Language Death and Language Maintenance: Theoretical, Practical and Descriptive Approaches (Current Issues in Linguistic Theory)

LANGUAGE DEATH AND LANGUAGE MAINTENANCE

AMSTERDAM STUDIES IN THE THEORY AND HISTORY OF LINGUISTIC SCIENCE General Editor E. F. KONRAD KOERNER (University of Cologne) Series IV – CURRENT ISSUES IN LINGUISTIC THEORY

Advisory Editorial Board Raimo Anttila (Los Angeles); Lyle Campbell (Christchurch, N.Z.) Sheila Embleton (Toronto); John E. Joseph (Edinburgh) Manfred Krifka (Berlin); Hans-Heinrich Lieb (Berlin) E. Wyn Roberts (Vancouver, B.C.); Hans-Jürgen Sasse (Köln)

Volume 240

Mark Janse and Sijmen Tol (eds) Language Death and Language Maintenance Theoretical, practical and descriptive approaches

LANGUAGE DEATH AND LANGUAGE MAINTENANCE THEORETICAL, PRACTICAL AND DESCRIPTIVE APPROACHES

Edited by

MARK JANSE Linguistic Bibliography, Ghent University, University of Amsterdam

SIJMEN TOL Linguistic Bibliography With the assistance of Vincent Hendriks

JOHN BENJAMINS PUBLISHING COMPANY AMSTERDAM/PHILADELPHIA

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences — Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.

Library of Congress Cataloging-in-Publication Data Linguistic Bibliography and the Languages of the World (2000 : Hague, Netherlands) Language death and language maintenance: theoretical, practical and descriptive approaches / Edited by Mark Janse and Sijmen Tol. With the assistance of Vincent Hendriks p. cm. -- (Amsterdam studies in the theory and history of linguistic science. Series IV, Current issues in linguistic theory, ISSN 0304-0763 ; v. 240) “The present collection of papers derives from a symposium “Linguistic Bibliography and the Languages of the World” held on November 2-3, 2000 at the National Library of the Netherlands (Koninklijke Bibliotheek)...” -- Preface. Includes bibliographical references and index. 1. Language obsolescense--Congresses. 2. Language attrition--Congresses. 3. Language maintenance-Congresses. I. Janse, Mark, 1959- II. Tol, Sijmen III. Title. IV. Series. P40.5 L33L56 2003 417'.7--dc21 2003041787 ISBN 90 272 4752 8 (Eur.) / 1 58811 382 5 (US) (Hb; acid-free paper) © 2003 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. • P.O.Box 36224 • 1020 ME Amsterdam • The Netherlands John Benjamins North America • P.O.Box 27519 • Philadelphia PA 19118-0519 • USA

CONTENTS Preface

vu

Introduction: Language death and language maintenance: Problems and prospects Mark Janse

ix

The endangered languages issue as a hopeless cause Paul Newman

1

The language situation and language endangerment in the Greater Pacific area Stephen A. Wurm

15

Language endangerment in Indonesia: The incipient obsolescence and acute death of Teun, Nila and Serua (Central and Southwest Maluku) Aone van Engelenhoven

49

Sibe: An endangered language Giovanni Stary

81

The gradual disappearance of a Eurasian language family: The case of Yeniseyan Stefan Georg

89

The endangered Uralic languages Rogier Blokland & Cornelius Hasselblatt

107

Endangered Turkic languages: The case of Gagauz Astrid Menz

I43

Loss of linguistic diversity in Africa Maarten Mous

157

VI

CONTENTS

Ongota (Birale), a moribund language of Southwest Ethiopia Graziano Savà

171

An endangered language: The Gùrdùrj language of the Southern Bauchi Area, Nigeria Andrew Haruna

189

Resian as a minority language Han Steenwijk

215

Index of languages

227

Index of names

237

Index of subjects

241

PREFACE

The present collection of papers derives from a symposium "Linguistic Bibliography and the Languages of the World", held on November 2-3, 2000 at the National Library of the Netherlands (Koninklijke Bibliotheek) to celebrate the publication of the 50th volume of Bibliographie Linguistique / Linguistic Bibliography. The editors wish to thank the directors and personnel of the Koninklijke Bibliotheek for substantial financial and practical support and the Netherlands Organisation for Scientific Research (NWO) for additional fund­ ing. A special word of thanks is due to Inge Angevaare, Theo Horstman and Vincent Hendriks for their dedication and efforts in the preparation of the sym­ posium and the present volume. We are also grateful to the series editor Kon­ rad Koerner, an anonymous referee and to John Benjamins Publishing Com­ pany, especially to Anke de Looper, for their support and encouragement. This volume is dedicated to the memory of Stephen A. Wurm, former president of CIPL, linguist extraordinaire, polyglot, and defender of endangered lan­ guages. Mark Janse & Sijmen Tol Bibliographie Linguistique/Linguistic Bibliography The Hague, January 2003

INTRODUCTION LANGUAGE DEATH AND LANGUAGE MAINTENANCE PROBLEMS AND PROSPECTS

MARK JANSE Linguistic Bibliography / Ghent University / University of Amsterdam

It has been estimated that more than half of the world's languages have dis­ appeared in the last 500 years (Sasse 1990). Of the remaining 6,809 languages listed in the latest printed edition of the Ethnologue (Grimes 2000) more than half are believed to be in danger of disappearing in the present century. Ac­ cording to one pessimistic view, only 600 languages stand a fair chance of sur­ viving in the long run (Krauss 1992). The conclusion is inescapable: languages are dying at an alarming rate all over the world. Leaving aside such cases as restricted languages (otherwise dead languages used exclusively in restricted domains, e.g. liturgical languages such as Latin, Coptic or Ge'ez) and residual languages (otherwise lost languages preserved in isolated words, phrases, songs or sayings to mark group membership, particu­ larly in minority groups), a language is dead when it no longer has any speak­ ers. Language death is defined by Campbell as "the loss of a language due to gradual shift to the dominant language in language contact situations" (1994:1961).1 Such situations involve an intermediate stage of bilingualism in which the subordinate language is employed by a decreasing number of speak­ ers in an equally decreasing number of contexts, until it ultimately disappears altogether. The process is typically accompanied by a gradual attrition of the subordinate language along a continuum determined mainly by age (although attitude and other factors may play an important part). Languages in the process of dying are endangered languages. Wurm (this volume) distinguishes five levels of language endangerment. A language is potentially endangered if the children start preferring the dominant language and learn the obsolescing language imperfectly. It is endangered if the young1 There are of course cases of abrupt language death caused by genocide, natural disasters or epidemics.

x

MARK JANSE

est speakers are young adults and there are no or very few child speakers. It is seriously endangered if the youngest speakers are middle-aged or past middle age. It is terminally endangered or moribund if there are only a few elderly speakers left. A language is dead when there are no speakers left at all. The factors determining language death are typically "non-linguistic" (Swadesh 1948:235). A long list of such factors can be found in Campbell (1994:1963). The most commonly cited are socioeconomic and sociopolitical. Socioeconomic factors include lack of economic opportunities, rapid economic transformations, on-going industrialization, work patterns, migrant labor, re­ settlement, migration and so on. Among the sociopolitical factors are official language policies, discrimination, stigmatization, repression, war etc. Official language policies can be and have been a particularly decisive factor in lan­ guage death. Western colonialism has proven extremely efficient in this re­ spect, as can be gathered from the use of the term "glottophagie" in Calvet (1974). Another term frequently encountered in this context is "linguicide", a concept analogous to genocide (Skuttnab-Kangas & Phillipson 1996:2212). The classic example is the "English Only" policy of the United States govern­ ment in the 19th century, designed to force Native Americans to learn English (still echoing in the "English Only" amendment adopted in 1988 in Arizona and the proposed "English Only" bill in Utah). Many modern parallels can be adduced, such as the repression of Kurdish in Turkey, Albanian in Kosovo or Aromanian in Greece.2 The official status of languages crossing borders may vary according to the statutory laws of the various countries. Catalan and Basque, for instance, have official language academies in Spain, but not in France. As much as linguicide and linguistic discrimination may add to language death, they are at the same time powerful forces in the reawakening of ethnic identity feelings among speakers of endangered minority languages, which appears to have become a global trend from about 1970 onwards. Ethnic identy is often accompanied by an increased interest in language maintenance. This was, curiously enough, matched from 1970 onwards, by a switch of negative, or very negative, governmental policies towards minority languages in their orbit to positive, or very positive, ones in many countries such as Australia, Japan, Taiwan, Canada, Scandinavia, Russia after the collapse of communism, Papua New Guinea, and others. This has led to the maintenance and revitalization of many endangered languages in parts of the world, and even to the re­ vival of some extinct languages, e.g. among Australian Aborigines. Other countries switched at least to neutral attitudes, whereas some, such as the USA, 2

Other cases are discussed in Skuttnab-Kangas & Phillipson (1995).

INTRODUCTION

xi

most African countries, some countries in South America, a few European and Middle Eastern countries, Indonesia etc., still adhere to their negative attitudes. Language death is of course not a new and not even a recent phenomenon. Let us start with a well-known myth of language birth in the days of yore. Ac­ cording to the biblical story of the tower of Babel, the whole world had at one time "one language and a common speech" (Genesis 11. 1). When man tried to build a tower that would reach to the heavens, God decided to "confuse the language of the whole world" and to "scatter the people over the face of the whole earth" (Gen. 11. 8). The peoples that eventually spread out over the earth after the flood were named after Noah's sons: the Japhethites, the Hamites and the Semites (Gen. 10. 1-32). The historical reality behind the story can of course be seriously questioned, as well as the rough correspondency with the Indo-European ("Japhethic") and Afröasiatic (Hamito-Semitic) language families. Yet even a quick glance at the number of extinct Indo-European and Afroasiatic languages should suffice to give an impression of the extent of language death in ancient times. Among these are a number of major literary languages such as Akkadian, Ugaritic, Ancient Hebrew, Ancient Aramaic, Ancient Egyptian and the Ancient Greek dialects. Other languages are less well-known such as the following from the Indo-European language family: Pahlavi, Sogdian, Khorasmian, Khotanese Saka and Tumshuqese (Middle Iranian), Luwian, Palaic, Lycian, Lydian, Carian, Sidetic and Pisidian (Anatolian), Faliscan, Oscan, Umbrian, Paelignian, Marrucinian, Vestinian, Venetic and South Picene (Italic), Gaulish, Lepontic and Celtiberian (Celtic), Thracian and Dacian, Illyrian and Messapic, Phrygian, Ancient Macedonian and still others. All of these are known from written tes­ timonies, but other extinct languages are known by name only and probably even more are not and will never be known at all. One of the better-known linguistic "killing fields" is Asia Minor (Janse 2002:347-359). Practically all the indigenous languages of Asia Minor became extinct under the pressure of Hellenization: Hatti, Human, Hittite and the other Anatolian languages, Phrygian, Galatian, Gothic, and a number of other lan­ guages known by name only such as Mysian, Lycaonian, Cataonian, Cilician, Bagdaonian and Cappadocian.3 The prestige of a politically and culturally su­ perior lingua franca was such that in the Persian Empire of the Achaemenids Greek language and even constitutional forms were adopted by satraps such as Ariarathes I of Cappadocia and Mausolus of Caria (both 4 th . ). Hellenism was used by the Romans to impose their own authority in Asia Minor. The 3

Ancient Cappadocian is not to be confused with the Modern Greek dialect of the same name (Janse 2002:352-355).

XII

MARK JANSE

Pontic geographer Strabo, who died shortly after the Roman annexation in 17 AD, noted that "under their reign most of the peoples had already lost both their languages and their names" {Geography 12. 4. 6). The case of Greek in Asia Minor shows that socioeconomic and political circumstances are neither sufficient nor necessary causes of language death (Dressier 1988:190-191). Neither the Persians nor the Romans were socioeconomically and politically inferior to the Greeks, yet both adopted Greek lan­ guage and culture for their own purposes in Asia Minor. In this respect it is particularly interesting to note that the Persians chose the Aramaic language for communication in other parts of their empire, while the Romans naturally used Latin in the western provinces. The imposition of Greek in the East and Latin in the West did not lead to abrupt language death. Most, if not all, of the indigenous languages went through an intermediate stage of bilingualism (Adams, Janse & Swain 2002). The classic example is Thucydides' "bilingual Carian" {Histories 8. 85). Bilin­ gualism inevitably leads to borrowing and according to Campbell language death is an extreme case "where an entire language is borrowed at the expense of another" (1994:1960). There are, however, cases where the bilingual stage does not lead to language death stricto sensu, but where the subordinate lan­ guage is maintained and subjected to what Thomason and Kaufman call "heavy borrowing" (1988:50). In its most extreme form the linguistic result of heavy borrowing is what Thomason calls a "contact language", defined as "any new language that arises in a contact situation ... identifiable by the fact that its lexicon and grammatical structures cannot all be traced back primarily to the same source language" (Thomason 2001:158). Contact languages are also called "mixed languages" {ibid.). According to Strabo, Carian would qualify as such: "it has extremely many Greek words mixed up with it" {Geography 14. 2. 28).4 Strictly speaking, the original language has not died, but has been transformed into a new language.5 The linguistic differences between dying languages and mixed languages are important. Dying languages generally exhibit morphological and syntactic reduction (Dressier 1988:184-188; Campbell 1994:1962-1963), whereas mixed languages, with the notable exception of pidgins and creoles, generally retain and often combine the complexities of the source languages.6 Language death, in other words, is normally characterized by attrition, leaving in the final stages 4

The notion of mixed languages is well attested in antiquity (Janse 2002:333-334). Examples of contact languages can be found in Thomason (1997). 6 Compare, for instance, the grammatical complexity of Michif (Bakker & Papen 1997) or Media Lengua (Muysken 1997). 5

INTRODUCTION

XIII

only "forgetters" and "rememberers" (Campbell 1994:1960-1961).7 Needless to say, the degree of attrition will have serious consequences for the description of the language, especially if it has never been described before. The description of endangered languages is an urgent task for various rea­ sons. First, every language expresses thoughts and ideas in unique ways, both grammatically and semantically. The quest for universal grammar tends to obliterate the diversity of natural languages, even though studies of hitherto undescribed languages tend to reveal "same but different" structures time and again. The study of such languages is therefore of the utmost importance for our general understanding of the sum total of the possibilities of the formal and semantic expression of human thought patterns. Second, the study and descrip­ tion of endangered languages saves them from oblivion after the death of their last speakers. This is not only of interest to future linguists, but equally impor­ tantly, it may enable the descendants of the last speakers to acquaint them­ selves with and even to relearn their ancestral language. A good example is the Aboriginal Kaurna language: its last speaker died in 1927, but the language has been revitalized on the basis of earlier descriptions (Wurm, this volume). Third, every language is the guardian of its speakers' history and culture and its extinction represents "the irretrievable loss of a portion of our own humanity" (Campbell 1994:1966). The conservation of oral traditions in endangered lan­ guages will help us understand more about human values, culture, world view, verbal art, oral literature, and much more. The question of language maintenance and revitalization is too complex a matter to go into in detail. It has been discussed most extensively by Wurm (1997; 1998; 2002). Suffice it to mention some key factors. On the community level, language endangerment can be reversed if the children are encouraged to relearn the language with the help of the surviving speakers in playing situa­ tions. Literacy programs and mother tongue education are of course essential as well, especially if they are backed up with language attitudes such as ethnic identity awareness. The success of such programs depends in no small measure on national and international language policies such as official language status and linguistic human rights in general (Skuttnab-Kangas & Phillipson 1995). Although language death is not new, its study is fairly recent. Apart from pioneering works like Cust (1899), Vendryes (1933; 1951; 1954), Swadesh (1948), Terracini (1951), Ellenberger (1962) and Pande (1965), language death started drawing serious attention in the 1970s, culminating in a special issue of 7 A note of caution is in order here, as dying languages can experience both generalization of unmarked features and overgeneralization of marked features (Campbell 1994:1962).

XIV

MARK JANSE

IJSL (Dressler & Wodak-Leodolter 1977), the first and definitely not the last collective volume on the subject. The 1980s witnessed the start of a veritable explosion of workshops, conferences and publications on language death, in­ cluding a recent encyclopedia of endangered languages (Moseley 2001), a spe­ cialist journal to be published by Mouton de Gruyter and the first "popular" books on the subject (Crystal 2000; Hagège 2000). These were followed, in the 1990s, by the establishment of the first com­ mittees, societies, and foundations such as the Linguistic Society of America's Committee on Endangered Languages and their Preservation, the German So­ ciety for Linguistics' Society for Endangered Languages (Gesellschaft für bedrohte Sprachen), the British Foundation for Endangered Languages and the International Clearing House for Endangered Languages (ICHEL) of the Uni­ versity of Tokyo. A bibliography on endangered languages is maintained on the website of the ICHEL. The LINGUIST List is currently setting up its own Endangered Languages Homepage and many linguistic areas in the world have their own list on the internet. At the 15th International Congress of Linguists, held in August 1992 at La­ val University, Quebec, the Comité Permanent International de Linguistes (CIPL) put language endangerment on top of the agenda. A collective volume was edited by the then president and secretary-general of CIPL in preparation of the plenary session on "Endangered Languages" (Robins & Uhlenbeck 1991) and the same title appeared emblematically on the cover of the proceed­ ings of the congress (Crochetière, Boulanger & Ouellon 1993). During the con­ ference the General Assembly of CIPL approved the following resolution: As the disappearance of any-one language constitutes an irretrievable loss to mankind, it is for UNESCO a task of great urgency to respond to this situation by promoting and, if possible, sponsoring programs of linguistic organizations for the description in the form of grammars, dictionaries, and texts including the recording of the oral lit­ eratures - of hitherto unstudied or inadequately documented endangered and dying languages.

With financial support from UNESCO CIPL is now actively involved in the organization and coordination of the survey and study of some seriously en­ dangered languages of the world. This work includes fieldwork, collecting and recording appropriate language material and documentation, linguistic research and other activities. Important publications include UNESCO's Red Book of Endangered Languages and the Atlas of the World's Languages in Danger of Disappearing (Wurm 2001). As a matter of course CIPL's initiative was extended to its main publication Bibliographie Linguistique / Linguistic Bibliography (BL): "an important part

INTRODUCTION

xv

is reserved for BL to provide the linguistic community with exhaustive lin­ guistic information on the many endangered languages of the world" (Janse & Tol 1993:vii). Since BLonline, the on-line version of BL, is now, thanks to the generous support of the National Library of the Netherlands, available for free on the internet, we hope to serve the linguistic community even better in this respect.8 When BL celebrated its 50th volume in 2000 at the National Library of the Netherlands, a symposium was organized around the theme "Linguistic Bibli­ ography and the Languages of the World". With the exception of the then president of CIPL, Stephen Wurm, the invited speakers were all selected from among BL's contributors, many of whom specialists in endangered languages. The present volume includes some of the papers presented at the symposium as well as a number of invited contributions, three written by former contributors (Newman, Mous and Steenwijk) and two by relative outsiders from BL's point of view (Haruna and Savà). The resulting collection tries to strike a balance between theoretical, practical and descriptive approaches to language death and language maintenance. It is our hope that it will provide a useful addition to the ever-growing body of literature on endangered languages. References Adams, J. N., Mark Janse & Simon Swain, eds. 2002. Bilingualism in ancient society: language contact and the written word. Oxford: Oxford Uni­ versity Press. Asher, R. E. & J. M. Y. Simpson, eds. 1994. The encyclopedia of language and linguistics. Oxford: Pergamon Press. Bakker, Peter & Robert A. Papen. 1997. "Michif". Thomason 1997. 295-363. Bradley, David & Maya Bradley, eds. 2002. Language endangerment and lan­ guage maintenance: an active approach. London: Curzon Press. Calvet, Louis-Jean. 1974. Linguistique et colonialisme: petit traité de glottophagie. Paris: Payot. Campbell, Lyle. 1994. "Language death". Asher & Simpson 1994, vol. 4. 1960-1968. Crochetière, André, Jean-Claude Boulanger & Conrad Ouellon. 1993. Les lan­ gues menacées: actes du XVe congrès international des linguistes, Québec, Université Laval, 9-14 août 1992. Sainte-Foy: Presses de l'Université Laval. 8

At the time of this writing BLonline provides all the entries of the printed volumes covering the years 1993-1998, plus an ever-increasing number of more recent references. BLonline is updated every month and available at .

XVI

MARK JANSE

Crystal, David. 2000. Language death. Cambridge: Cambridge University Press. Cust, Robert Needham. 1899. Language its I. birth, II. development and life, III decay and death. London: Spottiswoode & Co. Dressier, Wolfgang U. 1988. "Language death". Newmeyer 1988. 184-192. —, & Ruth Wodak-Leodoter, eds. 1977. Language death. The Hague: Mouton {International Journal of the Sociology of Language, 12). Ellenberger, Henri F. 1962. "Mort et résurrection des langues". Revue de Psychologie des Peuples 17. 430-441. Grimes, Barbara F., ed. 2000. Ethnologue. 14th ed. Dallas, Texas: SIL Inter­ national. Hagège, Claude. 2000. Halte à la mort des langues. Paris : Odile Jacob. Janse, Mark. 2002. "Aspects of bilingualism in the history of the Greek lan­ guage". Adams, Janse & Swain 2002. 332-390. —, & Sijmen Tol, eds. 1993. Linguistic Bibliography for the year 1991 and supplement for previous years. Dordrecht: Kluwer Academic Pub­ lishers. Krauss, Michael. 1992. "The world's languages in crisis". Language 68. 4-10. Matsumura, Kazuto, ed. 1998. Studies in endangered languages. Tokyo: Hituzi Syobo (ICHEL Linguistic Studies, 1). Moseley, Christopher, ed. 2001. Encyclopedia of the world's endangered lan­ guages. London: Curzon Press. Muysken, Pieter. 1997. "Media lengua". Thomason 1997. 365-426. Newmeyer, Frederick J., ed. 1988. Linguistics: the Cambridge survey. Vol. IV. Language: the sociocultural context. Cambridge: Cambridge University Press. Pande, Govind Chandra. 1965. "The life and death of languages". Diogenes 51. 193-210. Robins, Robert H. & Eugenius M. Uhlenbeck, eds. 1991. Endangered lan­ guages. Oxford: Berg (Diogenes Library). Sasse, Hans-Jürgen. 1990. Theory of language death. Köln: Institut fur Sprach­ wissenschaft, Universität zu Köln (Arbeitspapiere N.F., 12). Shoji, Hiroshi & Juha Janhunen, eds. 1997. Northern minority languages: problems of survival. Osaka: National Museum of Ethnology (Senri Ethnological Studies, 44). Skuttnab-Kangas, Tove & Phillipson, Robert. 1994. "Linguicide". Asher & Simpson 1994, vol. 4. 2211-2212. —, & —, eds. 1995. Linguistic human rights: overcoming linguistic dis­ crimination. Berlin: Mouton de Gruyter.

INTRODUCTION

xvii

Swadesh, Morris. 1948. "Sociologic notes on obsolescent languages". Inter­ national Journal of American Linguistics 14. 226-235. Terracini, Benvenuto. 1951. Conflictos de lenguas y de cultura. Buenos Aires : Imán (Panorama de la filosofía y de la cultura, 2). Thomason, Sarah Grey, ed. 1997. Contact languages: a wider perspective. Amsterdam: John Benjamins Publishing Company (Creole Language Library, 17). —, 2001. Language contact: an introduction. Edinburgh: Edinburgh Uni­ versity Press. —, & Terrence Kaufman. 1988. Language contact, creolization, and genetic linguistics. Berkeley: University of California Press. Vendryes, Joseph. 1933. "La mort des langues". Conférences de l'Institut de Linguistique de l'Université de Paris 1. 5-15. —, 1951. "La mort et la résurrection des langues". Hesperia 6-1. 79-101. —, 1954. "Comment meurt une langue". Vie et Langage 31. 449-452. Wurm, Stephen A. 1997. "Prospects of language preservation in the North". Shoji, Hiroshi & Janhunen 1997. 35-53. —, 1998. "Methods of language maintenance and revival, with selected cases of language endangerment in the world". Matsumura 1998. 191-211. —, 2001. Atlas of the world's languages in danger of disappearing. 2nd ed. Paris: UNESCO Publishing/Pacific Linguistics. —, 2002. "Strategies for language maintenance and revival". Bradley 2002. 1123. —, Peter Mühlhäusler & Darreil T. Tryon, eds. 1996. Atlas of languages of intercultural communication in the Pacific, Asia, and the Americas. Berlin: Mouton de Gruyter.

THE ENDANGERED LANGUAGES ISSUE AS A HOPELESS CAUSE* PAUL NEWMAN Department of Linguistics and West African Languages Institute, Indiana University

1 Introduction The figure often bandied about, taken from statements by Michael Krauss (1992), is that there are now some 6,000 languages in the world, half of which are likely to be lost within the present century. According to Krauss, the situa­ tion is even worse: of these 3,000 remaining languages, only 600 have a good chance of real survival over the long term, i.e., if the trend persists, some 90% of the world's languages will be lost. The question Krauss (1992:7) poses is: "What are we linguists doing to prepare for this or to prevent this catastrophic destruction of the linguistic world?" It is only within the past ten or so years that linguists have begun to focus on this issue and to stress the point that the disappearance of languages and linguistic diversity is a major loss to linguistic scholarship and science. How­ ever, since the endangered languages issue was brought to the fore, it has caught the attention of the linguistics profession and has stimulated much activity. There have been a variety of conferences on the subject, both in America and in Europe, and a number of major publications have appeared, including Brenzinger (1998), Brenzinger, Heine & Sommer (1991), Fishman (1991; 2001), Grenoble & Whaley (1998), Matsumura (1998), and Robins & Uhlenbeck (1991). The lead article in the March 1992 issue of Language,

* Acknowledgements: This is a revised version of an article originally published as, '"We has seen the enemy and it is us': The endangered languages issue as a hopeless cause", Studies in the Linguistic Sciences 28(2): 11-20 (1998). The paper was first presented at a symposium on "The Linguistic Sciences in a Changing Context" organized by Braj Kachru, Adele Goldberg, and Jerry Morgan at the University of Illinois (1998). The paper was subsequently the subject of a heated discussion at a workshop on endangered languages organized by Simon Donnelly at the LSA summer institute (1999).

2

PAUL NEWMAN

which was written by a group of distinguished linguists, was devoted to the endangered languages question (see Hale et al. 1992). Even more striking has been the creation of organizations and activities devoted to the topic. For example, the Linguistic Society of America (LSA) has a standing Committee on Endangered Languages and their Preservation; there is a Foundation for Endangered Languages at the University of Bristol, UK, and an International Clearing House for Endangered Languages at the University of Tokyo. UNESCO is involved in the preparation of an Atlas of the World's Languages in Danger of Disappearing. Non-profit foundations are also springing up, e.g., the Endangered Languages Fund (New Haven) and Terralingua, Partnerships for Linguistic and Biological Diversity (Hancock, Michigan). In addition, the Lisbet Rausing Charitable Fund has recently given the University of London School of Oriental and African Studies (SOAS) a large grant to set up an Academic Project on Endangered Languages. Although I think that Marianne Mithun (1998:163) is overstating it when she asserts, "At long last the tragedy of language loss worldwide has begun to enter the public conscious [sic]," it is true that awareness of the issue now extends beyond the narrow confines of professional linguists. Here one can cite the front page articles that appeared in the New York Times (Brooke 1998) and the Washington Post (Gugliotta 1999) and the short piece found in News­ week (Raymond 1998). My intention here is not to raise the question of why languages disappear (see Mufwene 1998; 2001). Nor do I want to get into the sensitive question of whether it makes any sense philosophically or practically to try to renew or revive dying languages (see Ladefoged 1992; Dorian 1993; Fishman 2001). Once one leaves the realm of emotional hand twisting by sentimental scholars, the question is much more debatable than appears at first sight. However, I think that professional linguists can agree that the disappear­ ance of a language without documentation is a huge scientific loss. Our lin­ guistic scientific enterprise depends on the multiplicity of languages and the knowledge of linguistic diversity. It is only through knowledge of diverse languages with different structures and belonging to different language fami­ lies that we can truly begin to gain an understanding of universal grammar, i.e., the nature of the human language capacity. Similarly, our understanding of linguistic typology and our ability to classify languages accurately and reconstruct proto-forms depends on the availability of a wide array of lan­ guages. If one believes this, if one takes the position that no language should be allowed to become extinct without having been scientifically preserved, then one has to acknowledge that the task is urgent. Speakers of endangered lan-

THE ENDANGERED LANGUAGES ISSUE AS A HOPELESS CAUSE

3

guages are not only dying away - the most obvious and final loss - but they are also forgetting their languages and losing command of the richness that defined that language as opposed to the one down the road. As Dixon (1997:147 note 5) correctly points out, "A sad lesson that has been learnt from the study of language-death situations is that a community does not realise its language is threatened until it is too late to do anything to remedy the situa­ tion." Dixon's view about what needs to be done is stated in unequivocal terms (1997:144): "The most important task in linguistics today - indeed, the only really important task - is to get out in the field and describe languages, while this still can be done. Self-admiration in the looking glass of formalist theory can wait; that will always be possible. Linguistic description must be under­ taken now." Even if one puts some of the hyperbole aside, the truth is that the problem is real and we linguists are doing very little about it, apart from discussing the matter among ourselves so as to assuage our guilt. This is clearly a case where we cannot shift the blame to someone else: the failure to tackle the endangered languages crisis is not due to some budget dean nor to some philistine of a congressman nor to a CEO of some big corporation. The fault lies with us linguists, the people who should be up in arms about the problem. In essence, to quote from Pogo, the popular American comic strip created by Walt Kelley (1913-1973), "We has seen the enemy, and it is us." What I shall do in this paper is discuss significant areas in which we as linguists exacerbate rather than solve the problem. The discussion falls under three headings: First, we linguists don't care; second, we linguists care too much; and third, our non-western colleagues don't care and would be unpre­ pared to help out even if they did. 2 We linguists don't care 2.1 Theory One hates to make blanket generalizations about a discipline as varied and with so many subfields as linguistics. Nevertheless, it is probably fair to say that in terms of overall world view and intellectual orientation, linguistics as a field is fundamentally theory driven as opposed to data driven. There was a time when linguistics was inextricably tied up with the study of non-written, non-western languages, but this is not the case today. General linguists aren't opposed to the study of these languages; it's just that it isn't important to them. What is viewed as important is trying to characterize the species-shared human language capability, i.e., linguistics has diverged from its anthropological and

4

PAUL NEWMAN

philological roots and has become in effect a branch of cognitive psychology. The lack of concern about the endangered languages problem is an extension of the general lack of interest in descriptive empirical research, whatever the language might be. This lack of interest is reflected in the structure of graduate linguistics curricula, the content of linguistics courses at the introductory and advanced levels, and in professional hiring practices. The low valuation of fieldwork in our discipline is shown by the marginal position of field methods classes in most linguistics programs. (See Newman 1992 for a now out-of-date but probably still accurate compilation of departmental practices.) The appearance of two recent books on fieldwork (Vaux & Cooper 1999; Newman & Ratliff 2001) might be heartening and suggest a reawakening of interest in field lin­ guistics. But against this one can point to the depressing fact that out of almost ninety different courses offered at the 2001 LSA Linguistics Institute in Santa Barbara, there was not a single course on field methods! Someone might legitimately ask whether there is any objective evidence to document the claim that the empirical study of "exotic" languages occupies a minor position in linguistics. To check this out, I decided to look at the topic of Ph.D. dissertations completed at American universities, since what students focus on is probably a reasonable reflection of the current ethos in a field and the interests of their teachers. The information was derived from an analysis of the linguistics section of Dissertation Abstracts International covering the period from January 1997 through December 2000, i.e., four years' worth of entries. Based primarily on the titles, with a quick glance at the abstracts them­ selves, I sorted the dissertations into a number of crude categories, defined both in terms of linguistic subfields, e.g., theoretical (English), computational linguistics, language acquisition (including second language learning), and sociolinguistics, or in terms of language families and geographical areas, e.g., Romance, Semitic, African, Native American, etc. Granted that the methodol­ ogy was a bit haphazard and unsystematic, nevertheless the results were strik­ ingly clear. In the four years covered, there were a total of some 1,860 dissertations. Of these, 1,152 were concerned either with English, including English as a second language, or matters of a general linguistic or theoretical nature; 325 were on languages of Europe (including Turkish), of which 274 were on the big three, namely, Romance, Germanic, and Slavic; 215 were on Asian languages, most of which (192) were on Chinese, Japanese, or Korean; and 40 were on Semitic and Ancient Egyptian. These four macro-categories accounted for 1,732 of the dissertations, i.e., 93% of the total. That left a mere 128 dissertations (i.e., 7%) devoted to the languages of Native America (51), Africa (41), Austronesia

THE ENDANGERED LANGUAGES ISSUE AS A HOPELESS CAUSE

5

(including India) (33), and Australia (3). But even these 128 dissertations cannot all be said to represent fieldwork on small "exotic" languages, since (a) they include studies of major languages such as Malay, Hindi, Navajo, Hausa, and Swahili, and (b) some of these theses were theoretical studies drawing on secondary materials or descriptions of the student's own language. In short, 5% at most of the Ph.D. dissertations written during the past four years could be said to represent primary linguistic research involving significant fieldwork. 2.2 The culture of linguists (as opposed to anthropologists) When linguistics was a part of anthropology, as it was for Boas, Lounsbury, Sapir, Voegelin, et al. (see Lounsbury 1968), fieldwork was a natural compo­ nent of work in the discipline. Anthropology graduate students have tradition­ ally been expected to go into the field; a student who wanted to do an "arm­ chair" dissertation was viewed as a professional misfit. Crediting Kroeber, Geertz (1984:265) speaks of the "centrifugal impulse of anthropology - distant places, distant times, distant species ... distant grammars." Clearly there are problems with basing scholarly pursuits on the appeal of the "exotic" (con­ sider, for example, the concerns expressed by Said 1978), but what is striking about linguistics nowadays as opposed to anthropology is its almost total separation from fieldwork. My personal experience with linguistics graduate students over the years is that they display a singular lack of venturesomeness. Students aren't attracted by the idea of fieldwork for the simple reason that they don't want to go to the field. I suspect, that if I had funding to send a dozen graduate students to remote places to do research on dying languages, I would have trouble giving the money away. The students whom I have met would much rather stay in the comfort of a university setting working within the confines of the latest (and thus non-risky) linguistic theory than undertake basic descriptive research, i.e., they have no desire to subject themselves to the practical rigors and intellectual uncertainties and frustrations of fieldwork (see Hyman 2001). Dixon's charge to linguists to get out into the field will fall on deaf ears because it runs counter to the prevailing culture and personality of the people who now make up the discipline of linguistics. 3 We linguists care too much When Emmon Bach, a well-known and distinguished theoretical linguist, was working on Wakashan, an endangered language of British Columbia, he was challenged by one of the elders as to why he and members of his commu­ nity should care about the linguistic work that was being done. Bach's re­ sponse was to formulate the following principle (Bach 1995): "I will try to put at least half of my time and effort in working in a community into things that

6

PAUL NEWMAN

make sense for the community. What that work might be can range from things as simple as copying tapes for people who want them, through prepar­ ing texts, etc. in ways that are accessible, to helping out with language pro­ grams, etc." This quotation has subsequently been repeated with approbation and without a hint of challenge - by any number of linguists. Whereas fieldwork does entail real ethical and professional responsibilities to the people whose language one is studying (see Greaves 1994; Newman 1992), I am troubled by the notion that researchers have an obligation to spend half of their time doing what I would call linguistic social work. I know that this is an unfashionable position nowadays, but I would argue that there is value in pure basic research and that as scientists we have to resist the everpresent pressure to justify our work on grounds of immediate social relevance. The primary justification for doing research on an endangered language has to be the scientific value of providing that documentation and in preserving aspects of that language and culture for posterity. The purpose cannot be to make the few remaining speakers feel good. Having said this - and fundamentally I do believe strongly in the correct­ ness of this viewpoint - the reality is that it is impossible to escape the practi­ cal and emotional pressures to behave like a caring human being in the field, nor would one want to (see Grinevald 1998:157). In many cases, languages are dying because communities are dying, and they are often dying because they are poor and have been neglected, if not directly exploited. The linguist who is welcomed into such a situation will either fail to establish rapport, in which case the research will be a failure, or will establish rapport, in which case he or she will increasingly acquire social and professional responsibilities that will compete for research time. The result is that the good-hearted, well-meaning linguist, to whom we can all extend our admiration, will do less of a job of basic documentation than one would have hoped for. One might argue that in the case of endangered languages, the intertwining of language preservation as a social goal and language documentation as a scientific goal is, if not beneficial, at least harmless. I think otherwise. To begin with, language preservation projects drain resources from the important linguistic task of primary documentation, both in terms of personnel and in terms of research funds. An example that for me is close to home is the important American Indian Studies Research Institute at Indiana University. Over the past years, this Institute, which is staffed by a number of technically competent and highly motivated linguists, has devoted a major portion of its efforts into designing and preparing language teaching materials in Native American languages for use in the schools. These materials (for example, the ones on Arikara) are truly impressive; they are masterfully done with beautiful

THE ENDANGERED LANGUAGES ISSUE AS A HOPELESS CAUSE

7

typography and excellent graphics, accompanied by interactive recordings. The negative side, however, is that manpower that could have been spent working on basic linguistic description of endangered languages has been occupied with what are essentially ethnic awareness, cultural heritage projects. One cannot deny that socially relevant issues like language survival/revival have more "sex appeal" than pure linguistic documentation. As a result, pres­ ervation projects are bound to be given preference by funding agencies, espe­ cially when large groups that have political visibility and can speak out for themselves are involved. A few years ago, for example, the Administration for Native Americans announced the availability of substantial grants (up to $125,000 per year for three years) in support of projects that would "promote the survival and continuing vitality of Native American languages" and en­ courage the "establishment and support of community Native American lan­ guage projects to bring older and younger Native Americans together to fa­ cilitate and encourage the transfer of Native American language skills from one generation to another . . ." (e-mail distribution from SMARTS grantline, 1998). Given the paucity of funds for linguistic research, one can understand why linguists would be attracted to apply for such grants and if successful would gladly embark on the work. But, the reality is that the Administration for Native Americans is not going to fund revival projects on essentially mori­ bund languages spoken by the last three or four octogenarians - the money is much more likely to go to support seemingly viable languages such as Navajo and Lakota. Moreover, even if funds were to be provided for work with truly endangered languages, which would be unlikely, the applied nature of the projects and the extent of commitments to the relevant communities would leave almost no time for whatever pure research the linguist might hope to carry out on the side. Once one leaves the realm of North America, there is also a troublesome question regarding the appropriateness of activist policies regarding preserva­ tion and revitalization of minority languages. A westerner who gets permission to conduct basic linguistic research in Africa (or Asia or Latin America) is a guest in someone else's country, one who has been allowed to go there for specific scholarly purposes. Language policy in fragile multi-ethic states is not a simple sociolinguistic matter; rather, it is a serious, highly contentious politi­ cal matter with which a foreigner should not become embroiled. This is espe­ cially a problem because in many cases the cause of the language endangerment is not the spread of some European colonial language, but rather the spread of large indigenous languages. For example, in many parts of Africa, "big-population, officially and functionally recognised languages ... constitute a threat to small-population languages because their speakers are the power

8

PAUL NEWMAN

brokers and decision makers. Thus they make decisions that favour big lan­ guages and threaten the very existence of small ones" (Adegbija 2001:286). If as part of its educational and economic policy, a country such as Nigeria should choose to promote its big languages (e.g., Hausa, Yoruba, and Igbo) at the expense of the "minor" ones, the western linguist who takes a "proactive" role in defense of the smaller, endangered languages is not only being pre­ sumptuous but is also being personally reckless, thereby risking the continua­ tion and success of the field research project, not to mention his or her own safety and welfare. As linguists, we can attempt to educate and inform respon­ sible persons in government, education, and business about the significance and value of linguistic diversity in their countries, but we have no right to intervene in domestic policy matters nor to undertake linguistic social work under the guise of scientific research. 4 Our non-western colleagues don't care and would be unprepared to help out even if they did We are now in the 21 st century, but generally speaking we American (and European) linguists function in many ways as if it were 1960 or 1940 or even 1920. That is to say, although languages are dying in Brazil and India and Nigeria and Indonesia, we operate as if both the problem and the solution were ours and not that of the Brazilians, the Indians, the Nigerians, or the Indone­ sians. We're way off the mark. As active practitioners in the discipline of linguistics, with all the rights and interests and responsibilities thereof, our international colleagues should have been brought into the endangered lan­ guages fold, but unfortunately they have not. Colette Grinevald (1998:151) has written: "To accept the fact that South American linguistics should be carried out as much as possible by South Americans has in fact deep implications for the way we conduct our business and the way we basically conceive of our role as linguists." One could argue whether this necessarily "should" be the case, but for very real practical mat­ ters, this has to be the case, and part of our inability to address the endangered languages problem in any meaningful way is due to the failure to recognize this point. Even if we - by which I again mean we Americans and western Europeans - had the will to carry out the needed empirical research on endan­ gered languages around the world, there is no way that we could do it because of political and economic impediments. Most scholars are fully aware of the political and social realities of working in the developing world, namely the persistent hostility to foreign researchers. In many countries, it is a major hassle to get a visa, not to mention official permission to conduct research, and even if these are forthcoming, there are problems in getting in-country coop-

THE ENDANGERED LANGUAGES ISSUE AS A HOPELESS CAUSE

9

eration and support. A more serious problem, however, is research funding: it just costs too much money for an American scholar to go abroad to carry out field research. One might be lucky in getting funds for one person to work one year on one endangered language, but who is going to attend to the other ten or twenty or thirty languages? The only way endangered languages in Africa, for example, are going to get described is if African linguists and their African students do the work. Otherwise it cannot get done. In some sense, linguistics in the African area, to which I will limit myself for purposes of the discussion, is already falling into the hands of Africans. Anyone who now attends the Annual Conference on African Linguistics (which has been going on now for over thirty years) cannot help be struck by the shift in the balance of the participants as opposed to twenty or so years ago. At that time, most of the participants were white (and white male at that!); nowadays Africans, some of whom are established scholars, many with per­ manent positions in the U.S., others of whom are Ph.D. students, generally constitute at least half of the attendees. Further evidence of the importance of Africans in African linguistic re­ search can be gathered by looking at Ph.D. dissertations produced over the past decade. A convenient source here is the list published regularly in the ASA (=African Studies Association) Newsletter, which includes theses from Canada and the United Kingdom as well as the United States. (This listing is more inclusive than that provided in Dissertation Abstracts International, which was analyzed above.) The entries from January 1990 through December 2000 provide a total of 145 theses on African languages and linguistics. Interest­ ingly, 92 (over 60%) of the theses were by Africans whereas 53 (under 40%) were by non-Africans. At first sight, these figures might appear to be promising and suggest that there is a new generation of young African scholars emerging who can under­ take the fieldwork that needs to be done. However, there is a factor of real significance for the endangered languages question that does not appear in the raw numbers. When one looks at the languages treated by the African Ph.D. students, it turns out that almost all of the theses are descriptions of the writer's own language. There is the occasional exception - and one cannot be sure in all cases since the claim is based on surmise depending on the authors' names - nevertheless, based on the many instances where the authors are easily identified, there is good reason to believe that the exceptions are rare indeed. Typically when it comes to African Ph.D. students, a native speaker of Yoruba writes on Yoruba, a Wolof speaker writes on Wolof, and a Xhosa speaker writes on Xhosa. In most cases, the writer serves almost exclusively as her or his own informant. There is nothing intrinsically wrong with native

10

PAUL NEWMAN

speakers working on their own languages - and for many purposes it is of tremendous linguistic value - but one has to face the fact that little or no fieldwork was involved in the research leading to the preparation of the Ph.D. dissertation. In effect, having said that the study of endangered languages in Africa has to be done by Africans, we find that our African colleagues are no more qualified, no more trained, and no more ready to undertake the task than the most abstract, theoretically-oriented linguist. Assuming, as I do, that this is a most unfortunate state of affairs, we have to ask, "What went wrong?" I would suggest that we western linguists have unwittingly distorted the intellectual development and orientation of nonwesterners studying in the United States (and Europe) with the result that we have excluded them from involvement in empirical descriptive linguistics in general and in the endangered languages issue in particular. For an African to write on her own language, e.g., for an Igbo speaker to do a dissertation on Igbo, is essentially the same as for an English speaker to write on English. We descriptive field linguists who have little patience with the English speaking Ph.D. student who does the umpteenth study of reflex­ ives or what have you in English - obviously in light of the latest ephemeral theory - have failed to recognize that what characterizes our work is the ex­ citement of discovery with regard to a language that is outside of ourselves, and that the Igbo linguist who writes on Igbo is not partaking of the same enterprise. Those of us who are quick to say, "Who needs another study of English?" or "Why can't that person go to the field and do something of real value such as describing a poorly known language?" never pass judgment on our African students for what they are doing. We forget that whereas Igbo may be exotic for us, it is not for the Igbo speaker. By allowing African students to work exclusively on their own languages, we fail to communicate the real nature, not to mention the importance, of fieldwork, which is essential if the person is ever going to do basic research when he returns home. In effect, we never encourage or cajole our African students who speak major languages such as Hausa, Yoruba, Lingala, or Swa­ hili that what they must do when they finish their degrees and return home is undertake the study of minority languages and, moreover, that they must pres­ sure their own students in their home universities to do the same. Since we don't require that they do first-hand research on other (small) African lan­ guages when they are students, how can we expect them to do such work later? For a variety of reasons, the foreign students attending universities and studying linguistics (whether in their own countries or abroad) are rarely members of these minority communities themselves; it is people from the

THE ENDANGERED LANGUAGES ISSUE AS A HOPELESS CAUSE

11

dominant groups who most often have these opportunities. Thus, in the ab­ sence of visionary scholars who fervently believe that language loss is indeed a culturally and intellectually catastrophic matter, language centers in Africa (and Asia and Latin America) will continue to devote their energies to the promotion and development of large national and regional languages, with scant attention to the languages that are speeding towards extinction. Apart from the matter of attitude is the fact that we here in America do not properly train our African students in fieldwork procedures. Since most of our African Ph.D. students are writing on their own languages, generally using themselves as informants, we fail to give them instruction in empirical scien­ tific methodology. They are not given solid training in phonetic transcription witness the fact that those who do not speak a tone language, and even many who do, are seldom trained in hearing and transcribing tone. They are not taught how to manage a corpus (since they are basing their theses on personal introspection) nor how to elicit and preserve primary data. In addition they are not taught how to collect and transcribe texts nor what one should do with them once they are available. In short, even if we could convince our African colleagues of the seriousness of the endangered languages question, the im­ practical Ph.D. education that we have provided them, with its heavy dose of modern linguistic theory and elegant formalism, has not equipped them to undertake the task. 5

Conclusion I am afraid that I have to close on a somber note. Despite the best intentions of many well-meaning and dedicated linguists, the rapid disappearance of languages throughout the world is likely to continue unabated. Those of us who are concerned about the endangered languages question and the problem of language extinction are up against a formidable enemy - and that enemy is our own discipline of linguistics and the individuals who make it up. We can continue to talk about the matter, as surely will be done ad nauseam at one international meeting or workshop after another; but given the odds against us, the chances of concrete results are pitifully small. References Adegbija, E. 2001. "Saving threatened languages in Africa: A case study of ". Fishman 2001. 284-308. Bach, Emmon. 1995. Posting. Linguist List (posting 12 February 1995). Brenzinger, Matthias. 1998. Endangered languages in Africa. Cologne: Rüdi­ ger Köppe.

12

PAUL NEWMAN

—, Bernd Heine & Gabriele Sommer. 1991. "Language death in Africa". Robins & Uhlenbeck 1991. 19-44. Brew, J.O., ed. 1968. One hundred years of anthropology. Cambridge, Mass.: Harvard University Press. Brooke, James. 1998. "Indians striving to save their languages". New York Times, April 9, Al, A20. Dixon, R. M. W. 1997. The rise and fall of languages. Cambridge: Cambridge University Press. Dorian, Nancy 1993. "A response to Ladefoged's other view of endangered languages". Language 69. 575-79. Fishman, Joshua A. 1991. Reversing language shift: theoretical and empirical foundations of assistance to threatened languages. Clevedon: Multilin­ gual Matters. —, ed. 2001. Can threatened languages be saved? Clevedon: Multilingual Matters. Geertz, Clifford. 1984. "Distinguished lecture: anti anti-relativism". American Anthropologist 86. 263-78. Greaves, Tom, ed. 1994. Intellectual property rights for indigenous peoples: a sourcebook. Oklahoma City: Society for Applied Anthropology. Grenoble, Lenore A. & Lindsay J. Whaley, eds. 1998. Endangered languages: current issues and future prospects. Cambridge: Cambridge University Press. Grinevald, Colette. 1998. "Language endangerment in South America: a pro­ grammatic approach". Grenoble & Whaley 1998. 124-59. Gugliotta, Guy. 1999. "Saying the words that save a culture". Washington Post, August 9, Al A8. Hale, Kenneth et al. 1992. "Endangered languages". Language 68. 1-42. Hyman, Larry M. 2001. "Fieldwork as a state of mind". Newman & Ratliff 2001. 15-33. Krauss, Michael. 1992. "The world's languages in crisis". Language 68. 4-10. Ladefoged, Peter. 1992. "Another view of endangered languages". Language 68.809-11. Lounsbury, Floyd G. 1968. "One hundred years of anthropological linguis­ tics". Brew 1968. 153-225. Matsumura, Kazuto, ed. 1998. Studies in endangered languages. Papers from the international symposium on endangered languages, Tokyo, Novem­ ber 18-20, 1995. Tokyo: The International Clearing House for Endan­ gered Languages, University of Tokyo. Mithun, Marianne. 1998. "The significance of diversity in language endan­ germent and preservation". Grenoble & Whaley 1998. 163-91.

THE ENDANGERED LANGUAGES ISSUE AS A HOPELESS CAUSE

13

Mufwene, Salikoko S. 1998. "The ecology of languages: new imperatives in linguistics curricula". Studies in the Linguistic Sciences 28:2. 135-45. —. 2001. The ecology of language evolution. Cambridge: Cambridge Univer­ sity Press. Newman, Paul. 1992. "Fieldwork and field methods in linguistics". California Linguistic Notes 23:2. 1-8. — & Martha Ratliff, eds. 2001. Linguistic fieldwork. Cambridge: Cambridge University Press. Raymond, Joan. 1998. "Say what? Preserving endangered languages". News­ week, September 14, 14. Robins, Robert H. & Eugenius M. Uhlenbeck, eds. 1991. Endangered lan­ guages. Oxford and New York: Berg. Said, Edward W. 1978. Orientalism. New York: Pantheon Books. Vaux, Bert & Justin Cooper. 1999. Introduction to linguistic field methods. Munich: LINCOM Europa.

