestanian family is exogamous, i.e. the marriage partners are representatives of different ethnic groups), lead to the state in which many traditional forms of culture are not associated with ethnic identity any more.

Bibliography Abdulaev, Arsen K. & Isa K. Abdullaev. 2010. Didojskij (cezskij) fol’klor. Makhachkala: Lotos. Comrie, Bernard, Khalilov, Madzhid and Khalilova, Zaira. 2015. Grammatika bežtinskogo jazyka: fonetika, morfologija, slovoobrazovanie (A grammar of Bezhta). Leipzig-Makhachkala: ALEF. Forker, Diana. 2013. A Grammar of Hinuq. Berlin: De Gruyter. Gamzatov, Gadži G. 1995. Bespis’mennyi, no Zhivoi, Real’nyi (Predislovie k serii «Bespis’mennye jazyki Dagestana») // Khalilov, Madžid Š. “Bežtinsko-russkij slovar’. Makhachkala: Nauka, p. 12. Gamzatov, Gadži G. 2005. Lingvističeskaja planeta Daghestan. Etnojazykovoj Aspekt Osvoenija. Moskva: Nauka, p. 63. Ganieva, Faida A. 2002. Khinalug-Russian Dictionary. Makhachkala: Nauka.

Unwritten Minority Languages of Daghestan

107

Isakov, Isak A. & Khalilov, Madžid Š. 2012. Gunzibskiy Jazyk (A grammar of Hunzib): Fonetika. Morfologiya. Slovoobrazovanie. Leksika. Teksty). Makhachkala: Nauka. Jazyki Narodov Rossii. Krasnaja Kniga. Enciklopedičeskij slovar’-spravočnik. Moskva: Academia, 2002. Karimova, Raisat Š. 2014. Fol’klor xvaršincev. Makhachkala: ALEF. Khalilov, Madžid Š. 1995. Bežtinkso-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Khalilov, Madžid Š. 1999. Cezsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Khalilov, Madžid Š. 2014. Bežtinsko-russkij frazeologičeskij, fol’klorno-etnografičeskij slovar’. Makhachkala: ALEF. Khalilov, Madžid Š. 2015. Slovar` bežtinskogo jazyka. Makhachkala: Nauka. Khalilov, Madžid Š. & Isakov, Isak. 2001. Gunzibsko-russkij slovar’. Moskva: Nauka. Khalilov, Madžid Š. & Isakov, Isak A. 2005. Ginuxsko-russkij slovar’. Makhachkala: Nauka. Khalilova, Zaira. 2009. A grammar of Khwarshi. Utrecht: LOT. Kibrik, Aleksandr E., Sandro V. Kodzasov, Irina P. Olovjannikova, and Džalil S. Samedov. (1977). Opyt strukturnogo opisanija arčinskogo jazyka. Tom 1. Leksika. Fonetika (in Russian). Moscow: Izdatel’stvo moskovskogo universiteta. Kibrik, Aleksandr E. et al. 2001. Bagvalinskij jazyk: Grammatika. Teksty. Slovar’ (A grammar of Bagwalal). Moskva: Nasledie. Magomedova, Patimat T. 1999. Čamalinsko-russkij slovar’. Makhachkala: Nauka. Magomedova, Patimat T. 2003. Tindinsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Magomedova, Patimat T. 2004. Bagvalinsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Magomedova, Patimat T. 2012. Tindinskij jazyk. Makhachkala: Nauka. Magomedova, Patimat T. & Rašidat Š. Xalidova. 2001. Karatinsko-russkij slovar’. SanktPeterburg: Scriptorium. Magomedova, Patimat T. & Abdullaeva, Indira I. 2007. Axvaxsko-russkij slovar’. Makhachkala: Nauka. Meilanova, Unejzat A. 1984. Buduxsko-russkij slovar’. Moskva: Nauka. Radžabov, Ramazan N. 1999. Sinstaksis cezskogo jazyka. Moskva: Academia. Saidova, Patimat A. 2006. Godoberinsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. Saidova, Patimat A. & Magomed G. Abusov. 2012. Botlixsko-russkij slovar’. Makhachkala: Institut JaLI DNC RAN. van den Berg, Helma. 1995. A Grammar of Hunzib (with texts and lexicon). Lincom Europa, München.

chapter 9

Report on the Fieldwork Studies of the Endangered Turkic Languages Yong-Sŏng Li

1

Introduction

Having realized the importance of the research and the documentation of the endangered Altaic languages, 8 scholars of the Altaic Society of Korea carried out at their own expense or on a personal level field researches at least 15 times from 1972 to 2002, on Dagur, Sibe, Uilta, spoken Manchu, Ewenki, Orochen, and Hezhe.1 All of these 8 scholars are male and graduated from the Department of Linguistics at Seoul National University. There is no Turkologist among them. These field researches are tabulated as follows:2

No.

Year

Language(s)

Participant(s)

1 2 3 4 5

1972 1972 1981 1986 1997

Dagur Sibe Dagur Uilta Spoken Manchu

6 7

1998 1999

Ewenki, Orochen Ewenki, Dagur

8 9 10

1999 1999 2000

Ewenki Sibe Ewenki, Dagur

11

2000

Ewenki

Baeg-in SEONG Baeg-in SEONG Baeg-in SEONG Ju-won KIM Jae-il KWON, Dong-ho KO, To-sang CHUNG, Gyu-dong YURN Ju-won KIM Baeg-in SEONG, Jae-il KWON, Ju-won KIM, Dong-ho KO, Gyu-dong YURN, Jae-mog SONG Ju-won KIM, Dong-ho KO Ju-won KIM Baeg-in SEONG, Jae-il KWON, Che-mun CHONG, Ju-won KIM, Jae-mog SONG Ju-won KIM

1  See Seong 2008: 24-33 and Yu 2014: 23-24. 2  See Seong 2008: 25.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_0�0

FIELDWORK STUDIES OF THE ENDANGERED TURKIC LANGUAGES No.

Year

Language(s)

Participant(s)

12 13 14 15

2000 2001 2001 2002

Ewenki Hezhe Dagur Sibe

Ju-won KIM Ju-won KIM, Dong-ho KO Baeg-in SEONG, Che-mun CHONG Dong-ho KO, Ju-won KIM

109

Members of this small group began to work to organize the ASK REAL (Altaic Society of Korea, Researches on Endangered Altaic Languages), the fullscale research and documentation project on endangered Altaic languages.3 Supported by the Korea Research Foundation Grant, the following two successive research projects were carried out:4 (1)

The 1st research project subject: Fieldwork Studies of Altaic Languages for Genealogy of Korean period: September 1, 2003-August 31, 2006 (2) The 2nd research project subject: Building Digital Archive of Altaic Languages for the Study of Genealogy of Korean period: July 1, 2006-June 30, 2009 These two research projects are called ASK REAL unofficially. The director of them was Ju-won KIM.5 These two research projects had the following two main goals:6 (1)

Accumulating the extensive data on Altaic languages for the future study on the genealogy of the Korean language; (2) Joining the world-wide efforts for the documentation of endangered languages.

3  See Yu 2014: 24. 4  See Seong 2008: 24, Kim 2008: 63-64, Kim 2011: 31-32, http://altaireal.snu.ac.kr/askreal_v25/, and http://www.cld-korea.org/eng/archives/archives_1.php. 5  See http://altaireal.snu.ac.kr/askreal_v25/ and http://altaireal.snu.ac.kr/askreal_v25/mem bers.html. 6  See Kim 2008: 31-34 and Yu 2014: 24.

110 2

Li

Field Researches

2.1 Determination of the Language or Dialect to Survey The ASK REAL classified the Altaic languages into 55 (34 Turkic, 10 Mongolic, and 11 Manchu-Tungusic) languages. The Altaic languages are distributed widely across Eurasia, and the speakers of those languages are under multilingual environment. And in most of the areas, the Altaic languages are not official languages. Therefore, in those countries, the Altaic languages are in danger of extinct.7 The most important criteria for the determination of the language or dialect to survey were “the degree of endangerment” and “the possibility of survey”. The first category is designated for those languages which are seriously endangered. The second category is for those languages whose users are easily and safely accessible.8 2.2 Organization of the Fieldwork Teams The ASK REAL had three fieldwork teams: one for Turkic, one for Mongolic, and one for Manchu-Tungusic. These three teams usually carried out the fieldworks at the same time and two together, if possible. The author was responsible for the Turkic team. A fieldwork team was composed of mainly four people: one team leader (professor) who controlled overall process of the field research, one specialist (with a doctor’s degree) for Manchu-Tungusic, Mongolic, or Turkic languages who did transcription; one questioner who was a fluent speaker of Khalkha Mongolian, Chinese, or Russian; one person for sound and video recording.9 In the beginning of the 1st research project, a fieldwork team was so. But, later the specialists (with a doctor’s degree) for the languages in question assumed also the role of team leader in the Turkic and Mongolic teams. The persons for sound and video recording were graduate students, the next generation of the discipline, of the Department of Linguistics at Seoul National University. The questioners were usually foreign graduate students of the same department.

7  See http://altaireal.snu.ac.kr/askreal_v25/altlang_main.html. 8  See Kim 2008: 136, Kim 2011: 51, and Yu 2014: 24. 9  See Kim 2008: 139, Kim 2011: 52, and Yu 2014: 24. For field research equipment, see Kim 2008: 143151, Kim 2011: 71-77, http://www.cld-korea.org/eng/fieldwork/fieldwork4_1.php, http://www .cld-korea.org/eng/fieldwork/fieldwork4_2.php, and http://www.cld-korea.org/eng/field work/fieldwork4_3.php.

FIELDWORK STUDIES OF THE ENDANGERED TURKIC LANGUAGES

111

According to the circumstances, the roles of the team members could overlap and the number of the team members increased or decreased one or two.10 2.3 Questionnaire Referred to Èwēnkèyǔ Jiǎnzhì (鄂温克语简志 “Brief record of the Ewenki language”) of China, the first questionnaire of the Altaic Society of Korea was prepared and used for Ewenki and Dagur in 2000.11 This questionnaire consisted of vocabulary, grammar and conversational components. It was revised and enlarged. In the version of September 2003, i.e. the first version for the first research project, it had 1,625 vocabulary items, 219 conversational sentences, and 182 grammatical sentences.12 Based on this version for China, the versions for Russia and Mongolia were prepared in 2003 and in 2005 respectively.13 These questionnaires were revised and enlarged continuously. Now they typically list around 2,750 vocabulary items, 340 conversational sentences, and 380 grammatical sentences.14 The Altaic Society of Korea has selected 24 lexical item classifications, seven grammatical categories, and 17 situations for conversational settings.15 These questionnaires are fit for the surveys which typically last 3-4 days for 6 hours a day.16 The informants were requested to answer twice for each entry. In addition, spontaneous speech was often gathered in an impromptu manner.17 The phonological portion is withheld from these questionnaires, because in many cases it is not possible to know the phoneme inventory due to the lack of description of phonemes for the language to survey. Research into the inventory is later performed using lexical data.18 When considering content placement, it is important to record information about the native speaker informant in the front of the questionnaire. The informant’s name, age, gender, ethnicity, family situation, information about

10  11  12  13  14  15  16  17  18 

See Kim 2008: 139. See Seong 2008: 30, Kim 2008: 153, and Kim 2011: 57. See http://altaireal.snu.ac.kr/askreal_v25/enqpdf/enq_0004.pdf and http://altaireal.snu .ac.kr/askreal_v25/field_research.html. See Kim 2008: 154, 159 and Kim 2011: 57. 2004 is given instead of 2005 due to an editorial error in Kim 2011: 57. See http://www.cld-korea.org/eng/fieldwork/fieldwork3_1.php. See Yu 2014: 25 and http://www.cld-korea.org/eng/fieldwork/fieldwork3_1.php. See Yu 2014: 25 and http://www.cld-korea.org/eng/fieldwork/fieldwork3_1.php. See http://www.cld-korea.org/eng/fieldwork/fieldwork3_2.php. See Kim 2008: 153 and http://www.cld-korea.org/eng/fieldwork/fieldwork3_2.php.

