The Vanishing Languages of the Pacific Rim

The Vanishing Languages of the Pacific Rim This page intentionally left blank The Vanishing Languages of the Pacific ...

Author: Osahito Miyaoka | Osamu Sakiyama | Michael E. Krauss

943 downloads 2302 Views 6MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

The Vanishing Languages of the Pacific Rim

This page intentionally left blank

The Vanishing Languages of the Pacific Rim Edited by

Osahito Miyaoka Osamu Sakiyama and Michael E. Krauss

1

3

Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © 2007 editorial matter and organization Osahito Miyaoka, Osamu Sakiyama, and Michael E. Krauss © 2007 the chapters their various authors The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2007 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by SPI Publisher Services, Pondicherry, India Printed in Great Britain on acid-free paper by Biddles Ltd., www.biddles.co.uk ISBN 978-019-926662-3(hbk) 1 3 5 7 9 1 0 8 6 4 2

Contents

The Background to This Volume

vii

PART I Diversity, Endangerment, and Documentation 1 Keynote—Mass Language Extinction and Documentation: The Race against Time

3

Michael E. Krauss

2 Documenting and/or Preserving Endangered Languages

25

Bernard Comrie

3 Linguistic Fieldwork among Speakers of Endangered Languages

35

Colette Grinevald

4 Language Policy and Language Rights

77

David Bradley

5 Using Written Records to Revitalize North American Languages

91

Toshihide Nakayama

6 Indigenous Voices and the Linguistics of Language Revitalization

107

Marcellino Berardo and Akira Y. Yamamoto

7 Pidgins and Creoles in the Pacific

118

Sabine Ehrhart and Peter Mühlhäusler

8 Linguistic Diversity in Decline: A Functional View

144

Osahito Miyaoka

PART II Areal Surveys The South Pacific Rim 9 Languages of Middle America

165

Yoshiho Yasugi

10 Languages of the Pacific Coast of South America

183

Alexandra Y. Aikhenvald

11 Fuegian Languages Oscar E. Aguilera

206

vi

contents

12 Indigenous Languages of Australia

221

Michael Walsh

13 Languages of New Guinea

239

Alexandra Y. Aikhenvald and Tonya N. Stebbins

14 Languages of the Pacific Region: Malayo-Polynesian

267

Osamu Sakiyama

South-East Asia 15 Indigenous Languages of Formosa

285

Naomi Tsukida and Shigeru Tsuchida

16 Languages of Mainland South-East Asia

301

David Bradley

17 Minority Languages of China

337

Dory Poa and Randy J. LaPolla

18 Japanese Dialects and Ryukyuan

355

Shinji Sanada and Yukio Uemura

The Northern Pacific Rim 19 Nivkh and Ainu

371

Hiroshi Nakagawa and Osami Okuda

20 Siberia: Tungusic and Palaeosiberian

387

Toshiro Tsumagari, Megumi Kurebito, and Fubito Endo

21 Native Languages of Alaska

406

Michael E. Krauss

22 Languages of the Northwest Coast

418

Honoré Watanabe and Fumiko Sasama

23 Languages of California

442

Leanne Hinton

24 Languages of the South-West United States

460

Kumiko Ichihashi-Nakayama, Yukihiro Yumitani, and Akira Y. Yamamoto The Authors Index of Subjects Index of Place Names Index of Names Index of Languages and Language Families

475 485 503 507 515

The Background to This Volume

Vanishing languages in the world

Deterioration of the physical environment and the biological ecosystem is proceeding rapidly on a global scale. There is hardly any area on earth that is pristine and not spoiled by civilization. Day after day we are made aware of the pollution of the land, sea, and air, the depletion of natural resources, the excess of waste materials, and the extinction of species of animals and plants, only to recognize begrudgingly that the physical and global environment (all through the Anökumene and even the ozone layer) are fragile, rapidly changing, and already in grave danger. The other contemporary issue that we may safely assume is directly connected with environmental deterioration and diminishing biological diversity is linguistic extinction, which is threatening an ever-increasing number of unique and traditional languages developed by the world’s peoples from hoary antiquity. This extinction is accelerating the loss of linguistic diversity in the face of incomparably slow processes of diversification as well as rediversification (e.g. into ‘Englishes’) and creolization (like Tok Pisin), which yield new languages. Some advocates have taken linguistic diversity to be inextricably linked to biological diversity, with each constituting a system in itself. Some observers, particularly in moist tropical regions (including Melanesia and southwest Asia), taking both linguistic and biological diversity as greatest near the equator and decreasing as one moves away from it, have suggested a direct correlation between the two diversities and even their “coevolution”. Depending upon how one grasps linguistic diversity, however, the correlation could be put into a caveat, since many linguists would believe that diversity should properly refer to deep genetic or “phylogenetic” divergence (cleavage) coupled with structural difference rather than simply to the sheer number of languages with less than moderate depth and minor differences and that value of linguistic diversity lies in the weight of time which the diversification has taken.

viii

t h e bac kg ro u n d to t h i s vo lu m e

As far as the diminishing phases of these diversities are concerned, they seem more positively correlated through the human and societal factors involved which came to the fore in the post-Columbian era and with everdeveloping technology. It would be more readily accepted that the humancentred deterioration of biological diversity on one hand and the diminution of linguistic diversity on the other should spring from a common soil and move in tandem through the socioeconomic (rather than ecological) mechanisms. It is now a familiar scene in countless areas around the world that people from civilized and industrialized countries have intruded—either through expeditions, trade, evangelization, or colonization—into the lands which indigenous peoples cherished and coexisted with for many centuries, forcedly expropriating the nature and recklessly exploiting its resources, while suppressing their traditional languages and cultures side by side. The same root notwithstanding, the issue of environmental deterioration together with the loss of biological diversity would much more easily come home to us as sources of anxiety or crises, while the loss of linguistic diversity seems much less likely to be perceived as something dire, or is perceived, if ever, in different—and often opposite—ways among people, barely penetrating public awareness in general. Linguists, however, started to organize around the growing concern about worldwide linguistic diversity and language endangerment in the late 1980s—especially under UNESCO, which had gradually developed a programme on languages from the 1960s (see Aikawa 2004). The Comité International Permanent de Linguistes (CIPL) has been taking a great interest in this highly topical worldwide problem. At the 1991 Linguistic Society of America (LSA) meeting, an incisive claim was made by Michael E. Krauss, our co-editor, in Krauss (1992), which has turned out to be probably one of the best-cited articles on language endangerment—see his Keynote Chapter 1 (this volume, based on Krauss 2001a) as well. As the International Decade of the World’s Indigenous People (1995–2004) also spurred stronger interest in the issue of endangered languages, together with the urging of UNESCO in particular, concrete actions started in various areas around the world, with conferences and symposia mushrooming all around. UNESCO’s Atlas (Wurm 2001) includes a concise but persuasive exposition about endangered language issues in the world, while important opinions on the issue from a global perspective have since been voiced by, to cite just a few, Dixon (1997), Crystal (2000), and Nettle and Romaine (2000), as well as many contributors in edited works and in conferences, reflecting varied linguistic, sociolinguistic, and language-policy points of view about


ix

the mechanism by which languages become endangered, the revival of endangered languages, and what researchers should be doing about the situation. The most recent (fifteenth) edition of the Ethnologue (edited by Gordon (2005), replacing the fourteenth by Grimes (2000)) catalogues all of the world’s 6,912 known living languages, 516 of which are “nearly extinct” (i.e. moribund) languages at the far end of the endangerment scale, so classified when “the speaker population is fewer than 50 or when the number of speakers is a very small fraction of the ethnic group”. Although the decline and extinction of languages is a worldwide phenomenon, both the UNESCO Atlas and the Ethnologue show that the distribution of endangered languages is not geographically even across the globe. Areal survey of endangered languages in the world was attempted as early as 1991 by Robins and Uhlenbeck, who concluded the preface to their work with the words “in the last decade of the twentieth century there can be an upsurge of descriptive activity in many countries” (1991: p. xiv). There have since appeared a number of surveys of the actual situations in particular areas of the world, for instance, Brenzinger (1998) on the situation in Africa, Shoji and Janhunen (1997) on the northern minorities which includes Krauss (1997) on the North Pacific Rim (though only from the Columbia River on the American side to the Ussuri River on the Asian side). Given the very rapid changes from which minority languages of the world have suffered, the decade and a half since Robins and Uhlenbeck (1991) is far from an insignificantly short period of time. A new areal survey on endangered languages, which gathers and systematizes accurate information on the situation, would provide a benchmark for our work in the coming decades. The present book is centred upon the Pacific region where the density of endangered languages is the highest in the world. The Pacific as it stands

The “Pacific” in this volume, not being coterminous with its geographical or oceanographical concept nor with the Ethnologue, encompasses not only Oceania (Island Pacific together with Australia and New Zealand), but also its bordering Pacific Rim areas (of Asia and the Americas) with the (sea) coast and its fringe area inland (or the “first tier” inland; see Krauss 2003: 211). Historically, a great part of the region was home to aboriginal peoples living there before the Mongoloid migrations as well as the crisscross of their migratory routes since prehistoric times. It is assumed on the basis of their linguistic interrelationships and from linguistic history, though the documentary evidence is very meagre, that numerous languages were born and

x


died out here. The intriguing hypothesis recently proposed by Nichols (1995), concerning the spread of language around the Pacific Rim on the basis of the distribution of what she regarded as diagnostic structural features in language, would deserve to be well corroborated. Given so vast a region on earth, with a multitude of languages, it is not surprising that the Pacific does not make up a linguistic world of its own, distinct from other areas. It constitutes, however, a whole that has evolved a complex pattern of diversification into, movement of, and interaction (mixture and shift) among languages or linguistic families since prehistoric times, and, importantly, still remains an area which, unlike elsewhere in the world, has not been levelled out by a single (or few) very widespread linguistic family or structural type such as Indo-European or “Altaic” languages. Campbell (1997), leaving aside the (macro)phylophiliac “lumping”, counts about 150 genetic units, i.e. independent language families and isolates (the only known survivors of a family), in the New World (about 60 in North America), with a genetic diversity of the same range as that of the Americas being found in the Pacific Ocean and Australia between 85 and 90 units (over two-thirds of those in New Guinea), while his comparable figures for Eurasia and Africa are 37 and 20 respectively. This reckoning gives us some idea of the incomparable linguistic diversity in the Pacific where approximately 60 per cent of the world’s genetic units are found. Considering the different states of our knowledge available, genetic and typological, from different areas, however, it still remains a question whether the genetic diversity of an area is simply comparable on equal terms with another: say, New Guinea, which is very often claimed to be the highest in the world, versus North America with the comparable figures, which is generally more favoured by sufficient information (see below). Notwithstanding this much diversity in the Pacific, very many of the languages spoken by both the aboriginal and the migrant peoples in every area have now become endangered, some “potentially” or “seriously”, while others are “moribund” or on the verge of extinction; these languages are incessantly followed by many others heading toward these categories, thereby eroding the pre-existing diversity at an alarming rate—see below for the scale of endangerment and the terminology. For several reasons, proper and particular attention should now be paid to where the Pacific stands linguistically. 1. The geographically unbalanced concentration of endangered languages in the Pacific is striking.The direst endangerment of the area in the whole world is reflected by the fact that the Pacific area (in our sense) includes 72 per cent of the 516 “nearly extinct” languages in Gordon’s Ethnologue (2005). Most


xi

importantly, the greatest majority of these small to infinitesimal languages are not well or hardly documented. 2. As anywhere else, the Pacific has seen linguistic replacements and absorptions to various extents by prestigious languages, foreign and aboriginal (including lingua francas), which led to greatly reduced pre-existing diversity in its various areas. But the region as a whole still remains highly varied and complex in terms of both linguistic genealogy and typology. Language Isolates and small families with no demonstrable kin abound here much more densely than anywhere else in the world, with the genealogical diversity generally being accompanied by the typological. Such diversity should weigh heavier than the sheer number of languages involved, given the weight of time it reflects. The region is thus a valuable repository of indigenous languages with genuine diversity. 3. Within the region, the most critical (not necessarily qualitative) situation of language endangerment obtains at its southern and northern extremities—Australia (Ch. 12) and the southern part of South America (Fuegian; Ch. 11), and the North Pacific Rim (Chs. 19 to 23). The major factors responsible for this situation should include at least: (i) the precontact population in the region was in general very thinly scattered chiefly because the natural environment with limited resources is too inhospitable to support a high density of exclusively subsistencedependent population; and (ii) on top of this, the originally shaky subsistence pattern sustained in a fragile ecosystem was readily liable to destruction by outside factors, from insinuated money economy to recent regime changes (e.g. the fall of the Soviet Union leading to abrupt corruption of reindeer herding in Siberia—cf. § 20.3.2.2); (iii) despite limited local bilingualism or multilingualism (far from the kind in Melanesia—cf. Ch. 7, § 13.1), there were few regional lingua francas, pidgins, and creoles for inter-ethnic communication which could possibly have served as a cushion against the invading languages (Chinook jargon in northwest North America stood no comparison with Tok Pisin in New Guinea); (iv) the cultural (especially material and technological) gap was overwhelmingly wide between the aboriginal hunter-gatherers and the civilized intruders who generally did not tolerate the cultural diversity; and (v) the latter’s political, educational, and religious attitudes toward native languages were in general suppressive and assimilative except for short periods of (apparently) weakened suppression.

Despite the precious store of linguistic—and intellectual—diversity with which the Pacific is endowed and despite its dire crisis, languages with decent

xii


documentation available are very few. In the face of this, we believe that the most urgent and strongest task for linguists should be their documentation, with high priority being assigned to those under-documented languages that have no known or no close relatives, and are on the very edge of extinction (see § 1.8.1). South Pacific. In terms of areal distribution, the South Pacific with its many islands (conveniently including those parts of Polynesia and Micronesia which are north of the equator) is outstanding in having a particularly large number of small (to very small and infinitesimal) languages. On the Latin American side of the Pacific, Middle America (Ch. 9) and the Pacific coast of South America (Ch. 10) also have remarkable diversity with a great many families, although the picture of actual situations still seems hazy for the latter, while only two (not known to be mutually related) isolates have survived in Fuegian (Ch. 11). The number of languages in Australia (Ch. 12)—which presumably had several hundred languages in the eighteenth century when the Europeans arrived—has now dwindled down to less than half that (i.e. about 250 or less). New Guinea (Ch. 13) is home to some 850 Papuan (non-Austronesian) and Austronesian languages, the latter of which are also distributed elsewhere (Chs. 14 to 17), with the total number of languages amounting to 1,000. Though still having a large number of languages, Australia and New Guinea have suffered an abrupt decline in their number of languages and speakers since the early days of contact, apart from several languages which still have over 100,000 speakers (Engan, Kuman in New Guinea, Samoan, Tongan, Tahitian, and Maori in New Zealand). Great phylogenetic diversity has been suspected of New Guinea, although the genealogical and typological information available still remains slim in general, despite the rapidly growing amount of documentation in Papua New Guinea (but not in Irian Jaya; see § 13.3.4). Another obscurity seems to prevail in Australia, with the difficulty of producing a phylogeny after a long “period of equilibrium” (Dixon 1997) when much language change may be of the diffusive, convergent type. South-East Asia. This area also has many small languages (Chs. 15 to 17) which have been threatened not only by pressure from dominant languages like Chinese (esp. Putonghua) and European colonial languages (Dutch, Spanish, French, English), but also from more locally long-established languages such as Bahasa Indonesia/Malay, Burmese, Thai, and Vietnamese. By contrast with the other regions of the Pacific, however, they belong to a relatively small number of linguistic families each with many sister languages (Mon-Khmer, Thai-Kadai, Tibet-Burman, Miao-Yao, and Austronesian),


xiii

and most of the inland languages (apart from Austronesian and some exceptions) are tonal, monosyllabic, and thus more or less analytical (if not that simple). By contrast, Japanese-Ryukyuan1 (or “Japanic”; Ch. 18) and Korean (see § 4.1.1) remain genealogically isolated despite the unflagging efforts devoted to unravel their respective historical root or mutual affiliation. The view that Japanese was the end product of some single lineage (e.g. Dravidian) persisted for many years, but an alternative theory of polygenetic formation—i.e. of mixed Austronesian-Altaic origin, primarily suggested and pursued by such scholars as Evgeny Polivanov (1891–1938) and Hisanosuke Izui (1905–83)—has again earned recognition in recent years (Sakiyama 1990), while renewed attention has also been directed to possible relatedness with Manchu-Tungusic, conceivably through a Koreo-Japanic relationship (see Vovin and Osada 2003). In this context, for the overview of the area, it would be important to note that the two isolates are surrounded in almost all directions by widely distributed families and language groups comprising the Austronesian family, the tonal-monosyllabic languages, and the so-called Altaic languages (subsumed as a group in view of typology). The apparent exception lies in the direction of north-east Asia. North Pacific Rim. Starting from Japanese and Korean, north-east Asia is characterized by isolates such as Ainu, Nivx (Gilyak), and Yukaghir as well as small families (containing only a few sister languages) such as the ChukcheeKamchatkan and the Eskimo-Aleut family, apart from a number of representatives of Tungusic languages (Chs. 19 to 21). This narrow belt of isolates and small families extends across the Bering Strait and along the Northwest coast of North America, with language distribution becoming more dense and typological diversity increasingly greater (Chs. 21 to 24). All the languages, including even Chukchee and Central Yup’ik, each with a substantial number of about 10,000 speakers, are (more or less) endangered or moribund. The most recent and detailed survey of the whole North Pacific Rim (barring the two ends just below) is Krauss (2003). The North Pacific Rim—defined as an arc bordering the North Pacific from the Japanese archipelago and the Korean peninsula at one end to California (at roughly the same latitude) at the other—is a region neither so varied nor so high in its number of languages as the Pacific. But it is a corridor teeming with languages of diverse and often obscure lineage 1 Given that Ryukyuan has probably greater internal diversity than Japanese, the two-branch family may better be called Ryukyuan (or even Ryukyuic) rather than Japanese-Ryukyuan (Krauss 2001b: 203).

xiv


comprising over forty genetic units, far more in the North American side (90 per cent)—an obscurity that led Boas (1911: 58) to the idea of “uncertainty of definition of linguistic families” as he got absorbed in the Northwest Coast Indian languages. Of approximately 150 genetic units in the Americas, North America north of Mexico contains some sixty, of which about 60 per cent are concentrated in the small region of northwest North America2 (Campbell 1997: 107–55; Miyaoka 1992a: 1048–66; Mithun 1999: 326–586). The centre of gravity of the indigenous languages lying in the northwest would arguably be a linguistic support for the coastal migration theory concerning North America (as first suggested in the early 1960s by C. J. Heusser et al. and recently strengthened by Canadian archaeologists—cf. Miyaoka 1992b; Nichols 1995) rather than the conventionally held “ice-free corridor” theory. This genealogical diversity and obscurity is naturally paired with the typological diversity of the North Pacific Rim, where each isolate or family seems more typologically fascinating than another. Typologically diverse as it is, the area is dominated by polysynthetic languages (§ 8.5, n. 10), which is a remarkable contrast with South-East Asia (above). Renewed interest in endangered languages

The year 2003 saw two important documents presented at the International Expert Meeting on UNESCO Programme Safeguarding of Endangered Languages (10–12 March 2003, Paris) by the Ad Hoc Expert Group on Endangered Languages, i.e. Language Vitality and Endangerment and Recommendations for Action Plans. These are expected to be the basis for international cooperation in works on language endangerment in coming years. In the first document, language situations are categorized into seven degrees with regard to intergenerational language transmission, the most commonly used factor in evaluating the vitality of a language (Fishman 1991). The seven-degree classification, basically derived from M. E. Krauss’s presentation at the Colloquium (12–17 February 2000, Bad Godesberg) for Language Endangerment, Research and Documentation—Setting Priorities for the 21st Century (Brenzinger ed. 2007), with some naming changes, will be followed in this book (except for some chapters) together with the terminologies, and the 2 In terms of the culture areas the region comprises the Arctic and Subarctic, the Northwest Coast, a portion of the Plateau, and California. The only other region in North America that comes rather close in linguistic diversity is, though out of the scope in this book, the Southeast, centred on the Gulf of Mexico. Outside these two regions one finds a relatively small number of widely distributed language families (Algonquian, Siouan, Iroquoian, etc.).


xv

five categories, excluding the two (5) safe and (5−) stable yet threatened which are not “endangered”, are indicated by the number:

• •

•

•

• •

•

Safe (5): The language is spoken by all generations. There is no sign of linguistic threat from any other language, and the intergenerational transmission of the language is uninterrupted.3 Stable yet threatened (5−): The language is spoken in most contexts by all generations with unbroken intergenerational transmission, yet multilingualism in the native language and one or more dominant languages has usurped certain important communication contexts. Potentially endangered (4): Most but not all children or families of a particular community speak their language as their first language, but it may be restricted to specific social domains (such as at home where children interact with their parents and grandparents). Endangered (3): The language is no longer being learned as the mother tongue by children in the home. The youngest speakers are thus of the parental generation. Parents may still speak their language to their children, but their children do not typically respond in the language. Seriously endangered (2): The language is spoken only by grandparents and older generations. While the parent generation may still understand the language, they typically do not speak it to their children. Moribund (1): The youngest speakers are in the great-grandparental generation, and the language is not used for everyday interactions. These older people often remember only part of the language but may not use it, since there may not be anyone to speak with. Extinct (0): There is no one who can speak or remember the language.

Background to this volume

This attempt at surveying the linguistic Pacific Rim region, modest as it is, is largely based upon research carried out over the period of 1999–2004 by the Endangered Languages of the Pacific Rim project (ELPR—director: Osahito Miyaoka) funded by the Ministry of Education, Culture, Sports, Science, and Technology of Japan (Grant-in-Aid for Scientific Research on Priority Areas).4 The project, which encompassed the South Pacific Rim and the Pacific, the North Pacific Rim, East and South-East Asia, and Japan (cf. Miyaoka 2001: 15), has so far published about 120 volumes, mostly of linguistic documentation (many with CDs) in the ELPR series (2001–5: 3 Even “safe”, however, does not guarantee language vitality, because at any time speakers may cease to pass on their language to the next generation, given the ever-increasing economic and cultural pressure to switch to a prestigious language of the area. 4 See Miyaoka (2001: 14–15) for a brief description of organized efforts by Japanese linguists before ELPR.

xvi


editors: Osamu Sakiyama and Fubito Endo), as listed in http://www.elpr.bun. kyoto-u.ac.jp/ The core of this volume is based upon work carried out by participants of the project including papers presented and discussions made at its Conferences on Endangered Languages held in the autumn of each year in Kyoto (Sakiyama and Endo (eds.) 2001 [ELPR C-002], Sakiyama et al. (eds.) 2004 [ELPR C-004], Sakiyama and Endo (eds.) 2004 [ELPR C-005]). As such, the great majority of contributors to this volume are participants and researchers from Japan and abroad who took part in the symposia. Others, however, were invited to write specifically for this publication so as to cover the whole region (as much as possible) and to highlight a number of important topics that would otherwise have been left untouched. The volume consists of two parts. Part I (Chs. 1 to 8) discusses a number of fundamental problems in language endangerment with special reference to the Pacific Rim. Part II (Chs. 9 to 24) is comprised of regional surveys of vanishing languages therein, with the chapters arranged by starting from Middle America clockwise down to South America and Australia, and up toward the north of Asia and turning back into the New World as far as the southwestern United States, although this last is the region with “the oldest continuous record of human habitation on the continent outside of Mesoamerica” (§ 24.1). The chapters in Part II are not uniform in coverage and in focus, depending upon the regional situations and the author’s judgement and preference. It is clear that there are far too few linguistic workers committed to the documentation of the endangered languages of the area, given the current situation in which more and more languages are nearing extinction year by year while the amount of time needed to complete a sufficient description of one language is so great. It is to be feared that ten years from now the number of vanishing languages will have increased at an even faster rate. If this book is of some help in giving its readers an understanding of and sympathy for the plight of these languages, then the editors’ goal will have been attained. O. M. and O. S. (ELPR) December 2005 Acknowledgements

This inexperienced editor would like to thank every author for his or her important contribution to this work. This volume would not have materialized as it has without the collaboration of Australian scholars (in particular


xvii

from La Trobe University). Endless thanks go to John Davey of OUP for his continuing patience and assistance with this book from its initial planning stages right through to its publication. Thanks are likewise due to his crew at OUP, and notably to Chloe Plummer, for all the hard work and conscientious effort undertaken in the final stages of production. My heartfelt gratitude goes to Setsuko Ikuta, Mitsuyo Takeuchi, and Tamami Shimada for their generous assistance in secretarial and technical work. Last, but not least, my thanks to Fumiko Sasama for going beyond the call of duty as a contributor by assisting with the editorial work on this volume. O. M. July 2006

References Aikawa, N. (2004). “UNESCO’s Programme on Languages”, in Sakiyama et al. 2004: 29–35. Boas, F. (1911). “Introduction” to F. Boas (ed.), Handbook of American Indian Languages, Part I, Bureau of American Ethnology, Bulletin 40. Washington, D. C.: Government Printing Office, 1–83. Brenzinger, M. (ed.) (1998). Endangered Languages in Africa. Cologne: Rüdger Köppe Verlag. —— (ed.) (2007). Language Diversity Endangered. Berlin: Mouton de Gruyter. Campbell, L. (1997). American Indian Languages: The Historical Linguistics of Native America. New York: Oxford University Press. Crystal, D. (2002). Language Death. Cambridge: Cambridge University Press. Dixon, R. M. W. (1997). The Rise and Fall of Languages. Cambridge: Cambridge University Press. —— (2002). Australian Languages: Their Nature and Development. Cambridge: Cambridge University Press. ELPR: Endangered Languages of the Pacific Rim (Publication Series). Osaka: Osaka Gakuin University. Fishman, J. A. (1991). Reversing Language Shift: Theoretical and Empirical Foundations of Assistance to Threatened Languages. Clevedon: Multilingual Matters. Gordon, R. G., Jr. (ed.) (2005). Ethnologue: Languages of the World, 15th edn. Dallas: SIL International (ed. B. Grimes until the 14th edn., 2000). Online version: www. ethnologue.com/. Krauss, M. (1992). “The world’s languages in crisis”, Language, 68/1: 4–10. —— (1997). “The indigenous languages of the north: a report on their present state”, in H. Shoji and J. Janhunen (eds.) 1997: 1–34. —— (2001a). “Mass language extinction, and documentation: The race against time”, in Sakiyama and Endo (eds.) 2001: 19–39. —— (2001b). “Linguistic arrogance and endangered languages: Japanese and Ryukyuan comments to Uemura’ paper” (Uemura 2001), in O. Sakiyama and F. Endo (eds.) 2001: 203–5.

xviii


Krauss, M. (2003). “The languages of the North Pacific Rim, 1897–1997, and the Jesup expedition”, in L. Kendall and I. Krupnik (eds.), Constructing Cultures Then and Now: Celebrating Franz Boas and the Jesup North Pacific Expedition. Washington: Smithsonian Institution: National Museum of Natural History, 211–21. Ladefoged, P. (1992). “Another view of endangered languages”, Language, 68/4: 809–11. Meillet, A., and Cohen, M. (eds.) (1952). Les Langues du monde, 2nd edn. Paris: Campion. Mithun, M. (1999). The Languages of Native North America. Cambridge: Cambridge University Press. Miyaoka, O. [宮岡伯人] (1992a). “North American Indian languages [北米インディアン諸語]”, in T. Kamei et al. (eds.), The Sanseido Encyclopaedia of Linguistics [言語学大辞典], vol. iii. Tokyo: Sanseido, 1004–78. —— (1992b). “Linguistic world of the North Pacific Rim [環北太平洋の言語世界]”, Monthly Journal of Language [月刊言語], (Tokyo: Taishuukan Shoten), 21/8: 30–7. —— (ed.) (1992c). Languages of the North Pacific Rim: Types and History [北の言語類型と歴史]. Tokyo: Sanseido. —— (2002). “The North Pacific Rim as a Linguistic ‘Old World’”, presented at the International Symposium “The Raven’s Arch: Jesup North Pacific Expedition Revisited”, Sapporo, 24–28 Oct. Nettle, D., and Romaine, S. (2000). Vanishing Voices: The Extinction of the World’s Languages. New York: Oxford University Press. Nichols, J. (1995). “The spread of language around the Pacific Rim”, Evolutionary Anthropology 3/6: 206–15. Robins, R. H., and Uhlenbeck, E. M. (1991). Endangered Languages. Oxford: Berg. Sakiyama, O. [崎山理] (ed.) (1990). Formation of the Japanese Language [日本語の形成]. Tokyo: Sanseido. —— and Endo, F. [遠藤史] (eds.) (2001). Lectures on Endangered Languages ii: From Kyoto Conference 2000. ELPR C-002. —— Endo, F., Watanabe, H. [渡辺己], and Sasama, F. [笹間史子] (eds.) (2004). Lectures on Endangered Languages 4: From Kyoto Conference 2001. ELPR C-004. —— (2004). Lectures on Endangered Languages v: From Kyoto Conference 2002. ELPR C-005. Shoji, H., and Janhunen, J. (1997). Northern Minority Languages: Problems of Survival. Senri Ethnological Studies, 44. Osaka: National Museum of Ethnology. UNESCO Intangible Cultural Heritage Units’ Ad Hoc Expert Group on Endangered Languages (2003). Language Vitality and Endangerment. 10 March 2003, Paris. Voegelin, C. F., and Voegelin, F. M. (1977). Classification and Index of the World’s Languages. New York: Elsevier North Holland, Inc. Vovin, A., and Osada, T. (eds.) (2003). Perspectives on the Origins of the Japanese Language. Kyoto: International Research Center for Japanese Studies. Wurm, S. A. (ed.) (2001). Atlas of the World’s Languages in Danger of Disappearing. 2nd edn., rev., enlarged, and updated. Paris: UNESCO.

PART I Diversity, Endangerment, and Documentation

1

Keynote—Mass Language Extinction and Documentation: The Race against Time Michael E. Krauss

1.1. Introduction: recognizing language endangerment as a global reality 1.2. Global total number of languages spoken 1.3. Classifying and counting languages as “safe”, endangered, or moribund 1.4. Language size distribution, extreme instability 1.5. Prevention of language loss, linguists’ role 1.6. Reasons for preventing language loss 1.6.1. Ethical argument 1.6.2. Scientific argument 1.6.3. Biological argument, “logosphere” 1.7. A tragedy quite unnecessary 1.8. Documentation and the linguist’s role 1.8.1. Documentation goals and priorities 1.8.2. Priorities within linguistics 1.8.3. Costs of documentation

3 5 7 10 12 13 14 14 16 17 18 19 21 22

1.1. Introduction: Recognizing Language Endangerment as a Global Reality I realize that there must be many people here who are attending a conference on the subject of Language Endangerment for the first time, and who may not have heard any of the information or opinions I shall begin this presentation with; on the other hand there are some people here, especially old friends and colleagues, who have heard this before, who I hope will be patient with me. To summarize: in a 1991 paper entitled “The world’s languages in crisis” addressing the Linguistic Society of America, I presented my views on the impending global mass extinction of languages with these statistics: of the world’s 6,000 languages, 20 to 50 per cent already are or soon will be no longer spoken by children, i.e. are moribund, to become extinct during the

This paper was written as a keynote address to the Kyoto Conference 2000 (held at Kyoto Kokusai Kaikan, 24 and 25 November). The author makes no apology for the informal and personal tone throughout. With minor updating, it expresses the author’s current views. I wish first to thank Professor Miyaoka for the great work he has done to arrange this conference and for inviting me to speak to this audience.