THE LANGUAGE SITUATION AND LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA STEPHEN A. WURM Australian National University

1 General remarks The term 'Greater Pacific area' is to be understood as comprising the entire world of very large and smaller to very small islands in the Pacific Ocean including marginal parts of it such as the China Sea, the Tasman Sea, etc. Marginal areas in the north, i.e. Japan, the Kuril and Aleut Islands, will receive short mention in an appendix to the main body of this contribution which cov­ ers Taiwan, the Philippines, Indonesia, Papua New Guinea, the Solomon Is­ lands, Vanuatu, New Caledonia, Fiji, Micronesia, Polynesia and Australia. The main part of the island world of the Greater Pacific apart from Austra­ lia, contains two very large groups of languages. One of them is the vast Austronesian group which consists of about 1200 genetically interrelated languages and covers most of the island world of the Greater Pacific area - one of the largest and geographically very widespread language groups of the world. The other large group is that of the 838 Papuan languages which cover the greater part of the large island of New Guinea and some parts of the large islands adjacent to it in the west and east such as parts of Timor, most of the Alor and Pantar Islands, the northern half of the Halmaheras, and New Britain, New Ireland and Bougainville Islands in the east. There are also some Papuan lan­ guages in the Solomon Islands chain as far as the Santa Cruz Archipelago which is its easternmost extension. Of the 838 Papuan languages, 702 or per­ haps more are members of five major groups of apparently genetically interre­ lated languages with a sixth eastern geographical group probably constituted by several small groups of languages, with these small groups not related to each other, but containing a total of thirty-four languages. The five large groups are not related to each other, except for two of them which may possibly be. The smaller groups, scattered in other parts of the Papuan language area, including the sixth eastern geographical group mentioned above, number nineteen in total, and thirteen additional isolated languages form the remainder of Papuan languages. The Austronesian and Papuan languages number together about 2000. When adding the about 300+ still surviving or recently extinct Australian languages to this, the 2300+ languages in the Greater Pacific area are well over

16

STEPHEN WURM

a third of the about 6000 languages in the world. This area is therefore one of the geographical areas of the world richest in languages. At the same time, the Greater Pacific area is, except for Australia and to a lesser extent Taiwan, New Caledonia, New Zealand and Hawaii which are recovering or have recovered to some extent, the part of the world until now least affected by the worldwide phenomenon of language endangerment and death. Of the 2000 or so Austronesian and Papuan languages, only forty-nine languages (thirty-eight Austronesian and eleven Papuan) have become recently or relatively recently extinct, and about 305 are threatened, with fifty-nine of them seriously or terminally endangered. Details of this are given below in the Austronesian and Papuan section, and in regions in which Austronesian and also Papuan languages are found, details on the extinction and endangerment of the latter are given alongside those of Austronesian languages of the same areas. Language endangerment leading to the eventual extinction and disappear­ ance of languages constitutes the gradual disappearance of the speakers of a language, usually beginning with children, continuing with young adults, mid­ dle-aged speakers, aged speakers, until only a few very old speakers are left with whose death the language becomes extinct. The main reason for this is the switching of the speakers to another, usually dominant, language under cultural pressure. If a proportion of the children starts giving preference to another language and gradually forgets their own, their own language is potentially endangered; if the youngest speakers are young adults, the language is endan­ gered', if they are middle-aged, the language is seriously endangered', and if there are only a few old speakers left, the language is moribund (or terminally endangered). 2 The Austronesian language group (background information) 2.1 Past migrations of the Austronesians and the history of their languages Judging from the results of historical and comparative linguistics, the Aus­ tronesian languages had their oldest home in Taiwan from where their speakers migrated southwards from about 7000 years ago. The most archaic Austrone­ sian languages are still found in Taiwan today, where they constitute three branches of the Austronesian language group, which comprise together twentythree living, recently extinct and long extinct languages of which fourteen are still spoken today. All other Austronesian languages, close to 1200, belong to a fourth Austronesian group, named Malayo-Polynesian. From Taiwan, the speakers of languages of this fourth group migrated to the Philippines and from there to various western parts of Indonesia and into Malaysia. In some areas,

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA17 they reached parts of continental South-East Asia, and also Hainan Island. About 1000-2000 years ago, a group speaking an ancestral form of the Ma'anyan language of southeastern Borneo, migrated westwards, and eventu­ ally reached the Madagascar Island off the southeastern coast of Africa via India. Their descendants today are the Malagasy speaking a daughter language of ancestral Ma'anyan, which is the Malagasy language of today, with 11 million speakers. One part of the Malayo-Polynesians moved eastwards and reached areas west of New Guinea about 5500 years ago. Malayo-Polynesians continuing eastwards from there seem to have reached the large islands of New Britain and New Ireland. There the Malayo-Polynesians had contact with dark-skinned Papuans and intermarried with them. Later they migrated from there to the north shores of New Guinea and its small off-shore islands. A part of them migrated further east to the western Solomon Islands, and also reached the islands of Vanikoro and Utupua in the Santa Cruz Archipelago, and southern Vanuatu, the Loyalty Islands, and New Caledonia. It seems that they settled in those areas about 4500-4000 years ago. Most of them are dark-skinned. Ma­ layo-Polynesians who stayed west of the New Guinea area remained lightskinned. A part of those moved eastwards, probably after the east and southeast migration of the dark-skinned Malayo-Polynesians, and settled on the south­ eastern half of the Solomon Islands, and on most of the northern and central islands of Vanuatu. They must have had very much less contact with Papuans and have generally remained more light-skinned than the dark-skinned Ma­ layo-Polynesians. A part of them occupied the Micronesian Islands world, with the exception of Guam, the Chamorro-speaking Mariana Islands, the Palau Islands, and the island of Yap, where today there are speakers of Western Malayo-Polynesian languages. The lighter-skinned Malayo-Polynesians also settled in the easternmost part, and a portion of the south-eastern coast, of the New Guinea mainland. They remained more light-skinned than their Papuanspeaking neighbours, and their languages are similar to those of the lighterskinned Malayo-Polynesians of the south-eastern Solomon Islands, though some of them show Papuan linguistic influence in their languages. The Ma­ layo-Polynesian languages spoken by the lighter-skinned Malayo-Polynesians differ in various ways from those spoken by the dark-skinned MalayoPolynesians mentioned above, and have developed towards the Remote Oce­ anic type of Oceanic Malayo-Polynesians (see below). About 3500 years ago, lighter-skinned Malayo-Polynesians migrating from central Vanuatu reached the islands of Fiji and Roturna. Much later, a migration of dark-skinned Ma­ layo-Polynesians reached the Fiji Islands, resulting in the present-day Fijians being in part dark-skinned and curly-haired though otherwise their physical

18

STEPHEN WURM

anthropological setup is close to that of the Polynesians. Before this last mi­ gration, a major migration from the Fiji Islands reached the easternmost islands of present-day Polynesia, i.e. Tonga and Niue, about 3000 years ago, and Sa­ moa a little later. On these islands, the proto-Polynesian language developed. About 2500 years ago, a number of westward migrations from the Samoan islands took Samoan-type Polynesian languages to central Vanuatu, into the Santa Cruz Archipelago, to far-flung small islands adjacent to the main Solo­ mon Islands chain, and to the Wallis and Futuna Islands lying between Samoa and Roturna. In the north, such languages reached the islands of Tuvalu, Tokelau and Pukapuka. About 2000 years ago, major eastward migrations reached the Cook Islands, Tahiti, the Tuamotu Archipelago, and the Marquesas Islands, where the Eastern Polynesian proto-language developed. About 1000 years ago, westward migrations from the Cook Islands and Tahiti reached New Zealand where the Maori language arose. About the same time, northward migrations from the Marquesas Islands reached the islands of Hawaii, where the Hawaiian language developed. The development of the Austronesian language picture is thought to have been as follows: from proto-Austronesian, the Taiwanese languages and protoMalayo-Polynesian developed. From the latter, the Western Malayo-Polyne­ sian languages and the Central and Eastern Malayo-Polynesian proto-language originated. From this, the Central Malayo-Polynesian languages and the East­ ern Malayo-Polynesian proto-languages developed. From the latter, the South­ ern Halmahera Islands and Northeast New Guinea Malayo-Polynesian lan­ guages and the Oceanic proto-languages were derived. The present-day Ma­ layo-Polynesian languages of Melanesia, Micronesia and Polynesia are daughterlanguages of the latter. 2.2 Classification of the Austronesian languages A summary classification of the present-day languages derived from protoAustronesian is as follows: 1) Atayalic; 2) Tsouic; 3) Paiwanic (these are all Taiwanese language groups); 4) Malayo-Polynesian languages. The latter consist of two large groups: A) the Western Malayo-Polynesian languages; B) the Central-Eastern Malayo-Polynesian languages. Group B) comprises two large subgroups: 1) the Central Malayo-Polynesian languages; 2) the Eastern Malayo-Polynesian languages. The latter has two member groups: a) the Southern Halmahera Islands and Northwestern New Guinea; b) the large Oce­ anic group which consists of 18 member subgroups. The 18th of these, which is called Remote Oceanic, has four sections: Micronesian, Southeastern Solo­ mons languages, Central and Northern Vanuatuan, and Central Pacific. The last

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA19 of these has two sub-sections: one is the Rotuma-Fijian, and the other com­ prises all the Polynesian languages. 3 Endangerment of Austronesian languages (with statistics on endanger ment of Papuan languages where relevant) 3.1 Introductory remarks The Greater Pacific area is the part of the world least affected by language endangerment and language death, considering the very large number of lan­ guages in it. The only exception to this are Australia, which is one of the most heavily affected in the world, and New Caledonia - both of them because of the overwhelming presence of an outside dominant population with a European metropolitan culture, speaking English and French respectively. New Zealand and Hawaii had moved a very long way in the same direction for the same reason, but both have recovered to some extent, their once almost extinct local indigenous languages having been reinvigorated and now being protected and maintained up to a certain point. Indonesia is rapidly moving in the direction of getting a bad record for language endangerment, whereas New Caledonia and Australia are making efforts to improve their bad record, though this is almost too late as far as Australia is concerned (cf. infra). Taiwan also had a bad rec­ ord, but changed its policies to positive and supportive ones a few years ago. The nature, progress and characteristics of language endangerment are not the same in the various parts of the Austronesian world. The same applies to the reasons for this endangerment, and to factors favourable for the mainte­ nance or reinvigoration and revival of the languages (cf. Wurm 2001b). There­ fore, the various regional and political parts will be treated separately from north-west progressing far to the east. 3.2 Taiwan From 7000 years ago to around 1600 AD, Taiwan was entirely Austronesian-speaking with twenty-three Austronesian languages known to have existed there. Eight of these are now extinct, three of them recently or relatively re­ cently, and six of the surviving fifteen languages are moribund. The first reason for this decline of the Taiwanese Austronesian languages was a strong immi­ gration, beginning around 1600 AD, of Chinese from southern China who spoke Fukien (Hokkien) Chinese. They occupied the western lowland areas of Taiwan, which resulted in a gradual sinicization of the Austronesians who lived there. In the 1950s, a further strong influx of Chinese took place, resulting in Mandarin Chinese becoming the official language of Taiwan. The pressure of the Chinese language on the Austronesian Taiwanese languages increased

20

STEPHEN WURM

further, and now all the former lowland Austronesian languages are extinct or almost extinct, with the remaining highland languages receding further. Gov­ ernmental policies towards the Austronesian languages were negative until a few years ago, and contributed to their gradual demise. However, during the last few years, a strong reversal of governmental language policies took place (in line with developments in other parts of the world such as Australia, Can­ ada, Japan, Scandinavia, Russia, England, Paraguay etc.), and the government is now actively supporting the maintenance, reinvigoration and revitalization of the Austronesian minority languages. 3.3 The Philippines All of the 165 Philippine languages are Austronesian languages belonging to the Malayo-Polynesian branch. There is a very uneven distribution of speaker numbers: ten very large languages have together 50 million speakers, i.e. about 90 percent of the total Philippine population. Tagalog, the national language officially named Pilipino, has 10 million first-language and 15 mil­ lion-second language speakers, and is understood by a total of over 40 million. There is widespread bilingualism, which constitutes a good bulwark against the loss of minority languages whose speakers simply add the knowledge of one of the large lingue franche and/or of Tagalog to their repertory of languages, without losing the others. The absence of language policies hostile to minority languages also assists the maintenance of small minority languages. Only three, perhaps now four, small languages have become extinct, and fourteen other small languages are threatened, due to pressure from large neighbouring lan­ guages. 3.4 Indonesia 3.4.1 Introduction Indonesia contains very many small languages and a number of large lan­ guages. The national language of Indonesia, Indonesian (Bahasa Indonesia), is based on the high form of Malay used by the former Dutch administration as the administrative language of what is today Indonesia. Indonesian is in part a simplified and in part a developed form of that language. It was nobody's first language when adopted as the national language of the newly created state of Indonesia, and it was chosen to prevent linguistic jealousies within Indonesia, which would certainly have erupted immediately if a large local language, e.g. Javanese (which has over 80 million speakers) had been chosen as the national language. Indonesian has now about 7 million young first-language speakers. These sometimes also know the local language of their parents, but usually

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA21 prefer Indonesian even when talking to them. The pressure of Indonesian, especially on not very large, and smallish to very small local languages, is very great and leads to their endangerment. Indonesian is officially used everywhere as the only language of education, and the pupils are indoctrinated to use it even when returning to their homes and families. The use of local languages for any purpose outside the family level is strongly discouraged, though with some of the very large languages, i.e. Javanese, Madurese, Sundanese, Batak on Sumatra, Gorontalo on Sulawesi, etc. some unofficial use for educational purposes occurs. The Indonesian language policies are similar to those for­ merly applied by monolingual speakers of metropolitan languages in Europe, the USSR, Australia, Canada, the USA, South American countries, Taiwan, Japan etc., but these policies have turned to more positive ones in the last decades or years in most of these countries except for at least some of these negative policies continuing in the USA, parts of Europe, some South Ameri­ can countries, etc. The effects of the negative Indonesian policies on local and minority languages are so far less devastating than those of comparable previ­ ous policies in some of the countries mentioned above. The reasons for this are especially the much shallower time depth of the existence of such policies in Indonesia when compared with those practised before (and in part continuing today) in the countries mentioned above as examples. Also, widespread multilingualism in especially eastern parts of Indonesia, and strong resistance to Indonesian pressure in some areas, slow down the effects of Indonesian lan­ guage policies. In addition, the international Summer Institute of Linguistics (SIL) organization has established some literacy and produced small educa­ tional publications especially in several smallish languages in West Papua and some other parts of eastern Indonesia. However, even very large languages are gradually losing out to Indonesian, especially among members of the young generation who have been indoctrinated during their official education. 3.4.2 Borneo For Borneo, which is politically split between Malaysia (with Sarawak and Sabah) and Indonesia (with Kalimantan), and has a total of about 140 lan­ guages, an astonishingly low number of one extinct and five threatened lan­ guages has been reported. This is probably due to a very incomplete knowledge of the language situation in some parts of Borneo, especially of western parts of Kalimantan. The number of threatened languages is likely to be considerably higher, especially when taking into account the negative attitudes of the Indo­ nesian government towards local languages in other parts of Indonesia as men­ tioned above. The reason of language death and endangerment in Borneo is

22

STEPHEN WURM

pressure from Indonesian in Kalimantan, and from Malay in Sarawak and Sabah, and also from some large local languages in both areas. 3.4.3 Sumatra There are very few languages on Sumatra which would be likely targets for endangerment from Indonesian and/or large local languages. One language on one of the islands located off the south-western coast of Sumatra is endan­ gered. Another one on an island off the northeastern coast of Sumatra has been reported to be moribund, or even already extinct. 3.4.4 Sulawesi The endangerment situation on Sulawesi is considerably worse than on the large more western islands of Indonesia, as far as is known (see Borneo). Of the 114 languages on Sulawesi, which are all Malayo-Polynesian, one is defi­ nitely extinct, and another is also possibly extinct now, both under pressure from large neighbouring languages, and from Indonesian. Thirty-six other languages are endangered to a varying extent, and several are moribund. One problem is the language of the Bugis who settled in many parts of Sulawesi as traders, and whose language is an important lingua franca which puts pressure on small languages, in addition to the general pressure exercised by Indone­ sian, mainly on the children and teenagers through the educational system, but also otherwise. There is little to stimulate the maintenance or reinvigoration of Sulawesi languages, except the high level of culture and civilization of speak­ ers of several large languages in various parts of Sulawesi who are very proud of their languages. This is of no help to culturally less sophisticated speakers of smaller languages whose only defence against pressure from large languages and their speakers is their multilingualism, which is beginning to be breached by the exclusive use of the Indonesian language in education, administration and all aspects of public life, and is made attractive to children and the young population through Indonesia's language policies and political pressure. 3.4.5 Maluku Maluku covers the world of small and more largish islands situated between Sulawesi and the New Guinea area in the east and the Timor-Flores and BimaSumba areas in the south. It comprises regions known as Central Maluku, southeast Maluku, and also other areas in the north, northeast and south. There are 102 languages in Maluku, many of them small, some larger, with thou­ sands, and tens of thousands of speakers. Eighty-seven of these are MalayoPolynesian, and fifteen are Papuan. Of the Malayo-Polynesian languages, ten are extinct, three moribund, nine are beginning to get endangered, and eleven

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA23 are endangered or seriously endangered. Of the Papuan languages, one is probably extinct, and another endangered. Therefore, of the 102 languages in the Maluku area, eleven are extinct, and twenty-four threatened to various degrees. The main reason for the extinction and endangerment situation has been and continues to be the pressure from Indonesian which causes even languages with large numbers of speakers to become endangered, like the Malayo-Polynesian Alune language on Seram Island, which has over 12,000 speakers. This is large in an area where very many languages have only a few hundred speakers, and others a few thousand. Some languages succumbed or became threatened as a result of special circumstances. For instance, the large island of Buru has been used as a place of deportation of criminals which has led to the extinction of a number of small speech communities there. Other speech communities were transferred by the Indonesian authorities to areas far removed from their homeland, often with disastrous consequences. For in­ stance, the speakers of the Malayo-Polynesian Nila language who lived on a volcanic island deep in the south of the Maluku area, were transferred by the Indonesian authorities to the south-central part of the large Seram Island, the main island of Central Maluku. Diseases and adverse conditions associated with their resettlement have taken their toll and led to a reduction of the num­ ber of the speakers of the Nila language which is now potentially endangered and heading towards being endangered (cf. van Engelenhoven, this volume). 3.4.6 Timor-Flores and Bima-Sumba areas These two areas lie to the south and southwest of the large Maluku area. They comprise West Timor, the Alor and Pantar Islands, the Lombon and Flores Islands, and the Roti and Sumba Islands. They are known as Nusa Tenggara on Indonesian maps. There are twenty-two Malayo-Polynesian and thirteen Papuan languages in them, totalling thirty-five languages. Of the Ma­ layo-Polynesian languages, one is potentially endangered, and one is endan­ gered. Only one Papuan language is potentially endangered. Indonesian pres­ sure is relatively weak in much of the area, where there is fairly strong resis­ tance to its influence, and most languages are quite large and less vulnerable. 3.4.7 West Papua (formerly Irian Jaya) West Papua occupies the western part of the very large island of New Guinea, and several of the islands lying west of the western tip of the Bird's Head Peninsula of West Papua, such as the Raja Ampat Islands and Waigeo Island. There are just over 200 languages in West Papua, forty-one of them Malayo-Polynesian, and over 160 Papuan. Many of the languages are small to very small, but under the Dutch administration before the Indonesian takeover,

24

STEPHEN WURM

there was very little pressure on local languages and they were not threatened. The local people had, and still continue to have, a very strong sense of ethnic identity. They regard their local languages as an important symbol of it. How­ ever, this has been insufficient to protect them from the relentless pressure of the Indonesian language which operates through the children. In addition, under the strong transmigration programme of the Indonesian government, a large number of immigrants were resettled from overpopulated Java to West Papua to the disadvantage of the local population. Of the forty-one MalayoPolynesian languages one is extinct, and sixteen threatened. Of these, eight are potentially endangered, seven endangered or seriously endangered, and one moribund. Of the over 160 Papuan languages, none is extinct yet, though one or the other of six moribund ones is likely to be extinct very soon. Twenty-one are portentially endangered, twenty-two endangered or seriously endangered, and six moribund. Therefore, of the a little over 200 languages in West Papua, one is extinct and sixty-five threatened to a varying extent. There are enough reasons for languages in West Papua getting threatened, and unfortunately inadequate and insufficient reasons for the maintenance and reinvigoration of threatened ones. There is a strong and increasing independence movement in West Papua, which, if ever successful some time in the future, should provide a better climate for the maintenance and reinvigoration of the local languages than the present negative one. For the time being, only the efforts by the SIL organization constitute a positive step towards the maintenance of local lan­ guages in eastern parts of Indonesia. 3.4.8 East Timor On East Timor, which has recently been separated from Indonesia, the situation is different from the remainder of the Timor area. As a former Portu­ guese colony, it was occupied by Indonesia in the 1970s, with cruel oppression by Indonesia lasting until early 1999, when in a UNO-sponsored vote, 78 percent of the population voted for independence from Indonesia. This was followed by an even more cruel period of devastation by the Indonesianbacked and trained militia with widespread massacres and the destruction of many settlements, until a UNO military contingent under Australian leadership took control of East Timor and ousted the militia from it. Until then, the militia had driven out over 200,000 East Timorese into Indonesian West Timor, into militia-controlled camps where they lived under awful conditions, and were prevented by the militia from returning to East Timor. A number of them were eventually able to go back to East Timor, but over 100,000 still remained. Their living conditions became even worse after the militia murdered three UNO relief workers active in West Timor upon which the UNO withdrew all

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA25 their relief workers, and made the dissolution and disarmament of the militia a condition for their return. Indonesia continued to make excuses for not doing so until the time of writing this in 2001. Before these catastrophic events, the linguistic situation in East Timor was as follows: there were seven Malayo-Polynesian and four Papuan languages, with one of the latter of doubtful status - its structure shows Papuan features, but much of its lexicon is distorted Malayo-Polynesian. It is probably an origi­ nally Papuan language, with strong Malayo-Polynesian influence. One of the Malayo-Polynesian languages was potentially endangered and one of the Pap­ uan languages (the doubtful one) was moribund. It was unclear after the cata­ strophic events mentioned above whether some language(s) may have become extinct as a result of those. A linguistic survey undertaken at the end of 2000 showed that this was fortunately not the case, but a decrease in speaker num­ bers was recorded. The reason for the endangerment of the two languages has been pressure from large neighbouring languages in the first place, with Indonesian pressure during the cruel reign of Indonesia over East Timor as an additional reason. With the independence of East Timor, this second reason has no validity any longer, though the first one continues. 3.5 Papua New Guinea Papua New Guinea occupies the eastern part of the large island of New Guinea, and several of the large and small islands lying to the east, north-east and north of it. With approximately 850 languages, of which 231 are MalayoPolynesian and well over 600 Papuan, it is the most polylingual area of a com­ parable small size in the world. Many of the languages are small to very small: fifty-six have less than 100 speakers, 225 between 100 and 500, 134 between 500 and 1,000 speakers, 330 between 1,000 and 10,000, but mostly in the low or very low thousands, and only seventy-two more than 10,000. The speaker numbers of about thirty other languages are unknown (Nekitel 1998). There­ fore, nearly half of the Papua New Guinea languages have less than 1000 speakers! At the same time, until about twenty years ago, Papua New Guinea had the lowest level of language endangerment of all those areas in the world which contained many small languages. The reasons of this were the following: 1) English, the metropolitan language of the colonial power Australia, had a very low role because of the very low number of native speakers of English, and the almost exclusive use of the main lingua franca, Tok Pisin (formerly called New Guinea Pidgin) in contacts between members of the colonial ad­ ministration and locals, and between Papua New Guineans who did not have a language in common, in spite of the widespread multilingualism in the country.

26

STEPHEN WURM

In the southern part of the country, another lingua franca, Police Motu (Hiri Motu) played a part, and there were some other smaller lingue franche in some parts of the country. Only a small number of Papua New Guineans had a good command of English, and their number is still not very significant today: less than 200,000 at best out of close to 5 million. 2) Papua New Guineans regarded their local languages as the main symbol of their local identities, which they held very dear. 3) The main lingua franca Tok Pisin which was known by about half the population over twenty years ago (now it is by far more than two thirds of it), was no threat to the local languages then. There was, and still is, very wide­ spread multilingualism, and the locals simply added a knowledge of Tok Pisin, and of other lingue franche, and occasionally also of English, to their repertory of languages without endangering their own language(s), as this is usually the case in many parts of the world when speakers of a local minority language acquire a good knowledge of a dominant language, and lose their own. 4) There was little mobility of the local population, and intermarriage between persons who did not have a local language in common within the range of their multilingualism, was rare and occurred only in the major towns, where Tok Pisin became the family language, constituting the initial creolization of it. The very few languages which became endangered or extinct until about twenty years ago, all succumbed to pressure of large neighbouring languages, except for the Admiralty Islands lying to the north of the Papua New Guinea mainland, where an already partly creolized Tok Pisin exerted pressure on some local languages along with such pressure by large local languages. This favourable situation started changing about twenty years ago, after Papua New Guinea became independent in 1975. The pride of many Papua New Guineans, especially young ones, in their local languages as the main symbol of their ethnicity weakened, young speakers of many small and even larger languages started to prefer using Tok Pisin instead of their own lan­ guages, and many small and some quite large languages, began to be endan­ gered. These unfavourable developments were caused by the following events: 1) Rapidly increasing mobility of the population, with many locals leaving their villages and moving to population centres, mainly large towns. 2) Intermarriages between speakers of different languages outside the ranges of their traditional multilingualism have become more and more frequent in an increasing number of parts of the country. 3) Electronic media was becoming increasingly important, with the use of Tok Pisin, Hiri Motu, fifteen large local languages, and some smaller languages in some areas, as well as English in some programmes. This means that about twenty-five major and local languages are becoming prestigious, and 820 or so

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA27 are neglected by the media, and are losing their prestige even more than through what has been said above under 1). 4) Until the 1950s, local languages and mission and church lingue franche played a major role in elementary education. Later, the Australian administra­ tion used English more and more, with poor results, because of the ignorance of most elementary school pupils of English, and it discouraged the use of local languages and especially of Tok Pisin in elementary education. At the same time, the international SIL organization, became highly active in Papua New Guinea as from the mid-fifties. Close to 300 mostly not very small languages have been thoroughly studied by its members until now, with the speakers becoming literate in their languages. The SIL produced copious literacy materials in these languages, including booklets for use in elementary education, in addition to their translating the Scriptures into these languages. After the independence, the small Englisheducated elite tried to continue the ill-conceived education-in-English-only of the former Australian administration, with meagre results. Some time later, the Papua New Guinea government recognized the importance of local traditional cultures and languages, and more and more provincial local language prepara­ tory schools were established in the various provinces, numbering about 100 in the early 1990s. Programmes in well over twenty languages were used in them. This, and the efforts of SIL resulted in more use of local languages by children (Mühlhäusler, Philpott & Trew 1996), but still left out over 500 languages from elementary education, with a resulting lowering of the prestige of such lan­ guages in the eyes of Papua New Guineans. To remedy this, there is a proposal that in an elementary school with multilingual children with several mothertongues, in which the chief language of education and literacy is one of the languages used as mentioned above, the pupils be split into various mothertongue groups and be spoken to for an hour or so by a suitable person in their respective mother-tongues, giving them purely oral information about their culture, and other useful subjects such as health matters, gardening etc. This would contribute to the maintenance of their languages, and raise the respect for them in the eyes of the children (Wurm 2001). Of the Papua New Guinea languages, sixteen are now extinct, largely as a result of pressure from larger neighbouring languages. Nine of these are Malayo-Polynesian, and seven Papuan. Seventy-seven other languages are threat­ ened to a varying extent, or moribund. Of these, thirty are Malayo-Polynesian, and forty-seven Papuan. Of the thirty Malayo-Polynesian languages, eleven are potentially endangered, and nineteen endangered or seriously endangered. Of the forty-seven Papuan languages, nine are potentially endangered, thirty-one endangered or seriously endangered, and seven moribund. From what has been

28

STEPHEN WURM

said above in this Papua New Guinea section, the reasons for the extinction of sixteen and endangerment of seventy-seven languages in Papua New Guinea should be clear. Similarly, factors contributing to the maintenance and reinvigoration of threatened languages have been mentioned. It may be added that efforts are being undertaken by the University of Papua New Guinea to reverse the progress of language endangerment, especially aiming at persuading young Papua New Guineans to use their mother-tongues, not Tok Pisin, when talking to speakers of their own language. This has had varied results. 3.6 Solomon Islands 3.6.1 Main Solomon Islands Chain The main Solomon Islands Chain extends southeastwards from the eastern­ most large island of Papua New Guinea which is Bougainville Island. Fortytwo Malayo-Polynesian languages and four Papuan languages are located on islands. Three other Papuan languages formerly spoken on one of the islands are now extinct. Nine of the Malayo-Polynesian languages and one of the Papuan languages are threatened. Of these nine Malayo-Polynesian languages, three are potentially endangered, four are endangered, one is seriously endan­ gered, and one moribund. The threatened Papuan language is endangered. The main causes of the endangerment of languages of the main Solomon Islands Chain have been several large indigenous lingue franche. These were formally mission lingue franche. As such, cultural and deliberately directed pressure by them on other languages has been greater than that of non-mission lingue franche and trade languages in the main Solomon Islands Chain which are included in the repertory of multilingual speakers of local languages, and do not exert cultural and directed pressure. There is also increasing pressure from the general Solomon Islands lingua franca, the English-based Solomon Pijin which is different from the Tok Pisin of Papua New Guinea. The attitudes and policies of the Solomon Islands government toward local languages are essen­ tially negative, and the only factor favouring language maintenance and reinvigoration in the Solomon Islands is the activity of the SIL organization there. 3.6.2 The Santa Cruz Archipelago The Santa Cruz Archipelago is politically a part of the Solomon Islands, but is geographically situated far to the east of the main Solomon Islands Chain. In the Archipelago, three not endangered Papuan languages and originally six smallish Malayo-Polynesian languages are located. Two of the latter are now extinct, and three more are threatened. Of these three, two are endangered and one seriously endangered. Their extinction and endangerment is attributable to

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA29 pressure from the sixth not yet endangered Malayo-Polynesian language and of Solomon Pijin. There are no factors favouring the maintenance and reinvigoration of these threatened languages. 3.7 Vanuatu All Vanuatuan languages are Malayo-Polynesian. About 100 of them are living languages. The introduction of ravaging new diseases in the early years of the twentieth century, especially smallpox, influenza and leprosy, decimated the population and greatly reduced the number of the speakers of individual languages, but led to the extinction of only twenty languages or so. In living memory, three more languages became extinct, and thirty-one are now in vari­ ous stages of endangerment. It seems likely that what is usually listed as the Southwest Bay language contains two different languages in addition to what is now recognized as the moribund Naati language, one probably moribund and the other perhaps endangered. Of the thirty-one threatened languages men­ tioned above, twelve are potentially endangered, eight endangered, five seri­ ously endangered, and six moribund. The threatening factors are the fact that these languages, which are mostly very small languages, are under pressure from larger neighbouring languages, and from the lingua franca and national language Bislama, an English-based pidgin with a Melanesian MalayoPolynesian type grammar, and predominantly English-derived vocabulary. There is not much to encourage maintenance and reinvigoration of threatened languages, though speakers of some of them take an active interest in it. The SIL organization is not as active in Vanuatu as in the New Guinea area and in the Solomon Islands.

3.8 New Caledonia As in Australia and New Zealand, New Caledonia is an area in the south­ western Pacific in which the majority of the inhabitants are monolingual speakers of a dominant metropolitan language, viz. French. That language and the culturally aggressive attitudes of its speakers has resulted in only about 20,000 of the 55,000 local Melanesians today knowing one or several of the aberrant Malayo-Polynesian languages of the island. Formerly, and in part still today, there was and is stable multilingualism in New Caledonia, but it was not adequate to protect the local languages against the very strong pressure of French and the negative attitudes of its speakers. However, a while ago a strong reawakening of indigenous ethnic identity feeling has taken place in New Caledonia which has led to the revitalization of some local languages, and also to a positive change of French attitudes. However, contrary to the former

30

STEPHEN WURM

stable multilingualism and mutual respect for indigenous languages in the local population, small languages are now subject to pressure from larger local languages in addition to some continuing pressure from French. Of the total number of thirty languages two are now extinct and thirteen threatened to a varied extent, two of them being moribund. Of the remaining eleven threatened languages, three are potentially endangered, four endangered, and four seri­ ously endangered. There is now a strong interest in the local Melanesian population in the maintenance and reinvigoration of their languages, and one of the two extinct languages has been revived to some extent. 3.9 Fiji Area There are two local Malayo-Polynesian languages in the Fiji area. One of them is the dialectally strongly varied Fijian, and the other Rotuman. Neither of them is threatened. 3.10 Polynesia There is a total of thirty-seven Polynesian languages in the Pacific area, all of them Malayo-Polynesian. Sixteen of them are located outside geographical Polynesia, four of them in Vanuatu, eleven of them in main Solomon Islands Chain and in the Santa Cruz Archipelago, and one in New Zealand. Seven Polynesian languages are threatened. Of these, two are potentially endangered, four endangered, and one seriously endangered. The state of the threatened Polynesian languages is so special, that some explanation is needed. Hawaiian was almost extinct about twenty years ago, under the pressure from Pidgin English and English, with the exception of Niihau Island in the northwest of the Hawaiian Islands. No visitors were allowed on it, because it has been a private property for a very long time. About two decades ago, a strong revival of Hawaiian started. It has now over 2,000 young speakers, but no parent speakers. This, and American monolinguals' attitudes, makes its ultimate survival somewhat doubtful. It is still regarded as endangered. The Maori language in New Zealand was also close to extinction at the beginning of the twentieth century. It has been strongly revived after that and has many thou­ sands of speakers. In recent years, young Maoris have been losing interest in it, which puts its survival into doubt. It is regarded as moving towards being endangered. In French Polynesia, Tahitian was strongly receding before French in the capital Papeete. However, during the last ten years or so, it has been strongly reinvigorated and is now threatening three languages in the Tu'amotuan, Marquesan and Austral Islands. One language there is now po­ tentially endangered, one is endangered, and one is moving towards being seriously endangered. The Polynesian languages in the Solomon Islands and

LANGUAGEENDANGERMENTINTHEGREATERPACIFICAREA31 the Santa Cruz Archipelago are isolated from the pressure of large languages because of their remote locations, and are not threatened. Of those in Vanuatu, Emae, is potentially endangered under the influence of a Melanesian MalayoPolynesian language. The Rapanui language of Easter Island became almost extinct early in the twentieth century, as a result of the removal of most of its speakers to South America to collect guano. Very few survived this and re­ turned to their island, but the speaker numbers increased again and reached 2,500. However, the young generation is no longer interested in it and it is getting endangered. 3.11 Micronesia Of the twenty-one Malayo-Polynesian Micronesian languages, one which is outside geographical Micronesia, north of West Papua, has become extinct under pressure of the large Malayo-Polynesian Biak language. Of two other Micronesian languages also located outside Micronesia in the Mariana Islands, and in Palau, one is endangered and the other seriously endangered under pressure from the Western Malayo-Polynesian Chamorro and Palau languages. 4 The Papuan languages Information on the location and distribution of the various unrelated groups of Papuan languages was given in section 1. Statistical information on the endangerment of Papuan languages was included in section 2 for the areas in which they occur. 4.1 Probable migrations and the history of Papuan languages The first evidence of human occupation in New Guinea, the main area of Papuan languages, dates back well over 50,000 years, in northwestern Papua New Guinea. Those humans were probably Australian Aborigines, because until about 8000 years ago, New Guinea and Australia were a single large continent, and the Australian Aborigines are believed to have penetrated into Australia via New Guinea. There is bloodgroup, cultural and also some lin­ guistic evidence of the former presence of Australian Aborigines in New Guinea (Foley 1986:271-275). The linguistic one is in the Eastern Highlands, geographically not far from the first evidence of human occupation mentioned above. It seems likely that the Australian Aborigines in New Guinea were forced out of it by the first (and perhaps subsequent) migration(s) of speakers of Papuan languages into New Guinea. It is not known when these have pene­ trated the New Guinea area - it may well have been close to 40,000 or more years ago, because evidence of human occupation in New Britain and New Ireland to the East of the New Guinea mainland at least 35,000 years ago

32

STEPHEN WURM

(Spriggs 1997:47) is likely to have been from ancestors of present-day Papuans there, not from Australian Aborigines. There may have been several initial Papuan immigrations into the New Guinea area, and the present speakers of Papuan languages which belong to the various eastern groups and the small scattered groups in New Guinea mentioned in section 1 may be descendants of these earlier Papuan immigrants. The same goes for the two possibly related groups of the five large groups mentioned there. One of these is the so-called Torricelli Phylum of Papuan languages which has a total of forty-eight lan­ guages in northwestern Papua New Guinea along the Torricelli Ranges, inland from the northwest coast. They seem to constitute the result of an old Papuan immigration, because their territory has been split into two by a part of another of the five large groups, the Sepik-Ramu Phylum, with the Sepik-Ramu Phy­ lum languages in the area of that split showing obvious Torricelli Phylum languages substratum (Wurm 1982:262). In spite of their assumed great antiq­ uity, the Torricelli Phylum languages are fairly similar to each other, which would militate against their antiquity. However, it seems that most of the origi­ nal Torricelli Phylum languages have disappeared under the pressure of other languages, and that the present-day Torricelli Phylum languages are the de­ scendants of only one language, or of several closely related ones. There is linguistic evidence for this, and also the presence of the very typical bloodgroups of Torricelli Phylum language speakers in populations not speaking Torricelli Phylum languages today. The languages of the other group possibly related to those of the Torricelli Phylum are those of the so-called West Papuan Phylum (thirty-one languages) found today in the Bird's Head Peninsula of West Papua, and in the northern part of the Halmahera Islands to the north of the Bird's Head Peninsula. It seems that a Papuan immigration from the west reached those areas, and stopped there. Another immigration of speakers of quite different Papuan languages brought the present-day Sepik-Ramu Phylum languages (104 languages) into a large area in northern and northeastern Papua New Guinea. They obviously came from the west, but evidence of their former presence further west appears to have been blotted out by a further immigra­ tion, probably by that of the recently discovered languages of a group in the western lowlands of West Papua, known as the Geelvinck Bay Phylum (thirtyfour languages), which is also one of the five large groups mentioned in section 1. The main immigration of Papuan languages was comparatively recent, probably only 5000-6000 years ago. It brought into the New Guinea area the widespread very large language group of now 485 more or less closely related languages known as the Trans-New Guinea Phylum which occupy three to four fifth of mainland New Guinea and much of the Timor-Alor-Pantar islands in the west of the West Papua mainland. The core languages of it occupy the large

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA highlands chain stretching from central Papua New Guinea with one interrup­ tion to the language picture, to the far west of West Papua and all of the north­ east of Papua New Guinea, a total of 239 fairly closely related languages, and a secondary core of 171 languages less closely related to those of the first core, though many of them are closely related to each other. They are located in central western and southwestern areas of Papua New Guinea and adjacent lowlands of West Papua as far west as the Bomberai Peninsula, and also in some northern near coastal areas of eastern West Papua. Furthermore, also in the central highlands areas of West Papua all the way from the Papua New Guinea border far to the west, the southern part of the Bird's Head Peninsula, the western part of the Bornerai Peninsula, and most of the Timor-Alor-Pantar Islands to the southwest of West Papua. Another large group of fifty secondary core languages less closely related to the first core languages, but relatively closely related to each other, are located in the large eastern tail section of Papua New Guinea, from the capital Port Moresby to the north, northeast or east to close to the tail-end which is occupied by Malayo-Polynesian lan­ guages. This gives a total of 460 core Trans-New Guinea Phylum languages. The remaining twenty-five languages which may be somewhat doubtful mem­ bers of the Trans-New Guinea Phylum, are located in fringe areas of it in southwestern Papua New Guinea and in eastern central West Papua giving a total of 485 Trans-New Guinea Phylum languages established to date. In the far and wide spreading of the speakers of Trans-New Guinea Phylum lan­ guages through New Guinea, it seems that they absorbed a part of the earlier Papuan immigrants and their languages. The twenty-five last potential TransNew Guinea Phylum languages may be such incompletely absorbed, originally different, Papuan languages. It seems possible that the ancestors of speakers of Trans-New Guinea Phylum languages who moved towards the tail-end of Papua New Guinea, forced some of the Eastern Papuan languages out of New Guinea into the island world beyond it in the east and north-east, because there is a substratum of the latter in a part of the Trans-New Guinea Phylum lan­ guages in the eastern tail-part of the New Guinea mainland, though the great majority of these is likely to have occupied part of the island world to the east of the New Guinea mainland for tens of thousands of years earlier (see section 4.2). In addition to the main west-to-east migration of the speakers of TransNew Guinea Phylum languages, there is evidence of a later east-to-west movement of their languages from the main central core area, especially from the Huon Gulf area, all the way into Western West Papua, and into the TimorAlor-Pantar area. In Trans-New Guinea Phylum languages, even in those which today are far removed from regions occupied by Malayo-Polynesian languages, there are Austronesian, i.e. Malayo-Polynesian, loanwords which

33

34

STEPHEN WURM

reflect proto-forms which may have been used in regions not far to the west of New Guinea before Malayo-Polynesians reached New Guinea about 5,500 years ago (see section 2.1) (Wurm 1982:102-108). This makes it seem likely that the ancestors of speakers of present-day Trans-New Guinea Phylum lan­ guages lived for a long time in regions to the west of New Guinea where they got into contact with Malayo-Polynesians migrating westwards, and borrowed loanwords from them. Pressure from the Malayo-Polynesians seems to have made them leave the regions in the west, and made them penetrate the New Guinea area where they multiplied. While it has been mentioned that TransNew Guinea Phylum language speakers may have forced speakers of the East­ ern Papuan language groups into the island world to the east of New Guinea, the question of the presence of some Papuan languages in the Solomon Islands Chain and the Santa Cruz Archipelago requires explanation, because Papuans, with very few exceptions, are not seafarers. The answer comes from archae­ ology. Golson (1966) states that the unpolished waisted stone axe, a typical characteristic of Papuan culture, has been found in parts of New Guinea, and also in New Britain, and the western Solomon Islands of Bougainville (on Papua New Guinea territory) and Vella Lavella - a proof that Papuans reached these islands which required relatively little navigational skill. However, the presence of only polished waisted axes in excavations further east is a proof that Papuans living there today had assistance and stimulus from MalayoPolynesian newcomers for open ocean voyaging or, as in the case of the Pap­ uan languages spoken in the Santa Cruz Archipelago, were brought there as slaves by Malayo-Polynesians, for which there is also other archaeological evidence. 4.2 Classification of the Papuan languages Much of the classification of the Papuan languages has been subsumed under section 4.1. Table 1 is a summary survey classification of them. 4.3 East Papuan languages The Papuan languages located in the island world to the east of the New Guinea mainland are likely to be descendants of the ancestral languages of the original immigrants there (but see the last part of section 4.1 for the Solomon Islands and the Santa Cruz Archipelago) who may well be those humans for whose presence there is archaeological evidence dating back at least 35,000 years for New Britain and New Ireland, and 29,000 years for Buka Island, northwest of Bougainville Island (Spriggs 1997:47). There are twenty-nine living languages, with four more extinct. Taking into account very strongly aberrant forms of one of the New Britain languages (Baining or Aaqet), and

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA 1

Trans-New Guinea Phylum languages

485

1.1

Main central core languages

239

1.1.1

Huon-Finisterre languages

67

1.1.2

Madang languages

60

1.1.3

Adelbert Range languages

44

1.1.4

East New Guinea Highlands languages

1.2

Western secondary core languages

68 171

1.2.1

Angan languages

13

1.2.2

Central and South New Guinea languages

67

1.2.3

Dani languages

11

1.2.4

Wissel Lakes Kemandoga languages

4

1.2.5

Kayagar languages

3

1.2.6

Marind languages

6

1.2.7

Moraori language

1

1.2.8

Kolopom languages

3

1.2.9

Gogodala-Suki languages

3

1.2.10

Inland Gulf languages

4

1.2.11

Tirio languages

4

1.2.12

Teheran languages

2

1.2.13

Pawaian language

1.2.14

Oksapmin language

1.2.15

Pauwasi languages

1.2.16

Uhunduni language

1.2.17

Dem language

1.2.18

Mek languages

1.2.19

Tanahmerah language

1.2.20

Mor language

1.2.21

South Bird's Head languages

1.2.22

West Bomberai languages

1.2.23

Timor-Alor-Pantar languages

18

1.3

Eastern secondary core languages

50

1.3.1

Binanderean languages

15

1.3.2

Dagan languages

4

10 3

8

35

36

STEPHEN WURM 1.3.3

Goilalan languages

5

1.3.4

Koiarian languages

6

1.3.5

Kwalean languages

3

1.3.6

Mailuan languages

6

1.3.7

Manubaran languages

2

1.3.8

Yareban languages

5

1.4

Languages whose membership to the Trans-New Guinea Phylum is uncertain

1.4.1

Kiwaian languages

7

1.4.2

Turama-Kikorian languages

3

1.4.3

Eleman languages

7

1.4.4

Kaure languages

4

1.4.5

Molof language

1

25

1.4.6

Tofamma language

1

1.4.7

Usku language

1

1.4.8

Morwap language

2

Sepik-Ramu Phylum languages

1

2.1

Main language stocks

91

2.1.1

Sepik languages

49

2.1.2

Ramu languages

42 13

104

2.2

Aberrant language stocks

2.2.1

Gapun language

1

2.2.2

Leonard Schultze languages

6

2.2.3

Nor-Pondo languages

6

3

Torricelli Phylum languages

48

4

Geelvinck Bay Phylum languages

34

5

West Papuan Phylum languages

31

6

Trans-Fly and Bulaka River Phylum languages

18

6.1

Eastern Trans-Fly languages

4

6.2

Pahoturi River languages

2

6.3

Morehead and Upper Maro Rivers languages

6.4

Yelmek-Maklew languages

2

Sentani Phylum languages

4

Nimboran Phylum languages

5

7

8

10

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA

37

9

Sko Phylum languages

8

10

Left May (Arai) Phylum languages

6

11

Kwomtari-Baibai Phylum languages

6

12

AmtoMusian Phylum languages

2

13

Border Phylum languages

14

14

Senagi Phylum languages

2

15

Tor Phylum languages

16

Kwerba Phylum languages

13 8

17

Mairasi Phylum languages

18

Isolated Papuan languages not belonging to any group and unre­ lated to each other

3 13

Table 1

one on East Bougainville (Buin), there are thirty-thi'ee languages, including the extinct ones. In 1975, an attempt was made on the basis of some very inade­ quate information, to include them into a highly tentative preliminary East Papuan Phylum (Wurm 1975a). A new look at the pronominal forms of nine­ teen of the living Papuan languages of that area was taken by Ross on the basis of very much extensive and informative material (Ross 2001) to establish whether that material gave proof of the existence of that Phylum. Ross con­ cludes that this comparison does not support its existence, and assigns the languages to eight different groups, admitting that the enormous time-depth may not allow for a definite conclusion he points to the presence of certain common characteristics such as the almost universal distinction of masculine and feminine in the pronominal system which he regards as archaic, but indi­ cates that this may be an areal linguistic feature and need not support the as­ sumption of common linguistic ancestry. Features shared by languages now geographically very far apart from each other such as the identity of singular subject suffixes on the verbs and of bound morphemes in the verbs which signal whether an action directed against an object damages the object or leaves only a harmless mark on it, in the Löndäi language in the Santa Cruz Archipelago and the Buin language in East Bougainville (Laycock, personal communication) are more likely to support the assumption of a common an­ cestry of the languages involved. The following are the nine by now presuma­ bly unrelated groups of East Papuan languages as proposed by Ross (2001), with numbers of extinct languages and languages not taken into account by him added by the present author to one or the other of his groups where he feels that they may presumably also belong to them.

38

STEPHEN WURM 1 Yele-West New Britain group

3

2 Kol isolate

1

3 Sulka isolate

1

4 East New Britain group

3 (+4)1

5 Kuot isolate

1

6 West Bougainville group

2(+l+l?) 2

7 East Bougainville group

4(+l) 3

8 Central Solomons group

4(+3) 4

9 Reef Islands-Santa Cruz group

4

Table 2

The total number of East Papuan languages is thirty-three. One extinct lan­ guage in New Britain (Makolkol) is of unknown affiliation (Yele-West New Britain or isolate?). It brings the total number of East Papuan languages to thirty-four, with five of them extinct. 4.4 Endangerment of Papuan languages What has been said towards the beginning of this contribution about the endangerment of Austronesian languages is all fully applicable to Papuan languages. In the parts dealing with the endangerment of Austronesian (mainly Malayo-Polynesian languages, except for Taiwan) languages in the Pacific area, all relevant information on the endangerment of Papuan languages has also been added for those in which Papuan languages also occur. 5 The Australian languages 5.1 The picture and history of Australian languages The Australian languages all belong to one very large group of genetically interrelated languages. Their territory was originally the entire Australian continent, including most islands off the north coast of Australia in Torres Strait. The inhabitants of the western Torres Strait Islands are racially different from the Australian Aborigines and are likely to have originally spoken a 1 East New Britain Group 3 (+4) = four languages closely related to one of the three are added (one of the three is extinct). 2 West Bougainville Group 2 (+1+1?) = +one living + one doubtful added. 3 East Bougainville Group 4 (+1) = one closely related to one of the four is added. 4 Central Solomons Group 4 (+3) = three extinct ones added.