112

Li

language use, birthplace, residence history, and other relevant pieces of information should be recorded.19 2.4 Processing Collected Data We digitalized the collected data, and marked and extracted each item with Sony Sound Forge for audio data and Window Movie Maker for video data.20 These works were taken by the graduate research assistants in almost all cases.21 2.5 Publications The ASK REAL published three books for the consolidated reports of the project in Korean:22 (1)

The Altaic Society of Korea (ed.) (2006). Fieldwork Studies of Endangered Altaic Languages – For the Genealogical Study of Korean and the Preservation of Endangered Languages –. The Language and Cultural Studies Series 2. Paju: Taehaksa. [한국알타이학회 (엮음) 2006. 절멸 위 기의 알타이언어 현지 조사 – 한국어 계통 연구와 알타이언어 보존을 위하여 –. 알타이학회 언어 문화 연구 2. 파주: 태학사.] This book is a general report on the ASK REAL project done from September 2003 to February 2006 (Fieldwork Studies of Altaic Languages for Genealogy of Korean). (2) Kim, Ju-won, et al. 2008. Documentation of Endangered Altaic Languages. Paju: Taehaksa. [김주원 외. 2008. 사라져가는 알타이언어를 찾아서. 파 주: 태학사.] This book is a general report on the ASK REAL project done from September 2003 to August 2006 (Fieldwork Studies of Altaic Languages for Genealogy of Korean). (3) Kim, Ju-won, et al. 2011. Documentation of Altaic Languages for the Maintenance of Language Diversity. Paju: Taehaksa. [김주원 외. 2011. 언어 다양성 보존을 위한 알타이 언어 문서화. 파주: 태학사.] This book is a general report on the ASK REAL project done from July 2006 to June 2009 (Building Digital Archive of Altaic Languages for the Study of Genealogy of Korean). 19  20  21  22 

See Kim 2008: 156, Kim 2011: 58, and http://www.cld-korea.org/eng/fieldwork/fieldwork3_1.php. See http://www.cld-korea.org/eng/archives/archives_1.php. For detail, see Kim 2008: 176193, Kim 2011: 77-87, and Yu 2014: 27-28. See Kim 2008: 176. See Yu 2014: 28 and http://www.cld-korea.org/eng/fieldwork/fieldwork3_1.php.

FIELDWORK STUDIES OF THE ENDANGERED TURKIC LANGUAGES

113

The Members of the ASK REAL published seven descriptive grammars on some endangered Manchu-Tungusic, Mongolic, and Turkic languages in English as Altaic Languages Series:23 (1) (2) (3)

(4)

(5) (6) (7)

Kim, Ju-won, et al. 2008. Materials of Spoken Manchu. Altaic Languages Series 01. Seoul: Seoul National University Press. Yu, Won-soo, et al. 2008. A Study of the Tacheng dialect of the Dagur language. Altaic Languages Series 02. Seoul: Seoul National University Press. Li, Yong-Sŏng, et al. 2008. A Study of the Middle Chulym dialect of the Chulym language, Altaic Languages Series 03, Seoul: Seoul National University Press. Yu, Won-soo. A Study of the Mongol Khamnigan spoken in northeastern Mongolia. Altaic Languages Series 04. Seoul: Seoul National University Press. Li, Yong-Sŏng. 2011. A Study of Dolgan. Altaic Languages Series 05. Seoul: Seoul National University Press. Kim, Ju-won. 2011. A grammar of Ewen. Altaic Languages Series 06. Seoul: Seoul National University Press. Ko, Dong-ho & Gyu-dong Yurn. 2011. A Description of Najkhin Nanai. Altaic Languages Series 07. Seoul: Seoul National University Press. The first three books were selected among the excellent scholarly book (and the only books in a foreign language) for 2009 by the National Academy of Sciences of Korea.

The Members of the ASK REAL also published two monographs on the methodology of field work researches and documentation in Korean:24 (1)

Choi, Moon-jeong, et al. 2011. Linguistic Questionnaire for Investigation of Altaic Languages. Paju: Taehaksa. [최문정 외. 2011. 알타이언어 현지 조 사 질문지. 파주: 태학사.] (2) Choi, Woon-ho. 2011. The Practice of the Text Material Analysis and the Digital Archive Constructions for Altaic Languages. Paju: Taehaksa. [최운 호. 2011. 알타이언어 텍스트 자료의 분석과 디지털 아카이브 구축의 실 제. 파주: 태학사.] 23 

24 

See Yu 2014: 29-30. These books do not contain spontaneous speech. The 8th volume of this series was also published in 2015 as follows: Kim, Ju-won, Dong-ho Ko, Antonina Kile & Moon-jeong Choi. 2015. The Life and Rituals of the Nanai People. Altaic Languages Series 08. Seoul: Seoul National University Press. See Yu 2014: 29.

114

Li

Papers on the findings of the fieldwork studies have been published in domestic and foreign academic journals, or presented in conferences in Korean, or in English.25

3

Fieldwork Studies of the Endangered Turkic Languages

3.1 The 1st Research Project The original plan of the Altaic Society of Korea was to select two languages for each team and to survey each language three times for three years, i.e. once a year. The following languages were selected for the Turkic team: (1) Fuyü Kyrgyz in China, (2) the Oghuz dialects of Amu Darya region in Turkmenistan. In the regions where the endangered Altaic languages are spoken, it is usually very cold in winter, whereas the ground is so muddy that the contact with the native speakers is very difficult in summer. Therefore, in the beginning of the research project these three teams carried out fieldwork studies during the semester. The entire period of a fieldwork including departure and return was 7-10 days. So, it was not possible to survey many informants. The first fieldwork for Fuyü Kyrgyz was carried out during September 23-24, 2003 in the villages of Wujiazi and Qijiazi. The questionnaire had 1,625 vocabulary items, 219 conversational sentences, and 182 grammatical sentences. It was only possible to obtain the following data: 216 words and 22 conversational sentences from Wujiazi 305 words and 32 conversational sentences from Qijiazi. The data obtained was very small in number. Therefore, the second fieldwork for Fuyü Kyrgyz was carried out during January 15-16, 2004 in the village of Qijiazi. However, it was only possible to obtain 363 words and 43 conversational sentences. It was also impossible to obtain any data on grammatical sentences. Upon the result of the fieldworks on Fuyü Kyrgyz, the original plan of the research project was changed to survey as many Altaic (Turkic in particular) languages as possible. Each team tried to save expenses. Consequently it was possible to carry out more fieldwork studies than in the original plan. It was impossible to get a visa for Turkmenistan in Seoul at that time. Therefore, Prof. Dr. Mehmet Ölmez, one of the foreign collaborators, carried

25 

See Yu 2014: 30-31.

FIELDWORK STUDIES OF THE ENDANGERED TURKIC LANGUAGES

115

out with his colleagues the fieldwork studies on the Oghuz dialects of Amu Darya region in Turkmenistan during May 7-17, 2004. The following Turkic languages were surveyed in the 1st research project.26 (Information is given in the following order: name of the object language (or dialect); place of the fieldwork study; time period of the fieldwork study): (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18)

26 

Fuyü Kyrgyz; Wujiazi and Qijiazi, Fuyu County, Qiqihar, Heilongjiang Province, China; September 23-24, 2003 Fuyü Kyrgyz; Qijiazi, Fuyu County, Qiqihar, Heilongjiang Province, China; January 15-16, 2004 Shor (Mrass dialect); Myski, Kemerovo Province, Russia; April 20-23, 2004 Oghuz dialects of Amu Darya region; Ashgabat, Turkmenistan; May 12-17, 2004 Tuvan (Kök Monchak dialect); Aqqaba, Altay Prefecture, Xinjiang Uyghur Autonomous Region, China; October 21-22, 2004 Kazakh; Almaty, Kazakhstan; January 4-7, 2005 Yakut; Yakutsk, Sakha Republic, Russia; February 16-20, 2005 Chuvash; Cheboksary, Chuvash Republic, Russia; April 19-22, 2005 Tuvan (Kök Monchak dialect); Qanas and Aqqaba, Altay Prefecture, Xinjiang Uyghur Autonomous Region, China; April 28-May 4, 2005 Tuvan (Tsaatan dialect); Khatgal, Khövsgöl Aymag, Mongolia; June 23-26, 2005 West Yugur; Hongwansi, Sunan Yugur Autonomous County, Gansu Province, China: October 17-21, 2005 Gagauz in Kiev, Ukraine: February 5 and 17, 2006 Urum (Oghuz dialect); Mariupol’, Donetsk Province, Ukraine; February 7-11, 2006 Urum (Kypchak dialect); Mariupol’, Donetsk Province, Ukraine; February 8-9, 2006 Krymchak; Simferopol’, Autonomous Republic of Crimea, Ukraine; February 12-14, 2006 Karaim (Crimean dialect); Jevpatorija and Simferopol’, Autonomous Republic of Crimea, Ukraine; February 15-16, 2006 Chulym Tatar (Middle Chulym dialect); Tomsk, Tomsk Province, Russia: May 15-18, 2006 Siberian Tatar; Tomsk, Tomsk Province, Russia; May 18 and 20, 2006

Two co-researchers performed a preliminary survey of Altai and Khakas during February 17-19, 2004 in Novosibirsk, Russia. This survey is not included here.