4

m i c h a e l e . k r au s s

twenty-first century; at the other end of the scale, only 5 to 10 per cent (300 to 600) are “safe”; and all the rest, 40 to 75 per cent are “merely” endangered, may cease being spoken by children during this century. The subject of language endangerment, i.e. languages ceasing to be spoken, becoming extinct (with or without documentation), is certainly not new, but in fact is as old as human history—and prehistory. But as a subject for conferences, however, particularly international conferences (rather than local conferences for patriots or loyalists concerned with the survival of particular languages or language groups, e.g. Ainu, Celtic), conferences concerned with the subject of survival or extinction of large numbers of languages on a massive scale as a field of enquiry and/or action, this is a fairly recent development, not much more than ten years old. Organized concern began on the part of the Permanent International Committee of Linguists, under UNESCO, in the late 1980s. That resulted in the first book with the title Endangered Languages (Robins and Uhlenbeck 1991). The first plenary conference session on the subject as such that I know of was at the Linguistic Society of America, January 1991, arranged by the late Ken Hale. The first whole conference that I attended on “Endangered Languages” was at Cornell, February 1995, and the second was that same year in Japan, the “International symposium on endangered languages”, Tokyo, 18–21 November 1995 (see Matsumura 1998). Though that was only five [now ten] years ago exactly it already seems like a long time! I dearly wish that my basic statistical statements of ten [now fifteen] years ago had by now proven to be a terrible mistake, but I’m afraid I see no reason to change them very fundamentally. They have met with little or no challenge as to their accuracy or realism, incidentally, even though obviously I can make no claim to be an “expert” (except perhaps merely for Alaska), and there are many experts, especially at this conference, who have studied for many years in this way many parts of the globe. Of these, I mention especially Barbara and Joseph Grimes as compilers of the Ethnologue, in a category by themselves, and to whom I shall return soon below. In another sense, however, my statements on the mass language extinction which we face this century, on an entirely unprecedented scale, have indeed met with another kind of challenge, namely that far from being the tragedy or catastrophe I call it, we should welcome it as progress and a benefit to mankind. That is a very widespread opinion among the public—though probably not the audience gathered here—and that is an all-important challenge to us to answer. I shall therefore also include my reasonings that it is not good for us to lose languages at the end of the first part of this presentation.

m a s s l a n g u a g e e x t i n c t i o n a n d d o c u m e n t at i o n

5

1.2. Global Total Number of Languages Spoken First, for the number of languages still spoken (or, in some cases, no longer actively spoken, but only still remembered by a few old persons), there have been several global survey estimates over recent decades. The trend is definitely for higher and higher estimates—but for how long? In the 1930s to 1950s linguists guessed there might be 2,500 to 3,500 languages (e.g. Meillet and Cohen 1952: p. xxix); here we have not only to consider the perennial problem of definition between language and dialect, but also to remember that exploration was at a relatively preliminary stage, especially in some parts of the world where diversity is greatest. Major advance was realized in the Voegelin’s Classification and Index of the World’s Languages (1977); I count in that book about 4,430 languages listed, not including about 370 more listed as extinct. I note further Stephen Wurm’s well-informed and representative statement that “well over five thousand languages are known to exist or to have existed in the world” (Robins and Uhlenbeck 1991: 1). However, by far the most extensive survey and inventory of mankind’s languages worldwide has long been conducted by the Summer Institute of Linguistics/Wycliffe Bible Translators, as published in the Ethnologue (1st edition 1951, most recent printed edition the 14th in 2000); I have at hand the 8th (the first “complete” or worldwide one, with language counts) to 14th editions, all with Barbara Grimes as editor and including Joseph Grimes as consulting editor, from the years 1974, 1978, 1984, 1988, 1992, 1996 2000; the total number of (non-extinct) languages listed in each of those, respectively, is 5,687, 5,103, 5,445, 6,170, 6,528, 6,703, 6,809 [6,912 in the 15th edn. by Gordon].1 The drop between the first two totals is related to the earlier reports of the Voegelin project and their final publication of 1977, as the Grimes project basically takes over from there. Note the lessening in the increase between the 10th–14th editions, each four years, 725, 350, 175, 106. How high their total might get before the rate of extinction starts to exceed the rate of discovery or definition of new languages is an interesting question. In any case, I personally believe that the Ethnologue numbers are on the high side, from the examples I know, given a fairly high standard for defining different languages as opposed to dialects, perhaps in consideration of immediate intelligibility of Bible translation. The worldwide total figure I have been using is

1

The 15th edn. (Gordon 2005) lists languages, with an increase of 103, but perhaps over five years.

6


6,000 extant languages, a nice round figure that happens to be one millionth of the human population, a kind of middle figure that I have not heard anyone object to, not only because no exact such figure could be established, but most importantly, because what matters is of course not the hypothetical figure itself, but rather the dynamics and proportions in what is happening to that figure. It is perhaps of interest to provide some further detail on Ethnologue totals in tabular form including breakdown by continent: Europe, Middle East, Asia, Africa, Pacific, Americas (after 1988 Middle East is included in Asia) (see Table 1.1). Table 1.1. Ethnologue language totals Edition

Year

Total

Eur.

ME

Asia

Afr.

Pac.

Am.

8 9 10 11 12 13 14 [15

1974 1978 1984 1988 1992 1996 2000 2005

5,687 5,103 5,445 6,170 6,528 6,703 6,809 6,912

46 55 83 185 209 225 230 239

42 40 45 67

1,956 1,562 1,666 1,846 2,034 2,165 2,197 2,269

1,741 1,600 1,727 1,919 1,995 2,011 2,058 2,092

1,039 1,034 1,053 1,216 1,341 1,302 1,311 1,310

803 812 871 938 949 1,000 1,013 1,002]

One factor which may enter into some of the increases, in addition to the obvious one of improved information or coverage, is certainly not proliferation or diversification of languages, but, especially in these SIL counts, bilingualism in dominant lingue franche, replacing ability to understand and interact in broad dialect continua, so that such traditional competence range shrinks and what might have been functionally one language might now be a dozen, not by linguistic change but only by sociolinguistic change. Extinct languages are not included in these table figures. However, the number of (known) extinctions is not (yet) nearly so high as is sometimes claimed. For example, extinct languages are very liberally listed in Voegelin and Voegelin (1977), including both ancient and recent cases, not only the well-known ones, but many that are only mentioned as having existed, without any documentation; yet of an approximate total of 4,800 languages reported in the book, only 370 are extinct. Thus, though the total number of language extinctions since history began is surely higher, the number of historically recorded extinctions (in this broad sense) is probably not far over 400—yet. We could soon be losing that number every decade.


7

1.3. Classifying and Counting Languages as “Safe”, Endangered, or Moribund Just as the term endangered languages is adopted from the biological term endangered species, the UNESCO project began with the concept of Red Book to list what is classified as endangered, as an alarm, a call for protection. However, just as many biologists have pointed out that, given the situation that is so alarming in so many parts of the world, including huge sectors for which we have very limited information or access, especially because of regimes that judge such information and access to be in economic or political conflict with the “national interest”, it would be far more realistic to maintain instead a Green Book of species that are not endangered, so that any known extant species is assumed to be endangered unless shown otherwise. The biological species endangerment rates of the very worst biological alarmist estimates, e.g. that 50 per cent (not just 5 per cent) of mammals may be endangered (or threatened), are still better than the very best language endangerment scenarios I can imagine. Therefore all the more reason I ended up resorting first to the much easier calculation of how many languages are “safe” (in quotation marks), a sort of term I coined for the purpose. To define this term, I took then and still take now the time-frame of this twenty-first century. Though it is barely imaginable to me what the world will be like in the year 2100, there are people alive now who will still be alive then, and some calculation principles are pretty clear. One is that, generally, any language which is now no longer being spoken to children, barring radical change and reversal of language loss, will be extinct by the year 2100. Calling languages no longer spoken by children—i.e. moribund languages—calling them merely “endangered” is a kind of euphemism, but that usage is already well established, probably for good humanistic reasons. Let us continue that, but with the special qualification of “seriously endangered”, and set that special category for the moment aside. For classification as “safe”, I can envision only two factors definitive for our purposes. One is sheer number of speakers, a number large enough that even under the worst political or economic conditions, e.g. genocides that governments can get away with for long periods these days, it is still probable that at least some children will be learning the language as mother tongue in the year 2100. With the consideration that Breton, with over a million speakers in living memory, is now highly endangered under French rule, likewise Ryukyuan under Japanese, whereas Greenlandic and Faeroese with only forty-odd thousand speakers each, both geographically isolated and under Danish rule,

8


at the other extreme, are “safe”, it is clear that safety in numbers is relative to the size and nature of the dominant society, and to geography, but it is also clear that Greenlandic and Faeroese are somewhat more the exception, with both their geographic isolation and the relative lightness of the Danish yoke. In other words, the lower threshold of safety in numbers, on the average, is probably much closer to one million speakers than to 50,000. The number of languages with a million or more speakers has generally been listed as 200odd, but the current total could be 330 (Grimes 2000),2 increased of course not so much by a sudden spectacular growth of languages with under a million speakers as by subdivision of very large languages. Perhaps it is fair to say that 300 out of 6,000 languages, i.e. most importantly 5 per cent, are “safe” by sheer number of speakers. Ten years ago I calculated that by lowering the threshold of safety in numbers from one million down to 100,000, one might double or actually quadruple that percentage, but such a limit seems perilously low in most cases for the conditions of this new century. The other most important type of factor contributing to language maintenance is socioeconomic and/or attitudinal, a factor surely much more abstract and difficult to define, though probably no less important than sheer number of speakers in itself. For relatively small numbers of speakers to maintain their languages indefinitely, certain minimal conditions would be required, e.g. a viable literacy, and social conditions favourable enough to support or at least tolerate maintenance of that language. The only languages qualifying as realistically “safe” under those conditions, I believe, would be those that enjoy recognition and support as national languages of nation-states or at least as regional languages thereof. This adds only a very small number of languages to even those with 100,000 or more speakers, let alone 1,000,000. There are currently about 200 sovereign states in the world (a number gradually increasing), but of those a very large proportion recognize only English or Spanish or Arabic, for instance, and of course the state languages of most of the 200 are already very safely in the category of over a million speakers to begin with. Thus very few languages are added to the “safe” category by this factor of state support, e.g. Icelandic with 280,000 speakers, or in the regional class and under 100,000 speakers the unusual Greenlandic and Faeroese examples. Perhaps so factoring this kind of status seems simplistic to some, and there may well be categories of languages in Africa and Asia where no state or regionally recognized languages dominate over certain types of relatively strong tribal languages enough that some kind 2

The 15th edn. of Ethnologue: 347.


9

of balance may allow the tribal languages to be spoken to children through the century. Unless shown otherwise on a significant scale, for the moment I still estimate that the number of “safe” languages, those which will still be spoken by (some number of) children at the end of this twenty-first century, is somewhere between 300 and 600, or 5 to 10 per cent of today’s 6,000. That means, of course, that we stand to “lose”, in some sense, 90 to 95 per cent of the 6,000 languages we still have. With that I raise two types of question: one is whether a language is to “die”, become extinct, i.e. cease to be spoken as a living language, and the other is whether that language is to be preserved at least in the sense that it is documented before it becomes extinct. The documentation question I shall consider later, and here consider the more probable scenarios for how many of the 6,000 will be extinct by the year 2100 and how many more will be on their way to extinction, still spoken or remembered, but with no children speakers, so as to become extinct during the next century. I take pleasure in noting two important favourable changes from my earlier estimates of the number of languages already moribund and to become extinct during this century. One is that ten [now fifteen] years ago I had the impression that somewhere between 20 and 40 or even 50 per cent of the 6,000 were already not spoken by children, or soon would not be, by the year 2000. My present impression is that the century has begun with a figure of more like 15 to 30 per cent moribundity, perhaps most likely 20 to 25 per cent. (My minimum guess of 20 per cent moribundity some years ago matched or at least approached Joseph Grimes’s maximum guess [p.c.], and my current minimum of 15 per cent comes still closer.) The second favourable change is that there has been significant increase in programme activity to reverse the loss of languages no longer spoken by children in at least some parts of the world, such as North America and Australia, where the loss has been especially severe—increase in the incipient development of serious “immersion”-type programmes. These significant exceptional cases, perhaps soon to multiply rapidly, now clearly show that the process of language loss does not inevitably have to progress unidirectionally toward extinction. The two perhaps most spectacular cases of loss reversal are the closely related Maori of New Zealand and Hawaiian. Both of these languages, incidentally, have the advantages of (1) relatively large numbers of people, (2) a history as a state and literary language, and (3) are the only aboriginal language of the entire state, with (4) relatively little internal dialectal diversity. I do not see any clear way to predict just how rapidly such success at reversing loss may spread in various parts of the world, so can only emphasize still more insistently the qualifying phrases in my previous estimates, “will become

10


extinct during the 21st century, at the rate things are going, unless radical measures are taken”. In any case, ten [now fifteen] years ago and still by far the largest category of the 6,000 languages lies with those that fall between the two categories described above, that are neither moribund nor “safe”, i.e. languages that are “merely endangered”. Those languages are still spoken by children, but for how much longer? Given the maximum and minimum figures for the two other categories above, 5 to 10 per cent and 15 to 30 per cent, there are left in between 60 to 80 per cent, or 3,600 to 4,800 languages whose future hangs in the balance. But what “balance”?

1.4. Language Size Distribution, Extreme Instability We need to put that question of “balance” first into long-term perspective, and to recognize that the language extinction crisis we now face is certainly quite unprecedented. Mankind’s history has been one long process of technological change from hunting-gathering society to agrarian to industrial (and post-industrial) society with ever-increasing rates of population growth that accompany such development. Before the food-production revolution beginning about 10,000 years ago, human population was not only quite low and sparse, c.5,000,000 to 10,000,000 worldwide, but also very large societies would not have been possible. Size distribution of societies was relatively uniform, and the difference between the average-size language and the mediansize language was not enormous. The number of languages was therefore relatively stable, and perhaps somewhat higher than it is now, though not enormously so. If any pre-agrarian societies had developed so far beyond others in power that they eliminated others on the scale that is now about to happen, there would be historical or archaeological evidence that we don’t have. Now, however, the size distribution is such that the top ten languages account for close to half of the world’s population, the top fifty or sixty for maybe three-quarters of it, and the top 300 or so (i.e. the languages with a million or more speakers) must account for over 90 per cent, leaving still 5,700 languages to be spoken by something like 5 to 7 per cent. Though the average-size language (6,000,000,000/6,000) has about 1,000,000 speakers, the median-size language is now far smaller, with about 6,000 speakers. The upper 3,000 languages must account for something on the order of 99.9 per cent of the world population. In the process of preparing this paper, I had received a most interesting document from Joseph Grimes, a listing of the current Ethnologue’s


11

(14th edn.) languages in order of their number of speakers, which I have used to adjust somewhat the above figures.—Here, even though I digress, I cannot pass up the chance to remark on the bottom end of that list, the 229 languages listed with ten or fewer speakers. Of those 229, 105 are in Australia and 48 are in the USA. Starting at the bottom, 52 languages are listed as having one speaker only (28 in Australia, 10 in the USA), 29 with two (19, 3 respectively), 28 with three (21, 0), 12 with four (4, 2), 26 with five (7, 8), 18 with six (10, 4), 10 with seven (only 2, 1), 7 with eight (3, 2), 2 with nine (1, 1), and 45 with the round number of ten speakers (9 in Australia, 17 in the USA). One suspects that a part of the reason why most of these languages are in Australia and the USA is that information and access is much better in these two countries than in, for example, Indonesia, Myanmar, Sudan. To return to my original text—such a pattern of size distribution as noted above is extremely unstable, reflecting of course the ever-increasing inequality in power of the societies the languages represent. Given the current trends of accelerating industrialization, urbanization, deforestation, expulsion, resettlement, compulsory assimilationist education, genocide, globalization, pervasiveness of expanding communication technology, mass media, one can only expect most factors favouring the languages of more powerful societies to expand at the expense of the less so to take effect with increasing rapidity, with the result that a graph showing total human population growth exploding almost vertically upward will also show the number of different languages still spoken imploding almost vertically downward with but little delay. The United States and Australia are relatively “advanced” in this process—as implied in my digression above—both already having shown over 90 per cent loss (extinction or moribundity) of their aboriginal languages. Areas where linguistic diversity is at its densest, such as parts of Africa, and especially Papua New Guinea and Indonesia, which have the very highest (over 800) and second-highest numbers (over 700) of different languages, have not yet “advanced” nearly so far in language loss (or elimination), but those areas are rapidly heading in that direction. I therefore fear that whatever favourable change I might see in the moribundity figures presented above, and in the prospects for reversal of language loss through community commitment and external encouragement or at least tolerance—I fear that such improvement may be greatly exceeded by the rate at which the “merely endangered” majority, the 60 to 80 per cent of the 6,000, will cease to be spoken to (and therefore by) children during this century, to become extinct in some cases already during this century, and surely during the next. Others have interpreted statistics such as those I suggested ten [now fifteen] years ago to imply that with the higher moribundity estimates, negligible

12


amount of loss reversal, and rapidly increasing cessation of intergenerational transmission adding significantly to extinctions during this very century, we might well find that by the year 2100 already 50 per cent or 3,000 languages will be extinct, and another 40 to 45 per cent will by then be well on their way to extinction, so that we do indeed stand to lose 90 to 95 per cent of our present human languages. I agree that is possible, that is certainly what is at stake. In some senses it is a relatively trivial point then, to determine more precisely which percentage will become extinct during this century and which percentage the next, at the rate things are going or will go. Perhaps 50 per cent extinction this century—a rate of about one language every twelve days, by the way—is a pessimistic scenario, and an optimistic but possible figure might be 30 per cent, for the rest to be lost during the next, giving us at least a little more time to consider and act.

1.5. Prevention of Language Loss, Linguists’ Role Obviously a more important point than the statistical speculation is not only what can be done to reverse language loss, but in at least twice as many cases, i.e. of the “merely endangered” languages (60 to 80 per cent as opposed to 15 to 30 per cent), what can be done to prevent language loss. It is certainly far easier and more effective for parents to speak the language to their children in the home, in the miraculous supremely human universal traditional way, than by any method linguistics can devise. If only people could be brought to remember this before they abandon speaking their traditional language in the home to their children, rather than regret it themselves too late for that, or for their children to regret it still later. Here I shall take just this moment to answer the phonetician Peter Ladefoged’s very important rhetorical question (“Another view of endangered languages”, 1992, in direct opposition to mine printed in the previous issue). He refers to an elderly speaker of Dahalo in Uganda, who was proud and happy, without regret, that his son spoke only Swahili. Ladefoged asks, “Who am I to say he was wrong?” I think I can answer, on the basis of what I have seen, not of speculation, that at the very least the Dahalo father himself may live to regret that his son does not speak Dahalo, and even more likely, the son will regret it, sooner or later. Ladefoged himself would probably have to agree with at least that much; so we narrow the question to the role of linguists in the issue—and who, if not linguists? I insist that linguists have a crucial role in fact at three levels in the matter of language endangerment: (1) in calling attention to it, increasing as much


13

as possible public awareness of it, that it is happening, and that individuals, whole peoples or nations, and even the human race as a whole will live to regret it, or at the very least—and even Ladefoged would have to agree—may live to regret it; (2) in documenting languages, all languages of course, but with logical priority systematically assigned to languages according to criteria we need to agree upon; (3a) in providing technical help to language communities which are trying to reverse loss after cessation of traditional intergenerational transmission, and (3b), the most difficult but with the highest numerical stakes by far, providing help to communities in strengthening the position of endangered languages both technically, e.g. with good writing systems, lexical expansion, expansion of use in education, media, and even activism in sociopolitical support where possible, given the often advantageous position of the professional linguist. This last role, (3b), is by far the most controversial one, which I realize many linguists disagree on. Role (3a) is increasingly being recognized as important and appropriate for linguists as more essentially theoretical linguistics becomes less fully the predominant or exclusive type in academe, and in fact paying jobs are multiplying rapidly for role (3a), helping with endangered languages. I shall for the moment concentrate my continuing remarks on roles (1) and (2).

1.6. Reasons for Preventing Language Loss As I noted above, response to my efforts in role (1), in raising public awareness of the impending loss of so many languages, has never included any objection to the statistics of its magnitude, but only objection to my claim it is a tragedy or a disaster. At all educational and economic levels in the USA especially, the most common objection and probably prevalent attitude is that the decrease in number of languages should in fact only be welcomed as a benefit to the human condition. It is therefore of the utmost importance that we give the reasons why we believe the loss to be bad rather than good for mankind. I now classify such reasons into three types, which I shall summarize here, not in order of importance. However, all these reasons presuppose one point which it is also essential to explain, that all known human languages are at an equal human intellectual, spiritual, logical, aesthetic, linguistic level. There is no such thing as a “primitive” language, or inferior or superior language, any more than there is such a race. True, there are languages with larger literatures, larger vocabularies, representing larger, more powerful societies, but this has only

14


to do with the technology, especially weaponry, of the society, and has nothing to do with the inherent linguistic nature of the languages themselves. All languages in that sense are equal.

1.6.1. Ethical Argument First, then, is the human rights or ethical argument. If the number of languages is to decrease, on what principles do we allow this to happen or help it to happen? So far the only principle that appears to be operating is social Darwinism, “survival of the fittest”, whereby languages of the most powerful, best-armed societies replace those of the weaker. This is getting increasingly frowned upon when it operates between defined nation-states, which have armies and navies, but it seems to be still rather well tolerated when it happens within nation–states. (That can also lead to trouble where linguistic boundaries do not coincide with those of the nation–states, taking Kurdish and Albanian as some recent notorious examples.) Language and “reason” are gifts which define humanity as unique, different from all other animals or brutes, giving us the ability to do conscious planning for future conditions to our best advantage. It is ironic indeed that, endowed with reason, we have not yet learned to apply such reasoned principles to our language situation. Instead, brutishly, we still allow the future of that to be determined essentially by length of claw or missile trajectory control. The United Nations has since the1960s, and increasingly in the last decade, recognized the right to speak one’s own ethnic language as a basic human right, therefore at least questioning the right of Turkey to forbid its citizens to use Kurdish, protecting Basque in Spain and France, for example, and literally thousands of other languages, the majority of which are spoken in only one nation–state and are therefore especially vulnerable to violation of human rights.

1.6.2. Scientific Argument Second, I list the scientific argument, at three levels, of increasing importance. At the lowest level, surely, I consider the importance of languages for the science of linguistics. It could in fact be argued, as Chomsky is reported to say, that to a Martian all human languages essentially are the same, with only trivial variation, and that, to understand the basic nature of human language, any single such language will suffice, so the famous phrase “take English, for example”, and if you insist on double-checking, all right, take Japanese too. In biology, to understand the basic nature of life, it would be stupid to


15

use panda bears and coelacanths for their DNA, when white rats and fruitflies are so much handier and cheaper. That is an important argument, for an entirely different purpose, to which I’ll return. My point here is that the argument that language loss is indeed bad for the science of linguistics, coming from us linguists, can be mostly counterproductive, sounding self-serving and trivial to the average hard-working taxpayer: “Ah, so you linguists want to keep people speaking their own languages just so you can study them, like animals in a laboratory or a zoo, not caring about the people at all. Their interest and the national interest are far more important than the science of linguistics.” I believe it very important to argue the point within linguistics, but not outside the profession. In fact, there it could still be claimed that lost diversity is not considered an important issue by mainstream academic linguists, given the theoretical definition and claim cited above. The second level of scientific argument, far more important to the public, is what I call the informational, that embedded in every language is a vast body of information that has been accumulated by the countless generations of its speakers regarding e.g. their own special history, geography, and especially environment, animals and their behaviour, and plants and their medicinal uses, that “you chew on the bark of this bush and the lump in your breast goes away”, but the last person who remembers that is now 87 years old. I suppose at that level, perhaps beyond knowledge of a useful root like potato are also realms of ideas such as those represented by kayak, igloo, or omelette, sushi, bagel. Every time we lose a language, mankind is impoverished. As Ken Hale put it, it’s “like dropping a bomb on the Louvre”. Now for the highest level of scientific argument, which I call the abstract, that every language has embedded within it not only information and ideas unique to its history and environment, for example, but also its own unique interpretation of universal human experience, e.g. of human emotions, human relations, kinship systems, or especially in its grammar, e.g. of time and reality, subordination, evidentiality in its verb system, or of space and direction in its demonstatives, to pick just some examples, not necessarily the most abstract either. This becomes quite obvious to anyone who seriously tries to learn a new language. To sum up the second and third levels of the scientific argument, when we lose a language, we certainly lose not “just another set of words (for the same things)” but vast treasures of information, ideas, and even beyond that, we lose different ways of thinking, in fact ability, even freedom some would say, to think in different ways. Just as it is in our interest to be able to assign difficult or important decisions to juries or parliaments rather than to single individuals or dictators, because “two heads are

16


better than one”, and it is not better that all people think alike, the same claim should be made for the importance of linguistic diversity.

1.6.3. Biological Argument, “Logosphere” This then leads to the third type of argument, which I now call the biological. This is the most abstract and difficult to claim, but potentially the most important of all. I claim first that each language is of inestimable value, and as complex as a life-form, a species of human creation that is so complex that, as I’ve put it before, “a hundred linguists working a hundred years could not get to the bottom of (fully document and explain) a single language”. I’ve never heard a linguist disagree with that claim, though many non-linguists have, so it behoves us linguists to explain the biological argument first at that level. The final level of the biological argument is, I claim, that in the same way that individual languages can be compared with living species, so also can the human world of languages, mankind’s linguistic diversity, also be compared with biological diversity as constituting a system in itself, a structure of some kind, with interrelationships, interactions, interdependences, a system which we are barely beginning to recognize as such, or to understand. In fact we have no name for it, so on the model of the term biosphere which we use for the ecosystem of biodiversity that is the delicate membrane around our little planet in which we have evolved and on which our survival as living organisms utterly depends, I construct the term logosphere for the ecosystem of linguistic diversity that is the delicate environment of cultural and intellectual and linguistic diversity in which we have evolved and on which, I herewith claim, our survival as humanity utterly depends. I confess that is a challenging claim to defend, and many people believe the opposite, that linguistic diversity is a curse, but I consider that the most important argument of all. For me, furthermore, there are two quasi-mystical connections to it. One is what I used to call the aesthetic argument, and made the tactical mistake of listing that first, that the world would be a less beautiful and less interesting place if the only flower were tulips or chrysanthemums, or the only language were English or Japanese. That could seem like a trivial or subjective issue, aesthetics being a luxury where people are starving, and beauty being in the eye of the beholder. Or worse, since journalists’ deadline is tonight, their space is limited, likewise their time and their readers’ attention span; as I once made the mistake of listing that aesthetic argument first, that’s all that got printed, and my whole argument would be merely that with so many languages lost the world would be less interesting and beautiful


17

(to some privileged persons). Yet “man does not live by bread alone” and for some important reason a beautiful landscape is a biologically healthy landscape, so the aesthetic argument is somehow part of the biological. The other quasi-mystical connection I call the religious argument, which for me is even ancestral, as I would not be the first of my ethnic group to argue that our linguistic diversity was given to us or imposed upon us by God, as explained in the story of the Tower of Babel, the interpretation of that story being that our creator, our father-figure, in his wisdom, which far surpasses ours, gave us that linguistic diversity, or punished us with it, for our own good. There must be many people who believe that we have now suddenly somehow reached that pinnacle of godly wisdom, so that we may now presume to rebuild the tower and stop speaking so many languages. I personally do not share that conviction, and believe instead that we are at the very least much safer not to stop speaking so many languages before we are really sure we know what we are doing, and, with the logosphere as with the biosphere, we have the responsibility to hand it down as best we can to the next generation in the condition it was handed down to us.

1.7. A Tragedy Quite Unnecessary To conclude this basic section on language endangerment, its causes and consequences, I wish to emphasize one further principle, now to distinguish language endangerment sharply from endangerment of biological species. Whereas there is in fact often or usually a genuine conflict of economic interest between human groups and the maintenance of animal species, e.g. the logging industry’s need to harvest redwood trees and the maintenance of spotted owl habitat, or agriculture and panda bamboo supplies, or demand for decorative knife handles and rhinoceroses’ reluctance to donate their horns, to take just a few well-known examples, which require genuine sacrifice or compromise on the part of the human groups, conflict between language groups as such is totally avoidable; loss of any language is a totally unnecessary tragedy, maintenance of any language along with acquisition of one or more others requires no sacrifice or compromise whatever, of anything of value. Though instead in many parts of the world arrogance, paranoia, bigotry, and stupidity seem to prevail, it is really only those qualities that need to be sacrificed in order to prevent the loss of language. It has been shown routinely over and over again that very ordinary persons can learn two, three, four, or more languages quite well, with no harm to the brain

18


or any mental abilities—quite the contrary—and that such skills are so natural to the human brain that even children can learn multiple languages, more easily in fact than adults. There is no limit in principle to the capacity of a human brain for language, learning whereof can be a lifelong process. So there is no real or justifiable reason to sacrifice Navajo in order to add English, or Ryukyuan or Ainu for Japanese, or Basque for French and/or Spanish. Even the political reasons for thinking so are false and unnecessary, as the time is rapidly coming when we will have to understand human organization as something more complex than the mere exclusive taxonomy of nation-states left of defunct colonial empires. For us to survive that phase it will have to become possible for one to be linguistically and competently both a Kurd and a Turkish citizen, an Ainu and Japanese, a Basque and Spanish or French, Navajo and American, and a loyal one too. Our organizational principles must become intelligent and flexible enough to adapt to the complex nature of our humanity.

1.8. Documentation and the Linguist’s Role It remains for me to append some remarks on the linguist’s role in documentation of endangered languages. Who, of course, can do this job, if not linguists? It is true and important to remember that audio- and videotape recording can best be done by anyone who has a good knowledge of the language and its culture, a good working relationship with the speakers, access to good recording equipment, and skill in using it. Linguists should obviously do everything they can to facilitate the work of such people, but the skills of linguists are themselves not required for that role. If two persons had been standing next to Moses, say at the shore of the Red Sea, one the linguist who had learned how to scratch onto skin or stone symbols for the consonants of what God spake there, and another person next to him with a video camera recording all that, which document would be valued more today? Even though linguistics has made some progress since then so that now we presumably know how to transcribe vowels too, and even vowel lengths and tones, linguistics cannot and now should not have to substitute for the recording. Yet without linguistic information and skills the mere recording of an extinct language, no matter how good, would remain as indecipherable as an Etruscan inscription, so obviously there remains a crucial role for linguistics. In the Hebrew case, if all living


19

tradition of the language had been lost, the consonant transcription and comparative semitics would be the key. Obviously with both the technology and the improved linguistics, we should be very grateful to be today in such a good position to document endangered languages, very fortunately now good enough just in time to meet the challenge of accelerated language loss. We do indeed have a chance, if we get our priorities right, to minimize or greatly compensate for the loss in the form of “adequate” documentation which can preserve permanently a record of many or perhaps even most of the languages we stand to lose—whether we lose them or not as living languages being a different matter, treated above; the permanency of the record also being a different matter, for the policy and technology of safe archiving, also an important subject, which we must never fail to take to account.

1.8.1. Documentation Goals and Priorities Here I shall try to describe the linguist’s documentation role in a relatively narrow and traditional way, in fact the old-fashioned Boasian one of “grammar, texts, dictionary”, but first remark, of course, that though no doubt many endangered languages may not or may never have speakers who could become the linguists, or linguists who could become speakers, obviously such a combination is ideal. Linguists should certainly strive at all times to find and train speakers who have the motivation and talents to learn the linguistic skills, and themselves to learn the languages. I am also aware that there are other priorities in type of documentation, especially high-quality electronic recording of language use in a broad range of domains, which would have to concentrate more on languages still in vigorous use, such as those the Volkswagen Foundation has had meetings about these past few years in Germany, which have been very fruitful and stimulating. Here I confine myself to the narrower traditional view, which I justify with the following hypothetical claim: that a perfect student could learn from a perfect grammar and set (of some size and variety) of written texts, and a perfect dictionary, from those three books alone, how to speak perfectly an extinct language or language that student had never heard spoken. True, of course, no such perfection will be achieved, but what can one aim for? Surely we can come somewhere near, and even a remote approach can be precious indeed. Hebrew had mainly just the consonants, various oral traditions of vowels and prosody, no systematic grammatical or lexical elicitation, but only an extensive text which happens to have a remarkably wide

20


vocabulary. The oral traditions and assiduous cultivation of those, inspired by passion for the content of the text, are what made the revival of Hebrew possible, a language that had had no native speakers for most of two millennia. It is not wrong to compare the resources or potential resources for today’s endangered languages with those for Hebrew, in some important ways very favourably, in fact, so this is another consideration for the importance and potential of documentation, even very imperfect documentation. It may take twenty years, or two hundred, or it may never happen that documentation may result in language revival, but two things are completely clear, first that here again the linguistic and biological parallels do not hold, that an extinct language can at least hypothetically be revived, a species (so far) not, and second, that it is certain that no extinct language can ever be revived if it is not documented. Given today’s situation, number of endangered languages, and number of linguists able to work with them, prioritization on a wide scale is obviously necessary—especially with meetings like this, involving people and organizations who can play a major role in coordinating the prioritization. Obvious criteria are (1) degree of endangerment of a given language (from the Bad Godesberg meeting in 2000, February, we have been working on a scale of such degrees—cf. the Background to this volume); (2) quantity and quality (adequacy) of extant documentation, for which criteria need to be established; (3) genetic and/or typological isolation of the language (e.g. a dialect of a well-documented language or member of a well-attested branch of a family rated lower than a relative isolate, or language of a relatively unusual type); and one other type of criterion, namely (4) degree of accessibility, which could eliminate some languages ranking very high by the above criteria but which are situated in countries or districts where access or contact is not permitted or is too dangerous, threatening even the life of the linguist and/or speakers, often even in areas which are not war-torn. Finally, I shall offer a few thoughts on criterion (2) above, adequate documentation, and the financial costs thereof. Taking first dictionaries, what is a “perfect” or “complete” dictionary of a language? I believe such a dictionary would enable the (perfect) student (who also has a perfect grammar) to interpret correctly all words and all possible sentences or texts in its language, likewise to construct them. However, no finite corpus of texts could guarantee the occurrence of all possible lexemes, and a far more efficient way to assemble the dictionary would have to include extensive systematic elicitation on a culturally informed semantic basis, also on a phonological basis (elicitation through possible phonological sequences), and most efficiently


21

of all on a combined phonological and semantic basis, from lexica of genetically related languages, i.e. on a comparative basis, where possible. Obviously, even under the best of circumstances, no dictionary will ever be complete (though it may become so for the opposite reason, closure of the corpus by the extinction of the language). At the Alaska Native Language Center we have used the term “comprehensive dictionary” to mean fairly full coverage, such that in a new or newly discovered text in the language on any traditional subject(s), one would on the average probably find one word at most, say, every five pages which could not be correctly interpreted from the dictionary. In my experience, the minimum-size such dictionary, i.e. for the lexically minimum-size language, turns out to have about 6,000 lexemes. That is for languages of very small populations, spoken over very narrow geographical areas, with minimum dialectal variation, and no lexical expansion beyond the aboriginal (e.g. few or no trade-language loans, or Bible translations), such as my Eyak dictionary (mainly three speakers, of one village), or Larry Kaplan’s King Island Iñupiaq Eskimo (one village on a small island). Comparably comprehensive dictionaries of languages like Central Alaskan Yupik or North Alaskan Inupiaq, with base populations in the ranges of 10,000 to 20,000, and significant dialectal variation, might contain 10,000 or more lexemes, before including development like loans from trade languages or neologisms e.g. from pedagogical texts or Bible translations.