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA

39

Papuan language before incompletely adopting an Australian language. The inhabitants of Darnley Island in eastern Torres Strait are Papuans speaking a language very closely related to others of the Eastern Trans-Fly Family of languages in southwestern Papua New Guinea (section 4.2). The several lan­ guages of Tasmania got extinct during the 19th century, with the last speaker dying in 1877, and materials available on them are meagre and poor. It is not clear whether or not they were related to the Australian mainland languages or not. Their relationship to the latter, if any, was certainly not close, but very distant lexically, structurally and phonologically, as far as it can be judged from the inadequate material. At the same time, the separation of Tasmania from Australia dates back to 12,000 years, when a rise in the sea levels cut off Tasmania from Australia. This time depth is enough to wipe out traces of a distant relationship, especially when taking into account the restricted and unreliable nature of the material available for the Tasmanian languages. The Australian Aborigines are thought to have first reached what is today New Guinea more than 50,000 years ago. Australia, New Guinea and Tasmania were a single continent then until the separation of Tasmania 12,000 and of New Guinea 8000 years ago. The Australians appear to have spread from New Guinea into Australia quite quickly, because recently evidence of their pres­ ence in southeastern Australia about 50,000 years ago has been established. The first Papuans are believed to have penetrated into New Guinea at least 40,000 years ago, and they may have put pressure on the Australians to move south. There are still physical, cultural and even linguistic traces in eastern New Guinea of their former presence there as mentioned in (section 4.1). The Australian continent was formerly much lusher, wetter and more fertile than today, especially in its now dry and desert-like interior. This makes it probable that the distribution of Australian language groups was much more even, and quite different from that at the time of contact, until the drying-up of the inte­ rior of the continent started more than 5000 years ago. Today there are twenty rather different, but still related, smallish Australian language groups in the north and northwest of the continent in the tropics, with one huge group, called Pama-Nyungan, occupying the other parts of Australia, seven-eighths of it, including the northeast, at the time of contact. A certain number of lexical items are found in many Australian languages, but regional vocabularies occur in parts of Australia which are mainly near the coasts, especially in the east and southeast, but also elsewhere. In the same areas, structural features appear in languages containing such regional vocabularies, which differ from some which are widely found in other Pama-Nyungan languages. Also, lexical cog­ nates exist in widely separated areas, with no such cognates appearing in lan­ guages spoken in between. All this may have the following explanation: the

40

STEPHEN WURM

speakers of the ancestral languages of the present Pama-Nyungan language subdivisions who appear to have lived in the formerly lush and green areas full of food and water resources in the far interior of Australia, but which were first, and most, affected by the drying-up process of the interior of Australia, especially of its western part, may have started to move outwards quite exten­ sively and forcefully. In the course of this, they absorbed speakers of other languages and pushed them further away towards marginal areas and the coasts in the east, southeast and west, splitting speakers of closely related languages far apart, superimposing upon them much of the vocabulary and structural features of their languages, and influencing many of them, until only remnants of the earlier languages remained, such as regional vocabularies, vocabulary cognates in languages thousands of kilometres apart, with no such cognates in between, and some structural characteristics not usually found in PamaNyungan languages, or features reflecting Pama-Nyungan characteristics in unusual ways. The number of different, though in many areas quite similar, Australian languages before contact may well have been 500 or even more, but was re­ duced to about 400 as a result of a catastrophic smallpox epidemic which broke out in 1789 and which, as it became evident later, had spread right through Australia from tribe to tribe and killed an unknown number of speakers of many languages who were not even aware of the existence of white man. Most of these different languages had, with very few exceptions, only very few speakers each, usually in the hundreds. During the 19th and the early 20th cen­ tury, not much less than 100 languages became extinct, as a result of the impact of English, and especially the hostile attitudes, actions such as the poisoning of waterholes of the Aborigines, and harsh policies of the white Australians, the removal of Aborigines from their traditional lands, the crowding together of Aborigines from many different language backgrounds into missions and gov­ ernment-controlled settlements where they had to use pidgin English or per­ haps one larger Aboriginal language understood by all of them in one settle­ ment in view of the widespread multilingualism which used to be, and in some areas still is today, a typical cultural feature of Aboriginal Australia. This led to the disappearance of most, if not all, of the languages represented in individual settlements. Children often did not learn an Aboriginal language, but pidgin English and later English instead. Pressure in this direction was exerted by the controllers of the settlements and missions. This situation continued until the 1970s, with very strong assimilation policies in Australia (which were also applied to migrants from non-English speaking countries). These were very much to the detriment of the maintenance of Aboriginal languages. Then there was a complete about-face of governmental policies from assimilation to mul-

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA

41

ticulturalism and support for the maintenance of Aboriginal languages, which still continues today, but it was too late for most of them. In the early 1980s there were still at least a few elderly speakers of well over 300 Aboriginal languages. Many languages with between one, a few or up to a dozen speakers, some with a little more, were listed and indicated on maps in Wurm & Hattori (1981-83), but the few elderly speakers of most of those languages have now died, resulting in the extinction of about 170 Aboriginal languages in the last few decades, and some a little earlier. Today, there are about 24-25 still fully functional Aboriginal languages in Australia, and about 120 other languages in various stages of endangerment. Of the latter, perhaps 50 or more are now in the final stages of disappearance (Wurm, 2001a). This is all that is left of the original 400 languages or so a short time after contact, and perhaps more than 500 before that time. In addition to the still continuing pressure from English, there is a relatively recent further threat to the maintenance of Aboriginal languages in northern Australia. This is from Kriol, an English-based Australian creole language which is very widespread in north Australia, with over 10,000 first language speakers and well over 10,000 second language speakers. It has caused many Australian languages to fall into disuse, especially with children. The even more widespread Aboriginal English which Aborigines regard as an important symbol of Aboriginal identity and which is used by them all over Australia, has also worked against the maintenance of Aboriginal languages. It is a form of English deceptively similar to colloquial English, but reflecting Aboriginal speaking attitudes and discourse patterns which differ markedly from those adhered to by native speakers of (Australian) colloquial English, and have also other special features. All this leads to English speakers frequently misunder­ standing Aboriginal English, to such an extent that for instance in court cases in which Aboriginal English speakers are involved, the presence of a qualified interpeter is needed. As an example, silence in response to a question addressed to an Aboriginal English speaker signals "yes", i.e. agreement, not "no" or ignorance, as this would be the case between speakers of English. At the same time, there have been two positive features and developments in addition to the replacement of the negative official language policies in Australia by positive ones, including bilingual education in English and major Aboriginal languages in northern Australia, though the latter has come under threat again recently as a result of ill-informed and shortsighted new policies of the government of the Northern Territory of Australia. One of these two positive features of the cata­ strophic Aboriginal language situation is that good speakers or at least semispeakers of languages believed extinct can be found, usually a long way away from their traditional homeland. So, for instance, in late 2000 two good speak-

42

STEPHEN WURM

ers of an almost unstudied language believed extinct since the 1970s were discovered, one of them not far from his traditional homeland in far northern Australia near the Gulf of Carpentaria, but the other in far southwestern Queensland, near the border of New South Wales, more than 2000 kms from his traditional homeland. Two linguists specialising in Australian languages brought them together into a settlement in southern Queensland and made a good rescue study of their language with support from UNESCO and CIPL (the Permanent International Committee of Linguists which is the world organiza­ tion of linguists). Quite a few such events have taken place in the last decade. The discovery of speakers of languages believed extinct is a sequel to the former adverse language policies in Australia and the large-scale displacement Anindilyakwa

1

(n)

Bunaban

2

(le-s, ls)

Burarran

4

(2n, ls-m, le)

Daly

20

(ln, 2e, le-s, 2s, ls-m, 9m, 4d)

Djamindjungan

3

(ls, lm, ld)

Djeragan

3

(le, ls-m, lm)

Gagadjuan

1

(m)

Garawan

2

(le, lm)

Gungarakanyan

1

(d)

Gunwingguan

14

(ln, l p , le, 2e-s, 4s, 3m, 2d)

Laragiyan

2

(2d)

Mangerrian

3

(3d)

Maran

3

(ls, ls-m, ld)

Nyulnyulan

9

(2e-s, 7d)

Tiwian

1

(n)

Umbugarlan

3

(3d)

Unclassified non-Pama Nyungan North Australian

3

(3d)

West Barkly (Djingili-Wambayan)

3

(2m, ld)

Wororan

8

(le, 3s, ls-m, 3d)

5

(2p, 3m)

Yiwaidjan

Table 3: Non-Pama-Nyungan

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA

Arandic

6

(4n, lp-e, lm)

Baagandji

4

(lm, 3d)

1

(m)

anjalang Barrow Point

1

(d)

Durrubulan

2

(2d)

Dyirbalic

3

(lm, 2d)

Flinders Island

1

(d)

Galgadungic

2

(2d)

Kala Lagaw Ya

1

(P)

Karnic

20

(2m, lm-d, 17d)

Kulinic

1

(d)

Kumbainggiric

2

(2d)

Maric

20

(20d)

Muruwaric

1

(d)

Ngarinyeric-Yithayithic

3

(3d)

Nyawaygic

2

(2d)

Paman

47

(ln, l p , 4e, 8s, ls-m, 2m, 30d)

South-West Pama-Nyungan

55

(8n, 3p, 2e, 5e-s, 2s, ls-e?, ls-m, 7m, 2m-d?, lm-d, ld?, 22d) (3m, 3d)

Tangkic

6

Waka-Kabic

7

(7d)

Wakaya-Warluwaric

3

(3d)

Warumungic

1

(e)

Wiradhuric

5

(2m, 3d)

Yalandyic

4

(ln, le, 2d)

Yanyuwan

1

(e)

Yidinic

2

(ls-e, lm)

Yuin-Kuric

13

(13d - 5 of them long dead, but in living memory)

Yuulngu

10

(4n, 4p, le, ls) Table 4: Pama-Nyungan

43

44

STEPHEN WURM

of speakers of many languages in the past, with usually none left in their tradi­ tional areas, but some still living in areas far removed from them. Another positive feature has been a general reawakening of a very strong ethnic identity feeling among Australian Aborigines in the last two decades or so, with a very much increased interest in their traditional languages on their part. This has, with the help of linguists, led to the reinvigoration of a number of languages which were very much on the way to disappearing, and even resulted in the revival of some long extinct Aboriginal languages. A good example of this is the Kaurna language of the area of Adelaide, the capital of South Australia, which ceased to be a language of daily life in the 1890s, with its last speaker dying in 1927. It had, however, been studied extensively before its demise, and at the request of the surviving Kaurna ethnic community, which still lives in the Adelaide area, linguists from Adelaide University established teaching courses of the Kaurna language for them, with the result that there are now quite a few speakers again who use it in daily life. 5.2 Classification of the Australian languages It has been mentioned above that there are twenty small families of Austra­ lian languages in the north and northwest of Australia, i.e. in Arnhem Land and south of it, and in the Kimberleys and areas close to it, and one huge family, called Pama-Nyungan, occupying the remaining seven-eighths of the continent. In table 3, the twenty non-Pama-Nyungan families will be enumerated first, each with the number of its languages added, excluding, in most cases, long extinct languages, but including recently extinct ones. In parenthesis, the num­ ber of threatened and recently extinct languages is given after that, with the symbols p = potentially endangered, e = endangered, s = seriously endangered, m = moribund, d = extinct, n = not threatened (cf. section 1). Then, in table 4, the Pama-Nyungan family languages are given. The family has twenty-eight subdividions of which fourteen consist only of one or two languages.

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA

45

Appendix: The northern islands of the Pacific world 1 Japan, Sakhalin and The Kuril islands In these areas, except for Sakhalin, there is, or was, only one language which was until recently seriously endangered, but is now revived in part of Hokkaido, the northern island of Japan. This is the Ainu language which was formerly spoken widely in Hokkaido, on much of Sakhalin Island, and on the Kuril Islands which link northeastern Hokkaido with the Kamchatka Peninsula of the Chukchi Peninsula of northeastern Siberia. Until late in the nineteenth century, Ainu was in part a lingua franca in the Kuril Islands, which had changed hands between Russia and Japan several times. By the first decades of the twentieth century, Ainu ceased to be spoken on the Kuril Islands. On Sa­ khalin Island, Ainu v/as still spoken in the southern third of it at the beginning of the twentieth century and used as a lingua franca between speakers of the local languages, and also used as a lingua franca between them, the local ad­ ministration and the Japanese fishing industrialists. However, the Ainu lan­ guage gradually disappeared. At the beginning of 1949, less than 100 Ainus remained on Sakhalin Island, and in the late twentieth century, the last speaker of the Sakhalin Ainu dialect died. The other main local languages on Sakhalin Island were the Palaeo-Asiatic Nivkh (formerly known as Gilyak) and the Tungusic Orok and Evenki. How­ ever, as a result of intensive Russification of the local population, only 447 speakers of Nivkh, fifty-five of Orok and thirty-three of Evenki have remained now (Gruzdeva 1996:1007-1012). On Hokkaido, the Ainu language had been under intense pressure from Japanese as a result of very adverse language policies of the Japanese government, until only eight elderly speakers remained in the early 1990s. A short time after that, the Japanese language policies re­ lating to Ainu made a total about-face and changed to very positive ones. A magnificent Ainu museum with Ainu language teaching facilities was estab­ lished near Sapporo town on Hokkaido, a considerable number of semispeakers who had been afraid of using the language for fear of repraisals, were encouraged to reuse and improve their language knowledge with help from the surviving old speakers, young Ainus were encouraged to learn the language until they became fluent speakers etc. 2 Aleutian islands Another language, spoken on the Aleutian Islands connecting southern Alaska with the central part of the Kamchatka Peninsula, is the Aleut language, related to Eskimo, spoken as Eastern Aleut on the USA Islands, and as Western

46

STEPHEN WURM

Aleut on the Russian ones. On the former it is endangered and becoming seri­ ously endangered, on the latter it is moribund, in both cases as a result of pres­ sure from English on the eastern and from Russian on the western islands. On Copper Island in the Russian part, it is spoken in an Aleut-Russian hybrid form which is also moribund (Wurm et al. 1996, Arctic North America map by Hein van der Voort). 3 St. Lawrence Island Finally, on St Lawrence Island, in the Bering Sea, Bering Strait Eskimo is spoken. It is now potentially endangered. References Foley, William A. 1986. The Papuan languages of New Guinea. Cambridge: Cambridge University Press (Cambridge Language Surveys). Golson, Jack. 1966. 50,000 years of New Guinea history. Paper read at the meeting of the New Guinea Society on July 26, 1966. Gruzdeva, Ekaterina. 1996. "The linguistic situation on Sakhalin Island". Wurm, Mühlhäusler & Tryon 1996, vol. II .1007-1012. Moseley, Christopher, ed. 2001. Encyclopedia of the world's endangered languages. London: Curzon Press. Mühlhäusler, Peter, Malcolm Philpott & Rachel Trew. 1996. "Modern media in the Pacific area and their role in intercultural communication". Wurm, Mühlhäusler & Tryon 1996, vol. II .1389-1454. Nekitel, Otto. 1998. Voices of yesterday, today and tomorrow. Language, culture and identity. New Delhi: UBS Publishers' Distributors Ltd, in association with Nekitelson Pty Ltd, Port Moresby, Papua New Guinea. Pawley, Andrew, Malcolm Ross & Darrell Tryon, eds. 2001. The boy from Bundaberg. Studies in Melanesian linguistics in honour of Tom Dutton. Canberra: Australian National University, Pacific Linguistics. Ross, Malcolm. 2001. "Is there an East Papuan Phylum? Evidence from pro­ nouns". Pawley, Ross & Tryon 2001. Spriggs, Matthew J.T. 1997. The Island Melanesians. Oxford: Blackwell. Wurm, Stephen A. 1975a. "The East Papuan Phylum in general". Wurm 1975b. 783-804. —, ed. 1975b. New Guinea area languages and language study. Vol. I: Pap­ uan languages and the New Guinea linguistic scene. Canberra: Austra­ lian National University, Department of Linguistics (Pacific Linguistics 38). — & Shirô Hattori, eds. 1981-83. Language Atlas of the Pacific Area. Can­ berra: Australian Academy of the Humanities in collaboration with the Japan Academy (Pacific Linguistics Series 66-67).

LANGUAGE ENDANGERMENT IN THE GREATER PACIFIC AREA

47

—. 1982. Papuan languages of Oceania. Tübingen: Narr. —, Peter Mühlhäusler & Darrell T. Tryon, eds. 1996. Atlas of languages of intercultural communication in the Pacific, Asia, and the Americas. Berlin: Mouton de Gruyter. —, ed. 2001. Atlas of the world's languages in danger of disappearing. 2nd ed. Paris: UNESCO Publishing/Pacific Linguistics. —. 2001a. "Language endangerment in the Pacific, especially in the New Guinea area". Pawley, Ross & Tryon 2001. —. 2001b. "Australasia and the Pacific". Moseley 2001.

LANGUAGE ENDANGERMENT IN INDONESIA THE INCIPIENT OBSOLESCENCE AND ACUTE DEATH OF TEUN, NILA AND SERUA (CENTRAL AND SOUTHWEST MALUKU) 1

AONE VAN ENGELENHOVEN Leiden University

1 Introduction The province of Maluku comprises about 1000 islands in the east of Indone­ sia. The number of languages found on these islands is estimated 92 (Taber et al. 1996), depending on the definition of language versus dialect. The indige­ nous languages seem to share many typological features that are not found elsewhere in Indonesia, thus suggesting that insular East-Indonesia be a typo­ logical Sprachbund. The languages are either Austronesian or belong to one of the non-Austronesian or Papuan stocks, which makes this province one of the linguistically most complex parts in Indonesia from a comparative point of view. Florey & Van Engelenhoven (2001) found that, in terms of language ecology, Maluku forms one of the most severely endangered linguistic regions of Indonesia. In this paper we focus on the isolects of Teun, Nila and Serua in Southwest Maluku which show patterns of language endangerment explanatory for entire Island Southeast Asia. Most examples are taken from the Serua isolects. 2 Genetics The isolects of Teun, Nila and Serua (TNS, or TNS language for short) are originally spoken on three islands bearing the same name in Southwest Maluku. Although no consensus yet exists whether the reconstructed direct 1

The research underlying this paper has been conducted within the framework of the project Endangered Languages of West and Central Seram Island, eastern Indonesia, funded by the UNESCO Endangered Languages Fund. I wish to thank the following persons for their support while writing this paper: Mr. Semol Pormes and Mr. Mozes Workala of the Letra Natu Foundation, Margaret Florey, who kindly provided me with data from her Amahai fieldwork, Don van Minde and Johnny Tjia, Otjep Rahantoknam and above all Patrick Sullivan, who kindly provided the sociolinguistic information about Waipia that I needed.

50

AONE VAN ENGELENHOVEN

ancestor of TNS, Proto Central Malayo-Polynesian, really has existed or not (cf. Blust 1993 versus Hull 1998), most linguists agree that the TNS languages are genetically related to the Austronesian languages of Timor, either directly (Taber 1993) or indirectly (Hull 1998). Based on shared phonological innovations, Collins (1982) concludes that the isolects on the two northern islands of Nila and Serua are dialects, whereas exclusive sound shifts set off the southernmost Teun isolect as a separate lan­ guage. The dialect relation between Serua and Nila is confirmed by informant reports that consistently stress the mutual intelligibility of both isolects. Serua informants suggest that the lesser intelligibility of the Teun isolect be linked to the relative isolation of its speech community rather than to a divergent gram­ mar or lexicon. Because a decisive classification awaits further comparative research, we adopt here the indigenous perception that all three isolects are dialects of a single TNS language. Proto Central-Malayo-Polynesian

Proto Timor

Proto Teun-Nila-Serua

Teun

Nila

Serua

Figure 1: Genetic affiliations of the Teun, Nila and Serua isolects (after Collins 1982)

From an ethnolinguistic point of view, the TNS islands are typical members of the Southwest Malukan ethnolinguistic zone, in which the islands of Kisar and Luang function as cultural centers. Their society distinguishes between an

LANGUAGE ENDANGERMENT IN INDONESIA

51

aboriginal and a migrant population, while local folklore acknowledges clan relations based on trade and war alliances between Banda and Serua, Leti and Nila, and between the Babar Islands and Nila. As on the other islands in Southwest Maluku, the migrant clans emphasize the memorization and retell­ ing of narratives that explain their place and function inside the society. Al­ though further fieldwork is required, we hypothesize that the TNS language like all other Southwest Malukan languages features extensive lexical pairing in its oral traditions (Van Engelenhoven 1997).

Map 1: The islands of Teun, Nila, Serua and Seram in Maluku (Indonesia).

The TNS communities are traditionally seafaring and established an economic network that connected the Banda Islands in Central Maluku to the Kei and Aru archipelagos in Southeast Maluku and to the other Southwest Malukan islands. Informants generally consider the TNS dialects to be heavily influ­ enced lexically by languages from islands with which a trading relation was maintained, for example Leti, one of the Luangic-Kisaric languages off the coast of East-Timor (Van Engelenhoven 1995). Indeed, the speech in one village on Teun Island, Layeni, has been identified as a dialect originating from Wetan Island in the Babar archipelago (Collins 1982). Interestingly, while Taber (1993) analyzed it as another Wetan dialect, informants in the Nether­ lands identify the speech spoken in Isu on Teun Island rather as a Serua dialect. The Teun dialect was spoken in three villages on Teun Island and in one village, Bumei, in Southwest Nila. The Nila dialect (or North-Nila in Collins'

52

AONE VAN ENGELENHOVEN

(1982) analysis) was spoken in at least three villages in Northeast Nila, whereas the Serua dialect was spoken in three villages in Southeast Nila and on entire Serua Island. Beside that, Serua has been reported to function as a lingua franca among the three islands and among TNS islanders on sea-travels.

Map 2: The islands Teun, Nila and Serua.

3 Typology From a typological point of view, Southwest Maluku is part of the larger Timor region. The TNS language therefore displays features that are typical for this region (cf. Van Engelenhoven & Hajek 2000). Like all Austronesian lan­ guages in the region it displays a subject-verb-object word order and has two first person plural pronouns signaling whether the hearer is implied or not. Its phoneme inventory is relatively simple: /p, m, f, w, t, n, s, d, 1, r, k, ', i, u, e, o, a/. No instances of geminate vowels or consonants have been attested. Plurality is indicated on the noun by means of a suffix -rV. Table 1 shows that the final vowel of the stem is copied into the vowel slot of the suffix. On stems ending in Va#, the suffix is simply added. Stems ending in VCV# or VV# in which the final vowel is not /a/ delete the final vowel. In stems ending in VCCV# the final vowel metathesizes with the preceding consonant. Stems ending in ra# also trigger metathesis between /a#/ and /r/, and additionally delete the suffix's consonant. This is displayed in Table 1.

LANGUAGE ENDANGERMENT IN INDONESIA

GLOSS

+ -rV

GLOSS

wia-ra

'bananas'

'banana'

wia

'child'

wakoi

wako-ri

'children'

'pig'

wawi

waw-ri

'pigs'

'stone'

watu

wat-ru

'stones'

'forest'

letna

letan-ra

'forests'

'lip-

iwti

iwit-ri

'lips'

'mountain'

wura

wuar-a

'mountains'

'bamboo'

temra

temar-a

'bamboos'

53

Table 1: Plural inflection in Serua

The metathesis attested between final vowels and preceding consonants in plural inflection (e.g. letan-ra 'forests' / iwit-ri Tips' from letna 'forest' / iwti Tip' + -rV 'PL', respectively) also occurs on the phrase level between words. This type of metathesis groups the TNS language with most Luangic-Kisaric languages of Southwest Maluku and the Austronesian isolects of Central and West Timor. The rules for metathesis remain unclear for the time being. The phonological context that triggers metathesis seems to be that the right side morpheme begins in #CV and the left side morpheme ends in VCCV#. GLOSS

+ lapna 'big'

GLOSS

'boat'

wroa

wro=lapna / wroa lapna

'big boat'

'work'

karei

karei lapna

'lot of work'

'goat'

pipi

pip=lapna /pipi lapna

'big goat'

'island'

nusa

nus=lapna

'big island'

'fish'

ina

ian=lapna

'big fish'

'rain'

usna

usan=lapna

'huge rain'

Table 2: Metathesis, apocope or adposition in Serua phrases

However, some stems ending in VCV# also metathesize into VVC# (e.g. wura ~ wuar 'mountain' in Table 1 and ina ~ ian 'fish' in Table 2). In most cases, however the final vowel is deleted (e.g. nusa ~ nus 'island' in Table 2). No rule has been found yet for VV#, where final high vowels seem always preserved, whereas final /a/ is sometimes deleted. The TNS language features two possessive constructions that distinguish between alienable and inalienable possession. Table 3 shows that in construc­ tions designating alienable possession, the possessor noun or pronoun precedes

54

AONE VAN ENGELENHOVEN

the possession noun. Possessor pronouns directly procliticize onto the posses­ sion noun and delete their final vowel before nouns with initial single conso­ nants (e.g. ruma 'house') or vowels (e.g. asu 'dog'). 2 PRON

POSS

X's boat

X's house

X's dog

lsg

a'u

sa'

sa ' wroa

sa=ruma

sa '=asu

2sg

oa

mu

mu wroa

m=ruma

m=asu

3sg

ia

ni

ni wroa

n=ruma

n—asu

lplinc

ita

tita

tita (ni) wroa

tit=ruma

tit=asu

lplex

ami

sama

sama wroa

sam=ruma

sam=asu

2pl

imi

mira

mira (ni) wroa

mir=ruma

mir=asu

3pl

ira

rira

rira (ni) wroa

rir=ruma

rir=asu

Table 3: Possessive pronouns and alienable possessive construction in Serua

The third person singular possessor pronoun may be inserted as an interme­ diate possessive particle - indicated by brackets in the table - between the possession noun having an initial consonant cluster and first inclusive, second and third plural possessor pronouns. The third person singular possessive pro­ noun also functions as an intermediate particle between nominal possessors and possession nouns, which corresponds with most Austronesian languages found in this region3, e.g. mani n=ruma (man 3sg.POS=house) '(the) house of the man/(the) man's house'. Most Austronesian languages of Southwest Maluku formally distinguish alienable nouns from inalienable nouns by inflecting an obligatory possessive pronominal suffix on the latter. A few lemmas in Anonymous (1937)4 and Workala (1993) suggest that a similar pattern may exist or has existed in Nila and Serua. However, ongoing research and a comparison with Taber's (1993) wordlist rather suggest that the same construction, originally exclusively for alienable possession is now also used for inalienable possession. In this list, many Teun, Nila and Serua nouns that are eligible for inalienable status display 2

The TNS language generally disallows V'CV, hence deleting the glottal stop. Notable exceptions are the Luangic languages, e.g. Leti (Van Engelenhoven 1995) and Luang (Laidig 1993:340-341) that only feature possessive suffixes and the Wetar languages that only feature prefixes, e.g. Tugun (Laidig 1993:342). These languages therefore do not formally distinguish between alienable and inalienable possession anymore. Rather, inalienability is indicated by the permanent possessive affixation of nouns. 4 The Anonymous 1937 'Niala' wordlist compiled at Riring, nowadays at the west bank of the Sapalewa River in Northwest Seram (Taniwel District), and republished in Stokhof (ed.) 1981 was identified as Nila by Collins (1982). 3

LANGUAGE ENDANGERMENT IN INDONESIA

55

the 3sg pronominal proclitic. Interestingly, some kinship terminology occurs unmarked in the wordlist, whereas others display a possessive suffix. 'bone'

'fat'

'tongue'

'mother'

'grandrelative' 5

Teun

ro

ni=mina

ni=mau

yina

pai

Nila

n=kui

mina

ma

ina

wem-ni

Serua

n=kui

n=mina

n=ma

ina

up-nu

Table 4: Alienable and inalienable possessive marking in TNS (Taber 1993)

Table 5 shows that possessive suffixation is gradually becoming obsolete in Serua. Only the first and second singular inflections were attested in Workala (1993), whereas under elicitation the alienable construction was used, lplinc and 2pl combine their pronominal proclitics with the one of 3sg. This suggests that the TNS language undergoes the neutralization of the alienable-inalienable distinction that is attested on several places throughout Maluku (Laidig 1993). 'X's mother' (ina) Serua

Atauru6

1sg

ina-ku

sa'=ina

ánu

2sg

ina-mu

m=ina

(o) miam

3sg

?

n=ina

(ni) ina-n

lplinc

?

tita n=ina

(ita) ina-kre

lplex

?

sam=ina

(im) ina-m

2pl

?

mira n=ina

(mi) ina-m

3pl

?

rira n=ina

(si') ina-kre

Table 5: Possessive suffixes and inalienable possessive constructions in Serua and Atauru

For comparative purposes the possessive paradigm has been added of the Atauru language on nearby Atauru Island (East-Timor) where possessive suf­ fixation is still productive.

5

Like most other Southwest Malukan languages the TNS language uses the so-called 'Hawaiian' system, which only stresses generational differences, thus using the same term for grandchildren and grandparents. Gender differences may additionally be indicated by either adding 'man' or 'woman' (e.g. Nila wem-ni mana (grandrelative-POS man) 'grandfather', wem-ni fata (grandrelative-POS woman) 'grandmother'). 6 Duarte (1990) reports separate lexical forms for the lsg and 2sg possessive forms of 'mother' and 'father'.

56

AONE VAN ENGELENHOVEN

Another Southwest Malukan regional feature is that subjects are crossreferenced on the verb by means of pronominal prefixation or 'iconic linking' (Collins 1991). The data in Collins (1991) indicate that verbal stems featuring an initial consonant cluster take CV prefixes. Ongoing fieldwork suggests that there are two parallel inflections in Serua, as is displayed in Table 6. I

II

1sg

u-klei

u-klei

2sg

mu-klei

mu-klei

3sg

ne-klei

ne/ni-klei

lplin

te-klei

tit=ne-klei

lplex

ame-klei

?

2pl

mi-klei

mir=ne-klei

3pl

re-klei

rir=ne-klei

Table 6: Pronominal inflection of #CC stems in Serua: 'to see'

Column II in Table 6 shows that alternatively in plural conjugation the alienable possessive pronouns are procliticized to the 3sg-inflection. This is also attested on verbs beginning with a vowel or a single consonant. Collins' (1982) claim that pronominal inflection involves stem-internal vowel changes Collins 1991

Workala 1993 I

II

lsg

a= '-ena

ti= '-ena

a '-n-ena

2sg

o=m-ona

ti=m-ona

m=n-ena

3sg

i=n-ena

ti=n-ena

n=n-ena

lplin

i=t-ena

?

tit=n-ena

lplex

a=m-ena

a=ti=m-ena

?

2pl

i=m-ena

ti=m-ena

mir=n-ena

3pl

i=r-ena

?

rir=n-ena

Table 7: Pronominal inflection and procliticization on 'to sleep' in Serua

has only been attested sporadically but has not yet been confirmed solidly in fieldwork sessions. This is exemplified in Table 7 by the Serua verb ena 'to sleep', which takes prefixes. Collins (1991) reports an /e/ to /o/ change for 2sg, which in Workala's (1993) manuscript is attested only after the verb ti 'to

LANGUAGE ENDANGERMENT IN INDONESIA

57

go'. Elicitation, however, rather produced the II column, in which the verb form inflected for 3sg is either once again inflected with a singular prefix, or procliticized by a possessive pronoun. Although Collins (1991) reports that all verbs feature a pronominal marker, pronominal inflection on verbs seems partly constrained in Serua. Verbs be­ ginning in #CV that are preceded by lexical subjects are never inflected. (1)

kok=ea tia noa me? man=this go where Q 'Where does this man go to?'

(Serua)

The fact that the aspectual verb ti 'to go' in Table 7 separates /a/ from m-ena 'lplex-sleep' indicates that a is not part of the pronominal inflection, but rather a proclitic functioning as a subject, like in Leti (Van Engelenhoven 1995:127). As has been attested in Meher (Kisar Island, Blood 1992), inflection is only possible, when the C-prefix docks as a coda on a vocalic proclitic (e.g. pro­ nominal subjects, negators, aspectual verbs, etc.). (2)

ta=n-tia n=ruma NEG=3sg-go 3sg.POS=house 'Did he not go home?'

me? Q

(Serua)

For the time being we uphold the above hypothesis that the constraint is lexico-syntactically motivated in Serua. We assume that pronominal inflection requires pronominal subjects that are procliticized to the verb, and that lexical subjects suppress cross-referencing. The formal similarity between prefixes and proclitics in the 2sg and 3sg sets enables the reanalysis of inflected verbs (3 a) as procliticized verbs instead (3b). (3) a. b.

mtiwait na Walanta m-tia=wait na Walanta DIR Holland 2sg-go=PERF m=tia=wait na Walanta DIR Holland 2sg.POS=go=PERF 'Have you gone to Holland (yet)?'

naye? naye? Q naye? Q

(Serua)

Our data show that this formal similarity, which started between the 2sg and 3sg sets gradually has spread to the other pronominal sets. The following sen­ tence was attested only with the 1 sg possessive proclitic and not with the con­ forming prefix ('-).

58

AONE VAN ENGELENHOVEN

(4)

sa'u payer=wua-ra lsg.POS pay=fruit-PL 'I pay those fruits'

sea DEM

(Serua)

The use of possessive proclitics is also featured by predicatively used adjec­ tives7, as can be seen in example (5). All adjectives occur reduplicated in predicates in Seraa. However, whereas possessive marking in Luangic Leti, where a similar construction by means of possessive suffixation exists (Van Engelenhoven 1995:110), creates nominal predicates, the occurrence of the possessive proclitic in Serua before dynamic verbs as payer 'to pay' suggests that the pronominal set in Serua has entered a process of simplification in which the possessive notion of the proclitics is bleaching. This element is elaborated in section 6. (5)

saina ni road 3sg.POS 'The road is long'

k-lo-lola RED-long

(Serua)

4 Demographic history in Indonesia and in the Netherlands As already said above, the TNS language was originally spoken on the islands of Teun, Nila and Serua in Southwest Maluku. These islands in fact emerged from volcanic activity, which explains the fertility of the soil. It counterbalanced the small size of the islands and surely increased their eco­ nomic value in the region as a permanent food source. Throughout Southwest Maluku it is acknowledged that the TNS islands were frequented especially during periods of famine and drought. Their small size, however, prevented the population to grow. For example, a census taken in the late 19th century men­ tions a total of 3,940 individuals who were distributed among the islands in like amounts. Informants acknowledge the social division of society into a dominant group of 'newcomer' clans, who came from surrounding islands by boat, and a submissive group of 'aboriginal' clans.8 Although Riedel (1886) does not provide any evidence, the small size of the islands and their important contri­ bution in interinsular trading certainly must have had a permanent influence on the demographic and social composition of their societies. The small popula­ tion required that marriages were contracted with people outside the own is7

Many Austronesian languages in the region categorize adjectives as a type of intransitive verb rather than as an independent word class. 8 The term 'submissive' in fact is deceptive, since on many Southwest Malukan islands it is the 'aboriginal' clans that own the land in most cases. Social structure, however, is generally considered to be imported into the insular communities by the 'newcomer' clans (cf. Van Engelenhoven 1999).

LANGUAGE ENDANGERMENT IN INDONESIA

59

land, as is attested on most Southwest Malukan islands, with the notable ex­ ception of Wetar. This type of ethnic diversity explains why most Leti and Kisar islanders are bilingual: as children they had the opportunity to learn the languages of both parents - and sometimes of one of the grandparents too, if one of them originated from a different ethnolinguistic group. However, it remains unclear in how far this Southwest Malukan demographic picture ap­ plied to the TNS islands. A census taken among the households of the initial TNS migrants in the Netherlands rather reveals that the partners in most mar­ riages originated from the same island (Van Engelenhoven 2002). As has been explained elsewhere (Van Engelenhoven & Hajek 2000), lan­ guage is not emblematic for ethnic identity in Southwest Maluku. If no cultural restrictions apply9, Southwest Malukans easily swap languages. According to Hull (1998), who reports the same phenomenon in East-Timor, it is the gram­ matical similarity of the regional languages that facilitates this type of multilingualism. Around 1980, the TNS islands were involved for the first time in the trans­ migration policy of the Indonesian government. This policy, which was intro­ duced during the Dutch colonial rule, intends to relieve the demographic ten­ sion in densely populated regions - as for example the island of Java - by redistributing the population in regions with few inhabitants. In the case of the TNS islands, the relocation of the population was motivated by possible vol­ canic hazard. Ellen (1993) reports that in 1979 1175 households from the is­ lands were transported to the Northeast coast of the Elpaputih Bay in Amahai District (South Seram, Central Maluku). In 1982 and 1983 the remaining population on the TNS islands was moved to Seram. The islanders kept their transmigration

households

Location

1964

60

Letwaru (Masohi)

1976

50

Ruatan Valley (Makariki)

1977

50

Ruatan Valley (Makariki)

1979

1175

Waipia

1982

150

Waipia

1983

remaining individuals

Waisiru (Waipia)

Table 8: Transmigration from the TNS islands to Amahai District (after Ellen 1993)

9

Non-Austronesian Oiratans in the Netherlands, for example, are denied access to learning materials of Austronesian Meher on Kisar Island (Florey & Van Engelenhoven 2001).

60

AONE VAN ENGELENHOVEN

municipal independence and a separate TNS District around Waipia was carved out of the Amahai District. The initial construction of the TNS District at Waipia reflected the original geography on the islands. The 17 original villages are represented as quarters in three island clusters. Three Serua-speaking villages, clustered in Northwest Serua, are grouped together south of the Pia River. Waru, which was located on the east shore of Serua, is found on the north bank of the river. To its west it borders on the quarters that represent the Nila-speaking villages of North Nila. The river runs through the Amet quarters, one of the Serua-speaking villages of South Nila. The two remaining Serua-speaking Nila quarters, Sifluru and Wotai, are located at the south bank of the Pia River. The Teun villages are en­ closed by the Serua-speaking quarters of Terana and Lesluru (from Serua Island) in the East and of Sifluru, Wotai and Bumei (from Nila Island) in the North. The latter is a Teun-speaking exclave. North of the river the fourth Teun quarter, Mesa, is found, which on Teun Island was isolated in the extreme West.

Map 3: Seram island.

Although we presume the distribution of the quarters into island clusters to be motivated on traditional village alliances, it additionally enabled the initial maintenance of the TNS dialects in Waipia. By accident or on purpose, the location of Teun-speaking Mesa and Serua-speaking Waru outside the own island cluster prevented possible mutual exclusions based on 'incrowd senti­ ments'. Waipia inhabitants perceive themselves as TNS people, rather than Teunese, Nilanese or Seruans. The mutual comprehensibility of the Nila and Serua dialects enabled their initial maintenance, albeit probably under heavy mutual influence. Thanks to its relative isolated position in Southwest Maluku,

LANGUAGE ENDANGERMENT IN INDONESIA

61

Teun Island managed to function as a 'linguistic haven' for the Teun dialect. In Waipia, however, the three indigenous Teun-speaking quarters, Watuludang, Yefila and Mesa were crushed between the Serua-speaking quarters. The situa­ tion turned out worse for the Wetan-speaking enclaves Isu and Layeni from Teun Island. Whereas in the original setting they were adjacent to each other, in Waipia they are separated by Teun-speaking Yefila and surrounded by Serua-speaking quarters. In 1980 only in nearby Awaya, on the north coast of the Elpaputih Bay, a related vernacular was spoken by settlers from Southwest Malukan Leti and Moa. The only language-related demographic information on TNS is provided by Taber (1993)10, based on fieldwork in the late 1980s and early 1990s. Taber has incorporated the figures of the Teun-speaking Bumei quarter of Nila Island into his calculation of Teun speakers. Although they are mentioned in passing in another paragraph, Taber does not mention the amount of speakers of the Luangic isolects of Isu and Layeni. He is not clear about the composition of the figure of Nila speakers, which probably was motivated by the ambivalence of his informants on the mutual comprehensibility and status of the Nila and Serua dialects. Taber does not relate to the correspondence between Serua /v/, South-Nila /f/ and North-Nila /h/ that Collins mentions as the chief phonetic difference between the three isolects (Collins 1982:124). His Nila figure there­ fore seems to reflect ethnic sentiments rather than linguistic motivation. A comparison with the census in Riedel (1886), which was confined to a count of island inhabitants and did not consider their ethnolinguistic background, sug­ gests a slight increase of 6% and 5% of Serua and Nila people, against a de­ crease of 11% of Teun people. As already said earlier, it is unclear what the positions of Isu, Layeni and Bumei are in these calculations. The ethnic and religious riots throughout Indonesia in 1998 and 1999 drasti­ cally changed the demographic composition of the TNS District. The World Food Programme (WFP 2000) reports a massive influx of fugitives into the Waipia area, from February 1999 through July 2000. These 'Internally Dis­ placed Persons' or IDPs as they are labeled in the report, acknowledge to have come from the adjacent districts Tehoru, Amahai (specifically the city of Masohi) and Taniwel on South and Northwest Seram, and from the Banda Islands District about 150 kilometers south of Seram. All fugitives acknowledge to be Christians on the run for the ChristianIslamic violence in their home districts. According to the WFP, the known Southwest Malukan allegiance to the Protestant Church motivated them to seek refuge in the TNS District. Thus far it remains unclear, whether the IDPs fled 10

This information was adopted unchanged in Taber et al. (1996) and Grimes et al. (1996).

62

AONE VAN ENGELENHOVEN

in groups or individually - in August 2000 individuals from Wahai (North Seram District) have been reported to have crossed the inland jungles and mountains in order to arrive in the TNS District. 1886 Serua 34%

Nila 36%

Figure 2: 'Ethnolinguistic' proportion of the TNS population in 1886 and 1990

The fear for Muslim violence, perhaps in combination with a feeling of obligation towards fellow-Christians, made the TNS people provide shelter for these fugitives. The WFP (2000) reports that IDPs stay with volunteer host families. All households in the district are said to accommodate one or more refugees. The impact on the TNS community is obvious. The arrival of a re­ ported 8,650 IDPs increased the population with 50% and as such completely destabilized society. The small district (35 km2) appeared not to be able to provide enough food for everybody, while riots in the surrounding districts cut off the connections across land and sea with the Provincial capital on Ambon Island. Food and medication became difficult to supply, which explains the recent reports of starvation in the region.

LANGUAGE ENDANGERMENT IN INDONESIA

63

2000

Serua 20%

Nila 18%

Figure 3: Proportion of 'Internally Displaced Persons' (IDP) in the TNS District

When in 1950 the government of the Republic of East Indonesia acceded into the unitary Republic of Indonesia, an independent Republic of the South Moluccas, Republik Maluku Selatan (RMS), was proclaimed on Ambon Island, on April, 24 th. At that time, units of the Dutch colonial army, the KNIL11, were still encamped on the island of Java. In order to disband its colonial army, the Dutch government proposed its soldiers to enlist in the Indonesian Army. Most soldiers of Malukan origin, however, refused to do so. Because the Dutch Court of Justice disallowed any demobilization of Dutch army units abroad, the Dutch government had to transfer the remaining soldiers to Dutch territory. As a result, 12,500 Malukan soldiers and their families were transported to the Netherlands in 1951. A majority of these soldiers originated either from the Central Malukan islands (76%o) or from the Southeast Malukan islands (21%). A small minority of 3%), 330 households, came from Southwest Maluku. As has been explained elsewhere (Van Engelenhoven 2002), their numerical insignificance compared to the other Malukan households, and their ethnolinguistic complexity deter­ mined the linguistic behavior of these families. Especially the Babar house­ holds typically reflected the ethnolinguistic diversity of the home archipelago. In most Southwest Malukan marriages, both partners came from different language areas in the home region. Malay, the official medium in the colonial army and the traditional contact vernacular in Southwest Maluku, was the natural candidate to gain first language status in these households. The TNS families (approximately 110 households) appear to differ drasti­ cally from the other Southwest Malukan families in the Netherlands. Firstly, most marriages in the TNS group were contracted between partners Koninklijk Nederlands-Indisch Leger, 'Royal Netherlands-Indies Army'.

64

AONE VAN ENGELENHOVEN

Figure 4: Ethnic composition of Malukan migrants in the Netherlands 12

from the same dialect group.13 Beside that, the similarity between the Serua and Nila dialects - and the first mentioned functioned as a contact vernacular among the islands - made cross-dialect communication possible. A quick glance at the ethnolinguistic composition of the Babar and TNS groups in Figure 5 immediately shows the unfeasibility for any Babar language and the principally good chances for the Nila and Serua dialects to survive. Secondly, the migration history of the TNS people in the Netherlands slightly differs from the other Malukan soldiers. The latter were initially ac­ commodated in temporary hostels, which had either been concentration camps during the German occupation, or former convents and the like. The soldiers and their families were distributed among these hostels according to religious conviction and geographic origin14 rather than ethnolinguistic background. In Van Engelenhoven (2002) it is explained how the initial accommodation of Protestant Southwest Malukans among Roman-Catholic Southeast Malukans caused the relocation of the first mentioned in Protestant Central Malukan residences. When in 1964 the Indonesian army had eliminated the final hearths of RMS resistance in Maluku, this resulted in a political segregation in the Malukan exile community of supporters and adversaries of an independent 12

Central = Central Malukans, SE = Southeast Malukans, SW = Southwest Malukans. Admittingly, the TNS figures in Van Engelenhoven (2002) represent island affiliations rather than linguistic allegiances. It is therefore still not entirely clear what the proportions among the dialect speakers are. Whereas the Teun people (9%) almost all returned to Indonesia, at least one Teun-speaking family from Bumei has remained in the Netherlands. 14 Only in 1999 the Southwest Malukan islands became an independent regency (van Dijk 2000:25, footnote 2). At the time of the arrival in the Netherlands, the Southwest Malukan islands belonged to the Regency of Southeast Maluku, because of which Southwest Malukans were categorized as Southeast Malukans. 13

LANGUAGE ENDANGERMENT IN INDONESIA

65

Malukan republic. These contrasts were crystallized in the populations of the respective Malukan hostels and quarters, because of which several families were forced to move to other locations. As a consequence most Southwest Malukan families now live scattered throughout Babar Households

TNS Households

Nila 52% Figure 5: Ethnolinguistic composition of the Babar and TNS groups in the Netherlands. 15

the Netherlands. Most TNS soldiers, however, joined the navy and not the land-forces as the majority of Malukan exiles did. As such, they were not accommodated along with the other Malukan soldiers in the initial hostels, but instead were concentrated in the quarters of navy personnel in Den Helder, the main harbor of the Dutch Royal Navy. Both elements, the shared ethnolinguistic origin of the marriage partners, implying a monocultural household, and the geographic concentration of the TNS households gave the TNS people more possibility to maintain their own cultural identity. The other Southwest Malukan families, scattered throughout the Netherlands, either had to adopt and adapt to the 'Alifuru Concept' - the 15

Legend: D: Dawlor-Dawra; E: Emplawas; I: Imroing; M: Marsela; N: North Babar; SEB: Southeast Babar; T: Tela; W: Wetan.

66

AONE VAN ENGELENHOVEN

political and cultural framework of the Central Malukan majority in the hostels - or, alternatively, had to integrate into the Dutch society. 5 Language economy in the Indonesian and Dutch settings The different demographic histories of the TNS people caused separate language economies in the Indonesian and the Dutch settings. As already men­ tioned earlier, we hypothesize that the TNS language was a typical Southwest Malukan language in its original setting on the islands of Teun, Nila and Serua. Like in most Southwest Malukan island communities, traditional knowledge about society must have been stored in the lexicon as word pairs, or lexical parallelisms (Van Engelenhoven 1997). Verbal arts, story-telling and songs primarily functioned as a means to divulge this specific type of knowledge (Van Engelenhoven 1999). Ongoing research shows that the Serua lexicon has special items for planting, harvesting etc., which demonstrate the importance of agriculture in the traditional TNS communities. Seafaring, however, had a far more important economic impact on society. Highly specialized terminology in the Serua lexicon, for example tomra wroa (stare boat) 'watch a departing boat until it has passed the horizon', emphasizes the importance of seafaring in insular economy and its impact on the communities. Malay, both the local variant and its standard counterpart Indonesian, functioned as the official ver­ nacular in schooling, sermons and government as everywhere throughout Indonesia. The complete removal of the TNS population to South Seram from 19791983 may have had little direct influence on the communities in first instance. As has been mentioned above, the original villages were rebuilt as separate quarters in the Waipia area. The quarters were clustered into island groups as an asset to safeguard the traditional alliances between the villages. For the Wetan-speaking Teun enclaves of Isu and Layeni, however, the new location appeared to be a set-back. In the original setting on Teun Island both villages were isolated together from the Teun-speaking villages, whereas a strait be­ tween Teun Island and Nila Island barred any possible direct influence from the latter. In the new setting the Wetan enclaves are completely surrounded by Serua-speaking quarters. In the case of Layeni, the speakers of the Serua dia­ lect (Jerili and Wotai) and the Teun dialect (Yefila) live across the street. In a personal communication, Mr. P. Sullivan, who stayed in Waipia for his PhD research from 1998 through 1999, indicates that the TNS language in that period was confined to gatherings where representatives of the original villages met to discuss traditional topics. Interestingly, Mr. Sullivan specifically men­ tions the use of the indigenous language in interinsular meetings. This means that each speaker either used his own dialect, or a 'creolized' or koine variant

LANGUAGE ENDANGERMENT IN INDONESIA

61

of the TNS dialects. Another option is that the Serua dialect functioned as an intermediate vernacular, since it had that function in the original setting too. In a Southwest Malukan context, traditional affairs in an interinsular setting are highly formal, even when the vernacular is a creole (cf. the 'Sung Language' in Southwest Maluku, Van Engelenhoven & Hajek 2000:120). Indonesian is confined to the formal registers of the government, church and school domains, as it already was in the original setting. Ambonese Malay is the vernacular used in informal speech in school, in church and on television and radio. This has been epitomized in Table 9. Mr. Sullivan reports that, just before the massive inflow of refugees, TNS people started to reconsider their ethnic background. Should they identify themselves as TNS, or as either Malukans from the Central or Southeast Re­ gency, or perhaps even as North Malukans? Affiliation to the original grounds on Teun, Nila and Serua encourage inhabitants of Waipia more and more to return to their islands, either legally or illegally. A census held by Mr. Sullivan in three quarters revealed that nearly 50% of the population has been born on Seram. Whereas Malay in the original setting was confined to the church, school and government, it now entered domains traditionally belonging to the TNS language. Dutch informants report that the TNS language in Waipia is still used in the home domain among those who experienced the removal to Seram as adults. P. Sullivan reports that the strong influence of Malay in daily life frustrates the transfer of the TNS language from these adults16 to their children. The Waipia district is located where the North and East coast of Elpaputih Bay meet. Two indigenous languages are spoken in the coastal region: Amahai and Yalahatan. Amahai is reported to be known by a few of the oldest genera­ tion in nearby Makariki. Collins (1982:110) reports that Yalahatan, or Atamanu as he labels it, is still spoken by a few elderly in Haruru near Masohi. Both languages are highly endangered and respectively had 50 and 1,000 speakers around 1996 (Taber et al. 1996:55, 66). The only remaining indigenous lan­ guage in the interior is South Nuaulu. Ellen (1993) describes how its speakers, nowadays a reported 1,000 (Taber et al. 1996:59), are found in several small settlements at the Banda Sea coast and in the area further east where Amahai District borders on the Tehoru District. The east coast of Elpaputih Bay was developed as the initial transmigration area of Seram. For that purpose, a new administrative center for entire Seram and the Uliase Islands (Ambon, Haruku, Saparua and Nusalaut) was set up at Masohi. This explains the high amount of immigrants in this region. The major 6

P. Sullivan specifically mentions mothers trying to teach the TNS language to their children.

68

AONE VAN ENGELENHOVEN

immigrant group originates from Buton (Southeast Sulawesi) and is found especially in the town areas of Makariki, Masohi and Amahai. They have also settled among the other immigrants in most locations along the eastern shore of Elpaputih Bay. The other settlers came from the Selayar archipelago (South­ west Sulawesi), Java, Saparua, the Banda Islands, and the Kei and Tanimbar archipelagos (Southeast Maluku). In this complex ethnolinguistic scenario, Ambonese Malay, the traditional contact language in the region, naturally functions as the vernacular for interethnic communication, pushing the use of the indigenous languages back into the household. The WFP (2000) report identified the fugitives in the Waipia District in 2000 to have come from Masohi, the adjacent Districts of Tehoru (around Teluti Bay) and Werinama (Southeast Seram), the Taniwel District (Northwest Seram) and the Banda Islands District, approximately 150 kilometers south of Seram. Teluti (or East Littoral in Collins 1982, 1983) and Bobot are the biggest indigenous languages in the first two districts with about 17,000 and 4,000 speakers, respectively (Taber et al. 1996:57, 64). The major indigenous lan­ guages in the Taniwel District are Alune and North Wemale17, which, however, are mostly spoken in the interior. On the Banda Islands a Malay dialect is spoken, which varies slightly from Ambonese Malay (Adelaar et al. 1996). WAIPIA

NETHERLANDS Dutch

Malay

* * *

* *

TNS

Domain

Register

Domain

state govern­ ment

state govern­ ment

tradition

tradition

church

Formal

church

Ind

Malay

* * *

school

school

*

radio televi­ sion

radio televi­ sion

*

local govern­ ment

local govern­ ment

*

tradition

*

* * *

tradition

* *

Informal

church

church

home

home

TNS

*

*

*

*

Table 9: Language use in Waipia and the Netherlands in 1998-99 17 Taber et al. 1996 reports 13,000 and 5,000 speakers for Alune and North Wemale, respectively (pages 42 and 52). These figures, however, do also include speakers in the adjacent Districts of Kairatu and North Seram.

LANGUAGE ENDANGERMENT IN INDONESIA

69

Above it was already explained that in 1998 and 1999 Ambonese Malay had taken over most domains in the Waipia community and hindered the teaching of the TNS language to younger generations. The TNS language - besides being used among elderly people - was at that time already confined to tradi­ tional topics like marital law and reciprocal village obligations. We hypothe­ size that the massive influx of refugees into the Waipia District will further restrict the use of the TNS language. The WFP (2000) elaborates that each TNS household in Waipia hosts at least one up to five fugitives. In the case of Isu, for example, this implied a population increase from 546 to 2,352 people. Before their arrival the TNS language already was becoming obsolete in the home, being used only between spouses who were adults at the time of emi­ gration from Teun, Nila and Serua. In this context in which households were expanded with speakers from other ethnolinguistic regions, the use of the TNS language can only have come to a complete stand-still. The linguistic economy of the TNS people in the Netherlands is partly different from the other Southwest Malukan migrants. As elaborated above, unlike the other Southwest Malukan migrants, the TNS group is fairly homo­ geneous culturally and linguistically. Most TNS soldiers had joined the Dutch Royal Navy and as such were housed in the naval base of Den Helder. Other Southwest Malukans were scattered throughout the Netherlands and stayed in hostels where the majority was either ardent adherents or adversaries of an independent Malukan republic. Van Engelenhoven (2002) explains how initially an anticipated return to a liberated Republik Maluku Selatan (RMS), the Republic of the South Moluc­ cas, dominated daily life. A special committee, the CRAMS18 maintained law and order in the hostels and had constituted itself as the official representative of the guerilla government on Seram Island. It developed an RMS ideology, which was based on the cultural conceptions of the Central Malukan majority. The main function of the 'Alifuru Concept' as it is labeled by Van Engelenho­ ven (2002) was to present the Malukan migrants to the Dutch community as one people having one language and one culture. This would facilitate a swift return to Maluku after it was liberated from the Indonesian occupying forces. According to the 'Alifuru Concept', Seram is the origin of all Malukan islands and the cultures and societies found on them. The Seramese mountain tribes (called Alifuru in Ambonese Malay) are acknowledged as the original inhabi­ tants of Seram. Because of their relative isolation from the coastal peoples their societies are perceived as reflecting the purest form of Malukan culture. Simi­ larly, their languages, narratives and songs are expected to contain original 18

Commissie voor de Rechtspositie van Ambonese Militairen en Schepelingen 'Committee for the Legal Status of Ambonese Military and Naval Soldiers'.