116

Li

(19) Chulym Tatar (Lower Chulym dialect); Tomsk, Tomsk Province, Russia: May 19-21, 2006 (20) Tuvan (Uriankhai dialect) in Tsagaan-Üür, Khövsgöl Aymag, Mongolia: July 1-3, 2006 The author participated in all the fieldwork studies listed here except for 4-7. The informant of Lower Chulym was the last speaker of this dialect and died in 2011. Therefore, the linguistic data collected by the author is probably the last one for this dialect. 3.2 The 2nd Research Project The original plan of the Altaic Society of Korea was to select six languages for each team and to survey one language every six months. The following languages were selected for the Turkic team: (1) Gagauz, (2) Dolgan, (3) Bashkir, (4) Salar, (5) Kyrgyz, and (6) Tofa. All the fieldwork studies were planned to be carried out during the summer or winter vacation. The following Turkic languages were surveyed in the 2nd research project. (Information is given in the following order: name of the object language (or dialect); place of the fieldwork study; time period of the fieldwork study): (1) Salar; Xining, Qinghai Province, China; August 19-26, 2006 (2) Kyrgyz (Talas subdialect of Northern dialect); Bishkek, Kyrgyzstan; December 22-25, 2006 (3) Kyrgyz (Ičkilik subdialect of Southern dialect); Bishkek, Kyrgyzstan; December 28-30, 2006 and January 2, 2007 (4) Kyrgyz (Narïn subdialect of Northern dialect); Bishkek, Kyrgyzstan; January 2-3, 2007 (5) Kyrgyz (Čüy subdialect of Northern dialect); Bishkek, Kyrgyzstan; January 3-5, 2007 (6) Kyrgyz (Ïsïk-Köl subdialect of Northern dialect); Bishkek, Kyrgyzstan; January 6-8, 2007 (7) Dolgan; Yakutsk, Sakha Republic, Russia; January 30-February 5, 2007 (8) Yakut; Yakutsk, Sakha Republic, Russia; February 3-5, 2007 (9) Karaim (Trakai dialect); Trakai, Lithuania; July 17-28, 2007 (10) Bashkir (Ĕyĕk-Haqmar subdialect of Southern dialect); Ufa, Republic of Bashkortostan, Russia; August 8-10 and 21, 2007 (11) Bashkir (Urta subdialect of Southern dialect); Krasnousol’sk, Republic of Bashkortostan, Russia; August 11-13, 2007 (12) Bashkir (Zilim subsubdialect of Urta subdialect of Southern dialect); Krasnousol’sk, Republic of Bashkortostan, Russia; August 14, 2007

FIELDWORK STUDIES OF THE ENDANGERED TURKIC LANGUAGES

117

(13) Bashkir (Dim subdialect of Southern dialect); Rajevka, Republic of Bashkortostan, Russia; August 14-16, 2007 (14) Bashkir (Qïδïl subdialect of Eastern dialect); Asqar, Republic of Bashkortostan, Russia; August 19-20, 2007 (15) Chuvash; Ufa, Republic of Bashkortostan, Russia; August 22, 2007 (16) Tatar; Ufa, Republic of Bashkortostan, Russia; August 22, 2007 (17) Khakas (Sagay dialect); Abakan, Republic of Khakassia, Russia: August 2-4, 2008 (18) Khakas (Shor dialect); Chernogorsk / Abakan, Republic of Khakassia, Russia; August 5-7 / 15, 2008 (19) Khakas (Koibal subdialect of Kacha dialect); Abakan / Bejskij rajon selo Kojbaly / Abakan, Republic of Khakassia, Russia; August 7 / 12 / 13, 15-16, 2008 (20) Khakas (Kïzïl dialect); Abakan, Republic of Khakassia, Russia; August 8-11, 13, 16, 18, 2008 (21) Khakas (Beltir subdialect of Sagay dialect); Askizskij rajon selo Apchinajev / Abakan, Republic of Khakassia, Russia: August 12 / 13, 18, 2008 (22) Khakas (Kacha dialect); Abakan, Republic of Khakassia, Russia; August 13-16, 2008 (23) Altai (Telengit dialect); Gorno-Altajsk, Altai Republic, Russia; January 18-20/28, 2009 (24) Altai (Altai-kizhi dialect); Gorno-Altajsk, Altai Republic, Russia; January 20-21/24/26/27, 2009 (25) Altai (Kumandy dialect); Gorno-Altajsk, Altai Republic, Russia; January 22/25, 2009 (26) Altai (Chalkandu dialect); Gorno-Altajsk, Altai Republic, Russia; January 23-24, 2009 (27) Altai (Teleut dialect); Gorno-Altajsk, Altai Republic, Russia; January 25, 2009 (28) Altai (Tuba dialect); Gorno-Altajsk, Altai Republic, Russia; January 28, 2009 The author participated in all the fieldwork studies listed here except for 4. The Tofa-speaking region was not easily and safely accessible. It would be very expensive to go there. Moreover, it was informed that there was no fluent Tofa speaker. Therefore, Khakas was selected instead of Tofa. Altai was selected instead of Gagauz, because it was almost impossible to find informants at that time in winter.

118

Li

3.3 The Unrealized 3rd Research Project To carry out fieldwork studies on Altaic languages from July 1, 2009 to June 30, 2014, the members of the ASK REAL applied in March 2009 for a grant of the National Research Foundation of Korea (formerly the Korea Research Foundation) with the subject ‘Documentation of Endangered Altaic Languages – For the Study of Genealogy of Korean –’ (called ASK DEAL unofficially). The Altaic languages to survey were as follows: (1) Turkic: Fuyü Kyrgyz, Shor, Tuvan, Kumyk, Gagauz, Karaim, Urum (2) Mongolic: Mongolian, Buryat, Kangjia, Oirat-Kalmyk, Bonan (3) Manchu-Tungusic: Manchu, Uilta, Nanai, Ewenki, Ewen, Sibe, Orochi, Solon However, this research project was rejected by the examiners of the National Research Foundation of Korea on the pretext that the members of the ASK REAL did nothing in spite of the enormous support of the Korea Research Foundation. This is simply not true at all. As mentioned above, the members of the ASK REAL wrote a lot of books, articles, etc. Moreover, three books were among the excellent scholarly book (and the only books in a foreign language) for 2009 selected by the National Academy of Sciences of Korea. The real reason is the common practice of funding two successive research projects at most regardless of good result in Korea. From July 2009 to January 2010, Seoul National University supported the members of the ASK REAL to prepare a web site in English providing information on the ASK REAL. The result is the web site http://altaireal.snu.ac.kr/ askreal_v25. With funding from the National Research Foundation of Korea, Ju-won KIM and his colleagues carried out the “Languages and Culture of the Indigenous Peoples of the Amur River” Project from May 1, 2010 to April 30, 2013. The author participated in this project for the first six months. Sponsored by the Ministry of Culture, Sports and Tourism of Korea, Ju-won KIM and his colleagues including the author prepared the web site of the Center for Language Diversity (http://www.cld-korea.org/index.php) from October 2010 to March 2011. This site has also English version consisting of 4 sections of About, Fieldwork, Archives, and Bibliography. The materials of the sections Fieldwork, Archives, and Bibliography are related with the ASK REAL.

FIELDWORK STUDIES OF THE ENDANGERED TURKIC LANGUAGES

4

119

Conclusion

Using homogenous questionnaires and excellent audio/video equipment, the members of the Altaic Society of Korea carried out two successive research projects with the unofficial name ‘ASK REAL’ and published some books and papers in connection with these projects. The following points are to be mentioned: (1)

The period of fieldwork studies was not sufficient. So, it was not possible to survey many informants. However, it would cost too much to survey for a long time, for example, one or two months. Moreover, there was almost nobody to do so, especially among the graduate research assistants. (2) The number of informants was usually very small for each language. In many cases, there were only one or two informants. Therefore, the reliability of the collected data may be questionable. However, the members of the ASK REAL did their best in a given situation. (3) It was not possible for the members of the ASK REAL to continue their fieldwork studies on (endangered) Altaic languages due to the common practice of funding two successive research projects at most regardless of good result in Korea. Financial support is needed for those fieldwork studies. (4) The members of the ASK REAL digitalized the collected data, and marked and extracted each item with Sony Sound Forge for audio data and Window Movie Maker for video data. These works were taken by the graduate research assistants in almost all cases. However, the most of these data are not transcribed by the experts. Practically no one among the graduate students, the next generation of the discipline, wants to devout himself/herself to Altaic studies in Korea. Because there is little possibility for finding any chance in Altaic studies.

References The Altaic Society of Korea (ed.) (2006). Fieldwork Studies of Endangered Altaic Languages – For the Genealogical Study of Korean and the Preservation of Endangered Languages –. The Language and Cultural Studies Series 2. Paju: Taehaksa. [한국알 타이학회 (엮음) 2006. 절멸 위기의 알타이언어 현지 조사 – 한국어 계통 연구 와 알타이언어 보존을 위하여 –. 알타이학회 언어 문화 연구 2. 파주: 태학사.] Baskakov, Nikolaj A. (ed.) (1966). Jazyki narodov SSSR: 2. Tjurkskije jazyki. Moskva: Nauka.

120

Li

Choi, Moon-Jeong, et al. 2011. Linguistic Questionnaire for Investigation of Altaic Languages. Paju: Taehaksa. [최문정 외. 2011. 알타이언어 현지 조사 질문지. 파주: 태학사.] Choi, Woon-ho. 2011. The Practice of the Text Material Analysis and the Digital Archive Constructions for Altaic Languages. Paju: Taehaksa. [최운호. 2011. 알타이언어 텍스 트 자료의 분석과 디지털 아카이브 구축의 실제. 파주: 태학사.] Deny, Jean, et al. (eds.) (1959). Philologiae Turcicae Fundamenta; T. 1. Wiesbaden: Steiner. Grenoble, Lenore. 2008. Endangered Languages. In Peter K. Austin (ed.), One thousand languages: living, endangered, and lost, 214-235, Berkeley & Los Angeles: University of California Press. Johanson, Lars & Éva Á. Csató (eds.) (1998). The Turkic Languages. London & New York: Routledge. Kim, Ju-won, et al. 2008. Materials of Spoken Manchu. Altaic Languages Series 01. Seoul: Seoul National University Press. Kim, Ju-won, et al. 2008. Documentation of Endangered Altaic Languages. Paju: Taehaksa. [김주원 외. 2008. 사라져가는 알타이언어를 찾아서. 파주: 태학사.] Kim, Ju-won. 2011. A grammar of Ewen. Altaic Languages Series 06. Seoul: Seoul National University Press. Kim, Ju-won, et al. 2011. Documentation of Altaic Languages for the Maintenance of Language Diversity. Paju: Taehaksa. [김주원 외. 2011. 언어 다양성 보존을 위한 알 타이 언어 문서화. 파주: 태학사.] Ko, Dong-ho & Gyu-dong Yurn. 2011. A Description of Najkhin Nanai. Altaic Languages Series 07. Seoul: Seoul National University Press. Li, Yong-Sŏng. 2008. Endangered Turkic Languages – Preliminary Report on Fieldwork Studies –. Sibirische Studien 3/1. 1-25. Li, Yong-Sŏng, et al. 2008. A Study of the Middle Chulym dialect of the Chulym language, Altaic Languages Series 03, Seoul: Seoul National University Press. Li, Yong-Sŏng. 2011. A Study of Dolgan. Altaic Languages Series 05. Seoul: Seoul National University Press. Seong, Baeg-in, Ju-won Kim, Dong-ho Ko & Jae-il Kwon. 2010. Grammar and lexicon of Dagur and Ewenki Spoken in China. Daewoo Academic Series 597. Seoul: Acanet. [성백인, 김주원, 고동호 & 권재일. 2010. 중국의 다구르어와 어웡키어의 문 법∙어휘 연구. 대우학술총서 597. 서울: 아카넷.] Tekin, Talat. A New Classification of the Turkic Languages. 1991. Türk Dilleri Araştırmaları (Researches in Turkic Languages) 1. 5-18. Yu, Won-soo, et al. 2008. A Study of the Tacheng dialect of the Dagur language. Altaic Languages Series 02. Seoul: Seoul National University Press. Yu, Won-soo. A Study of the Mongol Khamnigan spoken in northeastern Mongolia. Altaic Languages Series 04. Seoul: Seoul National University Press.