1.8.2. Priorities within Linguistics In 2000, I was in Washington to testify at a US Senate hearing on amending the 1990–2 Native American Languages Act now to support total-immersion type programmes. The hearing was chaired by Hawaii Senator Daniel Inouye, who asked me how much it would cost to produce such dictionaries for all remaining Native US languages, which required me to think very hard. Luckily in that connection I was also in contact with Katherine Ball, then head of the National Science Foundation linguistics programme, who provided me with some very useful information and also some very good news, which I’d like to share here. First, for some relevant historical background: when I first presented my sermon on the magnitude of the language endangerment crisis at the Linguistic Society of America plenary in January 1991, criticizing the discipline, then so heavily preoccupied with theory under the Chomskyan impetus, for “presiding obliviously over the disappearance of 90% of the very field to which it is dedicated”, one of the most important responses from the floor came from Paul Chapin, Kathy Ball’s long-time

22


predecessor as head of the NSF linguistics programme. He pointed out, very justifiably, that NSF had never stopped placing very high funding priority on documentation of endangered American (or other) languages (no matter how humble and marginal a role the theoretical mainstream had relegated such work to, as barely being “linguistics” at all). As a result of that 1991 meeting I found myself chair of a new LSA committee on endangered languages; I set the first major goal for the committee to be not to change the whole world but, more modestly, at least to influence the discipline to include much more concern for endangered languages and documenting them. NSF’s Kathy Ball told me that, in 2000, now by far the largest category of proposals NSF receives from American linguists is for the documentation of (especially American) endangered languages! So the last decade has indeed produced a profound change in priorities for American linguistics.

1.8.3. Costs of Documentation Perhaps it would not be inappropriate to conclude with some very practical, i.e. financial, considerations. Kathy Ball further told me that the average documentation (dictionary) proposal for an endangered US language was for $300,000 over three years ($100,000 a year, including all salaries, expenses, and maybe 40 per cent overhead charged by the typical university), and that they usually don’t get finished in those three years, and have to come back and ask for two years more, so it costs $500,000 for a dictionary per language. To answer Senator Inouye’s question, then, I estimated that out of 155 remaining US languages, about 100 have nothing like a comprehensive dictionary but have speakers from whom such could be assembled, so the total project at that rate would cost $50,000,000. Furthermore, the time-frame for that should be within the next ten years, as probably a third of the 155 languages will become extinct or at least severely reduced that soon, thus $5,000,000 a year for the next ten years. (That information now at least stands in the US Congressional Record, though I cannot say there has yet been any such appropriation!) Note that the above is for dictionaries only. Professor Osahito Miyaoka has been working for many years on a comprehensive grammar of Central Alaskan Yupik, so to find out about grammars I recently asked him how much more quickly such a grammar can be done than a dictionary, to which he of course answered “longer, much longer”. So we’ll have to agree for the moment that for a dictionary and a grammar, we should double the cost, to $1,000,000 per language. This does not include the texts, but for the moment let us at least imagine that that can be done “on the side”, and/or


23

also by others. Furthermore, the USA, Japan, and Germany are expensive countries to work in, so that the cost in dollars or yen or marks could be a lot less insofar as the work is conducted in other parts of the world, or on the economic scale of other parts of the world, where ideally much of the work we might hope will eventually originate and be based. In that way the average total per language might well be lowered to half that million, to begin with. Permit me, accordingly, since I’ve come this far, to indulge in a little more arithmetic and speculation. To get as much work done as possible, it would save significant time if a given linguist spends his or her documentation career on related languages, so that though the first language will cost $500,000 average, the next would go a lot faster, for maybe $200,000 average. If the linguist starts at the age of 25, finishes first language at the age of 35, and then spends an average of four years for five languages until age 55, that linguist can document six languages in a thirty-year career, costing $1,500,000, at $250,000 per language. I further speculate that if such work were done for 3,000 languages, properly prioritized, over the next thirty years, or with lead time and training time even over the next fifty, most of the moribund and more immediately endangered languages could be comprehensively documented. This would require a worldwide total of 500 linguists, costing $750,000,000 distributed over the next thirty years (at $25,000,000 per year), or over the next fifty (at $15,000,000 per year). Compared to the cost of certain military hardware and many other programmes, certainly compared to the gross national product of countries like the USA, Japan, or Germany, these language documentation costs seem quite reasonable. The more I think about it the more optimistic I become that funding on such a scale is indeed attainable. Granted, this does not allow for “a hundred linguists working a hundred years”, as every single language deserves, but it will make a vast difference in the record and future of mankind’s intellectual heritage—not a bad way for linguistics to serve.

References The following includes literature referred to in the foregoing, and also other relevant articles by the author, especially those available in Japanese. Dorian, N. (1993). “A response to Ladefoged’s other view of endangered languages”, Language, 69: 575–9. Grimes, B. (ed.), (2000). Ethnologue, 14th edn. Dallas: SIL. Harmon, D. (1995). “The status of the world’s languages as reported in Ethnologue”, Southwestern Journal of Linguistics, 14: 1–28. Krauss, M. (1992a). “The world’s languages in crisis”, Language, 68/1: 4–10.

24


Krauss, M. (1992b). “A loss for words”, Earthwatch, 11/3: 10–12. —— (1992c). “Status of Eskimo languages in world perspective”, in N. Graburn (ed.), Language and Educational Policy in the North. Berkeley and Los Angeles: University of California Press. —— (1993a). “The language extinction catastrophe just ahead: Should linguists care?” Actes du Xve Congrès International des Linguistes (Quebec), 43–6. —— (1993b). “Gengo zetsumetsu no kiki mokuzen ni semaru – wareware gengogakusha wa nani o nasubeki ka” (Japanese translation of variant of above), Gengo, 22/8: 12–7. —— (1994). “Language endangerment”, in O. Miyaoka (ed.), Languages of the North Pacific Rim. Hokkaido University Publications in Linguistics 7. Sapporo: Hokkaido University, 283–93. (Translated by F. Sasama into Japanese as “Gengo no kiki”, in Ainu-go no tsudoi – Chiri Mashiho o tsugu. Sapporo: Hokkaido shuppan kikaku centre, 249–63. —— (1995). “Language loss in Alaska, the United States, and the world”, in Frame of Reference (Alaska Humanities Forum), 6/1: 1, 3–5. —— (1996a). “The status of Native American language endangerment”, in Stabilizing Indigenous Languages. Flagstaff: Northern Arizona University, 16–21. —— (1996b). “Linguistics and biology: threatened linguistic and biological diversity compared”, Chicago Linguistic Society (CLS), 32: 69–75. —— (1997). “The indigenous languages of the North: a report on their present state”, (lead article) in H. Shoji and J. Janhunen (eds.), Northern minority languages. Senri Ethnological Studies 44. Osaka, 1–34. —— (1998a). “The scope of the language endangerment crisis and recent response to it”, in K. Matumura (ed.), 101–12. —— (1998b). “The condition of native North American languages: the need for realistic assessment and action”, Special issue of International Journal of Sociology of Language, 132: 9–21. —— (2000). Statement in Hearing before the Committee on Indian Affairs of the United States Senate. S. Hrg. [06–648], 106th congress, 2nd session, on S.2688 to amend the Native American Languages Act, to provide for the support of Native American language survival schools, 20 July 2000. US GPO 20000. 29–33, 111–28. Ladefoged, P. (1992). “Another view of endangered languages”, Language, 68: 809–11. Matumura, K. (ed.) (1998). Studies in Endangered Languages. Tokyo: Hituzi Syobo. Meillet, A., and Cohen, M. (eds.) (1952). Les Langues du monde, 2nd edn. Paris: Compion. Robins, R. H., and Uhlenbeck, E. M. (eds.) (1991). Endangered Languages. Providence, RI: Berg Publishers Ltd., published with the authority of the Permanent International Committee of Linguists. Voegelin, C., and Voegelin, F. (1977). Classification and Index of the World’s Languages. Amsterdam: Elsevier.

2

Documenting and/or Preserving Endangered Languages Bernard Comrie

2.1. Introduction 2.2. Language documentation and preservation 2.3. The case of Welsh and its implications 2.4. Language revitalization and linguistic diversity 2.5. Conclusions

25 25 28 32 34

2.1. Introduction This chapter originated as a “general remarks” response to the individual papers presented at the conference on endangered languages held in Kyoto at the end of 2000; some of these papers appear in revised form in the present volume. In my presentation, I have decided to concentrate on one topic that emerged recurrently in the papers presented, and this is the relationship between the documentation of endangered languages and the preservation of endangered languages.

2.2. Language Documentation and Preservation By documentation, I refer to archival materials on a language such that at least some aspects, and preferably as many as possible, of that language will be available to future generations even if the language dies out. By preservation, I mean prevention of the demise of a language by ensuring that it remains as a viable means of communication in a speech community. At first sight the relationship between preservation and documentation might seem to be straightforward, and indeed in my early thinking on this matter I would myself probably have adhered to the dictum that we should preserve wherever we can, and document in those cases where we cannot, no doubt allowing a certain amount of overlap as a kind of “insurance policy” in case we overestimate the viability of a particular language, or for purely scientific or educational reasons.1 The

1 After all, both general linguistics and mother tongue education have benefited from the “documentation” of languages like English and Japanese that are far from endangered.

26

bernard comrie

crucial aspect of this naive—as I would now call it—view is that preservation is seen as potentially pre-empting documentation: if we know that a language will be preserved, then it is at least in less urgent need of documentation. In my remarks here, I want to suggest some reasons why this view is misplaced, and why the apparent hierarchical relation between preservation and documentation may need to be rethought, or at least taken more carefully into consideration as we develop priorities for the documentation of endangered languages. Actually, the term “preservation” is perhaps a little too vague in this context, and I would like further to distinguish among three possible scenarios. The first, which we might refer to as preservation proper, is where a language is still the basic medium of communication of a community, acquired as first language by all or virtually all children, with no signs of any diminution of its role in the speech community. In such a situation, the question of an active policy of “preservation” hardly arises, since preservation of the status quo will simply be the continuation of the language’s healthy state, although of course if one can foresee future dangers from new technologies (television, Internet, etc.) to which the language of the community has not yet adapted one might want to take pre-emptive measures to ensure the language’s viability in these new domains. The second is revitalization, where a language is still spoken as a first language without attrition by some members of the community, but is not being widely acquired in this form by children—indeed it may even be the case that the middle generation consists primarily of semi-speakers and no children are acquiring the language without attrition. Here, the question of “preserving” the language does arise, in the sense of ensuring that, despite the possible generation gap, children once again start acquiring the language in its unreduced form, for instance from those of their grandparents’ generation, or from those of the middle generation who happen to have acquired the full version of the language—since such language acquisition would normally take place outside the immediate family, specific community structures, such as schools, would need to be initiated. Third, if a language no longer has any native speakers (or at least, any speakers of the language in its unreduced form), one might attempt language revival, i.e. reviving a language that has become extinct. Such revival does, of course, depend on good documentation, if the language is to be revived in anything like its original state: Features of the language that were not well documented will not be reproduced in the revived variety, except perhaps by chance.2 2 And in the absence of documentation, one will have no way of knowing which chance solutions are correct.

d o c u m e n t i n g a n d / o r p re s e rv i n g l a n g uag e s

27

Indeed, the role of such a revived language may be more in the nature of an “emblem” for the community in question than a resurrection of the language in its original form. This is not to belittle the importance of the emblematic function of a language, but rather to emphasize that a revived language may indeed satisfy this emblematic function without approaching a faithful continuation of the now extinct language. An example of this would be Revived Cornish in its relation to the original spoken language, which died out probably in the early nineteenth century. Since language revival depends crucially on documentation, rather than being a potential alternative, I will not consider language revival except briefly in what follows. Clearly, the recognition of three levels (preservation, revitalization, revival) is something of a simplification, given that in reality there can be any number of intermediate stages, and different parts of the same speech community may be at different stages. A more detailed typology can be derived from, for instance, Fishman (1991: 395). The concerns of the preceding paragraph bring up a related question, namely: for whose benefit is the documentation, revitalization, or revival taking place? First, it is clearly for the community that speaks or spoke the language in question, since without the cooperation of the speech community none of documentation, revitalization, or revival is possible, with the likelihood of success diminishing as we go through the list. And nothing in what I will say subsequently goes against the primacy of this requirement. But second, it is also for the benefit of linguistics as a science, since this science depends on access to linguistic diversity. There are at least two aspects to this. The one concerns linguistic structure, e.g. grammar, where only access to languages of different types gives us a chance of identifying the limits on the range of potential human cross-linguistic variation, and this is the area that probably most linguists think of first in this connection. But there is another, at least equally important area, especially now that linguistics is once again playing an important role in discussions about human prehistory, and that is that endangered languages often have an important part to play in resolving historical questions with implications going far beyond linguistics. Even the documentation of the lexicon of a language can go a long way to resolving questions of genetic affiliation and language contact, and we can only regret that the Etruscan dictionary which the Emperor Claudius is reported to have compiled is no longer extant. As shown by D. Bradley (§ 16, this volume), an endangered language may well provide crucial archaic phonological material that aids in our understanding of the history of a language family and its speakers.

28

bernard comrie

While linguists will surely never match geneticists’ success in retrieving DNA from 30,000-year-old Neanderthal remains, at least our discipline can make its contribution to our understanding of human prehistory by documenting, if not preserving, as much as possible. In the ensuing discussion, by documentation I will mean primarily documentation for linguistic purposes, while by revitalization I will mean revitalization for the purposes of the community.

2.3. The Case of Welsh and its Implications I want now to turn to a particular case study, where the potential for tension between documentation and revitalization does, I think, come to the fore. It deals with one particular language, Welsh, although it reflects a situation that is, I suspect, typical of many, if not most or even all, situations of language revitalization.3 My attention was drawn to the issues I am about to discuss by the detailed study carried out by Mari C. Jones on the revitalization of Welsh in two communities in present-day Wales (Jones 1998), although I should emphasize that my own interpretations with regard to the relation between revitalization and documentation are not to be attributed to the author of this monograph. It may be worth briefly summarizing the basic facts concerning the contraction of the Welsh language over the last two and a half centuries. Around 1750, Wales was still overwhelmingly a Welsh-speaking country (Jones 1998: 8), indeed overwhelmingly a monolingual Welsh-speaking country. By the 1991 census, the percentage of the population able to speak Welsh had fallen to 18.7 per cent, with perhaps the most significant drop between 1931 (36.8 per cent) and 1971 (20.8 per cent); there are now effectively no monolingual speakers of Welsh. Indeed, my own discussions around 1970 with relatively impartial observers of the linguistic situation in Wales suggested that the language might well die out as a basic medium of communication in the whole of Wales within one to two generations. In fact, in the period since the 1971 census the percentage of Welsh speakers has more or less stabilized, and there is even some evidence that it may actually be increasing, given that there are younger fluent speakers whose

3 For instance, Lynch et al. (2002: 28) refer to “the demographically swamped Maori and Hawaiian languages, which are spoken by many young speakers according to patterns that clearly reflect those of English”, and this despite substantial support for the languages in the local educational systems.


29

parents do not speak Welsh; the increase is largely attributable to the educational system, in particular immersion schools (Jones 1998: 15–24). Perhaps even more important is that the decline of the Welsh language is perceived as having been at least halted, if not reversed, so that Welsh-speaking society, and indeed Welsh society as a whole to the extent than even non-Welshspeakers regard the continued existence of the language as an important part of Welsh culture, consider the language to have been saved, i.e. as being in a situation where we can speak of an ongoing successful revitalization programme. It is clear that in terms of the emblematic value of the Welsh language to Welsh society, Welsh is no longer (barring future changes in attitude) an endangered language. But something else is clear from the detailed description of varieties of revitalized Welsh from Jones (1998), which indeed forms the bulk of the discussion in her monograph: the variety of the language that is emerging as a result of revitalization is one that lacks some of the distinctive characteristics of the Welsh language, including its various dialects, as this language was previously passed from generation to generation. I will mention just two examples. Welsh, like most European languages, has a system of grammatical genders, distinguishing two genders, masculine and feminine. Much as in French, the assignment of some nouns is readily semantically predictable (e.g. nouns denoting male humans are masculine, those denoting female humans are feminine), while more specific semantic and formal criteria can be adduced for ever smaller groups of non-human nouns, leaving a residue of nouns whose gender simply has to be learned. Different linguists will no doubt adopt different stances on just how many words have arbitrarily assigned gender, with respect to Welsh as with respect to other languages having grammatical gender that is not completely predictable on a semantic basis (as in English). But those who are acquiring or learning Welsh under the language revitalization programme bear witness to serious attrition of the gender system. One of the manifestations of the gender system is the presence of different third person singular pronouns, with the feminine pronoun hi being required not only for female human referents (English “she”), but also for non-humans (English “it”) that happen to be of feminine gender. In her study, Jones found that in the largely Anglicized Rhymney area only speakers aged 75 and over regularly used hi with feminine non-human nouns, while those under 75 varied between just under 40 per cent and scarcely over 0 per cent correct usage of the feminine pronoun under these circumstances (Jones 1998: 65). In the more strongly Welsh-speaking area of Rhosllannerchrugog, the dividing age between those who control this aspect

30

bernard comrie

of pronoun usage and those who do not is lower, at 40, but the difference between those 40 and above and those below 40 is equally striking (Jones 1998: 170). Other features related to the gender distinction show, if anything, an even clearer decline, in terms of adherence to the traditional norm, correlating with younger age, though some of these interact with another of the typical features of traditional Welsh. Welsh, like the other Insular Celtic languages, is characterized by initial “mutations”, i.e. changes in the initial consonant of a word depending on what precedes it or what syntactic position it finds itself in. Thus the word “father” is tad in isolation, but voices its initial consonant (“soft mutation”) in dy dad “your father” (and would also do so as direct object of a finite verb, whether or not adjacent to that verb), fricativizes it (“spirant mutation”) in ei thad “her father”, and nasalizes it (“nasal mutation”) in fy nhad “my father”. There are dialect differences within traditional Welsh as to the details of the mutations, and some dialects have lost the so-called spirant mutation, but all varieties of traditional Welsh have systematic use of mutations. Again, Jones’s study finds a significant drop in the correct usage of the mutations correlating with younger age, the drop being more marked in the Anglicized Rhymney district than in Rhosllannerchrugog, though even in the latter the 7 to 19 year olds failed to use soft mutation after the possessive pronoun ei “his” in about 60 per cent of cases.4 But perhaps the most “worrying” aspect of the disruption of traditional patterns in revitalized Welsh in the schools is that these disruptions not only characterize the speech of children from non-Welsh-speaking family backgrounds, but seem also to be feeding back though the school into the speech of those who do come from Welsh-speaking backgrounds (Jones 1998: 147–50), so that the disruptions of the traditional patterns come to characterize not only the speech of “new” speakers of the language, but rather the language as a whole in the usage of the younger generation. It should also be noted that most of the features examined by Jones are what might be regarded as relatively “superficial”, in the sense that they are easily recognizable and almost inevitably occur with considerable frequency in any reasonable stretch of linguistic performance; this is, of course, necessary given the need in her study to perform statistical comparisons of 4 Since they failed to make use of the aspirate mutation after the homonymous possessive pronoun ei “her” in about the same percentage of cases, the crucial use of mutation to differentiate “his” from “her” was thus significantly eroded. Incidentally, as these examples show, the mutations are not synchronically phonologically motivated, since homonymous items can behave differently as mutation triggers.


31

different speakers. One wonders what is happening, in revitalized Welsh, to more subtle features of traditional Welsh, such as some of the syntactic complications involved in subordination, for instance in complement clauses and relative clauses. There is another kind of change noted by Jones in the shift from traditional Welsh to revitalized Welsh, and that is the loss of dialect diversity. Indeed, Jones’s discussion of dialect-specific features for both Rhymney (Jones 1998: 90–103) and Rhosllannerchrugog (Jones 1998: 188–206) shows that most dialect features investigated are being levelled out in favour of a generalized South Welsh in the case of Rhymney, a generalized North Welsh in the case of Rhosllannerchrugog, with even some of loss of distinctions between North and South Welsh, for instance in the case of lexicon. This is, of course, something that is perhaps inevitable in the process of language standardization—witness, for instance, the amount of variation that has been lost within English dialects of Great Britain in the last 150 years—but it does perhaps point to an important aspect of the documentation/revitalization issue. In today’s world, a speech variety with a larger number of speakers has, other things being equal, a better chance of survival than one with a small number of speakers. This means that those interested in preserving or revitalizing a language to meet a reasonably full range of today’s communicative needs are almost inevitably pushed in the direction of greater standardization, of centralization with respect to regional variation, as seen in other cases parallel to Welsh, for instance in the development of Standard Basque. Such standardization also facilitates the task of teaching the language to members of the community who are not speakers of the corresponding language, since it provides a clear norm that can be used across the educational system, without having to provide materials for each regional variety of the endangered language. But this can also threaten to lead to the loss of structurally or historically interesting features, for instance of aspiration or pitch accent in Basque, both of which are found in some (not all) dialects, but are not part of the standard language. At worst, such standardization can lead to an artificial compromise variety that comes to be felt as characteristic of “new” speakers of the language and is rejected by traditional speakers. Indeed it may even lead, as in the case of Breton, to “new” speakers creating a whole batch of neologisms based on derivational possibilities (or impossibilities) of the language in order to differentiate the endangered language clearly from the language that is threatening to engulf it, but which are rejected

32

bernard comrie

by traditional speakers (Jones 1998: 301), thus creating an even greater rift between speakers of the traditional and revitalized languages. From Jones’s discussion, it seems that Welsh may have avoided the worst of this scenario, since at least some regional variation, in particular the rather clear divide between North and South Welsh, is being maintained. But it is equally clear that language revitalization is leading, among “new” speakers, to considerable loss of the variation that was once there. Once again, this factor may not be evaluated equally by linguists and by members of the speech community, given the former’s interest in diversity, although here the interests of the speech community may themselves be divided: loyalty to a particular language can be consistent both with a positive attitude towards dialect diversity (on the grounds that preserving a variety of the language is equally preserving the language) and with a negative attitude (on the grounds that dialect diversity can impede the resistance of the language to the forces that threaten it).

2.4. Language Revitalization and Linguistic Diversity But let us return to the main line of our argument, in particular the effect that language attrition is having on linguistic diversity. A counter-argument that may well be made, and has indeed been made in such cases as the revitalization of Welsh, is that the kinds of changes that are being observed in the passage from traditional to revitalized language are simply the kinds of changes that would be expected to take place anyway by normal language change, certainly by language change in a situation of language contact. Perhaps the changes are somewhat more rapid under revitalization. But since language change has characterized all known instances of living languages at all time periods, and since virtually all languages have almost always been in some situation of language contact, decrying the changes that have taken place between a traditional and a revitalized language would be tantamount to trying to stop language change. After all, English has gone much further than Welsh in the loss of grammatical gender, although grammatical gender was a full-fledged feature of Old English, and this without any threat of endangerment. According to Ross (2001), Takia, an Oceanic language that shares the island of Karkar (Papua New Guinea) with the Papuan language Waskia, has almost completely remodelled its syntax after that of its nonAustronesian neighbour—a process that Ross calls “metatypy”—while retaining its Oceanic lexicon and morphology, and its distinct identity. And


33

Maltese, historically an offshoot of vernacular Arabic, has lost almost all the idiosyncrasies of the phoneme inventory of Arabic (almost any other variety of Arabic), no doubt in large measure through close contact with speakers of Romance languages, though without losing ground as the basic medium of communication on the Maltese archipelago. Might not the radical simplifications in the phoneme inventory of revitalized Lummi, as documented by Timothy Montler and mentioned by H. Watanabe and F. Sasama (§ 22.5) be placed under the same heading? As an individual phenomenon, perhaps, but not in terms of the overall effect on the world’s linguistic diversity. At present, and in contrast to the situation during the rest of the past tens of thousands of years of human linguistic history, the languages threatening to engulf endangered languages are relatively few in number and encompass a remarkably narrow range of known cross-linguistic diversity.5 In previous periods, regular language change gave rise to some of the highly idiosyncratic features of particular languages that have played a major role in the construction of subparts of linguistic theory, such as clicks in the Khoisan languages and near-consistent syntactic ergativity in such languages as Dyirbal (northeastern Australia). Indeed, language contact even permitted clicks to spread from Khoisan to neighbouring Bantu languages. What we are witnessing now is the loss of idiosyncratic features in favour of the levelling influence of the world’s dominant languages. A hard-hearted scientist might not shed too many tears over the loss of grammatical gender in Welsh, since there are still plenty of other languages that provide rich information on grammatical gender systems, including major languages like French and German. But the loss of the initial mutations should cause more concern, since there are rather few languages that have rich systems of morphophonemic alternations of initial segments, especially such as are as far removed from their phonetic-phonological origins as in the case of Welsh. Basically, the well-documented cases are the other Insular Celtic languages, all of them more seriously endangered than Welsh, and a number of West African languages of the Atlantic branch of Niger-Congo—in addition to some Austronesian languages whose documentation is in general at a less advanced stage.

5 Note that even in the prototypical laboratory of linguistic diversity, Papua New Guinea, for which Ethnologue (www.ethnologue.com, consulted on 31 May 2005) lists 820 languages, according to the same source ten of these languages are extinct, while a further twenty-two (including one creole) are “nearly extinct”.

34

bernard comrie

2.5. Conclusions Let me now try to draw some conclusions from the above discussion, in particular with regard to linguists’ work on endangered languages. Perhaps the most salient lesson for linguists is that the revitalization of a language does not obviate the need for documentation of the traditional language, since a revitalized language may differ quite extensively from the traditional language to which it corresponds, in particular through the loss of precisely those distinguishing features that make the traditional language of such paramount importance to linguistics. Let me emphasize that by this I do not mean to construct an argument against language revitalization (or revival); indeed I believe that the emblematic value of a language to its community is an important part of the cultural legacy of that community and one that should not be controlled by outsiders, not even in the name of science. But the present discussion does suggest that even in cases of language revitalization, even in a case as successful as that of Welsh, documentation of the traditional variety should remain a matter of the highest priority.

References Fishman, J. (1991). Reversing Language Shift. Multilingual Matters 76. Clevedon: Multilingual Matters. Jones, M. C. (1998). Language Obsolescence and Revitalization: Linguistic Change in Two Sociolinguistically Contrasting Welsh Communities. Oxford Studies in Language Contact. Oxford: Oxford University Press. Lynch, J., Ross, M., and Crowley, T. (2002). The Oceanic Languages. Richmond: Curzon. Ross, M. (2001). “Contact induced change in Oceanic languages in north-west Melanesia”, in A. Y. Aikhenvald, and R. M. W. Dixon (eds.), Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics. Oxford: Oxford University Press, 134–66.

3

Linguistic Fieldwork among Speakers of Endangered Languages Colette Grinevald

3.1. Introduction 3.2. Working on endangered languages: linguistic description at the core 3.2.1. Basic linguistic descriptive work 3.2.2. Language documentation projects 3.2.3. Preservation-revitalization projects 3.2.4. Conclusion: multiple demands but language descriptions at the core 3.3. Issues embedded in the complex situations of fieldwork on endangered languages 3.3.1. Considering fieldwork in time line: past–present–future 3.3.1.1. Past 3.3.1.2. Present 3.3.1.3. Future 3.3.2. Dealing with loss 3.3.2.1. Loss of natural context for language learning and participant observation 3.3.2.2. Loss of a sense of norm, and increased variation in the language 3.3.2.3. Loss of choice of speakers 3.3.2.4. Loss of linguistic confidence of certain types of speakers 3.3.2.5. Loss of speakers retold 3.3.2.6. Loss of the speakers 3.3.2.7. Conclusion: a pervasive sense of loss 3.3.3. Dealing with language attitudes 3.3.3.1. Helping revalorize the language and the speakers 3.3.3.2. Dealing with normative attitudes of language activists 3.3.3.3. Facing language ownership 3.3.4. Conclusions: on the extra complications of fieldwork on endangered languages 3.4. Working with speakers of endangered languages

36 38 39 39 40 41 42 42 42 43 43 44 44 45 45 45 45 46 46 46 47 47 48 48 49

Colette Grinevald was previously Colette Craig. The title of this chapter was suggested to me by Nancy Dorian whom I wish to thank here for her generous sharing of ideas and materials, and whose pioneering work in the field of endangered languages, including its issues of fieldwork, I wish to acknowledge here. I also want to thank Roberto Zavala for all the brainstorming time and effort he invested in the production of the original version of this chapter. While I think of myself simply as a spokesperson for the fieldworker colleagues from various continents with whom I know I share the concerns expressed here (in particular North American, Latin American, European, and Australian colleagues on career tracks parallel to mine over the last decades), I am also sure others could have been more eloquent and I will therefore take full responsibility for the likely awkwardness and roughness of my own statements. What should be clear is that the issues raised here need to be integrated in any public debate on endangered languages in the interests of those who might consider joining in the work.

36

c o l e t t e g r i n e va l d 3.4.1. About speakers of endangered languages 3.4.1.1. Typology of speakers of endangered languages 3.4.1.2. About a typology in the first place 3.4.1.3. Types of speakers 3.4.1.4. Projected revisions to the typology 3.4.2. Working with this great variety of speakers 3.4.2.1. Counting the last speakers 3.4.2.2. Evaluating speakers’ knowledge of the endangered language 3.4.2.3. Collecting data from speakers of endangered languages 3.4.2.4. Conclusion: Adapting methodologies for data collection and analysis 3.4.3. About linguists working on endangered languages 3.4.3.1. A certain personality profile 3.4.3.2. Energy, time commitment, and professional risk 3.4.4. An invitation to take up the challenge 3.4.4.1. A collective responsibility 3.4.4.2. More about why we do it 3.4.4.3. A last warning

3.5. A case study of fieldwork on an endangered language 3.5.1. The sociopolitical circumstances of the project 3.5.1.1. Regional autonomy project and linguistic rights in time of war 3.5.1.2. The Rama people and the Rama language 3.5.2. The search for speakers 3.5.2.1. Finding three Rama speakers to work on the Rama language 3.5.2.2. The different skills of the three main consultants 3.5.2.3. Looking for more speakers 3.5.3. Trying to work with more Rama speakers 3.5.3.1. Other native speakers that did not work out as language consultants 3.5.3.2. Working at the margins: rememberers and semi-speakers 3.5.3.3. Cycles of expectation and frustration 3.5.3.4. Dealing with a major contradiction for the language revitalization part of the project 3.5.4. Some figures: time, (wo)manpower, and grant support 3.6. Conclusions

49 49 50 50 52 53 53 53 53 56 57 57 58 59 59 59 60 60 60 61 61 61 62 62 64 65 65 67 68 69 70 71

3.1. Introduction Beyond being convinced of the importance of documenting the diversity of the world’s languages before it is too late, and beyond advocating the involvement of the linguistic scholarly community in the task, it is

s pe a ke r s o f e n da n g e re d l a n g uag e s

37

important that we also address the various dimensions of the nature of the fieldwork enterprises for such a task. In the wake of a newly orchestrated dedication to carrying out linguistic work on endangered languages, it would seem essential to consider some of the specifics of fieldwork in such circumstances to be taken into account in the planning and carrying out of such field projects. The position taken here is that while linguistic fieldwork is never an easy task, it happens to become, more often than acknowledged, an especially complex endeavour in the particular case of fieldwork on endangered languages. By considering here some aspects of the fieldwork part of the enterprise, the hope is to contribute to what Fishman (1991) has called the need for “intellectualizing” the developing subfields of linguistics that concern themselves with endangered languages, from their documentation and description to their potential revitalization. The focus here will therefore be on the relation of linguists to the practice of fieldwork on endangered languages, and in particular on some of the psychological, strategic, and methodological dimensions of such types of projects, highlighting certain aspects of the relation between linguists and speakers of endangered languages. The chapter will consider some essentials of projects dealing with endangered languages, such as the fact that (1) field linguists working on endangered languages today often find themselves involved in field projects of wider scope than just the linguistic description they feel best prepared to handle; (2) the complexity of endangered language field situations means dealing with a multidimensional sense of loss, and diverse and strong language attitudes; and (3) working with the many types of speakers of endangered languages leads to a reconsideration of data-collecting methodologies seldom carried out in field methods courses of linguistics departments. The chapter will close with (4) a case study of such a project with the Rama language of Nicaragua, a Latin American country of the Pacific Rim (cf. Ch. 10, this volume), illustrating some of the major points raised earlier. It might be worth underlining from the start that much of what will be made explicit here will most likely appear to be no more than common sense to many experienced fieldworkers familiar with this type of field situation and sensitive to their particularities. But it is assumed that, for readers of vital language communities unfamiliar with such situations and curious of them, articulating what some of this common sense consists of, and what it is meant to respond to, is worth putting down in writing.