70

AONE VAN ENGELENHOVEN

Malukan knowledge that dates from before the time Christianity and Islam were introduced in the region. Exactly these properties are seen as displays of retardedness and ignorance in the Indonesian setting, where Alifuru became a denigrating term. In the Dutch setting, however, they enhanced their function as role model for the citizen of the future RMS. The imposition of this Central Malukan model was the main drive behind the maintenance of the own language and culture among the Southeast Malu­ kan Keiese, who rejected the 'Alifuru Concept'. Like the Southeast Malukans, the cultural patterns of the Southwest Malukans diverged from the Central Malukan model. Instead of openly disassociating from the RMS ideology and the 'Alifuru Concept' as Southeast Malukans did (e.g. the collective Indonesian naturalization of Keiese in the 1970s), most Southwest Malukans chose to conceal their cultural identity, a phenomenon typical for their home region (Van Engelenhoven 1999, Van Engelenhoven & Hajek 2000). Numerical insignificance, internal ethnolinguistic pluriformity and the tradition-based culture concealment barred the transfer of the own languages and cultures to younger generations. Songs and dances were not performed and children en­ couraged their parents not to speak their own language in public, often para­ phrased in Malay as Jangan bikin malu ('Do not embarrass me'). Their relative isolation as naval soldiers shielded the TNS soldiers from the 'Alifuru Concept' as it had developed in the RMS-oriented hostels of the other Malukan soldiers. TNS people always were aware of the fact that their culture diverged from the RMS model. Up to the 1980s, therefore, they were in general reluctant to demonstrate their cultural heritage outside the own group, which was considered to be 'inappropriate'. Whereas cultural concealment among other Southwest Malukans meant a complete ban on 'un-Alifuru' dances and songs, these kinds of cultural displays became typical 'in-law' events not to be performed outside the family. Disassociation with radical expressions of RMS ideology (e.g. the train hijackings (1975, 1977) and especially the occupation of a primary school (1978)), the independence of the only remaining Dutch colony, Surinam, in 1975 and the demographic consequences of the govern­ ment's family reunion policy for foreign laborers caused TNS people to more and more disseminate their own cultural heritage in the 1980s. In the 1990s there was at least one foundation, Letra Natu 'The Four Villages' that guarded the interests of especially the Seruans in the Netherlands. After 50 years of exile the 'Alifuru Concept' has lost its impact on the Malukan community. Ethnolinguistic diversity is now considered to be an asset to the Malukan iden­ tity. The public performance of a TNS dance group on the commemoration of the arrival of the first Malukan families in Rotterdam (April 21 st , 2001) would have been impossible during the 1960s and 1970s.

LANGUAGE ENDANGERMENT IN INDONESIA

71

A comparison of the language use in the Indonesian and Dutch settings in Table 9 reveals that the TNS language completely disappeared in the Nether­ lands, whereas it was mainly confined to formal discussions on traditional topics in Waipia up to 1999. Informally it was used only among adult peers. In the Dutch setting marriage partners may have spoken a TNS dialect with each other in the home. Outside the home it was completely suppressed during the first thirty years of their stay in the Netherlands. The use of Malay in the formal register in the Netherlands slightly distorts the overall language use picture. Dutch primary schools indeed offer Malay language courses to Malukan children. This Malay variant, bahasa Melayu Maluku, 'Malukan Malay' was developed at a Dutch teaching program as an alternative for Indonesian, which among Malukan migrants was still perceived as the 'language of the enemy'. These courses, however, are additional to the school curriculum, in which Dutch is still the only language of instruction. This Malay variant is confined to schools. The Malay used in sermons depends on the political affiliation of the church. RMS-oriented churches, as for example the Malukan Evangelic Church, use bibles and song books that display the Malay orthography of the Netherlands-Indies period, while Ambonese Malay and Dutch alternate in sermons. In the Southeast Malukan Protestant Church and the Malukan Protestant Church Indonesian is the official medium in printed materials. The use of Indonesian in sermons is promoted by flying-in ministers from Indonesia. Already in the barracks in Indonesia Malay had taken over the lingua franca function among TNS people that traditionally was assigned to the Serua dialect. It therefore does not come as a surprise that Malay and not the TNS language became the first language among adults in the Netherlands, just as in the other Southwest Malukan groups. 6 Language obsolescence or linguistic innovation? As already explained in section 4, TNS people are traditionally bilingual. This factor must be taken in consideration in order to appreciate the type of obsolescence that their dialects are and have been subjected to. In a Southwest Malukan context a language does not necessarily occupy each register and each domain. Rather, one language, for example Malay, may be used for one or two domains (education and religion), whereas in Oslepan (North-Nila) the Nila dialect would be used for daily speech and the Serua dialect for trading pur­ poses or tradition-based politics with villages on Teun and Serua. This is why in the original setting the TNS language was not considered emblematic for the ethnolinguistic identity. A speaker's command of a language in question re­ flected his involvement in and acquaintance with a certain domain, for example trade, rather than his representativeness as a member of his people. The previ-

72

AONE VAN ENGELENHOVEN

ous section outlined how language allegiance has shifted from the indigenous language to either Malay in Waipia or to Dutch in the Netherlands. Are there formal features in the language that corroborate this obsolescence, if in fact one might question whether the TNS language ever had an 'ideal speaker' in the Chomskyan sense? In section 3 it was suggested that the pronominal system in Serua had en­ tered a process of simplification. Table 5 showed that in Serua pronominal suffixation of nouns indicating inalienable possession became obsolete. In elicitation no possessive suffixes where found and in Workala's (1993) manu­ script only the words for mother and father were attested with the suffix of 1 sg or 2sg. More interestingly, possessive pronouns are more and more used as subject markers instead of the appropriate pronominal prefixes. Tables 6 and 7 showed that in plural conjugation possessive pronouns are procliticized onto verbs that are inflected with the 3sg subject prefix. Possessive pronouns are also procliticized onto predicative adjectives where they cross-reference with the subject (6): (6)

sa'u sa' lsg.POS 'My son is

rapi rapi boy already

ni ni 3sg.POS big.'

lapanwait. lapna=wait big=PERF

(Serua)

For comparison example (7) displays an attributively used adjective: (7)

tamkonianlapna a'o? ta=m-konu=ina=lapna a'a=o NEG=2sg-eat=fish=big DEM=Q 'Do you not eat that big fish?'

(Serua)

This use of possessive pronouns deviates from what is generally known about Southwest Malukan languages. Luangic languages, as for example Leti, only have a few adjectives, which are always verbalized by means of a pro­ nominal prefix when used predicatively. Process verbs are derived into adjec­ tives through reduplication. Such deverbal adjectives may fill the predicate in equational clauses (8), which, however, is not very common. Normally, proc­ ess verbs feature a pronominal prefix in the predicate. Only when they indicate a state instead of a process they are nominalized by means of a pronominal suffix (9):

LANGUAGE ENDANGERMENT IN INDONESIA

(8)

asu asu dog:DET 'The dog is

mèmètme mè-mètma=e RED-black=DET a black one'

(9)

asu asu dog:DET 'The dog is

mètamne mètma-nV black-3sg.POS black'

73

(Leti)

(Leti)

The construction is not attested in the indigenous languages of the Waipia area. In South Nuaulu adjectives may function as predicates in equational clauses and as such are intonationally marked (10, taken from Bolton 1990:106-107). By means of a phrase-final demonstrative subjects can for­ mally be separated from predicates. Ongoing research suggests this to be obligatory in Amahai non-verbal clauses as shown in the following Amahai example provided by Margaret Florey (p.c.): (10)

asu mainae (Nuaulu) dog big 'The dog is big'

(11)

asuro ne ahia kio! dog DEM rotten EXC 'This dog is bad, you know! '

(Amahai)

Ambonese Malay does feature a similar construction. These, however, are mere expletive phrases and not clauses, as in (12): (12)

kapal pung kancang! (Malay) ship POS fast 'Is this ship fast! (literally: the fastness of the ship)'

Simplification of the pronominal paradigm is attested throughout Maluku and need not necessarily indicate language obsolescence. None of the Luangic languages (e.g. Leti, Luang, Wetan etc.) formally distinguish between alienable and inalienable possession, but they are just as vital as their relatives to the West (e.g. Meher, Roma and East-Damar) where alienable possession is en­ coded differently from inalienable possession. The severely endangered Laha language on Ambon Island on the other hand only has one suffix for all inal­ ienably possessed nouns and cannot distinguish person, but it still has a full set of pronominal proclitics to indicate alienable possession (Laidig 1993). Collins (1980) explains this type of language erosion to be induced by the pressure of Ambonese Malay. All Laha speakers are bilinguals who are just as fluent in Ambonese Malay as in their native tongue. A similar type of language erosion has been attested in the Keiese migrant community in the Netherlands.

74

AONE VAN ENGELENHOVEN

Among recently immigrated Ewaw speakers only one of the fourteen numeral classifiers is still used. Keiese of the second and third generations born in the Netherlands no longer speak the language and use semantically bleached cli­ ches like Fel be? 'How are you?' as shibboleths of Keiese identity (Van Enge­ lenhoven 2002). Florey (1997) noted a comparable simplification among Alune speakers in Taniwel District (Northwest Seram). Here, speakers in the coastal communities either no longer distinguish formally between alienable and inal­ ienable possession, or use the 3sg suffix for all persons on inalienably pos­ sessed nouns. Collins' (1991) conservative Serua data were collected during several fieldwork trips from 1977 to 1979 in the Letwaru village near Masohi (Collins 1982:135, footnote 2, 1983:135-6). Because this settlement dated before the TNS transmigration, we can therefore safely consider Collins' data to reflect the original grammatical structure of these isolects. Workala's (1993) data have been collected in the early 1990s in Waipia and consequently reflect the Serua grammar about ten years after the foundation of TNS District (see Table 7). Florey (p.c.) mentions the removal of the villages to the coast from their origi­ nal setting in the mountains as an important activator of the simplification in genitive constructions in Alune. From this perspective the complete removal of the TNS people from their islands and their relocation in South Seram probably also initiated the attested structural changes in the Serua isolect: the loss of the pronominal suffixes indicating inalienable possession and the restructuring of possessive pronominal proclitics as markers cross-referencing subjects on verbs (see section 3). Subject cross-referencing by means of possessive pronouns in adjective clauses has been attested neither in the languages indigenous to South Seram nor in the local lingua franca, Ambonese Malay. Although a similar construc­ tion has been attested in Leti, albeit that this language uses possessive suffixes rather than proclitics, this construction may very well be an innovation that is exclusive for the Serua dialect. Elicitation revealed that unlike Leti (13), it is impossible for adjectives, or deverbal adjectives for that matter, to function as adverbs. In order to encode a similar notion in Serua, the verb is placed in the subject slot of an adjective predicate (14). Further research is required to de­ termine whether this is an original grammatical structure or another 'migrationinduced' innovation. (13)

n-kari pèppèrte. n-kari pè-ppèrta=e 3sg-work RED-heavy=DET 'He works hard'

(Leti)

LANGUAGE ENDANGERMENT IN INDONESIA

(14)

sa'u karei ni f erf erta. sa'u karei ni fer-ferîa lsg.POS work 3sg.POS RED-heavy 'I work hard (lit. my work is hard).'

75

(Sema)

7 What to salvage in acute language death? Steinhauer (1994) explains how the shift of an indigenous language, or regional language as he labels it, to Indonesian is effectuated not only by edu­ cation, but also through exposure and interethnic contacts in which Indonesian or Malay is the lingua franca. The importance and influence of Indonesian and Ambonese Malay in South Seram is evident. Although Waipia was founded as a separate district for TNS people, it has always been economically dependent on the nearby port of multiethnic Makariki. Indonesian has always been the language for the official register. Ambonese Malay, the lingua franca of the region, easily replaced the Serua dialect as the vernacular for interethnic com­ munication among TNS people themselves, leaving the home as the only place to speak the TNS language. The fact that around 1998 about 50% of the popu­ lation had been born in Waipia pushed it even further back in the home to communications between grandparents, and perhaps between parents that were born on the islands. Mr. Sullivan's observation that mothers tried but not suc­ ceeded in teaching the TNS language to their children indicates that around 1998 this language was about to take the path of obsolescence as Ewaw had done in the Netherlands. The confinement of the TNS language to the formal register in traditional affairs on the one hand safeguards a high status for this language. Because it is not equipped for most topics in discourse, the preoccu­ pation with their ethnic identity as reported by Mr. Sullivan predicts a scenario in which Malay-speaking TNS youth will use stock-phrases in the TNS lan­ guage as a linguistic means to express their non-Seramese or non-Central Malukan ethnicity. The TNS people are aware that they lose their language. Already in 1996 during a conference at Pattimura University (Ambon), a lecturer of Nilanese origin informed the author that the elderly in Waipia were looking for aca­ demic help to save their language. In 2000, before the first reports of starvation caused by the ongoing instream of fugitives in Waipia, the author received a letter from Mr. M. Workala on behalf of the Letra Natu foundation, proposing a project among TNS people in the Netherlands that was similar to the one

76

AONE VAN ENGELENHOVEN

begun with Ewaw-speakers in Zwolle19, of which I want to quote a few lines here: 'After 25 years the Serua language is threatened. Elderly people who can still speak the language are dying. Indonesian is taught to youngsters at school. ... The language used to be transferred from generation to generation. Nothing has ever been reported about our language. ... We hope to maintain our language for the generations of Seruans to come and for everybody who is interested in the Serua language. ... Our means are limited; therefore we address you in order to record our endangered language.'

The ongoing state of kerusuhan 'disturbance', as the ethno-religious civil war in Central and Southeast Maluku is labeled by Dutch and Indonesian Malukans, has changed an incipient process of language shift among TNS people into an acute thread of language death, simply because its speakers are either killed in the riots or dying because of starvation (see section 5). What can a linguist do in such a context? In how far is linguistic support and research legitimate? Obviously, the main concern of the inhabitants of Waipia now is how to stay alive, while their present weal and woe worries their relatives in the Netherlands. Securing the lives of the TNS people must now be done before anything else. Nevertheless the fear of losing their language is an element where lin­ guistics can help. M. Florey (p.c.) observed that inside the village people could speak either Malay or Alune or both as one liked it. On the shore, outside the village, however, the main function of speaking the language was to mark oneself off as an Alune from the other non-Alune road-users. Whereas the 'Alifuru Concept' in the Netherlands precluded such language attitude, it can be anticipated in Waipia. In Central Maluku, especially on the Uliase islands off the Elpaputih Bay, indigenous language now only functions as a vernacular for traditional affairs, which is either sung or recited. It is, however, not longer used for daily communication. The loss of knowledge of their language may lead young TNS people to perceive their own language also as a source of traditional knowledge, which must be cherished and protected. Speakers of the TNS language in Indonesia and in the Netherlands therefore need to be taught that 'freezing' their oral traditions in the end will only add to the extinction of their language. The TNS language has always shared the sociolinguistic stage

19

The Zwolle project was executed within the framework of the international research project Language Maintenance for Endangered languages: an active approach, funded by the Australian Research Council (A59803475). The result of the Zwolle research, a learner's grammar of Ewaw plus texts is planned to be published at the Utrecht-based Landelijk Steunpunt Educatie Molukkers (National Malukan Education Support).

LANGUAGE ENDANGERMENT IN INDONESIA

77

with other languages, like Indonesian and Ambonese Malay, and as such was not the only receptacle of knowledge. Nevertheless I acknowledge the need to document the dialects of Teun, Nila and Serua. Steinhauer (1994) pointed out that a fair knowledge of the lan­ guages in use is necessary for a better understanding of the present linguistic mosaic of Indonesia. Certainly within the narrower scope of East Indonesia, the TNS language is indispensable for a better understanding of the many histori­ cal linguistic relationships and the many typologies in the region. Suppose, the civil war may one day settle down, what can linguists do to save the TNS language before it is too late? One of the things that can be done now is commencing research in the Netherlands among the few remaining initial migrants who still know parts of their language. Another option is to flyin people to the Netherlands from either the war zone or, perhaps much easier, people who migrated to other areas in Indonesia that are not yet affected, for example Kupang (Timor), or Surabaya (East Java). These informants, who certainly would have a better command of their language than the Dutch mi­ grants, could help in compiling a dictionary that would contain all lexical pairs in the TNS language. These are the main leads of indigenous knowledge man­ agement in the Southwest Malukan conception (van Engelenhoven 1997) and would be directly acknowledged as the major task for any language salvage program by TNS people. Alongside this dictionary project grammatical re­ search could be easily conducted. The amount of speakers of the TNS language has always been too small to guarantee its existence in the modern Indonesian society. However, as has been elaborated by Slomanson (1996), the high esteem that TNS people have of their own language gives the language a chance to survive in some domains next to Malay or Indonesian. The compilation of a dictionary containing all lexical data for three dialects would only emphasize the importance of the TNS language for its speakers and would certainly contribute to its maintenance.

78

AONE VAN ENGELENHOVEN

Appendix: List of quarters in Waipia & their island and language affiliation Dialect/language

Teun Island

Wetan

įsu Layeni

Teun dialect

Mesa Yerfilai Watuludang

Nila Island

Sema Island

Bumei

Serua dialect

Wotai Sifluru

Nila dialect

Amet Kokroman Kuralele

Waru Jerili Lesluru Terana

List of quarters in Waipia and their island and language affiliation

References Adelaar, K. Alexander, et al. 1996, "Malay: its history, role and spread". Wurm, Mühlhäusler & Tryon 1996. 673-693. Anonymous. 1937. "Mala". Stokhof 1981. 161-174. Blood, Cindy. 1992. "Subject-verb agreement in Kisar". Burquest & Laidig 1992. 1-21. Blust, Robert. 1993. "Central and Central-Eastern Malayo-Polynesian". Oce­ anic Linguistics 32. 241-294. Bobaljik, Jonathan David, Rob Pensalfini & Luciana Storto, eds. 1996. Papers on language endangerment and the maintenance of linguistic diversity. Cambridge, Mass.: Department of Linguistics, MIT (MIT Working Pa­ pers in Linguistics, 28). Bolton, Rosemary Ann. 1990. A preliminary description of Nuaulu phonology and grammar. MA thesis, The University of Texas at Arlington. Bradley, David & Maya Bradley, eds. 2002. Language endangerment and language maintenance. London; New York: Routledge Curzon. Burquest, Donald A. & Wyn D. Laidig, eds. 1992. Descriptive studies in lan­ guages of Maluku. Jakarta: Badan Penyelenggara Seri Nusa, Universitas Katolik Indonesia Atma Jaya (NUSA, 34). Collins, James T. 1980. "Laha, a language of Central Maluku". Indonesia Circle 23. 3-20.

LANGUAGE ENDANGERMENT IN INDONESIA

79

—. 1982. "Linguistic research in Maluku: a report of recent field work". Oce­ anic Linguistics 21. 73-146. —. 1983. The historical relationships of the languages of Central Maluku, Indonesia. Canberra: Australian National University (Pacific Linguis­ tics, D 47). —. 1991. Towards a description and classification of the language of Serua. Paper presented at the Sixth International Conference on Austronesian Linguistics, Honolulu, Hawaii, May 20-24, 1991. Dijk, Toos van. 2000. Gouden eiland in de Bandazee: socio-kosmische ideeën op Marsela, Maluku Tenggara, Indonesië. Leiden: Centre of NonWestern Studies, Leiden University. Duarte, Jorge Barros. 1990. Vocabulario Ataúro-Portuges, Português-Ataúro. Lisbon: Instituto Português do Oriente. Ellen, Roy. 1993. "Rhetoric, practice and incentive in the face of the changing times: a case study in Nuaulu attitudes to conservation and deforesta­ tion". Milton 1993. 126-143. Engelenhoven, Aone van. 1995. A description of the Leti language (as spoken in Tutukei). PhD thesis, Leiden University. —. 1997. "Words and expressions: notes on parallelism in Leti". Cakalele Maluku Research Journal 8. 1-25. —. 1999. "Epítetos e epítomes: uso e perda do conhecimento narrativo no sudoeste das Molucas (Leste da Indonêsia)". Encontros de divulgaçao e debate em estudos sociais 4. 99-107. —. 2002. "Concealment, maintenance and renaissance: language and ethnicity in the Moluccan migrant community in the Netherlands". Bradley & Bradley 2002. 272-309. — & John Hajek. 2000. "East Timor and the Southwest Moluccas: language, time and connections". Studies in Languages and Cultures of East Timor 3. 107-124. Florey, Margaret J. 1997. "Skewed performance and structural variation in the process of language obsolescence". Odé & Stokhof 1997. 639 - 660. — & Aone van Engelenhoven. 2001. "Language documentation and mainte­ nance programs for Moluccan languages in the Netherlands". Interna­ tional Journal of the Sociology of Language 151. 195-219. Grimes, Barbara F., Richard S. Pittman & Joseph Grimes. 1996. Ethnologue: languages of the world. Dallas: Summer Institute of Linguistics. Hull, Geoffrey. 1998. "The basic lexical affinities of Timor's Austronesian languages: a preliminary investigation". Studies in Languages and Cultures of East Timor 1. 97-202. Laidig, Wyn D. 1993. "Insights in Larike possessive constructions". Oceanic Linguistics 32. 311-352.

80

AONE VAN ENGELENHOVEN

Milton, Kay, ed. 1993. Enviromentalism: the view from anthropology. London: Routledge. Odé, Cecilia & Wim Stokhof, eds. 1997. Proceedings of the seventh interna­ tional conference on Austronesian lingustics, Leiden 22-27 August 1994. Amsterdam: Rodopi. Riedel, J.G.F. 1886. De sluik- en kroesharige rassen tusschen Selebes en Papua. The Hague: Martinus Nijhoff. Samloy, Semuel et al. 1998. Struktur bahasa Kisar. Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikan dan Kebudayaan. Slomanson, Peter. 1996. "Explaining and reversing the failure of the Irish language revival". Bobaljik, Pensalfini & Storto 1996. 115-135. Steinhauer, Hein. 1994. "The Indonesian language situation and linguistics: prospects and possibilities". Bijdragen tot de taal-, land- en volkenkunde 150. 755-784. Stokhof, Wim A.L., ed. 1981. Holle lists: vocabularies in languages of Indone­ sia. Vol. 3/2 Central Moluccas: Seram (11). Canberra: Australian Na­ tional University (Pacific Linguistics, D 44). Taber, Mark. 1993. "Towards a better understanding of the indigenous lan­ guages of Southwestern Maluku". Oceanic Linguistics 32. 389-441. — et al., eds. 1996. Atlas bahasa tanah Maluku. Ambon: Universitas Pattimura, Pusat Pengkajian Maluku, Summer Institute of Linguistics. World Food Programma (WFP). 2000. South-Central Seram Mission, 12-13 August. Retrieved at http://www.idpproject.org. Workala, Mozes. 1993. Serua-taal-bundel. Leiden: manuscript. Wurm, Stephen A., Peter Mühlhäusler & Darrell T. Tryon, eds. 1996. Atlas of languages of intercultural communication in the Pacific, Asia, and the Americas. Berlin: Mouton de Gruyter.

SIBE: AN ENDANGERED LANGUAGE GIOVANNI STARY University of Venice

1 Historical Background Information relative to the origins of the Sibe people is limited exclusively to Chinese sources, which link the ethnonym to the Shiwei of the Sui-Tang period (589-906) and to the Xianbi of the Han period (206 BCE-220 CE). With their present name "Sibe" they are found, in Manchu sources, from 1593 on­ wards among the components of a coalition of nine tribes fighting against the ever stronger domain of Nurhaci, the founder of the Latter Jin dynasty (re­ named Qing in 1636). For reasons which are still unclear, the Sibe already in that period (or immediately after) were subordinated to the Khorchin Mongols, and remained in this state of semi-subjection even after the submission of the IChorchin to the Qing and their consequent adaptation to the system of the Eight Banners. Due to this subjection to a Mongol tribe, the Sibe have often been erroneously viewed as belonging to a Mongol ethnic group. This has also given rise to rather imaginative speculations concerning their original lan­ guage: Mongol or Tungus? Rhoads (2000:53), quoting from a recent Chinese study, writes: "The Xibe, a New Manchu group of Mongol origin, were so adept at learning and absorbing Manchu that it eventually replaced Mongolian as their 'native' language." That the Sibe rather spoke a Manchu (Tungus) dialect is made clear in Chinese sources: during an audience which in 1692 the Emperor Kangxi granted to the Khorchin, in whose following were also people of the Sibe, the Emperor noticed that these latter ones spoke "his same language" among them. It was precisely due to this "Imperial discovery" that the Sibe were freed from Khorchin's rule and incorporated into the Eight Banners. This late incorpora­ tion and their consequent difficult integration within the Eight Banners, is probably the reason why the Sibe always had a special kind of administrative treatment which, in more recent times, has led to their recognition as a "na­ tional minority", separate from the Manchus. Therefore, quite rightly Janhunen

82

GIOVANNI STARY

(1996:113) concludes saying that "...even if the Sibe may have been culturally influenced by the Khorchin, there is no factual basis for claiming that they would ever have spoken Mongol, or any language other than Jurchen (Man­ chu)." The most significant event in Sibe history took place in May 1764, when 4,295 persons, among whom soldiers and their families - as according to an Imperial order - left their native lands and from Mukden headed towards the Ili Valley in Xinjiang, with the task of defending that newly conquered land and protecting the new frontiers of the Qing Empire. Thus began the so-called "Sibe diaspora", which led to the present subdivision in "Sinicized Sibe" (i.e. those who have remained in Manchuria), and those who have found a new settlement in distant lands inhabited by Turkish-speaking populations, an ele­ ment which is commonly indicated as the cause which has favoured the preser­ vation of the Sibe language up to this day. According to the 1990 census, Sibe population amounts to 172,847, with an increase of 106,68% if compared to the 1982 census (total population: 83,629). Of these, 27,315 persons (in 1982, 17,362 persons) reside in the Sibe Cabcal Autonomous District, which was instituted in 1954 on the southern banks of the Ili river, on the borders with Kazakhstan. It is precisely on these inhabitants of Cabcal (Chábùcháěr in Chinese) that scholars have concentrated their linguistic research, in the past as well as in the present. This because they represent the most compact and ho­ mogeneous group, maintaining their traditional customs along with the use of the Sibe language. Is Sibe an "independent" Tungus Jurchen language in its own right, or a simple "Manchu dialect"? An answer is found in the history of this people which, as we have already seen, appears with this name for the first time in 1593, together with other Mongol and Tungus tribes - the Yehe, Hada, Hoifa etc. These tribes, however, submitted directly to Nurhaci's rule, while the Sibe became subjects of the Khorchin Mongols and were therefore not included into the process of "ethnic unification" officially called "Manchu" only in 1635, i.e. an ethnonym which from that time on substituted by Imperial order the native term "Jurchen" (jušen). The linguistic situation of the young Manchu khanate was characterized by several Jurchen "tribal variants", and only one of them the dialect spoken at the khanate's court (the language of the Aisin Gioro clan, Jianzhou area) - was codified in written form and became the Manchu lan­ guage1. Side by side with this dialect, now the "state language", other dialects 1

If the Manchu khanate was not founded among the Jianzhou Aisin Gioro Jurchen, but among other jurchen tribes (for instance among the Jurchen from Nunjiang river), the Manchu language would appear with quite different characteristics: for example; "water" (in "classical" Manchu - m ) - would have been codified in written form in the Nunjiang spelling muku; "seed" (classical Manchu use) would be ucu, "shoulders" (meiren) would be mirin etc.

SIBE: AN ENDANGERED LANGUAGE

83

continued to exist in oral form (in some cases up to the present day). Among these dialects the variant of the Sibe people is found, which was not "in good time" to be amalgamated with other Tungus tribes already called "Manchu". As seen from these historical events, Sibe should be considered a Manchu dialect, where the definition "Manchu" should be regarded as an innovative synonym of "Jurchen". Written Sibe, indeed, is almost identical to classical Manchu. 2 The Evolution of the Linguistic Situation Sibe sources underline the fact that during the last Qing period, when the Manchus had already abandoned their mother-tongue in favour of Chinese, in many sectors of the Xinjiang bureaucratic administration Sibe officials were almost exclusively utilized precisely because they had perfectly preserved their language. In those days, the Sibe population was subdivided into eight settle­ ments or villages, numbered from one to eight according to the military com­ panies {niru). These settlements or villages have preserved their names up to this day, with the only exception of the Ningguci niru (i.e. the Sixth Com­ pany/Village) which has become the main town and has taken up the name "Cabcal" within the district bearing the same name. Already at that time, de­ spite the precarious economic conditions noted by Radloff (1866:88-89, 250264), there existed spontaneous initiatives for preserving the Sibe language which actually multiplied in number after the fall of the dynasty. In 1946 the newspaper Sulfanjilgan ("The Voice of Freedom") was started, which in 19361938 was preceded by the publication of six numbers of the literary journal Ulden ("Light") published in Urumqi, and in 1938-1939 by the publication of another newspaper called Fuldun. All in all, we know (although quite fragmentarily) of the publication of the following magazines and newspapers: Ulden, 1-6(1936-1938) Fuldun, (around 1938-1939) Sulfanjilgan, (1946-1949) Su wen maturu, 1-8 (1953-1954) Cabcal šu hūman, 1-6 (in the 1950s) Efin mutun, 1-2 (in the 1950s) Ice banjin, (1946-1966), which stopped with the "Cultural Revolution" and was started again in 1974 with the title Cabcal serkin (see below) Sabdan ci šanggaha boobei, 1-6 (-1979?) Gisun tamin folon, (1980-1983), from 1984 changed into Kuren folon Kuren folon, from 1984 (stenciled in Cabcal) Cabcal serkin, semi-weekly, continuation of Ice banjin and still in existence.

84

GIOVANNI STARY

No information is available on the magazine called Dergi Turkestan dashalangge serkin ("Revolutionary Messenger of Eastern Turkestan"), predecessor of the Sulfan jilgan and of the Sinlu bao ("Newspaper of the New Road"). Impossible to verify are also rumours asserting the existence of Sibe editions of the "Xinjiang Daily" (Xinjiang eìbào). Of the magazine Sibe su wen sibkin, in Chinese except for the title, only a single number was published in 1994. The best known magazine, which is still in existence, is the Sibe su wen (in Chinese: Xībó wénhuà), a half-yearly publication started in 1987. In 1947 there was a first reform of the Manchu alphabet in order to adapt it to the characteristics of the Sibe language. In the same year, a school was created in which all teachings were imparted in the Sibe language. On March 25 th , 1954, with the foundation of the Sibe Cabcal Autonomous District, the Sibe language was proclaimed to be the official language of the District. In this same year, the publication of the first books in Sibe was inaugurated. To this day, more than 400 books in the Sibe language have been published. Quite significant, in this regard, is the number of copies which have been printed: from the initial 1,000 copies per book, these have passed nowadays to only 350-400 copies. In the 1950s, there have been two projects aiming at substi­ tuting the Manchu alphabet with the Cyrillic one (in 1957) and subsequently with the Latin one (in 1959): both projects, however, never became operative. The "Cultural Revolution" of the 1960s put an end to all publications in the Sibe language. Moreover, the teaching of this language was prohibited in ele­ mentary schools. Although this situation ended in 1978, it had quite negative consequences on the very survival of the Sibe language. Indeed, eighteen years of prohibition determined, according to Sibe intellectuals themselves, "une generation entière de jeunes Sibe grandit, illettrée dans sa propre langue. Ce furent les années les plus noires de l'histoire de l'écriture sibe, menacée alors de totale disparition" (Su Chengzhi (Kichengge) 1995:200). With the reintro­ duction of the teaching of the Sibe language in the primary schools of Cabcal, new manuals were published which - besides the teaching of the mothertongue (Niyamangga gisun, 11 volumes) - also taught Geography (Natacin, 2 volumes), Mathematics (Bodocin, 10 volumes), Natural Sciences (Banjitan, 6 volumes), History (Suduri, 2 volumes) and Ideology (Sesiyang doro tacin, 12 volumes). The teaching in Sibe varies from school to school, as is shown in the following table elaborated in 19942:

This and all following tables are taken from Qi Cheshan (2000:12-19).

85

SIBE: AN ENDANGERED LANGUAGE

Primary school

classes

%of Sibe students

no. of Sibe teachers

weekly hours per year

1

2

3

4

5

6

1st (uju) niru

12

78

3

6

6

6

4

4

2 nd (jai) niru

6

45.2

3

3

3

3

19

59.5

5

-

3

3 rd (ilaci) niru

-

4

3

3

3

4 (duici) niru

15

68.4

5

5

5

5

5

4

4

5th (sunjaci) niru

12

83.3

4

4

4

4

4

3

3

th

th

6 (ningguci) niru

11

88.5

3

5

4

4

4

3

3

7th (nadaci) niru

8

55.6

3

4

4

4

3

3

3

8th (jakûci) niru

18

43.6

4

-

4

4

4

3

3

In general, the linguistic situation (in percentage) in the eight niru (compa­ nies/villages) is as follows: language utilized in the family

language utilized in external affairs

Sibe

Han

Sibe

Chinese

Sibe

Chinese

st

61.8

20.9

100

7.2

92.8

41.2

6.2

16.6

nd

72.2

15.5

100

12

85.6

73.6

30.4

34.4

1 (uju) niru 2

Ethnic component

(jai) niru

Uighur

Kazakh

3 rd (ilaci) niru

73.2

12.3

100

6.9

100

43.4

13.1

13.8

4 th (duici) niru

32.9

28.4

100

8.2

88.2

64.5

7.3

15.5

5th (sunjaci) niru

38.2

15

100

7

96

52

19

14.4

6th (ningguci) niru

29

32

96.9

44.9

88.6

72

28

31.1

7th (nadaci) niru

41.6

32

98.1

17.6

82.4

66.7

27.8

26

8th (jakûci) niru

72.2

15.5

99.1

14.5

86.4

69.1

26.4

29.1

3 The Present Situation The end of territorial isolation and the ever increasing contacts with the Chinese cultural world linked to the laws of market economy and technological development are all factors which necessarily and in an almost natural and automatic way favour the supremacy of the Chinese language, since the utili-

GIOVANNI STARY

86

zation of Sibe continues to be relegated within the boundaries of the Cabcal District. If one analyzes the knowledge and utilization of the Sibe language accord­ ing to the pattern of age, one gets the following table: Age

Family environ­ ment

External relations

Sibe

Chinese

Sibe

Chinese

Uighur

Kazakh

10

100

-

95.7

38.1

28.6

28.6

20

100

16.7

100

43.8

29.2

31.3

30

100

8.3

93.1

48.6

29.2

29.2

40

100

16.4

98.2

65

32.7

31.8

50

100

8.6

96.7

55.7

21.4

22.4

60

100

12

92

64

18

24

70

100

11

82

70

6

3

Again differentiating on the basis of age, the knowledge of the written language and its utilization may be deduced from the following table:

4

Age

Command of written language

Reading Sibe Purchase Sibe newspaper & books newspaper & books

Good

average

absent

often

occas.

never

yes

no

10

47.6

38.1

14.3

42.9

23.8

33.3

47.6

52.4

20

47.9

33.3

18.8

47.9

27.1

25

50

50

30

63.9

33.3

4.2

44.4

38.9

16.7

55.6

44.4

40

44.5

44.5

10.9

25.5

49.1

25.5

41.8

58.2

50

11.4

44.8

42.9

8.1

28.6

62.9

24.3

75.5

60

8

37

55

7

12

81

18

82

70

23

58

19

11

45

44

21

79

Conclusion Sibe experts agree in saying that there are two factors which threat the survival of the Sibe language and script: the necessity of a thorough learning of Chinese, and the interruption of the teaching of Sibe in secondary schools. As a consequence, Su Chengzhi (Sibe name: Kicengge) (1995:202) notes that after the primary school students tend to forget the written language in order to

SIBE: AN ENDANGERED LANGUAGE

87

focus on Chinese. Thus, after some years they are not able to read Sibe any­ more: "... les élèves ne savent plus ni lire ni écrire le sibe. Une enquête que j'ai effectuée montre que 95% des Sibe âgés de vingt à trente ans ne connaissent que la langue parlée". Moreover, according to Kicešan (in Chinese Qicheshän) (2000:19), the emigration of the Sibe population from Cabcal makes it so that "... the Sibe language, with lesser population using it, assumes a degenerating trend under linguistic and environmental pressures". Alarming signs are the decrease in the birth rate (in 1990, the Sibe of Cabcal were 27,315; in 1993 only 20,364) and the slow but constant drop of the number of publications (as well as number of book copies, see above) in the Sibe language. A good exam­ ple of this trend is represented by the Sibe su wen / Xíbó wénhuà magazine, which when it first came out in 1987 published 1,700 copies, whereas of its late 36th number, issued in 2002, only 500 copies were published. Moreover and quite eloquently, starting with number 15 (which came out in 1993) the maga­ zine has become bilingual, half of it being written in Sibe and the other half in Chinese. Officially, this is said to have been done so as to make the magazine accessible to the sinicized Sibe living in Manchuria. However, isn't this pre­ cisely one more sign of a gukume hamika gisun hergen, that is, of a truly "endangered language"? References Cing gurun-i dangse ci sonjome banjbuha Sibe-i suduri mutun [Historical materials on the Sibe chosen from Qing archives], I-II, 1987. Urumqi. Hoppe, Th. 1995. "Die ethnischen Gruppen Xinjiangs: Kulturunterschiede und interethnische Beziehungen". Mitteilungen des Instituts für Asienkunde 258. 374-399. Ikjintai. 2000. "Gukume hamika Sibe gisun hergen be aitubuci acambi" [The endangered Sibe language and script must be saved]. Sibe Su wen /Xibó wénhuà, 34. 1-10. Janhunen, Juha. 1996. Manchuria. An ethnic history. Helsinki: The FinnoUgrian Society. Kusunoki, Yoshimichi. 1989. "Horuchin-Mongoru shihaiki no Shibo zoku / The Sibes under the rule of the Korcin Mongols". Töyö gakuhö 70:3-4. 27-50. Qí Chēshān [recte: Qicheshän]. 2000. "Xinjiang Chábùcháĕr Xibó zizhixian Xíbózú yuyán wénzi shïyông xiànzhuàng" [The present situation of spoken and written Sibe in the Sibe Cabcal autonomous county in Xin­ jiang]. Mănyŭ yánjiū /Journal of Manchu Studies 2. 12-19. Radloff, W. 1866. "Das Ili-Thal in Hoch-Asien und seine Bewohner". Peter­ manns geographische Mitteilungen 3. 88-89, 250-264.

88

GIOVANNI STARY

Rhoads, Edward J. M. 2000. Manchus & Han. ethnic relations and political power in late Qing and early republican China, 1861-1928. Seattle: University of Washington Press. Stary, Giovanni. 1985. Geschichte der Sibe-Mandschuren. Wiesbaden: Harrassowitz. —. 1990. Manchu studies. An international bibliography, 2. Wiesbaden: Harrassowitz. Su Chengzhi (Kicengge). 1995. "Situation linguistique des Sibe du Xinjiang". Etudes mongoles et sibériennes 26. 197-207.

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY THE CASE OF YENISEYAN

STEFAN GEORG Leiden University The Yeniseyan languages are generally referred to as one of the so-called Palaeoasiatic (also known as Palaeosiberian) language families. This designa­ tion was introduced in the 18th century as a catch-all term for those indigenous languages of Asiatic Russia/Siberia, which do not belong to one of the greater language families which dominate this part of the world, and which extend far beyond the borders of this territory. Palaeoasiatic is thus a purely negatively defined term, referring to languages which are spoken exclusively in Siberia, between the Ural mountains in the West, the Ocean in the East and North, and the borders of Kazakhstan, Mongolia and China in the South. This criterion, together with the second one of not having demonstrable genetic relatives outside of this territory, defines the following language isolates and small families as belonging to this group1: Chukotko-Kamchatkan, Nivkh, Yukaghir, and Yeniseyan. The logical antonym of Palaeosiberian would be, of course, Neosiberian, and, though rarely used these days, this term has been coined by Czaplicka (1914) to unite all indigenous languages of Siberia belonging to larger, wellestablished language families, which extend beyond the confines of Asiatic Russia, viz. Uralic, Turkic, Tungusic, and Mongolic languages. The criteria which led to this dichotomy are only partly valid today2, yet all "Palaeoasiatic" languages do have in common that they 1 More precisely, this would define the group of Palaeosiberian languages; some linguists, as e.g. the editors of the Palaeoasiatic volume of the Russian series Jazyki mira ("Languages of the World", Volodin et. al. 1997) extend the latter term to include other language isolates of Asia as well, such as Burushaski (in Pakistan), or Ainu. Conventionally, Soviet and Russian linguists often add Eskimo-Aleut languages to their number of Palaeoasiatic families. 2 Thus, the theory that Yukaghir may not be a language isolate after all, but forms a part of Uralic (or, then, Uralo-Yukaghir) is gaining ground in specialist circles (cf. Nikolaeva 1988);

90

STEFAN GEORG

- are at best distantly related to their immediate neighbours3; - are found on the very fringes of the vast Siberian territory (Nivkh and Chu­ kotko-Kamchatkan languages on the eastern shores of the continent, Yukaghir in the Far North, Ket just below the Polar Circle), which lends support to the hypothesis that they - represent the remnants of formerly more widespread populations, which gradually lost ground to (or were pushed into residual areas by) speakers of Neosiberian languages, as the latter advanced into Siberia from more southerly areas (particularly Turkic and Tungusic). I will try to show below that various observations strengthen the assumption for such a scenario especially for the Yeniseyan languages. Few indigenous languages of Siberia may be viewed as not endangered at the beginning of the 21 st century. In fact all of them constantly lose ground to Russian, the one language which offers prospects for higher education, eco­ nomic success in urban life etc. to native Siberians. The number of persons who claimed one of the Palaeoasiatic languages of Siberia as their native lan­ guage in 1989, the date of the last general census held in the USSR, is given in table l 4 . population

speakers

%

Chukchi

15,106

11,163

73.9

Koryak

8,942

5,168

57.8

Eskimo

1,703

933

54.8

Ket

1,084

589

54.3

Yukaghir

1,112

398

35.8

Aleut

644

184

31.4

Nivkh

4,631

1,199

25.9

2,428

563

23.2

Itelmen

Table 1

this family may very well form part of a still larger grouping comprising Eskimo-Aleut and Chukotko-Kamchatkan, cf. Fortescue 1998 and, for a very thorough reasoning in favour of such a grouping on the basis of non-trivial correspondences in morphology, Seefloth 2000. 3 Nivkh is certainly a true isolate, theories arguing for external relations of Yeniseyan are extremely problematic, cf. Georg 2000. 4 Adapted from Janhunen 1991 ; the first column lists the number of persons who claimed membership in the respective ethnic group (or "nationality"), followed by the number of native speakers of the language and the percentage of speakers among the ethnic group.

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

91

In this list, the only Yeniseyan language mentioned is Ket (from ke 't p1. de' "person, human being"), and, indeed, it is the only language of the family sfill spoken. However, at least six different, and clearly distinct, languages make up the whole family: apart from Ket, Yugh, Arin, Assan, Pumpokol, and Kott have been, at least partially, recorded and identified as forming the closeknit language-family, which came to be known as Yeniseyan, since all varieties belonging to it are or were spoken on the banks of the Yenisey river or within the area of its large tributary system. Their genetic relationship has always been clear and uncontroversial, so we need not dwell on this here5 (the question whether the genetic unit Yeniseyan may, after all, form part of a larger genetic grouping together with other languages and families of the Old (or even the New) world is, however, very controversial and continues to be debated vigor­ ously). In the following, I will try to summarize the history of the Yeniseyan lan­ guages in a rough chronological order, beginning from reconstructed ProtoYeniseyan and its discernable interactions with other language families of Inner and Northern Asia down to the present day. While the history of every language which finds itself in the situation of serious endangerment in our times will naturally be a history of decline and retraction, the case of Yeniseyan is somewhat special, among the languages of Northern Asia at least, since it can be shown that this language family once not only occupied a con­ siderably greater territory than its remnants do now (or did as late as at the time of the Russian conquest), but that its speakers may once have played a quite important role in one of the more influential theatres of world history, namely the South Siberian/Inner Asian steppe zone. Arguments in favour of this may be drawn from two types of sources, toponymy and etymology. Ever since it was first developed in the context of Indo-European linguistics, the study of toponymy has been regarded as a potentially powerful device for the determination of prehistoric ethnolinguistic territories and boundaries; toponyms, especially hydronyms, are generally known as diachronically highly stable, often surviving repeated changes of populations, which continue to use the names for rivers and other natural objects in spite of the fact that they may be semantically meaningless for them. The wealth of place-names in the Americas which have been accepted and taken over by European settlers is a well-known testimony to this fact. Outside of Europe and the Americas, the historical study of place-names as an aid to determine the ethnolinguistic

5

For the reconstruction of the parent language ("Proto-Yeniseyan") cf. Starostin 1982, Verner 1990.

92

STEFAN GEORG

makeup of a region for the time before written records appear has not been employed widely, but a remarkable exception is Western and Central Siberia. Due to the efforts of A.P. Dul'zon, one of the pioneers of Ket/Yeniseyan stud­ ies in the Soviet Union after World War II and the founder of the Ketological school at the Pedagogical University of Tomsk, thousands of place-names of the region have been collected and linguistically analyzed. One of the most remarkable results of this large-scale project was that a large territory in Cen­ tral Siberia is characterized by place- and river-names, which are not analyzable from the languages spoken there in historical times, but rather from Yeniseyan. This territory extends well over the great rivers Ob' and Irtysh in the west, reaches the Irkutskaja Oblast' in the east, and finds its southern boundary only in the Sayan/Altay mountain range in Southern Siberia/Western Mongolia. The toponomastic "leading fossils", which reveal the Yeniseyan origin of these names are frequently found suffixal elements, which are matched by Yeniseyan words for "water" and "river", such as -ses (cf. Ket ses "river"), -ces (Yugh), -tet (Pumpokol), -set (Arin), -set (Kott), or -ul (cf. Ket ul "water"). This vast area is now mostly inhabited by Uralic-speaking peoples, Ob'Ugrians and Samoyeds (now extinct Samoyedic languages - Kamass, Koibal, Tawgy, Mator - were also spoken here, Kamass well into the 20th century); the other ethnolinguistic group which is now dominating this territory is Turkic, with Tuvan and Tofa, Khakas, Shor, Oirot/Gorno-Altai, and the ChulymTurkic language being present in the territory where Yeniseyan place-names dominate. One of the earliest groups of Turkic written documents comes from exactly this region - the Yenisey-Runic inscriptions. Yeniseyan loanwords in Southern Samoyedic and Turkic languages like Khakas and Shor, which cannot have been in contact with Yeniseyan for at least half a millennium, further add to this picture. Cf. Čispijakov 1976, Čispijakova 1992. However, neither the reconstructable Yeniseyan proto-language nor the names of rivers and villages do provide any clue to answering the question of dating, i.e. when did Yeniseyan speakers come to be present in this region, when did they begin to exert linguistic influences on other languages of the region and when did they cease to do so. In the absence of direct historical sources, or the clear and unambiguous identifiability of an archaeologically observable culture with early Yeniseyan speakers this question may well remain unanswerable forever, but one indirect hint that we may be dealing with a considerable time-depth indeed comes from areal-typological observations.

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

93

It is well known that Turkic languages do not tolerate initial liquids, like /l-/ or /r-/, which is in fact a phenomenon very widespread in Eurasia, and it should not surprise us that Yeniseyan, to a degree, shows almost the same restriction; while it does tolerate initial /l-/, initial /r-/ is never found in native words. Another phonological restriction of Turkic is, however, not so widespread, neither in Eurasia, nor on a global scale, and this is the Turkic horror nasalium: neither /n-/ nor /m-/ may occur ir 'ly in native Turkic words, unless another nasal consonant further in the words present, which may assimilate an initial stop regressively. Actually, the same rule holds for Yeniseyan, initial nasals are restricted to loan words and words with a second nasal in inlaut position. This pattern seems too specific, and typologically too unusual to think of anything else here than of the direct consequence of early language contact, already at the level of the respective proto-languages. While it remains unclear, which language family was the donor and which the recipient of this highly marked feature, one further observation, from the domain of lexical etymology, may shed some light on early Yeniseyan-Turkic contacts, strengthening the as­ sumption that an early variety of Yeniseyan - or undifferentiated ProtoYeniseyan - was not only present in the region, when the earliest historically datable nomadic confederations entered documented history, but also that its speakers may at times have played a rather important political and cultural role in some of them. As argued in detail in Georg (2001), one of the most widespread and cultur­ ally significant cultural terms of early Turkic (and Mongolian), which for internal reasons has always been regarded as a foreign loan in Turkic (and a Turkic loan in Mongolian) finds a satisfying etymological explanation in Yeniseyan. The Turko-Mongolic word for the "sky", the "sky-god", or "God", /tängri/, attested from the earliest Turkic sources (8th century CE) onward, is doubtlessly a loan from Proto-Yeniseyan *ti gir "high"6. The earliest attestation of this term in any source, however, antedates the era of documented Turkic considerably. The word occurs in Chinese sources of Han times, and is there explicitly labelled as belonging to the language of the Xiongnu, a nomadic confederation of as yet undetermined, but doubtlessly heterogeneous, ethnic elements, which threatened Northern China from the 4th . to the 2nd CE. The language of the Xiongnu7 (if there ever was a

6

The vowel of the second syllable is somewhat unclear, the difference in form can be accounted for in terms of well-documented processes internal to Turkic, the shift of meaning "high" > "sky" has many parallels in Inner and Northern Asian languages. 7 The Xiongnu of Chinese sources are often equated with the Huns of European history, but cf. Doerfer 1973 for the penumbra of difficulties connected with such a rash identification.