FIELDWORK STUDIES OF THE ENDANGERED TURKIC LANGUAGES

121

Yu, Won-soo. 2014. Documentation of Altaic Languages. In Workshop on Language Documentation and Typology, 21-36. [This workshop was held at Seoul National University with the same name on 12-13 September, 2014 by the Department of Linguistics at Seoul National University.]

chapter 10

Empire, Lingua Franca, Vernacular: The Roots of Endangerment Nicholas Ostler

Endangered languages are one kind of outcome of a historical process which is fundamental to the story of humanity. Recounting this process illuminates the causes of language endangerment, and may give rise to a more realistic understanding of what it is, why it matters, and how any policy may be designed to affect it or reduce it. When we speak of language endangerment, we are not suggesting that the human language faculty itself is in danger: what is endangered is one or more specific languages, languages which stand to cease being used without being replaced by any other equally distinct languages. In short, and to generalize over many individual cases, the danger is that language diversity will be diminished: the number of languages in the world will decrease, and the average population of languages remaining will increase correspondingly. The first question to ask, then, in order to give the backgound to language endangerment, is: Where does language diversity come from? How was it created? The answer is that it is joint result of the development of the human language faculty (presumably a single event, approximately 100,000 years ago), and the spread of humanity, a single species which might be called Homo Loquens, to dominate ecological niches in every part of the earth. This process of spread, which is believed to have taken place between the eras 100,000 and 12,000 years ago,1 totally outpaced any ability of the human race to keep in touch with its constituent tribes as they moved across and out of Africa, and ultimately into every part of the earth’s land surface except Antarctica. Since human tribes were able to survive and flourish independently without mutual contact, there was no common constraint on how the language which each community spoke would develop. The process of language change will have accompanied this vast spread of humanity, and will have continued after all the niches had been filled with peoples, each making a living out of local conditions and resources. Inevitably, 1  Burenhult 2000.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_0��

Empire, Lingua Franca, Vernacular

123

they will have developed different vocabularies to support their very different life-styles; but they will also have perpetuated their languages by means that are as old as language itself: the young learn to speak and understand from the old. This process of learning requires imitation of sound in the context of stimulus, but also rational reconstruction of the system and the meanings it conveys. As a process, it does not give perfect copies of the elders’ speech, and there is always the potential, in every generation, for the language to be a little different from how the elders spoke it. And since there is no general opportunity for distant tribes to contact one another and communicate, this process of imperfect learning will have led – over 100,000 years – to all the diversity of language, in phonetics, vocabulary and structures that we now discover when we review the languages humanity still speaks and (where documented in writing) has spoken. This account of the origin of human language diversity is not just a rational reconstruction. The purported route of humanity round the world is confirmed by the pattern of gene mutations (Y chromosome for men,2 mitochondrial DNA for women),3 and also the gross positioning of detectable families of languages.4 The ability to use language must have been important in making this explosion of humanity’s homelands the success that it was. Language makes thoughts, ideas and plans discrete and potentially explicit: therefore, they become available for inspection, and for discussion with, and transmission to others who use the same code. Joint planning is possible, and so is co-ordination of action beyond the instinctive or habitual. Imagined possibilities can be explored, loyalties pledged, memories recalled. This makes a tribe with language much more powerful and effective in the game against nature. But this does not exhaust language as it was experienced. It was not just language as such: rather, it was a multitude of languages. Given the spread of people out beyond the bounds of regular contact and the iron law of imperfect transmission from generation to generation, different, incompatible codes of language came to be used. This had little ill-effect in these early days of human spread and residence: by its nature, the people who spoke languages other than yours were the people you were unlikely to meet, or to collaborate with. But even within a single code well within your understanding, regional groups would speak with perceivable variation, so that 2  https://en.wikipedia.org/wiki/Human_Y-chromosome_DNA_haplogroup#Major_Y-DNA_ haplogroups. 3  https://en.wikipedia.org/wiki/Human_mitochondrial_DNA_haplogroup. 4  http://emeld.org/workshop/2004/bibiko/bibiko-original.html.

124

Ostler

speech became a marker of tribe or kin-group: how much more so, when one encountered people who were incomprehensible. The way one spoke became a badge of belonging – both reinforcing solidarity among comrades, but also – when among enemies – providing a give-away shibboleth of one’s true loyalties. This badge function of language patterned with, and perhaps sometimes also against, other markers in dress and custom. But it also fed into the reserves of remembered lore which came with an upbringing in a particular language, so that each language grew its own traditions and literature, recall of which served to reinforce, in a different way, the conscious identity of speakers. However, this scattering of humanity into small, self-conscious and self-loyal groups ran up against another characteristic of the environment, sometime after the human species had taken effective occupation of all the landmasses of the earth. This was man’s ability to domesticate and intensify the food animals and food plants: pastoralism and agriculture. Rather than living off the land, seeking whatever might be hunted or gathered, some groups began to adapt their life-style to control of foodstuffs. This had various effects on their lived world. Importantly, they became more sedentary, staying with their fields and pastures, and growing their population in place, since the new methods meant more food could be produced with the new approach to applying human effort. Sedentary bases needed to be protected, both from weeds and wild beasts, but also from other human beings applying the world’s oldest labour-saving device: robbery. Protection and predation – defensive and aggressive warfare – became specialist occupations; and the utility of language in organizing large numbers of people was increasingly valued. Large groups came to be safer – and more powerful – than small ones; and the large groups began to define particular roles – ranks, functions and jobs – for individuals. It became possible to progress through a career, but even more to destine people for particular roles – including priests to preside over tradition, and rulers to preside over the whole structure of agriculural and pastoral society. All in all, the economic principle of falling marginal costs began to assert itself. When controlled units of production were large, the cost of adding to them was less than where the units were smaller: growth fed on itself, making bigger units natural winners – and conversely the traditional, small selfsufficient units, natural losers. At least, this was the case where resources were relatively easy to access: so the rich lands fell to larger and fiercer tribes, with marginal land being left to the traditional hunter-gatherers. Two of the problems of running large units is the threat of anonymous theft, and the difficulty of enforcing distributive contracts when they rely solely on the memory of (corruptible) participants and witnesses. A useful response to

Empire, Lingua Franca, Vernacular

125

this was the development of tally systems, and as their flexibility was appreciated, they changed into complete system writing, which could represent distinctly anything that could be said. Once this was available (defining another rank, function and job – the scribe) it soon became clear that there were many other uses to which this frozen speech could be put, not least correspondence which would make rulers’ decisions knowable, and enforceable, wherever their language was understood. Writing made possible the control of large-scale empires, the use of written orders and intelligence overcoming a natural limitation on an oral command structure. It also facilitated long-distance trade, by clarifying the content of orders and payments. But the effectiveness of these could only be realized if the recipients of messages could understand them. There was now a premium on understanding the language of the ruler or the merchant – in other words, the need for a common language between instructor and agent. People now had a motive to learn other languages, at least the language of those in power. This was the simplest case, but in fact, there was no pressing need for the language of communication to be the instructor’s own, as long as there was some language – a lingua franca – which was accessible both to instructor and agent. Hence the Achaemenid empire of the Persians, founded by Cyrus and destroyed by Alexander the Great, had for 400 years kept Persian as the language of the rulers, while imperial communications over their massively multilingual domain “from Hōdu to Kush” (i.e from India to Sudan) were conducted in Aramaic, which had been the native language of the (preceding) Babylonian empire. A further use that was discovered for written language during this period of the mid first millennium BC (called by German philosopher Karl Jaspers “the Axial Age”). This was the writing down of religious scriptures, making permanent and evident (just like the ruler’s commands) the tradition passed down by the priests. Since writing was highly explicit in its form, this also had the effect of sanctifying a particular style and dialect of language. Once written down, language does not change, even if it is read by successive genrations whose spoken languge is different, and the result was to give scriptural language an authority that surpassed any vernacular dialect. The institution of human empire, then, with its three clear prongs of military command (including taxation), commercial exchange and religious worship, all created a need for a large scale super-language – a lingua franca – with currency that extended beyond the range of any vernacular, and inevitably a language which could be written down fully and clearly. With this established, the tendencies which have since made for language endangerment were all in place. As the empires expanded, and became

126

Ostler

hardened by inherited loyalties, there was less space left for communities which relied only on their own traditional vernaculars; inevitably, these would only survive in less favoured sites, such as marshes, mountains and deserts. As the empires tended to use their power to monopolize wealth, there would be little left for the free peoples oustide. The hierarchy of command and wealth in the empires would be reflected in a class system with defined échelons of rank: but the outsiders who spoke their own language would be seen as beyond the pale. And with the establishment of religions, notably the missionary faiths Buddhism, Christianity and Islam, the denizens of the large empires would also see themselves as having virtue too that was superior to the “lesser breeds without the law”. Although the small tribes had shown the variability and innovations which once provided the wherewithal to spread humanity into every corner of the habitable earth, they were in effect disrespected and humbled when humanity had turned the world – and above all, its sources of food – into its own domain. Language endangerment, then, which can been defined as the pressure to give up on the use of the languages of smaller communities, is a tendency with a long tradition. Looking, for example, at the area of western Europe, which was largely conquered by Rome up to the first century AD, the 16 or more languages spoken in 500 BC had declined to some half dozen by 500 AD. Most had been replaced by Latin, the languages of a large-scale empire; but there was still a lot of use of Gothic, which had come in latterly with marauding tribes. The old languages that survived in Europe (Aquitanian aka Basque, and Celtic in Brythonic and Goidelic forms), together with Punic and Albanian slightly further afield, were all spoken in areas which had been on the edge of, or quite beyond, the land occupied by Rome. Their territories were also mountainous, making them of less economic value. In the east of the Mediterranean, the Greek language, which had spread as a lingua franca with Alexander’s empire (following on from a widespread Greek commercial network round the whole sea), proved quite able to resist the conquering might of Rome: this was a case of one lingua franca (Greek) ranged against another (Latin), and Greek won: it was already in place, already widely known to many Romans (in both elite and slave classes), and did not yield in prestige to Latin. It may also have benefited from the fact that there was relatively less settlement of Roman colonies in the eastern Mediterranean (though there was some – e.g. Corinth, sacked by Rome in 146 BC but re-settled only a century later, in 44 BC, and with many Romans. It would become the capital of the Roman province of Achaea, but still reverted to Greek.) Since we have represented the tendency to language consolidation – the flip-side of language endangerment – as almost a social necessity of human