38

c o l e t t e g r i n e va l d

3.2. Working on Endangered Languages: Linguistic Description at the Core The field of “endangered languages” has seen a rapid expansion in recent years, and several major syntheses of general aspects of the issue have come out in the last few years (see for instance Grenoble and Whaley 1998; Crystal 2000; Nettle and Romaine 2000; Hagège 2000). Linguists have also started recounting their field experiences (see papers in Newman and Ratliff 2001 and Austin 2003, for instance) as part of a more or less concerted effort at reflecting on the nature of linguistic fieldwork, in particular fieldwork on endangered languages.1 It is now fairly common for linguistic fieldworkers working on endangered languages to find themselves involved in a variety of types of projects in which the actual component of the linguistic description of the endangered language is considered as more or less central. As opposed to traditional fieldwork of the past century that by and large was concentrated on the single track enterprise of a linguistic description, today the enterprise of producing a mere linguistic description is often embedded into a wider scope project. And although some academic programmes are now gearing up to handle such challenges,2 it remains that most field linguists today are more or less prepared to deal with, or embrace such wider scope projects. This expansion of the scope of many fieldwork situations on endangered languages may actually be more pronounced and widespread in certain parts of the world than in others. It certainly is a common condition of field situations encountered today in the American continent, from its northern to its southern parts (including the Pacific Rim side of the continent, from Alaska and the west coast of Canada and the USA, down along the various Latin American states facing west, such as Mexico, Guatemala, Nicaragua,

1 This chapter is one of a series of papers by the author on the theme of fieldwork on endangered languages, in the context of Latin America in particular. Craig (1993) was an early consideration of the ethical issues of such fieldwork, Grinevald (1998, 2000) an earlier discussion of the relation of foreign linguists to national and regional institutions and their linguists, as well as their own academic institutions, and Grinevald (2003a) was a brief treatment of the variety of speakers of endangered languages, with an introduction of the notions of fieldwork as an art (as per Wolcott 1995) and of fieldwork frameworks defined by the power relations established between field linguists and the community of speakers of those languages. 2 Such as the Hans Rausing Endangered Languages Academic Program of London started at SOAS in London in the autumn of 2003.


39

Colombia, Ecuador, and Peru, all the way to Chile—Chs. 9–11 and 21–24, this volume).3 The point to be made here is that no matter what the expanded scope of most projects on endangered languages today, the linguists will always be the academic researchers primarily responsible for the analysis of the linguistic structure of those languages.

3.2.1. Basic Linguistic Descriptive Work Basic linguistic descriptive projects are the kind of projects most familiar to field linguists and the ones most easily validated by the linguistic profession. They involve primarily work in synchronic linguistics which typically (ideally) deals with the triad: grammar + texts + dictionary (GTD). Since the linguistic description of the endangered languages will always have to be the most original contribution of the linguists, who are the only professionals trained for this work, this chapter will focus on this admittedly narrow scope but essential and unique contribution of the linguistic profession. It is from this narrower focus that a sketch of some of what there is to think about when doing fieldwork on endangered languages will be considered here. Such a choice of focus is mostly strategic and certainly does not mean to underestimate other goals; it is a reminder not to forget the challenges of basic linguistic fieldwork at the core of all projects.

3.2.2. Language Documentation Projects For definitions of “language documentation”, see for instance Himmelman (1998) and Woodbury (2003). It would seem that the impetus for the development of this new type of field project stems as much from the increasing availability of new documenting and archiving technologies as from an increasing awareness of the rapid loss of much of the linguistic wealth of the world. The two are now thoroughly intertwined, with the technological part receiving perhaps more attention today than the human relations side of the enterprise considered here.

3 See Grinevald (1997) for an overview of the situation of language endangerment in South America, and the start of a discussion on the possible relations, in that part of the world, of foreign linguists to national linguists and institutions, and to language communities. See also Queixalos and Renault-Lescure (2000) for a thorough introduction to the situation of Amazonian languages today by country, in which Grinevald (2000) is an attempt at articulating the nature of the conflicting pressures put on fieldworkers by their academic ties on the one hand and their commitment to language communities on the other.

40


Today, the term “language documentation” seems to cover two conceptions of field projects, which may be distinguished by the scope of the enterprise and the relation that holds between language documentation and linguistic description. A narrower scope approach of description-for-documentation may take the form of an edited and annotated version of the field database which has been collected primarily for the production of the traditional triad grammar/texts/dictionary, while a wider scope approach of documentation-for-description means a radically expanded primary data collection, aided by the descriptive activity of linguists but essentially carried out by a multidisciplinary team of fieldworkers (linguists, anthropologists, ethno-botanists, musicologists, historians …).4 The position taken here is that these two approaches should be viewed as successive cycles of a major process, one that naturally starts with an initial documentation that produces an initial description, this description becoming essential for a wider type of documentation, which itself allows for a more sophisticated and more comprehensive description, in ever widening and deepening cycles. Proposals of documentation projects need to be assessed on the basis of what is feasible for a particular situation at a particular time, a more encyclopedic documentation only conceivable on the strength of pre-existing extensive linguistic description, and long-term working relations of the linguist with (members of) the community that have active participation in the project.

3.2.3. Preservation-revitalization Projects The third type of project in which field linguists working on endangered languages may find themselves involved today is language preservation and revitalization projects, which are, at best, generated and managed by the linguistic communities themselves. See for instance the collection of articles produced by two of the leading North American linguists involved with such projects, and pointedly entitled The Green Book of Language Revitalization in Practice (Hinton and Hale 2001).5

4 Nothing will be said here of the newly developing field of archiving, but it is to be kept in full view as a complementary component of all documentation projects today. 5 Ken Hale, the MIT linguist who helped create summer linguistic institutes for speakers of indigenous languages and trained the first Native American speakers through Ph.D.s in linguistics, and Leanne Hinton, from the University of California at Berkeley, and her Master Apprentice Program for the native languages of California. For a sense of the diversity of ongoing language revitalization programmes in which linguists are involved, see for instance the SSILA newsletter.


41

As convincingly discussed by Gerdts (1998) and clear from general work on endangered languages by Nettle and Romaine (2000) and Crystal (2000), the role of linguists in the overall dynamics of such projects may need humbling re-evaluation and readjustment, even though, and once again, one must keep in mind that the original and indispensable contribution of linguists remains the analytical study of the language. It may well be that, in such contexts, the most productive approach to the description of the language is one channelled through the training in descriptive linguistics of linguistic community members, for self-sustaining language work of the kind that can be of use to the community. This means that the field linguists double up as linguistics teachers, or are hired actually as full-time teachers and supervisors of linguistic work done by speakers themselves (the with and by fieldwork frameworks mentioned below on p. 44 at n. 9).6

3.2.4. Conclusion: Multiple Demands but Language Descriptions at the Core The main point of this first section was therefore that linguistic fieldwork on endangered languages may well be cast today within more encompassing documentation and revitalization projects, in which case one of the major challenges for the linguistic fieldworker is to manage a demanding balancing act between multiple demands. This issue has been vividly described by Nagy (2000), who describes her fieldwork experience as wearing different “hats”, among which are the sociolinguist hat, the theoretical linguist hat, the applied linguist hat, and the “techie” hat, with others yet to be considered. This balancing act may well be in fact one of the major field issues to face for linguists working on endangered languages today.

6 See the example of Guatemala and the development of a cohort of Mayan native linguists, and the dedication of a field linguist like Nora England to the training of Guatemalan native Mayan linguists: this included contribution to the production of pedagogical material (England 1992b, 2001) and the creation of a native linguistic research institute, OKMA. See also Yasugi (2003) for ELPR-financed volumes of linguistic materials based on a questionnaire (that of the Archive of Indigenous Languages of Mexico organized by Swadesh) produced by the linguists of OKMA. The special case of linguistics in Guatemala is described in England (1992a, 1997) and Grinevald (2003b). The training effort has continued into the creation of CILLA (the Center for the Indigenous Languages of Latin America— N. England director), to provide graduate training for the likes of the indigenous graduates of university programmes in Latin America (for Mayan speakers, the programmes of CIESAS in Mexico, or of the Landivar and San Carlos Universities of Guatemala).

42


3.3. Issues Embedded in the Complex Situations of Fieldwork on Endangered Languages Unlike other disciplines relying on fieldwork, such as anthropology, ethnology, or sociology, linguistics has not had a tradition of discussing much the phenomenon of fieldwork. What is generally lacking is an addition to the standard course on “field methods” with some discussion of the wider context of doing “fieldwork”. This becomes crucial when one is heading for fieldwork in the midst of endangered language communities, with all the specific extra complications such projects may entail. Among the themes that could be raised in such fieldwork sessions, three will be considered below: the importance of grasping the dynamics of a fieldwork project in a time line, from past through the present to the future; understanding and anticipating having to deal with various forms of loss sometimes in unexpected and powerful ways; and the need to cope with the sometimes insidious impact of particular language attitudes on even the simplest linguistic elicitation sessions.

3.3.1. Considering Fieldwork in Time Line: Past–present–future This is a basic issue to always take into consideration when doing fieldwork, but one that is not very familiar to linguists: to become aware of the community’s past experience with linguistically oriented outsiders, of its present concerns and activities about the language, and to work for some possible continuation of the work in the future by the community itself or other outsiders on other projects.

3.3.1.1. Past With respect to the past, the issue is that one may or may not be the first one in the field. Earlier on, when fieldwork was basically a very individual enterprise and field linguists were very scarce, basic attention to this issue consisted in checking whether another academic had already done or was still doing fieldwork at the same site.7 Today, it is best to assume that communities have already had experience with any number of development projects, including some dealing with the language in some fashion. The community’s memory of our predecessor(s) has potentially either a positive or a negative 7 In the case of Latin America one factor considered in the choice of site was and remains to a great extent the established presence of missionary Bible translator linguists, for instance.


43

impact. It can actually take a while to figure out, as communities often do not share that information at first with new outsiders, particularly when negative feelings linger on.

3.3.1.2. Present One needs to assess the level of vitality of the language, which is not a simple endeavour, if one considers all the variables judged important (as spelled out in UNESCO 2003). This includes evaluating the level of ethnic consciousness of the community and the level of politicization of its relation to the language. All field projects can benefit from such an assessment, to better evaluate the desirability and feasibility of description and documentation, as well as revitalization efforts. As will be seen below in § 3.4 on speakers of situations of endangered languages, defining the linguistic community of such languages and simply counting its members is already a challenge.

3.3.1.3. Future In the case of work on endangered languages, it is important to consider how one may be the one and only, and, crucially, the last one to work on that particular language. If one is to be the one and last linguist to work on an endangered language, and one should always assume so, it means that the data one will be collecting may well be all that there will be of documentation of the language, unless some native people can be trained to continue collecting material after the departure of the field linguists. Therefore, as spelled out by Mithun (2001), a major issue to keep in mind in collecting data is that one cannot tell what will be of theoretical interest later in the field of linguistics, which means that one becomes accountable for collecting all the data one can, even data in which one may not be personally interested because of one’s own theoretical leaning and interests. A future perspective in terms of the community also means considering the sustainability of the work done on the language, through empowerment of members of the community, particularly in the form of continued training of speakers and semi-speakers capable and interested, and participation and support to the production of language materials, with a view to producing material that is actually usable in the field and by the community.8 8 This author can be critical of too much effort, energy, and funding being spent on the production of technologically highly sophisticated materials of no use at all to communities, in plain truth, for lack of access to electricity, or computers, or computers with the kind of memory capacities required by programmes first world academics have created, or simply for lack of able bodies to be trained locally in using them.

44


Such concerns are now integrated in ongoing discussions of what has been labelled as “good practice”, including legal, ethical, and practical issues in a sociopolitical context of support to sustainable community development.9

3.3.2. Dealing with Loss Working on endangered languages has another dimension rarely mentioned in the literature: that it means dealing with ongoing loss. The notion of loss is pervasive in fieldwork on endangered languages, both in a practical and in a psychological, or even an emotional sense. The sense of loss may take many shapes, all with some impact on the experience of fieldwork. This makes the work essentially different from fieldwork in dominant language communities where language is everywhere, choice of speakers is wide open, participant observation is easy to come by, and attitude of speakers toward their language is one of relative confidence. There are losses that have to do with the language to be studied itself, and others that are more tied to the persons of the speakers.

3.3.2.1. Loss of Natural Context for Language Learning and Participant Observation The loss of varieties of language due to the loss of contexts of use, which is the other side of the phenomenon of “language shift”, means fewer opportunities to capture the language in its various forms. It becomes from difficult, to impossible, to record certain varieties in their natural settings, since, by definition, fewer children are learning it—if any at all—fewer elders are passing on the traditional culture, fewer ceremonies are performed so that fewer traditional performing arts can be documented. The loss of the critical mass of speakers necessary to maintain a vital linguistic community translates into less of a chance to observe the language in use, to hear it in its natural use, to learn it by immersion, to practise it. In general, there are fewer opportunities, often no more opportunities, for the last speakers to gather, certainly no more traditional night gatherings typical of winter nights in many places. 9 The notion of “good practice” corresponds to the “empowerment’” framwork discussed in Cameron et al. (1992), who summarized the progression of fieldwork frameworks in the course of the second half of the 20th century in the formula of fieldwork having been earlier on a language, then later on a language and for the community, to develop into fieldwork on a language and with the community, where these frameworks are labelled the “ethical”, “advocacy”, and “empowerment” frameworks, respectively. Actually a further model of work on a language but by the community is the one requested by more and more indigenous communities today, in America at least, as discussed in Grinevald (2000, 2003a).


45

3.3.2.2. Loss of a Sense of Norm, and Increased Variation in the Language These are losses typical of those situations that linguists can be very sensitive to. Normally oral tradition languages naturally function with a certain amount of variation that is already hard to handle for linguists, who come from normative language traditions and have been raised in linguistic traditions that primarily consider data of standardized long written tradition languages. But in the case of language obsolescence an additional layer of variation pervades everything: it is the variation caused by the lack of norm enforcement from lack of vitality of the speaker community.10

3.3.2.3. Loss of Choice of Speakers The limitations are obvious when there are few speakers left, and among them older people with physical ailments, although the limitation in number is often compensated by the fact that some of those last speakers are extremely attached to their language, eager to work with a linguist, and excellent speakers. One can evoke for instance the talent of the extraordinary California Yahi man they called Ishi, the last survivor of his tribe and last speaker of his language, who was the unique source of a very rich documentation of his language and culture (Kroeber 1961). Working with as many of the last speakers as possible can help piece together what is left of the language and the knowledge it conveys. It has often been observed that what remains of knowledge of a language can be distributed across speakers, so that it might take working with multiple semi-speakers to complete the study of some aspect of the language.

3.3.2.4. Loss of Linguistic Confidence of Certain Types of Speakers This is a major characteristic trait of the semi-speakers of very endangered languages, as will be considered below in the section on types of speakers of endangered languages. This type of loss has a direct impact on the conduct of elicitation sessions, as it becomes obvious that simple questions may trigger renewed sense of loss, of shame, or confusion in the speakers.

3.3.2.5. Loss of Speakers Retold It is probably worth mentioning another aspect of the sense of loss likely to shock the unprepared fieldworker. It is the possible traumatic catharsis 10 For a discussion of the nature of the multidimensional linguistic variation inherent to such a situation of endangered languages, see Dorian (1982, 1986, 2001 in particular).

46


effect of asking last speakers of some language for some personal narratives, and of hearing tales of decimation of their people. One can think of the haunting voice of Ishi recorded on wax cylinders telling the story of the successive massacres that decimated his people. Such an experience can still happen today, as happened, for instance, on several occasions during the campaign for the “normalization” of the alphabets of the Amazonian languages of Bolivia in 1995 and 1996, when several groups of speakers of seriously endangered languages recalled from living memory episodes of the decimation of their people. This is not what graduate students in linguistics usually expect to hear when they ask for a simple personal narrative from native speakers in order to have textual material to study the basic morphosyntax of the language.

3.3.2.6. Loss of the Speakers And there is always the ongoing loss of the old speakers to whom the linguists had become attached as if they were family members, and on whom they were dependent in order to learn about the language. There are common cases of plain mourning one must deal with, but there is also the added dimension for linguists of mourning the death of the language itself. And before the actual death of the old speakers and of their languages with them, there may be painful dealings with the loss of memory or language ability that can accompany old age and ill health.

3.3.2.7. Conclusion: a Pervasive Sense of Loss There is therefore a high likelihood, during fieldwork on endangered languages, of being confronted at some point or another with any combination of these feelings of loss. They may be diffuse enough for us not to be able to put labels on them but their accumulation definitely gives to this kind of fieldwork a very different personal dimension. These feelings of loss can be acute and dramatic when speakers die, but also when our questioning and listening to some of them gives them an opportunity to express the depth of some of their own emotions about it, making us direct witnesses of the human and personal tragedy of language death in some parts of the world.

3.3.3. Dealing with Language Attitudes Actually one always has to deal with language attitudes when working on languages, but these attitudes are not always easy to identify either. In speakers of dominant languages they tend to be a cluster of pride in the language


47

and quiet personal assurance of its value and claimed superiority. But with speakers of un-empowered languages, including those of endangered languages, the attitudes are often very mixed: negative attitudes absorbed from the dominant culture that can lead to serious alienation, combined with often hidden positive attitudes that underlie a strong need to identify with that language. Maintenance and revitalization projects must take into account these attitudes because it is clear that they determine, more than scientific studies and modern technologies, the quality of the documentation and the ultimate fate of the language. Dealing with language attitudes can take many forms.

3.3.3.1. Helping Revalorize the Language and the Speakers Linguists contribute to the revalorization of the language itself, by their own scientific work, the proof that the language can be written, is worth studying and is rich in grammar and vocabulary. This revalorization of the language must be addressed both to the community members themselves and to the members of the dominant culture. But revalorization of a language must also pass through the conscious revalorization of its last speakers. Language endangerment created by language shift creates a varied population of speakers with different levels of bilingualism and degrees of alienation to the dominant culture and to their own language and culture. In such situations, a discriminatory attitude is common toward the last speakers of an endangered language, often seen as backward, if not primitive, partly for speaking a denigrated language. If they work with linguists, these speakers can additionally be considered as traitors to the community, accused of selling a good that belongs to the community, even when the community has very mixed feelings about that good. Such contradictory attitudes of non-speakers in a love-hate relation to the language are a disturbing and frustrating attitude for the Cartesian minds of most linguists.

3.3.3.2. Dealing with Normative Attitudes of Language Activists Another aspect of language attitudes that linguists have to deal with is the double challenge of handling the natural variation of the data collected in situations of endangered languages while being confronted with the extreme normative attitudes of some of the leaders of the revitalization projects. Those leaders are generally among the people with the most formal (colonial) education in the community, and they have identified the notion of a language with that of a set of norms to be taught, as was taught to them in the

48


dominant colonial language. This normative bend is further complicated by often acrimonious discussions of whose language is the most representative and therefore should be documented and then hopefully taught, between that of the old traditional speakers and that of the young fluent speakers. Attitudes toward language change can also be a real challenge for the field linguists caught between community factions, and navigating between the puristic and folklorizing tendencies of some and the claimed needs of others to modernize the language in order to make it adapt to the present world of its speakers.

3.3.3.3. Facing Language Ownership More extreme yet, one may also encounter community attitudes towards the language that will severely limit data collection, such as communities that have a very developed sense of language ownership and that do not want to teach their ethnic language to outsiders. This is known of some linguistic communities of the south-west of the USA, and is also the basic attitude of some Amazonian communities, for instance.11 Language ownership issues in terms of diffusion of information on the language are also now systematically and pressingly raised by the new archiving projects and use of web technologies.

3.3.4. Conclusions: on the Extra Complications of Fieldwork on Endangered Languages The point of this section of § 3.3 was therefore to reveal some of the hidden dimensions of fieldwork on endangered languages that can make this kind of linguistic work more challenging than most outsiders can envision. The issues faced by field linguists are varied, from understanding their place in the history of the native community’s dealings with outsiders, to facing, sometimes in the most intimate and personal way, the multiple aspects of loss characteristic of such work on endangered to moribund languages, to 11 Ospina Bozzi (2002) is such an instance of fieldwork with one of the last nomadic groups of the Colombian Amazonian region, the Yuhup Maku. It was a very difficult monolingual fieldwork setting that demanded she follow tenaciously the group moving through the forest, leaving her no other recourse than integrating herself in daily activities of subsistence and learning to speak the language by observation and self-teaching. It is difficult to conceive of some major documentation project with teams and cameras under such circumstances. The grammatical analysis found in the doctoral thesis is the result of years of contact with the group and numerous field trips, some of them having been unproductive because the group could not be located in its displacements in the forest: conditions that render any field project difficult to time, and yet the language presents extremely interesting linguistic characteristics well worth any effort at describing it.


49

being confronted with the realities of complex and commonly very contradictory attitudes of members of the community, toward speakers, toward language variation and language change, and toward outsiders. A further issue for field linguists is to be confronted with the wide variety of types of speakers one encounters in such situations, which is the topic of the next section.

3.4. Working with Speakers of Endangered Languages As already mentioned in passing, in a situation of endangered languages, not only are there fewer and fewer speakers but, in addition, there are many different types of speakers, of the kind that are not found in situations of full vitality of a language. Following Dorian (1982), it is in addition the position of this chapter that field linguists working with communities of endangered languages should consider taking a wider rather than narrower scope and conception of what constitutes the linguistic community at hand, and include in the process of documentation and description as many types of speakers as possible, including those that, from our common experience with vital language communities, would appear to be at the margins of the linguistic community. They too have their place and their contributions to make, particularly in contributing to the basic social fabric a language needs to survive.

3.4.1. About Speakers of Endangered Languages Speakers of endangered languages are characterized by particular traits which linguists need to take into account in order to handle appropriately the various types of interactions likely to arise, and to identify better the nature of the data they are collecting.

3.4.1.1. Typology of Speakers of Endangered Languages Speakers of a vital language normally present great diversity in their knowledge, attitude, and talent for working on their language. All field linguists know that some speakers can be superb linguistic consultants, while working with others can be difficult, sluggish, and frustrating. The situation is always much more complex when dealing with speakers of endangered languages, both because of the inherent limitation of choice of speakers with whom to work that has already been mentioned, but also, as this section intends to point out, because of the types of speakers one is likely to encounter in these

50


situations. What follows is a quick consideration of the types of speakers of an endangered-language-speaker community and the ways in which this diversity relates to the process of data collecting.

3.4.1.2. About a Typology in the First Place Various attempts at building a typology of speakers of endangered language communities are available in the literature, such as Campbell and Muntzel (1989), Dorian (1982, 1989), Dressler (1978), and Sasse (1992). Grinevald Craig (1998) was an overview of this literature at that point. The intrinsic difficulty in establishing a workable typology of speakers resides in the nature of the linguistic community, in particular in the effects of the progressive state of decay of the linguistic social networks and of the reduction of the domains of use of the language. Part of the difficulty in building a typology of speakers of endangered languages comes from deciding whether to approach the task from a linguistic competence perspective or from a language use perspective, i.e. by how well the speaker knows the language, vs. how often and regularly she or he still uses it. The approach taken here is one that attempts to categorize speakers first on the basis of their knowledge of the language; it places the different types of speakers on a continuum from monolingualism in the minority language to practical monolingualism in the socially dominant language, with all degrees of progressive bilingualism in between minority and dominant language. In some ways such an approach would not be very different from a study of language shift in immigrant communities. However, what makes the situation different in the case of endangered languages is the complex interlocking of multiple factors beyond the level of language competence of a particular speaker, such as his or her mode and extent of acquisition, length and type of exposure to the language, community and personal attitudes. The result is that no two speakers will have the same language history, although a categorization of some recurring prototypes may be worked out.12

3.4.1.3. Types of Speakers There seems to be a consensus that the major prototypes of speakers to be reckoned with in situations of language endangerment by language shift can be identified on a primary distinction of three levels of competence,

12 Bert (2001) offers a discussion of the need to consider all these variables based on extensive interviews with over a hundred of the last speakers of Franco-Provençal.


51

yielding the types of fluent speakers, so-called semi-speakers, and terminal speakers. Fluent speakers. Among fluent speakers one needs to distinguish two subcategories that have been labelled “old fluent speakers” and “young fluent speakers”, although the labels may be confusing, since they do not appeal directly to the age of the speaker. “Old fluent” are the traditional speakers raised in that language alone, and most secure in it. The expression “young fluent” refers to bilinguals who are still fluent in the endangered language but speak it in a somewhat changed form. By the time a linguist arrives, the language may be so endangered that those speakers are in fact some of the older people of the community. Characteristically the new form of language spoken by these “young fluent” speakers is accepted by the community. As it turns out, discussions of standardization and revitalization often involve choosing between older and younger fluent forms of speech to be taught to the learners. Semi-speakers. The category of semi-speakers, prominent in Dorian’s writing, is the category most emblematic of situations of endangered languages. It is a large category which includes all members of the community with appropriate receptive skills, but varying levels of productive skills. The category includes semi-speakers, who can be fluent but whose changed forms of the language are considered mistakes, and weak semi-speakers with a limited ability to produce speech—speech which tends to be made mostly of frozen expressions. It is worth noting that it is from this generally larger semi-speakers group that some of the most involved activists of language revitalization emerge. Terminal speakers and rememberers. These are members of the linguistic community with very limited productive skills, but some passive knowledge. This very limited knowledge can either be the result of a very partial acquisition of the endangered language, with the effect of producing some form of substratum influence on the dominant language, or the result of an advanced level of language attrition on the part of once very good childhood speakers. Such speakers should not be overlooked in fieldwork, particularly in efforts at gathering speakers, since they may gain back or reacquire some partial active use and can always help reconstitute a sense of community at organized gatherings. They are bound to derive deep satisfaction from the renewed contact with the language, provided they are not too psychologically scarred and scared about that language (such as in the case of survivors of massacres).

52


3.4.1.4. Projected Revisions to the Typology This is a working typology and terminology that identifies some major types of speakers of endangered languages that fieldworkers may encounter. The projected revisions will take into account the need to review some of the terminology, such as the contrast between “old vs. young” fluent speakers, often misunderstood; or the key term of “semi-speaker” itself, sometimes taken so literally that it seems to mostly evoke incompetent speakers, even if the category explicitly includes fluent speakers; or the terms “terminal” and “language death”, which have been criticized as being politically incorrect.13 The other revisions of this typology will underline the limitless variety of speakers, in fact their uniqueness in the case of the very last speakers of a language, and work out a set of variables to describe the attributes of those speakers rather than types to box them in. Here is not the place to sort out this problem of terminology and typology, but just to acknowledge the existence of a wide variety of speakers and to consider their interactions with field linguists in the building of a database for the description of an endangered language. As already mentioned, to the extent that the knowledge of the language may linger on in a fragmented way among the various types of speakers, it is important to consult as many speakers of as a many types as possible. Their contributions will be of different types too, but all are valuable, in terms of time depth, coverage of topics, levels of retention of certain aspects of the language, and eventually the study of the process of language degenerescence itself.

13

Dressler (1978) had originally talked of: healthy speakers weaker speakers preterminal speakers better terminal speakers worse terminal speakers

while Campbell and Muntzel (1989) worked with the following categories of speakers: S I W R

= strong, nearly fluent speakers = imperfect but reasonably fluent speakers = weak speakers = rememberers

which Sasse (1992) reorganized as: S = rusty speakers I and W = semi-speakers R = both from rusty speakers and of semi-speakers.


53

3.4.2. Working with this Great Variety of Speakers Working with speakers of endangered languages brings about a number of challenges, from identifying them to working with them, some of which are mentioned below.

3.4.2.1. Counting the Last Speakers The nature of the social fabric of the linguistic community is such that it will take time to identify all the speakers, particularly the isolated ones and the ones who have not been claiming to be speakers. It is likely that the speakers themselves do not know who are all the other speakers to the extent that some know the language but do not use it ever, and are not identified as speakers. Identifying speakers is a slow process best carried out over a period of time, as there are usually more last speakers than said.

3.4.2.2. Evaluating Speakers’ Knowledge of the Endangered Language There may be surprises in the evaluation of the knowledge of a speaker. Self-evaluation is based on identity criteria, with common over- and underestimation of that knowledge. Self-claimed speakers can find themselves confronted with their limitations when asked for complete grammatical paradigms, for instance, while self-claimed non-speakers can be heard correcting those speakers from the back of the room without accepting being identified as speakers.14 Dynamics can also be set off in such a way that renewed contact with the language may either reactivate some knowledge in the case of language attrition of rememberers, or provide opportunities for new (re)acquisition in the case of some semi-speakers.

3.4.2.3. Collecting Data from Speakers of Endangered Languages Data-collecting methods need to be rethought for field situations of oral tradition languages in the first place, but the extra dimension of a situation of endangered languages often demands further consideration from a methodological standpoint. Most standard elicitation methods taught in university field method courses need to be adapted, to obtain natural data and to avoid confronting speakers with their limitations, which potentially results in 14 This is the category of “ghost speakers” identified by Bert in his fieldwork on Franco-Provençal of France. They were wives of self-proclaimed speakers who stood in the back of the kitchen watching the interviews and the taping sessions and would correct or complete the answers of their husbands without wanting to be identified as speakers. This points to the fact that there is a social identity of “last speaker” that some endorse and others do not, independent of their level of fluency in the endangered language (Bert 2001).

54


psychologically difficult and even painful situations unlikely to arise with speakers of vital languages.15 What follow are some notes on methodological issues to be entertained. Recreating settings for natural language use. Natural data basically means data that are not the product of translation from direct elicitation. They may be spontaneously produced, or they may be produced on the basis of certain kinds of verbal, visual, or manipulated stimuli. Although this requirement is not really specific to endangered languages, the reality of endangered languages is that possibilities to collect natural data may have become so limited that it becomes an absolute necessity to think of how to create settings for natural language use. The basic practice is to bring speakers together in order to provide new opportunities for social gatherings and language interaction, and therefore promote the production of natural speech. This is much easier said than done actually in many extreme cases of language endangerment, because it requires identifying the speakers in the first place, then networking among them to gather them, and strategizing the encounters (like organizing transportation of sometimes geographically distant or disabled people, and providing strategic support for the hosting party). However, the efforts always pay off, as such gatherings may come to mean a lot to the speakers who yearn for some social encounters of the sort in order to have a chance to reuse their language and (re)activate their relation to other speakers. This is valid for all types of speakers, and as valued by semi-speakers as by fluent speakers. When doable, it becomes the most productive approach to (video)taping different kinds of language data.16 These social gatherings create reasons to talk, through shared activities, such as telling stories, including personal narratives, or listening and

15 However, one must cite the now classical and still relevant guide to the experience of linguistic fieldwork by Samarin (1967), which considers the basic issues of fieldwork in situ with good sense, in contrast to the recent book by Vaux and Cooper (1999), which strikes one as being extremely Euro/ USA- (even Harvard-) centric and cast into a surprisingly traditional approach to field methods. Bouquiaux and Thomas (1976) are valuable for ethnolinguistic fieldwork but from quite outdated to downright odd in their grammatical questionnaires. More useful as guide for grammatical work are Shopen (1985) and Payne (1997). The author’s two favourites for good reading on the general topic of fieldwork are from ethnographer Wolcott (1995) already mentioned, and folklorist Jackson (1987). 16 In some circumstances, as in urban settings, the impossibility of physically gathering last speakers may be partly compensated by using some technology such as conference telephone calls and video links, obviously only in the regions where that is feasible. If nothing else is available, one can also play back recordings of some speakers to others, although this will not necessarily work immediately for the reasons discussed below.


55

commenting on recordings or videos of other speakers from other places or other times, or talking about still pictures, videos, or handled objects.17 Using a variety of material for stimulating language production. A variety of stimuli can be used to trigger language production in general, and to work on targeted aspects of the language, although, as fieldworkers know, when they are used in communities with no literacy tradition of the kind that relies on formal education and training, those stimuli, particularly the ones relying on purely visual stimuli, may not produce by and large the kind of data they are meant to elicit on very specific topics. But they can be very productive, nevertheless, for simply eliciting natural language material.18 For general linguistic analysis it should not matter that speakers seem to often ignore the story line implicit in the sequential arrangement of videos and books alike, or the main event of a picture, and express much more concern and interest for minute details of the pictures (like musings about the dress of the protagonists, the kinds of flowers and animals of the backdrop, the time of day or season of year it might have been). While attending to those details (in our eyes) they are producing actual coherent sentences that are good data for a descriptive grammar of the language. It is only with longstanding familiarity with the language and its speakers’ community, and once efficient working relations have been developed with some of the speakers, that one could be expected to create appropriate stimuli for a particular type of data, and then to collect reasonably reliable data, in order to produce desired reliable and comprehensive analysis of the data.19 About direct elicitation methods. Although the use of direct elicitation is probably the data-collecting method still most used in the field today and most 17 Since the time of the original conference and the first version of this chapter, the field of language documentation and archiving has largely developed, so that today a discussion of the use of video and new computer technologies would have to be considered, although the focus here is still on the human and social interaction of field linguists with the last speakers of an endangered language. 18 Examples of such stimuli used in cross-linguistic research on specific linguistic traits have been, for instance, the 1970s pioneer “Pear Story video” of W. Chafe (University of California at Berkeley) that was meant to track discourse features, or the 1980s “Chicken video” of T. Givón (University of Oregon) that was meant to elicit serialization data. More recently, the wordless Mercer “Frog Story” children’s books have been widely used for the study of adult and children’s narrative skills, in particular by teams supervised by D. Slobin (University of California at Berkeley). 19 There is also a difference between different types of visual stimuli, from line drawings of simple objects like dictionary illustrations, to more complex drawings of scenes, to still pictures which, when just black and white, can disconcert the speakers. Beyond visual stimuli that are always likely not to produce the expected in illiterate communities, one can turn also to the manipulation of objects as a trigger for the production of language. Some of the better-known cases of the use of varied stimuli are the ones that have been developed by the researchers of the Language and Cognition Group of the Max Planck Institute of Nijmegen for instance, in particular those that were targeting the expression of spatial relations.