94

STEFAN GEORG

single, or a single dominant Xiongnu language) remained unwritten, but the possibility that some of the isolated words and phrases8 of this mysterious language might be Yeniseyan in origin has been discussed since the 1950s. Cf. Ligeti 1950, Pulleyblank 1962. These, now together with Itängri/, provide a strong indicator of Yeniseyan presence - and importance - among the ethnic groups which formed one of the most powerful nomadic states of early Inner Asia. After that the Yeniseyans and their language(s) disappear from documented history until the advent of Russian fur-traders (and later settlers), who first appear east of the Ural mountains with Yermak's famous quest in 1581. The Yenisey river was reached by Russians in 1610 (the important fortresses and later cities of Yeniseysk and Krasnoyarsk were founded in 1619 and 1628 respectively, right within the territory of Yeniseyan-speaking tribes). At this time, the name of "Yenisey Ostyaks" was coined for the Kets, which lingered on in the literature until the 20th century. Russian colonists transferred the name of the Ostyaks of the Ob' 9 to peoples further eastward, which happened to resemble West-Siberian Ob'-Ugrians in terms of outward appearance, ways of life etc. Language did not play a role in that10. Few accounts from these earliest days of Russian-Yeniseyan contact are known which could clarify the ethnolinguistic situation, but Yasak ("tax") lists for the various regions which came under Russian control are available from the middle of the 17th century. Using these sources, . . Dolgich (Dolgich 1960) could ascertain that speakers of Yeniseyan languages were living in an area much smaller than that defined by the extension of Yeniseyan-type toponyms. Though in the South East, the confines of the Irkutskaja Oblast ' were

8

The only "text" in "Xiongnu" is a two-line poem found in a Chinese chronicle from the 1st c. CE, which now is read and analyzed as Yeniseyan by A. Vovin 2000. This identification is not without problems, but likely to be basically on the right track. 9 Nowadays known as the Khanty nationality, living, together with the Mansi (formerly known as Vogul) in the Chantimansijskij Nacional'nyj Okrug (cap. Chantimansijsk); both Khanty and Mansi speak Ob'-Ugrian languages, which form, together with Hungarian, an independent branch of the Finno-Ugric language family. 10 Another ethnic group which received the name of "Ostyaks" this way were the OstyakSamoyeds, now known as Sel'kups. In the 1930s, the official Soviet terminology for the various ethnic groups of the Union was changed, introducing self-designations in place of traditional names, some of which were felt as derogatory terms. Thus, e.g., Zyryenes became Komi, Lamuts became Ewens, Lapps became Saami and so on. As a curiosity, it should be mentioned that, of all groups which went by some form of the "Ostyak" name in the 19th century, the Kets were the only ones who embraced it themselves. Thus, alongside the official self-designation ke't (often also Keto, which is originally the vocative form), Ket speakers even today refer to their language as ostigan '.

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

95

still reached, and in the South the city of Krasnoyarsk, the western boundaries of the language family had been withdrawn largely to the watershed which divides the tributary systems of the Ob' and the Yenisey. Only in the Far North, where Yeniseyan toponyms are rare and recent, the language family was still winning ground. While the northern boundary of Ket roughly coincided with the Yeloguy river and the settlement of Verkhneimbatsk in the 17th century (where the Ket dialect spoken today is called Southern Ket), Ket speakers advanced as far as the Kureyka river (a right tributary of the Yenisey, north of Turukhansk), some 300 km north of Verkhneimbatsk and well beyond the polar circle, during the following centuries. Apart from language-shift to the idioms of neighbouring Samoyed, Turkic, and Tungus groups, a further factor which led to the reduction of speakers of Yeniseyan and other native Siberian languages during the 17th century - a close analogy to the situation in post-columbian America - was the spread of im­ ported diseases. Epidemies of smallpox are recorded for the Yeniseyan terri­ tory in 1630-1631, and the 1660s. No exact figures of victims are known, but the impact on Kets, Sel'kups, Khanty, Mansi, and Ewenks is reported as dev­ astating. The first reports about the actual languages spoken by these people reached the scholarly world only considerably later. The names of D.G. Messerschmidt (1723), J.E. Fischer (1730-47), G.F. Müller and his co-traveller J.G. Gmelin (1733-43), who travelled through remote Siberia, often with more botanical than linguistic interests in their minds, but nevertheless keeping their ears open for Siberia's linguistic diversity, must be mentioned. From their notes and travel diaries, sometimes published during their lifetimes, but often only con­ siderably later unearthed in Russian archives and analyzed by Yeniseyanists of the 20th century, we have linguistic information on the six Yeniseyan languages mentioned above11. For Arm (on the left bank of the Yenisey, south of Yeniseysk), Assan (on the right bank of the river, south of the Angara) and Pumpokol (NW of Arin), all we possess is lexical lists. By the end of the 18th

11 Or five, if we stick to the traditional terminology, which viewed Yugh as a mere dialect of Ket (= Sym-Ket). The current view, that Ket and Yugh have to be regarded and described as seperate languages was only developed from the 1960s onwards. The language now generally referred to as "Ket" is the Imbatsk-dialect of the earlier literature.

96

STEFAN GEORG

century, all these three languages had completely disappeared12, leaving only Ket, Yugh and Kott as the sole surviving Yeniseyan idioms13. For Kott, we are more fortunate. The Finnish pioneer-linguist, and true founder of Yeniseyan linguistics, Matthias Alexander Castren, brought an exhaustive grammatical description of Ket, Yugh, and Kott home from his Siberian journey (1845-48). The Kott language was, already then, on the brink of extinction. Castren found only five persons, who still spoke this southern­ most Yeniseyan language, which was earlier found on the shores of the Kan' and Biryusa rivers. In his university lectures (Castren 1857:88) he writes: Diese fünf Personen waren übereingekommen ein kleines Dorf am Agul anzulegen, wo sie ihre Nationalität aufrecht erhalten wollten, theils aus Liebe zu derselben, theils auch aus der Ursache, weil Sibiriens Eingeborne der russischen Regierung geringere Abgaben als die Russen zahlen. An diese Colonisten haben sich später einige von den Kotten herstammende Familien angeschlossen, welche bereits ihre Muttersprache vergessen haben und Russen geworden sind. Indessen liegt es auch diesen gegenwär­ tig sehr am Herzen, sowohl sich selbst als ihren Kindern die kottische Sprache beizu­ bringen und es ist möglich, dass die kleine Colonie noch lange ihre Nationalität, welche bereits als erloschen angesehen wurde, beibehalten werde 14 .

Nothing is known about the further fate of this village, but by the end of the 19th century the Kott language had finally ceased to exist. No continuous texts in the Kott language have been recorded, but, with the glossary and the numerous morphological paradigms Castren managed to save,

12 Most of the surviving lexical material has been collected and edited by Dul'zon 1961 (approx. 550 lexical entries in total). 13 But the process of the gradual disappearance of Yeniseyan languages has started earlier. Of course it was well underway already before contact-time, as evidenced by the discrepancy between the territory of Yeniseyan toponymy and the ethnolinguistic picture drawn by 17th century sources. Additionally, several Yeniseyan tribes and their languages disappeared between the first Yasak-records and the advent of non-fur-trading, scientifically minded Europeans. Thus, we have indirect indications that the tribes of the Baikot, Yarin, and Yastin spoke Yeniseyan languages or dialects, too, albeit no data at all of these hypothetical languages have survived. One of the first dialects of Ket ever recorded, however superficially, that of the Eed-shesh river - the exact location of which is unknown - vanished shortly after Messerschmidt's journey, cf. Werner 1997a, 4. 14 "These five persons had agreed to found a small village on the Agul river, where they intended to keep their nationality alive, partly out of love for it, partly because the indigenous peoples of Siberia pay less taxes than Russians. These settlers were later joined by some families of Kott origin, who have already forgotten their native language and have become Russians. However, they, too, have the strong wish to learn the Kott language and to pass it on to their children, so that it seems possible that this small settlement might preserve their nationality, which had already been regarded as extinct".

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

97

Kott is considerably better known than the abovementioned Southern Yeniseyan languages. These data suffice to allow for the hypothesis that the main factor which led to the disappearance of Kott was language shift to some Southern Siberian variety of Turkic. For Kott, especially when compared to Ket and Yugh, shows strong traits of areal influence from Turkic. Not only do Turkic loanwords abound in Kott15, but several of its structural-typological characteristics suggest that the language had been already under prolonged and intensive Turkic areal pressure at the time of its recording. Thus, instead of the complicated vowel system of Ket with three central vowels, differentiated by height (/ /, /ә/, / / ) , Kott shows only one central vowel, namely /i/, just like most Turkic languages. More striking is the transformation of the verbal system. In both Ket and Kott, the predominant morphological technique is agglutination, but, whereas in Ket and Yugh affixes of person agreement can, as a rule, only be prefixes16, in Kott these affixes are exclusively suffixes. Especially in the singular, the typological parallels between Kott and, say, Khakas (South-Siberian Turkic), are quite striking, when compared to a (rather untypically simple) prefixing Ket paradigm (see table 2). Kott

Khakas

Ket

"to wash"

"to throw"

"to go"

urka:k-

atča-m

- γ tn'

2.P.Sg.

urka:k-u

atča-zi

ku-γ tn'

3.P.Sg.

urka:k-ø

atča-ø

1.P.P1.

urka:gan-to

atča-bis

aΛ-γ tn'

2.P.P1.

urka:gan-o

atča-zar

kΛ- otn'

3.P.P1.

urka:gan-ø

atča-lar

- otn'

l.P.Sg.

-γ tn' ( .), u-γ tn' (f.)

Table 2

15 These were studied chiefly by Timonina 1978, 1985. Cf. also Werner 1997b, a work which assembles all available knowledge on Kott. The recorded Turkic loans of Kott are often also found in extinct Southern Samoyed languages, which themselves have been absorbed by Turkic languages during the last two centuries, possibly by closely related Turkic-speaking groups (cf. Joki 1952, Georg 1999). 16 More precisely, they can only occur before the root, since an intricate system of preverbs and other affixes may leave the actual personal affixes somewhere in the middle of the affix-chain, rather than always at the leftmost position.

98

STEFAN GEORG

Few scholars, and even fewer linguists, visited the Yenisey region after Castren, and we have to wait well into the 20th century to learn more about the state and fate of the Yeniseyan languages. In the 1920s and 1930s H. Findeisen17 and K. Donner18 collected ethnographical data, as well as a few texts, but the first grammatical description of Ket after Castren was published only in 1934 by Karger (Karger 1934a). Karger was also responsible for the first at­ tempt to introduce a Latin-script writing system for the Ket language, based on the Central Ket dialect. At least one school primer of Ket got published (Karger 1934b), before the project of alphabetizing the Ket language was discontin­ ued19. From ca. 1950 onwards, Tomsk and its Pedagogical Institute (later Peda­ gogical University) became the centre for Yeniseyan studies (as well as for studies on most indigenous languages of the Tomsk and Krasnoyarsk oblasti, such as Khanty, Sel'kup, Chulym Turkic, Enets, Nenets and others). The ener­ getic personality of A. P. Dul'zon initiated large-scale linguistic field-work activities (as well as archaeological investigations and the thorough study of Siberian toponymy, s.a.) on the vanishing languages of the region, and put Yeniseyan studies on a firm methodological and factual footing. All aspects of Ket and Yugh grammar were studied extensively by Dul'zon and his pupils. The impressive scholarly output of the Tomsk school is epitomized by Dul'zon's own Ket grammar (Dul'zon 1968) and the encyclopedic series of monographs published by Heinrich Werner (G.K. Verner) in the 1990s20. Un­ der the supervision of the latter, a second attempt to create a written language for Ket was launched in the late 1980s. This time, the dialectal basis is the Southern Ket dialect (as spoken in Kellog, which is informally known as the Ket "capital") and a modified Cyrillic alphabet is used. So far, three school primers and some readers have been published. Instruction in Ket has begun in a few elementary schools, but the prospects of success may be regarded as shaky, since the Ket language now seems to be rather irreversably on the de­ cline, as a closer look at some statistical data will show.

17

E.g. Findeisen 1929. Donner 1933, and especially 1955; another valuable ethnographic description from these times is the rare booklet by Dolgich 1934. 19 Karger himself did not survive the Soviet purges of the late 1930s; he got executed under the charge of having propagated "Ket nationalism". 20 See the list of references; apart from Tomsk-based scholars, E.A. Krejnovič must be mentioned as one of the most prolific and influential Ketological scholars in the second half of the 20th century, cf. i.a. Krejnovič 1968. 18

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

99

For the 19 century, the development of the Ket nationality presents itself as shown in table 321.

"Yenisey Ostyaks" of Upper and Lower Imbatsk and the Podkamennaya Tunguska

1836/39

1859

1897

740

911

888

"Yenisey Ostyaks" of the Symsko-Kasov-Tribe

164

161

94

total

904

1,072

982

Table 3

The "Yenisey Ostyaks" of the Symsko-Kasov-Tribe were speakers of the Yugh language (Sym-Ket); later Soviet censuses did not differentiate these from Ket speakers. The decline of Yugh is obvious, and by the time of the 1980s the last speakers of this language had died in Vorogovo and Jarcevo. During the Soviet era, the census data as shown in table 4 became available. 1926

1959

1970

1979

1989

Kets (general)

1,428

1,017

1,182

1,122

1,084

number of Ket native speakers

?

?

?

language retention ratio

885

589

78,9 %

54,3 %

Table 4

On the whole, the number of individuals claiming Ket ethnicity for themselves seems to be remarkably stable, but the language retention rate is decreasing rapidly. This observation is considerably strengthened by the wealth of data we find in the work of V.P. Krivonogov (1998), who visited every single settle­ ment where Kets still live between 1991 and 1995 and managed to assemble a great amount of fine-tuned sociolinguistic data. According to Krivonogov, Ket was spoken in villages and settlements as shown in table 5 (from N to S)22.

21

These data are adapted from Patkanov 1912. The difference between the number of Kets given in the most recent census reports and the overall number of Kets found in the region by Krivonogov is, of course, accounted for by the fact that a certain number of Kets now live in Siberian cities like Krasnoyarsk, or in places as far afield as Moscow or St. Petersburg.

22

100

STEFAN GEORG

population

number of Kets

% of Kets

Kureyka

600

8

1.3

Maduyka

86

53

61.6

Svetlogorsk

2,000

16

0.8

Goroshikha

212

58

27.4

Turukhansk

8,400

23

0.3

Star. Turukhansk

330

3

0.9

Kostino

85

10

11.8

Baklanikha

84

23

27.4

Vereshchagino

216

25

11.6

Surgutikha

299

91

30.4

Kangatovo

58

21

36.2

Verkhneimbatsk

820

25

3.0

Kellog

405

247

61.0

Bakhta

283

44

15.5

Sumarokovo

93

16

17.2

Sulomay

243

154

63.4

Bor

4,500

15

0.3

Vorogovo

1,300

17

1.3

Zotino

800

5

0.6

Sym

135

10

7.4

Table 5

Regarding the internal dialectal division of Ket, all settlements south of (and including) Kangatovo are home to Southern Ket, in Baklanikha, Vereshchagino and Surgutikha Central Ket is spoken, and north of this is the territory of Northern Ket23. Only in three of these settlements Ket speakers form the majority, which in no case exceeds 2/3 of the population. Maduyka, with its very few inhabitants, does not really count, while the two villages with the greatest percentage of

23

The dialectal differences between these variants are quite sharp and may at times stand in the way of mutual comprehension. Northern Ket is the least studied Ket variety, most extant data pertain to Southern Ket.

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

101

Kets, Kellog and Sulomay, are situated in quite inaccessible regions on tribu­ taries of the Yenisey. It seems highly likely that this remoteness itself is a decisive factor for the ethnic composition of these settlements (Maduyka, too, is on a right tributary of the Yenisey); the villages right on the bank of the Yenisey, which is a major water-road between the densely populated south around Krasnoyarsk and the Far North with the industrial city of Noril'sk, always attracted more Russian and other non-local settlers than places, which are rarely served by regular means of transportation and offer little economic prospects. All in all, only 454 ethnic Kets live in settlements, where this nationality forms the majority. In addition to the decreasing language retention rate (48,3% in 1989), the future prospects of Ket as a living language looks even less promising when we take into account Krivonogov's data on language proficiency. Of all ethnic Kets in the Yenisey regions, the ratios (in %) as shown in table 6 are found regarding their ability to use Ket and other lan­ guages in all situations of daily life: language

fluently

with difficulties

with great difficulties

passively

not at all

Ket

21.9

7.6

12.5

23.1

34.9

Russian

95.8

2.8

1.2

0.1

0.1

Sel'kup

2.0

1.3

1.1

3.3

92.3

Ewenki

0.8

0.4

0.1

2.0

96.7

Table 6

The low figure of 21,9% full speakers (not more than 190 persons) is al­ ready alarmingly low, but, as might be expected, if we have a closer look at the percentage of speakers in various age groups, it becomes clear that Ket may be fairly alive among persons over the age of forty, while for individuals under thirty years of age the Ket language is certainly no longer a means of daily communication. Language abilities by age groups (Ket only) are shown in table 7. One word of caution might be added, lest the percentage of persons who speak Ket "with difficulties" is viewed with too much optimism. All too often, at least in the places which the present writer was able to visit in 1999, 2000, and 2001 (Baklanikha, Vereshchagino, and Kellog), this label has to be tacitly translated into "with considerable difficulties". It is very hard to find fluent

102

STEFAN GEORG

age group

full Ket speakers

with difficulties

not at all

70+

92.9

7.1

0

60-69

88.1

11.9

0

50-59

58.2

27.3

14.5

40-49

55.3

36.2

8.5

30-39

36.4

54.5

9.1

20-29

14.0

59.8

26.2

10-19

1.3

54.7

44

0-9

1.3

29.6

69.1

Table 7

speakers, and it is even harder to overhear a spontaneous Ket conversation between people under forty. Even some elderly persons, who classify them­ selves as native speakers and have not begun to use Russian before entering primary school (some of whom have served as language consultants in their youth and keep lively memories of Ketologists like A. P. Dul'zon working with them in the 1950s) may now at best be classified as semi-speakers, whose Ket speech is patched with long stretches of Russian. Krivonogov offers some further interesting statistics, which show that it is mostly the older generation which still uses Ket spontaneously, or which can expect being addressed in Ket these days. Asked which language they use most in their daily lives, Ket speakers gave the answers as shown in table 8 (in %). Ket

Russian

both

70+

50

28.6

21.4

60-69

45.2

38.1

16.7

50-59

5.5

70.9

23.6

40-49

6.4

78.7

14.9 14.4

30-39

2.3

83.3

20-29

1.1

86.6

12.3

10-19

0

97.3

2.7

0-9

0

97.8

2.2

Table 8

The most dramatic caesura is obviously that between the age cohorts of the 50-59 years old (born between 1930 and 1940) and older persons. While some

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

103

of the latter have been exposed to elementary schooling in Ket in their youth, the next cohort had to grow up in the turmoil of World War II, when fathers fought at the front and a great number of deported persons (mainly Volga Germans and Baits) were resettled in villages of the North (where they and their descendents still live) with whom Russian was the only possible means of communication. As for the question when and with whom Ket is spoken, and when other languages are used, Krivonogov obtained the figures as shown in table 9 (in % of those who speak Ket). language spoken

Ket

Russian (and other lg.)

both

with parents

30.3

23.1

23.1

with spouse

14.8

14.5

14.5

with siblings

10.4

12.5

12.5

with children

9

16.5

16.5

when working

3.2

19.9

19.9

Table 9

Ket, thus, keeps losing its social functions. It is mainly used by the elderly, or by (adult) children communicating with their parents. From Krivonogov's data we also learn that no single monolingual Ket speaker exists24 (and that a minuscule ratio of 2.8 and 1.2% admit to speak Russian only "with (great) difficulties"); the prospects of keeping Ket alive for more than the lifespan of the present middle generation thus seem to be very low. The most threatened of all Ket varieties still spoken is certainly the Central Ket dialect, especially its subvariety which is known in the literature as the dialect of Pakulikha. While it was chosen as the basis for the short-lived writ­ ten language of the 1930s, it is now spoken only by very few individuals, and used on a daily basis by almost none of them. Though this dialect is now spo­ ken in Baklanikha and Vereshchagino, on the banks of the Yenisey, it was formerly the main language of the settlement of Pakulikha, which was situated on the left tributary of the Yenisey of the same name, until it was officially dissolved by Soviet authorities in the early 1960s as one of the numerous Sibe­ rian villages "without perspective". The transfer of Pakulikha's population to their present homes is felt, by those who still remember, as the major caesura of their lives. Shortly after their resettlement, in the early 1970s, they were

The 0.1 %, who do not speak Russian have Sel'kup as a second language.

104

STEFAN GEORG

forced to abandon their traditional activity of reindeer breeding. After the end of the Soviet Union, most adults lost any opportunity to work and stayed un­ employed ever since. Pensions are payed erratically only. Fishing, hunting and gathering wild berries and mushrooms are thus the only means left to ensure survival. Needless to say, these circumstances do not encourage younger peo­ ple to stay in their villages, and most of them plan to move to one of the Sibe­ rian cities, or even further away. Among those who stay, the omnipresent abuse of alcohol takes its toll, too. Languages are, first and foremost, problem-solving devices. For most Ket speakers, situations where only the use of Ket will allow them to solve real-life communicative problems will be extremely rare in their daily lives. However intensively linguists and local teachers strive to save the Ket language from oblivion, the one indispensable prerequisite for maintaining a native language and passing it on to the next generation is the need to use it. When in each communicative situation with which a potential but restricted Ket speaker will be confronted, a different language will be available, namely Russian, a lan­ guage spoken by all members of the community, while only the elderly remain full speakers of Ket, little can be done to prevent the end of the use and the eventual death of Ket, and, with it, of the Yeniseyan language family. References Alekseenko E.A. 1982. et al. eds. Ketskij sbornik. Antropologija, etnografija, mifologija, lingvistika. Leningrad: Nauka. Anžiganova, O.P. et al. ed. 1992. Chakasskaja dialektologija. Abakan: s.n. Castrèn, M.A. 1857. Ethnologische Vorlesungen über die altaischen Völker nebst samojedischen Märchen und tatarischen Heldensagen. St. Peters­ burg: Akademie. —. 1858. Versuch einer jenissej-ostjakischen und kottischen Sprachlehre nebst Wörterverzeichnissen aus den genannten Sprachen. St. Petersburg: Akademie. Čispijakov, Ė.F. 1976. "Šorsko-ketskie parallell v leksike". Jazyki i Toponimija 3. 73-76. Čispijakova, F.G. 1992. "Istoričeskie plasty v leksike kondomskogo dialekta". Anžiganova et al. 1992. 144-154. Czaplicka, M.A. 1914. Aboriginal Siberia. A study in social anthropology. Oxford: Clarendon Press (repr. ibid. 1969). Doerfer, G. 1973. "Zur Sprache der Hunnen". Central Asiatic Journal 17. 1-50. Dolgich, B.O. 1934. Kety. Moskva/Irkutsk: Ogiz. —. 1960. Rodovoj i plemennoj sostav narodov Sibiri v XVII v. Moskva: Nauka.

THE GRADUAL DISAPPEARANCE OF A EURASIAN LANGUAGE FAMILY

105

Donner, . 1933. Ethnological Notes about the Yenisey-Ostyak (in the Turukhansk Region). Helsinki: Suomalais-Ugrilainen Seura. —. 1955. Ketica. Materialien aus dem Keuschen oder Jenissei-Ostjakischen. Helsinki: Suomalais-Ugrilainen Seura. Dul'zon, A.P. 1961. "Slovarnye materialy XVIII v. po ketskim narečijam". Učenye Zapiski Tomskogo Gosudarstvennogo P edagogičeskogo Insti­ tuta 19:2. 152-189. —. 1968. Ketskij jazyk. Tomsk: Tomskij gosudarstvennyj pedagogičeskij in­ stitut. Findeisen, H. 1929. Reisen und Forschungen in Nordsibirien: 1927-1928. Skizzen aus dem Lande der Jenissejostjaken. Berlin: s.n. Fortescue, M. 1998. Language relations across Bering Strait. Reappraising the archaeological and linguistic evidence. London/New York: Cassell. Georg, St. 1999. "Rev. of. Werner 1997b". Ural-Altaische Jahrbücher N.F. 16. 304-309. —. 2000. "Methodologische Bemerkungen zum Problem der äußeren ge­ netischen Beziehungen der jenissejischen Sprachen". Osipova 2000. 128-139. —. 2001. "Türkisch-mongolisch tengri 'Himmel, Gott' und seine Herkunft". Studia Etymologica Cracoviensia 6. Janhunen, J. 1991. "Ethnic death and survival in the Soviet North". Journal de la Société Finno-ougrienne 83. 111-122. Joki, A. 1952. Die Lehnwörter das Sajansamojedischen. Helsinki: SuomalaisUgrilainen Seura. Karger, N.K. 1934a. "Ketskij jazyk". Krejnovič 1934. 223-238. —. 1934b. Bukvar (na ketskom jazyke). Leningrad: Gosudarstvennoe učebnopedagogičeskoe izdatel'stvo. Krejnovič, E.A. 1934. Jazyki i pis'mennost' paleoaziatskich narodov. Lenin­ grad: Nauka (Jazyki i pis'mennost' narodov Severa, 3). —. 1968. Glagol ketskogo jazyka. Leningrad: Nauka. Krivonogov, V.P. 1998. Kety na poroge III tysjačeletija. Krasnojarsk: KPI. Ligeti, L. 1950. "Mots de civilisation de Haute Asie en transcription chinoise". Acta Orientalia Academiae Scientiarum Hungaricae 1. 141-188. Nikolaeva, I.A. 1988. Problema uralo-jukagirskich svjazej. Diss. Kand. fil. nauk, Moskva. Osipova, O.A. ed. 2000. Sravnitel'no-istoričeskoe i tipologičeskoe izučenie jazykov i kul'tur. Tomsk: TGPU. Patkanov, S. 1912. "Über die Zunahme der Urbevölkerung Sibiriens". Keleti Szemlé 9. 54-94. Pulleyblank, E. 1962. "The consonantal system of Old Chinese". Asia Maior New Series 9. 58-144; 204-265.

106

STEFAN GEORG

Seefloth, U. 2000. "Die Entstehung polypersonaler Paradigmen im UraloSibirischen". Zentralasiatische Studien 30. 163-191. Starostin, S.A. 1982. "Praenisejskaja rekonstrukcija i vnešnye svjazi enisejskich jazykov". Alekseenko et al. 1982. Timonina, L.G. 1978. "Tjurkskie zaimstvovanija v kottskom jazyke". Sovet­ skąj a Tjurkologija 1978:3. 7-13. —. 1985. Tjurkskie zaimstvovanija v enisejskich jazykach v sravniteVnoistoričeskom osveščenii. Avtoref. Kand. fil. nauk. Leningrad. Verner, G.K. 1990. Sravnitel'naja fonetika enisejskich jazykov. Taganrog: TPL Volodin, Aleksandr P., N.B. Vachtin & A.A. Kibrik eds. 1997. Jazyki Mira. Paleoaziatskie jazyki. Moskva: Indrik. Vovin, Aleksandr. 2000. "Did the Xiong-nu Speak a Yeniseyan Language?". Central Asiatic Journal 44:1. 87-104. Werner, H. 1995. Zur Typologie der Jenissej-Sprachen. Wiesbaden: Harrassowitz. —. 1997a. Die ketische Sprache. Wiesbaden: Harrassowitz. —. 1997b. Abriß der kottischen Grammatik. Wiesbaden: Harrassowitz. —. 1997c. Das Jugische (Sym-Ketische). Wiesbaden: Harrassowitz. —. 1998. Probleme der Wortbildung in den Jenissej-Sprachen. München: Lincom-Europa.

THE ENDANGERED URALIC LANGUAGES ROGIER BLOKLAND & CORNELIUS HASSELBLATT University of Groningen

1 General information on the Uralic language family The Uralic language family is - after Indo-European - with approximately 23 million speakers the second largest language family in Europe. One Uralic language - Finnish (5 million speakers) - is an official language of the Euro­ pean Union, two other - Estonian (1 million) and Hungarian (15 million) - are the national languages of two future members of the Union. The other Uralic languages are spoken in the Russian Federation, in the Scandinavian countries (the Saami languages are spoken in Norway, Sweden, Finland, and also in Russia) and in Latvia (Livonian) and are endangered to some degree. The three mentioned official state languages are not regarded as endangered but, instead, even classified by the U.S. Department of Education (published in 1985; Crystal 1997:344) as 'Critical languages'. 'Critical' means here that 'knowl­ edge of them would promote important scientific research or security interests of a national or economic kind.' (Crystal 1997:344). We can exclude these three from the actual question of survival, as they have long written traditions stemming from the (late) Middle Ages, flourishing literatures, fully developed terminology for almost all fields of technology and they serve the language of higher education (although there are fears that even they will eventually have to give way to English). A fourth Uralic language, Mordvin, had been included in this list of 'critical languages', which is an obvious mistake. The relatively high number of Mordvins - with ca. 1.15 million they even outnumbered the Estonians in the last Soviet census from 1989 - has misled the compilers of the list who thought that mere figures would say something about the importance of a language. This mistake concerning the Mordvins (cf. below on the situation of Mordvin) makes, however, the data of the whole list of 169 critical languages somewhat doubtful.

108

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

The genetic relationship between the Finno-Ugric languages was detected in the 18th century (Stipa 1990). Half a century later it was convincingly shown that the Samoyedic languages are related to this group, those two subgroups thus forming the Uralic language family. Genetic affiliations to other families have been suggested but never unequivocally proven; they must still be con­ sidered uncertain (Altaic, cf. Sinor 1988b; Yukaghir, cf. Collinder 1940, Angere 1956; Rédei 1999 comes to a negative conclusion; or even Nostratic: Bomhard & Kerns 1994). Some Finno-Ugric languages have died out in the last 1000 years: Merya ( ±1000 AD), Muroma ( ±1000 AD) and Meshchera ( ?1500) though the last two might have been dialects of Mordvin (Décsy 1965:146-149; Tkačenko 1985). The Samoyedic branch consists today of the three northern Samoyedic languages Nganasan, Enets, and Nenets and the remaining southern one, Selkup. Several Samoyedic languages have become extinct during the last centuries, e.g. the northern Samoyedic language Yurats ( 1800) and the southern Samoyedic languages Kamassian, whose last speaker died in 1989 (cf. Künnap 1999a; on her language see Künnap 1971:9-12); Mator (Helimski 1997) became extinct in the first half of the 19th century; of Karagass M. A. Castren last wrote down a few words in 1848 and of Koibal in 1847; of Taigi and Karagass it is not even sure whether they were independent languages or dialects of Mator. Sojot died out in the 18th century. The speakers of these languages have been Turkicized, Russianized and Mongolized, though the present-day Turkic Karagass and Sojot are of mixed Samoyedic-TurkicYenissey origin (Joki 1952:25-30). The Finno-Ugric languages are usually divided into the following subbranches: the Ugric languages (Hungarian, Khanty, and Mansi), the Permian languages (Udmurt, Komi, and Komi-Permyak), the Volgaic languages (Mordvin and Mari), the Finnic languages (Finnish, Estonian, Karelian, Vepsian, Ingrian, Votian, and Livonian) and the Saami languages. Due to the im­ possibility of differentiation between languages and dialects it is difficult to name a concrete figure for the number of living Uralic languages, but it would not be too far off the mark to speak of roughly 30 Uralic languages. 2 Geographical distribution, figures The spread of the Uralic languages is shown in the map on page 3. Note that the marked areas display the widest distribution and do not tell the whole truth about the actual settlements. A look at the map can therefore be misleading, since it suggests huge territories covered by Uralic-speaking populations. The actual situation and state of endangerment becomes clear from the statistical data. The last exact figures base on the last Soviet census from 1989 (Künnap

THE ENDANGERED URALIC LANGUAGES

109

110

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

1992; Tiškov 1994), but in some cases more recent data has been possible to obtain (cf. Kolga 1993, Salminen 1993):1 People/language

native speakers

place

Mordvin: 1,153,987 (Erza: ca. 2/3, Mokša: ca. 1/3)

773,827

Russia

Mari: 670,868 (Hill/Southwest 10 %, Meadow/East 90 %)

542,160

Russia

Komi: 344,519

242,515

Russia

Komi-Permyak: 152,060

106,531

Russia

Udmurt: 746,793

520,101

Russia

Khanty: 22,521

13,615

Siberia

Mansi: 8,474

3,140

Siberia

Karelian: 130,929

65,542

Finland, Russia

Vepsian: 12,501

4,800

Russia

Ingrian: 820

302

Russia

Votian: 62

31

Russia

Livonian: 226

15-35

Latvia

South Saami:

?500

Norway, Sweden

Ume Saami:

?20

Sweden

Pite Saami:

?20

Sweden

Lule Saami:

? 2,000

Norway, Sweden

North Saami:

30,000

Norway, Sweden

Inari Saami:

400

Finland

Skolt Saami:

300

Finland, Russia

Akkala Saami:

8

Russia

Kildin Saami:

800

Russia

Ter Saami:

6

Russia

Nganasan: 1,278

1,063

Siberia

Nenets: 34,665

26,730

Siberia

Enets: 209

95

Siberia

Selkup: 3,612

1,721

Siberia

Latest available figures from Tapani Salminen, http://www.helsinki.fi/~tasalmin/fu.html.

THE ENDANGERED URALIC LANGUAGES

111

Development of written literatures in the Uralic languages (after Taagepera 2000:83): First alphabet

First printed book

New Testament

Bible

Fiction

Estonian

Latin

1525

1686

1739

19th

ntury

Hungarian

Latin

1533

1541

1590

19th

tury

th

ntury

Finnish

Latin

1543

1548

1642

18

West Saami

Latin

1619

1755

1811

1900

Udmurt

Cyrillic

1770?

1997

Cyrillic

1788

1820s

Karelian

Cyrillic

1804

Komi

Cyrillic

1815

-

Mari

Cyrillic

1821

1821

Mokša

Cyrillic

1861

-

-

1904

Erza

Livonian

Latin

1863

1943

Komi-Permyak

Cyrillic

1865

-

Mansi

Cyrillic

1868

Khanty

Cyrillic

1868

Kola Saami

Cyrillic

1878

Selkup

Cyrillic

1879

Nenets

Cyrillic

1895

Vepsian

Latin

1931

Ingrian

Latin

1930s

1920s 1920s 19th

ntury

1907 1920s 1930s 1930s 1930s 1930s

1930s 1930s 1994

-

Not all specialists are of the opinion that literacy is necessarily a good thing and an important indicator of cultural development: "By introducing written skills into such a society (the Ob-Ugric community, RB-CH) we belittle their ancient ways of passing information, remove such tones of sacrality as may have been present in the oral traditions, and instead of preserving a traditional culture, we succeed only in promoting a new, bastardized one." (Nikolaeva 1995:127). 3 Degrees of endangerment There is no one definition of 'endangered'. The most general classification has three levels: safe, endangered and extinct. Krauss (1992:4) adds the term 'moribund', describing a stage between 'endangered' and 'extinct'. "Endan-

112

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

gered languages: are spoken by enough people to make survival a possibility, but only in favourable circumstances and with a growth in community support" (Crystal 2000:20-21), but moribund languages, though still spoken, are no longer being transmitted to children. Wurm (cited after Crystal 2000:21; cf. also Wurm, this volume) further divides the weaker languages into 5 levels: 1. potentially endangered languages: socially and economically disadvantaged, under heavy pressure from a larger language, and beginning to lose speak­ ers; 2. endangered languages: have few or no children learning the language, and the youngest good speakers are young adults; 3. seriously endangered languages: have the youngest speakers age 50 or older; 4. moribund languages: have only a handful of good speakers left, mostly very old; 5. extinct languages: have no speakers left. Note that the first three definitions do not actually say anything about the number of speakers. Many Caucasian languages, for example, have few speak­ ers but seem to be in no danger of dying out. For our purposes we can, though, on the basis of the number of native speakers form a rough division into three subgroups of different degree of endangerment, using Wurm's terminology. Firstly, there are five medium Uralic languages with a considerable number of native speakers: Mordvin, Mari, Udmurt, Komi (-Zyryan), and Komi-Permyak. These are potentially endangered according to the definition above. Secondly a number of smaller languages with tens of thousands, or at least several thousand native speakers: Karelian, Northern Saami, Nenets, Khanty, Mansi, and Vepsian. These range from potentially endangered to endangered. Finally we come to the minor languages spoken by about thousand or even less native speakers: all other Saami languages, Nganasan, Enets, Selkup, Ingrian, Votian, and Livonian. These range from endangered to moribund. If we scrutinize the specific situation of each language, we will see that there are certain problems with the border between the second and third group, as the mere number of speakers does not say enough about the possibility of language maintenance. Nevertheless, we will treat the three groups subse­ quently. The following questions have to be answered: what is the political status of the language? What is the social position of the language? Where - and by whom - is the language used? To which degree does the language function as language of (higher) education? How large is the text production? How old is the written tradition? The last question seems to be merely historical, but there is, of course, a certain correlation between future and history.

THE ENDANGERED URALIC LANGUAGES

113

In pre-revolutionary Russia Russian was the only official language (there were some exceptions, notably Swedish and later Finnish in Finland and Ger­ man in the Baltic provinces), but other languages had no rights. Some were used for missionary activity (the Orthodox Missionary Society in Kazan' was especially active in producing priests of Finno-Ugric origin and published some religious material in the Finno-Ugric languages). After the revolution the commitment to the equality of all peoples forced the Soviets to pay more at­ tention to the languages of the Soviet Union. Officially the Soviet Union did not have an official language and in principle all languages spoken in Soviet territory were equal. Thus it could easily be claimed that there was no linguis­ tic discrimination. That Russian had to fulfil the task of lingua franca was and is of course unavoidable. Written forms were then created for many languages, at first using mostly the Latin alphabet, to belie the accusation of Russification. When in the 1930s use of some of these languages for which a written language had been created was discontinued, the official reason was that they were not viable because of the small number of speakers. The Cyrillic alphabet was then adopted, suppos­ edly to facilitate the learning of Russian, but an untold reason was the fact that the USSR was then politically more isolated and thus consolidated its internal unity and delimited itself from outside forces (Comrie 1981:32-33). Later official Soviet policy was to encourage smaller peoples' consolidation into larger groups {sbliženie) and their eventual merging (slijanie) into a 'So­ viet' people, whose language would obviously be Russian. The main route towards this was through Uralic-Russian bilingualism, then Russian-Uralic bilingualism to Russian monolingualism: 'The mother tongues of the small ethnic groups and communities continue to function as a means of communi­ cation in everyday life. In cases of this type bilingualism should be viewed as a transitional stage to monolingualism, which will be reached by the small ethnic groups when their assimilation into the corresponding nations is complete.' (Isayev 1977:199-200). Large numbers of Uralic peoples were on their way to sbliženie: the percentage of those who consider themselves to be of a Uralic nationality but Russian-speaking varies from 15.3% for the Nganasan to 62.7% for the Mansi (Tiškov 1994:442). Whether it is possible for the Mansi to re­ main Mansi whilst speaking only Russian and living in towns is unlikely. According to the laws of the Russian Federation everyone is entitled to use his or her own language (Article 26 of the constitution of the Russian Federation): 1. Each person shall have the right to determine and indicate his or her nationality. No person may be forced to determine or indicate his or her nationality. 2. Each person shall have the right to use his or her native tongue and to choose freely the language of communication, upbringing, education, and artistic creation.

114

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

As the republics have their own constitutions and laws (Article 5, paragraph 2) the "...republics shall have the right to establish their own state languages. In the bodies of state authority, bodies of local self-government and the state institutions of the republics they shall be used alongside the state language of the Russian Federation" (Article 68, paragraph 2) the republics can declare which language(s) are to be official languages. The Karelian, Komi, Mari, Mordvin and Udmurt republics have done so. Russian remains the state lan­ guage (Article 68, paragraph 1). For the smaller ethnic groups the following articles apply: Article 69. The Russian Federation shall guarantee the rights of indigenous ethnic mi­ norities in accordance with universally acknowledged principles and rales of interna­ tional law and the international treaties of the Russian Federation. Article 72. 1. The joint authority of the Russian Federation and the members of the Russian Federation shall comprise: 1) the protection of the ancestral habitat and tradi­ tional way of life of small ethnic communities.

Under law the Uralic-speaking communities seem to be well protected, whether in actual terms they will be given those rights is not yet clear. At local level these laws in any case are often not implemented. The negative and belit­ tling attitude of many Russians towards any local language is still the cause for many speakers of Uralic languages to be ashamed of their own language. 4 Medium Uralic languages: Mordvin, Mari, Udmurt, Komi, and KomiPermyak These languages (cf. generally Lallukka 1990, 1997) all have a (semi-) official status since early Soviet times but have been described by travellers and scholars already in the preceding centuries. The first texts or descriptions date, respectively, from 1775 (Mari and Udmurt grammar), 1815 (Komi), 1838 (Mordvin grammar), 1865 (Komi-Permyak). Komi, however, did have its own ancient alphabet from the 15th century. This was mostly confined to religious inscriptions and has not been in use for over 300 years (see 'Komi'). After the October revolution the settlement areas of these peoples obtained the status of an ASSR (Autonomous Soviet Socialist Republic). This meant that certain aspects of their culture were protected and guaranteed, but gener­ ally social and political life was completely Russian (according to Stalin's vision of socialism). Basic education was held in the native language only for the first three or four years, in higher degrees only subjects as the own lan­ guage and literature were taught in the native language. Higher (university) education was - and is today - almost only in Russian. The Russian language

THE ENDANGERED URALIC LANGUAGES

115

was the key to all higher positions in the economic, political and social life of the respective republic. The political changes in Moscow in the middle of the 1980s under Michail Gorbachev led to a national revival movement among all Uralic peoples. After the final fall of the communism the Republics were renamed, national symbols like flags were officially used, language laws were issued etc. From the point of view of survival, there is no real danger of sudden extinc­ tion for the four medium ones: Mari, Komi, Udmurt and Mordvin - all of them (except for Udmurt) with two more or less different dialects with their own literary language and a living literature - will be spoken throughout the entire 21 st century. However, although they have literary languages and the number of speakers may be over half a million today, there are several reasons which make them differ from the 'large three' mentioned above or, say, Icelandic with its quarter of a million speakers: they are only partly official languages of their respective republics, their field of usage is quite restricted, and they func­ tion only to a less developed degree as languages of (higher) education. 4.1 Mordvin: Erza and Mokša The 1989 census showed 1,072,939 Mordvins (740,048 mother tongue speakers) in the Russian Federation, of whom only 313,420 live in their titular republic, i.e. only 29.12%, the lowest such figure for all Uralic peoples in the Russian Federation. The Mordvin Republic is situated some 600 km southeast of Moscow. Its vicinity to the capital and centre of expansion has been cited as a reason for the high assimilation of the Mordvins. There have been many concentration camps in the Mordvin Republic and at hearing the word 'Mordva' this is often what comes to mind first to many Russians. The Mordvins themselves do not use the name 'Mordvin' (though they may have done so before the split into Erza and Mokša; Zaicz 1998:185): there are two main ethnic groups, the Erza and Mokša. In the Mordvin Republic they make up 32.53% of the population, Erza and Mokša in about equal numbers, though overall the Erza make up some 65% of all Mordvins, the Mokša some 35%). There are smaller ethnic subgroups too; the most important of which is the Šokša; Erza who have undergone strong Mokša influences. There are smaller groups of Mordvins who have recently lost their language, but who are still aware of their Mordvin origin: the Karataj in the Tatar Republic (Paasonen 1903), who now speak Tatar, and the Terjukhan in the Nizhny Novgorod Oblast who have been Russified (Zsirai 1937:255-256). Erza and Mokša are quite close. Only scarcely do speakers of Erza not understand Mokša and vice-versa (Feoktistov 1966:175), though at present many Erza and Mokša do their utmost to emphasize the differences between

116

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

them: the Mordvin department at the University of Saransk recently split into an Erza and a Mokša department. There have never been many Erza-Mokša marriages; Erza-Russian or Mokša-Russian marriages are much more common. From 1788 onwards there were Erza- or Mokša-language publications; after 1917, when the Mordvins were given more possibilities to develop their lan­ guage, each writer wrote in his own dialect, and up to now the situation has not really changed. In the 1930s literary languages were created for both Erza and Mokša, supposedly in response to the peoples' wishes, but the divide et impera principle must have played a role. There was an attempt to create a panMordvin language based on a Mokša dialect (Gheno 1995:58). Despite lin­ guists (Zaicz 1995b; Keresztes 1995; Gheno 1995) who state that the formation of one unified Mordvin language might be possible, and is indeed most desir­ able to counteract their assimilation and to foster a feeling of greater unity, the Mordvin themselves stick to their native dialect with its own, if short, written traditions. Indeed, many activists say there is no possibility of working to­ gether with members of the other ethnic group. By 1988 there were no schools in Saransk, the capital of the Mordvin Re­ public, that taught either Erza or Mokša, but in 1990 these were started up again, as well as language courses for Russians (very few Russians learn a local language in the Russian Federation; according to Tiškov (1994:442) only 0.6%o of all Russians spoke another language of the USSR; for a definition of 'language of the USSR' see Comrie 1981:8). Officially Russian, Erza and Mokša are all official languages in the Repub­ lic, but their actual use is limited. No one actually tries to speak Erza to the doctor or Mokša at the town hall. Outside the Mordvin Republic, where thus the majority of Mordvins live, there are only 178 schools were either Erza or Mokša is taught, and availability of books is very low (in the Penza Oblast, where some 43,000 Mordvins live, there is not even a single book in a Mord­ vin language in local libraries; Salo 1991:70). 4.2 Mari The Mari (self-designation: mari) live in two main groups, one in the Mari Republic some 700 km northeast of Moscow, the other group further east in Baškortostan and the Perm Oblast. Only some 50.38% of the Mari live in their titular republic. In 1989 there were 643,698 Mari, of whom 81.9% spoke Mari as their mother tongue. 69.5% speak Russian. Many eastern Mari have some knowl­ edge of Tatar or Baškir. There are two main dialects, Hill Mari and Meadow Mari; in addition there are the rather different Eastern and Northwestern dialects spoken outside the

THE ENDANGERED URALIC LANGUAGES

117

borders of Mari El. Russian census figures do not differentiate the two; maybe some 10% of all Mari speak Hill Mari, the majority of the rest Meadow Mari. There are two written languages. In the 8th century AD the Mari were probably under Khazar domination, since about the same time they have lived in close contact with the Turkic Chuvash, and their languages have influenced each other strongly. In the 13 th century they came under Tatar domination, which lasted until 1552, when the Russians finally managed to conquer Kazan'. The Mari had gained a reputation amongst the Russians as fearsome warriors, leading to a saying in Russian: S odnoj storony čeremis, s dragoj beregis' 'From one side the Cheremis, from the other, beware' (Lallukka 1990:293-294). The Mari wars lasted till the 16th century. After their subjugation and exploitation by the Russians a new saying entered the Russian language: sobirat' kokury (from Mari kok 'two' and ur 'squirrel; kopeck') 'to cheat the Mari and Chuvash, to swindle money out of (Dal' 1881:135a). In the 18th century considerable numbers of Mari wandered east, partly volun­ tary, partly in response to Russian oppression. In 1821 the first Mari-language publication appeared; a translation of the New Testament in Hill Mari (in 1775 a Mari grammar in Russian had appeared). As with the other Uralic peoples after the Russian revolution a familiar pattern is to be seen: 'a blossoming of national culture and education in the 1920s, the murder of its proponents by 1940, a degree of relaxation in the 1950s, and a relentless slow smothering of Mari-language schooling thereafter' (Taagepera 1999:220). In the 1930s some 2000 Mari intellectuals were mur­ dered, and literary works and dictionaries in Mari were burned. Nowadays there is a renaissance in Mari El, probably the most active Uralic republic in Russia in terms of organizing 'Uralic' events. There are Mari peri­ odicals, some that appear five times a week, though most are in Meadow Mari; there are simply too few Hill Mari to support more than an irregularly appear­ ing newspaper and a literary quarterly (a Hill Mari in Helsinki publishes a literary quarterly Cikmä which often contains Hill Mari translations of Finnish and Estonian literature). Print runs of the Mari newspapers in Mari El were optimistically high at first, but have dropped due to the recession. In their titular republic the Mari even have some political power, as they make up some 43% of the population. Meadow Mari, Hill Mari and Russian are all official languages, and from 1991 to 1996 the president of the republic was a Hill Mari. Mari activity has been so successful that local Russians felt threatened: they demanded the abrogation of an Estonian-Mari cultural coop­ eration agreement, evidently fearing that an independent Uralic republic would give the Mari wrong ideas (Taagepera 1999:244).

118

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

In some schools all subjects can be taught in Mari from grades 1-4. From grades 5-8 there is some Mari history, and later Mari literature is added. But many parents still prefer to put their children in Russian schools, as their qual­ ity is slightly higher, and many parents are still convinced that it will give their children a better future. There are three institutes of higher education in Joshkar-Ola, the capital, where the Mari make up some 37% of all students (thus less than their percent­ age in the republic overall). Mari is the language of instruction only in the department of Mari philology. As in the Udmurt republic there have been suggestions of converting to the Latin alphabet, but so far nothing has happened apart from the publishing in 1992 of a couple of Latin-based articles in the magazine Kugarnia. 4.3 Udmurt The Udmurts (self-designation: udmurt) live in a relatively compact area some 900 km east of Moscow. In 1989 there were 714,833 Udmurts, of whom 496,522 lived in the Udmurt Republic (Udmurt Udmurt Elkun, Russian Udmurtskaja Respublika). The Udmurt language, closely related to Komi and Komi-Permyak, has been a literary language since the 19th century (though some Soviet sources claimed it was not written till after 1917, by that date about 200 Udmurtlanguage works had appeared). The first Udmurt publication was a poem ex­ tolling Catherine II, which appeared in 1767 (Csùcs 1990:19). In 1775 a Rus­ sian-language grammar of Udmurt appeared; it was reprinted in 1967 in the United States (Décsy 1967). The Udmurt language has four main dialects, though these are all mutually intelligible to a high degree. The literary language is based on the central dia­ lect, with both northern and southern elements. Udmurt-language religious publications were printed especially towards the end of the 19th century, in an attempt to widen and deepen Christian belief. Only in the 1930s did its intro­ duction as a school language take place. By 1958, though, only in the first four years of school was Udmurt taught. In 1970 all Udmurt-language books were destroyed, but due to administrative oversight only in the Udmurt Republic; in the surrounding republics like Baškortostan (then the Baškir ASSR) and Tatar­ stan (then the Tatar ASSR) there were Udmurt schools too, where these books survived. Udmurt scholars could later make reprints from these books (Taagepera 1999:275-276). From then on all instruction was in Russian; only in the second half of the 1980s was Udmurt-language instruction started up again. Now, according to the constitution of the Udmurt Republic, Udmurt and Russian are the official languages (in the Russian-language text 'Russian and

THE ENDANGERED URALIC LANGUAGES

119

Udmurt' are the official languages; in the Udmurt text 'Udmurt and Russian'; Konstitucija Udmurtskoj Respubliki 4:47). Alongside renewed interest in the language the need has been felt to create neologisms instead of blindly taking over Russian words (earlier attempts to create native vocabulary were branded as nationalistic). As in Mordvin (see above) even native words for dictionary (kïll' , kïlbugor, kïlsuzjet) were replaced with Russian /slovaŕ/, and written as such (where Udmurt has no /ŕ/). In 1995 an Udmurt-Russian/Russian-Udmurt pocket dictionary appeared where once more kälsuzjet is used. Booklets are issued with neologisms (cf. Komi). There have been suggestions of substituting the Cyrillic alphabet with the Latin alphabet (Krasilnikov 1995), but if this will ever happen is doubtful. 4.4 Komi and Komi-Permyak The Komi and the Komi-Permyak live in separate administrative territories (the Komi in the Komi Republic, the Komi-Permyak in the Komi-Permyak Autonomous Okrug [ ]), though there is no real reason to keep them sepa­ rate. This was done in the time-honoured Soviet divide et impera tradition, but since then no steps have been undertaken to change this. Komi and Udmurt linguistic areas border on each other, but there is a 30 km wide strip of the Perm Oblast (of which the is part) between them. From a linguistic view­ point Komi and Komi-Permyak are so close as to be dialects of each other; as with Moldovan (Comrie 1981:188) irrelevant differences were reflected in the orthography to distance the two from each other. The 1989 census gave a figure of 336,309 Komi and 147,269 KomiPermyaks. Of the Komi 71.0% spoke Komi as their mother tongue, for the Komi-Permyak this was 71.1%. The Komi Republic has more than a million inhabitants. Komi make up is unique amongst only some 23% of the population. The Komi-Permyak the Finno-Ugric regions in Russia in that there the original inhabitants are in the majority: they make up some 60% of the population. The ancestors of the Komi and Komi-Permyak arrived in their present-day region some 1000 years ago. When the Russian missionary Stepan of Perm baptized part of the Komi in the 14th century he supposedly created an alphabet for them, the so-called Komi alphabet. Why, though, would a missionary create a new alphabet and not use Cyrillic, as Cyril and Method did for the South Slavs? Taagepera (1999:295; 2000:321) claims that it was in use already before his arrival, but that the use of a 'pagan' alphabet would not have been acceptable; that is why he had to pretend he had invented it himself. This alphabet was regrettably only in use till the 18th century, though it was not used

120

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

exclusively by the Komi: in the 15th and 16th centuries Muscovian scribes used it as a secret language (Lytkin 1952:75-87). Continual immigration by Russians forced the Komi further north, to Ne­ nets and Mansi areas. Some of the Komi adopted reindeer breeding. Nenets disliked the Komi for their trading and dealing, where the Nenets were often the victims of disadvantageous transactions. The Russian Revolution brought big changes. During the first party con­ gress of the Komi Communist party in 1920 there were calls for fusion with Komi-Permyak areas (Russians from Perm were against this and imprisoned Komi activists from 1922 onwards). Moscow, however, was against a common Komi republic and in 1921 the Komi Autonomous Oblast was founded, leav­ ing the Komi-Permyaks out. They were given a National Okrug in the Perm Oblast. In 1936 the Komi was renamed the Komi Autonomous Soviet Socialist Republic. A year later Stalin's purges also caught up with the Komi and their cultural elite was decimated. Taking natural increase into account the KomiPermyaks lost 24% of their members (Lallukka 1995:82). There were even charges of wanting to make Komi a protectorate of Finland (Lallukka 1995:59). After World War II immigration increased once more, with both non-Komi workers moving to work the Vorkuta coalmines as well as uninten­ tional immigration to concentration camps. Later oil was discovered; that too had to be removed from the tundra with no regard to the environment. In 1815 the first Komi-language publication appeared, not with religious content, but a handbook on vaccination, but possibly some religious works had already been printed in the 18th century (nothing was of course published in the old Komi alphabet). In the 19th century the Komi produced a literary giant in Ivan Kuratov, a man whom the censor allowed to have only a few poems published, but who translated Burns, Schiller and Horatius in Komi. In 1920 the so-called Molodtsov alphabet (a system of heavily modified Cyrillic graphemes) was introduced, in 1929 the switch was made to the Latin alphabet (in 1933 for Komi-Permyak), only to be changed to the Molodtsov alphabet again in 1936; then in 1939 it was changed yet again, this time to the Russian Cyrillic alphabet, with two extra graphemes (ö and i). In the 1930s Komi needed new words for new concepts and many neologisms were in­ vented (as is being done once again). Teaching in Komi was diminished after World War II every few years or so, until by the 1980s there was no Komi ; in the Komi ASSR it was taught as an teaching in the Komi-Permyak optional subject in elementary schools. Nowadays Komi is being taught again in schools (there are more applicants than places) and once in a while you can hear Komi in the street.