Empire, Lingua Franca, Vernacular

127

development, at least in the period of increasing empires, it is worth pausing a moment to consider why it might be seen as regrettable. As we have painted it, language concentration, or at least the widespread use of lingua francas, is a side-effect of the requirements of imperial administration, which itself is an effect of falling marginal costs as economic units are organized on a larger and larger scale. If so, perhaps the increasing population of lingua-francas , and decline, even loss, of minority languages is inevitable. Part of the regret at language loss is undeniably sentiment, and empathy for what is lost as one human condition makes way for another: as the Roman poet Virgil puts it – his work the very pinnacle of all-conquering Latin – Sunt hic etiam sua praemia laudi;
 sunt lacrimae rerum et mentem mortalia tangunt.
 Solve metus; feret haec aliquam tibi fama salutem. Triumphs have their rewards here, but events provoke tears too, and the mortality of things touches us. Still, dissolve your fears: the renown of this will bring you some relief.5 These words are spoken by Virgil’s (Anatolian) hero Aeneas, once a Trojan, but soon to be Roman, as he contemplates a monument to his own city’s destruction. The fact that we may understand the vast forces that annihilate some of our valued possessions will not lessen the regret at their loss: and loss is just that – it does not necessarily lead to a corresponding, or even greater, gain further down the line. Look at the loss of the Celtic languages, once spoken all over the west of Europe and Iberia, but all overwhelmed in Gaul and in Spain by Latin. (In fact the modern Romance language are spoken almost exactly in the same places where once was Celtic.) An artefact like the Gundestrup Cauldron from Denmark, which shows a variety of strange details associated with the antlerbearing god Cernunnos, makes it evident that he had a highly complex myth: but we shall never know its content, since Greeks and Romans saw no reason to write it down – neither in their own languages, nor (much less) in its own Gaulish. In fact, the loss of a language, and the community that was identified by it, is not actually necessary in these cases, since the lingua-franca which has endangered it could just as well be acquired in bilingualism with it. What causes the loss is the lack of respect or affection for the old order – the

5  Virgil, Aeneid, i. 461-3; translation mine.

128

Ostler

class-based contempt for others’ way of life and web of old relationships, all articulated in a language which has fewer speakers. Indeed, the survival of diversity can be a source of strength, though it is not typically seen as such by those who have no concern for endangerment. Andaman Islanders showed amazing savoir faire, and lack of casualties, when confronted with the vast tsunami that devastated coasts all round the Indian Ocean; and in both world wars, the USA benefited from having access to American Indian language speakers to use as “code-talkers” – since their languages (such as Navajo, Choctaw and Cherokee) were effectively undecipherable to intelligence officers in enemy powers such as Germany and Japan. It takes an extreme situation to bring out the salutary value of diversity and flexibility. But it is evident to those who value their own traditions that something important stands to be lost. There is in fact a curious irony about knowledge and the value of diversity. Typically, those who represent, by their language and their lives, a minority culture will not see its value as stemming from its diversity: they just want to hold on to what has been their own. It is the representatives of empire – often native speakers of the lingua franca – who are in a position to be conscious of diversity as a value in itself, indeed a form of cultural wealth. We are presently in Ardahan, a city now in the extreme north-east of Turkey, but in a land which has previously formed part of various language zones – and indeed empires. Together with the neighbouring Caucasus, it is a vivid example of how not only imperial re-organization, but also the movements of peoples can radically alter the linguistic profile of a territory. They show that the general secular trends towards endangerment of minority languages which we have identified as diminishing linguistic diversity in the modern word, may be difficult to discern in the language history of a specific region. Imperial lingua-francas can cloak the existence of the common people’s vernaculars in the historical record; yet this situation may be unstable in the longer term. Such lingua-francas can ultimately either be eliminated in competition with some other elite language, allowing the vernacular to surge up as the dominant language again, or else be replaced by another invading vernacular. All of this has happened around Ardahan. When we look as far back as we can for the language of this region, we get back to the early first millennium BC. In that era, Ardahan lay in the northern reaches of the kingdom of Urartu (a name which is also seen in the name of Mt Ararat, not far to the south-east). This people the Greek historian Herodotus calls the Alarodioi, as one of many tribes ranged against the Greeks in Xerxes’ army of 480 BC – their

129

Empire, Lingua Franca, Vernacular

Russia

Georgia Tbilisi Trabzon Rize

Ordu

Kars

Azerbaijan

Armenia

Erzurum

Iḡdir

Turkey Van

Malatya

Urfa

Tabriz Diyarbakir Siirt Batman Mardin

Mosul Erbil

Syria map 1

Iran

Hamedan

Iraq Kenmanshah Ardahan in west Asia today. source: After Google Maps data.

last appearance in recorded history – although the army of Orondas, says Xenophon (5th century), besides Armenians, included Mards and Khaldian mercenaries – who would probably be Kurds and Urartians.) Urartu was centred on Lake Van, and after successfully opposing the Assyrian empire to its south in the 8th century, seems to have become a vassal state ca 705 BC, through agreement with the Assyrian king Sennacherib. This seems to have led to a period of peace and prosperity, which gradually degenerated through increasing attacks from the west and south, from marauding Cimmerians, Scythians and Medes, all of whom would have been speaking Iranian languages of the Indo-European family. These people are perhaps the ancestors of the (equally Iranian) Kurds who dwell in a large area extending

130

Ostler

36○

38○

40○

42○

44○

46○

48○

50○ Scale 0 50

Terek

C

A

0

U

100 km. 50

100 ml.

C S 42○

k)

Irma

O N T (Keik I C it)

M T S .

rates

Era

Euph (Firat)

U

U

N

Euphrates

Jey ha n GREAT SEA OF AMURRU

N

KARKEMIS

HARRAN

ALALAH

UGARIT 36○

map 2

Sea ANIASTANIA iri Na per n) Up (Va TUSHPAH (VAN)

T A Tigris

I

E

40○

S

Tig ris ISANA QULHU 42○

38○

Grea t Za b ARBAILU (ERBIL) IMGUR-ENLIL 44○

Sahand

SIBAR

MEYSHTA

NINIVE

SHNGAR RATSAPA

38○

N

ARDINI/ MUSASIR

P I R

TIL BARSIB

ARPAD

Erase (Araxes)

M. Sipan

fid

A

E M

U

40○

Se

M

S

ia) (Urm

A

O

R T

a iri Se

Pu ra

na

de

Seyh an

O

N

L.Gagham (Sevam)

ARZASKU

NIHIRIANI

M

MARASH . S A S S Y T R I A N M KUMAHA S

I

EREBOUNI M. Ararat

R A

l cani Ara

MALATYA U S

A

S E A

R

ARGISHTINHINELE

) axes (Ar se

r Na Lowe

U

T

42○

M. Aragatz

KIA A

N

Kur

(Hal DOUCKAMMA

T

O Ala U zon i

Ku r

ak) ys Kizil Irm

SAMAL

M

G R E A T S E A )

40○

36○

S Kur

U P P E R S E A (W E S T E R N G R E AT S E A ) (BLACK SEA)

P

38○

U

GANI

(Kizil

E A S T E R N ( C A S P I A N

A

36○

IZIRTU 46○

48○

The Kingdom of Urartu, 9th-6th centuries BC. source: Licensed under CC BY-SA 3.0 via Commons – https://commons .wikimedia.org/wiki/File.

east and west to the south of lake Van. It finally yielded to the Mede Cyaxares in 612 BC, but shortly afterwards it was incorporated by Cyrus into the Persian (Achaemenian) empire. Urartian is a language attested in cuneiform inscriptions which have survived to the present day (also called Haldian, after the empire’s ruling deity H̬ aldi, a warrior god who usually depicted standing on a lion), and is related both to Hurrian to its west, and Lezgian and Avar to the north in the Caucasus. But it did not survive the collapse of the Urartian state. Instead, the Armenians, with their Indo-European language separate from Iranian, appear to have replaced it. The first Armenian-speaking dynasty was called Orontid (properly Eruand, later Yervanduni). It is clear, however, that there was considerable bilingualism between Armenians and Iranians in this area, since the vocabulary of the Armenian language (called Hai by its speakers) is full of Iranian loans. This replacement seems comparable to the emergence of English from under French in 14th-century England, after that country recovered from the demographic (and social) shock of the Black Death. Perhaps because there was

131

Empire, Lingua Franca, Vernacular

little social mobility before the shock, it had not been possible for the elite lingua-franca (respectively Urartian or French) to permeate into lower orders; but when the aristocracy were shattered, the vernacular could emerge. In fact, there is no literacy to attest the actual language being spoken until two more conquests had passed over and around the Armenians: first the Macedonian Greek conquest from the late 4th century (led by Alexander, reinforced by his general Seleucus, and under Antiochus III finally deposing the Orontids in 212 BC); then the Roman conquest (the general Lucullus defeated Pontus and Armenia at the battle of Tigranocerta in 66 BC). One longlasting effect of the Roman domination of Anatolia was the penetration of Christianity. This reached Armenia at the end of the 3rd century AD (hitherto largely a Zoroastrian country), and the country has the distinction of being the first officially Christian state. This led after a century (in 405) to the invention of the Armenian alphabet by Mesrop Mashtots, and the translation of the scriptures and liturgy into Armenian. 36○

38○

40○

42○

42○

S A R M A T I A C A U S A KOL KH I S S NAKALAKEVI KUTATISI U S IBERIA

PHASIS

B L A C K

E

SEBASTEIA

M

CAESAREA

E

R

U

S

M

Tig

ri s

Eu ph rat es

n

Gr

NISIBIS

38○

P E R S I A N

ea

t Za

AFRAZA GANJAK

S

A

S

uta

b

E M P I R E SHIZ

M

NAL S E

N

O

SHNGAR

URBILLUM 44○

ZANJAN 36○

S.

42○

T

INTER

I

Kap

40○

PAYTAKARAN

NAKHICHEVAN

R

map 3

38○

A

Tig ri s

G

36○

T

40○

A r a xe s

A r a xe s

A

36○

N

LPNATS

k HER uni B z nVa n ) ( VAN/SHAMIRAMARKERT

MARTYROPOLIS

U

N S

am

ZARISHAT

O

T

A I I A KAPALAK

A R M E N I A

????

GERMANIKEIA AMIDA ANAZARBA HIERAPOLIS BIRTHA MOPSOUESTIA ZEUGMA TELA EDESSA KYRRHOS SELEUKIA APAMEIA ANTHEMOUSIA EUROPOS ???? KARRHAI RHOSOS HIERAPOLIS ANTIOKHIA BEROIA BARBELISOS KHALKIS

TCHOR

N

DVIN VAGHARSHAPAT ARTASHAT

G R E A T E R

MELITENE

U

YERVANDASHAT

42○

Z

????

A

IR

tes Euphra

gh

ZAREHAVAN

????

T 38○

P

es

N

100 ml.

S E A

A ra x

U

??

N

? ???

Ge

A

50

???

A

B

100 km.

0

?

M

URBNISI MTSKHETA UJARMA M A ????O RUSTAVI ? L

50○ Scale 0 50

??

O

?

48○

?????

??

40○

S E A

TRAPEZOUS NEO CAESAREA P O S N T I C M O U N T A I N ????? NIKOPOLIS

AMASEIA

R

?

? ??

AMINOS

?

????