56


commonly demonstrated in field methods courses in university campus settings, it should clearly be limited to a complementary role in situations of endangered languages at first. Particularly unreliable is the method most used, that of direct translation, which consists of the linguist asking the speaker: “how do you say X?”, X being some artificial sentence out of context built around a grammatical point of interest to the linguist.20 There is no reason why such type of questioning should make any sense to most speakers of oral traditions in their own home environment. Some can be trained to perform, if they happen to be in the small percentage of the natural linguists of a language, but most will never really be. And it can in fact become particularly morally objectionable to confine oneself to this method with linguistically insecure semi-speakers who are then made to feel like failures if they cannot come up immediately with a translation.21

3.4.2.4. Conclusion: Adapting Methodologies for Data Collection and Analysis Much of what was said above is actually, if one thinks about it, common sense, but it is probably useful to articulate these methodological issues because they are still not part of most training for linguistic fieldwork. In order to collect reliable data in situations of endangered languages, the methods to be used must be diversified and adapted for a relatively special population of speakers, because in situations of endangered languages, the last speakers are few by definition, and they are whoever they are and not often likely to become trained to respond to data-collecting strategies developed for vital languages. Even good speakers may not be able to

20 This method is always objectionable in terms of reliability of data when imposed on neophyte linguistic consultants from illiterate cultures. It is only because they have done the adjustments necessary to survive in that dominant culture, including a test approach to knowledge, that linguistic consultants of field methods courses on campuses may give the appearance of responding appropriately to direct elicitation. 21 This is not to say that direct elicitation cannot be very useful, but only as a secondary method, at the service of analysing naturally produced language material, and mostly as an exploration of the glossing process of naturally collected text. It must always be handled with great care, controlled with multiple checking and attention to non-verbal information such as body language cues, and only with speakers with whom one has established a productive working relation. It follows from the above that the use of field elicitation guides in the forms of questionnaires consisting of standard lists of sentences to translate may have something intrinsically inappropriate in approaching a new language, particularly in the case of the type of endangered languages considered here. This is not to say that questionnaires are of no use, but that their use is limited. They make sense for instance when checking the particularities of a variant form of a language or group of languages for which there is already a solid linguistic knowledge, for organizing already collected and analysed data, and for checking for gaps in data on basic grammatical topics which are of interest to other linguists interested in typological, areal, or genetic issues.


57

produce narratives, may never really learn to give an exact translation, and will forever ignore the value of a paradigmatic organization of knowledge. As suggested, much of the effort, in the first place, must aim at triggering the production of natural sequences of language, to obtain a certain quantity of data and to ensure the reliability of this data.

3.4.3. About Linguists Working on Endangered Languages Two more aspects of doing fieldwork on endangered languages which are not often made explicit but would certainly deserve more attention will be taken up here. One touches on the kind of personality profile best suited in fieldworkers for this kind of job, and the other on implications of carrying out such field projects from a professional and academic career point of view.

3.4.3.1. A Certain Personality Profile By now it should be obvious that the part of the fieldwork experience which consists in working directly with speakers of endangered language communities in the manner suggested here actually calls for a certain personality profile. The essential personality trait is connected to the fact that much of this approach to fieldwork relies on the ability of the linguist to accept not being in control of the situation, a lack of control which takes many forms. Beyond the usual lack of control of basic fieldwork, which is handled differently by different field linguists, who develop different types of work relations, there is, in the case of work on endangered languages, a much more pervasive lack of control. There are first limitations on when, where, and with whom one can work, and later limitations on what one can do with the speakers with whom one can work. As already mentioned when talking of the issue of data collecting, monitoring the process of data production can be a challenge, once the linguist has managed to get together with speakers. And this process is often more a matter of triggering data production rather than of controlling that process. One must be patient and allow data to trickle in, and one must bear with data one does not know what to do with at first. Because of the complex relation that can hold between those speakers and their ethnic language, it is also both ethically and strategically sound to be particularly aware of the balance of power between the parties and to give to the speakers as much of a sense of control as possible. Relinquishing control is a mark of respect for the knowledge they have of the language, an invitation made to them to become

58


invested in the work in whichever way they want or can, and it signals a conscious attention to keeping them feeling as comfortable and successful as possible. Relinquishing control is probably one of the most difficult postures to accept for academics. First-world academia tends to naturally select for, and then preferably promote, highly individualistic, self-motivated, and determined free spirits. This world values most highly the pursuit of “basic research” carried out within paradigms partly defined by a culturally bound sense of efficiency and productivity. All of this can be counterproductive in endangered language field situations, may well actually work against the production of reliable and comprehensive linguistic descriptions of those languages in the first place, and may not respond at all to the needs of the community itself (much of the above has already been discussed in Grinevald 1997, 1998, 2000, and 2003a).

3.4.3.2. Energy, Time Commitment, and Professional Risk Doing fieldwork on an endangered language implies a basic energy and time commitment that also needs to be acknowledged. A reasonable estimate of the length of time for the amount of work it takes to build a basic database with which to produce a reasonably accurate and complete description of an endangered language probably runs around three years minimum, between field time and months of data processing and analysis. It probably also needs to be said here how dangerous fieldwork on a very endangered language may be to the career development of a linguist, particularly graduate students and junior faculty, who are in fact the most likely to commit to it. If this line of fieldwork is time, money, and energy demanding (think of fieldwork in the Amazon for instance), it is also basically risky in many ways. Risky in the sense of the dissonance and alienation mentioned before between academic, financing foundation, field and community pressures. This is particularly the case for linguists involved in major documentation/revitalization projects with strict timetables. Risky in the basic sense of not being able to collect enough data for a doctoral thesis or for publications of the sort valued in promotion procedures, due to the death of speakers or any other factor rendering fieldwork difficult to impossible. Risky in the kind of data collected not providing the materials expected to enter the theoretical debate arena the way it is being set up by linguists working on major vital languages. Risky in the sense of relations to the community always being subject to quick turns because of misunderstandings or disappointments. The academic community of linguists needs to consider those risks and see


59

how to best minimize them to protect those that it sends off to do the work of describing and documenting endangered languages.

3.4.4. An Invitation to Take up the Challenge This piece of writing is not meant in anyway to discourage linguists from contemplating doing work on endangered languages, but rather to provide some realistic insights into the nature of the enterprise, for all parties concerned, potential fieldworkers in the first place of course, but also foundations that finance such work, and members of the linguistic profession that advocate such work.

3.4.4.1. A Collective Responsibility The work is important and urgent, and it ought to be the business of all sectors of the profession. Those field linguists available, interested, and willing to take on a part of the daunting task of documenting as many endangered languages as possible before it is too late need to be nurtured by the profession. They should be first adequately prepared and trained; then, while they are doing the work, they should be as well supported as possible—financially, psychologically, and academically—and their place in academia should be assured so they can pursue this line of work. Unless we commit collectively to all those aspects of nurturance we really have no business making much of a fuss advocating saving and documenting endangered languages.

3.4.4.2. More About Why We Do it I would therefore strongly encourage those engaged in such work to tell those interested in doing such work what sense of profound satisfaction and what occasional exhilaration obliterate all the moments of frustration, confusion, and heartache that are an inextricable part of the enterprise. We should all tell our future colleagues how it feels to be opting to be a linguist in the real world, dealing concretely with the very real catastrophe of the accelerated loss of the majority of the languages of the world. In so doing we are privileged to discover the riches of yet undescribed languages by working with speakers and communities that can be profoundly grateful and proud to contribute to the salvage of their endangered ancestral language. Some of the last speakers may have dreamed of it and may have hoped for it for a long time and will convey to us how extremely relieved they are to be given a chance to participate in some project to describe and document it. Those are strong moments in a linguist’s life, the priceless

60


human dimension that keeps fieldworkers going back in spite of it all. Any fieldworker familiar with this kind of fieldwork is certainly able to communicate, to anyone interested in hearing about it, how this human dimension of fieldwork amply makes up for all the kinds of headaches and heartaches that have been spelled out in the previous sections of this paper.22

3.4.4.3. A Last Warning However, in the end, and to be honest, fieldworkers also need to warn about the challenge of becoming a tightrope walker between the ivory tower of academia and the first-world values of financing foundations where the discipline of linguistics develops and is supported, and the realities of often embittered linguistic communities of endangered languages, with their complex sociopolitical set-ups where the last speakers can be themselves very marginalized. And this needs to be said maybe more so as time passes and the topic of language endangerment reaches beyond academia and acquires a certain veneer of hype within a developing public discourse of saving the biolinguistic diversity of the world.

3.5. A Case Study of Fieldwork on an Endangered Language What follows is a descriptive account of the development of one such field project on a very endangered language of Nicaragua. It is the story of a salvage linguistic project for a moribund language, but in fact a linguistic project embedded in a revitalization project of much larger scope, itself conceived within major political dynamics at the time in the country. The narrative will first situate the project in its specific sociopolitical context and will then focus on the main topic of this chapter, that of working with speakers of endangered languages. It will recount moments of the search for speakers and try to give a flavour of the strong mixture of contentment and frustrations that characterized the data-collecting phase of the project.

3.5.1. The Sociopolitical Circumstances of the Project The Rama Language Project (RLP) of Nicaragua took place in the context of the Sandinista Revolution of Nicaragua, as described in Craig (1992b), and stretched over a period of ten years starting in 1984. 22 Newman and Ratliff (2001) is a good place to hear about linguistic fieldwork and the relation of linguists to speakers, in a number of chapters on speakers of endangered languages.


61

3.5.1.1. Regional Autonomy Project and Linguistic Rights in Time of War The Rama project was initiated in response to demands expressed to the Sandinistas by Rama community leaders claiming their new linguistic rights granted to all ethnic groups of the region by the new Autonomy laws. That region was then called the “Atlantic Coast” and was the half of the country where indigenous populations were still speaking various indigenous languages (Miskitu, Sumu, Rama). The Ramas were concerned that they had lost their ethnic language and were asking for help to bring it back for ethnic identity purposes. The linguistic description component of the Rama Language Project was therefore cast from the start at the core of a much wider project, conceived by the Rama community and the authorities at all levels as a language and culture revitalization project. This was happening at the time of the so-called Contra War, a war waged against the Sandinista government by anti-Sandinista forces (largely financed, supported, and led by the USA government). The Rama population, as often happens in such conflicts, was caught between the two sides, and the Rama language project was conceived in part as a gesture of peace.

3.5.1.2. The Rama People and the Rama Language The Rama people themselves were considered then as the most marginalized population of a multiethnic autonomous region, below Spanish-speaking Mestizos and English-speaking Creoles, and below other indigenous people of the regions, the Miskitus and the Sumus they dominated. There were less than a thousand Ramas in total. The language of the Rama population was a form of the English Creole of the region known as Miskitu Coast Creole (MCC), and very few could speak Spanish. The Rama language is a Chibchan language, of a large family of languages spoken from Honduras in the north to Colombia in the south, and no study of it was available. At the time of the request made by the Ministry of Culture to the present linguist in 1985, it was said to be spoken by only three old men of the island of Rama Cay, the only Rama community known to the outside.

3.5.2. The Search for Speakers As it turned out, for complex political and sociological reasons, and to the lingering dismay of the linguist, none of the three old men of Rama Cay participated in the Rama Language Project and their actual knowledge of Rama was never assessed.

62


3.5.2.1. Finding Three Rama Speakers to Work on the Rama Language The first speaker of the project was actually located indirectly through academic networking, by connecting to a fellow Amerindianist, Lyle Campbell, who had done a survey of the endangered languages of Central America some years before and connected the linguist to an ex-research assistant of his field project, Barbara Assadi, who had stayed behind in Nicaragua and had integrated into the Rama community for a while. She in turn provided information and contact for a good speaker of the language, said to be eager to tell about her language. An initial visit to Bluefields a year into the process allowed for a first contact with this speaker that sealed the fate of the project. Miss Nora, as she is widely known now, who was first located as a destitute refugee from the war in the grounds of the Moravian church of Bluefields, later turned into the inspired leader of the whole project (Craig 1992a; Grinevald 2003b; Grinevald and Kauffmann 2004). Work on a grammar of the language was finally initiated with that one speaker the following summer, with research funds secured from the National Science Foundation. The first summer of linguistic work with Miss Nora revealed that she was a fluent but semi-speaker of the language, and an excellent language consultant. So, for the next field trip, Miss Nora had recruited her daughter-in-law Cristina Benjamins to help her out with the linguistic study of the language. This native speaker from the mainland who used Rama every day of her life became the main source of data for the study of the grammar of Rama. The third main language consultant has been Walter Ortiz, one of Miss Nora’s nephews. He was a young fluent speaker, in his fifties then, who was considered by all to be the one “intellectual” and scholar among the community of Rama speakers because he was (and still is) the only Rama speaker with some literacy skills (and at that only in Spanish, the national language). He was the hope for back-up and leadership of the team of speakers of the project who eagerly awaited his return from Costa Rica for several years. He joined the project after the completion of the grammar, at the time of the dictionary project.

3.5.2.2. The Different Skills of the Three Main Consultants The Rama language project leader, known as Miss Nora (her full name being Leonora Rigby), lived for her dream of documenting the language so it would not disappear without leaving a trace. A dignified woman with a vision for what could and should be done in the reticent community of Rama Cay, she was the daughter of the last Rama shaman who “talked to the


63

tigers in the Tiger language”. She had learned Rama at the age of 10 when she had gone to live in the jungle with her father and stepmother, both monolingual Rama speakers; she later married a pure Rama and raised her children on the mainland. She produced short narrative texts after some coaching, and regained higher fluency throughout the project. She was an excellent linguistic consultant, a natural linguist type of speaker. She corresponded to the type of native person known in the literature on fieldwork as “foreigner seekers”: in her case it was her third serious try in ten years at having the Rama language documented by a foreigner.23 Cristina Benjamins was a young fluent speaker who uses Rama to this day on a daily basis. She was a challenging language consultant at first, because she was too much of a natural language user to accommodate easily to the artificiality of linguistic elicitation. She initially had great difficulty at a number of tasks. One, for instance, was to sit still in the office during work sessions, a non-activity she was totally unaccustomed to, being the mother of eight children and spending her days in numerous physical activities to attend to the needs of her family. She furthermore had great difficulty concentrating on repeating exactly what she had said on the tape recordings (why repeat the same since it had clearly not been understood! was her attitude), and did not have patience with the task of translating exactly what she had actually said (preferring to add additional information to complete the narration). The first work sessions were very intense (because of her rapidfire native way of talking) and very tense (because she did not see the purpose of staying since she felt she was not doing what was wanted of her). She was by far the best speaker of Rama available, and luckily she stayed (at the urging of Miss Nora) and became over time a very valuable linguistic consultant, including for direct elicitation work. She contributed the bulk of the narrative collection which is the basis for the grammar description; she was best with personal narratives, in the course of which she talked of the way of life of the Ramas of the mainland, but had only fragmented knowledge of the main oral tradition of the Rama known as the Adam cycle. Walter Ortiz is a native speaker from the mainland, with very basic literacy skills. Unlike Miss Nora, he never developed much ease with grammatical elicitation, and, unlike Cristina Benjamins, he did not manage to contribute any narratives for the textual database. He was, on the other hand, invaluable for dictionary work, patient and detailed in his terminology

23 Previous attempts were with Barbara Assadi, and a German volunteer early in the Sandinista times (see Craig 1989).

64


and descriptions, particularly of his surrounding world and male survival activities. He is a quiet and private man, and his dream has been that some day a younger man will approach him and tell him he wants to learn to really speak Rama, the way he himself had learned the intricacies of it from his uncle, living with him alone in the jungle.24

3.5.2.3. Looking for More Speakers Identifying the last speakers of Rama took several field trips over several years. First, Barbara Assadi on a trip to Bluefields helped work out with the first two Rama speakers available a list of the other last speakers of Rama. Some of those last speakers were brought to the Bluefields CIDCA office to be interviewed,25 and trips around the Bluefields lagoon were organized to visit others. Later, as a considerable number of the mainland Ramas had taken refuge from the war in neighbouring Costa Rica (the Rama land borders on Costa Rica and reaching Costa Rica by sea is relatively easy), among them a number of the native speakers of the language, a trip was arranged to make contact with them in Costa Rica.26 Finally, as the project took on more visibility with work sessions on Rama Cay, and with Miss Nora beginning to fulfil her dream of teaching some Rama to the kindergarten children of Rama Cay, rememberers and terminal speakers began to make themselves known on Rama Cay. Three years from the start of the project, thirty-six fluent speakers had been identified (thirty-two mainlanders and four islanders), with an additional twenty-two limited speakers, for an actual total of fifty-eight speakers. These figures were a matter of both good news and bad news. The good news was that there was a much higher number of speakers than expected, and not

24 A dream of a one-dyad master–apprentice programme of the kind Leanne Hinton has been setting up in California for some of the most moribund languages that some communities want saved. Today Walter Ortiz has taken over from Miss Nora the teaching of some Rama in several grades of the elementary school of the island; he shares his time between Rama Cay and the mainland where he likes to retreat from everything and cultivate land for as long a time as his duties in the school and the increasingly dangerous situation on the mainland permit. 25 The Rama Language Project was locally sponsored from the start by a research centre specializing in the affairs of the region, the Center for the Documentation and Research of the Atlantic Coast (CIDCA). An office for the project was set up in Bluefields, not far from the Rama Cay community. 26 Easier said than done. Costa Rica was the base for the Contra activities from the south, and some of the Ramas were involved in Contra war activities. This meant negotiating permission for the trip, and official support to try to convince Rama speakers to come back to Nicaragua, at the expressed request of their relatives involved in the project. Offers for amnesty and recognition of refugee status with corresponding material support upon return, from Nicaraguan officials supporting the Rama Language Project, were transmitted to the Ramas of Costa Rica. A good number of them did return. But it did not translate into more speakers for the study of the Rama language, with the exception of Walter Ortiz.


65

just an old population of speakers (fourteen speakers were below 44 years of age, eight below 64, and only five above 65). The bad news for the linguistic description of the language was that work turned out to be impossible with most speakers. A striking characteristic of the profile of the last fluent and even native speakers was, for instance, that the male speakers (in a striking dominant ratio of male to female speakers of 22M/9F) were also by and large single men living by themselves deep in the jungle, with no descendants, and inaccessible in all senses of the word.27 An interesting feature of this sociolinguistic profile of Rama speakers, arrived at through many work sessions and interviews of speakers themselves, coupled with multiple checks and counterchecks over the years, is that it has never been accepted or acknowledged by the wider community, neither local, regional, nor national, in spite of repeated publications of the figures and numerous public interventions. To this date, the Nicaraguan myth of the Rama language spoken by only three old men of Rama Cay tenaciously lives on.

3.5.3. Trying to Work with More Rama Speakers Of the fifty-eight speakers recounted, twenty-five were contacted and visited, and thirteen were interviewed in some capacity; of those interviewed, only eight provided some language data, and generally only very little, so that, in the end, the description of the language had to rely on the three speakers already presented. A rundown of all the speakers considered will give below a sense of the diversity of speakers contacted and of the types of language information they could and could not give.28

3.5.3.1. Other Native Speakers that did not Work Out as Language Consultants As already mentioned, the three famous old men of Rama Cay said everywhere to be the last speakers of the Rama language apparently only spoke Rama anymore when drunk. Although they had apparently participated some years before in a previous attempt at documenting the language with a

27 The majority of these had fled to Costa Rica and did return to Nicaragua but went back to live in the jungle. 28 While the three major speakers that have been the main consultants of the project are identified by their names because they have been identified in all writings on the language, in recognition of their active contribution to the salvaging of their language, the other Ramas mentioned will not be personally identified to respect their privacy.

66


young German male anthropologist, they ended up having no contact with the project and their actual fluency could not be assessed. Two mainland native speakers eager to participate in the project were considered but their contributions could not be processed and did not become part of the database for the grammar. One was a young fluent speaker, sister and daily companion of Cristina Benjamins, whose texts did not cohere and whose ability as a linguistic consultant remained too limited for language analysis. The other was an old monolingual man who had lived for decades isolated in the jungle and spoke what the others labelled “real Rama”. He was considered by all to be the “best speaker” and agreed to tape interviews on the “old ways” and traditional place names.29 Those interviews took place at the CIDCA office, and turned into major social gatherings for Rama speakers who had not been together for a very long time and had not talked to him in a long time either. It first appeared that these hours of tapings would provide the core of the database. Unfortunately that could not be the case, as they turned out to be too difficult to transcribe and to translate, for reasons that remain unclear to this day. None of the speakers present, including the two main consultants of the project, could repeat what he had said, although his rapid slurred and toothless speech had not seemed to be an obstacle to their enjoyment of his recountings. The interaction had been joyful and intense, communication seemed to have been happening, as witnessed by the laughing and questioning captured on tape, but it sadly yielded nothing for the linguistic analysis of the language.30 More disappointment for the linguist in search of good data ensued when several mainland native speakers in the old fluent speaker category took part in some of the project activities but watched and said nothing, or so little so whispered and so mumbled that the data were not usable either. For instance, the brother of the old monolingual speaker of the gatherings just mentioned clearly enjoyed immensely being there, smiled and laughed, but never uttered a word. And a married couple of old fluent monolinguals, whose participation had been eagerly anticipated, looked bewildered and apprehensive all through the interviews they had agreed to. Direct interaction between them 29 As a matter of fact, these interviews took place in a dramatic way. He was on his way back down to the jungle after a long hospitalization in Bluefields, as he had asked to be taken back to Rama land feeling he was soon going to die. He did not die, and lived a few more years, settled at the mouth of a major creek where Ramas would stop and care for him before going further up the creek to their settlements. 30 Years later, when asked also to listen to the tapes, Walter Ortiz, the third consultant, could hardly make out some stretches of recording now and then. The mystery therefore remains as to what really happened then.


67

and the linguist turned out to not be possible, and even coaching by Rama speakers did not help either. In the midst of the Contra War, the trip to Costa Rica had provided direct contact with one more native speaker, beside Walter Ortiz already mentioned. He was Miss Nora’s elder brother; he listened attentively to her taped message in Rama asking him to return and join the project, but he did not return to Nicaragua for years. When he did, he went back deep into the jungle and was never interviewed again.

3.5.3.2. Working at the Margins: Rememberers and Semi-speakers The Rama speakers from the island of Rama Cay included a number of rememberers and terminal speakers. The female rememberers identified were relatives of Miss Nora, who described them as native speakers who had been fully fluent decades before, but who had been traumatized into hiding their knowledge of the language. One was an old woman who, as a teenager, had accompanied Miss Nora, her younger cousin of 10 years of age then, for the first months she had gone to live with her monolingual father and stepmother, in order to help her as an interpreter. Although this elder cousin denied knowledge of Rama for the first years of the project, she later became eager to become involved, but she could never recover much fluency and died a few years into the project. Two young rememberers, nieces of Miss Nora, also denied knowledge of any Rama for years, in spite of Miss Nora vouching that they had been very good speakers; they never agreed to join any activities of the project.31 A considerable number of other Rama speakers with more limited knowledge and practice of Rama participated actively in the revitalization part of the project. They included schoolteachers eager to introduce some Rama in their classrooms, high school youngsters who wanted some Rama phrases to impress people with them in the streets of Bluefields, community members with some links to the Rama language, rememberers, and maybe even shy semi-speakers who filled the schoolroom on Rama-language Sunday meetings. One of the semi-speakers worth mentioning, because of his key support to the project from the start and his new leadership role now, after the death

31 The two of them were typical cases of trauma-induced severe language attrition, the trauma being a case of a dramatic family situation, and discrimination against tiger people on Rama Cay, rather than a case of outside attack on the community (as in some cases of ethnic massacres).

68


of Miss Nora, is Pedro MacCrea, her eldest son and Cristina Benjamins’s husband. He was key to the project as boat captain, providing transportation for speakers between jungle, island of Rama Cay, and town of Bluefields; he was also an excellent spokesperson for the project. Interestingly he became more fluent in Rama as the years passed, reaching the point of carrying out interviews with monolingual speakers. He is today the most visible and charismatic representative of the Ramas in their battle to protect their land, dealing with authorities and international consultants, and appearing in different media speaking (sentences of) Rama.32

3.5.3.3. Cycles of Expectation and Frustration What this rundown on speakers has emphasized is the difficulty of finding a number of speakers who could function as linguistic consultants, even if there were a certain number of speakers still. The point is that this case study is not necessarily a very special situation for a case of extremely endangered language. On this backdrop of a constant search, one can probably imagine the cycles of excitement and frustration, of tediousness and confusion common to most of these field situations. But one has to also mention, hidden behind such a list of helpful and not so helpful speakers, the weaving of strong bonds with a community in search of its last speakers. Somehow it is the interaction of all those speakers that produced eventually a more or less accurate picture of the linguistic community of this very endangered language, a picture quite different from the popular view outsiders had of that community, in fact. Although the complexity of the Rama field situation exposed above is probably fairly typical of many cases of extreme language loss, it feels most important in closing to underline one aspect of this project. It is how both the production of the descriptive grammar of the language and the apparent success of the larger Rama language revitalization project of which it was a part are a tribute to the intelligence and the tenacity of one old woman, herself a fluent semi-speaker but with a vision and a natural linguist talent. There is no doubt in the mind of anybody that came close to her that Miss Nora was the real Rama language rescuer and that nothing much would have happened without her, as argued in Grinevald and Kauffmann (2004).

32 His leadership status of today is reminiscent of the phenomenon described by Evans (2001), of the emergence of new leaders and new last speakers, as the most visible and prestigious one leaves the scene and others find their place.


69

3.5.3.4. Dealing with a Major Contradiction for the Language Revitalization Part of the Project The sociolinguistic study showed that the Ramas were divided into two distinct communities: the majority of the population, the only Ramas recognized at that time, lived on an island in the lagoon of Bluefields called Rama Cay, while a much smaller group of a few dozen still lived on the mainland, along small creeks and on the ocean coast. The community of native speakers of Rama that still existed was from the mainland, and it is with members of that community that the description and documentation of the Rama language took place. As it turned out, for the whole duration of the project, this division of the Rama community into two communities (island and mainland) was a major dimension to be constantly factored in. The prejudice the islanders felt toward the mainlanders was extreme, and resulted in the former systematically ignoring the existence of the latter in their report on the Rama community to the Sandinistas. The other way around, the mainlanders, including members of the last community of speakers of Rama, avoided contact with the islanders for that same reason. The scenario was therefore complex: the demand for the revitalization of Rama came from the leaders of Rama Cay, islanders who only spoke English Creole and actually deep down had very mixed feelings about the Rama language. They were inclined to despise it as being a “primitive” language, as they had been told it was, and referred to it as the “tiger language”, saying of the speakers of Rama that they were “tiger people”. They had therefore wanted the revitalization to come from people of Rama Cay and were for that reason at first unwilling to accept language materials coming from the mainlanders, at the same time as the mainlanders did not want to set foot on the island, for that same reason. 33 Therefore to the already negative attitudes of the general Nicaraguan population towards indigenous languages, and the particularly negative 33 Today, twenty years later, the two communities are banding together, partly around the issue of a shared Rama language and culture revitalization vision that has been sharpened by a sense of doom and threat to their land base and all traditional ways of living. There is both severe threat of cultural (Rama Cay) and even physical extinction (mainland). The acculturation of islanders to the linguistic community of Bluefields is fast accelerating, as more families come to take refuge or settle there, many in hopes of providing secondary education to their children. But most acute today is the threat to the mainland community of more traditional Ramas created by the advance of the agricultural frontier from inland and of major international construction projects coming from the sea. A plan to build a railway (called the “dry canal”, a replacement for the now obsolete Panama canal) across the traditional (and officially protected) Rama land threatens to cut off the last speakers to the south from the more numerous Rama community of Rama Cay to the north. Land speculation has already resulted in physical attacks on the Rama population (Mueller 2001).

70


attitudes of the people of the region for the Rama people, the Rama themselves added another layer of self-discrimination that rendered the project particularly difficult at first.

3.5.4. Some Figures: Time, (Wo)manpower, and Grant Support To the extent that fieldwork projects on endangered languages are somewhat of a mystery to outsiders, a few figures from the Rama Language Project will be given here. It was of course just one in an infinity of possible projects, but it included major ingredients always necessary in such projects. Some of the dimensions of the project will be outlined in figures, and some of its characteristics will be underlined. The production of the Rama grammar took six field trips, about ten months of actual fieldwork in total after the initial exploratory trip. It took three years of steady work by a team of three linguists. The principals of this team were three academics coming from the USA: a project coordinator and main fieldworker (Grinevald, at that time Craig), a data-processing specialist, linguistic fieldworker, and coordinator of the production of materials for the community (Bonny Tibbitts), and a community contact (Barbara Assadi) who offered fieldnotes, conducted the census of the last speakers (among whom she had lived for several years), and checked the data for cultural accuracy. A number of graduate and undergraduate students processed data. The funding sources for the project were a combination of academic grants, from the National Science Foundation, Wenner-Grenn for Anthropological Research, and University of Oregon Research Funds for the grammar, to the National Endowment for the Humanities for the dictionary. As often calculated, applying for such grants, monitoring them, and reporting on the activities and results required about a third of the total time and energy invested in the project.34

34 To be honest to the end, for the sake of the reality check that is the purpose of this chapter, and for the benefit of those embarking on such kinds of projects and taking risks, it is also necessary to say that the actual production of the dictionary was not completed within grant time, due to a combination of personal and political changes of circumstances. The Hans Rausing Endangered Language Documentation Project is therefore hereby acknowledged for its willingness to financially support the completion of the dictionary project, and the archiving of the materials of the first phase of the project, ten years later. Independent of the reasons for not completing this particular project as per grant deadline, it might be worth taking the occasion to raise a taboo subject, which is that such things happen more often than admitted with field projects, particularly complex ones like this one, which demand more time and energy than is realistically factored into the grant schedules. It may be one of the responsibilities


71

Beyond numbers, this project was clearly cast in a particular fieldwork framework worth noting in the end. Although the list of speakers given above did not do justice to the level of involvement of the community of speakers and non-speakers, the enterprise was clearly cast into an “empowering fieldwork framework” with a clear ethical view of working on the language for and with the speakers. The circumstances of this project, one of the projects of “Linguists for Nicaragua” of Sandinista times, were described in Craig (1992b). It might be worth mentioning finally that the bulk of the material produced for the community was the work of student volunteers from the University of Oregon,35 and that the community part of the project received financial back-up from private human rights and solidarity groups, active in defence of the development programmes for the indigenous population at the time of the Contra War.36

3.6. Conclusions This chapter has concentrated on the data-collecting issue for the description of endangered languages, taking into consideration three aspects of it in turn. The first point was that linguistics fieldwork projects aiming at producing a description of an endangered language are today in most parts of the world more and more often woven into wider types of projects, which themselves may largely overlap: projects of linguistic description, language and culture documentation and language and culture maintenance or revitalization. Admittedly, this particular view of the likely type of field situations field linguists encounter today may be strongly biased by the author’s experience with Native American situations of the Americas (North, Central, and South), and may not apply as much to certain other parts of the world (yet). It would seem necessary that the new granting agencies preoccupied with the fate of endangered languages have a clear grasp of the realities on the of senior members of the profession, such as the present author, to help raise the issue of probably very unrealistic output expectations of granting agencies now engaged in documentation of endangered languages on a large scale, particularly if the projects are indeed to be of equal benefit to the academic and the speaker community. There is much pressure on junior members of the academic linguistic profession, if to boot the documentation has to be pluri-disciplinary and multimedia. 35 By the dozens over the years of the project, recruited in undergraduate courses such as “Introduction to linguistics” or “Languages of the world”, skilled and eager to participate in a research project. 36 Most noticeably the Council for Human Rights in Latin America of Oregon, Linguists for Nicaragua, and Witness for Peace.

72


ground of such larger scope projects, beyond facing the already strenuous concerns of the more technological aspects of documentation and archiving. And what was not raised explicitly but remains to be fully considered is the challenge to some collective professional thinking of how to best support the fieldworkers who have to balance the demands of both academia and the field, demands that are often contradictory and pressing, and always time and energy consuming beyond what is usually acknowledged by office- and library-bound linguists. The second point was that it will always remain the professional linguists’ responsibility to produce an analytical study of the language. That is what the profession is about at heart and what we must not lose sight of, particularly once we embark on wider community projects. This requires preparation in appropriate linguistics and field methodology. As mentioned also, training speakers wherever and whenever possible should always be a priority, particularly to empower those speakers with native talents, as it is fundamentally more ethical to share our knowledge with those who are interested in it. And if one needs to drum up other arguments, one can also say that it opens up the possibility of sustainable work, particularly in wider scope documentation projects. Such projects are best conceived as genuinely collaborative, best handled by members of the community at the ground level, who will always have better opportunities to work with speakers that field linguists may not have access to. It would therefore appear to be an all around sound strategy. But it is unfortunately still too rarely embraced as the main approach, probably because it calls for a heavier time investment initially, because it may appear at first to slow down the process of gathering data for a description of the language, and because it demands much more personal investment on the fieldworker’s part. The third major point was the existence of a wide variety of speakers in endangered language communities, and the need to consider them all for the different types of information they can provide on the language, whether linguistic or sociolinguistic. It was noted that some awkwardness remains in the terminology in use, such as the terms old and young fluent speakers, semi-speakers, and terminal speakers, and intrinsic difficulty in establishing a typology was pointed out, as partly due to the many variables to be handled to account for the uniqueness of each speaker. It crucially called for great care in handling the pervasive condition of linguistic insecurity of the semi-speakers, for human reasons as much as for the sake of the reliability of data being collected.