THE ENDANGERED URALIC LANGUAGES

121

5 Smaller Uralic languages: Karelian, Saami, Nenets, Khanty, Mansi, Vepsian 5.1 Karelian Karelians (self-designation: karjalaiset ~ karjalažet; liüdi ~ liügi, livviköit; lüüdiköit) live mostly in the Karelian Republic (KR) in the northwest of the Russian Federation. The census of 1989 shows 124,921 Karelians in the RF, of whom only 48.6% speak Karelian. The Karelians are divided into four main groups, three of whom, the North-Karelians, the Olonets and the Ludes live in the KR; the fourth group, the Tver Karelians, live in the Tver Oblast. There are smaller groups of Karelians in the Leningrad and Novgorod Oblasts too. Dialect divisions do not correspond to the above geographical divisions. The Karelian dialects are divided into four groups: Northern Karelian (in the north of the KR); Southern Karelian (in the south of the KR and in Tver); Olonets in the southwest of the KR and in Finland, some 30,000 speakers in the KR and Lude in central and southern parts of the KR, with 7,000 ~ 10,000 speakers. Lude, close to Vepsian, is sometimes also deemed to be an independ­ ent member of the Finnic languages. Karelian itself, though, is sometimes considered to be a dialect of Finnish (e.g. Décsy 1965:41-45). There is in fact a dialect continuum from Finnish to Vepsian. In the KR Karelians make up only some 9.98% of the population, compared to 42.3%) in Russia in 1897. The Karelians of Tihvin in the Leningrad Oblast and Valdai in the Tver Oblast have been almost completely assimilated to the Russians already. A good knowledge of Russian was so common that those who did not speak it well were called 'wild Karelians' (Vaba 1993a:171). Karelians have for a long time been on the border between East and West; belonging to Sweden/Finland and then again to Russia: in the 12th century they came under Novgorod and became Orthodox. The border between Sweden and Russia was delineated in 1323, cutting across Karelia, but despite that the border wars continued. After the Russo-Swedish wars in the 17th century and the Treaty of Stolbovo in 1617 many Karelians fled the now Swedish area and forcible Lutherization to the north, others to the southeast (the descendants of this last group are the Tver Karelians). After the Russian revolution there were attempts to create an independent Karelia, but despite a positive resolution by the League of Nations it was never realized. Soviet Russia made Karelia into a Karelian ASSR, which became the Karelo-Finnish SSR in 1940, though de­ moted to ASSR again in 1956. Since 1991 it has been the Karelian Republic. The history of the written language of the Karelians has been an involved one. The oldest written text in any Finnic language was found scratched on a piece of birchbark from the 13th century, found in 1957. This seems to be in

122

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

Karelian, though it is in some respects close to Vepsian (see Laakso 1999 for the most recent discussion). In 1804 the first Karelian-language book was published. The Karelians, though, have never had one literary language: in the Karelian ASSR Finnish was the official language from 1921 to 1932 alongside Russian, from 1932 to 1937 there existed a compromise 'Karelo-Finnish', in 1937 Karelian was declared the third official language, the literary language based on the northern dialects and written in Cyrillic, then in 1938 a new liter­ ary language based on the Olonets dialect and written in Cyrillic. Now this language and Russian were the two official languages. This new written lan­ guage displayed very strong influence from Russian, not only in the vocabu­ lary: the case system was even reduced from 12 to 9 (Sarhimaa 1999:39). In 1939 Finnish was once more restored as an official language, despite the war with Finland. In 1940 there was a return to standard Finnish, and when Finland occupied parts of Karelia between 1941 and 1944 all schooling there was in Finnish. In the Tver areas a different Karelian was created in 1931, using the Latin alphabet, changing to Cyrillic in 1937 and abandoned altogether in fa­ vour of Russian in 1939. Since then Karelian had not been used in Tver till the 1990s, when publishing in Tver Karelian restarted. During the last few years various dialects of Karelian have been used in both the press and in kindergartens and schools, using the Latin alphabet, though Karelian is still not an official language of the KR. There is also a newspaper Oma mua ('Own land'), which has been published now for over ten years. In Tver Karelian there is a bilingual monthly newspaper Karielan sana ('Karelian word'), with a print run of only 500, and radio programmes. Russian and Finnish are the only official languages of the KR, though Rus­ sian is the actual working language. In September 1996 a draft law on the languages of Karelia was published. In it Karelian, as the language of the titular nationality, was given official status for the republic, but so far it is still not an official language. In Finland there are some 140,000 people who were born in Karelia in areas ceded to the Soviet Union after World War II; the percentage of Karelian speakers amongst them is not known, but cannot have been very high. Nearly all of these have been assimilated. There are also some villages in Finland near the Russian border where Karelian is spoken. 5.2 Saami The situation of the Saami languages is exceptional in several respects: most of the ten Saami languages are spoken outside Russia and are located in the Scandinavian countries with intensive contact to Western nations which led to quite enhanced minority rights; they partly have a literary tradition dating

THE ENDANGERED URALIC LANGUAGES

123

back to the early 17 century; today they partly have their own literature, newspapers, radio programmes and even a Saami college in Guovdageaidnu (Kautokeino, Norway) where the language of instruction is only Saami. An­ other point is that the languages obtain a prominent position within the FinnoUgric curricula of Finnish universities as the nearest cognate languages outside the Finnic group. The University of Helsinki, e.g., has launched Saami studies in 1993 (cf. Seurujärvi-Kari 2000), and there has been a native lecturer for many years. Also at other universities (Oslo, Oulu, Rovaniemi, Tromsø, Umeå, Uppsala) Saami is taught. Finally, also the shift from foreign researchers to native researchers of Saami (cf. Larsson 2001:86) may indicate that the chances of survival of different Saami languages are better than the relatively small figures suggest. However, the situation was much worse 30-40 years ago, when Saami speakers were suppressed in many respects, e.g. until 1959 it was the task of Norwegian teachers to control whether the domestic language of their pupils was Norwegian and not Saami (Magga 1993:47) and the right to buy land was restricted to those who spoke Norwegian. In the early 1960s attitudes changed and in Finland a Saami parliament was founded as early as in 1973. Since the 1970s Saami instruction in schools has increased. Since 1992 language laws in Norway and in Finland guarantee in regions with a certain Saami percentage the right of using Saami in communicating with the authorities and in court. Official documents have to be bilingual. In Finland this right includes all the three varieties of Saami (North, Inari and Skolt), while in Norway only North Saami benefits from this guarantee as this is the language spoken in the respective areas. Today we have six standardized Saami languages, viz. South, Lule, North, Inari, Skolt and Kildin. The last uses nowadays the Cyrillic alphabet but used to apply the Latin alphabet in the 1930s. All the other languages are written with the Latin alphabet (including some diacritics). The degree of endangerment depends largely on the political status. The Saami languages spoken in Norway and in Finland have a real chance to sur­ vive, above all North Saami with its 20,000 ~ 30,000 speakers (estimated figures: see table above and Sammallahti 1998:1-2). In Sweden, where the legal framework is not as positive as in the other two countries, some lan­ guages might already be extinct (Urne, Pite?) and the number of Lule Saami speakers is declining. Even worse is the situation in Russia, where Akkala and Ter face an almost certain death, Kildin being under heavy pressure.

124

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

5.3 Nenets The Nenets (self-designation: nenec '), by far the largest group of Samoyeds, live in a large area from Western Siberia to the Kola peninsula, mostly in the (capital: Narjan-Mar), the Jamal-Nenets (capital: Salekhard) Nenets and the Taymir (capital: Dudinka). The 1989 census records 31,490 Nenets. In the Nenets they make up 11.91% of the population, in the Jamalo-Nenets 4.23% and in the Taymir 4.38%). Nenets has two main dialects, Tundra (± 94% of all Nenets) and Forest Nenets (± 6%). The number of Nenets is steadily increasing, unusual among the Samoyeds, though the percentage of those who consider Nenets their mother tongue is steadily decreasing (80% in 1979, 77.1% in 1989). Most Nenets speak Russian (62.1% in 1989), some speak Komi or Khanty too. The northern Samoyeds split up from the southern Samoyeds some 2000 years ago; on their wanderings in north-eastern Europe they had contacts with the Ob-Ugrians and even with the Vepsians (Viikberg 1993c:235). In the 14th to 16th centuries they were under Tatar domination, and from 1668 under Rus­ sian domination, though not without resistance: the last Nenets attack on a Russian fort took place in 1746. Russian missionaries arrived in Nenets areas to convert them to Christianity; traders to cheaply buy up furs. Russia popu­ lated Nova Zembla with Nenets to dissuade the Norwegians from showing too close an interest in the area; they were also displaced from their original homelands to the Kola Peninsula. After the Russian revolution the Nenets also collectivized; this process lasted until nearly 1950. Reindeer breeders rose up in arms and were shot from airplanes. After 1950 many Nenets were forced to change to a sedentary life­ style. In 1957 a national maintenance system was introduced and children were sent to boarding schools, away from the tundra. The Nenets have undergone the same destruction of their land as the Khanty and Mansi (see below). The Nenets have for centuries used a pictorial sign system of tamga 's to denote ownership of reindeer or other things. The first written texts, of relig­ ious content, were published in the 1830s. A new literary language was created in 1931, written with the Latin alphabet (in 1937 this was changed to the Cy­ rillic alphabet). Most of the Nenets-language printed works were schoolbooks. At present Nenets is the language of instruction in the first three grades in some primary schools and is taught as a subject in some secondary schools. Much new Nenets-language material has been published in the last few years. The only newspaper in Nenets, Nyaryana Ngyrm ('The Red North'), is pub­ lished in Salekhard.

THE ENDANGERED URALIC LANGUAGES

125

5.4 Khanty The Khanty (self-designation: (west) căntə. (east) kăntəγ 'person') live mostly in the Khanty-Mansi Autonomous Oblast in Western Siberia, an area of some 523,000 km2, which belongs administratively to the Tyumen Oblast. There are also some Khanty in the Jamal-Nenets in the extreme northeast is Khanty-Mansijsk, of European Russia. The capital of the Khanty-Mansi but most Khanty live in villages along tributaries of the Ob. In 1989 there were 22,283 Khanty in the Soviet-Union, of whom 60.8% spoke Khanty as their mother tongue (Neroznak (1994:60), though, gives a figure of about 30%!). In 1989 they made up only 1.8% of the population of the (compared with 9.2% in 1959). The Khanty language is divided into three main dialect groups, Northern, Eastern and Southern Khanty, each of which is further divided into many dialects. Many are mutually unintelligible and indeed sometimes one speaks of the Khanty languages. The Khanty's feeling of ethnic unity is the main reason for speaking of one Khanty language. Southern Khanty died out in the 1950s. Due to its dialectal diversity the language has had 5 literary languages: East (Vach and Surgut dialects) and North (Middle-Ob, Kazym and Šuryškary dialects). The Khanty, though, use Russian in schools, learning Khanty only in some preparatory classes. Northern Khanty sometimes speak Nenets or Komi. According to the 1989 census 60.3% speak Khanty as their mother tongue (urban population 42.9%, rural population 68.3%), 38.5% speak Russian as their mother tongue (urban population 56.3%, rural population 31%). The ancestors of the Khanty moved from what is now eastern European Russia to Siberia some time before 1000 AD. In the 14th century Novgorod directed campaigns in the area. Due to those incursions and later Moscow's the Khanty moved further east. There they were forced to pay tribute to the Tatars. After the Russians defeated the Siberian Tatars in 1582 the animist Khanty were forced to accept Christianity. Russian traders exploited the Khanty with vodka and easy credit. The Khanty lost their land, income and died of diseases brought in by the invaders. At first Soviet power seemed to be a change for the better. In 1925 the Northern Committee was founded in Tobolsk to bring the Khanty, Mansi and Nenets on the road to progress. In 1931 and 1933 primers were published in the Obdorsk and Kazym dialect, using the Latin alphabet. Not all aspects of Soviet life were so positive. The Khanty were forced to move to collective farms, shamans were repressed, sacred groves destroyed and children sent to boarding schools, where they were taught in Russian only and lost their native tongue. The Khanty associate these boarding schools with the 'Kazym war' of

126

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

1931-1934: in 1931 a recreation centre was opened in Kazym and Khanty children were brought to the boarding school by force. After an outbreak of food poisoning at the school the children were not allowed to go home. The parents thought their children were being kept from them on purpose and brought them home by force. The next year they did not let their children go back. In 1933 a collective fishery was founded on a sacred island; the Khanty stopped its activities. In December of the same year a group of Soviet workers was murdered by the Khanty. In February 1934 the Khanty fought a punitive expedition sent in by the Soviets but with their old-fashioned guns they were no match for the soldiers equipped with machine-guns. Later practically all male Khanty in Kazym were arrested and never seen again. The punitive expe­ ditions were active till at least the end of 1934. After losing their men-folk and thus their breadwinners hundreds of women and children died of hunger (Leete 1996, 1997; Barkalaja 1999:72-73; Taagepera 2000:383). In the 1950s natural gas and oil was discovered in Western Siberia. Massive immigration followed; grazing lands, lakes and rivers in the fragile Arctic environment, essential to the Khanty's survival, were quickly polluted by leaking pipelines (more than 20,000 tonnes of petroleum seeps into the ground each year; Viikberg 1993a: 114). Khanty were not employed because they 'might break something' and 'would earn too much' (Viikberg 1993a: 115). Major causes of death among Khanty males are alcoholism and suicide. In 1991 the assembly of delegates of the Nordic, Siberian and Far Eastern indigenous peoples protested a decision to exploit the oil deposits at Tyanovsk in the Khanty-Mansi . In response 35 Khanty families (211 persons) were deported from their homes. In 1993 a Khanty activist, a deputy of the Jamalwith parliamentary immunity, was murdered by the Russian police Nenets (Taagepera 2000:26-28). To this date the officer in question has not been reprimanded, arrested or even questioned. In Khanty-Mansiysk, a newspaper Khanty Yasang is published in Khanty. Radio stations in Khanty-Mansiysk, Beloyarsk and Salekhard have pro­ grammes in Khanty. Dialects of Khanty (depending on the teacher and avail­ able material) are taught in some schools in the Khanty-Mansi from the first grade to the ninth grade (Jääsalmi-Krüger 1998:103). 5.5 Mansi The Mansi (8479 in 1989, of whom only 37.1% speak Mansi; cf. 59.2% in 1959 and 98% around 1900) live in the Khanty-Mansi Autonomous Oblast. , which is Their self-designation is mänsi ~ mäńś. In the Khanty-Mansi approximately the same size as France they make up only 0.51% of the popu-

THE ENDANGERED URALIC LANGUAGES

127

lation (1989). There seem to have been Mansi on the European side of the Urals up to the 1950s (Kàlmàn 1989:145). Together with Khanty Mansi forms the Ob-Ugric subgroup of the Ugric branch (of which Hungarian is the third member) of the Finno-Ugric lan­ guages. Mansi and Khanty probably split up around the 13th century. Mansi is now divided into three main dialect groups: Northern, Eastern and Southern. Western Mansi probably died in the 1950s, though exact figures are not avail­ able. The dialects are to some extent mutually unintelligible. Under the leadership of their chiefs the Mansi fought against the Tatars and later against the Russians, who eventually converted them to Christianity by force and exploited them. After the Russian revolution the Ostyak-Vogul was created (changed to Khanty-Mansi in 1940). After the kulaks, i.e. the more successful fisher­ men and reindeer herders had been liquidated the shamans as figures of authority became the object of the Russians' enmity. In the 1950s ~ 1960s oil and natural gas was discovered in West-Siberia. The exploitation of these resources caused and is causing untold damage to the Mansi traditional way of life: in the 1960s 6 million hectares of reindeer breeding ground was destroyed, 200,000 hectares of rivers and lakes were polluted. Oil pipelines prevent reindeer from moving freely about and leak oil into the ground. Mansi families have been collectivized into sovkhozes and their children sent to boarding schools, were after the first year only Russian is spoken. Alcoholism is rampant and the average life expectation is 4 0 ~ 4 5 . The first book in Mansi was a Gospel translation published in London by the British and Foreign Bible Society. In 1931 a new literary language was created based on the Sosva dialect of Northern Mansi, written in the Latin alphabet (changed to Cyrillic in 1937). Little was published, nothing whatso­ ever between 1957 and 1971. Well-known even outside the Russian Federation is Mansi writer Juvan Šestalov (born 1937), who started out writing in Mansi, translated his own work into Russian, then switched to writing in Russian only, and now trans­ lates his own work back into Mansi. Nowadays there is a slight revival: three times a week there is a 15 minute Mansi-language radio programme, the Mansi newspaper Luima serikos appears once a week with a print run of 240 and Mansi is taught at the local pedagogical institute. A popular movement Spasenie Jugry ('Jugra Rescue'), with both Mansi and Khanty members, has been set up to further the interests of the indigenous inhabitants.

128

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

5.6 Vepsian The orthodox Vepsians (self-designations: veps laine, bepslane, lüdinik, lüdilaine; now all Vepsians use the name veps in addition to their older names) live on the isthmus between lakes Ladoga, Onega and Beloje, some 100 km northeast of St. Petersburg. In 1897 there were 25,284 Vepsians, in 1939 some 32,000. This had dropped to 16,400 in 1959, by 1989 to 12,142. The Vepsian language belongs to the Finnic subgroup of the Finno-Ugric languages and has three mutually intelligible dialects, North, Central and South Vepsian, corresponding to the geographical distribution of the population. These three areas are split up by areas with Russian populations, and indeed many Vepsians till very recently did not know of the existence of the other groups. Roads between the three areas are very bad. The Russification of the Vepsians started in the 15th century; forcible Christianization and the vicinity of St. Petersburg were important contributing fac­ tors (the first imposed, the second not). The Russian revolution was to change this for the better. Two Vepsian national regions were created, Vepsian was made a school language in one of them (in the other Finnish was used, a lan­ guage the children did not understand), a Vepsian literary language based on the central dialect was created using the Latin alphabet. More than 30 books, mostly primers, were published in Vepsian. In 1937, however, all cultural activities were forbidden, Vepsian-language schools were shut, Vepsian books were destroyed and teachers deported or shot. The national regions were eventually disbanded and divided in seven different regions between the Kare­ lian ASSR, the Leningrad Oblast and the Vologda Oblast. As a result of this many Vepsians looked down on their own language and culture and stopped speaking Vepsian. Migration to the cities, especially St. Petersburg, increased significantly after World War II, along with the concur­ rent assimilation. In the Vologda Oblast in 1958-1959 people were invited to leave villages supposedly 'incapable of development' for new larger settle­ ments; when they refused to leave, all shops, schools etc. in the villages were simply shut. Many preferred to move to towns instead. Census figures, however, were often incorrect: in 1979 officially 8,094 Vepsians were registered; in 1983, however, ethnographers came to a figure of about 13,000 (of whom 5,600 in Karelia, 4,000 in the Leningrad Oblast and under 1,000 in the Vologda Oblast, the rest dispersed throughout the USSR) (Vaba 1993c:350). Census officials had been told to enter Vepsians as Rus­ sians (Kährik 1989:283; Taagepera 2000:154); many Vepsians themselves out of fear pretended to be Russians. There are thus more Vepsians than the census figures would have us believe. However, most Leningrad and Vologda Oblast

THE ENDANGERED URALIC LANGUAGES

129

speakers of Vepsian are over 40, the younger generation generally does not speak it anymore. In 1989 the Vepsian Cultural Society was founded, whose aims included raising the Vepsians' self-consciousness and esteem towards their own language and culture. This had drastically fallen in the last years (according to the 1989 census only 9 pre-school children in the Karelian ASSR said Vepsian to be their mother-tongue). There were plans for the formation of a Vepsian region, but local Russians objected strongly to this. Eventually a Vepsian region was created in Karelia, where now in 4 schools Vepsian is a compulsory subject, with some 200 pupils; in the Leningrad Oblast it is taught in 5 schools, but there participation in Vepsian lessons is voluntary. In the Vologda Oblast there is no teaching in Vepsian. Vepsian is thus taught only as a language, all other subjects are taught in Russian. There is a mixed Vepsian-Russian newspaper Kodima, which appears once a month, and there are Vepsian-language radio programmes on Petroskoi radio. Some literary work now also appears in Vepsian. Primers have also been writ­ ten using the Latin alphabet, based on the central dialect (as in the 1930s, though more notice has been taken of the other dialects than in the literary language of the 1930s) and a small Vepsian-Russian/Russian-Vepsian diction­ ary. Neologisms have been taken from other Finnic languages rather than from Russian. Parts of the Bible have also been translated into Vepsian. Contrary to what one might expect, older age groups, who speak Vepsian well, care less about the fate of the language, thinking that the youth does not need it. Some of the younger Vepsians, however, who are more fluent in Rus­ sian, are worried about its disappearance and think that teaching Vepsian in school should be obligatory. But for many the language is no longer an issue: through mixed marriages and a feeling of shame engendered by the Russians' attitude towards it they have given up their Finnic mother tongue. 6 Minor, vanishing Uralic languages: Ingrian, Votian, Livonian, Enets, Nganasan, Selkup. 6.1 Ingrian The Ingrians (self-designation inkeroine ~ ižora ~ ižoralaine 'Ingrian', also karjalain 'Karelian') live in the area between St. Petersburg and Estonia, mostly on the Kurkola and Soikkola peninsulas. They are the descendants of Karelians who in the 11 th century wandered south from the Karelian isthmus and the area around Lake Ladoga (this ex­ plains their ethnonym karjalain), from the 1270s onwards Ingrians belonged to Novgorods armed forces. Various foreign forces fought each other over Ingermanland: Swedes, Germans and Russians. It thus came about that the Baltic

130

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

Finns were had to fight each other, as Finns were forced to fight for the Swedes, the Estonians for the Germans and the Votians and Ingrians for the Russians. Till 1617 Ingermanland belonged to Russia; when Ingermanland was ceded to Sweden in that year many of the orthodox Ingrians moved to Russia. The empty areas were populated with Finns from southeastern Finland, the socalled savakot and äyrämöiset. These are nowadays called Ingrian Finns, not to be confused with the Ingrians (Finnish linguists also often consider the Ingrian language to be a dialect of Finnish. Ingrian Finns, however, are Lutheran and usually have Finnish names, whilst the orthodox Ingrians have Russian names, and there has been no intermarriage between the two groups, though their languages have influenced each other). The Ingrians went to Russian-language schools; only after the Russian revolution was an Ingrian literary language created. Altogether 24 books were published (Selickaja 1965) and Ingrian was taught in schools. In 1937, though, this all ended. The books were destroyed (a couple of these now very rare books are found in university libraries), teachers deported or shot and 'wealthy' farmers sent to Siberia and Central Asia. In 1834 there were 17,800 Ingrians in Ingermanland, by 1926 more than 26,000. So how to explain the official figure of just 1,062 Ingrians in 1959 Soviet census? Many had been killed in the war, and in 1942-1943 many had fled to Finland. The Soviet Union forced Finland to send them back; once on Soviet territory they were not allowed to return to Ingermanland but sent into exile to various other parts of the Soviet Union. After 1956, when they were allowed to return, there were only about 1,000 left. There was no question of Ingrian-language schooling or publishing anymore. By 1989 only 36.8% of the 820 Ingrians still considered Ingrian their mother tongue. The language is no longer being spoken to children and young Ingrians only speak Russian. The census figures, however, are not very reliable: after their tribulations many Ingrians pretended to be Russians; in 1979 there were officially 748 Ingrians, of whom only 315 lived in Ingermanland, but in some 15 villages on the Soikkola peninsula most of the inhabitants are actually Ingrians (Kurs 1998:90). Thus the exact number of Ingrians is unsure. 6.2 Votian The Votians (self-designation vadjalain ~ vad'd'alain 'Votian', maavätsi 'people of the land') live in the area between St. Petersburg and Estonia, his­ torical Ingermanland, now the Leningrad Oblast, in a multi-ethnic area popu­ lated by Finns, Ingrians, Russians and Estonians. The Votians are presumed to be the descendants of various groups of Baltic Finns; ethnically they seem to be closest to Northern Estonians. In the 11 th

THE ENDANGERED URALIC LANGUAGES

131

century they are first mentioned in Russian chronicles, when Novgorod started expanding into the Votians' territory. They fought as well with the Slavs against the Swedes and the Germans. In 1444-1448 German crusaders invaded Ingermanland and Votian prisoners of war were deported to Latvia, where their descendants and the Votian language died out in the middle of the 19th century (Winkler 1997). The Votians had converted to Russian Orthodoxy soon after their first con­ tacts with the Slavs; when Ingermanland was assigned to Lutheran Sweden in 1617 many Votians fled to Russia. Peasants from eastern Finland moved in great numbers to Ingermanland. In 1702-1703 Peter I conquered Ingermanland and founded St. Petersburg. To safeguard the city the surrounding areas were populated with Russians. The first reliable population figures date from 1848, when there were 5148 Votians. Kettunen (1930:4) estimated there to be about 1,000 in 1915. By the eve of World War II this had fallen to 500, partially due to deportations in 1938; after the war Ariste (1948:3) found only 100 Votians. The later Soviet censuses do not mention the Votians. According to the latest estimates the Votians now number approximately 35 people, of whom 5 ~ 10 are fluent speakers, most of them old women (personal communication from Eva Saar, Tartu, Estonia, april 4th, 2001). The youngest speaker alive today, Niina Ieniveno, was born in 1935. Votian has never been a written language; even in the 1930s, when alpha­ bets were created for small languages all over the Soviet Union the Votians were considered to be too small a people to need their own written language. All Votians speak Russian, many also speak Ingrian. The Votian language, as the language most closely related to Estonian, has been studied by Estonian linguists especially; this has raised the status of the language in the eyes of the Votians themselves, giving the language perhaps a few more years before it finally dies out. 6.3 Livonian The Livonians' (self-designations: Western Livonian: rāndalist 'coast dwellers', Eastern Livonian: kalami'ed 'fishermen') historical homeland was a wide area in northwestern Latvia on the eastern coast of the Livonian Bay and in northern Courland. Estimates of the numbers of Livonians in the 12th—13th century range from 15,000 to 28,000 (Boiko 1998:5). In 1201 bishop Albert founded the city of Riga on Livonian territory. In 1206 the Livonians were defeated by German crusaders and were forced to fight against the Estonians. Latvian tribes settled in the areas inhabited by Livonians; by 1835, when we get the first reliable population figures there were 2,074 Livonians (Vaba

132

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

1993b:216). Assimilation with Latvians had caused this decline; the reason why not all Livonians had been assimilated is probably due to the broad band of forest and marshland segregating the Livonian coast from the Latvian hin­ terland, which in turn explains the different occupations of the Livonians and the Latvians (fishing and agriculture respectively). Immigration of Estonians from Saaremaa, speaking a closely related language, was undoubtedly another factor contributing to their ethnic survival. The last Livonians of Livonia proper are generally said to have died in the middle of the 19th century (Décsy 1965:75; Boiko 1998:6), though Rudzïte & Karma (1975:354-356) claim that it might still have been spoken around the beginning of the 20th century. In 1861 a grammar and dictionary were written by the Estonian linguist F. J. Wiedemann, based on the materials of the Finnish linguist A. J. Sjögren, who had visited the Livonians in the 1850s. Portions of the Bible were translated in 1863. After World War I their numbers had slunk even more, but in the 1920s there was a Livonian national awakening. The Livonian language became an optional subject in schools in the area in 1923. There was an attempt to estab­ lish an ethnic Livonian district but the Latvian government did not grant its permission. A choir was founded, a Livonian Society created. The independent Finno-Ugric nations, Estonia, Finland and Hungary sponsored a Livonian Community Centre, which opened in 1939 in Ire. The interest shown by these nations, especially Finland and Estonia, was a driving force in the development of Livonian culture. Livonian-language books were printed in Estonia and Finland. Latvia's own contribution was small, and later hostile: from 1934 onwards all children with at least one Latvian parent had to be taught in Lat­ vian; a Livonian primer (Damberg 1935) was not allowed to be used as it was printed in Finland and not in Latvia and after a few years Estonian intellectuals involved in the Livonian movement were barred from entering Latvia (Loorits 1938:5). In 1936 the Shorter Catechism in Livonian appeared in Helsinki; the complete New Testament was printed in 1942, but by then it was too late and it never did reach the Livonians. After World War II and the forced emigration of the Livonians from the shores there was little left of all these efforts. The remaining living Livonians that returned were not allowed to fish in what was now a Soviet border zone; after 1955 the shore was completely closed off to civilians and the remaining Livonians (200 according to the 1959 census) were forced to move inland. All expressions of Livonian culture were proscribed by the Soviets after the war. In the 1970 census they were not counted anymore, as census officials had been told to note them down as Latvians (Vaba 1993b:218). In the 1970s, though, a few Livonians founded a singing ensemble ('Livlist'), singing

THE ENDANGERED URALIC LANGUAGES

133

Livonian songs. In the second half of the 1980s, after liberalization, a Livonian Cultural Society was founded and the teaching of Livonian was started up again. In 1990 there were about 35 speakers, of whom 15 spoke it fluently. Now there are about 5 fluent speakers in Latvia, one in Finland and possibly one or two in Sweden and the USA. There are also some 200 or so people who consider themselves Livonians but who do not speak the language. There is a renewed interest in Livonian and the Livonians nowadays; it is taught again in the University of Latvia in Riga (throughout post-war Soviet times the Esto­ nian linguists Paul Ariste and Eduard Vääri had done research on Livonian, collected material and made recordings) and in summer schools to children of Livonian origin. In 1998 a bilingual Livonian-Latvian collection of Livonian poetry and a book on the Livonians of Riga (Caune 1998) appeared; in 1999 a Livonian-Latvian-Livonian dictionary and a Latvian-Livonian-English phrasebook; in 2000 a Livonian textbook (Boiko 2000). Since 1992 a Livonian newspaper Lïvli has appeared irregularly (mostly in Latvian, however) and from 1994 onwards the magazine Õvā. A CD with popular adaptations of Livonian folksongs, sung by a young Livonian, has met with great success in Estonia. There are annual meetings of Livonians and Livonia-enthusiasts on the Livonian coast, but Latvian and Estonian are heard there more often than Livonian. The present Latvian government is more propitious to Livonian demands: in 1991 the Livonians were declared to be indigenous inhabitants of Latvia and in 1992 a special cultural-historical territory was created, the Līvõd Rānda ('Livonian Coast'). There have been problems, though, about its funding, with the Latvian government wanting to change the whole Livonian coast into a nature park. 6.4 Enets The Enets, with a population of 198 (in 1989; 300 in 1959, 400 in 1926) live in Siberia on the eastern bank of the Yenisey river. Less than half the population still speaks Enets; before many also spoke Nenets or Nganasan. They have never had their own literary language, and all schooling is in Rus­ sian. As all schools are in towns or villages the nomadic Enets are forced to send their children away from the tundra. After years in these schools many children will have forgotten their Enets and speak only Russian. Oil explora­ tion and nuclear testing are destroying the Enets' homeland and the average life expectancy is very low. The first written records on Enets stem from the 17th century (Künnap 1999b:4). As is often the case with lesser-known languages, the only dictionary

134

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

of Enets (Katzschmann & Pusztay 1978) is a specialist work for linguists and not for the people themselves. The latest work on Enets is Künnap 1999b. 6.5 Nganasan The animist Nganasan (1,262 in 1989, of whom 83.4% speak Nganasan), are the northernmost people of the Russian Federation. They live on the Taymir peninsula, an area of some 860,000 km2. Their high percentage of mother tongue speakers distinguishes them from the Enets. Their neighbours are Ne­ nets, Enets, Tunguz, Jakuts and Dolgans. Nganasan has two dialects, Avam and Vadejev; speakers of Vadejev speak Dolgan also. Nganasan has never been a written language and all schooling is in Russian. Most Nganasan are reindeer herders and hunters, who are thus forced to send their children to boarding schools from September to May. The children therefore do not hear their native language for months and end up speaking Russian better than Nganasan. The Nganasan are proud of their richly complex language and attach a good deal of importance to the skill and quality of shamanic narrative. According to Helimski this may be a reason that has contributed to the decline of the number of speakers: 'Let them rather not speak our language at all, rather than butcher it' (Helimski 1995:150). 6.6 Selkup The first Selkup-language books were published in 1897. These were relig­ ious works to promote literacy and religious education. They were nevertheless not used; when the compulsory schooling started up some 50 years later the language of instruction was Russian (Helimski 1983:7). Kuper & Pusztay (1993:6) mention that the older generation speaks the language, the middle generation understands it but does not speak it and the younger generation neither speaks nor understands it. Selkup written material from the 1930s was based on the Tas dialect (be­ cause it had more speakers and because it was better known); this cannot be understood by the Selkups of the Tomsk Oblast. The Selkup language from the 1930s was also based on the Latin alphabet with many additional diacritics. In the 1930s 7 books were published, mostly schoolbooks. Research on the southern dialects of Selkup is being carried out at the Tomsk Pedagogical Institute with a view to publishing learning material like primers and text­ books. So far the only Selkup dictionary is the Selkupisches Wörterverzeichnis by Istvan Erdélyi (1970), based on the Tas dialect and regrettably a linguistic work more easily available in Western second-hand bookshops than in Siberia.

THE ENDANGERED URALIC LANGUAGES

7

135

Conclusions As has been shown above most of the Uralic languages cannot yet be classi­ fied as seriously endangered or moribund according to the quoted definitions. But some of them are, indeed, moribund (e.g. Livonian, Votian, Akkala and Ter Saami), and a large number surely lives in the danger to become endan­ gered, i.e. they face the possibility of shifting to the group of irrecoverable languages. As the sociolinguistic situation for most of the languages in question has changed considerably during the past 10 years it is not easy to predict, which development the various languages will take. There are, however, some indi­ cations, which may help us answer the following questions (which consist actually of two sides of the same coin): Which factors enhance the decay of the languages? And, on the contrary, which factors stimulate the maintenance or revival of the languages? Rapid industrialization, connected with the spread of a 'ruling' people together with a lingua franca accelerates, of course, the decay of an indigenous language. This has been the case in the Americas (English, Spanish, Portu­ guese) as well as in the Russian empire (to name only the most prominent 'killer languages', cf. Nettle & Romaine 2000:5-6 (by the way: Nettle & Ro­ maine themselves are a sparkling example for the loss of diversity: their bibli­ ography consists of 221 titles which are all in English except for two French titles!). But in the case of Russia, we face today an economic decline in the entire country (with the exception of the metropoli of Moscow and St. Peters­ burg) which slows down the (cultural) unification of the country. It seems a paradox, but the absence of social and economic welfare seems to improve the chance for survival of a language/culture, as in the case of the Uralic languages in Russia this means also absence of Soviet (unification) pressure. The recent revival movements with different smaller peoples (e.g. Khanty, Mansi, Nenets) show that we should refrain from predicting an untimely death. On the other hand social welfare can also support the maintenance of a language as it is the case with Saami: the fact that (the larger varieties of) these language are spoken in rich Western countries where a certain respect towards minority cultures gained ground in the last third of the 20th century makes their chance to survive this century rather good. Russia, instead, is still not able to provide those circumstances for her colonies, therefore the future of all the Uralic languages in Russia is somewhat more uncertain than that of the Saami. Besides lacking social welfare also globalization might threaten the further development of the indigenous languages, even of the semiofficial 'larger minor' Uralic languages as Mari, Komi, Udmurt and Mordvin: still the chances

136

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

for social rise lie in learning the 'ruler' language and if, e.g., young Mari or Mordvins wish to improve their possibilities to gain a lucrative position in the society, they have, as a matter of fact, additionally to learn English. Thus the native language of their ancestors remains, if at all, a hobby for the weekend. But this is no solid basis for real language maintenance. That is why the situa­ tion of the medium Uralic languages may be more critical than the relatively high figures suggest. Obviously the following, first non-Soviet generation will be decisive for the future development of these languages. It is certainly not yet too late. References Abondolo, Daniel, ed. 1998. The Uralic languages. London: Routledge. Angere, Johannes. 1956. Die uralo-jukagirische Frage: ein Beitrag zum Problem der sprachlichen Urverwandtschaft. Stockholm: Almqvist & Wiksell. Ariste, Paul. 1948. Vadja keele grammatika. Tartu: RK Teaduslik Kirjandus. Barkalaja, Anzori. 1999. "Ortodoksisuus, taisteleva ateismi ja hantien kansanusko". Laitila & Saarinen 1999. 67-81. Bartens, Hans-Hermann. 2000. Die finnisch-ugrischen Minoritätsvölker in Europa. 2 nd ed. Hamburg: Finnisch-Ugrisches Seminar der Universität Hamburg (Mitteilungen der Societas Uralo-Altaica, 19). Bátori, István. 1980. Russen und Finnougrier: Kontakt der Völker und Kontakt der Sprachen. Wiesbaden: Harrassowitz (Veröffentlichungen der So­ cietas Uralo-Altaica, 13). Boiko, Kersti, ed. 1994. Lībieši: rakstu kräjums. Rīga: Zinätne. —. 1998. "Liivlased ja liivi keel". Õispuu & Joalaid 1998. 5-13. —. 2000. Līvõ kēl. Lībiešu valoda. Rīga: Līvu Savienïba (Lïvöd It). Bomhard, A. R. & J. Kerns. 1994. The Nostratic macrofamily: a study in distant linguistic relationship. Berlin: Mouton de Gruyter (Trends in linguistics. Studies and monographs, 74). Caune, Andris. 1998. Rigas líbiesu un viriu īpasumzīmes. Jumava: Latvijas Kultūras Fonds. Collinder, Björn. 1940. Jukagirisch und Uralisch. Uppsala: Almqvist & Wik­ sell (Uppsala Universitets Årsskrift 1940, 8). Comrie, Bernard. 1981. The languages of the Soviet Union. Cambridge: Cam­ bridge University Press. Crystal, David. 1997. The Cambridge encyclopedia of language. 2nd ed. Cam­ bridge: Cambridge University Press. —. 2000. Language death. Cambridge: Cambridge University Press. Csúcs, Sándor. 1990. Chrestomathia Votiacica. Budapest: Tankönyvkiadó.

THE ENDANGERED URALIC LANGUAGES

137

Dal', Vladimir. 1881. Tolkovyj slovar' živago velikoruskago jazyka II. S.Peterburg/Moskva. Damberg, Pētõr. 1935. Jemakīel lugdöbräntöz skūol ja kuod pierast. Helsinki: Suomalaisuuden Liitto. Décsy, Gyula. 1965. Einführung in die finnisch-ugrische Sprachwissenschaft. Wiesbaden: Harrassowitz. —. 1967. The first Votyak grammar. Bloomington, Ind.: Indiana University. Erdélyi, István. 1970. Selkupisches Wörterverzeichnis: Tas-Dialekt. The Hague: Mouton. Feoktistov, Aleksandr Pavlovič. 1966. "Mordovskie jazyki". Vinogradov 1966-68. 172-220. Gheno, Danilo. 1995. "'Mordwinisch' oder 'Mokschanisch und Erzanisch'?" Zaicz 1995a. 57-61. Haarmann, Harald. 1974. Die finnisch-ugrischen Sprachen. Soziologische und politische Aspekte ihrer Entwicklung. Hamburg: Buske (Fenno-Ugrica, 1). Hakamies, Pekka, ed. 1998. Ison karhun jälkeläiset: Perinne ja etninen identiteetti yhteiskunnallisessa murroksessa. Helsinki: Suomalaisen Kirjallisuuden Seura. Helimski, Eugen. 1983. The language of the first Selkup books. Szeged: Attila József University (Studia Uralo-Altaica, 22). —. 1995. "Nganasan as a literary language". Zaicz 1995a. 149-153. —. 1997. Die matorische Sprache: Wörterverzeichnis - Grundzüge der Gram­ matik - Sprachgeschichte. Szeged: JATE Finnugor tanszék (Studia Uralo-Altaica, 41). Huss, Leena. 1999. Reversing language shift in the Far North: linguistic revitalization in Northern Scandinavia and Finland. Uppsala: Uppsala University (Studia Uralica Upsaliensia, 31). Isayev, M. I. 1977. National languages in the USSR: problems and solutions. Moscow: Progress. Jääsalmi-Krüger, Paula. 1998. "Khanty language and lower school education: native, second or foreign language?" Kasten 1998. 101-111. Janhunen, Juha. 1991. "Ethnic death and survival in the Soviet North". Journal de la Société Finno-Ougrienne 83. 111-122. Joki, Aulis J. 1952. Die Lehnwörter des Sajansamojedischen. Helsinki: Suomalais-ugrilainen Seura (Mémoires de la Société Finno-Ougrienne, 103). Kährik, Aime. 1989. "Problemy buduščego vepsov i vepsskogo jazyka". Sovetskoe Finno-ugrovedenie 25. 281-287.

13 8

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

Kálmán, Béla. 1989. Chrestomathia Vogulica. 3 rd ed. Budapest: Tankönyvkiadó. Kappeier, Andreas. 1982. Russlands erste Nationalitäten. Köln: Böhlau (Bei­ träge zur Geschichte Osteuropas, 14). Kasten, Erich, ed. 1998. Bicultural education in the North: ways of preserving and enhancing indigenous peoples ' languages and traditional knowl­ edge. Münster: Waxman. Katzschmann, Michael & János Pusztay. 1978. Jenissej-Samojedisches (Enzisches) Wörterverzeichnis. Hamburg: Buske (Fenno-Ugrica, 5). Keresztes, László. 1995. "On the question of the Mordvinian literary lan­ guage". Zaicz 1995a. 47-55. Kettunen, Lauri. 1930. Vatjan kielen äännehistoria. Helsinki: Suomen Kirjallisuuden Seura. Kolga, Margus. 1993. Vene impeeriumi rahvaste punane raamat. Tallinn: Nyman & Nyman. Konstitucija Udmurtskoj Respubliki. Udmurt El'kunlen Konstitucijez. 1995. Iževsk, Udmurtija. Krasilnikov, A. 1995. "The transition to Latin script: a project of Udmurt cultural revitalization". Zaicz 1995a. 85-88. Krauss, Michael. 1992. "The world's languages in crisis". Language 68. 4-10. Künnap, Ago. 1971. System und Ursprung der kamassischen Flexionsuffixe. I. Helsinki: Suomalais-ugrilainen Seura (Mémoires de la Société FinnoOugrienne, 147). —. 1992. "Die uralischen Völker nach den Angaben der sowjetischen Volk­ szählungen 1959-1989". Linguistica Uralica 28. 50-54. —. 1999a. Kamass. München: Lincom Europa (Languages of the World. Materials, 185). —. 1999b. Enets. München: Lincom Europa (Languages of the World. Mate­ rials, 186). Kuper, Simon & János Pusztay. 1993. Sel'kupskij razgovornik (Narymskij dialekt). Savariae: Berzsenyi Dàniel Tanàrképző Főiskola (Specimina Sibirica, 7). Kurs, Ott. 1998. Etnilisi vähemusi ajas ja ruumis. 1. Tartu: Õpetatud Eesti Selts (Öpetatud Eesti Seltsi Kirjad, 6). Laakso, Johanna, ed. 1991. Uralilaiset kansat: tietoa suomen sukukielistä ja niiden puhujista. Porvoo: Söderstrom. —. 1999. "Vielä kerran itämerensuomen vanhimmista muistomerkeistä". Virittäjä 103. 531-555.

THE ENDANGERED URALIC LANGUAGES

139

Laitila, Teuvo & Tuija Saarinen, eds. 1999. U skonto ja identiteetti: suomalaisugrilaisten kokemuksia ja vaiheita Venäjällä ja Neuvostoliitossa. Hel­ sinki: Suomen Kirjallisuuden Seura. Lallukka, Seppo. 1990. The East Finnic minorities in the Soviet Union: an appraisal of the erosive trends. Helsinki: Suomen Tiedeakatemia (An­ nales Academias Scientiarum Fennicæ, 252). —. 1995. Komipermjakit - Perämaan kansa: syrjäytyminen, sulautuminen ja postkommunistinen murros. Helsinki: Venäjän ja Itä-Euroopan Instituutti. —. 1997: Vostočno-finskie narody Rossii: analiz ėtnodemografičeskich proc­ essov. Sankt-Peterburg: Evropejskij dom. Larsson, Lars-Gunnar. 2001. "Lappische Sprachforschung an der Schwelle zum 21. Jahrhundert". Finnisch-Ugrische Forschungen 56. 79-94 Leete, Art. 1996. "Märkmeid Kazõmi sõjast". Akadeemia 8:2. 392-405. —. 1997. "Uusi teateid Kazömi söjast". Akadeemia 9:2. 376-383. Loorits, Oskar. 1938. Liivi rahva mälestuseks. Tartu: Loodus. Lytkin, Vasilij Il'ič. 1952. Drevnepermskij jazyk. Moskva: Akademija Nauk SSR. Magga, Ole Henrik. 1993. "Saamelaiset kielet tänään". Salminen 1993. 46-51. Neroznak, V.P., ed. 1994. Krasnaja kniga jazykov narodov Rossii: enciklopedičeskij slovar '-spravočnik. Moskva: Akademia. Nettle, Daniel & Suzanne Romaine. 2000. Vanishing voices: the extinction of the world's languages. Oxford: Oxford University Press. Nikolaeva, Irina. 1995. "Obsko-ugorskie narody i literaturnyj jazyk: ličnoe mnenie". Zaicz 1995a.l23-127. Oispuu, Jaan & Marje Joalaid, eds. 1998. Kaheksa keelt, kaheksa rahvast. Tallinn: Tallinna Pedagoogikaülikool. Paasonen, Heikki. 1903. "Die sogenannten Karataj-mordwinen oder Karatajen". Journal de la Société Finno-Ougrienne 21. 1-51. Rédei, Károly. 1999. "Zu den uralisch-jukagirischen Sprachkontakten". Fin­ nisch-ugrischen Forschungen 55. 1-58. Rudzïte, Marta & Tõnu Karma. 1975. "Miliai oli Liivimaal veel kuulda liivi keelt?" Congressus Tertius Internationalis Fenno-ugristarum, Pars 1. 354-356. Salminen, Tapani, ed. 1993. Uralilaiset kielet tänään. Mikko Korhosen muistosymposiumin 14.10.1992 esitelmät. Kuopio: Snellman-Instituutti (Snellman-Instituutin julkaisuja A, 13). Salo, Merja. 1991. "Mordvalaiset". Laakso 1991. 156-184. Sammallahti, Pekka. 1998. The Saami languages: an introduction. Kàràšjohka: Dawi Girji.

140

ROGIER BLOKLAND & CORNELIUS HASSELBLATT

Sarhimaa, Anneli. 1999. Syntactic transfer, contact-induced change, and the evolution of bilingual mixed codes: focus on Karelian-Russian lan­ guage alternation. Helsinki: Finnish Literature Society (Studia Fennica Linguistica, 9). Selickaja, I. A. 1965. "Bibliografija literatury na ižorskom jazyke". Sovetskoe Finno-Ugrovedenie 1. 302-305. Seurujärvi-Kari, Irja, ed. 2000. Beaivvi mánát. Saamelaisten juuret ja nykyaika. Helsinki: Suomalaisen Kirjallisuuden Seura (Tietolipas, 164). Sinor, Denis, ed. 1988a. The Uralic languages: description, history, and for­ eign influences. Leiden: Brill (Handbuch der Orientalistik. 8. Abt., Handbook of Uralic studies, 1). —. 1988b, "The Problem of the Ural-Altaic relationship". Sinor 1988a. 706741. Stipa, Günter Johannes. 1990. Finnisch-ugrische Sprachforschung: von der Renaissance bis zum Neopositivismus. Helsinki: Suomalais-ugrilainen Seura (Mémoires de la Société Finno-Ougrienne, 206). Taagepera, Rein. 1999. The Finno-Ugric republics and the Russian State. London: Hurst. —. 2000. Soomeugri rahvad Venemaa Föderatsioonis. Tartu: Ilmamaa. Tiškov, V. A., ed. 1994. Narody Rossii: enciklopedija. Moskva: Bol'šaja Rossijskaja Enciklopedija. Tkačenko, Orest Borisovič. 1985. Merjanskij jazyk. Kiev: Naukova dumka. Vaba, Lembit. 1993a. "Karjalased". Kolga 1993.169-174. —. 1993b. "Liivlased". Kolga 1993. 216-219 —. 1993c. "Vepslased". Kolga 1993. 348-352. Viikberg, Jüri. 1993a. "Handid". Kolga 1993. 112-116. —. 1993b. "Isurid". Kolga 1993. 132-137. —. 1993c. "Neenetsid". Kolga 1993. 233-238. —. 1993d. "Vadjalased". Kolga 1993. 339-344. Vinogradov, Viktor Vladimirovič, ed. 1966-68. Jazyki narodov SSSR. III. Finno-ugorskie i samodijskie jazyki. Moskva: Nauka. Winkler, Eberhard. 1997. Krewinisch: zur Erschließung einer ausgestorbenen ostseefinnischen Sprachform. Wiesbaden: Harrassowitz (Veröffentli­ chungen der Societas Uralo-Altaica, 49). Zaicz, Gabor, ed. 1995a. Zur Frage der uralischen Schriftsprachen. Budapest: A Magyar Tudományos Akademia Nyelvtudományi Intézete (Linguis­ tica. Series A: Studia et Dissertationes, 17). —. 1995b. "Skol'ko jazykov nužno ėrze i mokše?" Zaicz 1995a. 41-46. —. 1998. "Mordva". Abondolo 1998. 184-218.