46○

I A N C A S P

PITYOUS SEBASTOUPOLIS

44○

46○

Greater Armenia in 5th century AD. source: Licensed under CC BY-SA 3.0 via Commons – https://commons .wikimedia.org/wiki/File:Armenian4thcenturies.gif#/media/ File:Armenian4thcenturies.gif.

48○

132

Ostler domain of the alans

Pontus Euxinus

IA Phasis GTephelis Sinopa R Daruband O Amysos Constantinople Heraclea PA P H L A G O N I A GE Trapezunt Gandza Nicomedia Kabala Artanos CH ALD I A Dvinis KH Nicaea AC Ancyra Amasya Anis H Theodosyopolis EN S H I R V A N Pergamon Sebastea A R M E NI A TA RO N Dorylaeon Mantsikhert lesbos Van n Smyrna A N A T O L I A a ak chios N ur p Iconion A s S A S U N a I IJA v C Siss A I Attaleia Edessa Hayas RB IL C Tarsos A S S Y R I A ZE Antiochia A rhodus Mosul TRACE

EA

ST

ROM

AN EM PIRE

crete

RUS

SY R I A

CYP

ABBASID CALIPHATE & OTHER MUSLIM STATES BY 1060

map 4

CONQUERED BY SELJUK TURKS BY 1081

TERRITORY OF EAST ROMAN EMPIRE AND GEORGIA AFTER 1081

J A B E L

MODERN GEORGIA

Armenia and Seljuq expansion 11th century. Source: AFTER http://www.conflicts.rem33.com/images/Georgia/geor_ histr%202.htm.

The Arsacids (a Persian dynasty) had gained control of Armenia in the first century AD, and may have reinstated elite use of Persian in this region, but this appears to have had little effect on continuing vernacular use of Armenian. Politically and religiously, Armenia was even autonomous from Persia after 451, hence effectively a monolingual country. This would only change in the 11th century. First the Byzantine Greeks invaded. This would have had little effect on language use, except in court circles. But then the Seljuq Turks (under Alp Arslan) invaded Armenia, as a first fruit of the Turkish invasion of Anatolia more generally, enabled by their victory at Manzikert in 1071. This invasion actually involved long-term settlement, hence penetration of Anatolia by the Oǧuz Turkish language, side by side with Armenian, but not as an elite lingua-franca, since for the Turks this role would be played by Persian (or the mixture of Persian and Turkish known as Lisân-ı Osmânî – “Ottoman”). So far, our story has extended over at most 2000 years. But for most of the following millennium, the pace of change slows down. Armenian and Turkish co-existed (with Greek) as vernacular languages within the domain of the Ottoman empire. The latest accident of history (in the last century) is to segregate languages much more into discrete states, so that Ardahan is capital of a Turkish-speaking

Empire, Lingua Franca, Vernacular

133

area, while Armenian is spoken in the state of Armenia, capital Yerevan (which some conjecture to be related to Yervand, the modern form of Oronta). Meanwhile to the north, in the mountainous Caucasus, Kartvelian languages unrelated to Armenian or Turkish, such as Georgian, Mingrelian and Laz, have largely maintained their age-old locations. In the Ardahan region over all these millennia, the languages which have been endangered – indeed which have ultimately disappeared – have not been the languages of down-trodden lower orders, but the reverse, the languages of elites. These are the languages which – ironically enough tend to survive in the written record, because they are used for royal inscriptions: Urartian from the 10th to 6th centuries BC, Old Persian and Imperial Aramaic from the 6th to 3rd, Greek from the 3rd BC to 10th AD, Persian and Osmani from the 10th to the 20th AD. But actual speakers and users have ultimately deserted them. It is the languages of the people, Kurdish, Armenian and Oǧuz Turkish which have shown permanence: Kurdish and Armenian throughout the 3000 years we surveyed, (though the presence of Armenian in eastern and central Anatolia have recently been restricted artificially); and Turkish ever since its arrival with Seljuq settlers in the 11th century. How has this divergence been possible, from the general human pattern outlined at the beginning of this paper? Are not dominant classes more capable of hanging on to, and indeed spreading use of, their languages? The answer seems to be that this only applies when dominant urban upper classes are in contest with an urban proletariat. By contrast, farming, i.e. settled cultivation of the land, is more effective in perpetuating a language community not only than the hunter-gather lifestyle which historically it replaced, but also than the elite urban cultures which grow up, ultimately supported by the surplus of food created by farming. It was the urbanized classes, upper and lower, which were at risk of being “driven out” when their cities are conquered and put under new rulers. As an Irish immigrant put it in Margaret Mitchell’s popular 1936 novel Gone with the Wind, “Land is the only thing in the world that amounts to anything, . . . for ‘tis the only thing in this world that lasts, and don’t you be forgetting it!” Empires, then, do have an effect of oppressing, and ultimately suppressing, minority languages within their domains, provided that there is a lively prospect of recruitment of members of lower orders into richer and more dominant classes. Demeaned classes may try to ape their presumed betters, leaving their own languages behind. If this kind of mobility is ruled out, then there

134

Ostler

may be a polarization of languages, with a serf or peasant class continuing to speak their language, and never finding means of promotion: this seems to have been the position in mediaeval England, but also ancient Anatolia. The language of the elite, though, may itself be under threat when power politics leads to an overthrow of the elite by others; and in such changes the language of the rural food producers is likely to be immune to change.

Reference Burenhult, Göran. 2000. Die ersten Menschen, Augsburg: Weltbild Verlag.

chapter 11

Endangered Turkic Languages from China Mehmet Ölmez

The Caucasus is perhaps the foremost region in the world in terms of endangered languages. When we analyze the linguistic connection between Turkey and the Caucasus, among the first languages that stand out are the Laz language, and Hamshin Armenian. What I shall be focusing on in this study however, takes us farther east to the Altai Mountains; a region that bears striking similarities in its topography to Ardahan and the Caucasus. A number of Chinese sources name Jin Shan (金山 “Gold Mountains”) as the region Turkic peoples settled in.1 On that note, let us mention the peoples that first appeared in this area and belong to the Turkic ethno-linguistic group. Those that remain to this day within the borders of China are: 1. Uyghur 维吾尔, 2. Kazakh 哈萨克, 3. Kirghiz 柯尔克孜, 4. Salïr 撒拉, 5. Tatar 塔塔尔, 6. Tuvan 图瓦, 7. Yellow Uyghur 裕固, 8. Uzbek 乌孜别克, 9. Fuyu Kirghiz; Kïrkïs 柯尔克孜. Only six group of peoples among those listed continue to effectively use their native language. Of these languages, the speakers of Uzbek have recently given up speaking their own language and started speaking Uyghur. Consequently, it is hard to consider Uzbek as one of the Turkic languages spoken in China. The same applies for Tatar and Fu-yü Kirghiz as well. Middle and Old generations speak Mongolian (“Ölöt Mongolian”) while the young generation speaks Chinese. The related details will be given below. In China, more than one language belonging to the Turkic language group are spoken. And notably, we see isolated Turkic languages spoken in China. In 1970s, nine Turkic languages existed in China. Unlike the above mentioned classification, these languages can be listed based on population/speaker as follows: Uyghur, Kazakh, Kirghiz, Salïr, Uzbek, Yellow Uyghur, Tatar, Tuvan and Fü-yu Kirghiz. Although not up-to-date, the demographics of China can be viewed on Chinese articles.2 The list shows the languages and the number 1  From 通典 tong dian (http://ctext.org/tongdian/zhs?searchu= 代居金山). 2  http://zhidao.baidu.com/question/97856583.html?si=7. Bold numerals belongs to the website.

© koninklijke brill nv, leiden, ���7 | doi ��.��63/97890043�8693_0��

136

Ölmez

of people that speak them: 6. 维 吾 尔族 The Uyghurs 8.399.393; 17. 哈 萨 克 族 The Kazakhs 1.250.458; 32. 柯尔克孜族 The Kirghiz 160.823; 36. 撒拉族 The Salïrs 104.503; 48. 裕固族 The Yellow Uyghurs (speakers of both Mongolian and Turkic) 13.719; 49. 乌孜别克族 The Uzbeks 12.370; 53. 塔塔尔族 The Tatars 4.890, respectively, on the basis of population density. The Fuyu Kirghiz and the Tuvans are not included in official censuses since they are not known or recognized as a separate community during the minority identification works executed in China. In this paper, Uyghur, Kazakh and Kirghiz which are spoken in Xinjiang and thus known closely by Turcologists; Uzbek, which has become almost Uyghur and Tatar, will not be dealt with in detail. Apart from that, another Turkic community whose language has not yet been studied sufficiently is the Tarbagatay Kirghiz living in Xinjiang. Their ethnology has been studied by M. Čertïkov3 and their language by Erkin Awgaly.4 The information I provided below is based on my previously published studies, related sources, and personal studies in the field, as well as the Tubi̇tak project numbered 108K413 which is connected with this study. My library, which I use for a detailed study to be included in a book in the future, has not been fully explored as of yet and thus a part of the important sources regarding the subject has not been used in this paper. Both the book of the project and separate books about various languages included in the project will be published soon. I would like to extend my thanks to Batubayr from Ürümči Pedagogy University, Toolay from Bowuršin & Kanas, Arslan from Sunen / Gansu, Yunus from Xining / Qinghai, TÜBİTAK and Korea Research Institute for the help and support they provided me during my studies in the field. Now let me say just a few words on already extinct or almost extinct Turkic languages from Xinjiang.

Uzbek The Uzbeks living in China do not have their own autonomous prefecture. Having been scattered to the north and south of Tian Shan 天山 (Uyg. Tengri Tagh; literally “God’s Mountains”) in Xinjiang Uyghur Autonomous Region, most Uzbeks live in the cities. The Uzbek population is centered in İli, Kashgar, Yarkant, Urumchi, Aksu, Uchturpan, Chöchek and Karghilik. The population is 3  M. A. Čertïkov, Tarbagataiskie Kırgızı. 4  According to private communication about topic with Erkin Awghaly after his field research.

Endangered Turkic Languages From China

137

about 20,000 (Özbek Edebiyati Tarihi, 2005, p. 9). In China, a written language or an alphabet peculiar to the Uzbeks do not exist. As mentioned above, the Uzbeks use Uyghur as a written language. Adalaiti Abdulla’s study will give us more detailed information (Abdulla 2013).

Tatar According to the 2006 census, the Tatar population in China is around 5400. Most of the Tatars live in Xinjiang Uyghur Autonomous Region, densely in Urumchi, Gulja and Chöchek. Like the Uzbeks, the Tatars do not have their own written language or alphabet. Ersin Teres’ study will give us new information (Teres 2011). Within the scope of my research, let us now analyze the four languages three of which are spoken outside of Xinjiang, and one in Xinjiang.