73

On the background of this diversity of speakers, the next section emphasized the need to reconsider data-collecting methods. This could be said of all linguistic projects in any case, and of all linguistic field situations, but it was argued that such a reconsideration becomes crucial when dealing with endangered languages. For if the database to be produced is to remain the main information on the language for the future, it needs to be as complete as possible; and if the description is to capture the genius of the language and to do so with reliable data, the task is to collect natural language data. But as stated, this is not so easily done with speakers of endangered languages, and a variety of methods were proposed: mainly those of creating gatherings of speakers and of using various types of stimuli. While all the above may sound like common sense not worth writing about, it is obvious that it is not yet the practice of the majority of linguists, as evidenced by the type of data used in publications. Of course, this overall methodological approach is nothing new, but has been forgotten rather, as the task of linguistic fieldwork has passed from an anthropological tradition more attentive to such issues, to strictly linguistic circles dominated by a certain approach to theory building bent on the notions of native intuition and ideal speaker. The last section presented a case study of one project of language description and revitalization that served the purpose of illustrating how some of the issues raised earlier in the chapter played themselves out in a particular situation, as unique as all endangered field situations can be, but also as universal as they are in their complexity. In the end, it is hoped that these notes on one of the many aspects of fieldwork on endangered languages will provide material for a necessary confrontation of our discipline with some of the realities of the work it wants to promote.

References Assadi, B. (1983). “Rama Cay Creole English: Sociolinguistic History; Text RCC: 1 ‘Rama Cay Neighbors’ ”, In J. Holm (ed.), Central American English: Varieties of English around the World. Text Series 12. Heidelberg: J. Gross, 115–23. Austin, P. (ed.) (2003). Language Documentation and Description, vol i. The Hans Rausing Endangered Documentation Project; London: SOAS. —— (ed.) (2004). Language Documentation and Description, vol. ii. The Hans Rausing Endangered Documentation Project; London: SOAS. Bert, M. (2001). “Rencontre de langues et francisation: l’exemple du Pilat”, Ph.D. thesis, Sciences du Langage, Université Lumière Lyon 2, France. Bouquiaux, L., and Thomas, J. (1976). Enquête et description des langues à tradition orale, i: L’enquête de terrain et l’analyse grammaticale; ii: Approche linguistique;

74


iii: Approche thématique. Paris: Société d’etudes Linguistiques et Anthropologiques de France. —— (1992). Studying and Describing Unwritten Languages. Dallas: SIL. (Translation of Bouquiaux and Thomas 1976.) Cameron, D., et al. (1992). Researching Language: Issues of Power and Method. London: Routledge. Campbell, L., and Muntzel, M. (1989). “The structural consequences of language death”, in N. Dorian (ed.), Investigating Obsolescence: Studies in Language Contraction and Death. Cambridge: Cambridge University Press, 181–96. Craig, C. (1989). Review of N. Rigby and R. Schneider, Dictionary of the Rama Language:Rama–English–Rama–Creole–Spanish/English–Rama. Speaking with the Tiger, vol. ii. Berlin: Dietrich Reimer Verlag, International Journal of American Linguistics, 56/2: 293–304. —— (1992a). “Miss Nora, rescuer of the Rama language: A story of power and empowerment”, in K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating Power: Proceedings of the Second Berkeley Women and Language Conference, Berkeley, i. 80–9. —— (1992b). “A constitutional response to language endangerment: the case of Nicaragua”, in K. Hale, M. Krauss, L. Watahomigie, A. Yamamoto, C. Craig, J. La Verne Masayesva, and N. England (eds.), 11–16. —— (1993). “Fieldwork on endangered languages; a forward look at ethical issues”, in A. Crochetière, J. C. Boulanger, and C. Ouellon (eds.), Proceedings of the XVth International Congress of Linguists, vol i. Saint Foy: Les Presses de l’Université Laval, 33–42. Crystal, D. (2000). Language Death. Cambridge: Cambridge University Press. Dimmendal, G. J. (2001). “Places and people: field sites and informants”, in P. Newman and M. Ratliff (eds.), 55–75. Dixon, R. M. W. (1984). Searching for Aboriginal Languages: Memoirs of a Field Worker. St Lucia: University of Queensland Press. —— (1997). The Rise and Fall of Languages. Cambridge: Cambridge University Press. Dorian, N. (1982). “Defining the speech community to include its working margins”, in S. Romaine (ed.), Sociolinguistic Variation in Speech Communities. London: Edward Arnold, 26–32. —— (1986). “Gathering language data in terminal speech communities”, in J. Fishman et al. (eds.), The Fergusonian Impact, ii: Sociolinguistics and the Sociology of Language. Berlin: Mouton de Gruyter, 555–75. —— (1989). “Introduction”, in N. Dorian (ed.), Investigating Obsolescence: Studies in Language Contraction and Death. Cambridge: Cambridge University Press, 1–10. —— (1992). “Working with endangered languages: privileges and perils”, in A. Crochetière, J. C. Boulanger, and C. Ouellon (eds.), Proceedings of the XVth International Congress of Linguists. i. Saint Foy: Les Presses de l’Université Laval, i. 11–22. —— (2001). “Surprises in Sutherland: linguistic variability amidst social uniformity”, in P. Newman and M. Ratcliff (eds.), 133–51. Dressler, W. (1978). “Language shift and language death: a protean challenge for the linguist”, Folia Linguistica, 15/1–2: 5–27. England, N. (1992a). “Doing Mayan linguistics in Guatemala”, Language, 68/1: 29–35. —— (1992b). Autonomía de los idiomas mayas: historia e identidad. Rukutamil, Ramaq’il, Rutzijob’al: Ri Layab’ Amaq’. Guatemala: Editorial Cholsamaj.


75

—— (1997).“Mayan efforts toward language preservation”, in L. Grenoble and L. Whaley (eds.), 99–116. —— (2001). Introducción a la gramática de los idiomas mayas. Guatemala: Editorial Cholsamaj. Evans, Nicholas (2001). “The last speaker—long live the last speaker!”, in P. Newman and M. Ratliff (eds.), 250–81. Fishman, J. (1991). Reversing Language Shift: Theoretical and Empirical Foundations of Assistance to Threatened Languages. Clevedon: Multilingual Matters. Gerdts, D. (1998). “Beyond expertise: the role of the linguist in language revitalization programs”, in N. Ostler (ed.) Endangered Languages: What Role for the Specialist. Bath: Foundation for Endangered Languages, 13–22. Grenoble, L., and Whaley, L. (eds.) (1998). Endangered Languages: Language Loss and Community Response. Cambridge: Cambridge University Press. Grinevald, C., and Kauffmann, M. (2004). “Meditaciones en la sombre”, Wani: Revista del Caribe Nicaragüense, (Managua), 38: 34–48. Grinevald Craig, C. (1997). “Language contact and language degeneration”, in F. Coulmas, (ed.), Handbook of Sociolinguistics. Oxford: Blackwell, 257–70. —— (1998). “Language endangerment in South America: a programmatic approach”, in L. Grenoble and L. Whaley (eds.), 124–60. —— (2000). “Los lingüistas frente a las lenguas indígenas”, in F. Queixalos and O. Renault-Lescure (eds.), 37–53. —— (2002). “Linguistique et langues mayas du Guatemala”, in J. Landaburu (ed.), MésoAmérique, Caraïbes, Amazonie: Faits de langues. 17–25. —— (2003a). “Speakers and documentation of endangered languages”, in P. Austin (ed.), 52–73. —— (2003b).“Educación intercultural y multilingüe: el caso de los ramas”, Wani: Revista del Caribe Nicaragüense, (Managua), 34: 20–38. Hagège, Claude (2000). Halte à la mort des langues. Paris: Editions Odile Jacobs. Hale, K., Krauss, M., Watahomigie, L., Yamamoto, A., Craig, C., La Verne Masayesva, J., and England, N. (1992). “Endangered languages”, Language, 68/1: 1–42. Himmelman, N. (1998).“Documentary and descriptive linguistics”, Linguistics, 36: 161–95. Hinton, L. (1994). Flutes of Fire. Berkeley: Heyday Books. —— and Hale, K. (eds.) (2001). The Green Book of Language Revitalization in Practice. San Diego: Academic Press. Jackson, B. (1987). Fieldwork. Urbana: University of Illinois Press. Kroeber, T. (1961). Ishi in Two Worlds: A Biography of the Last Wild Indian in North America. Berkeley and Los Angeles: University of California Press. Mithun, M. (2001).“Who shapes the record: the speaker and the linguist”, in P. Newman and M. Ratliff (eds.), 34–54. Mueller, G. (2001). “Defending Rama Indian community lands. Florida: Four Directions Geographic Consulting. 4DGC, 77, http://lic.law.ufl.edu.academics/conservation/ projects_i_rama.htm Nagy, N. (2000).“What I didn’t know about working in an endangered language community: some fieldwork issues”, in N. Dorian (ed.), Small Languages and Small Language Communities, 32. International Journal of the Sociology of Language, 144: 143–60. Nettle, D., and Romaine, S. (2000). Vanishing Voices: The Extinction of the World’s Languages. Oxford: Oxford University Press. Newman, P., and Ratliff, M. (eds.), (2001). Linguistic Fieldwork. Cambridge: Cambridge University Press.

76


Ospina Bozzi, A.-M. (2002). Structures élémentaires du Yuhup Makú, langue d’Amazonie colombienne: morphologie et syntaxe. Paris: Université Paris VII. Payne, T. (1997). Describing Morphosyntax: A Guide for Field Linguists. Cambridge: Cambridge University Press. Queixalos, F., and Renault-Lescure, O. (eds.), (2000). As linguas amazônicas hoje: The Amazonian Languages Today. São Paulo: IRD, ISA, MPEG Instituto Socioambiental (distribution: www.socioambiental.org). Rice, Keren. (2001). “Learning as one goes”, in P. Newman and M. Ratliff (eds.), 230–49. Samarin, W. (1967). Field Linguistics: A guide to Linguistic Fieldwork. NewYork: Holt, Reinhart and Winston. Sasse, H.-J. (1992).“Language decay and contact-induced change: similarities and differences”, in M. Brenzinger (ed.), Language Death: Factual and Theoretical Explorations with Special Reference to East Africa. Berlin: Mouton de Gruyter. Shopen, T. (ed.) (1985). Language Typology and Syntactic Description, vols. i–iii. Cambridge: Cambridge University Press. UNESCO (2003). Language Vitality and Endangerment. Report from the Intangible Cultural Heritage Unit’s Ad Hoc Expert Group on Endangered Languages. Approved 31 March 2003 by the Participants of the International Expert Meeting on the UNESCO Programme Safeguarding of Endangered Languages. Paris-Fontenay, MS, 25 pp. Vaux, B., and Cooper, J. (1999). Introduction to Linguistic Field Methods. Munich: Lincom Europa. Wilkins, D. (2000). “Even in the best of intentions: some pitfalls in the fight for linguistic and cultural survival”, in F. Queixalos and O. Renault-Lescure (eds.), 61–85. Wolcott, H. (1995). The Art of Fieldwork. London: Altamira Press, Sage Publications. Woodbury, T. (2003). “Defining documentary linguistics”, in P. Austin (ed.), SOAS, 35–52. Yasugi, Y. (ed.) (2003). Materiales de lenguas mayas de Guatemala, vols. i–iv. ELPR 2003 A1–007.

4

Language Policy and Language Rights David Bradley

4.1. Policy and reality 4.1.1. Western Pacific Rim 4.1.2. Southern Pacific Rim 4.1.3. Eastern Pacific Rim 4.2. Can policy change reality? 4.3. Conclusion

77 77 82 86 87 88

4.1. Policy and Reality In most countries along the Pacific Rim, there is an increasing mismatch between official policy, which has become progressively more liberal and supportive of minority linguistic and other rights, and the reality of continuing loss of indigenous minority languages. This is partly because assimilationist policies, explicit or implicit, had been under way for centuries or millennia before recent changes; and because members of minority groups themselves choose for economic and social reasons to continue to assimilate, despite now having nominal linguistic rights. As countries develop and economic security is approached, many governments allow their minorities the luxury of some language and cultural rights. This does not change the underlying attitude of the majority group, and it also does not necessarily lead to a positive change in the attitudes and choices of members of minority groups. The following sections briefly trace and compare recent language policy developments for minority languages in the countries of the Pacific Rim. For some more detailed studies of language policy in South-East Asia, see Bradley (1985).

4.1.1. Western Pacific Rim In Japan,1 the long-standing policy of separation between majority Japanese and the Ainu in the north of the country changed during the nineteenth century. Up to 1855, access to the Ainu in their main territory of Hokkaido I gratefully acknowledge the funding support of the UNESCO Endangered Languages programme and of the Australian Research Council (A59701122, A59803475, A00001357). 1 I would like to thank Theresa Savage and Hiroshi Nakagawa for information on the Ainu.

78

d av i d b r a d l e y

was mediated through the Matsumae regional feudal lords. By this time, many Ainu had doubtless already assimilated into the Japanese majority, especially in northern Honshu, and Ainu in Hokkaido began to learn Japanese after 1799. Especially after 1869 when Ainu land rights were eliminated, with further assistance from the assimilationist Aborigines Protection Law of 1899, there was very rapid in-migration of majority Japanese, catastrophic economic and cultural change, and breakdown of intergenerational language transmission in most areas. This is why the most fluent Ainu speakers are now very old people, the last generation to learn the language as children in a family setting in the early twentieth century. The Ainu from the Kurile Islands were moved to Hokkaido in 1884, and those in southern Sakhalin were moved to Hokkaido in 1945. With the Ainu Cultural Protection Law of 1997, the official policy has changed, and there is a small-scale cultural revival under way, led by official bodies (museums and cultural centres) funded by the central, Hokkaido prefecture, or local governments, as well as some commercial touristoriented ventures and individual craftsmen. Over the last twenty years, the number of language revival initiatives has multiplied, with fourteen centres now teaching Ainu language in a limited way, though not yet in school to Ainu children. For more details, see Chapter 19 (this volume, by Nakagawa and Okuda). Korean is the national language of the two Koreas, and is also the language of large minority groups in adjacent areas of China and Russia, as well as migrants in Japan and the United States. Koreans in China have national minority status, and there are special provisions for Korean education, including language maintenance, in various overseas countries. While there are lexical and other differences within Korean, the language is not endangered, though some Koreans feel that English influence is too great in the South. The vestiges of ancient and long-standing Chinese influence, as in Japan, remain extensive, both culturally and linguistically, though the use of Chinese characters is much less than in Japan. Not having any indigenous minority languages, the two Koreas obviously have no policy for them. The former Soviet Union implemented its nationalities policy from the 1920s, dividing its ethnic groups into 130 nationalities, including some which were more advanced in the western and central part of the state. Many others, including all those in the Pacific Rim area, were less advanced. For them, the Soviet period led to major advances: the creation and dissemination of writing systems, support for literary and cultural activities,

l a n g uag e p o l i c y a n d l a n g uag e r i g h ts

79

the development of an elite group within each nationality, and varying degrees of political autonomy. The nationalities policy in the Soviet period was initially formulated by Lenin in 1922, and implemented more or less immediately. A nationality required four characteristics: a language, a territory, a culture, and a means of economic production. There was an explicit policy of unifying small related groups into larger nationalities, which was carried even further in designating the national minorities of China in the early 1950s. In the Pacific Rim area of the Soviet Union, various Palaeosiberian and Tungusic nationalities were recognized, along with Nivkh (Gilyak; Ch. 19). Nivkh and the Tungusic languages, along with most Palaeosiberian languages in eastern Russia, are now vanishing, despite the efforts of Soviet linguists and educationists, as discussed in Chapter 20 (by Tsumagari, Kurebito, and Endo). One of the reasons for the break-up of the former Soviet Union was precisely the national aspirations of the larger nationalities. The fifteen nationality-based areas which had full Soviet Union Republic status, including Russia, are now independent states. Many of the smaller nationalities in the twenty autonomous republics, eight autonomous regions, and ten autonomous areas within Russia and some other independent states are in the process of seeking greater autonomy; for example, Chechnya in the Caucasus area of Russia, or South Ossetia in Georgia, among others. The nationalities policies of the various communist states in East and South-East Asia derive directly from those of the Soviet Union. Russian anthropologists and linguists were directly involved in the initial surveys and classification work. In China four of the nationalities were recognized under the Republic of China after 1911, but fifty-one more national minorities received their official recognition only during the 1950s, after the creation of the People’s Republic of China. In Vietnam the full classification and recognition of fifty-four minority ethnic groups took place during and after the struggle for independence and unification, in the North after 1954 and in the South after 1975. In Laos, the recognition of forty-seven ethnic groups including the majority Lao took place after the unification of the country under the Pathet Lao government in 1975; a new list of forty-nine has recently been approved, with some amalgamations and deletions. In these nations, the national language and main medium of education and administration is the language of the majority group: Mandarin Chinese in China, Viet in Vietnam, and Lao in Laos. For more details on vanishing languages in these and other other East Asian states, see Chapter 17 (Poa and LaPolla).

80


There are four other areas with a mainly Chinese population: Taiwan, Hong Kong, Macao, and Singapore. After its return to China in 1945, Taiwan became the sole remaining area controlled by the Republic of China when the People’s Republic of China was established on the mainland. The policy is that Mandarin2 is the national language, but in Taiwan this is called Guoyu (national language), while the similar variety chosen as the standard in China is called Putonghua (common speech). Another standard variety of Mandarin is one of the four official languages of Singapore, where it is called Huayu (Chinese speech). In China over 75 per cent use a Mandarin subvariety as mother tongue, but other important varieties of Chinese are used in the coastal provinces. Both in Taiwan and in Singapore, the vast majority of the population speaks other types of Chinese, not Mandarin, as their traditional mother tongue. The overall unity of Chinese, despite mutual unintelligibility between major spoken varieties, is maintained by its shared writing system, history, and culture. For further discussion of Chinese language policy, see Bradley (1991). The use of indigenous Minnanhua (Southern Min speech) in Taiwan has never ceased, and indeed seems to be spreading. Knowledge of Japanese from the pre-1945 period is receding. In addition there are fifteen indigenous Austronesian languages still spoken on the eastern side of Taiwan, as discussed in Chapter 15 (by Tsuchida and Tsukida); many of these are vanishing, and some others have already disappeared. The Austronesian languages of Taiwan are also spoken by very small groups on the mainland, where they are classified in a single Gaoshan (high mountain) national minority. Hong Kong was established under British rule in 1841, with a short interlude of Japanese occupation from 1941 to 1945. Since the return of Hong Kong to China in 1997, the local Yue dialect is continuing in use, but the use of Mandarin is spreading. Cantonese, the common English name of Yue, is derived from the English name of the nearby city Guangzhou, formerly also known as Guangdong, which is now the name of the entire adjacent province. In nearby Macao, since its return to China after many centuries of Portuguese rule in 1999, the local Yue speech also remains in use, with Mandarin on the rise. Unlike Hong Kong, where the former colonial language is still widely used and studied, in Macao the study and use of Portuguese has greatly decreased, and the indigenous Portuguese creole of Macao is rapidly vanishing. Also disappearing are the indigenous non-Yue 2 Mandarin or guanhua (official speech) is the mother tongue variety of the north, west, and southwest of China, spoken by the Han Chinese majority. The standard is based on the speech of Beijing.


81

speech varieties of Hong Kong, in particular Hakka. In Taiwan, Hong Kong, and Macao, the use of traditional full-form Chinese characters continues. However, China and Singapore now use simplified characters as implemented in China in 1957. Singapore, established by the British in 1819, rapidly became a majorityChinese trading port. Most (77 per cent) of its resident population is ethnic Chinese, with smaller Malay (14 per cent), Indian (8 per cent), and other groups. It was one of the British-ruled Straits Settlements surrounded by various Malay states in treaty relations with the British, like them was occupied by Japan from 1942 to 1945, joined the Malaysian Federation in 1963, but left it in 1965 for independence. The national languages are English, Mandarin, Malay, and Tamil. Singapore English is a New English, indigenized and very widely spoken. The main policy initiative in Singapore has been to spread spoken Mandarin and simplified characters at the expense of the mother tongue south-eastern Chinese “dialects” since 1978. The national language of Cambodia is Khmer, which is used for all purposes within the country. Cambodia has no explicit policy for minorities; this is because the minorities there are a small part of the population, live mainly in the most remote areas of the north-east of the country, and have been under the political and cultural sway of the Khmers for at least a thousand years. The French colonial period from the mid-nineteenth century to 1954 added French to the mix, but otherwise made little difference, since the French classified all non-majority groups in all of Indochina (including Laos and Vietnam) as “montagnards” (hill people). This is not to say that the French were not aware of the ethnic diversity of Indochina; rather, they chose to ignore it administratively, and govern through the majority groups. In the post-Khmer Rouge period since 1979, at least the minorities are not being targeted and hunted down, as some were, but there are still no policies or efforts to maintain minority languages. Burma (Myanmar) became a British colony gradually during the nineteenth century, with the process completed in 1885, and the British implemented a policy similar to that of India. This gave some political autonomy to local princes who had been in a tributary relationship with the Burmese monarchy. The British classified minority ethnic groups according to the over-general traditional categories of the majority Burmans. After independence in 1948, there was a nominally federal system with seven minoritybased states (Arakan, Chin, Kachin, Shan, Kayah, Mon, and Karen), but with continuous ethnic rebellions since 1948 and unbroken military rule since 1962, the nation now promotes only the Burmese language. The teaching of

82


minority languages has long been illegal during normal school hours. In the early 1990s a new classification of the nation’s ethnic groups gave a total of 135, including the majority Burmans, the seven with separate states, and 127 others. From the mid 1990s, a series of truce agreements with many of the ethnic rebel groups led to the spread of central administration and Burmeselanguage education into areas which had been off limits to the government for up to fifty years. Thailand has never been a colony, and the Thai ethnic group comprises the vast majority. Their language is one of the core elements of Thai national identity. In-school education in other languages, notably Chinese for ethnic Chinese, was long illegal but tolerated. Official figures used to recognize nine tribal groups, mainly in the north, but the most recent have increased this to sixteen, including three types of Thai-related languages from adjacent countries. The national language of Malaysia is Malaysian, derived from traditional Malay as spoken by the majority, especially in West Malaysia. Various types of Chinese are also spoken by urban groups; Portuguese creole is still spoken in Melaka (Malacca) by some descendants from the Portuguese colonial period (1511 to 1641). In West Malaysia, there are small indigenous Orang Asli (original people) groups, mainly speaking Austroasiatic languages, and in East Malaysia there are other indigenous Austronesian-speaking groups. Malaysia was created in 1963 by uniting various Malay states in treaty relations with the British, the former Straits Settlements colonies, as well as Sarawak (formerly run by a British family) and Sabah (formerly British North Borneo, run by a British company) in East Malaysia. Brunei remained outside this federation and is still an independent Malay-ruled state, and Singapore withdrew from Malaysia in 1965.

4.1.2. Southern Pacific Rim A long-term Dutch colony, Indonesia was granted nominal independence under Japanese occupation from 1942 to 1945. A struggle against the Dutch led to independence in 1949, at first in union with the Netherlands, but since 1954 complete. One of the main platforms of the national struggle, and one of the five pillars of pancasila (five pillars), the national philosophy, is the Indonesian language, a codified version of Malay. Initially this was the mother tongue of very few Indonesians, but it has been spread throughout Indonesia, with most people now either bilingual in Indonesian and a traditional language, or monolingual in Indonesian. A compromise spelling


83

which unifies the representation (but not the vocabulary!) of Indonesian and Malaysian was implemented in 1972, and also adopted for Malay in Brunei. Eight other major languages have also received limited official status in Indonesia, but these are not among the many Austronesian languages throughout the country and Papuan languages in the east of the country which are endangered. The spread of education and administration in Indonesian is increasing the pace of their disappearance, and even some languages with rather large numbers of speakers may be potentially endangered, for example Sasak of Lombok Island with over two million ethnic group members. Timor Leste (East Timor) was a Portuguese colony until 1975, was then invaded and annexed by Indonesia, then taken under United Nations stewardship in 1999, and achieved independence in 2002. The independent state has chosen Portuguese as its official language, but English is also widely used, and a distinct variety of indigenous Austronesian Tetun is used as a lingua franca in some areas, notably the capital, Dili. Hajek (2002) traces the history of language policy and language endangerment in Timor Leste, noting that several of the indigenous Papuan and Austronesian languages are endangered, and some may have vanished during Indonesian occupation. After centuries of Spanish rule, the Philippines were sold to the United States in 1898 and eventually achieved independence in 1946. All the indigenous languages of the Philippines are Austronesian, with a great deal of diversity. The language of the Manila area, Tagalog, with some input from regional languages, has been developed into the national language, Pilipino, which over the last fifty years has spread from an original quarter of the population to be nearly universal either as first or second language, especially among the young and the educated. Under American rule and with many American teachers, English also became very widespread, and now Philippine English is one of the major New Englishes. Tagalog/Pilipino uses an amazing number of English loan words, and there is also very frequent code switching between Tagalog/Pilipino and Philippine English. Nine other regional languages have local official status in their regions, but many other indigenous Austronesian languages are endangered to various degrees. Some Spanish creoles which had developed during the Spanish colonial period are now extinct. In the late nineteenth century, the island of Papua was divided between the Dutch in the west, the Australians in the south-east, and the Germans in the north-east. From 1914 to the independence of Papua New Guinea in 1975, all of the eastern half and nearby islands to the north-east were under

84


Australian administration. In addition to English, the Australians used Motu, an Austronesian language of the capital, Port Moresby, as a lingua franca along the south-eastern coast; they also took over from the Germans the use of the developing Pacific Pidgin, and this has since spread throughout the country. The western half of Papua passed to Indonesia after 1962 and is now known as Irian Jaya (West Irian). Indonesian has become the national language there, and is endangering many indigenous languages which have no official recognition. Papua and some adjacent islands to the west in Indonesia are the area of greatest linguistic diversity in the world, with many hundreds of indigenous Papuan languages in a large number of unrelated families, as well as many Austronesian languages. Papua New Guinea, the Solomons, Vanuatu, and adjacent areas are also the focus for the expansion of Pacific Pidgin, which as Tok Pisin has official status in Papua New Guinea, as Bislama in Vanuatu, and as Solomon Islands Pidgin in the Solomons. Pacific Pidgin arose during the plantation period from the late nineteenth century, then creolized and spread widely after workers returned home. In many cases, indigenous languages are being replaced by Pacific Pidgin, as is also the case among some Aboriginal populations in north-eastern Australia, where the creole Broken is spreading in the Torres Strait and the northern tip of Queensland, and Kriol is taking over in north-west Queensland and some northern parts of the Northern Territory. The numerous Pacific island states with their various Austronesian languages went through various colonial hands, and in each case the colonial period left an influence on language policy. Some were under German rule briefly before 1914. New Caledonia and French Polynesia remain under direct French rule, and Guam, Western Samoa, and Hawaii are under United States rule, so French and English are replacing indigenous languages there. Easter Island, whose Austronesian language has vanished, is under Chilean rule. The Federated States of Micronesia were first German, then Japanese, and finally American before their independence. Prior to independence in 1968, Nauru was under British and then Australian rule; some smaller islands such as Christmas, Cocos, and Norfolk are still administered by Australia. The Cook Islands, Niue, and (Western) Samoa were controlled from New Zealand up to the 1960s or 1970s, Fiji, Kiribati, the Solomons, and Tuvalu went through periods of British rule, with independence in the 1970s. Tonga remains a British protectorate under its indigenous monarchy, and Pitcairn is still a British colony. Vanuatu was a joint British/French condominium up to independence in 1980. The more easterly Pacific states with


85

single indigenous Polynesian languages are linguistically more unified, while there is extreme internal diversity even among the Austronesian languages of Micronesia and especially Melanesia (Vanuatu, New Caledonia, the coastal areas of Papua New Guinea, and so on). With official policies in favour of colonial or ex-colonial languages and the spread of Pacific Pidgin, Melanesia is an area where many indigenous languages are vanishing. However, some initiatives are under way to reverse this situation, such as the development of orthographies for over a hundred additional languages in Papua New Guinea (Easton 2003), the development since 1977 of the Vanuatu Cultural Centre and linguistic training for members of indigenous communities in Vanuatu, and so on. For more details on the language situation in the Pacific area, see Chapters 7 (by Ehrhart and Mühlhäusler) and 14 (by Sakiyama). English was finally recognized as the national and official language in Australia in 1988, but this had long been the policy in any case. Australia only gave full civic rights to its Aboriginal population in 1967. As Walsh outlines in his chapter (Ch. 12), the previous policy, with honourable exceptions at some missions which were based on one language area, was one of rapid assimilation and exploitation at best, extermination at worst. In particular, the Stolen Generation of mixed-race Aboriginal children were removed from their mothers and communities systematically up until the 1960s; a very effective way to break language transmission! In many areas, people from a variety of language backgrounds were mingled together on missions and reserves, and strongly discouraged from using any traditional language. There are now strong incentives for language maintenance. Since 1992 the basis for successfully claiming land rights under the Australian High Court’s Mabo decision3 is a continuous association with traditional land and its traditional language and culture. Since 1967, and especially with the initiatives of the national language policy from 1988, major efforts have been made to document surviving languages and provide two-way (bilingual) education where appropriate. Language revival efforts are also under way in some areas where the pre-1967 policy was sadly effective, and many communities are receiving support for initiatives such as dictionaries and other materials. The best-documented revival is of Kaurna in Adelaide (Amery 2000). For further details, see Chapter 12 (by Walsh). New Zealand has only one large indigenous group, the Maori with their Polynesian language. The policy in New Zealand since 1987 is that both 3 The late Eddie Mabo successfully claimed land rights over Murray Island in the Torres Straits in a decision of the Australian High Court released 3 June 1992.

86


English and Maori have official status. Due to its history of colonial tutelage of various other Pacific states to its north and east, there are also substantial migrant populations from other Pacific island groups (Cook Islands, Niue, Samoa, Tonga, and so on). Until the 1960s, assimilation was the official policy, and the proportion of Maori who could speak Maori was rapidly decreasing. From 1981, the 1840 Treaty of Waitangi4 was revisited, and the government started to support Maori language maintenance. The te kohanga reo (language nest), preschool centres where young Maori children are taught through language immersion by Maori elders who, unlike most of the children’s parents, can speak Maori fluently, are one successful initiative which has been widely imitated in language revival efforts around the world.5 Grandparents as secret weapon for language maintenance are a widely neglected resource, also used in many school-based and extracurricular activities in education for Australian Aboriginal communities, or in the United States master and apprentice system of intensive one-on-one training by a tribal elder.

4.1.3. Eastern Pacific Rim Policies along the eastern Pacific Rim have been even more assimilationist than the western Pacific Rim policies. In Canada, Original Nations now have increased legal and territorial rights, and in the United States, tribal rights originally based on territorial reservations have gradually improved in recent years. In both, the former catastrophic policy of forced assimilation by removal of generations of children to English-only schools away from traditional areas, paternalistically intended as an acculturation and linguistic assimilation device, has gravely damaged the transmission and viability of most indigenous languages. Now there is a major effort under way, partly through indigenous networks and partly through educational institutions, to reinvigorate many of these languages. However, the English Only movement in the United States is making such efforts more difficult. The four chapters (Chs. 21 to 24) discuss the North American situation. The original inhabitants of the fiftieth state of the United States, Hawaii, were speakers of the Polynesian language Hawaiian. Massive immigration

4 This treaty between the British crown and the Maori chiefs permitted and led to the settlement of large numbers of English-speaking settlers, soon greatly outnumbering the existing Maori population. 5 The first hundred te kohanga reo were established in 1982. By 1994 there were 800, with 14,000 children attending, and over 60,000 graduates.


87

from the western Pacific Rim and from the continental United States has made them a minority on the main islands, but some outer islands still have a Hawaiian majority. The language was in severe decline, but is undergoing a revival, with large numbers of young Hawaiians studying the language in school, at university, and elsewhere. They benefited from official bilingual education policies for a number of years, until these were revoked under the Reagan administration. There is also a substantial positive effect for attitudes about Hawaiian identity from cultural tourism activities. In western Latin America, the hegemony of Spanish has been extremely forceful, though not quite as locally intense as in the United States or Canada. This was partly due to the initial adoption of many indigenous languages by the Catholic Church. The remaining population of Indian descent is also a much larger proportion of the overall population than in North America, and in many cases they are geographically concentrated in areas with a much smaller population of Hispanic descent. For these reasons, some of the indigenous languages of Latin America are less endangered than those in North America. Recent policy changes are not helping, however; education and other government activities continue to be almost exclusively Spanish-medium, and centuries of intermarriage and the increasing development of mainly Spanish-speaking urban centres are continuing to erode the indigenous language speaker populations. Again, the three chapters (Chs. 9 to 11) provide more details for the Pacific Rim in Central and South America. Overall, increased minority linguistic and other rights have been officially implemented in most countries around the Pacific Rim, but so far with limited positive results. Languages nevertheless continue to vanish, and linguistic documentation remains an urgent priority.