THE ENDANGERED URALIC LANGUAGES

141

Zsirai, M. 1937. Finnugor rokonsàgunk. Budapest: A Magyar Tudományos Akademia.

ENDANGERED TURKIC LANGUAGES THE CASE OF GAGAUZ

ASTRID MENZ Orient-Institut der Deutschen Morgenländischen Gesellschaft, Istanbul

1 The Turkic language group In this section I will provide some general information on the Turkic lan­ guage group as a whole before going on, in a second section, to give an over­ view of the endangered Turkic languages in various regions and look at the case of Gagauz more closely. 1.1 Genetic affiliation Turkic is a member of the Altaic language family. The question of whether the members of this "family" - Turkic, Mongolic, and Manchu-Tungusic (ac­ cording to some scholars even Japanese and Korean) - are affiliated in a ge­ netic sense is, however, not yet solved (Golden 1992:16-19). The Turkic language group itself is, with respect to modern languages, traditionally divided into six subgroups: Oghuz, Kipchak, Uighur, Chuvash, Khalaj, and Siberian Turkic (Johanson 1998b:82).1 An alternative classification regards Yakut, which is included under Siberian Turkic in the aforementioned grouping, as a separate group (Doerfer 1990:18-19). 1.2 Geographical situation The territory over which Turkic languages are spoken extends from the Balkans to northern China and from northern Siberia to central Iran. The core region of this territory (labeled "Turcia" by Johanson 1988) reaches from Asia Minor to East Turkistan, including northern Iran, Transcaucasia, and West Turkistan (Johanson 1988:51). It is generally believed that nomadic Turkicspeaking groups originated in the Altay region, from where they spread first to Mongolia and subsequently south- and westwards to East Turkistan and west1

For a more recent proposal regarding the classification of the Turkic languages see Schönig (1997ab-1998a).

144

ASTRID MENZ

ern Eurasia. Groups of Kipchak Turks entered eastern Europe, whereas larger groups of Oghuz Turks migrated into Anatolia.2 With the expansion of the Ottoman Empire parts of these Oghuz Turks moved further west and finally entered into southeastern Europe, first in 1352 when they crossed the Dardanelles and conquered Gallipoli. In 1362 the Otto­ mans conquered Edirne (Adrianople) and from there began their expansion into the Balkans. Ottoman settlement of Anatolian Turks in the Balkans led to a constant increase in the Turkish speaking population up until the national movements that lead to the emergence of nation states in the 19th and beginning of the 20th century. An enclave of about 3 million Turkish speakers is situated in northwestern Europe. Since the early 1960s large groups of migrants from Turkey, linguisti­ cally belonging to the Oghuz subgroup, began arriving in northwestern Europe, mostly as workers. The largest portion, consisting of over 2.5 million people, lives in Germany. In the Netherlands they numbered about 240,000 in 1992 (Backus 1996:44), with smaller groups living in Sweden, Norway, Austria, Belgium, France and Britain.

2

A detailed account of the history of the Turks is given in Golden (1992).

ENDANGERED TURKIC LANGUAGES

145

1.3 Historical development Johanson (1998b) gives the following periodization for the historical devel­ opment of the Turkic languages:3 1. The Old Turkic Period, including the language of the "runic" inscriptions4 (the earliest written sources in a Turkic language dating from the 8th century) and manuscripts, Old Uighur and Karakhanid, and dating from the 8th century until the Mongol conquest. 2. The Middle Turkic Period, including Khorezmian Turkic, Early Chagatay, Kipchak Turkic, and Oghuz Turkic, dating from the 13th to the 16th century. 3. A Pre-Modern Period comprising Middle and Late Chagatay, Ottoman Turkic, Middle Azerbaijanian and successors of Kip­ chak Turkic, lasting until the nation-building processes in the late 19th and early 20th century. 4. The Modern Period, which contains 24 written languages and a number of non-standardized spoken languages. 1.4 Structural features common to all Turkic languages Among the shared structural characteristics of all Turkic languages are the following:5 Phonetics and phonology. All Turkic languages exhibit intrasyllabic vowel harmony with respect to frontness and backness, which also affects most of the suffixes attached to stems. Additionally most languages show harmony with respect to roundness, i.e. words with a rounded vowel in the last syllable take suffix variants with rounded vowels, whereas words with unrounded vowels take variants with unrounded ones. In most Turkic languages initial consonant clusters are avoided and final ones are restricted. Suffix consonants assimilate with the preceding stem with regard to voice. Morphology. The word structure is agglutinative, and both the derivational and inflectional suffixes are clear-cut. The order of suffixes is fixed, and the same in all Turkic languages.6

3

Johanson (1998b:84) mentions however the difficulties regarding such a periodization. He states that: "It is impossible to find linguistically meaningful criteria for a periodization of the development of the Turkic group as a whole". 4 This script is called runic because of the resemblance of some of its characters to Germanic runes. Their phonetic value and the orthographical systems, however, are different. For further details and an overview of the various alphabets used for Turkic see Róna-Tas (1998). 5 I have only given some of the features here; for more detailed descriptions see Bazin (1959) and Johanson (1998a). 6 Only Chuvash deviates here in that it shows an inversed order of plural and possessive after nominals.

146

ASTRID MENZ

Syntax. In NPs modifiers precede their heads. Turkic languages show predomi­ nantly SOV order in neutral declarative sentences with great variation for pragmatic purposes. Embedded clauses are based on non-finite predicates and precede their main clause. Conjunctions are thus limited and are all of nonTurkic origin. Due to language contacts deviations from these structural characteristics have occurred in several Turkic languages. Some of the endangered languages, in particular, diverge in several respects from the common Turkic patterns. This is most apparent in the lexicon, but all other language levels can be af­ fected by contact-induced change. 2 Endangered Turkic languages Most of the endangered Turkic languages are situated at the periphery of "Turcia".7 They either have a very poor and recent literacy tradition or none at all. For the smaller Turkic languages on the territory of the former Soviet Union, for example, literary languages were created artificially only in the first half of the 20th century.8 All of the languages and dialects in question are under the heavy influence of a surrounding, socially dominant language (Russian, Chinese, Persian etc.) and sometimes also under the influence of a different Turkic language or dia­ lect with a higher prestige. Some of the languages listed below are actually diaspora varieties of a language spoken by a larger group in another area. Contacts between languages of the former Soviet Union and their respective diaspora groups in the neighboring countries were minimal until the political changes of the 1990s cleared the way for easier contacts in at least some re­ gions. The numbers of speakers for each of the languages given below are taken from Boeschoten (1998), except for the languages of the former Soviet Union. These figures are taken from the census of 1989.9 In all other countries infor­ mation on ethnic minority membership and/or language usage is either incom­ pletely elicited or not elicited at all. All numbers, however, have to be taken as approximate values. From my own observations speakers sometimes either 7

An overview of linguistic field research on the peripheral Turkic languages is presented in Johanson (2001). 8 For Siberian Turkic languages missionaries produced the first alphabets and grammars in the 19th century. In addition to their scholarly activities the missionaries also opened nativelanguage schools in Siberian villages. 9 The Center for Russian Studies at the Norwegian Institute for International Affairs has built a large database including general information on all minorities in the Russian Federation and the figures of the 1989 census. Under http://www.nupi.no/russland/database/start.htm this database is accessible.

ENDANGERED TURKIC LANGUAGES

147

over- or underestimate their competence. At times they simply want to be counted as a native speaker of the ethnic group they belong to for political reasons. The opposite can just as often be true, too. Nevskaja (1998:256) re­ ports that urban speakers of Shor tend to hide their native language skills. Among the endangered Turkic languages and dialects are the following (numbers of speakers are given in table 1 below): 1. In south Siberia, all Turkic varieties with less than 100,000 speakers, such as Shor, Altay-Turkic, Tofalar, and Khakas. 2. In the Republic of Mongolia, a small diaspora group of speakers of Tuvan/Tuvinian (about 6,000 members).10 3. In China, Salar11 and Yellow Uighur,12 Fu-yü Kyrgyz,13 as well as diaspora groups of Uzbek and Tuvan/Tuvinian. 4. In Afghanistan, diaspora groups of Kazakh, Karakalpak and Kyrgyz. 5. In Iran, Khalaj and Khorasan-Turkic. 6. In the Balkans, Gagauz (approx. 200,000) and Turkish dialects as well as the diaspora group of Crimean Tatars in Bulgaria. 7. A seriously endangered language is Karaim, with less than 100 speakers, spoken in Lithuania and Ukraine. In the core of the linguistic area we find endangered or seriously endangered Turkic languages: in the Caucasus Nogay, Kumyk, Karachay, and Balkar among others, and Crimean Tatar with its large diaspora group in Uzbekistan. In his European section of the Unesco Red Book Report on Endangered Lan­ guages (www.helsinki.fiMasalmin/europe_index.html), Salminen also cites Chuvash and Bashkir, which both have over a million speakers, as among the endangered languages. In what follows I will focus on the situation of the Turkic languages of the former Soviet Union, with a strong emphasis on the situation of Gagauz, a language on which I have undertaken research myself. Due to the political changes across this region, research on minority languages is nowadays rela­ tively easy, especially compared with the situation in countries where minori­ ties are regarded as a source for possible political turbulences or research is impossible because of the general political situation.

10 On the relationship between this diaspora variety and the language of the Republic of Tuva see Taube (1998). 11 Extensive fieldwork among the Salars has been conducted by Dwyer (see references). 12 This minority group is linguistically diverse see Roos et al. (1999). 13 See Schönig (1998b).

148

ASTRID MENZ ethnic popula­ tion

native speakers in

Altay-Turkic

69,400

87%

Baikar

78,300

98%

Crimean Tatar

271,700

93%

language

%

Crimean Tatar/Bulgaria

number of speak­ ers

6,000

Fu-Yü Kyrgyz

1,500

Gagauz/former USSR

197,000

90%

Karachay

150,300

99%

Karaim

260

25%

Karakalpak/Afghanistan

2,000

Kazakh/Afghanistan

2,000 78,500

Khakas

76%

Khalaj

28,000

Khorasan-Turkic

400,000

Kumyk

277,200

98%

Nogay

76,733

95%

Salar14

88,697

Shor

15,700

Tofalar

700

Kyrgyz/Afghanistan

500

60,000 58%

Tuvan/China

400 6,000

Tuvan/Mongolia Uzbek/China

44% 15,000

Yellow Uighur (Turkic)

14

4,000

Table 1 : Numbers of ethnic population and/or speakers

Since the last decade research has been conducted on the Turkic languages of Iran (Doerfer 1998; Johanson 2001; Kiral 2001), China (Johanson 2001; Dwyer 1998; Roos et al. 1999) and to a lesser extent of Mongolia (Johanson 2001; Taube 1998). Regarding regions such as Afghanistan and Iraq, however, no information is available on the current linguistic situation.

14 Figures for speakers of Salar and Yellow Uighur are taken from Dwyer (1998:49) and (2000a:431).

ENDANGERED TURKIC LANGUAGES

149

2.1 Endangered Turkic languages on the territory of the former Soviet Union Soviet censuses included figures on the various ethnic groups as well as on language usage. Thus, compared to the situation in other countries, we have a relatively detailed picture of the size of the minority groups as well as language preferences among their members. The endangered Turkic languages of the former Soviet Union share a number of common problems: first of all, literacy in the native language traditionally came late and was in most cases interrupted from time to time due to the vagaries of Soviet language policy. Secondly, socio-economic changes led to destruction of the traditional way of life. The necessity to participate in modern economic life made a good knowledge of Russian much more important than a profound knowledge in a mother tongue spoken in a restricted area by only a few people. Nevskaja (1998) gives an overview of the history of the Shor language in which she describes the various difficulties in language development. Her observations for Shor can easily be applied to the other small languages of this region. In the following section I will give some information on the situation of Gagauz as an example of an endangered language in one territory of the former Soviet Union. 2.1.1 Gagauz Gagauz is a Turkish dialect spoken nowadays mainly in the southern region of the Republic of Moldova. According to the 1989 census,15 about 200,000 Gagauz live within the territory of the former Soviet Union, most of them (about 92%) in the Republic of Moldova and the Odessa Oblast of Ukraine, to where they migrated from Bulgaria in the late 18th and the beginning of the 19th centuries. Smaller groups of Gagauz live in Bulgaria, Romania and Greece, but in these countries their number is uncertain. Their language belongs to the western group of Oghuz Turkic and is closely related to Turkish (Doerfer 1990:19). According to the 1989 census, 80% of the Gagauz living in the former Soviet Union are bilingual, with Russian as their second language. Romanian, the language of the major ethnic group in the Republic of Moldova is spoken by 4% of the Gagauz people. Their religion is Orthodox Christianity. The ethnogenesis of the Gagauz remains uncertain. For an overview on the various theories, see Özkan (1996:10ff.) and Chinn & Roper (1998:88-89). In 1957, Gagauz was established as a written language in the Moldovan S.S.R. and was taught at school from 1959 to 1962. Today Gagauz is part of the cur­ riculum in all schools in the Gagauz Yeri or Gagauzia, the autonomous region

15

The ethnographic data from the 1989 census in the former Soviet Union is cited after Fane (1993).

150

ASTRID MENZ

in southern Moldova.16 Since 1957 several books have been published in Gagauz, but it has never become a real written language in the sense that in eve­ ryday life nobody, besides some professional writers, actually uses it for any kind of writing. Gagauz phonology and morphology are very similar to those of (Anatolian) Turkish, but within the lexicon and syntax many differences can be observed due to the fact that Gagauz has been spoken for centuries in regions where the Slavonic languages Bulgarian and Russian are dominant (Doerfer 1959). 2.1.2.1 Factors enhancing the decay of Gagauz An important factor contributing to the decay of Gagauz was the split in the group of speakers that began with the migration from Bulgaria around the turn of the 19th century. This first split saw a smaller group of speakers left behind in Bulgaria where, in the course of the nineteenth century, the former high prestige of Turkish (Grannes 1996:2-3) began to drop significantly with the "National Awakening" of the Bulgars. The negative attitude towards Turkish in Bulgaria resulted in the "Turkishness" of the Gagauz in that country becoming a contested point. Thus a widespread opinion among the Gagauz themselves in Bulgaria concerning their ethnogenesis is that they are in fact "pure" Bulgars (temiz Bulgar) who merely changed their tongue in order to avoid pressure irom the Ottoman government. This negative attitude toward their own lan­ guage resulted in Gagauz not being transmitted from one generation to the next as a mother tongue any more. During fieldwork in Bulgaria in 1994 I observed that most speakers of Gagauz were over 60 and that their language of everyday communication was Bulgarian. The fact that the Gagauz are Orthodox Chris­ tians makes linguistically mixed marriages easy. In such cases, the language of the family automatically becomes Bulgarian. Contact between the two groups in Bulgaria and Moldova did not take place until very recently and is still restricted. The immigrant group of Gagauz in Moldova was further divided during the agrarian reforms of Stolypin (1908-1914); a group of Gagauz migrated to Kazakstan and Uzbekistan. In the last years of World War II young Gagauz men were sent to labor camps in Kazakstan.17 As their stay took longer than expected their families followed and settled there for several years. 16 On December 23 rd , 1994 the Parliament of the Republic of Moldova adopted a law on the special juridical status of this region. The International Foundation for Electoral Systems published an abridged English translation of it in 1995. 17 For Crimean Tatar, a Turkic language of the Caucasus region, the deportation of the whole population from their homeland to Central Asia during World War II was a decisive factor for serious decay.

ENDANGERED TURKIC LANGUAGES

151

In 1946 and 1947 in the great famine which afflicted Moldova, as other parts of the Soviet Union, about 350,000 people starved to death (Beckherrn 1990:252). This famine, which holds a prominent place in the collective mem­ ory of the Gagauz, took half of the population in some Gagauz villages. The Gagauz population of Moldova decreased from 98,000 in 1930 to 96,000 in 1959 (Fane 1993:137). Another factor contributing to the decay of the Gagauz language is the fact that it is spoken in rural areas of a traditionally agrarian society. Literacy in pre-Soviet times was not widespread among the peasants. The first attempt to write Gagauz was made in the 1930s by Mihail Ciahir, a Gagauz priest who translated religious texts and wrote a history of the Bessarabian Gagauz and a Gagauz-Romanian dictionary. He used the Latin alphabet and Romanian spelling. With the Soviet era came literacy, first in Russian; finally, in 1957, a Cyrillic-based alphabet was introduced for Gagauz and in the same year it was announced as one of the official languages of the Soviet Union. In the follow­ ing years a grammar, a Gagauz-Russian-Romanian dictionary, and a number of textbooks were produced. As mentioned above, it was taught in schools for some years during the Khrushchev era. After this short period and up till the breakup of the Soviet Union all formal instruction was in Russian. With the necessity to learn Russian well to enter higher education and pursue a profes­ sional career some parents shifted to Russian and did not transmit Gagauz to their children. Some poetry and folklore books appeared during the Soviet era, but apart from the writers of these books no one used Gagauz as a written language. During my fieldwork in 1995 some of my informants told me that for the first time in their life for my benefit they wrote something in Gagauz. 2.1.2.2 Factors encouraging the preservation of Gagauz As the Gagauz were a traditionally sedentary society, their way of life did not undergo major changes after collectivization. Most of them are bilingual with Russian as their second language, but among the older generation monolingualism is still common, especially among women. Since in the rural areas three generations will often live together, the language is maintained within the family. With interethnic marriages in these villages the non-Gagauz spouse is often able to communicate in Gagauz and sometimes nearly at the level of confidence of a native speaker. A shift to the socially dominant language is thus not as widespread as in Bulgaria. Due to the fact that ethnically speaking the Gagauz are living in relatively homogenous villages in southern Moldova, Gagauz is also the language of communication in the kolkhozes, and thus preserved in this domain too.

152

ASTRID MENZ

2.1.2.3 The recent revival of the Gagauz language In the 1980s Gagauz intellectuals first started activities to promote their language and culture. These activities increased through the political develop­ ments in the perestroïka period and after Moldova's independence gradually led to a movement for political autonomy.18 Autonomy within the Republic of Moldova was finally gained in 1994 and Gagauz was established as one the official languages of the autonomous region. Within the framework of the autonomy movement ethnic consciousness among the Gagauz increased and with it the intention to maintain language as the major cultural distinction between the Gagauz and both the Russians and the Romanians. Publishing activities increased, Gagauz language lessons were implemented in the school curriculum for all pupils in the autonomous region and even television and radio programs were produced. Parents revived Ga­ gauz as a family language. After long discussions among the intellectuals a Latin-based alphabet was introduced in January 1996. It should be mentioned, however, that questions of ethnicity and language were essential mostly for the intellectuals. The village population was more or less indifferent. Economic problems had greater importance for them and good command of Russian was still regarded indispensable. Changes in the economy had an influence on language maintenance in that Turkey attracted a lot of seasonal workers from the former Soviet Union and Turkish businessmen started to work in Moldova. Knowledge of Turkish thus became more important. With an increasing influence of Turkish, however, the distinctive features of Gagauz - either archaisms or elements developed under the influence of Slavic languages - may be lost. An increasing influence of Turkish was already no­ ticeable in 1995. Most of the villages received Turkish private television chan­ nels. Though a Gagauz television program existed it was transmitted for only half an hour monthly. Newspaper writers who had been invited to Turkey and taken language courses began to write in a "mixed" language, incomprehensi­ ble for the "ordinary" Gagauz, as some complained to me. With the imple­ mentation of the Latin alphabet Turkish print media and literature has become accessible and will have its impact. 3 Concluding remarks The prognosis for the maintenance of all the above mentioned languages cannot be optimistic. Those minority languages with reduced communicative function in restricted geographical areas are likely to be given up in the long run. Even with the revival movements under the changed political circum18

On the details of these activities see Chinn & Roper (1998) and Fane (1993).

ENDANGERED TURKIC LANGUAGES

153

stances in the former Soviet Union, it is doubtful whether minority languages and dialects can survive in modern industrial societies. This situation makes it all the more important to gain complete language descriptions and documenta­ tion. In the list of references I have therefore included the results from several fieldwork projects of the last two decades on endangered Turkic languages. References Backus, Ad. 1996. Two in one: bilingual speech of Turkish immigrants in The Netherlands. Tilburg: Tilburg University Press (Studies in multilingualism, 1). Bazin, Louis. 1959. "Structures et tendances communes des langues turques (Sprachbau)". Deny et al. 1959. 11-19. Beckherrn, Eberhard. 1990. Pulverfaß Sowjetunion. Der Nationalitätenkonflikt und seine Ursachen. München: Knaur. Boeschoten, Hendrik. 1998. "The speakers of Turkic". Johanson & Csató 1998. 1-15. Bremmer, Ian & Ray Taras. 1993. Nation and politics in the Soviet successor states. Cambridge: Cambridge University Press. Chinn, Jeff & Steven D. Roper. 1998. 'Territorial autonomy in Gagauzia". Nationalities Papers 26. 87-101. Csató, Éva Ágnes. 2000. "A syntactic asymmetry in Turkish". Göksel & Kerslake 2000. 417-422. Deny, Jean, Kaare Grønbech, Helmuth Scheel & Zeki Velidi Togan, eds. 1959. Philologiae turcicae fundamenta. 1. Aquis Mattiacis: Steiner. Doerfer, Gerhard. 1959. "Das Gagausische". Deny et al. 1959. 260-171. —. 1990. "Die Stellung des Osmanischen im Kreise des Oghusischen und seine Vorgeschichte". Hazai 1990. 13-34. —. 1998. "Turkic languages of Iran". Johanson & Csató 1998. 273-282. Dwyer, Arienne. 1998. "The Turkic strata of Salar: An Oghuz in Chagatay clothes?" Turkic Languages 2:1. 49-83. —. 2000a. "Consonantalization and obfuscation". Göksel & Kerslake 2000. 423-432. —. 2000b. "Direct and indirect experience in Salar". Johanson & Utas 2000. 45-59. Ehrensvärd, Ulla, ed. 1988. Turcica et Orientalia. Studies in honour of Gunnar Jarring on his eightieth birthday 12th October 1987. Stockholm: Swed­ ish Research Institute in Istanbul. Fane, Daria. 1993. "Moldova: breaking lose from Moscow". Bremmer & Taras 1993. 121-153. Göksel, Ash & Celia Kerslake, eds. 2000. Studies on Turkish and Turkic lan­ guages: proceedings of the ninth international conference on Turkish

154

ASTRID MENZ

linguistics, Lincoln College, Oxford, August 12-14, 1998. Wiesbaden: Harrassowitz (Turcologica, 46). Golden, Peter. 1992. An introduction to the history of the Turkic peoples: Ethnogenesis and state-formation in medieval and early modern Eura­ sia and the Middle-East. Wiesbaden: Harrassowitz (Turcologica, 9). Grannes, Alf. 1996. Turco-Bulgarica: Articles in English and French con­ cerning Turkish influence on Bulgarian. Wiesbaden: Harrassowitz (Turcologica, 30). Hazai, György, ed. 1990. Handbuch der türkischen Sprachwissenschaft. Teil I. Budapest: Akadémiai Kiadó (Bibliotheca Orientalis Hungarica, 31). Johanson, Lars. 1988. "Grenzen der Turcia. Verbindendes und Trennendes in der Entwicklung der Türkvölker". Ehrensvärd 1988. 51-61. —. 1998a. "The structure of Turkic". Johanson & Csató 1998. 30-66. —. 1998b. "The history of Turkic". Johanson & Csató 1998. 81-125. —. 2001. Discoveries on the Turkic Linguistic Map. Stockholm: Swedish Research Institute in Istanbul (Publications, 5). — et al., eds. 1998. The Mainz Meeting: proceedings of the seventh interna­ tional conference on Turkish linguistics. Wiesbaden: Harrassowitz (Turcologica, 32). — & Eva A. Csató, eds. 1998. The Turkic languages. London: Routledge (Routledge Language Family Descriptions). — & Bo Utas eds. 2000. Evidentials: Turkic, Iranian and neighbouring lan­ guages. Berlin: Mouton de Gruyter (Empirical Approaches to Language Typology, 24). Kiral, Filiz. 2000a. "Reflections on -miš in Khalaj". Johanson & Utas 2000. 89101. —. 2000b. "Copied relative constructions in Khalaj". Göksel & Kerslake 2000. 181-188. —. 2001. Das gesprochenen Aserbaidschanisch von Iran: Eine Studie zu den syntaktischen Einflüssen des Persischen. Wiesbaden: Harrassowitz (Turcologica, 43). Laut, Jens Peter & Mehmet Ölmez, eds. 1998. Bahşi Ögdisi: Festschrift für Klaus Röhrborn anläßlich seines 60. Geburtstags. Istanbul: Simurg 1998 (Türk Dilleri Araştirmalari Dizisi, 21). Menz, Astrid. 1999. Gagausische Syntax: Eine Studie zum kontaktinduzierten Sprachwandel. Wiesbaden: Harrassowitz (Turcologica, 41). —. 2000a. "Indirectivity in Gagauz". Johanson & Utas 2000. 103-114. —. 2000b. "Analytic modal constructions in Gagauz". Göksel & Kerslake 2000. 151-158.

ENDANGERED TURKIC LANGUAGES

155

Nathan, David. 2000. "The spoken Karaim CD: Sound, text, lexicon and "ac­ tive morphology" for language learning multimedia". Göksel & Kerslake 2000. 405-413. Nevskaja, Irina. 1998. "The revival of the Shor literary language". Turkic Languages 2:2. 253-269. Özkan, Nevzat. 1996. Gagavuz Türkçesi grameri: Giriş - ses bilgisi - şekil bilgisi - cümle - sözlük - metin örnekleri. Ankara: TDK (Türk Dil Kurumu Yaymlari, 657). Pritsak, Omeljan. 1959. "Das Karaimische". Deny et al. 1959. 318-340. Róna-Tas, András. 1998. "Turkic writing systems". Johanson & Csató 1998. 126-137. Roos, Marti, Hans Nugteren & Zhöng Jínwén. 1999. "On some proverbs of the Western and Eastern Yugur language". Turkic Languages 3:2. 189-214. Schönig, Claus. 1997a. "A new attempt to classify the Turkic languages (1)". Turkic languages 1. 117-133. —. 1997b. "A new attempt to classify the Turkic languages (2)". Turkic lan­ guages 1. 262-277. —. 1998a. "A new attempt to classify the Turkic languages (3)". Turkic lan­ guages 2. 130-151. —. 1998b. "Bemerkungen zum Fu-Yü Kirgisischen". Laut & Ölmez 1998. 317-348. Taube, Erika. 1998. "Observations of a non-linguist concerning the Tuvinian language in Tuva and Western Mongolia". Johanson et al. 1998. 647655.

LOSS OF LINGUISTIC DIVERSITY IN AFRICA MAARTEN MOUS Leiden University

1 Introduction The aim of this article is to provide an assessment of the linguistic diversity that is endangered in Africa. The African continent has a large number of lan­ guages, roughly 2000, or, one third of the world's linguistic heritage. At the outset we should state that the situation of the indigenous languages of Africa is in general healthy, presently and in the immediate future. Most African lan­ guages are not on the verge of extinction and many smaller languages, i.e. languages with fewer than 50,000 speakers, are quite stable and do not show reduction in number of speakers. In this respect the African situation is mark­ edly different from that in Europe and in the Americas. The Ethnologue (Grimes 2000) names 37 African languages that are on the verge of extinction against 161 in the Americas or in percentages: roughly 0.02% of the languages of Africa is in immediate danger against 1.15% (a percentage that is 60 times higher) of those in the Americas. When we consider every language that has under 10,000 speakers endangered - which seems to be reasonable -, the num­ ber for Africa rises to roughly 300. In other respects, too, the African situation is different from that in the Americas and Australia, where endangered languages are associated with (in­ digenous) peoples who belong to the "lower" classes of society. This is not generally the case in Africa. A number of African languages are now in a stronger socio-political situation than they were twenty years ago and these languages have gained ground against the European official languages: Wolof in Senegal, Bambara in Mali, Setswana in Botswana to name a few. Their success is a force against the long term threat of loss of linguistic diversity in a world dominated by English, a perspective that is predicted by some globalists. This does not relieve us from the task of documenting the languages that are endangered. It does, however, have its consequences for the need of linguistic attention to be devoted to the growing languages as well, and in particular when

158

MAARTEN MOUS

their incomplete level of description is a drawback for the use of the language in education. Another reason why language endangerment is less dramatic in Africa is because, contrary to the situation in Latin America, in Africa former colonial languages are not a major factor in language loss. Colonisation has not led to the marginalisation of the original peoples as it has in the Americas, and peo­ ples speaking endangered languages in Africa are often not economically worse off than their neighbours speaking healthier languages. In the majority of Afri­ can states the colonial language is the official language but this has little impact on the every-day communicative situations compared to, say, Spanish in Latin America. A notable exception is Nigerian Pidgin English in Nigeria. In Nigeria English has become the language of interethnic communication and in that process the language has changed considerably under the influence of the vari­ ous African mother tongues, to the extent that it is now considered a different language. Proposals have been made to propagate Nigerian Pidgin English and let it take over the role of Standard English in education and administration (Emenanjo 1985:127). In the case of Nigerian Pidgin English we see the birth of a new language as a consequence of globalisation. Other examples of lan­ guage birth under globalisation in Africa are the cases of urban youth slang losing its stigma and serving the function of bridging ethnicity and taking over all communication situations in the big city, as is the case with Iscamto in Johannesburg, South Africa, (Childs 1997). Centralisation of economy and administration was not the norm in Africa until recently (that is the colonial and post-colonial era), and this differs from the history of Europe and large parts of Asia. 2 Situations of Language Loss In order to understand how linguistic diversity is lost we have to examine the situations of linguistic loss first. Coupled with the great diversity of languages, Africa also has a diversity of situations of linguistic loss. We can distinguish at least five: 1. shift to the non-colonial official and national language 2. shift to the language of wider communication often as a consequence of settlement in the urban centres 3. shift to the dominant regional language 4. shift as a result of giving up traditional economy that is central to the identity of the group 5. vocabulary loss without shift In addition, there is the rare case of language death as a consequence of extinc­ tion of the people due to genocide as is happening to various Kordofanian

LOSS OF LINGUISTIC DIVERSITY IN AFRICA

159

languages in the Nuba Mountains in Sudan. Wars do have the effect of dis­ placing and separating people with an adverse effect on language maintenance. We briefly discuss these five more common ways of linguistic loss. Examples of African official languages that replace other African languages are Amharic, Somali, and Swahili. In Ethiopia Amharic is and has been domi­ nant for centuries and in the regions where the influence is oldest this has led to languages disappearing. Several smaller Central Cushitic languages have died out or are on the verge of extinction. Appleyard (1998) and Zelealem (1998) provide accounts of how two Central Cushitic languages, Qwarenya and K'emant, have given way to Amharic. The dominance of Amharic was directly linked to the dominance of the Amhara rulers and strengthened by the Ethio­ pian Christian Church. As a consequence of the political changes in the last decade Amharic is now hindered in its further spread by its ethnic association. The present policy of the Ethiopian government is to endorse and support local initiatives for local languages. The southern part of Ethiopia is still one of the richest in linguistic diversity in Africa and many smaller languages are in need of such support. Somali is the national language of Somalia and few other languages are spoken in the country. One of these, the Swahili-related language Chimwiini has been influenced by Somali over centuries and is losing more and more ground to Somali. In the recent diaspora due to the civil war the speakers of the language are dispersed and, without a community, the future of the lan­ guage looks very bleak. Swahili, an official language of Tanzania and Kenya, never had such a clear ethnic association as Amharic and Somali. The spread of Swahili in pre-colonial times into the interior was linked to trade. Nowadays ethnic Swahili are a negligible minority among mother-tongue speakers of Swahili. Swahili is replacing several local languages in Tanzania. The process of Swahili replacing other Tanzanian languages is one that involves several stages of bilingualism, which gradually topples to Swahili dominance and eventually to restricted competence in the original mother tongue (Batibo 1992). In particular, people moving to the urban centres tend to switch to Swa­ hili and no longer use their mother tongue. Some of the larger Bantu languages near the capital Dar es Salaam, for example, Zaramo and Bondei, are under strong influence of Swahili; Legère (1992) illustrates how Swahili penetrates every aspect of Bondei, even though as a people the Bondei are demographically healthy with over 200,000 people and growing fast. Amharic, Somali and Swahili are official languages and as such have some extra prestige. These are not the only languages of wider communication that are replacing other African languages. The larger replacing languages do not need the status of sole official language to persuade people to give up their mother tongue in their favour. Bamana (Bambara) or Jula in Mali and Burkina

160

MAARTEN MOUS

Faso, Hausa in Northern Nigeria and Niger, Lingala in Congo, and Wolof in Senegal are replacing many languages, often as a consequence of settlement in the urban centres. Bamana (Bambara) is de facto the national language of Mali where it is replacing smaller languages such as Kakolo (Sommer 1992), and its vehicular form, Jula, is becoming more and more dominant in south-western Burkina Faso, threatening several smaller Mande and Gur languages, and one Kru language (Seme). Some of these Gur languages (Tyefo and Viemo) are crucial in terms of diversity since they constitute primary branches of Gur. One of the most common threats to languages in Africa is shift to a dominant regional language. This regional language need not be numerically very strong, see for example the shift of Ongota to Ts'amakko (Savà, this volume). Other such examples are the gradual shift of Logba (Togo Mountain language) to Ewe (Kwa) in Ghana, of Bayso (Cushitic) to Wolayta (Omotic) and Oromo (Cushitic) (Brenzinger 2001), of Gweno in the Pare mountains to Chasu. This last shift has been described in some detail by Winter (1992) and Mreta (2000). Language shift can also be the result of a change in the social situation when the language is felt to belong to a specific socio-economic role. An example is the case of the Aasáx, a Southern Cushitic language spoken by hunter-gatherers among the Maasai, who gave up their language for Maasai while giving up hunting for cattle keeping and thus becoming Maasai (Winter 1979). Loss of linguistic diversity is not always identical to language death. A common process is that of loss of lexicon. Many Bantu languages of Tanzania, for example, gradually replace non-basic vocabulary with Swahili items, now that education has become formal and in Swahili. An example from Bowe (Bantu F 34) shows that the original Bowe words are in the process of being replaced by Swahili words, which are adapted to Bowe phonology and mor­ phology: the Swahili class 15 prefix for infinitives is replaced by its Bowe counterpart and the v is changed into a stop after a nasal, as in chumbi from Swahili chumvi 'salt' in the following example taken from Mous (2001). new-Bowe

Swahili

original Bowe

"be angry"

okasirika

ku-kasirika

ovéna

"to breathe"

opumua

ku-pumua

ofwéréra

"to mould"

ofinyánga

ku-finyanga

otulatüla

"to bend"

opinda

ku-pinda

uünánya

"salt"

chumbi (class 10)

chumvi

tony o (class 10)

Table 1 : Swahili influence on Bowe

In the light of loss of diversity it is important to realise that the overall loss of certain types of languages results in the disappearance of certain types of

LOSS OF LINGUISTIC DIVERSITY IN AFRICA

161

contact situations and of special languages. Secret languages, initiation lan­ guages, languages of occupational "castes" and of people dependent on others in a client situation are, in general, in danger. An example is the loss of the languages of the blacksmiths and leatherworkers in the Mande domain (Kas­ tenholz 1998). It is to be expected that in time language death will eradicate this type of sociolinguistic situation while the mechanisms of it are not unimportant for historical and anthropological linguistics. Thus, some endangered languages are of special interest for the insight that they can provide into language contact phenomena. For example, the study of displaced Bantu languages such as Ngoni from South Africa to Tanzania and the Bantu Mushungulu from Tanza­ nia to Somalia is of interest because they show insight into gradual language adaptation. The study of mixed Swahili varieties along the Mozambican coast is also important because of loss of a special contact situation. 3 Loss of Diversity Language death is more often than not a social loss. It is a sign of an ethnic group's lack of self-confidence, which in itself is a prerequisite for sustainable development. In addition, language death entails loss of cultural diversity. Loss of diversity is regrettable not only from a philosophical or aesthetic point of view, it can also deprive us from data that are crucial for increasing our insight into the human language capacity. It limits our possibilities to recover history and with the language indigenous knowledge disappears as well. It is evident that the loss of linguistic diversity is an impediment for linguis­ tics as an empirical discipline. Hayward (1998) shows that some of the endan­ gered languages of Ethiopia have typologically rare phenomena. The loss of genetic diversity also adversely affects our knowledge of the early history and migrations of the speakers of the proto-languages. Obviously languages that defy classification (isolates) are of special concern in this respect as they may be evidence for a situation of more linguistic diversity millennia ago. On the African continent there are also some such language groups, which are in their totality endangered: the Kuliak languages, the Kordofanian languages and the Khoe and non-Khoe Khoi-San language families. We now present a fuller discussion of the genetic linguistic diversity in Africa and the endangerment thereof. 4 Genetic Diversity of African Languages The standard classification of African languages that is used is that of Greenberg (1963). The invaluable merits of this work are widely recognized and unquestioned. Two issues, however, are of importance to us here. One is the question whether there is any doubt that all African languages fit into one of

162

MAARTEN MOUS

the four Greenbergian African phyla: Afroasiatic, Khoisan, Niger-Congo, and Nilo-Saharan. The second issue is whether this represents a remarkable lack of linguistic diversity on the African continent. As regards the first issue, there are of course reservations about the validity of some of the genetic relationships proposed by Greenberg. As regards the Afroasiatic phylum there is consensus that all major branches are indeed related, the only discussions are on subclassi­ fication of these branches and on the position of the Omotic language family or families. The validity of the Niger-Congo phylum has been questioned, mainly by non-Africanists. Campbell (1998) has argued that the family is nothing but a hypothesis because there is no historical reconstruction of the highest level. Dixon (1997) claims that the proposals for Niger-Congo are primarily based on typological criteria, which are invalid as proof for genetic relationships. Among Africanists there is little concern about these reservations. Very few researchers are indeed working on reconstruction at a level higher than the subfamily, despite the fact that John Stewart's work shows that such an enterprise is not impossible. The fact that a language has a Niger-Congo type noun-class system or remnants thereof is considered not to be a typological criterion because the presence of a Niger-Congo type noun-class system does not make reference to just function but rather to function in connection with form and paradigmatic systematicity and hence this is a valid criterion to posit genetic relatedness. As a consequence there is little discussion about the membership of the Kordofanian languages in Niger-Congo given their noun-classes even if these languages are always considered the first branching off (Williamson & Blench 2000). On the other hand the inclusion of Mande in Niger-Congo is highly disputed, partly because it is difficult to find remnants of the noun-class system. Within NigerCongo the subfamily of Ijo is not disputed on lexical grounds but the origin of the many traits that are highly atypical for Niger-Congo (word order, gender system) remains a mystery. Voeltz (1990) has argued that the evidence for inclusion of Ubangi in Niger-Congo is actually very weak, but the issue has not attracted much attention since. Nilo-Saharan was the last phylum that Greenberg proposed and many are not convinced of the higher-level genetic relation­ ships. The genetic relationships of the major groups, Central Sudanic, Eastern Sudanic and Saharan are more or less accepted (Bender 2000), even if Saharan is considered to be very divergent, but the inclusion of Songhay is highly de­ bated (Nicolai' 1990), as is the membership of Kuliak. Common opinion is that Kuliak belongs to Nilo-Saharan. Bender (2000) has it as an independent pri­ mary branch of Nilo-Saharan, but Lamberti (1988) provides evidence for an Afroasiatic connection. The position of the remaining languages is very un­ clear. For example, Gumuz and Koman are in Bender's core group, while Ehret (2000) has them as a first, separate branch. There is a reason for this astonish-

LOSS OF LINGUISTIC DIVERSITY IN AFRICA

163

ing divergence of opinion in the classification of these languages, namely for many of these non-core Nilo-Saharan languages there is little documentation and even less diachronic study.1 The Khoisan phylum is most probably not a genetic unit. Sands (1998) has shown that there is not enough positive evidence to link Hadza to Sandawe or to the South-African Khoisan languages. Elderkin (1989) links Sandawe lexically to Khoe (Central Khoisan) and sees structural similarities with Northern Khoisan, but this is not accepted by Güldemann & Vossen (2000). Vossen (1997) provides ample evidence for the unity of Khoe or Central Khoisan, but questions the existence of a genetic relationship be­ tween Khoe and the rest of South-African Khoisan. Other puzzles are Kwadi (see below) and Dahalo. Dahalo is a Cushitic language with some clicks and it has been suggested that the origin of the lexemes with clicks lies in a former Khoisan language that they spoke before shifting to a Cushitic language (Nurse 1986). The second question that arises from the Greenberg classification is: is Af­ rica remarkably homogeneous linguistically? Human population started in Africa and came much later to the Americas and yet the number of language families in the Americas is much higher than in Africa. Is this because the language family relationships in the Americas have not yet been recognized or accepted? Is it because the African situation has been idealized due to a ten­ dency to incorporate every language into one of the four phyla proposed by Greenberg? Was there more contact in Africa and has a prior diversity con­ verged after millennia? Has the spread of the Niger-Congo family and in par­ ticular that of Bantu simply wiped out a former greater diversity? Obviously these questions are not easily answered but they warrant a special interest for those languages or pockets of languages that pose a challenge to classification and that are in danger of disappearance (Blench 1999). As an assessment of the endangerment of language diversity in Africa I present an overview starting with language isolates that are endangered and (sub-)families of languages that are endangered in their entirety. This in turn is followed by a list of the remaining language families in Africa. For each family I give (by means of an abbreviation) the phylum to which they belong accord­ ing to the Greenberg classification, information on how many languages are in the family, the geographical area of the family, and an impressionistic indica­ tion of the situation vis-à-vis language endangerment (that is, the number of languages with fewer than 10,000 speakers). It is not a list of endangered lan­ guages, for which the reader is referred to Sommer (1992), which contains 140 languages that are in the process of extinction or in immediate danger of exThat is, those languages that are not part of Central Sudanic, Eastern Sudanic or Saharan.

164

MAARTEN MOUS

tinction (her criterion is 500 speakers), and also to lists that will be published in the proceedings of a recent conference in Bonn (Brenzinger, forthcoming). Other important sources for the data in the following list are Grimes (2000), Crozier & Blench (1992) for Nigeria, and Dieu & Renaud (1983) for Camer­ oon. The number of languages is based on the Ethnologue (Grimes 2000), which tends to make distinctions where others would group lects into one lan­ guage. As a consequence the number of languages that I give for the various families is higher than what is usually cited. For example, the Ethnologue has 195 Chadic languages while most other sources quote much lower numbers. 5 Isolates An overview of remnant hunter-gatherer societies and linguistic isolates in Africa can be found in Blench (1999). The following endangered languages have been considered isolates. For some of them this may be due to a lack of material, while for others a substantial amount of deviant basic lexical material poses the puzzle of their classification. Ongota is an unclassified language with 8 speakers in southwest Ethiopia shifting to a minority Cushitic language, Ts'amakko (Savà, this volume). Shabo is an unclassified language of southwest Ethiopia with about 200-600 speakers. The proposals for its classification range from Nilo-Saharan to Omotic to unclassifiable. Some data are published in Anbessa (1991, 1995) and Daniel Aberra (Addis Ababa) has further material. Hadza is a linguistic isolate (Sands 1998) of roughly 200 speakers in north­ ern Tanzania. The published material on the Hadza language is very limited. The population seems to have been more or less stable during the last decennia, but these traditional hunter-gatherers have become more and more dependent on food aid from outside. Jalaa is spoken by 20 elderly people among the Cham of northeast Nigeria, (Blench 1999:54, based on data from Kleinewillinghöfer). Laal is an unclassified language with roughly 300 speakers in south-central Chad. Kwadi is an unclassified language formerly spoken by hunter-gatherers in Angola. Their language contains clicks but a relation to Khoe is not clear (Blench 1999). Güldemann & Vossen (2000) mark it as (presumably) extinct. Pre, spoken in northern Ivory Coast, is probably Kwa, but contains substan­ tial Mande material (Williamson & Blench 2000). Mpre (Ghana) is most probably extinct. The reported vocabulary defies classification (Williamson & Blench 2000).

LOSS OF LINGUISTIC DIVERSITY IN AFRICA

165

6 Endangered families Khoe (Central KS) is the undisputed core of the hypothetical Khoisan phy­ lum. The group contains a number of extinct languages, the largest and only viable Khoisan language, Nama, and about fifteen other languages spoken in Namibia, Botswana, Angola and South Africa that are all endangered to various degrees. Non-Khoe Khoisan (Northern and Southern Khoisan) primarily consists of extinct languages (Southern Khoisan) and some highly endangered languages, only the Ju|'hoan dialect cluster still has a respectable number of speakers. Kordofanian (NC): a group of 20 languages spoken in southern Sudan forming the first branch of Niger-Congo. All the languages in the group are endangered, partly by genocide. Kuliak (classification debated): a group of 3 languages in Uganda, all of which are nearly extinct. Togo-Mountain languages, also known as Togo-Remnant languages, are in fact two small sub-families within Kwa (NC), consisting of about seven lan­ guages each, all of which are in a state of endangerment. Koman (NS) is a group of several languages in the Sudan-Ethiopia border area. Some of them were displaced because of the prolonged war situation and none of them have more than a few thousands of speakers (Bender 2000). 7 Other f amities Adamawa (NC): a family of 86 languages spoken in Northern Nigeria, Northern Cameroon, Chad and the Central African Republic. Most of these languages are endangered. Many people are shifting to Fulfulde or the larger regional languages. The Adamawa languages in Nigeria and Chad are particu­ larly poorly known. Atlantic (NC): a family of 65 languages mainly along the coast from Senegal to Sierra Leone. It contains two large languages on which nearly all research has been concentrated, namely Fulfulde and Wolof. At least 13 Atlantic lan­ guages have fewer than 10,000 speakers. Wolof and Mandinka are dominant. A primary sub-branch of Atlantic consisting of one language, Bijago, is under pressure. Bantu (NC): a family of roughly 500 languages spoken in the southern half of Africa. The group forms a unique research area because of the sheer numbers of closely related but different languages. In the area of endangerment, the special contact situations are of interest, e.g. Chimwiini (Swahili in Somalia), Chimakwe (Makonde-Swahili mixed variety predominantly spoken by women)

166

MAARTEN MOUS

in Mozambique. There is substantial loss of specialised vocabulary in the Bantu languages in East Africa through the influence of Swahili. Benue-Congo (excluding Bantu) (NC): a family of over 200 languages mainly spoken in Nigeria. The group contains some of the larger languages such as Yoruba, Igbo, Efik, Ibibio, Tiv, and Nupe. But for the rest it contains large groups of smaller and for the most part poorly described languages, many of them endangered, especially in Northern Nigeria, due to the expansion of Hausa. Large sub-groups are Kainji (55), Plateau (63), Cross-River (63). Some languages that are of special concern due to their unique position within the family and their endangered status are Ukaan, an isolate within Eastern Benue Congo, and Akpes and Ayere-Ahan, which are isolates within West BenueCongo (Williamson & Blench 2000). Berber (AA): a family of 29 varieties. In terms of number of speakers, the languages of oases in Algeria and further east are endangered. The languages are threatened because of lack of prestige. Recently the status of Berber in Morocco has started to improve and an institute of Berber studies has been set up in Rabat. Variation among the Berber languages is played down for political reasons by Berber activists. Central-Sudanic (NS): a family of 64 languages spoken mainly in Chad, Central African Republic, Eastern Congo (DRC) and Southern Sudan. Most languages are small in numbers. Chadic (AA): a family of 195 languages (according to the Ethnologue) spoken in Northern Nigeria, Northern Cameroon, Chad and Niger. In Northern Nigeria the smaller languages are threatened by Hausa. Cushitic (AA): a family of 47 languages spoken in the Horn of Africa ex­ tending to Kenya and Tanzania. Several died out in recent years or are on the verge of extinction (Aasáx, Qwadza, Yaaku, Elmolo, Weyto, K'emant).2 En­ dangered are Arbore, Bayso, Dahalo and Ts'amakko.3 Dogon (NC): Dogon is a family of closely related languages spoken in Mali and Burkina Faso, which proves very difficult to classify within Niger-Congo. The languages are not in immediate danger except for the very intriguing secret languages that have been reported for Dogon. Gur (NC): a family of 100 languages spoken in northern Ivory Coast, Ghana, Togo, Benin, Mali and Burkina Faso. It contains many smaller endangered languages.

2

Most likely Aasáx is not dead as I met some speakers in 1995 north of Orkessumet in the Maasai plains of Tanzania. 3 The Ongota shifting to Ts'amakko are not numerous enough to stop Ts'amakko itself from being endangered.

LOSS OF LINGUISTIC DIVERSITY IN AFRICA

167

Ijo (NC): a group of 11 languages spoken in southeast Nigeria. The lan­ guages are highly divergent typologically from a Niger-Congo perspective. Defaka is the first branch within the group but also endangered. Kru (NC): a group of 41 languages spoken in Liberia and the Ivory Coast. Roughly 8 of these have speaker numbers under 10,000. One Km language, Seme, a primary branch, is spoken in Burkina Faso and is under pressure from Jula. Other endangered single language primary branches are Kuwaa in Liberia and Aizi in the Ivory Coast. Kwa (NC): a family of 78 languages in the southern half of the Ivory Coast, Ghana, Togo and Benin. A lot of research is going on, especially on syntax, and particularly on the Gbe and Akan clusters. The cluster of some 15 languages previously classified as Togo-Remnant are endangered. Ega (spoken in Ivory Coast) occupies a special position within Kwa in being highly divergent, and furthermore surrounded by Kru speakers. The language is endangered and has been studied in a project within the endangered languages programme of the Volks wagenstiftung. Mande (classification debated): a family of 58 languages spoken in all West African countries. Some ten languages are endangered. Nilotic (NS): a large family of 60 lects. Several languages are endangered: Ongamo, Okiek, Akiek, Omotik and others. Omotic (AA): a family of 28 languages, half of which have fewer than 10,000 speakers. Saharan (NS): a group of 6 languages (Kanuri, Kanembu, Teda, Daza, Zaghawa, Berti) spoken in the border area of Niger-Chad-Niger. Songhay (NS, debated): a group of closely related languages (not endan­ gered). Some mixed Songhay-Touareg varieties are of special interest for con­ tact linguistics. Semitic (AA): Arabic and its variants, Southern Semitic (Amharic, Tigre, Tigrinya, Gurage) in Ethiopia and Eritrea. The African Semitic languages are not in a situation of language endangerment, with the exception of Gafat. Ubangi (NC): a group of 71 languages spoken in the Central African Repub­ lic, Cameroon, Congo-Brazzaville, Congo (DRC) and Sudan. The larger ones are Sango, Gbaya, Ngbandi, Banda, and Zande. More than 20 with fewer than 10,000 speakers.