Fuyu Kirghiz The Fuyu Kirghiz are a turkic community whose language and society we have the least information and material on. The earliest information we have about the subject relies on the collected works and studies of Hu Zhenhua, faculty member of Kirghiz/Kazakh Language and Literature Department in Zhongyang Minzu Daxue (Minzu University of China; formerly known as Central University for Nationalities). The information I will be providing here, however, is mainly from the field work that the faculty members of Philology Department in Seoul University, my colleagues from Korea Altay Research Institute and I did in September, 2003. The Fuyu Kirghiz live in the villages of Fuyu County, which is under the administration of Qiqihaer City in China’s Heilongjiang (widely known as Manchuria) province. Named after the affluents of the Amur River (Heilongjiang “Black Dragon River”), Fuyu is a residential area of 30,000 people and 1 hour away from Qiqihaer. Although the Fuyu Kirghiz are scattered around a few villages, they mainly live in two villages. The research group I was a part of made a compilation in the villages of Wujiazi and Qijiazi in the September of 2003 (my studies on the Fuyu Kirghiz people and Fuyu Kirghiz was carried out within the frame of “Korea Research Institute” project numbered KRF-2003-072-AL2002). Since Hu Zhenhua visited this region, the number of speakers of Fuyu Kirghiz have decreased and as of 2000s no native speakers remain. Today, the

138

Ölmez

Fuyu Kirghiz use Ölöt Mongolian and Chinese in their daily lives. During the compilation work we did among the Fuyu Kirghiz, whose number reaches up to 1000, we could talk to only seven people. Two of them were over 50 and the others were over 70. Our best source who was over 70 stated that she used their native tongue when she was 18. The population of Wujiazi and Qijiazi villages consists of different communities and the Fuyu Kirghiz make up nearly half of it. Despite all our efforts, it was not possible to fill in the half of 3000 questionnaires. However, even the limited material compiled is enough to show that this language is a branch of Khakas and Shor. After all of the compilations are discussed and published, useful results will come up for South Siberian Turkic languages in particular and for all Turkic languages in general. The first studies related to the data we collected were published in 2007 and 2010/2011 (see. Li, Ölmez and Kim, 2007 and 2010/2011). With respect to the importance of Fuyu Kirghiz, we can safely conclude from our work based on limited language material that Fuyu Kirghiz is a Turkic language that falls into the same category as South Siberian Turkic languages and basically modern Khakas written language. On the basis of the data related to that language, we can reach several results about the phonetic properties and vocabulary of South Siberian Turkic languages and modern Turkic languages. First of all, the general belief among the local people is that they came to that region after the second half of the 18th century. In fact, after the Manchus started to rule over China, the Manchus had several Mongol clans settle (mainly Dagur Mongols) in the region. Among those clans coming from Southern Siberia, there was the Fuyu Kirghiz who today call themselves Kirghiz and speak the same Turkic language with today’s Khakas. When Fuyu Kirghiz is compared with Old Turkic and modern Turkic languages in the light of previous studies, a few characteristic features can be listed as follows: 1.

2. 3.

Old Turkic and Common Turkish sound y- becomes c-: OT yap- “to cover, close” = ǰap-; OT yay “summer” = ǰay; OT yė- “to eat” = ǰe-; OT yürek “heart” = ǰürüh; Old Turkic and Common Turkish d sound becomes z: uzï- “to sleep”, ɢizin“to get dressed”, güzi “groom”; Final -g of a word becomes -h: OT tag “mountain” = tah, OT sarïg “yellow” = sarïh, OT ulug “great, big” = uluh.

In order to make a comparison with modern Turkic languages, I can give a few short sentences from our field reseach result:

Endangered Turkic Languages From China

139

“where do you live?” “(You) Smoke (a cigarette) too!” “it has been three days since I came here.” aragï ïzïrtïr “(alcoholic) drinks make you drunk.” min at mïlbat gïlçïk ~ mïn at mïl gïlçïk “I rode a horse and came here” min bozïn aldïr “I am taking (it) on my own” sïn miniŋ piçeŋ buzïh “you are a little older than me” bïs sïnïŋ olïh “we older than you” sïn kaya tursïn sin mï tart! men mïnda gïlgïn uşgun boltïr

What we see here as tart- (to smoke) corresponds to that of most Turkic languages. Likewise, aragï “drink” has the same root as Turkish rakï (a kind of alcoholic beverage), now common in all Altaic languages as a borrowing from Arabic through Persian (?). ïzïrt- is related to Old and Middle Turkish esür-, esürt- “to become drunk” (cf. Anatolian dialects: esirik, esirikli “fractious; little bit crazy”). In the fifth sentence, gïlçïk is not anything else but Old Turkic < kelyük “has come” (for details see. Ölmez, 2007). boz is another form of Old Turkic bod “himself”. buzïh is Fuyu Kirghiz form of Old Turkic bedük “big”. olïh is the word ulug in Old Turkic. Regarding these examples, you can see Ölmez, 2001; Ölmez 2006a, b and c. The first work on the Fuyu Kirghiz and their social life was prepared by Gundula Salk and Mambet Turdi. The last and detailed work belongs to Mixail Čertïkov (Čertïkov, 2008a and 2008b). Below I provide the words “moon”, “day” and some number names in order to show the phonetic similarities or differences between Fuyu Kirghiz and Turkic languages and Turkish.

Salïr The first data on the Salïrs are based on the works conducted in the West which are higher in number when compared to Turkish sources. At this point we need to mention Russian scholar Potapov’s compilations based on the last decades of the 19th century (and based on that N. Poppe’s work), the compilations and publications of E. R. Tenišev, text publications of Zsuzsa Kakuk, the works of anthropologist Kevin Stuart, the articles of R. Hahn and recently the works of Arienne M. Dwyer.

140

Ölmez

We owe the primary information about the Salïrs to Kashgari, specifically the part where the Oghuz clans are mentioned and the words Turkmen and boy (clan) are explained. As we know, the Oghuz consist of twenty-four clans. One of them is the Salïr. Sufficient information regarding the Salgurs/Salurs can be found in the sources that give information about the Oghuz and the Oghuz clans. For the historical resource entries about the name and the clan of Salur, Salur Atabegs and Salur settlements in Anatolia, see Sümer, 1980, pp. 336-344; 447-448. When the Salïrs are discussed in Turcological studies, their language and geography are almost always mentioned and covered with those of the Yellow Uyghurs. However, both the culture (religion, clothing, production styles) and geographies of these people are different. The two languages and peoples have been covered under the same entry in basic reference sources and handbooks published since 1959. Yet, it takes approximately 10 hours of a trip with the vehicles found in the region from Xining, the capital city of Xunhua where the Salïrs live, to the Sunan region where the Yellow Uyghurs live (for similar information and opinions see Dwyer, 2007, p. 1, footnote 1). While the Salïrs live together on the “Salïr” plain surrounded by the Huizus (Muslim Chinese people) and Tibetan villages, nearly two hours away from Xining, the Yellow Uyghurs live on the tableland which is two hours away from the city of Zhangye of Gansu province. There is no precise information about their population, either. According to the records, their population was 40,000 in 1960s, 56,000 in 1978 and 69,000 in 1982 (Schwarz, 1984, p. 39). According to the 1990 census, they had a population of nearly 65,000. The data belonging to the year 2000 suggests a population of about 100,000. A considerable portion of this population lives in Salïr (Xunhua), in Salïr Autonomous County (循化撒拉族自治县 Xunhua Salazu Zizhixian), the city of Xining, Huizu Autonomous County of Gansu and Ili prefecture of Xinjiang (for Ili Salïrs see Bibliography, Yakup, 1988 and 2002). The Muslim Salïr people are very similar to the Huizus and the Muslim Han Chinese in terms of culture (religion, clothing, production styles). They give and take girls as wives; and share the same mosque (mišit, mišt). According to my colleague Ma Wei (Yunus), the most important difference is maybe the wedding traditions and ceremonies. The Salïr people are involved in commercial activities in the whole region (Qinghai / Tibet Autonomous Region and the capital of Tibet) and in many cities of the People’s Republic of China. Mainly engaged in restaurant management, the Salïr people run restaurants known as “Lanzhou Restaurant” which are in compliance with Islamic standards. Not all of these restaurant owners are Salïr however; most of them are Muslim Chinese, the Huizus.

Endangered Turkic Languages From China

141

As to the origin of the Salïr, we cannot find detailed information about their history in Chinese resources. Related to the subject, one-folio long data found and published by Tenišev is important and interesting (Tenišev, 1977); according to the story circulating among themselves, they came to the region from Samarkand at the end of the 14th century; they chose that place as their homeland since the soil and water was similar to those of their motherland and the accompanying camel, which was carrying a Qur’an on his neck, was found petrified exactly at that point (Altiūli) after getting lost: that event is the subject of the story named döye yül ~ döye yuli “camel spring” (Chin. 駱駝泉 luo tuo quan). The altitude of the land they live is about 1800 meters. In a region where the Yellow River originates and reaches the inner parts of China, the Salïr usually settle along the “river”. Since it is the sole and closest river, the Salïr call the Yellow River morun ~ morïn (< Mo.) for short; they never use the word “yellow”. They have very good relationships with the Tibetans, especially with those living on the mountainous parts. The Tibetan villagers usually shop from Salïr shops. They speak Salïr effectively among themselves and in their daily lives. They are engaged in agriculture relying on irrigation on vast plains. Apart from Qinghai, there are a few thousand Salïr living in the city of Gulja, in Xinjiang Uyghur Autonomous Region (I will deal with this issue in another paper; the details are in the book of Abdurishid Yakup, Yakup, 2002). Those living within the borders of Gansu (according to the information Yunus provided) have almost forgotten the Salïr language. Today, the Salïr call Xunhua as “Salïr” among themselves: Men salïrga (~ Salïːga) vargur (~ vaːgur) “I will go to Xunhua”. They don not use the name Xunhua. With intensive efforts to learn their history and language as of 1990s, the Salïr started to publish a periodical, which includes Latin lettered texts, issued twice a year since 2008. Distinctive Features of the Salïr Language When compared with other Muslim Turkic language groups in Central Asia, Chaghatay and Kipchak groups, Salïr language distinguishes itself in terms of both grammar and vocabulary. When we compare it with Turkmen and Turkish, however, it partially shows an archaic characteristic. On the other hand, it has lost many grammatical elements in comparison with Yellow Uyghur which has far less speakers (about 5000 people). For example, it is almost impossible to see the first person affix for present when conjugating verbs. Palatal and labial harmony may not be seen in case of vowels. When we consider the vocabulary, we see that local Tibetan and Chinese have enormously affected Salïr and the words borrowed from these languages have replaced the words of