4.2. Can Policy Change Reality? The most important factor in language maintenance, and an absolutely crucial one in cases of language revival from near-extinction such as Ainu in Japan, is the attitude of the minority community as a whole; see Bradley and Bradley (2002a) for fuller discussion. For a language to survive or reemerge and develop into wider use in its traditional community, a viable core group of individuals in the community must be willing to make the effort to implement a language programme, and enough motivated participants, including children, must be involved. This effort must be sustained,

88


and official resources should be drawn upon where possible. The strategies for such activities are discussed much more fully in other chapters of this volume. Of course it is much easier to have a positive attitude and support folkloristic heritage activities (songs, dance, crafts, festivals, and so on) than it is to maintain or revive a vanishing language. I do not wish in any way to denigrate the importance of such heritage activities, which have been so effective in keeping some Celtic communities strong in the face of the advance of English and French. They are also a powerful weapon in the revival of languages such as Manx and Cornish, or revitalization efforts for Welsh and Breton. It seems that it is usually only after increased economic and social security have been achieved that a minority group choose to go back to a language they had lost or were losing. Revival to date has mainly been successful in developed first world countries: Hebrew in Israel and Welsh in Wales are the prime examples. Hebrew was a liturgical language which became a modern spoken language over the last century, and the language of an independent nation since 1948. Welsh was a retreating minority language used in increasingly restricted domains until the 1960s, but its domains have now greatly expanded and it has a high public profile. Even those groups who are reclaiming their traditional languages in Australia and North America live within prosperous nations which give substantial support for such measures. The problem is that most endangered languages are spoken in less prosperous nations which not unreasonably choose to devote scarce resources mainly to education in the national language. Another fundamental problem is that many nations still have a covert if not overt assimilationist policy, and the negative attitudes to minority languages which often accompany this.

4.3. Conclusion Does it matter if a language dies? For the yes case, there are four main arguments (Bradley and Bradley 2002b: pp. xi–xii). 1. Linguistic: each language is a unique system of human communication; 2. Ethical: language is part of community heritage and access is an essential right; 3. Scientific: each language encapsulates unique knowledge of the natural world; 4. Symbolic: group identity and self-esteem is tied to language.


89

The most important of these to an individual community is probably the symbolic argument: preservation of a positive group identity and positive self-esteem. Also important is the scientific argument for maintenance of traditional cultural knowledge, and the ethical argument of connection with the community’s roots. Does it matter if the official policy is negative? Obviously yes, because this means that majority and minority attitudes are likely to be negative, and the minority languages will be used in increasingly restricted domains. Government resources will not be available, and in particular schools and dealings with the government will become a means of linguistic assimilation. In turn, many parents will make what they believe to be a rational economic decision, not to transmit the traditional minority language, believing that their children will do better in school and advance further in society with the national language. Sadly, the bilingual advantage—the fact that bilinguals tend to do better at a second, national language than monolinguals— is something we as linguists have not yet managed to spread into public consciousness. Once the national policy becomes more positive for language maintenance, as it now has in most countries in the Pacific Rim, each minority community with an endangered language is faced with a decision: do they take action and implement measures to preserve their language and culture? Do they insist that schools should become not just a place for assimilation, but also a place for keeping their language and culture alive? Do they make other efforts to maintain and develop their traditional life in as many settings as possible? Do they persevere and succeed? And how is such success to be measured? It is solely for the community itself to decide how to proceed. Ladefoged (1992) makes the case for allowing languages to die quietly, according to the wishes of their speakers. Thieberger (2002) points out that it is selfinterested and often unrealistic for linguists to insist on full-blown language maintenance programmes directed at all children, and that communities must decide themselves what to maintain and how to go about it. I believe our role as linguists is to document languages in cooperation with the speech communities, and try to help them to do what they want with our results and their own knowledge. Sometimes we must wait for the community to be ready for language maintenance. One such instance is further discussed in my chapter on South-East Asia (Ch. 16). If this kind of work is done well, fewer languages will vanish, knowledge about the past will survive, and positive feelings based on traditional heritage will persist in the hearts of many more communities.

90


References Amery, R. (2000). Warabarna Kaurna! Reclaiming an Australian Language. Lisse: Swets & Zeitlinger. Bradley, D. (ed.) (1985). Language Policy, Language Planning and Sociolinguistics in South-East Asia. Papers in South-East Asian Linguistics 9. Canberra: Pacific Linguistics A–67. —— (1991). “Chinese as a pluricentric language”, in M. G. Clyne (ed.), Pluricentric Languages. Berlin: Mouton de Gruyter, 305–24. —— and Bradley, M. (eds.) (2002) Language Endangerment and Language Maintenance. London: RoutledgeCurzon. —— and Bradley, M. (2002a). “Language attitudes: the key factor in language maintenance”, in D. Bradley and M. Bradley (eds.), 1–10. —— (2002b). “Introduction”, in D. Bradley and M. Bradley (eds.), pp. xi–xx. Easton, C. (2003). “Designing orthographies through community interaction: Alphabet Design Workshops in Papua New Guinea”, paper presented at Conference on Language Development, Language Revitalization and Multilingual Education in Minority Communities in Asia, Bangkok, Thailand. Hajek, J. (2002). “Language maintenance and survival in East Timor: all change now?”, in D. Bradley and M. Bradley (eds.), 310–28. Ladefoged, P. (1992). “Another view of endangered languages”, Language, 68/4: 809–11. Thieberger, N. (2002). “Extinction in whose terms? Which parts of a language constitute a target for language maintenance programs?”, in D. Bradley and M. Bradley (eds.), 182–202.

5

Using Written Records to Revitalize North American Languages Toshihide Nakayama

5.1. Introduction 5.2. Impact of unpublished materials on the study of endangered languages 5.2.1. Significance of the archived records: the case of Sapir’s Nuuchahnulth materials 5.2.1.1. Sapir’s work on Nuuchahnulth 5.2.1.2. Value of the fieldnotes 5.2.2. Challenges in utilizing archived materials 5.2.2.1. Issues in locating archived materials 5.2.2.2. Issues in extraction of information from archived materials 5.3. Maintaining a field philological chain: from the past to the future 5.3.1. Field philological approach to the past 5.3.2. Extending a field philological approach to the future 5.4. Conclusion

91 94 94 94 96 97 97 98 99 100 102 104

5.1. Introduction In research on endangered languages it is relatively common to downplay the value of written records or any other second-hand attestation of the actual uses of the language as a source of information. This tendency undoubtedly has to do with the fact that the collection of primary data has been felt to be more important and urgent than devoting time to secondary records of the language when endangered language communities are going through drastic and rapid acculturation. This feeling has been strengthened by the fact that many of the previous records were of inferior quality, especially in terms of phonetic accuracy. The tendency for the researchers to rely so heavily on the data which they themselves collect has led to a feeling that their field records are mainly notes for themselves. Moreover, within the tradition of descriptive linguistic studies, it is the completed analysis that has generally been considered the most I would like to thank Ives Goddard, M. Dale Kinkade, Thomas Hess, Victor Golla, Marianne Mithun, Osahito Miyaoka, and Michael Krauss for their helpful comments and guidance regarding earlier versions of this chapter. All omissions and errors are mine alone.

92

t o s h i h i d e n a k ay a m a

valuable—fieldnotes and other records have value only in their relation to the final analysis. Thus, the combination of a strong emphasis on fieldwork and the exclusive focus on completed analysis has contributed to the peripheral (underappreciated) status of field records, which is the case with the study of endangered languages of North America (a similar concern was raised for the research in Australia by Austin (2002)). The records themselves are not felt to be inherently valuable, nor is information regarding the creation of these records considered to be very important. Following this line of thinking, field records are essentially not much more than private records of the fieldworker’s research activities, and the primary value of the materials is historical or autobiographical in nature. However, as the seriousness of language endangerment has intensified around various parts of the world, the collection of data directly through fieldwork has become increasingly difficult. In many cases the traditional languages exist only in the memories of a small number of elders, the languages having long since passed from daily use. Stories and words might be remembered, but only as historical artefacts, rather than as part of a living tradition. Under such circumstances, it is difficult to collect rich information about the actual use of the languages. Consequently the records of previous fieldwork constitute an important, and sometimes the sole, source of information regarding a traditional language. In other words, such records cease to be a mere snapshot of reality and have become one-of-a-kind materials and a primary object of study. The validity of the records can no longer be examined easily against real world facts. Instead, the original facts have to be surmised on the basis of these records. Thus, the role of the records in understanding facts about the language shifts from that of “reflection of reality” to “basis of a (reconstructed) reality”. This change in the role of field records necessitates a change of emphasis in the study of endangered languages. It has become necessary to direct attention to the documents themselves and also to the process of creating and interpreting the documents. In interpreting such data, we need to adopt an appropriate methodology for evaluating documents and extracting information about a language without the possibility of directly observing its actual use. Essentially, descriptive linguists need to bring in a philological perspective and conduct their research in a correspondingly rigorous way. That the study of endangered languages has philological aspects has already been pointed out by some scholars in their analyses of the state of the study of North American Indian languages (Goddard 1973; Miyaoka 1992; Watanabe 1996). In this chapter I will

using written records

93

discuss the importance of a philologically enlightened approach to the study of endangered languages and also to language revitalization efforts.1 Here I will adopt Miyaoka’s (1992, 2001: 13) term “field philology” to refer to such an approach. All descriptive linguistic research is of course fundamentally philological in a broad sense. As Goddard (1973: 727) points out, “[w]hether the linguist works with his own fieldnotes or three-hundred-year-old manuscript dictionary, he must base his analysis not on the actual, internalized, language—nor even on the actual ‘performance’ of speakers—but on a record, a second-hand attestation, of a finite number of verbal manifestations of the language.” It is nonetheless necessary to discuss the need for a philological emphasis, especially in relation to the work on endangered languages. The attention and care that are required for proper handling of field records are far from trivial, and it takes conscious and organized efforts to bring philological rigour to the treatment of field materials. In addition, as the amount and range of information that can be collected through direct fieldwork become more limited, we need to explore ways to creatively maximize the usefulness of written records. Although the notion of philology typically suggests a relationship only to the past, it should be noted that the need for adopting a philological approach is not limited to the aspects of our research that look back to the past. Field researchers are users of previously collected materials, but at the same time they themselves are creators of field documents that are objects of philological research. They stand not only at the end of the material utilization cycle as users and information-miners, but also at the beginning of the cycle as creators of materials. As researchers engage in efforts to collect information in fieldwork, they create one-of-a-kind materials about a language, in some cases the only source of information, for future researchers and descendants of the community members. As contributors to future philological research, it is important for field researchers to be conscious of both the significance and the problems associated with field philological research. They should make a conscious effort to ensure that

1 Discussion in this chapter is limited to written field records, due to space constraints. However, basically the same discussion applies to audio and visual recordings created in fieldwork. Compared to written records, audio and visual recordings are less abstract forms of recordings and as such they may be less prone to ambiguities and misinterpretations. Still, any kind of recording, written or audiovisual, is inherently selective in terms of space and time, and what is captured is merely a slice of a life observed from a certain angle. Without accurate and adequate information regarding the situation of the recording, interpretation of the record can easily go amiss.

94


their contributions last as long as possible and are as helpful as possible for future users of their materials.

5.2. Impact of Unpublished Materials on the Study of Endangered Languages In many parts of the world, endangerment of indigenous languages is at a critical state, and the prospect of obtaining new primary data from these languages is rapidly diminishing. Previously collected field materials therefore play a crucial role in the documentation and revitalization of endangered languages. In this section I will first illustrate, citing Sapir’s Nuuchahnulth materials as an example, the potential of archived materials to make a significant impact on efforts to document and revitalize endangered languages of North America. I will then discuss practical challenges that are typically faced by researchers seeking information in archived field materials.

5.2.1. Significance of the Archived Records: The Case of Sapir’s Nuuchahnulth Materials Edward Sapir (1884–1939) is one of the most influential scholars in the linguistics and anthropology of Native American Indians. He carried out indepth studies on a wide variety of languages, including Wishram Chinook, Takelma, Yana, Southern Paiute, Nuuchahnulth, Navajo, and others. The quality of his data and the sharpness of his insights are well proven in his voluminous published work. However, the information ingeniously laid out in such publications only represents a portion of what this remarkable scholar accumulated in his notebooks and file slips. The rest, just like gold trapped in ore, is still waiting to be processed and utilized. His materials on Nuuchahnulth serve as a good example. I will illustrate below the richness of the unpublished materials left by Sapir by drawing on my own experience working with his Nuuchahnulth materials.

5.2.1.1. Sapir’s Work on Nuuchahnulth Sapir conducted fieldwork on Nuuchahnulth in the early 1910s right after he had become chief of the Anthropological Division, Geological Survey of Canada. In September 1910, Sapir spent several months in the Tseshaht community in Port Alberni, British Columbia, collecting information on


95

the language and ethnography of Nuuchahnulth. He returned to the field for a longer period from September 1913 to February 1914. During his two field trips, which amounted to approximately a year, Sapir filled twentyfour notebooks with numerous texts, lexical items, and ethnographic notes. In addition, on his second trip he successfully trained two young Nuuchahnulth interpreters, Frank Williams and Alex Thomas, to transcribe texts phonetically so that they could continue collecting information after Sapir had left the field. The contribution of Thomas was particularly great. Thomas continued to send texts to Sapir, which ultimately more than doubled the amount of Nuuchahnulth text materials available to him (Golla 1991). Moreover, Thomas’s materials added the perspective of a community member to the text collection. True to the Boasian tradition of field research, information collected by Sapir covers a wide spectrum of linguistic and cultural phenomena. Particularly notable are numerous ethnographic texts, including myths, legends (e.g. about extraordinary hunters and family origins), ethnographic accounts of various facets of traditions (e.g. rituals, marriages, feasts, warfare), and tales of extraordinary experiences. What makes these texts exceptionally valuable is the fact that these are accounts provided by native speakers in their own language. These texts, therefore, can project a fairly detailed and rich picture of the traditional Nuuchahnulth ways as the Nuuchahnulth people themselves understood them to be. This large collection of extended narrative texts is valuable also as a linguistic resource, providing a wealth of naturally occurring, in-context examples. The texts that Sapir collected were analysed and edited for publication in collaboration with Morris Swadesh (1909–67), who was a student of Sapir. Of these texts forty-four were published in 1939 along with a grammatical sketch and a glossary (Sapir and Swadesh 1939). Work on the Nuuchahnulth textual materials halted for some time due to the death of Sapir and to the Second World War. Then during 1946–8 Swadesh devoted much of his time to arranging and editing these texts for publication (Klokeid 1971). A second volume of Nuuchahnulth texts was published in 1955 with thirty-four texts, most of which were collected by Alex Thomas (Sapir and Swadesh 1955). There was a plan for a third volume that would have contained sixtyseven texts (thirty-four texts collected directly by Sapir and thirty-three texts written by Thomas), but that volume was never completed. Although the two published collections of texts made available a substantial amount of valuable material, over half of the texts still remained unpublished, along with virtually all Sapir’s ethnographic notes. There is, thus, much to recover from Sapir’s field records.

96


Fortunately, a large amount of Sapir’s unpublished materials on Nuuchahnulth is archived at the American Philosophical Society Library (in the Franz Boas Collection, Philadelphia) and at the Canadian Museum of Civilization (formerly the National Museum of Man, Ottawa). By far the most valuable items among them are Sapir’s own field notebooks, which were deposited in the American Philosophical Society Library by the Sapir family in 1972. The material consists of twenty-four notebooks with over 2,800 pages and approximately 750 file slips of Nuuchahnulth personal and place names. Nuuchahnulth forms in the fieldnotes were transcribed phonetically, and transcriptions are detailed and very accurate. Texts are written on every other line and for the most part they are accompanied by interlinear word-by-word translations. All the Nuuchahnulth data are put only on the right-hand pages, and relevant linguistic and ethnographical notes are on the facing pages. Also found in the archives are various secondary materials that derived from the fieldnotes. Among the important materials in this category are the typescripts of the texts that were prepared for the planned third volume of the Nuuchahnulth text collections. Although these typescripts do not have interlinear translations or free translations (with a few exceptions), they greatly facilitate efforts to complete the halted publication project. Another set of important materials is comprised of categorized ethnographical notes. These notes were extracted mostly from the fieldnotes, and they were typed and categorized into about 30 topics, probably in preparation for an ethnographical study of the Nuuchahnulth. In addition, there is a large collection of file slips (approximately 65,000) containing morphemes and word forms.

5.2.1.2. Value of the Fieldnotes The contribution that Sapir made through his numerous published papers and monographs to the study of Nuuchahnulth and beyond is both remarkable and important. Sapir’s analyses brought out his genius most dramatically and inspired many later studies. However, what he left in the form of fieldnotes is by no means any less significant than his published work. The field materials complement and enhance the published work in several ways. They contain a large amount of information that was left unpublished. They also retain the extra information that was filtered out in the process of preparing the data for publication. In particular, field materials are closest to the primary data in terms of the amount of phonetic details included in the representation of language forms. The phonetic details in textual data


97

are typically reduced in publications in favour of a phonemic representation. These details, however, are crucial for some studies, such as those on phonetics, language change, and dialectal differences. The major advantage (and disadvantage) of field materials as a source of information is that information contained in fieldnotes has not been processed and organized for a particular purpose. The materials, consequently, can be utilized for a wide range of purposes. Sapir’s field materials on Nuuchahnulth can make significant contributions in the following areas among others: Linguistic studies

• • • •

Detailed phonetic studies; Corpus-based descriptive and theoretical studies on grammatical phenomena; Studies on language change (through comparison with present-day data); Dialectal studies (through comparison with data from other dialects).

Ethnographical studies

•

Comprehensive ethnography.

Teaching materials

• •

Creation of language teaching materials; Creation of cultural readers.

5.2.2. Challenges in Utilizing Archived Materials It is relatively easy to see the ultimate value of field materials for language documentation and revitalization efforts. However, when one actually tries to use the materials, one quickly realizes that it is far from a straightforward process to locate relevant archived materials and harvest necessary information from them. Researchers face challenges both in locating the needed resources and in extracting information from the materials.

5.2.2.1. Issues in Locating Archived Materials Catalogues and finding aids that are well maintained and widely accessible are crucial for locating archived materials efficiently. Traditionally, even for preliminary surveys of contents, it was necessary for researchers to visit archives in person to consult card catalogues or finding aids. In recent years we have been seeing a notable improvement in this regard, as printed catalogues and finding aids are converted into electronic formats. Now several large-scale archives provide on-line access to their electronic catalogues and

98


finding aids, allowing focused searches in their collections.2 The on-line search services allow researchers to conduct preliminary surveys off-site and to make their visits to archives efficient and productive. Conversion of printed card catalogues and finding aids is definitely a general trend, and a rapidly increasing number of collections are becoming searchable over the Internet. However, as of this writing, fully searchable electronic catalogues and finding aids are still emerging services. Another factor that is hindering efficient searches for archived materials is the lack of a standard in archiving practices, including classification schemes and keywords. Different archives have developed idiosyncratic archiving practices that are totally incompatible with each other. Thus, researchers cannot transfer the knowledge and skills gained in the experience with one archive over to the work on another. This problem is also inhibiting development of a mechanism for allowing cross-archive global searches.

5.2.2.2. Issues in Extraction of Information from Archived Materials In addition to the challenges in searching and locating materials, researchers face challenges in finding and extracting information from the archived materials. Generally speaking, it is difficult to extract information that the researcher needs for a specific purpose from fieldnotes and other archived materials. Even with a fairly detailed index to the contents of the materials, it is still a very time-consuming task to work through materials that are not necessarily organized in a logical order. Moreover, information contained in the fieldnotes is not “tuned” for any specific uses; and, therefore, the user is required to do (possibly extensive) “preprocessing” before the information can be utilized for an actual project. This is a problem inherent in the nature of fieldnotes, and it is the flipside of the unique advantage of field materials (i.e. flexibility in use due to the unprocessed nature of the data in fieldnotes). This issue is particularly difficult to resolve. Accessibility to information can also be hindered at the most basic level due to indecipherable handwriting and to idiosyncratic transcription and abbreviation conventions. Some of these difficulties can be reduced dramatically by publishing the materials with prudent editing. In such projects it is important, although

2 Good examples include the National Anthropological Archives (Washington DC), the Australian Institute of Aboriginal and Torres Strait Islander Studies (Canberra), the Bancroft Library at University of California (Berkeley), and the Canadian Museum of Civilization (Ottawa).


99

difficult, to strike the right balance between the amount of processing done on the original in order to enhance usability of the materials and closeness to the original. Examples of excellent editing of field materials can be seen in the series The Collected Works of Edward Sapir (published by Mouton de Gruyter). Unfortunately, publication cannot be a widely available solution to the problem of inaccessibility of field materials. In order to assist potential users of archived materials, it is ultimately more practical either to have them form a partnership with those having the necessary skills and resources or to provide them with the necessary set of skills to work with the materials by themselves.

5.3. Maintaining a Field Philological Chain: From the Past to the Future Research on endangered languages is situated in social and academic dynamics that are very different from those surrounding research on major languages, and researchers of endangered languages are subject to special expectations and responsibilities arising from this unique situation. The most important facts are that there are only a very limited number of opportunities to collect data directly from people who lived using a traditional language, and that there are and will be a limited number of records available regarding these languages. In dealing with a major language with a healthy community of active speakers, data are typically collected for the exclusive use of the researcher. New data can be obtained easily, and it may even be desirable for researchers to collect data independently so that the description of the language as a whole will be based on a larger, more diverse database. When working with endangered languages, on the other hand, new language data are, or will soon be, a rare commodity. Since the sources of information on these languages are so limited, we cannot afford to overlook even information jotted down on a small corner of the beaten-up notebook of a fieldworker. Field researchers do not necessarily create their fieldnotes as public documents, but they should, nonetheless, be conscious of the fact that their notes will be one of the few and precious sources of information on the language in question. There is a strong possibility that knowledge about many endangered languages will be passed on to future generations only through these records and documents, and field researchers should take care to maximize the amount of

100


information that can be passed on successfully. Field documents provide the physical basis for a link of knowledge from the past to the future; but such documents themselves can never be an automatic guarantee of successful transmission of all the richness of a language that the documents attempt to capture. What we need are research strategies with a philological rigour that focus on the fact that there are a finite number of documents of the language available, and that it is becoming continuously more difficult to examine the information contained in the documents against the actual language facts. As Goddard (1973: 742) points out, “[i]t is necessary to make the fullest and most careful use of what there is, and to exercise the greatest diligence in preserving this corpus for the future in the most useful possible form.” Thus, a field philological approach consists of a battery of strategies to minimize loss and distortion of original knowledge and to maximize the chance of transmission of this knowledge in a form as close to the original as possible.

5.3.1. Field Philological Approach to the Past This section deals with a field philological approach to records from the past. Specifically, we discuss possible strategies to make maximum use of previously recorded data. (i) Careful treatment of the records. Written notes, as opposed to audio-visual recordings, are a highly selective and abstract form of recording. In addition, there are various conventions even now for writing down speech sounds. These two factors make the representational connection between the written symbols and the phonetic facts rather precarious. Written records are more like highly ambiguous clues than unambiguous, universal representations of the facts. Given the precarious nature of this connection, it is crucial not to contaminate the original records through careless editing or normalization. Even adjustments that seem innocent can add an extra layer of ambiguity and uncertainty into an interpretation of the records.3 Editorial adjustments can be justified when they bring out facts that are otherwise hidden, but the fact that the adjustments have been made should be clearly stated with indication of the original form. This point is crucial where it is difficult to verify data against primary facts, as is often the case with seriously endangered languages. (ii) Refining techniques of extracting information. In dealing with records, it is not possible to increase the amount of recovered information by improv-

3

Such danger is discussed in detail by Goddard 1973: 729 ff.


101

ing on the original. It is, however, possible to refine tools and techniques to deduce reliably more information than is directly represented in written records. In particular, computer-based processing of a large amount of data makes possible many new ways of extracting information from records. For example, the strategic use of a concordance can facilitate the identification of morphemes and their glosses, even when they are not directly provided in the original records.4 It can greatly extend usefulness of the original field material. (iii) Active, creative use of written documents in supplementing and enhancing the current database. It is possible to broaden the use of old records in creative ways. One example is to use old records to jog memories that have faded away into the periphery of a speaker’s consciousness. This type of prompting can greatly assist speakers who otherwise have trouble recalling the past, and it thereby increases the number of speakers who can contribute to the data collection. Such prompting is becoming increasingly important as the traditional languages fall into disuse. (iv) Training for utilization of field materials and information extraction. Another strategy to maximize use of field materials is to increase the number of people who can utilize the records. Typically, archived records are examined by only a limited number of professional researchers. Considering the value of information contained in archived records, it is obvious that such records are under-utilized. It is particularly regrettable that endangered language communities have not been able to make use of these archived records for their language documentation and revitalization efforts. Needless to say, this is not the fault of the communities. As discussed in § 5.2.2, access to archived materials is anything but straightforward in many ways, especially for non-specialists. As it is, there is not much that members of endangered language communities can do other than wait for the information buried in the archived records to trickle out through professional researchers’ publications. An ideal solution to this problem is to train members of endangered language communities to equip them with the necessary knowledge and skills so as to glean for themselves the information sought. Often the needs and priorities of language communities are different from those of researchers, and the exact kind of information that a community seeks may not

4 The Yahi translation project (directed by Leanne Hinton) at the University of California, Berkeley, provides a good example. The textual data of Yahi were collected by Edward Sapir in 1915 from the last and only-known speaker of the language, Ishi (Kroeber 1961). Due to Ishi’s minimal English proficiency and his sudden death, a large portion of the texts was left untranslated. However, in the Yahi translation project, translation and analysis of the texts are in progress through extensive use of a concordance built though computer-based text processing (Golla 2003).

102

t o s h i h i d e n a k ay a m a be obtainable through professional researchers’ publications when it is needed. Many endangered language communities are fighting against time, and they cannot wait for articles or books that some researcher may write some day. Thus, it is best if community members can study the field materials themselves. This is part of the effort to empower the communities of endangered languages with the academic and technical skills necessary for obtaining intellectual independence. Certainly document archives are not the most user-friendly resources, and it would undoubtedly be a challenge to set up such a training programme. The challenge, however, has been taken up by some. The Breath of Life workshop series organized by Leanne Hinton at the University of California, Berkeley, is an excellent example. This biennial workshop is offered by the Linguistics Department of the University of California Berkeley (since 1992) in partnership with a California native organization, The Advocates for Indigenous Language Survival (see § 23.4.5–6). “Approx. 40–50 native people representing 20 different languages come for an intensive week, to tour archives and get basic training in phonetics, grammatical analysis, and techniques for finding “useful language” in the materials here to use for language revitalization purposes” (cited from the website5 of the Linguistics Department of the University of California, Berkeley). This event in fact inspired another workshop of the same kind. Also named Breath of Life, the workshop was organized by Alice Taff of the University of Washington. Through a series of sessions stretched throughout a week, forty members of thirteen Native American communities of the Pacific Northwest Coast area were equipped with the basic knowledge and skills necessary to utilize the materials in the University of Washington archives for their indigenous language revitalization efforts (Broberg 2003). It is hoped that these workshops represent the beginning of a wave of similar programmes.

5.3.2. Extending a Field Philological Approach to the Future While engaging in descriptive research and other language revitalization activities, researchers naturally tend to focus on how they can utilize old records for their ongoing activities. However, as pointed out above, descriptive linguists conduct their own fieldwork, and therefore, they are also creators of source documents for philological research of the future. Adoption

5

http://linguistics.berkeley.edu/lingdept/Current/research/bol.html (accessed on 30 May 2005).


103

of a field philological approach by creators of documents amounts to assuming the responsibility of leaving field records for later generations in as useful a form as possible. It is not necessarily obvious or natural for field researchers to view their own fieldnotes as one-of-a-kind precious materials that need to be preserved for posterity. In an academic context the significance and value of research activities are measured and evaluated in terms of descriptive or theoretical synthesis made on the basis of data (i.e. grammars and papers). The data are, of course, important. However, they are important only as empirical support for syntheses, and they themselves cannot be the goal of research. Because of this academic paradigm, field researchers often feel that fieldnotes are nothing more than notes to themselves for writing grammars and articles, and they do not see further significant value in their fieldnotes, especially after they write a grammar. It is, therefore, important to reaffirm that the fieldnotes of researchers on endangered languages are part of a very limited number of irreplaceable records, and that these records are often as close as one can get to the primary facts about the languages in question. When researchers view their work and records within this wider context of documentation and revitalization of endangered languages, they should be able to understand that the value of their fieldnotes extends far beyond their private use. What is most important in the interest of future field philological work is for field researchers to deposit their fieldnotes ultimately in an appropriate repository for proper care and preservation. In order to make field records “philology-friendly”, field researchers should pay attention to two aspects of record making. One is record keeping of the data collection process. Ideally, records contained in fieldnotes are self-explanatory and self-contained, in the sense that they can be interpreted unambiguously and accurately without relying on knowledge other than that obtainable from the fieldnotes themselves. The meaning of a performance, linguistic or otherwise, is relative to space, time, and context (social, cultural, and interactional). Information regarding the surroundings of the recording, therefore, is crucial in accurate interpretation of the record. Such contextual information should be contained in the fieldnotes along with the data. Also important to be included in field records is an explanation about the spelling conventions and use of symbols. The other aspect of record making that requires careful consideration is the recording format and medium. With paper documents, quality of the paper and writing instruments can affect the longevity of the record significantly (Kenworthy et al. 1985). Especially, it is important to use acid-free

104


paper. The recording format is quickly becoming a major and complicated issue, as more and more data are acquired and stored in digital formats. Digital records revolutionize processing, storage, and utilization of information. They are undoubtedly making a significant positive impact on descriptive linguistic research and language revitalization efforts. At the same time, however, this digital revolution has introduced a major issue for record archiving with respect to longevity and reusability of the records.6 As pointed out by Bird and Simons (2003: 557), this is a problem of portability in a broad sense: across different software and hardware platforms; across different scholarly communities; across different purposes; and across time. This portability problem is succinctly summarized by Bird and Simons (ibid.): “Today’s linguists can access printed and hand-written documentation that is hundreds (sometimes thousands) of years old. However, much digital language documentation and description becomes inaccessible within a decade of its creation. The issue is acute for endangered languages. In the very generation when the rate of language death is at its peak, we have chosen to use moribund technologies, and to create endangered data. When the technologies die, unique heritage is either lost or encrypted.” As digital technologies constantly and rapidly evolve, it is not yet possible to determine the ultimate solution to this problem of portability of records. It is therefore crucial for researchers to stay focused on the portability of records and make informed decisions regarding the technologies to embrace. The Electronic Metastructure for Endangered Languages Data project (E-MELD), a five-year project funded by the National Science Foundation of the United States, represents an important initiative that specifically addresses this issue. The project aims to develop, support, and promote the best practices in language documentation and data archiving in order to make the data maximally durable, accessible, and useful.7

5.4. Conclusion None of the points or suggestions made in this chapter is necessarily new. However, they are yet to be integrated into a systematic approach that 6 The shift to digital technology has also affected audio-visual recording drastically in both positive and negative directions: digital recordings are less prone to degradation of quality over time; they can be reproduced, distributed, enhanced easily; but, on the other hand, the issue of format incompatibility is paramount, as the digital technology in the audio-visual domain is changing even faster than it is in the textual domain. 7 For further information, consult the project website (http://emeld.org/).


105

focuses on the limited availability of records. It is time to pay due attention to field documents as a major, and sometimes the sole, carrier of knowledge of endangered languages and cultures. In this chapter, I have suggested adoption of a field philological approach. In essence, this means to situate our research activities within the continuity of knowledge bridged by records from the past to the future. In order to maximize the chance of successful preservation and transmission of knowledge, we should utilize records from the past with the utmost care. At the same time, we should also strive to make every bit of our field research efforts count for as wide a range of uses as possible. After all, our field research represents a few precious opportunities to see a language in action, not only for field researchers ourselves, but also for any linguist, present or future. Each note a fieldworker takes in the field is a small window through which people can catch a glimpse of the language in action. In many ways, our research and the records resulting therefrom are not for ourselves alone.