168

MAARTEN MOUS

References Anbessa Teferra. 1991. "Sketch of Shabo grammar". Bender 1991. 371-381. —. 1995. "Brief phonology of Shabo (Mekeyir)". Nicolai & Rottland 1995. 2937. Appleyard, David. 1998. "Language death: the case of Qwarenya (Ethiopia)". Brenzinger 1998. 143-162. Batibo, Herman. 1992. "The fate of ethnic languages in Tanzania". Brenzinger 1992. 85-98. Bender, Lionel M., ed. 1991. Proceedings of the fourth Nilo-Saharan linguis­ tics colloquium. Hamburg: Buske. —. 2000. "Nilo-Saharan". Heine & Nurse 2000. 43-73. Biesbrouck, Karen, Stefan Elders & Gerda Rossel, eds. 1999. Central African hunter-gatherers in a multidisciplinary perspective: challenging elusiveness. Leiden: CNWS. Blench, Roger. 1999. "Are the African pygmies an ethnographic fiction?". Biesbrouck, Elders & Rossel 1999. 41-60. Brenzinger, Matthias, ed. 1992. Language death: factual and theoretical explo­ rations with special reference to East Africa. Berlin: Mouton De Gruyter. — ed. 1998. Endangered languages in Africa. Cologne: Köppe. —. 2001. "Language endangerment through marginalization and globalization". Sakiyama& Endo 2001. 91-116. — ed. forthcoming. Language diversity endangered. Berlin: Mouton De Gruyter. Campbell, Lyle. 1998. Historical linguistics: an introduction. Edinburgh: Edin­ burgh University Press. Childs, G. Tucker. 1997. "The status of Isicamtho, an Nguni-based urban vari­ ety of Soweto". Spears & Winford 1997. 341-367. Crozier, D.H. & Roger M. Blench. 1992. An index of Nigerian languages. 2nd ed. Dallas: Summer Institute of Linguists. Dieu, Michel & Patrick Renaud. 1983. Atlas linguistique del'Afrique centrale (ALAC): Atlas linguistique du Cameroun (ALCAM). Situation linguis­ tique en Afrique centrale. Inventaire préliminaire: le Cameroun. Paris: ACCT; Yaounde: CERDOTOLA/DGRST. Dixon, R.M.W. 1997. The rise and fall of languages. Cambridge: Cambridge University Press. Ehret, Christopher. 2000. "Language and history". Heine & Nurse 2000. 272297. Elderkin, Edward Derek. 1989. The significance and origin of the use of pitch in Sandawe. Ph.D. dissertation, University of York.

LOSS OF LINGUISTIC DIVERSITY IN AFRICA

169

Emenanjo, E.'Nolue. 1985. "Nigerian language policy: perspective and pro­ spective". Journal of the Linguistic Association of Nigeria 3. 23-131. Greenberg, Joseph H. 1963. The languages of Africa. Bloomington: Indiana Univ. Research Center in Anthr., Folklore and Linguistics (Publication, 25). Grimes, Barbara, ed. 2000. The ethnologue: languages of the world (14th edi­ tion) http://www.ethnologue.com. Güldemann, Tom & Rainer Vossen. 2000. "Khoisan". Heine & Nurse 2000. 99122. Hayward, Richard. 1998. "The endangered languages of Ethiopia: what's at stake for the linguist". Brenzinger 1998. 17-38. Heine, Bernd & Derek Nurse, ed. 2000. African Languages: an introduction. Cambridge: Cambridge University Press. Kahigi, Kulikoyela K., Yared Kihore & Maarten Mous, eds. 2000. Lugha za Tanzania /Languages of Tanzania: studies dedicated to the memory of Prof ClementMaganga. Leiden: CNWS. Kastenholz, Raimund. 1998. "Language shift and language death among Mande blacksmiths and leatherworkers in the diaspora". Brenzinger 1998. 253266. Lamberti, Marcello. 1988. Kuliak and Cushitic: a comparative study. Heidel­ berg: Winter. Legère, Karsten. 1992. "Language shift in Tanzania". Brenzinger 1992. 99-115. Mous, Maarten. 2001 "Paralexification in language intertwining". Smith & Veenstra2001. 113-123. Mreta, Abel. 2000. "The nature and effect of Chasu-Kigweno contact". Kahigi, Kihore & Mous 2000. 177-189. Nicolai', Robert. 1990. Parentés linguistiques: à propos du songhay. Paris: Éditions du CNRS. —. & Franz Rottland, ed. 1995. Proceedings of the fifth Nilo-Saharan linguis­ tics colloquium. Cologne: Rüdiger Köppe. Nurse, Derek. 1986. "Reconstruction of Dahalo history through evidence from loanwords". Sprache und Geschichte in Afrika 7:2. 267-305. Sakiyama, Osamu & Fubito Endo, ed. 2001. Lectures on endangered languages 2. From Kyoto Conference 2000. Osaka: Project Endangered Languages of the Pacific Rim. Sands, Bonny. 1998. Eastern and Southern African Khoisan: evaluating claims of distant linguistic relationships. Cologne: Köppe (Quellen zur Khoisan-Forschung, 14). Smith, Norval & Tonjes Veenstra, eds. 2001. Creolization and contact. Am­ sterdam: Benjamins (Creole Language Library, 23).

170

MAARTEN MOUS

Sommer, Gabriële. 1992. "A survey on language death in Africa". Brenzinger 1992.301-413. Spears, Arthur K. & Donald Winford, eds. 1997. The structure and status of pidgins and creoles, including selected papers from the meetings of the Society for Pidgin and Creole Linguistics. Amsterdam: Benjamins. Voeltz, Erhard. 1990. Paper presented at the VIII. Afrikanistentag, Wien. Vossen, Rainer. 1997. Die Khoe-Sprachen: Ein Beitrag zur Erforschung der Sprachgeschichte Afrikas. Cologne: Köppe (Quellen zur KhoisanForschung, 12). Williamson, Kay & Roger Blench. 2000. "Niger-Congo". Heine & Nurse 2000. 11-42. Winter, J. Christoph. 1979. "Language shift among the Aasáx, a hunter-gatherer tribe in Tanzania: A historical and sociolinguistic case-study". Sprache und Geschichte in Afrika 1. 175-204. —. 1992. "175 years of language shift in Gweno". Brenzinger 1992. 285-298. Zelealem Leyew. 1998. "Some structural signs of obsolescence in K'emant". Brenzinger 1998. 163-185.

ONGOTA (BIRALE), A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA GRAZIANO SAVÀ Leiden University

1 Introduction Ethiopia is not an exception in the African continent in having a multitude (about 80) of languages spoken within its boundaries. The linguistic situation of Ethiopia is interesting not only for the high number of languages but also for the diversity of linguistic groups found in its territory. These groups are Se­ mitic, Cushitic, Omotic1, Surmic and Nilotic. The first three represent main branches of the Afroasiatic family, while Surmic and Nilotic constitute the northeastern fringe of the Nilo-Saharan family. Half of the Ethiopian languages are found in southwest Ethiopia, the Federal State of the Southern Nations, Nationalities and Peoples (Federal Democratic Republic of Ethiopia 1994). As its name suggests, this is a multiethnic region made up of 44 peoples, varied both in number and cultural background. The languages in this region range over the five linguistic groups mentioned above. Moreover, in this area there is at least one language classification of which is still uncertain. It is a language on the brink of extinction, called Ongota by the people who speak it and Birale by neighbouring peoples. This paper discusses the sociolinguistic situation of this language and presents some of its elementary features. 2 The people Birale is a name that sounds despising to the Ongota. Hereafter, we will not use it, preferring the self-referent Ongota ['o gota]2 for both the people and the language. The Ongota people have probably already been mentioned in 1896 by Donaldson Smith who writes of Borali whose name he heard while travel­ ling in southwest Ethiopia. The Ongota live in the forest on the west bank of 1 This group formerly represented, and to some still represents, West Cushitic. 2 Note that this name should be written with an initial voiced pharyngeal /7, which has been omitted to simplify our spelling.

172

GRAZIANO SAVA

the Weyt'o River (known as Dullay in the area)3. Their main village, Muts'e, is a good hour's walk along the river from the bridge on the main Konso-Jinka road. The small Ongota settlement is surrounded by peoples who speak Dullay dialects, which form a branch of Lowland East Cushitic. More precisely, it is found inside the area of one of these peoples, the Ts'amakko. The Ts'amakko number about 10,000 and are the only Dullay speakers found on the west bank of the Weyt'o River. As we will see, they are directly related to the endangered status of the Ongota language4. This lowland region, known as the plain of the Weyt'o River, is included in a much wider Cattle Culture Area. Although sorghum and maize cultivation and bee keeping are widespread, pastoralism here is the main economic activity and the only means of production of wealth. This is valid for the Ts'amakko which sets them apart from the rest of the Dullay peoples and this is equally valid for their neighbours the Hamar (who speak an Omotic language). Among these cattle peoples, Ongota are the ex­ ception, being the only hunter-gatherers. Their support is based on wild ani­ mals, fish, wild edible plants, honey, although they also practice hoe agricul­ ture at a small scale. According to a local myth, the Ongota originate from the coexistence of different ethnic groups. This multiethnic and multilinguistic melting pot was composed of the North Omotic Maale, the South Omotic Banna and the East Cushitic Borana and Dishina (a Dullay group). They are linked to what are now the Ongota clans. The myth of the Ongota origin has been collected by Aklilu Yilma (p.c.) and by Savà & Tosco (2000). No specific historical or ethno­ graphic study of the Ongota people has been carried out, but precious informa­ tion is found in Melesse Getu's publications on the economy of the Ts'amakko (Melesse Getu 1995; 1997). He writes that the Ongota are not pastoralists, as are the other peoples in the area, because of the strong presence of the tsetse fly in their forest. Moreover, they are facing the huge problem of the impoverish­ ment of the fauna of the forest and the river. The number of animals is de­ creasing mainly due to the presence and use of firearms in the last decades as well as to the process of desertification. The latter has additional negative effects on bee keeping and incense collection which constitute the only goods the Ongota can sell. The drastic impoverishment of the Weyt'o River is mainly related to the construction of a dam which diverts the water to irrigate the immense "Birale Cotton Farm" situated nearby. It is during the dry season, when the water level is low and the river flows from the C'ew Bahir Lake 3 The 1980 version of the Ethnologue wrongly locates the Ongota North of Weyt'o Lake (sic). The thirteenth edition (Grimes 1996) gives the correct information. 4 The only published data on Ts'amakko are to be found in Hayward (1989) and Amborn, Minker& Sasse (1980).

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

173

towards the source, that the Ongota concentrate on their fishing activity. How­ ever, most of the fishes do not have the strength to pass the dam. The Ongota are known to have avoided contacts with other people in the past (information from Shelo Laybo, a Ts'amakko policeman of Weyt'o police station). At a certain point they have started to build relations and have slightly changed their life style, practicing some agriculture. Probably this attitude has been favoured by the worsening of the living conditions in their bush area. Their neighbours consider their "wild" life condition, their "incapacity" to produce food on their own, and their "funny" language ridiculous. They have to face the derision and the raids of more powerful and richer people. The only people with whom they have a better relation are the Ts'amakko. Nowadays the Ongota take part in the Ts'amakko Sunday market in the village of Weyt'o and intermarry with Ts'amakko5. The Ongota have felt and will feel the need to change their de­ spised status. It is therefore understandable that they have started to give up their ridiculed traditional language. Given the nature of their relations with the other peoples, it is not surprising that they adopted the language of the Ts'amakko. Because of the above-mentioned socio-economic reasons, the Ufa 'ongota ("mouth" = "language of the Ongota") is in an advanced process of extinction. According to the National Census, the Ongota were numbered at 85 in 1994. Counts made by researchers reaches 85 (Fleming et al. 1992-93), 75 (Aklilu Yilma p.c.) and about one hundred (Savà & Tosco 2000). The language, how­ ever, has not been taught to children probably since two generations. In fact, today only a residual group of 8 elders speak it. They are: Mole Sagane, his older brothers Aburre Sagane and Tabba Sagane (of the baritto clan), a group of four brothers, Dulo Korayo, Oydalle Korayo, Guya'o Korayo and lida Korayo (of the 'amad'd'o clan), and Gacco Olle (of the 'izamakko clan). We also heard of four women having a very limited knowledge of Ongota but we could not record their names. 3 The language So far, the important pioneering work by Harold Fleming and six other scholars (Aklilu Yilma, Ayyalew Mitiku, Richard Hayward, Yukio Miyawaki, Pavel Mikes and J. Michael Seelig, cf. Fleming et al. 1992-93) produced the only article containing a sketch of the Ongota language, providing a good description and a consistent glossary6. Unfortunately, the important part of the 5

According to Mole Sagane, our main Ongota informant, the cattle received as dowry are recycled as dowry for the family of a Ts'amakko girl who marries an Ongota. 6 Aklilu Yilma, a researcher of the Addis Ababa Institute of Ethiopian Studies, was in charge of collecting the data.

174

GRAZIANO SAVA

verbal system was covered only marginally and sentences are presented with­ out analysis. It does, however, give a fair idea of the structure of Ongota. In 1994 our knowledge of Ongota improved in the form of a wordlist consisting of 230 items collected by Dinota and Siebert in the context of the Survey of Little Known Languages of Ethiopia (S.L.L.E.) (Dinota & Siebert 1994). The items are in accurate phonetic transcription, but lack a phonological analysis. The same S.L.L.E. report includes comparable wordlists in Ts'amakko and Arbore. Afterwards, the Ongota have been the object of a sociolinguistic visit by Aklilu Yilma (p.c.) and of a linguistic field research by Savà and Tosco. Savà and Tosco initially encountered three Ongota speakers with the help of the manager of the local cotton farm (the Birale Cotton Company). After a small test, we realised that their knowledge of Ongota was not good enough for proper linguistic work. However, some Ts'amakko friends assured us they would provide a much better informant who turned out to be Mole Sagane, an Ongota elder who had already been the informant for Dinota and Siebert. Thanks to the great spirit of collaboration and patience of Mole, in ten days of full immersion elicitation we were able to collect enough material for a 76page sketch (Savà & Tosco 2000). The languages of elicitation were Ts'amakko and Amharic. Due to Mole's poor knowledge of Amharic, many sentences were translated from Amharic to Ts'amakko by Olle Fattale, a Ts'amakko policeman. One session had to be conducted through Borana Oromo, one of the languages of the area spoken by Mole, with the help of Arkano, whose first language, however, is Arbore. This procedure was work­ able due to Sava's knowledge of Amharic and his linguistic research on Ts'amakko. Due to space limitations we cannot provide a complete sketch, instead some salient features of Ongota phonology and grammar (for a more complete de­ scription the reader is referred to Fleming et al. 1992-93 and to Savà & Tosco 2000). The language is heavily influenced by Ts'amakko, although it shows some interesting original aspects. Most of the examples are extracted from Savà & Tosco (2000). In Table 1 we present the Ongota consonantal phonemes, their phonetic realisation and the orthography used in this article. When two consonants appear in one cell, the first is voiceless. The gap of the voiceless oral bilabial plosive is a common feature of the wider area. The Ongota consonantal distri­ bution is nearly identical to that of Ts'amakko. The only differences with Ts'amakko are the absence of a globalised (implosive) labial /bV in Ts'amakko and the lack of glottalisation in Ongota in the alveolar affricate /ts/ and in the voiced pharyngeal /7. The latter is a marked feature of Ts'amakko.

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

bilabial Plosives, oral

(p st) al­ veolar

labio­ dental

uvu­ lar q

t, d

k, g

d'

g'

m

n

Affricates Fricatives

f

ts

c,j

s, z

š

X

pharyngal

glot­ tal

',

hh,'

h

r

Trill Approx. Central

ve­ lar

b

P1 implosive Pl nasal

palatoalveolar

175

w

Approx. Lateral

1 Table 1: Phonological chart (modified from Savà & Tosco 2000)

The vowel system follows the Ts'amakko and the general Cushitic arrange­ ment with 5 long and 5 short vowels: i-ii, e-ee, a-aa, o-oo, u-uu. There is some height neutralisation between and and e and i both in Ts'amakko and On­ gota which still needs to be clarified7. Ongota is an OSV language in which independent subject pronouns can optionally appear in first position with emphatic meaning. In a declarative sentence subject clitics precede the verb. The Lowland East Cushitic lan­ guages, including the Dullay dialects, make use of such preverbal subject markers, in addition to the independent subject pronouns. The forms of these markers in Ongota are, however, different from those in the Lowland East Cushitic languages. Ts'amakko has retained only the 1SG and 1PL preverbal markers as shown in Table 2. 1SG

ka-

1PL ju-

2SG

i-

2PL gida-

3SG.M

ki-

3PL ki-

3SG.F

ku-

Table 2: Preverbal subject markers

7

This is an enigmatic phenomenon observed also in the Konsoid languages and in the rest of the Dullay dialects (Tosco p.c.).

176

GRAZIANO SAVA

These preverbal subject markers are partially related in form to other pronomi­ nal series such as the (optional) emphatic subject pronouns, the two object pronoun series (one is postverbal), the possessive pronouns and the ablative (prefixed to the adposition tu "from") and the locative "on" pronominal series (suffixed to uku, ugu, eke). The indirect pronoun series has only been recorded for the singular: subject emphatic

object

object postpos.

indirect

possessive

ablative

locative "on"

1Sg

kata

ka

ka

na

sinne

ka-tu

uku-ni

2Sg

janta ~ jaama

jami

jan

jata

siidu

jan-tu

ugu-du

3SgM

kita

ki

ki

wana

seena

kii-tu

eke-na

3SgF

kuta

ku

ku

wata

suu 'u

kuu-tu

uku- 'u,uku-wi

1P1

juta

ju

ju

sijju

ju-tu

uku-šijja

2P1

gidata

gida

gida

sigida

gida-tu

uku-gida

3P1

ki 'ita

ki 'i/-a

ki'i

-

suwaya

ki 'i-tu

uku-waya

Table 3: Pronouns (modified from Savà & Tosco 2000)

Ongota nouns do not distinguish gender and number overtly. These categories can be observed only in the agreement of the pronominals in Tables 2 and 3 above. The agreement in the preverbal subject markers does not distinguish between 3SG.M and 3PL, both being marked by ki. Only the context can show if reference is made to a masculine singular or to a plural noun. Small and unimportant animals are usually feminine. (1)

oxaya ki-tib lion 3SG.M/PL-die.PAST "a/the lion died" or "the lions died"

(2 a)

karbo ku-tib bird 3SG.F-die.PAST "a/the bird died"

(2b)

karbo ki-tib bird 3 SG.M/PL-die.PAST "the birds died"

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

177

In a few cases the plural has a different root: (3a) (3b) (3c) (3d) (3e)

Singular ayma inta jaaka juuka maara

Plural aaka yooba eela igire eela

"woman, female, wife" "man, male, husband" "child, baby" "girl, daughter" "boy, son"

Some adjectives show number marking, -uni (sg.) and -eta (pl.): (4a)

gadda '-uni big-SG "big, large"

gidde'-eta big-PL

(4b)

munnu '-uni small-SG "small, little"

min '-eta small-PL

Others end invariably in -a: (5a)

inta aadala man bad "bad man"

(5b)

ayma abba woman nice "nice woman"

Basic adjectives of Ts'amakko origin have kept the Ts'amakko gender and number endings: (6a)

gecc-ate old-F "old"

gecc-ayke old-PL

(6b)

kamur-ko rich-M "rich"

kamur-te rich-F

(6c)

zaar-akko zaar-atte crazy-M crazy-F "fool, crazy"

zaar-ayke crazy-PL

The endings -uni and those of Ts'amakko origin also appear in the derivation of adjectives from verbs: (7a)

inta ša 'at-uni man afraid-ADJ "scared man"

178

GRAZIANO SAVA (7b)

kata ka-ša 'atí 1SG-be.afraid.NON-PAST "I am scared"

(8 a)

ayma-ko 'erehh-te woman-DET pregnant "the pregnant woman"

(8b)

ku- 'érehhi 3SG.F-be pregnant.PAST "she became pregnant"

(9 a)

intadaaf-akko man blind "blind man"

(9b)

kitaki-dáaf 3SG.M-beblind.PAST "he became blind"

In addition to attribution in the form of an adjective, Ongota also had the pos­ sibility of stative verbs used attributively, a feature common in Ethiopia: (10)

hanca ki-šóoni wood 3SG.M-be hot.PAST "the hot wood"

Gender and number in verbs are not distinguished by means of suffixes as is the norm in Cushitic languages. The verbal form is invariable and person is indicated in the preverbal subject marker only: ka-šúb

I killed

i-šúb

you killed

ki-šúb

he killed

ku-šúb

she killed

ju-šúb

we killed

gida-šúb

you (p1.) killed

ki-šúb

they killed

Table 4: Verb conjugation

Ongota has tense distinctions in verbs. This is a feature not described in Fleming et a1. (1992-93). The distinction is between a past and a non-past which are not marked by morphological segments, but, interestingly, by the position of tone on the verb stem. A past verb has a high tone on the first (or

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

179

only) mora of the root. On long monosyllabic verb roots ( C W C ) this tone is realised as falling. In polysyllabic verbs the first syllable receives a high tone: (1 1a) ka-bú' 1SG-spit.PAST "I spat" (11b)

ku-xáàb 3SGF-scratch.PAST "she scratched"

(1 1c) ju-íški 1 PL-play.PAST "we played"

A non-past verb has no tone if it is monomoraic (CVC). In this case, the sub­ ject clitic gets high tone. If it is at least bimoraic, it has tone on the last mora and no high tone is on the clitics. In long monosyllabic verbs ( C W C ) a rising tone is realised. In bi- and pluri-syllabic verbs the last syllable gets high tone: (12a)

ká-bu' 1SG-spit.NON-PAST "I'll spit"

(12b)

ku-xàáb 3SG-scratch.NON-PAST "she'll scratch"

(12c)

ju-iškí lPL-play.NON-PAST "we'll play"

The distinction is neutralised in the negative paradigms. The negative form is based on the past tonal pattern and it is made up of a suffix -'i and by the nega­ tive marker ma between the subject clitic and the verbal form: (13)

ka-ma-eeni-'i 1SG-NEG-come.PAST-SUFFIX "I won't come"

The Ongota imperative forms are -á for singular addressees and -ta for plural addressees: (14a)

bud'-à spit.IMPV.2SG "spit!"

(14b)

búd'-ta spit.IMPV.2PL "spit!"

180

GRAZIANO SAVA (15a)

gufa'-á cough.IMPV.2SG "cough!"

(15b)

gufá'-ta cough.IMPV.2PL "cough!"

The negative imperative is expressed by the preverbal element 'intima. The plural form contains the 2PL subject clitic gida: (16a)

'intima qáfi NEG 2SG.fish.PAST "don't fish!"

(16b)

'intima gida-qáfi NEG 2PL.fish.PAST "don't (pl.) fish!"

The infinitive is formed by the suffix -o preceded by the gemination of the last consonant of the stem: (17)

kara šúb-bo ka-hháabini fish kill-INF lSG-want "I want to kill fish"

The same holds for Ts'amakko. Most of the verbal derivational suffixes are based on Ts'amakko. There are three causative extensions, -is and -as, bor­ rowed from Ts'amakko, and -san: (18a)

gutaldance "to dance"

(18b)

gutal-isdance-CAUS "to make dance'

(19a)

'adlick "to lick"

(19b)

'ad-aslick-CAUS "to make lick"

(20a)

muxilaugh "to laugh"

(20b)

mux-sanlaugh-CAUS "to make laugh'

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

181

The middle extension suffix is -i '. This is a frozen ending in several lexicalised verbs: (21a)

ho batwash "to wash"

(21b)

hobat-i'wash-MIDDLE "to wash o.s."

(22a)

(*mux-)

(22b)

mux-i'laugh-MIDDLE "to laugh"

The intransitive extension -am comes from the identical Ts'amakko passive extension, which is also very common in East Cushitic: (23a)

bulpull "to pull out"

(23b)

bul-ampull-INTRANS "to be pulled out"

(24a)

laxmix "to mix" (tr.)

(24b)

lax-ammix-INTRANS "to mix" (intr.)

Some verbs seem to show frozen suffixes with a dental ending in their basic stem: (25 a)

'ang-at"to mix" (tr.)

(25b)

hob-at"to wash"

(25c)

noq-ot"to look, aim at"

182

GRAZIANO SAVA

Syntax - like phonology - is very much influenced by Ts'amakko. This is proven by the following syntactic rules, which all have parallels in Ts'amakko. The position of the modifier is after the head: (26a)

inta attomuni man white "white man"

(26b)

hhanca gollanke tree nine "nine trees"

No verb "to be" appears in nominal sentences: (27)

kata 'ongotitta 1SG ongota "I am Ongota"

Adverbs can precede or follow an emphatic personal pronoun: (28 a)

barama kata ka-qafî tomorrow I will fish "I will fish tomorrow"

(28b)

kata barama ka-qafi I tomorrow will fish "I will fish tomorrow"

A specific characteristic of Ongota syntax is that no particle can separate the subject clitic from the verb stem, except the negative ma. Half of the Ongota lexicon can be claimed to be of Ts'amakko origin. The verbal lexicon seems to be more Ongota than the rest of the lexicon. Table 5 shows a list of some original Ongota roots 4

Classification There is a debate about the classification of Ongota. Many affiliations have been proposed. On the one hand this is because of lack of data, not only from Ongota but also from neighbouring languages. On the other hand comparative study in the relevant language groups and on the Ongota data is also insuffi­ cient. Fleming et a1. (1992-93) consider Ongota in itself a primary branch of Afroasiatic. According to Ehret (p.c.), Ongota represents a separate branch within South Omotic. Bender (p.c.) thinks that it is a hybridised Cushitic lan­ guage. Aklilu Yilma (p.c.) proposes a process of creolization in the formation of Ongota basing himself on the paucity of morphology (in particular the lack of plural formation) and the lack of (segmental) tense marking. The myth of the multhi-etnic origin provides additional non-linguistic arguments in favour of

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

abba

"good, beautiful, well"

gutal-

183

"to jump"

ali-

"to tell"

hhaabini-

"to want"

aka

"foot; leg"

kalbano

"one"

axaco

"sun"

kimiša

"crocodile"

binta

"animal"

kiti

"road"

burinki

"today"

ista

"eight"

carkamuni

"green"

oxoni

"fire"

­

"to eat"

roomini

"red"

'aw

"to drink"

siinsad-

"to smell"

ca 'awa

"water, river"

taq-

"to sleep"

cata

"meat"

tsanafa

"six"

d'akkamuni

"black"

wura

"house"

gabare

"snake" Table 5 : Original Ongota roots

this idea. In our opinion, however, the historical value of this myth is doubtful because the names of the clans generated by each immigrant group are also names of Ts'amakko clans. This shows at least that also in the oral history there is Ts'amakko influence. In this paper we do not propose a classification of Ongota, but only some elements for discussion. We assume that Ongota has original elements still surviving mostly in its lexicon and morphology and that these are the elements to be taken into consideration for the classification. Analysing the lexicon, the specific Ongota lexical roots are so unlike any of the other languages in the region that no common origin with other languages can be indicated. If we consider the structural presence of preverbal subject markers as original, this would mean that Ongota forms a branch of Lowland East Cushitic, since this is the strongest common feature that justifies the existence of this subgroup of East Cushitic (Tosco 2000). The idea of a creolised Ongota gets credit from the paucity of verbal and nominal morphology. However, Ongota has more mor­ phology than is usual in creoles, which again could be accounted for by the strong pressure exercised by Ts'amakko. Another explanation of the paucity of morphology is the attrition of this language which has also favoured the ab­ sorption of Ts'amakko elements. One could argue that Ongota has become a Cushitic language of the Dullay group. It would be interesting to compare Ongota to two Cushitic languages, Yaaku (Mogododo) and Elmolo, which

184

GRAZIANO SAVA

have become extinct in recent years (Heine 1973; 1974-75). The origin of Ongota remains undetermined under the heavy influence from a Cushitic lan­ guage that is taking over. 5 Ongota, a dying language The data in this article are the result of 10 days of intensive elicitation dur­ ing which we have met several kinds of difficulties, such as many irregulari­ ties, inconstancies, and ambiguities. These are clearly problems of lack of stability in a moribund language which is not transmitted anymore, and which is used in extremely limited contexts and not at all as the main means of com­ munication in a community of speakers. Ongota survives in fixed forms af­ fected by slow obsolescence in the minds of the few who speak it. In addition to the amount of irregularity in its original forms, Ongota is heavily influenced by the language which is completely replacing it, Ts'amakko. Ongota does not develop itself independently anymore and the Ts'amakko borrowings are more and more absorbed without any adaptation (cf. the adjectival forms maintaining the Ts'amakko suffixes). We cannot consider these mere cases of code switching: the elements of Ts'amakko origin are Ongota and have been absorbed by Ongota in a massive way. The usual situation that phonology and syntax are the most affected in con­ tact situations is confirmed in this case as well. Lexicon, as can be expected, is the second most influenced level in Ongota. However, a lot of basic and nonbasic lexicon is still of Ongota origin. Ongota morphology shows most inter­ esting and unique Ongota peculiarities, even if it, too, clearly converges to Ts'amakko. Among the presumably original Ongota morphological elements are: the forms of the preverbal subject markers and pronouns, the tonal realisa­ tion of tense, the adjectival and some of the verbal derivations. These elements presumably belong to the original Ongota once spoken as a first language and will remain there until they will die with the last speakers. 6 Levels of fluency We have reported that Ongota is spoken by eight people. However, we should note that they speak different kinds of Ongota, with different levels of regularity and of independence from Ts'amakko. Comparing the results of the elicitation with Mole and the interviews with the first three Ongota speakers met in Weyt'o, it was clear that the instability and influence of Ts'amakko is less in Mole's Ongota. The qualitative level of Ongota speech is clearly related to the wish to maintain Ongota. Mole, being an elder, charismatic and influen­ tial person, is proud of his being Ongota, and he is more prepared to associate himself to a separate, albeit despised, linguistic identity.

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

185

We have been told that there are also four women who have some knowl­ edge of Ongota. We could not interview them, but according to Mole, since they live in the Ts'amakko area with Ts'amakko husbands, they speak an extremely weak Ongota which absorbed a large amount of Ts'amakko ele­ ments. It is to be expected that other Ongota can speak and understand some On­ gota. Some people may be able to understand it, but are not willing to do so, since it is a language they do not use and do not want to use. 7 Arresting the extinction process? The ideal solution would be not only to describe but also to save Ongota. Unfortunately, this is utopia. Although the Government of Ethiopia is presently founded on the concept of ethnic federalism, it has not created a genuine inter­ est for ethnic variety among the Ethiopians. In particular the educated people, the students, and in general those who are more or less involved in the process of economic development have changed their life style towards a more west­ ernised one and they usually despise rural, wild and "uncivilized" peoples such as the Ongota. This attitude can be attributed to a sense of superiority that is in particular shown by those Ethiopians belonging to the two ethnic groups that have always ruled the country, the Amhara and Tigreans. Arresting this process of extinction would mean starting a difficult project of diffusion of the lan­ guage among people who have decided to give it up. If a new prestigious status could be achieved by the Ongota people maybe they could rethink the decision to replace their traditional language, keeping it as a mark of their new positive identity. However, a new social status for the Ongota can neither be realized nor planned. Moreover, the government, which is also interested in the eco­ nomic development of the South, seem to be absolutely insensitive to the dam­ ages made to the Ongota people by the Birale Cotton Company (see above). It owns 50% of this farm but it is not intending to modify the system of irrigation which is causing so many problems to Ongota economy and subsistence. This is another index of lack of general interest in the Ongota's life. Paradoxically, the farm itself could employ Ongota workers, who could thus effectively change their status, with possible beneficial effects on the condition of the Ongota language. However, attempts in this direction made in the past failed because, according to Mole, no Ongota could stand such a radical change in his/her life. Let us imagine that for some reason Ufa 'ongota will regain a first language status and that the language taught to the children will be the one Mole speaks

186

GRAZIANO SAVA

which we have documented. What kind of Ongota would it be?8. Of course it would not be pure Ongota since most of its structure has been replaced and/or lost. So, the children will speak a mixed language as first language. Will they be considered proper Ongota speakers or semi-speakers, using a common, but infelicitous expression? This thought experiment is to prove that dealing with Ongota means dealing with a language that no longer exists. The real Ongota cannot be recovered because a good part of it has disappeared forever. 8 Documentation plans What we can do is make the best description possible of Ongota before the last speaker passes away. However, as far as we know, at this moment there is no project to describe Ongota. Aklilu Yilma seems to be the only researcher who visits the area, but for sociolinguistic purposes. The importance and the many questions related to the enigma of Ongota deserve a full descriptive study. There are two sketches, Fleming et al. (1992-93) and Tosco & Savà (2000), which could be taken as a good starting point for a grammar of Ongota. Sensible elicitation and analysis of Ongota requires a sound knowledge of Ts'amakko and Amharic. References Amborn, Hermann, Gunter Minker & Hans-Jürgen Sasse. 1980. Das Dullay: Materialen zu einer ostkuschitischen Sprachgruppe. Berlin: Dietrich Reimer (Kölner Beiträge zur Afrikanistik, 6). Central Statistic Authority. 1998. The 1994 population and housing census of Ethiopia. Summary reports at country and regional levels. Addis Ababa: Office of Population and Housing Census Commission, Central Statistical Authority. Dinota, Kusia & Ralph Siebert. 1994. "Wordlists of Arbore (Irbore), Birayle (Ongota), Tsamai (Tsamaho)". Survey of Little-known Languages of Ethiopia Linguistic Report 20. 1-12 Donaldson Smith, Arthur. 1896. "Expedition Through Somaliland to Lake Rudolf'. The Geographical Journal 8. 120-137, 221-239. Federal Democratic Republic of Ethiopia. 1994. The Constitution of the Fed­ eral Democratic Republic of Ethiopia (unofficial English translation from the Amharic original). Addis Ababa: Bole Printing Enterprise. 8 This reflection has been generated by the question "What kind of Ongota did you describe?" made to the author by Dr. Felix Ameka in the discussion of the paper "Ts'amakko grammatical borrowings in Ongota". This paper was presented at the Symposium on Ethiopian Morphosyntax in an Areal Perspective (EMIAP) held in Leiden from 4 to 5 February 2001. The author and Christian Rapold are editing the articles for the proceedings which will be published in the Afrikanistische Arbeitspapiere.

ONGOTA, A MORIBUND LANGUAGE OF SOUTHWEST ETHIOPIA

187

Fleming, Harold C , Aklilu Yilma, Ayyalew Mitiku, Richard Hayward, Yukio Miyawaki, Pavel Mikes, & J. Michael Seelig. 1992-93. "Ongota or Birale: a moribund language of Gemu-Gofa (Ethiopia)". Journal of Afroasiatic Languages 3:3. 181 -225. Fukui, Katsuyoshi, Eisei Kurimoto & Masayoshi Shigeta, eds. 1997. Ethiopia in broader perspective: papers of the thirteenth international confer­ ence of Ethiopian studies (ICES XIII), Kyoto, 12-17 December 1997. Kyoto: Shokado. Grimes, Barbara F., ed. 1991. Ethnologue. Languages of the world. 11th ed. Dallas, Texas: Summer Institute of Linguistics. —. 1996 Ethnologue. Languages of the world. 13th ed. Dallas, Texas: Summer Institute of Linguistics. Hayward, Richard J. 1989. "Comparative notes on the language of the S'aamakko". Journal of Afroasiatic Languages 2:1. 1-53. Heine, Bernd. 1973. "Vokabulare ostafrikanischer Restsprachen. Teil 1. Elmolo". Afrika und Übersee 56. 276-283. —. 1974-75. "Notes on the Yaaku language (Kenya)" Afrika und Übersee 58. 27-61, 119-138. Melesse Getu. 1995. Tsamako women's roles and status in agro-pastoral production. Social anthropology dissertation series 3, Department of Sociology, Anthropology and Social Administration. Addis Ababa: University Printing Press. —. 1997. "Local versus outsider forms of natural resources use and manage­ ment: the Tsamako experience in Southwest Ethiopia". Fukui, Kuri­ moto & Shigeta. 1997. Vol. II, 748-767. Savà, Graziano & Mauro Tosco. 2000. "A sketch of Ongota: a dying language of Southwest Ethiopia". Studies in African Linguistics 29:2. 57-135. Tosco, Mauro. 2000. "Cushitic overview". Journal of Ethiopian Studies 33:2. 87-121.

AN ENDANGERED LANGUAGE THE GÙRDÙD LANGUAGE OF THE SOUTHERN BAUCHI AREA, NIGERIA*

ANDREW HARUNA Institute of African Studies, Humboldt University, Berlin

"I am always sorry when any language is lost, because languages are the pedigree of nations" (Samuel Johnson in J. Boswell, Journal of a tour to the Hebrides, 1786)

1 Introduction The death of a language is always distressing, since the cultural tradition connected to it, and the socio-cultural or even ethnic independence of the group that speaks it, very often perish with it. Yet it is a common phenomenon, the world over. Here we are concerned with the Southern Bauchi Area (henceforth SBA) of Northern Nigeria. The SBA contains a large number of languages and small ethnic groups which belong to different language sub-families. The Gùrdùη language which is located in this area is in the process of being re­ placed by Hausa, another Chadic language. The process might be said to have started in the early 18th century and it is probable that in the next few genera­ tions this language will completely disappear as a distinct linguistic form. As a contribution to studies in dying languages, this paper surveys the social factors that have contributed and still contribute to the Gurdurj people adopting the language of another group in place of their own.1 Sasse (1992:10) suggests three factors in the study of language death: (1) External Setting: i.e. extralinguistic factors (cultural, sociological, ethnohistorical, economic, etc.), which create a situation of pressure in a speech community, forcing it to give up its language, (2) Speech Behaviour; i.e. variables, which, in a given speech comI am very grateful to Joe McIntyre for comments on an earlier draft and for correcting the language of the paper. In keeping to tradition, he owes no responsibility for any error. 1 The Gùrdùrj case study, in a strict sense, is what might be called "language replacement or shift". This is because the descendants of members of the original speech community still exist, but they speak a replacing language (Hausa), which has superseded their mother tongue after an intermediate period of bilingual usage.

190

ANDREW HARUNA

munity, are bound to social parameters, e.g. the use of different languages in multilingual settings, the use of different styles of one language, domains of languages and styles, attitudes towards variants of language, and (3) Structural Consequences: i.e. a set of linguistic phenomena, e.g. changes in the phonol­ ogy, morphology, syntax and lexicon of the language threatened. I shall con­ sider only the first and second factors. The article will therefore report findings from a survey I made during fieldwork. Interviews were done with fluent (and semi-) speakers of the language whose death means the death of the language. Oral stories gathered have been compared with written historical events in and around the region of the SBA. I have been working on the languages of SBA since 1991 as part of the study of the minority languages of the area. My research has not been primarily concerned with language replacement/death, but it is not possible to study these languages today without acquiring some information on this topic. Many communities in the SBA are experiencing a shift from the predomi­ nant use of their particular language to a more widespread acceptance of Hausa. An important question facing the linguist working in this area today is whether this kind of language shift may be expected to lead ultimately to the loss of the language(s) of these speech communities. Lamenting on the lan­ guage situation of the SBA area, Shimizu (1978:8) said that, "Hausa is rapidly replacing the indigenous languages". Thus, he ominously predicted: In a generation or two, more or less sixty distinct speech communities will be merged together and ... their speech forms will become extinct. (Shimizu 1978:8).

Before coming to the main subject of this paper, it is necessary to make a few remarks about the linguistic classification and the socio-linguistic situation of the SBA. 2 History and present situation of the Gùrdùη language 2.1 Chadic internal classification According to Jungraithmayr & Leger (1993:162), the Chadic language family contains ca. 150 languages spoken in the area west, south and east of Lake Chad2, mainly in northern Nigeria and Cameroon, central Chad and the surrounding areas. Hausa is by far the largest and the most widely spoken 2

There is no agreement among scholars over the exact number of Chadic languages. Newman (1987) mentioned that the family consists of ca. 135 languages and in another publication, (1990:1) he put the number at 140. The reason for this disagreement is that new languages continue to be discovered and the status of a language or dialect is not established for some of the languages.

THE GÙRDÙ LANGUAGE OF THE SOUTHERN AUCHI AREA

191

Chadic language3. Apart from a few medium size languages most of the other languages are rather small and spoken by a few hundred to a few thousand speakers. The term "Chad" languages was first introduced by Greenberg in the 1950s to mean a family of genetically related languages developed from a common parent (Greenberg 1950:51). Greenberg arrived at this conclusion based on the recognition of lexical items and grammatical morphemes sounding alike. The basic principle of Greenberg's method is that of "mass comparison" of basic vocabulary. The next step towards proving the genetic relationship of the Chadic languages was undertaken by Newman and Ma (henceforth N/M 1966). In the words of the authors, their aim was to "demonstrate conclusively that the Chad family as postulated by Greenberg does indeed constitute a valid linguis­ tic unit" (N/M 1966:219). N/M merged Greenberg's groups 1 and 9 into a single major group, which they called "Plateau-Sahel", and his groups 3 and 6 into another group which they called "Biu-Mandara"; the other groups were left unclassified. The proof consists of the establishment of regular phonological correspondences between the two major divisions within Chadic and of the subsequent reconstruction of 144 proto-Chadic lexical items. Hoff­ mann (1971) merged Greenberg's groups 2, 4, 5, 7 and 8 into one single branch "Biu-Mandara". In another article, Newman (henceforth NM 1977) revised his and his colleague's earlier classification (N/M 1966) of the Chadic languages and the revised version consists of four branches (the fourth being provisional), namely, "West", "Biu-Mandara" or "Central", "East" and "Masa"4, each of these branches containing sub-branches, etc. These groupings according to him demonstrated "a greater and more precise internal structure (of Chadic) than provided in previous classification" (NM 1977:3). The development and different steps in the classificatory treatment of the internal subdivision of languages which belong to the Chadic family reflects the ongoing discussion on this matter, a discussion which indicates that the classification still does not have a final form5. Table 1 gives NM's (1990)

3

The demographic importance of Hausa is seen in the degree to which it has become an object of scholarly investigation. 4 One would think that by now the classification of Masa within Chadic would no longer be an issue, but this is not the case (Newman 1990; Tourneux 1990). 5 The development of the dichotomous and trichotomous views in the classification of Chadic languages quoted by Jungraithmayr (1991:62) may be summarized as follows: Dichotomic classification: Lukas (1934): "Chadohamitic" vs "Chadic" ("Mandara"); Newman & Ma (1966): "PlateauSahel" vs "Biu-Mandara"; Jungraithmayr (1981): "West-East" vs "Central". Trichotomic classification:

192

ANDREW HARUNA

classification of the Chadic family (simplified, with only a few representative languages listed)6. West Chadic

Central Chadic (Biu-Mandara)

East Chadic

A

A

A

Hausa

Bade

Bole

Bubburè 7

Angas

Gùrdùη

Kotoko

Somrai

(Masa?)

Dangla

Masa

Table 1 : The Chadic Family

2.2 Linguistic situation of the Southern Bauchi Area The SBA is unique because it is a meeting point of the three great African linguistic phyla found in Nigeria; the Afroasiatic (Hamito-Semitic) phylum, namely Chadic, e.g. Hausa and its nearest relatives; the Nilo-Saharan phylum, e.g. Kanuri; and the Niger-Congo phylum, represented by Benue-Congo lan­ guages, e.g. Jukun (Jukunoid), Jàar and Laßur (Bantoid). These groups are intermingled because they have had some kind of contact over a long period of time. The various contacts - peaceful and violent - took place in a region of Nigeria which, by virtue of its geography and topography, became unique in the country. As to the period of arrival of the various linguistic phyla in this area, there are good reasons to believe that it was many centuries ago (James 1987; Aliyu 1986; Adamu 1978, 1984). The speakers of Benue-Congo lan­ guages are believed to have dominated the whole area before the arrival of Chadic-speaking peoples. It was earlier hypothesised by Lukas (1936a; 1936b; 1939) that the Chadic-speaking peoples of the SBA reached their present areas by westward migration from the east. If this proposition is correct, then one might even ask with Hoffmann (1970:7): Whether the territory presently occupied by the Western Chadic languages might not partly or wholly have been Benue-Congo territory before the advent of the Chadic speaking peoples (and if such a view is valid which of course looks convincing), it would also help to understand a little better the present distribution of tribes and espe­ cially the high degree of interpenetration that can be observed in the area of the Jos and Bauchi.

Newman (1977): "West" vs "East" vs "Biu-Mandara" (vs "Masa"); Jungraithmayr & Shimizu (1981): "West" vs "East" vs "Central". See NM (1980) for further discussion. 6 For a complete classification the reader is referred to NM (1990). 7 Provisional language addition by Haruna (1993).

THE GÙRDÙ

LANGUAGE OF THE SOUTHERN

AUCHI AREA

193

2.3 The Gùrdù ) language 2.3.1 Classification Gùrdùrj8 belongs to the Saya (Hoffmann 1971 used the term "Zaar") "Southern Bauchi" group of the West Chadic-B sub-branch of Chadic (Table 1 above9. It is among more than a dozen languages of Chadic-speaking ethnic groups found in the south-eastern part of Bauchi State. The language is called byàu-gùrdùr "mouth of Gùrdùrj" by its speakers (and in Hausa is usually re­ ferred to as gùrùntùm); the village of the Gùrdùrj people is called Gùrdùrj. A speaker of Gùrdùη is called in the language mu-gùrdùη "Gùrdùη man/person". 2.3.2 Linguistic works on Gùrdùη While information on Southern Bauchi languages has increased in the last few years, our knowledge of Gùrdùrj is still limited. To the best of my knowl­ edge there is as yet no major linguistic research published on the language, only a few comparative studies, a few linguistic notes with some grammatical remarks and some unpublished field notes. Vocabulary lists have been col­ lected by Gowers (1907), Schuh (1978), Shimizu (1978), Kraft (1981) and Cosper (1994); further studies have been undertaken by Haruna (1981; 1993; 1996; 1997; 1998; forthcoming) and Jaggar (1988; 1991). Apart from my own collections (which are yet to be transcribed and published) Gùrdùrj oral litera­ ture is still unrecorded. 2.3.3 The language and its dialects There are variations in the speech forms of some of the villages but there is no difficulty in mutual understanding: people from one area can converse freely in any other. Very roughly, we can distinguish two major dialect areas. In a previous work (Haruna 1981), I named them after the main villages in and around which they are spoken, namely, the G/Kùukù and G/Gayar dialects. There are notable dialectal and stylistic differences (not treated here, but see Haruna forthcoming) between the two major dialect areas which are mainly phonological and lexical in nature. It is not unusual to come across speakers of G/Kùukù using the form of G/Gayar but not the other way round. However, I am aware of the fact, also attested by the speakers themselves, that G/Gayàr is the more conservative dialect and probably represents an older stage of the 8

Transcription system: Long vowel is indicated by double letters, low = â/âa, rising tone = ä/äa, high tone is left unmarked. 9 The Gùrdùrj people themselves tend to classify their language neighbouring Bùgàlàmbu language (another West Chadic-B language) more closely with the Bùgàlàmbu people along the borders and valleys

tone = à/àa, falling tone as being closer to the and identify themselves of the Gongola river.

194

ANDREW HARUNA

language. Nevertheless, the following points need to be borne in mind. First, no thorough dialectical survey has been carried out, the data being based on a spot check of speakers from various places in various villages. The second point follows from the first: the list is idealised in that not all speakers in the two dialect areas necessarily have the features ascribed to these areas in the exam­ ples found. Nonetheless, taking all these facts into consideration, it is clear that Gùrdùr is a single language with at least two dialects. 2.3.4 Location and statistics Although for the major part of their history the Gùrdùrj were a linguistic unit in a common political entity, today they occupy parts of three different presentday states of northern Nigeria. The majority of Gùrdùrj speakers live in several villages in the Pàali and Dùgùrì10 districts of the Al àalèeri Local Government Authority and in Gàlàmbi district of Bauchi LGA. Different villages, e.g. Dookà, Gayar etc., are referred to by the name of the language, e.g. G/Dooka, G/Gayar etc. (Haruna 2001 forthcoming). Statistics giving the exact number of speakers are lacking. The Al àalèeri Local Government office "Community Tax Book of 1978/79" estimates the number of Gùrdùrj in the LGA at 10,600 (Haruna 1981:1). More recently, in the national census of 1991, the Nigeria National Population Commission (NNPC) gives the location of the people without regard to the actual number of speakers of the languages spoken by these communities. However, recent research has revealed that the number is twice as many as in 1979 (Haruna forthcoming). My informants gave me the following villages (in alphabetical order): G/Bàadàra, G/Bwagwas, G/Gàji, G/Gayàr, G/Dookà, G/Duutsè, G/Kùukù, G/Gyà la , G/Kàrà-Kara, G/Jaalingôo. G/Kindir, G/Màrshàn, G/Mbaarù, G/Tora and G/'Yoolà. The villages in district can be identified on the map11. In the diaspora, Gùrdùrj people are also found in the Wase district of Plateau State and in some three villages in Taraba State. I obtained the following names of villages in Wase district of Plateau State (in alphabetical order): Bwagwayam, Dànkàm, Gwama -Kôogi, Jigaawa, Kambari, Lambà, Li gà, Mboorat and Tattumi. The villages in Taraba State are: Ceediya, Do ga and N m.

10

In Dùgùri district, the people are called gùrdù -mbaarù "people of the hill". The towns indicated on the map are intended to give the reader an approximation of the geographical distribution of the Gùrdùrj speaking people. 11

THE GÙRDÙ LANGUAGE OF THE SOUTHERN BAUCHI AREA

Alkaleri Local Government Area

195

196

ANDREW HARUNA

The Gùrdùrj living in the Wase district of Plateau and in Tàrabà States are among the least known of all the groups. Their number is still a mystery12. The younger generation of Gùrdùrj speakers in Pàali and Dùgùrì districts isprobably not aware of these sizeable Gùrdùrj populations. Nonetheless, their presence in these areas goes back at least several hundred years, the people having mi­ grated primarily from G/Karakara. In 1999, the civilian Government of Plateau State appointed a Gùrdùrj man to its cabinet. One Gùrdùrj man is also an elected Councillor in Al àalèeri LGA. In spite of the division of these peoples effected by developments in pres­ ent-day northern Nigeria, there is still a great deal of contact among them through cultural events, religious bonds, intermarriage and the undeniable bond of a common language and history, the language thriving today through the formation of the Fauwala association (see section 4). The neighbours of the Gùrdùrj in Pàali and Dùgùri districts are speakers of Jàar (and its various dia­ lect forms) and of Laßur (both Benue-Congo), Bùgàlàmbu (Chadic). There are also Hausa, Fulani and Kanuri settlers. 2.4 The Gùrdùrj people and their history 2.4.1 Migration history The origins of the Gùrdùrj people are still very unclear and there is no writ­ ten history. Thus oral accounts are supplemented by written sources gathered in the SBA (James 1987; Aliyu 1972; 1973; 1986). Oral accounts as told by elders suggest that the ancestors of the Gùrdùrj people were once Fulani who came from the east, probably around Borno13. In some accounts it is believed that, from the east, the people moved westward to Mallamawa (a village on the Kano border) and then later migrated southward to settle in Miiri (a village on the northern outskirts of Bauchi). At Miiri enmity arose between the two sons of the chief which resulted in the departure of the eldest son to an area south­ east of Bauchi town near Kàngèrè called G/Kàràkara (meaning Gùrdù -low land) in the Gàlàmbi district of Bauchi LGA. The town grew bigger as the people increased in number. After many years in this place, another power tussle arose between the sons of the Chief. One of the sons called Bùrùm left 12 Unfortunately, due to time constraints, it was not possible for me to conduct additional research during my fieldwork in these areas. Thus, many questions and the statistics of Gùrdùrj speakers living in Wase district of Plateau State and those in Taraba State will have to remain unanswered until there is an opportunity to conduct further investigation. The research area is therefore limited only to Gùrdùrj in Pàali and Dùgùri districts. 13 All the old people I interviewed strongly believe that there is a special historical relationship between them and the Fulani. When I asked for linguistic proof, I was referred to the word "come" which is wàri in Gùrdù and warù in Fulfulde. This is the only example I found.

THE GÙRDÙ LANGUAGE OF THE SOUTHERN BAUCHI AREA

197

with his supporters to settle in the surrounding mountains14. Bùrùm named their new settlement after himself as Bàrùmtùm which was later renamed by the Hausa as Gùrùntùm-Kùukù (in Gùrdùrj: G/Kùukù), Gùrùm (< gùrgùzà = "rocks" ) and Kùukù (