142

Ölmez

Turkic origin. Today, Tibetan gača ~ geče are used even for the words “word, utterance, language”: Yaxçux gačanï hemme kiš yišar, yaxçux išnï hemme kiš etmes “anybody (can) say a good word, not everybody does a good job / not everybody can exhibit good behaviors.” We separate the Salïr language from Kipchak and Chaghatay and approximate it to the Oghuz family of languages since the Salïr use the verb et- ~ ėt-, and the word el “hand” (in Kipchak and Chaghatay qol is used instead). Certainly, common use of the sound d- (dört “four” etc.) at the beginning of a word plays an important role in this approximation. (Several important features along with this will be discussed in detail in another paper; this paper is a report of a short field work carried out in August, 2009). The number systems do not go over “forty”. Especially the men, middle-aged people or the elderly do not know much about numbers. As to the women, I have heard the numbers forty, fifty, sixty, even from the young and middle-aged ones. We can relate this to the fact that the men are in more contact with other communities then women(the Chinese, the Tibetan). According to Musa Haji at the age of 64: Bir elli ma on dört “fifty and plus fourteen, fifty-four” (Ol Turkic *bir elig yme on tört) Below are the words which I heard on the first day and noted in a hurry in order of subject and hearing. More information on the material I recorded and the resources I collected will be included in another paper. Atuh (~kȯp) išse kursagïm agrïr. “if I eat a lot, I will become ill; if I eat a lot, I will have a stomachache” Saŋa bala neče (~neǰe?) var? segis, oːl dört, ane dört; suŋzï on segis vara. “how many kids do you have? Eight, son four, daughter four, grandchildren eighteen”. sïh “right, right hand side; well, healthy” dal “tree” ağaš “wooden, timber; wood” ağašli “forested; the forested village” emih “bread” hos “walnut” ǰiǰek “flower” ėt“to do, to make”

Endangered Turkic Languages From China

kini gadïn ame ape kiš

143

“spouse, wife” “woman” “mother” (ana “mam” is also known despite being an old word) “father” (ata “forefather” is also known despite being an old word) “person, human, man”

Salïr and Modern Uyghur can be compared in terms of similarities and differences with the following examples: Salïr sarï sağal satïhǰï satïhlï iş sensendersinih sïh soğan sorsorma suvasüt

Uyghur sėriq “yellow” saqal “beard” sėtiqçi “seller, merchant” soda “trade” öçmek “go out” söndürmek “extinguish” söŋek “bone” oŋ “healthy” piyaz “onion” sori-/sora- “to ask” harak “drink, alcoholic beverage” ussi- “to become thirsty” süt “milk”

For brief information on the Salïr see Ölmez, 2012.

Yellow Uyghur The population of the Yellow Uyghur living in the Gansu province of China, in Sunan Uyghur Autonomous Region close to the city of Zhangye and on the surrounding tablelands is over 10,000. Originally called Yellow Uyghur, this community refer to themselves shortly as Yogur ~ Yugur (Chin. Yugu). Those who speak a Turkic language are called Sarïg Yogur, while the others speaking Mongolian are called Shira Yogur. Their names are Western Uyghur (Xibu Yugu 西部裕固) and Eastern Uyghur (Dung Yugu 东部裕固), and their languages are Western Uyghur and Eastern Uyghur (Xibu Yuguyu and Dongbu Yuguyu 西部裕 固语 and 东部裕固语), respectively. The history and the language of the Yellow Uyghur belonging to old periods are known quite well compared to the other Turkic people in China. The

144

Ölmez

Yaglakar clan of the Uyghurs who established the khaganate and ruled for almost a hundred years (744-840) in Mongolia after the Uyghurs ended the Turkish rule (Tujue 突厥) live in the Mongolian-speaking region (see Ölmez 2012). We can look at Ariz, 2002, a work that gathers other works regarding the Yellow Uyghur. Living on the tablelands around Sunen, today the Yellow Uyghurs move to the tablelands of 2000-2500 metres high. With the coming of Spring, they go up to 3000-3500 metres high tablelands and to 4000 metres high tablelands with the coming of Summer (personal information received from Arslan). We need to mention G. N. Potanin as the first Western scholar doing research about the Yellow Uyghur. Having organized excursions to the Tangut region between the years 1884-1886, Potanin visited Yugur villages and residences; he collected data about the Mongolian-speaking Shira Yugurs and Turkic-speaking Kara Yugurs. According to Potanin, the Kara Yogurs are divided into two factions as Yaglak and Hurungut. These factions also are divided into smaller families (= otok). In the following years (1906-1908), C. G. Mannerheim arrived in the region and compiled texts both from the Yellow Uyghur and the Shira Yugurs. That was followed by Malov’s excursions (1909-1913) and detailed studies. After Malov, E. Tenišev took part in the activities of the People’s Republic of China about compiling Chinese minority languages; he made publications about text, dictionary and grammar studies. Afterwards, Lei Xuanchun and Chen Zongzhen made related publications in China. Today, Martti Roos, Erkin Ariz and Zhong Jingwen conduct studies in the field. With regard to the language of the Yellow Uyghur, they generally have been assumed to be the descendants of the Old Uyghurs presumably due to their names and some secondary language properties. As we mentioned above, it would be more accurate to classify them as the relatives of the Uyghurs who migrated to the region from Mongolia rather than assuming them to be the direct relatives of the Turfan Uyghurs. They have, naturally, connection with the Turfan Uyghurs. We can compare that with the migration of some Buddhist Uyghurs to the east, Dunhuang due to expansion of Islam in the Turfan region and with the Old Uyghur Altun Yaruk Sudur found by S. Ye. Malov. We should note that Yellow Uyghur shows similarities with Khakas language in terms of some phonetic evolution: OT -d-, -d becomes -z-, -z (OT adak “foot” > YUyg. azak, OT adgïr “stallion” > YUyg. azgïr, OT ïd- “to send” > YUyg. ïz-). However, it differs from Khakas in some aspects, for example OT vowel y- regularly becomes č- in Khakas, while sometimes it retains itself in Yellow Uyghur: OT yïl “year”, Khak. čïl, YUyg. yïl; OT yïltïz “root”, YUyg. yiltïs; OT yigit “young; strong” Khak. čit, YUyg. yïgït, yigit. OT consonant b- regularly becomes p- in

Endangered Turkic Languages From China

145

Khakas while it is seen both as b- and p- sound in Yellow Uyghur (see Ölmez, 1996 and 1998). To put it precisely, in Yellow Uyghur the OT consonant -g seen at the end of the polysyllabic words is retained as -k/-g but the consonant d becomes z as in the above given examples. Another old feature of Yellow Uyghur is seen in the number system. It shows similarities with Old Turkic: yidigirma “17” < yėti yėgirmi; sagïs yigirma “18” < sekiz yėgirmi, per otut “21” < bir otuz (see Clark, Geng & Clark). In some words, glottalization occurs before the unvoiced consonants k and t. YUyg. ahldï “altï” < OT altï, YUyg. tohɢïs “dokuz” < OT tokuz.

Tuvan Unlike the above mentioned Fuyu Kirghiz, Salïr and Yellow Uyghur, Tuvan is not a Turkic language exclusively spoken in China. Most Tuvans live in the Tuva Republic within Russia (in fact an autonomous region), in several cities of Russian Federation and in some regions of Mongolia. The Tuvan language spoken in China does not differ much from the original Tuvan; therefore, I will not dwell on the language of the Tuvan people of China in this short paper. I will just touch upon the differences between the Chinese Tuvan and the Tuvan spoken in Tuva. For Tuvan, see Ölmez, 2007, p. 25 and other. For the Tuvans living in Mongolia see Erika Taube’s works. The Tuvans of China became known better with the works of Geng Shimin, Talant Mawkanuli and Song Zengchun. The Tuvans living in Mongolia, their folklore, population and traditions have long been known thanks to the studies of Erika Taube. However, the villages and the towns and the population of the Tuvans in China are not documented as much when compared to Mongolia. Therefore, Marina V. Monguš went to the region, conducted field work and gave information about the Tuvan villages and towns with pictures. According to that information, the Tuvan living in China belong to the Ak Soyan (White Soyan) and Kara Soyan (Black Soyan) clans (2002, p. 21). Marina Monguš touches on the current situation of the Tuvan language and adds an interview at the end of her book. The interview made with Daš Čömblöv is given below. I would like to add that I interviewed the same source in Altay in 2004 (2002, p. 105): – Башкы, силер каш дыл билир силер? – МоолдуⱧ дылын билир мен. КазактыⱧ дылын билир мен. Мончак дыл билир мен. Кыдатча бичии-бичии билир мен.

146

Ölmez

– Teacher, how many languages do you speak? – I speak the language of the Mongols. I speak the language of the Kazakhs. I speak Tuvan. I speak very little Chinese. (Monguš, 2002) As is understood from the answer, Daš Čömblöv speaks three languages very well but speaks the fourth language poorly. The interview is from the year 1993. Yet, considering my personal observation I can safely state that the situation of the language in the Altay region of China has changed quite a lot since then; the Tuvan improved their Chinese; especially the Tuvan state officers speak Chinese very well. Those who receive education in Urumchi also speak Uyghur in addition to these four languages. To conclude, a well-educated “Monchak” speaks five languages including their mother tongue. This interview was made in 1993 with Daš Čömblöv who was born in 1962. The interview was written with present-day Tuvan alphabet but the punctuation reflects the punctuation of the Tuvans living in China. For example, in Tuva хꝋвей is used with x-, however, only кꝋвей is used in the interview made with Daš Čömblöv. Also, the word газыр which is derived from Kazakh qazïr is used to mean “now, at the moment, today” (Kazakh < Arabic). Another noticeable usage is the use of ивяааш instead of Tuvan эвээш. In the later parts of the interview, Daš Čömblöv Oronbayv briefly tells how their grandfather migrated from Russia to Kanas and met other Tuvans there; according to him, that migration occurred in 1913. Two women whom I talked to in Kanas told me a similar story. They, however, moved from Mongolia. As mentioned before, the Tuvan population concentrated in 3 villages: Kanas (Hanas), Akkaba and Kom. Kom is essentially a Tuvan village. Despite being mainly populated by the Tuvans, Kanas is also home to the Kazakhs and the Mongolians. Half of Akkaba’s population is Kazak and the other is Tuvan. The Kazakhs call the Tuvan as “Kök Monchak”. The Tuvans, however, call themselves diva or monchak. For population etc. see Mawkanuli, 1999, pp. 1-36; Ölmez, 2007, pp. 25-29; Yolboldi and Kasi̇, 1987, pp. 287-289. Among the places where the Tuvans live are the cities of Bowurǰin and Altay, and the towns of Köktogay and Lamajao. Since it is difficult to distinguish the Tuvans from the Mongols regarding lifestyles and beliefs, the Tuvans were accepted as Mongols and were not included among the 56 minority groups in China’s censuses. Chinese Tuvan was first made known by Geng Shimin. Russian Mongolist B. H. Todayeva met people with a language resembling Mongolian but unfamiliar to him while he was identifying and recording Mongol languages in the Altay region. With the opinion that the people were speaking a Turkic language, he informs Geng Shimin, and thus the research of Geng Shimin begins (this is what I was told by Geng Shimin).

Endangered Turkic Languages From China

147

Unlike Tuvan, words do not start with an h- sound but always a k- sound: kep, kerek, köl, kün, küreş (Tuvan hep “shape, form”, herek “necessary”, höl “lake”, hün “day, sun”, hüreš “wrestling”) etc. In Tuvan standard written language č- is systematically used instead of OT y-, but there are some words pronounced with ǰ- sound. In Chinese Tuvan, however, only c- sound occurs: ǰan-, ǰït-, ǰi- (Tuvan čan- “to turn”, čït- “to sleep”, či- “to eat”). The Tuvan language spoken in Chinese Altay includes many Kazakh (