References Austin, P. K. (2002). “Developing interactive knowledgebases for Australian Aboriginal languages – Malyangapa”, MS, University of Melbourne, 14 pp. Bird, S., and Simons, G. (2003). “Seven dimensions of portability for language documentation and description”, Language, 79/3: 557–82. Broberg, B. (2003). “Keeping their words”, Columns, Dec. Seattle: University of Washington Alumni Association (accessed on-line on 30 May 2005 at: www. washington.edu/alumni/columns/dec03/languages01.html). Goddard, I. (1973). “Philological approaches to the study of North American Indian languages: documents and documentation”, in T. Sebeok (ed.), Current Trends in Linguistics, x: Linguistics in North America. The Hague: Mouton, 727–45. Golla, V. (ed.) (1991). The Collected Works of Edward Sapir, Vi: American Indian Languages 2. Berlin: Mouton de Gruyter. —— (2003). “Ishi’s language”, in K. Kroeber and C. Kroeber (eds.), Ishi in Three Centuries. Lincoln: University of Nebraska Press, 208–25. Kenworthy, M. A., et al. (1985). Preserving Field Records: Archival Techniques for Archaeologists and Anthropologists. Philadelphia: The University Museum, University of Pennsylvania. Klokeid, T. J. (1971). “The West Coast (Nootka) text manuscripts: a report to the National Museum of Man”, MS, Massachusetts Institute of Technology, 19 pp. Kroeber, T. (1961). Ishi in Two Worlds: A Biography of the Last Wild Indian in North America. Berkeley and Los Angeles: University of California Press. Miyaoka, O. (1992). “Kiketsu Harinton o meguru kyoshotachi [Intellectual Giants surrounding Harrington]”, in C. Laird, Ikareru Kami to no Deai: Jonetsu no Gengogakusha Harinton no Shozo [Encounter with an Angry God: Recollections of my Life with John Peabody Harrington], trans. M. Ichinose. Tokyo: Sanseido, 261–99.

106


Nakayama, T. (2001). Nuuchahnulth (Nootka) Morphosyntax. University of California Publications in Linguistics 134. Berkeley and Los Angeles: University of California Press. —— (ed.) (2003). George Louie’s Nuu-chah-nulth (Ahousaht) Texts with Grammatical Analysis, ELPR A2–028. Sapir, E., and Swadesh, M. (1939). Nootka Texts: Tales and Ethnological Narratives with Grammatical Notes and Lexical Materials. Philadelphia: Linguistic Society of America. —— (1955). Native Account of Nootka Ethnography, International Journal of American Linguistics, 21/4, pt. 2. Watanabe, H. (1996). “Tekisuto no Shushu to Riyo [Collection and utilization of texts]”, in O. Miyaoka (ed.), Gengojinruigaku o Manabu Hito no Tame ni [An Introductory Reader on Linguistic Anthropology]. Kyoto: Sekaishisosha, 143–57.

6

Indigenous Voices and the Linguistics of Language Revitalization Marcellino Berardo and Akira Y. Yamamoto

6.1. Introduction 6.2. Voices of Native American teachers and parents 6.3. Community-wide teamwork in language revitalization 6.4. Issues in language revitalization 6.4.1. Ethical questions and issues of ownership: whose language is it? 6.4.2. Training members of the language team: introducing linguistics and language description 6.5. In closing: community voices

107 107 110 111 111 113 114

6.1. Introduction This chapter emphasizes the linguists’ task to provide diverse linguistic data for scientific purposes, and, equally important, to make available rich language materials to the endangered language communities from whom we obtain the very linguistic information. We also address issues that arise when working with endangered languages and in endangered language communities. When we work with endangered languages, we are committing ourselves not only to documenting and describing the languages, but also to meeting the linguistic, social, and educational needs of the communities. Thus, it is especially important for us to listen to the voices of indigenous peoples. For this reason, our discussion draws heavily from our experiences with endangered language communities in North America.

6.2. Voices of Native American Teachers and Parents A survey of American Indian opinion leaders by Indian Country Today on the state of Native languages found that, despite language programmes being offered by 71 per cent of respondents’ tribes, a substantial 64 per cent indicated they know only a few words of their Native language (Indian Country Today, 6 Dec. 2000). Under these conditions, teachers in endangered language communities are striving toward designing effective language programmes (including

108

m a r c e l l i n o b e r a r d o a n d a k i r a y. y a m a m o t o

different levels of immersion programmes), developing language curricula, and creating language-teaching materials. Here, the need is acute for linguists to go beyond documentation and description of endangered languages and to begin language revitalization with the communities. Faced with rapid language shift, Native peoples are reflecting on their attitude toward their languages and are reasserting the importance of their linguistic heritage. Many agree that they must stop blaming outsiders and that they themselves are responsible for the maintenance and perpetuation of their languages. The following points recently were made by Native American participants of the American Indian Language Development Institute (AILDI), the focus of which is to provide linguistic and educational training to the teachers of Native American languages. The AILDI participants believe that language is a part of being human. They say: We believe that our language is a gift from our Creator: • if we don’t use it, we are not fulfilling our responsibility, • if we don’t give life to it, we are neglecting to perform our duties. Native teachers know there has been enough talk about what to do and how to do it. Now, they declare it is time to start a programme to fulfil their “responsibility”. The goal of the teachers is high and they urge their community people to remember that, together, they need to plan for how, why, when, where, and for what purposes, they should revitalize their language. The teachers propose the following: Start small, • through children, get their parents involved; through parents, get their parents involved—start small and expand the circle, • start with your own family, get one more family to join you, and another; soon you will have your own speech community, but • don’t wait for someone else to start speaking the language. Native teachers have realized that it is not absolutely necessary to have a school to teach a language. Someone’s home can be an excellent place to begin language teaching. They say:

• • •

don’t wait until you have your own class, if you do speak your language, speak it to your children and your children’s friends, even if they say, “I don’t understand what you are saying,” don’t switch to English,

t h e l i n g u i s t i c s o f r e v i t a l i z at i o n

• •

109

if you don’t speak the language, ask someone to talk with you in the language, but if you don’t have a learner, speak to yourself, to your dog, your cat, and practice how effectively you can teach your language.

In many communities, the reality is that the only fluent speakers are the elderly. They will not wait for their community to get organized, obtain funding, recruit language workers, provide training in how to document, analyse, and describe the language, create curricula, and develop teaching materials. Native teachers realize that the situation is urgent and, against various odds, they try to obtain training in language teaching. They recognize that someone being a speaker of a language does not mean this individual is naturally a teacher. There must be some training in linguistics, language teaching, and education, which complement the ways teachers attempt to breathe life into their languages. The following statements are also from AILDI participants. When we teach a language in a situation of less than perfect immersion, • let’s get to know how our language is structured, • let’s get to know how language works, • let’s learn to plan how our language can be presented, • let’s use anything that will help us understand: gestures, things around us: pictures, books, computers, videos—anything, • let’s teach our language without using English or Spanish, and • let’s do things in our own language. Language teachers constantly engage in self-reflection on the manner in which they teach their languages. “Teasing”, for example, as a traditional way of teaching the young may not be the best method. They recommend that: When your learner tries to speak the language, • don’t criticize saying “you don’t speak our language correctly,” • don’t make fun of mistakes, • encourage the learner to be a risk-taker—look at our children and learn from them, and • be a model for good language use. They reassure: Learning is fun. Don’t kill it by making it “difficult” and “boring”, • use gestures, facial expressions, body movement, • use games, skits, songs, and dances, and • show you are enjoying your language.

110


They warn their people:

• • • • • • •

having our language in classrooms helps, but it is not enough, having community language programs helps, but it is not enough, putting our language in the computer helps, but it is not enough, videotaping elders speaking our language helps, but it is not enough, having a dictionary helps, but it is not enough, having a grammar helps, but it is not enough, and talking about our language helps, but it is not enough.

They urge their people:

• •

we must bring our language back to our homes and to our communities, and we must create an environment where our language is the language of thinking, creating, and communicating.

The teachers promise that when this revitalization is successful:

•

we will have two languages in which we think, create, and communicate. We want our Native language and English, not our Native language or English. Then, we will all be bilingual.

Listening to these voices, we elaborate on issues of language revitalization from the perspective of local communities with which we are familiar.

6.3. Community-wide Teamwork in Language Revitalization Revitalizing a language on a community-wide scale requires a teamwork approach. To facilitate the revitalization process, the team should include mutually supporting speakers, elders, community leaders, community language advocates, teachers, learners, and linguists. The team undertakes a task with at least three interrelated and ongoing parts (Berardo 2002). First, the language must be documented and described so the language team knows what teachers must include in order for the learners to have as much knowledge of the language as possible. The resulting grammar may be called a Language Revitalization Reference Grammar or LRRG. The LRRG is not intended to be used to teach learners to become fluent speakers. Instead, it catalogues and illustrates phonological and grammatical phenomena in the language. Second, the team turns to curriculum design, which may include a pedagogical grammar. The purpose of a pedagogical grammar is to teach learners to read, write, and speak the language rather than catalogue and illustrate the sound system and


111

grammar. Also considered in curriculum design are language materials for teaching all ages and levels, a dictionary and/or pictionary, texts, and teaching methods, which will efficiently and effectively produce new speakers of the language. Third, the language revitalization team needs to reintroduce the language into the homes, circles of friends, and community. This, of course, is the reason for the revitalization effort. Even when many community members are involved, there still are inherent difficulties with (re)introducing the language into the home and community. Today, interactions that family and community members have with one another often take place automatically and subconsciously in English. To reintroduce and reinforce the heritage language in the homes and community, the subconscious expectation of speaking English must be replaced by a conscious expectation of speaking the heritage language until there is an automatic, unspoken expectation that conversation with family and friends will be in the heritage language. Ultimately, families and community members will need to address this sociolinguistic process in order for the language to be revitalized.

6.4. Issues in Language Revitalization The desire to revitalize a heritage language must originate from within the local community and products of revitalization efforts must be relevant to that community. Previously, academic researchers wrote for other academic researchers and as a result local communities often did not benefit from such research because it was impenetrable, filled with jargon, and organized in ways that mostly benefited academia. Grammatical sketches, for example, were not guided by ethical issues of language ownership and by the need for an integrated team that includes speakers, non-speakers, teachers, political leaders, and other community members. The following sections briefly describe some of the ethical issues and linguistic training issues for those engaged in language revitalization. These issues are considered important to creating useful, relevant, and meaningful materials in an ethical way for the local community as well as academia.

6.4.1. Ethical Questions and Issues of Ownership: Whose Language is it? The language belongs to the people. It is a part of who they are. Thus, there arise some basic ethical questions of ownership of language that underlie

112


language revitalization work. Relevant questions include: (1) Who should and should not learn the language? (2) What needs to be documented and how? (3) Who should teach it? (4) How should teaching materials be prepared? (5) What should the materials consist of? (6) What variety or dialect of the language is to be documented and described? (7) What should the writing system represent: allophones, phonemes, syllables, lexemes, or a mixture of phonological and morphological elements? (8) Who makes the decision on these and related matters? There are no easy answers to these questions, in part, because they touch on highly complex issues of people’s attitudes toward language, identity, and the responsibility of language maintenance. (See Cameron et al. 1993 for further discussion.) Issues brought up by these questions require sensitivity, especially on the part of linguists coming from outside the community. The language does not belong to the linguist, but the methodology used by the linguistic researcher to discover how the language works is cast in linguistic terminology and shaped by linguistic and anthropological traditions of language description. Linguistic methodology, terminology, and tradition may give the impression that the linguist is imposing artificial categories on the language, which may seem unnatural, confusing, or inappropriate. In fact, the linguist does impose a linguistic approach to language description, which has been developed outside the values, attitudes, beliefs, assumptions, and traditions of the local culture. Because a linguistic approach to language description was developed specifically for an academic audience, some members of a language revitalization team may have concerns about its relevance, especially if local views on language are not consulted. For this reason, linguistics in a language revitalization framework focuses on learning the local perspectives on the language from the language community and educating the team about linguistics to make clear the intent of the linguist and underlying assumptions associated with a linguistic approach. It is especially important for the linguist to know the local community’s view on language ownership and how that perspective may impact academic research. Some communities may consider ownership of language as a ban on any public (non-tribal) discussion concerning the language, including presentations at professional conferences and academic publications as well as informal discussions on teaching methods or teaching materials the tribe uses. This understanding of ownership of language may seem to conflict with academic values of scholarship, including the idea that collaboration


113

with other academic colleagues, irrespective of their linguistic and ethnic heritage, is essential for high-quality, ongoing research in language description and teaching.

6.4.2. Training Members of the Language Team: Introducing Linguistics and Language Description As stated earlier, it is important for community members, including elders, speakers, political leaders, learners, language educators, and linguists, to participate as much as possible in the preparation of the Language Revitalization Reference Grammar. In this process, participants can learn to see how linguistic methods help describe language. Examples from the Algonquian language Potawatomi can serve to illustrate this point. In word-initial consonant clusters where the first consonant is as sonorous or more sonorous than the second, the perception of both sounds in the clusters may be difficult for untrained English speakers to hear. Consider the following Potawatomi words from Buszard-Welcher (2003: 15) and Potawatomilang.org (www.potawatomilang.org/Reference/Grammar/Morphology/Inflection/verbinflintro. html#anchor1450326, July 2005). mtek bkwezhgen gdewabdan

“tree” “bread” “You see it.”

English phonology does not allow word initial clusters such as mt-, bk-, and gd-. Untrained English speakers may hear these clusters as single consonants t-, k-, and d-, respectively. If members of the language team cannot hear all sounds in such clusters, they may reasonably object to documenting them in the description of the language. In such cases, a linguistic test that can tease off the initial consonants is needed. A linguistic test, here, refers to putting the relevant words in a linguistic context that will make all sounds in the cluster more salient. For the words above, the “determiner test” and “negation test” may re-syllabify the clusters, changing syllable structure to a more identifiable CVC syllable instead of the more complex CCVC or CCCVC syllables. Naming the tests makes for easy reference and ultimately the language team will be able to develop and apply appropriate tests to other phonological and grammatical constructions.

114


Determiner tests: The gi-test and i-test 1. gi-test

a. mtek b. gi mtegok c. [gim.tegok]

“tree” “these trees” “these trees”

(Word-final devoicing changes the final -k in (1a) to -g- in (1b) and (1c). Word-final devoicing is irrelevant to the point of this example.) 2. i-test

a. bkwezhken b. i bkwezhken c. [ib.kwezhken]

“bread” “the bread” “the bread”

When the determiners gi and i are introduced in (1c) and (2c), the initial consonant of the mt- and bk- cluster may syllabify with the preceding determiners resulting in the syllables gim and ib, respectively. These tests may make the initial sounds of such clusters more salient for the language team and therefore facilitate a more accurate description of the language. (Note that since only the initial syllable in (1c) and (2c) is relevant, the rest of the word is not separated into syllables.) Negation test: the “no/not-test” 1. jo-test a. gdewabdan “You see it.” b. jo gdewabdasin “You don’t see it.” c. [jog.dewabdasin] “You don’t see it.” The negative particle, jo, may also associate with the initial consonant of the following gd- string, resulting in the syllable jog that brings out the word initial g- in the gd- cluster. The jo-test is another test that can make the initial consonant in these clusters easier for the language team to hear and therefore contribute to the quality of the description of the language.

6.5. In Closing: Community Voices Throughout the twenty-six years of AILDI and ten years of collaborative projects with the Oklahoma Native Language Association, we have experimented with a variety of ways to work with language teachers, school administrators, community members, and political leaders. The key to success of


115

any language revitalization effort is for the people themselves to want their language back in their homes and communities. Here we see how some Native Americans regard their roles in revitalizing their own languages. The Native language teachers at AILDI say: A. As a speaker, 1. stimulate the desire to learn/teach by having fun with the language, 2. be creative in the use of the language—create stories, “turn-around” stories, songs, books, radio programs, TV programs, and others, and 3. keep using the language no matter what other people think and say. B. As a learner, 1. 2. 3. 4. 5. 6.

make a commitment to learn the language, continue to show enthusiasm, keep learning the language and/or keep talking, learn how to appreciate the language, show you’re having fun with the language, and try to use what is learned.

C. As a member of a family and a community, 1. start with your family—keep talking in the language even when they answer you in English—don’t switch to English, 2. keep nagging your family to speak in the language even when they say, “oh, not you again,” 3. keep talking in your language to your baby, to your mother, to your father, to your aunts and uncles, and 4. gradually enlarge your circle to your neighbours. D. As a teacher, 1. make a commitment to teach the language no matter what, 2. get training to learn what to teach and how to teach the language, and 3. work toward certification. E. As an advocate, 1. initiate formation of a language team or a support group that will promote the language, 2. be a part of the team, or 3. carry out a community survey and/or create strategies for raising community awareness, 4. plan community-wide language activities, and 5. just do it! F. As a program administrator, 1. administer a language survey:

116

m a r c e l l i n o b e r a r d o a n d a k i r a y. y a m a m o t o a. b. c. d. e.

to grasp the language situation, to raise awareness of the language, to recruit supporters, to recruit potential teachers, and to recruit language learners,

2. design a teacher recruiting program, 3. design teacher training programs, and 4. provide facilities. G. As a community leader, 1. make a commitment to funding, 2. recruit teachers and supporters, and 3. support all levels of language work—documentation, curriculum development, language materials development, use of technologies, etc. H. As members of a human community, they affirm: 1. our Creator has created for us the world through language. So, 2. if we don’t speak it, there is no world.

Linguists’ work with Native American languages has a long history. We are at a time when collaborative and cooperative efforts among Native language speakers, linguists, educators, and advocates can lead to successful Native language revitalization, maintenance, and fortification.

References Berardo, M. A. (2002). “An organizational model for community-wide language revitalization: a perspective from Native American languages”, Thule, 12/13: 21–47. Buszard-Welcher, L. A. (2003). “Constructions polysemy and mental spaces in Potawatomi discourse”, dissertation, University of California, Berkeley. Cameron, D., Frazer, E., Harvey, P., Rampton, B., and Richardson, K. (1993). “Ethics, advocacy and empowerment: issues of method in researching language”, Language and Communication, 13: 81–94. Dick, G. S., and McCarty, T. L. (1996). “Reclaiming Navajo: Language renewal in an American Indian community school”, in N. H. Hornberger (ed.), Indigenous Literacies in the Americas. Berlin: Mouton de Gruyter, 69–94. Dorian, N. (1998). “Western language ideologies and small-language prospects”, in L. A. G. and L. J. Whaley (eds.), Endangered Languages: Language Loss and Community Response. Cambridge: Cambridge University Press, 5–21. Grenoble, L. A., and Whaley, L. J. (eds.) (1998). Endangered Languages: Language Loss and Community Response. Cambridge: Cambridge University Press. Grinevald, C. (1998). “Language endangerment in South America: a programmatic approach”, in L. A. Grenoble and L. J. Whaley (eds.), Endangered Languages:


117

Language Loss and Community Response. Cambridge: Cambridge University Press, 124–59. Hale, K., Krauss, M., Watahomigie, L., Yamamoto, A., Craig, C., Jeanne, L. M., and England, N. (1992). “Endangered languages”, Language, 681: 1–42. Hinton, L. (1994). Flutes of Fire. Berkeley: Heyday Books. —— and Hale, K. (eds.) (2001). The Green Book of Language Revitalization in Practice. New York: Academic Press. —— with Vera M., and Steele, N. (2002). How to Keep your Language Alive: A Commonsense Approach to One-on-One Language Learning. Berkeley: Heyday Books. Indian Country Today (2000). “Ideal and reality”, Indian Country Today (6 Dec.). Linn, M., Berardo, M., and Yamamoto, A. Y. (1998). “Creating language teams in Oklahoma Native American communities: a process of team-formation”, International Journal of Sociology of Language, Special Issue on the Sociology of Indigenous Language Use and Change in the Americas. McCarty, T. L., and Zepeda, O. (eds.) (1998). Indigenous Language Use and Change in the Americas (International Journal of the Sociology of Language, 132). Berlin: Mouton de Gruyter. —— Watahomigie, L. J., and Yamamoto, A. Y. (eds.) (1999). “Reversing language shift in indigenous America: collaborations and views from the field”, Practicing Anthropology, 20/2: 2–47. Parks, D. R., Kushner, J., Hooper, W., Flavin, F., Bird, D. Y., and Ditmar, S. (1999). “Documenting and maintaining Native American languages for the 21st century: the Indiana University model”, in J. Reyhner, G. Cantoni, R. N. St. Clair, and E. P. Yazzie (eds.), Revitalizing Indigenous Languages. Flagstaff: Northern Arizona University, 59–83. Payne, T. E. (1997). Describing Morphosyntax: A Guide for Field Linguists. Cambridge: Cambridge University Press. Skutnabb-Kangas, T., and Phillipson, R. (eds.) (1995). Linguistic Human Rights: Overcoming Linguistic Discrimination. Berlin: Mouton de Gruyter. Yamamoto, A. Y. (1998). “Retrospect and prospect on new emerging language communities”, In Endangered Languages: What Role for the Specialist? Proceedings of the Second FEL Conference, University of Edinburgh, 25–7 Sept. Bath: Foundation for Endangered Languages, 113–20. —— (1999). “Introduction: reversing language shift in indigenous America: collaborations and views from the field”, Practicing Anthropology, 20/2: 2–4. Potawatomilang.org. www.potawatomilang.org/Reference/Grammar/Morphology/Inflection/verbinflintro. html#anchor1450326, June 2005.

7

Pidgins and Creoles in the Pacific Sabine Ehrhart and Peter Mühlhäusler

7.1. Introduction 7.2. Difficulties in compiling an inventory of existing pidgins and creoles 7.2.1. Historical circumstances of development 7.2.2. Substratum language(s) 7.2.3. Superstratum language(s) 7.2.4. Stages of development 7.2.5. Degree of vitality or the threat of extinction for the language 7.3. Pidgins in the context of language endangerment in the Pacific: social classification 7.3.1. Beach community languages 7.3.2. Administrative languages 7.3.3. Military pidgins 7.3.4. Shopkeeper-pidgins 7.3.5. Trade pidgins 7.3.6. Mixed types 7.3.7. Households and work situations 7.3.8. Boarding schools and missions 7.3.9. Varieties with lexifier languages other than English and French 7.3.9.1. Spanish 7.3.9.2. Pidgins based on local languages (especially inter-village trade languages) 7.3.10. Different stages of development 7.3.11. Vitality 7.4. Creoles in the context of language endangerment in the Pacific 7.4.1. Definition 7.4.2. Visualization of the dynamics 7.4.3. Examples of creoles in the Pacific context 7.4.3.1. Tok Pisin 7.4.3.2. Solomon Islands Pijin 7.4.3.3. Bislama in Vanuatu 7.4.3.4. Boarding schools and mission creoles 7.4.3.4.1. Tayo 7.4.3.4.2. Unserdeutsch 7.4.3.4.3. Australian Kriol 7.4.4. Torres Straits Broken 7.4.5. Cotabato Chavacao 7.4.6. Creoles of island utopias 7.4.6.1. Pitkern and Norf’k 7.4.6.2. Palmerston English 7.4.6.3. Ngatikese men’s language (Ngatik Pidgin English) 7.5. Conclusions

119 120 121 121 122 123 123 124 124 125 125 125 126 126 127 127 128 128 128 128 129 130 130 131 134 134 134 134 135 135 136 136 136 137 137 137 137 138 138

p i d g i n s a n d c r e o l e s i n t h e pac i f i c

119

7.1 Introduction During the twentienth century, there was a long tradition of neglecting the study of pidgins and creoles in linguistics and it is only in recent decades that publications in the theoretical domain have begun to fill the gap. Pidgins and creoles are still not highly regarded either by their speakers themselves or in publications on language rights. For example, Skutnabb-Kangas (2000) only mentions Nigerian Pidgin and refers to Hawaiian English as a regional accent. Pidgins are defined as structurally and functionally reduced second languages used for communication between interlocutors who have no language in common. As long as these contacts are not institutionalized, idiosyncratic interlanguages prevail. As the social context becomes better defined, social norms for lexicon and grammar develop. Where contacts

Figure 7.1. Changes in the ecology of Pacific pidgins and creoles

120

sabine ehrhart and peter mühlhäusler

become permanent, pidgins become creolized, i.e. they extend their functional range and acquire structural complexity. As in environmental biology, there is a danger of linguists focusing on charismatic “pure”, “authentic” species. Thus, the reasons for pidgins and creoles being under-documented include that they were associated with undesirable social processes such as slavery and disintegration of traditional societies. Generally, the speakers of pidgins and creoles have not had positive attitude towards their languages and, having internalized what Keesing (1990: 149 ff.) refers to as the “colonial ideology”, they often do not consider them as autonomous linguistic systems. Pidgins and creoles that emerged in colonial contexts have often been identified as the principal cause of disappearance of traditional vernacular languages, for instance by Tok Pisin researchers working in Papua New Guinea (Mosel 2002; Wurm 2001; Mühlhäusler 1993). The negative impact of these languages is only one part of the picture, however. As observed by Drechsel (1997), they can also fulfil an important role in maintaining diversity in complex language ecologies. Nowadays, with global languages making inroads into the former domains of pidgins and creoles, languages that could have been a threat to others are often threatened with extinction themselves. We shall focus on the pidgins and creoles of the Pacific Ocean and will not comment on sign languages such as “Hawaii Pidgin-Sign Language” (see Grimes 2001: ch. 5.9).

7. 2 Difficulties in Compiling an Inventory of Existing Pidgins and Creoles Knowledge of Pacific pidgins and creoles has increased rapidly since the publication of Hancock’s (1971) inventory of about twenty-five languages. More than ten times that number are listed in Wurm et al. (1996). It is difficult to compile an exhaustive inventory since the number of names available does not necessarily match the number of languages. Thus, some are almost invisible to their environment, partly because there are no proper names for them or because they are portrayed as inferior versions of other languages (e.g. broken English, français mal tourné, petit nègre). Ironically, the discovery of new pidgins and creoles has tended to coincide with their rapid decline. We shall not provide an exhaustive listing here but rather deal with the question of how different types of pidgins and creoles are differentially


121

susceptible to functional and structural attrition. In order to give a comprehensive classification of the existing linguistic contact varieties, several parameters need to be taken into account (Mühlhäusler 1997: ch. 3).

7.2.1. Historical Circumstances of Development Traditionally, pidgins and creoles are described according to successive stages of their development. Whether the endpoint of this developmental scale is reached depends on geographical, demographic, and sociological factors and to what degree learning and/or teaching is targeted.1 The distinction between société d’habitation (where white settlers outnumber introduced labourers) and société de plantation is central to the description of plantation creoles (Chaudenson 1992). Baker (2001: 44–5) distinguishes three types of creoles according to the linguistic attitudes and speakers’ objectives as language learners: homestead creoles,2 school creoles, and true creoles.

7.2.2. Substratum Language(s) Substratum languages are spoken by socially less powerful groups and superstratum languages by the dominant group. In the Pacific, the main substratum languages are the Oceanic (Austronesian) languages of the smaller islands and the numerous families of Papuan and Australian Aboriginal languages. Typically the many languages of the colonized are opposed to the single or few languages of the colonizers. The relative importance of substratum and superstratum languages in pidgins and creole development remains hotly debated. Two factors need to be considered: (a) the stage of language development at which influence from substrate or superstrate becomes important and (b) causes of similarities between a pidgin/creole and its substratum and superstratum languages. Longitudinal studies of the Melanesian pidgin Englishes appear to suggest that substratum influence was less in evidence in the early formative years than during their later expansion. Thus, in Tok Pisin’s direct predecessor, the Samoan Plantation Pidgin English spoken by Melanesian workers in Samoa before 1914, no distinction between inclusive and exclusive first person plural pronouns 1

See Mühlhäusler (1997: ii. 57) and Baker (2001: 42). “A common factor in the socio-historical circumstances in which these originated is that the first European male settlers had non-European consorts” (Baker 2001: 42). 2

122


is found nor are there dual pronouns. The inclusive–exclusive distinction was firmly established in New Guinea Tok Pisin from around 1910, though the full paradigm of dual pronouns took longer to emerge. Similarities between a pidgin/creole and substratum languages can have many sources: accident, inheritance from a common source, typological convergence, and others. As shown by Baker and Mühlhäusler (1996a, 1996b), many of the features of Melanesian Pidgin English labelled Melanesian “substratum” by Keesing (1988) were already found in New South Wales Pidgin English in 1820, i.e. fifty years before Melanesians acquired Pidgin English on the plantations of New South Wales and Queensland. Classification and naming by both creolists and general linguists tends to emphasize regional influence in the formation and development of a new language, as is evidenced by Hockett’s statement (1958: 424): “Note the pattern by which ‘Pidgin’ and ‘Creole’ are used in the designation of specific languages of the sort: X-Pidgin-Y or X-Creole-Y means a Pidgin or Creole based on Y as the dominant language which has supplied at least the bulk of the vocabulary, with X or the languages of the X-region, as most important second contributing factor.” However, pidgins and creoles have a very high mobility, “they can fly” (Mühlhäusler 1997: 15) from one place to the other.3

7.2.3. Superstratum Language(s) Generalizations about the role of superstratum languages in pidgin and creole development are difficult to make as such influences can vary greatly. Moreover, many past claims that a creole has directly developed from a European language have been ideological rather than factual. Traditionally, superstrate influence has been claimed for pidgins and creoles derived from European languages. This view does not fully satisfy as it does not consider the linguistic ecology as a whole (Mühlhäusler 1996d). There is another danger: as the majority of researchers stem from the linguistic areas of the main European languages, pidgins and creoles that are based on non-European elements tend to be overlooked. In order to construct family trees, the existence of one lexifier language per creole has been privileged by creolists and conventional classifications speak of English

3 So for instance, the contact languages from Pitcairn and Norfolk have an Atlantic influence due to the origin of one of the mutineers of the Bounty (St Kitts), and Tok Pisin of Papua New Guinea crystallized on the Polynesian island of Samoa among indentured Melanesian workers.


123

creoles, French creoles, etc. This does not account for creole development in complex multilingual societies. For instance, the many inter-village pidgins of Papua New Guinea (Mühlhäusler et al. 1996a) are lexically and structurally more mixed than pidgins “based” on a dominant European language. Relexification is a particularly important factor in pidgin/creole development and again illustrates the importance of political power. Relexification can occur when one dominant lexifier group is replaced by another one (e.g. in German New Guinea a pre-existing Pidgin English changed into a Pidgin German). Relexification appears to be a gradual process, but because of poor documentation, the transitional stages are rarely documented.

7.2.4. Stages of Development Not all contact-induced linguistic systems follow the life-cycle “jargon–pidgin– creole”;4 there are cases where creolization may have occurred without prior pidginization (Australian Koinés, Réunionnais) and others of jargons and pidgins that do not reach the stage of creoles (Fijian Pidgin). Moreover, a linguistic system which is used as a creole by one speaker may be a pidgin for another one.

7.2.5. Degree of Vitality or the Threat of Extinction for the Language Wherever linguistically mixed populations were brought together for joint tasks, pidgins emerge. The duration of their existence coincides with their usefulness. As long as German planters employed Melanesian workers in Samoa, Samoan Plantation Pidgin was a vigorous language. When in 1975 Mühlhäusler interviewed the last Melanesians in Samoa, the language was no longer actively used. The same scenario was encountered when Laycock interviewed New Guineans who had worked for the Japanese during the Second World War (Mühlhäusler and Trew 1996). The number of pidgins that must have disappeared is probably quite large and their existence can be verified only from fragmentary references in archival sources or remnants of language remembered by speakers once involved in enterprises such as phosphate mining on Clipperton and Matakea Island in French Polynesia or the pearl diving industry of Western Australia (Hosokawa 1987). One of the difficulties of identifying pidgins has been that they are not tied to well-defined user groups and territorial boundaries. Over 4

According to the life-cycle theory developed by Hall (1962).

124


a couple of decades the Aboriginal Pidgin English of New South Wales spread to Queensland (Mühlhäusler 1996a) where it was adopted by introduced workers from Melanesia and renamed Canefields English. This Canefields English was transported to the Torres Straits from the 1880s where it came into contact with Papuan Pidgin English, and eventually became creolized. To identify distinct varieties in such a situation is very difficult. Chinese Pidgin English (CPE) illustrates such problems. This label is conventionally applied to all varieties of Pidgin English spoken in China or by Chinese. However, the linguistic nature of CPE changed dramatically after the Opium War of 1842 without a change in substratum or superstratum languages (Baker and Mühlhäusler 1990), and the Pidgin English spoken by Chinese in places such as Nauru (Siegel 1988) or in Western Australia (Mühlhäusler and McGregor 1996) differs from China coast CPE. In short, the difficulties experienced in identifying, delineating, and naming pidgins and creoles in the Pacific are comparable to the problems universally experienced when trying to draw boundaries in complex multilingual ecologies. To give a definite number for endangered Pacific vernaculars is as impossible as giving a precise number for endangered pidgins and or creoles. It is essential, however, for language maintenance purposes, that the visibility of pidgins and creoles be enhanced by giving them names and stating boundaries, as support from governmental and other organizations depends on their being perceived as well-defined languages.

7.3. Pidgins in the Context of Language Endangerment in the Pacific: Social Classification Whereas structural properties have received most attention from creolists, they are poor indicators of the social life of these languages. Their viability is not determined by structural properties but by historical contingencies. The principal parameter for distinguishing different pidgin types is their domain of use.

7.3.1. Beach Community Languages These languages date back to the late eighteenth century when ships’ crews set up communities with indigenous women. The prototypical beach


125

community is that of Pitcairn Island where the Bounty mutineers and their Tahitian consorts created a new society (Mühlhäusler 2004). Other beach community languages are found on Ngatik (Tryon 2001), the Bonins (Long 1999), and Palmerston Atoll (Ehrhart 1996). In this group, very diverse degrees of vitality are represented, coinciding partly with the stability of the societies in which they were employed.

7.3.2. Administrative Languages Administration of highly multilingual areas such as Melanesia requires careful planning. Throughout the nineteenth century the colonizers generally took a dim view of pidgins derived from their own standard languages whilst not wishing to grant the colonized peoples access to the latter (Mühlhäusler 1996b, 1996c). As a consequence, they had to find other solutions to the problem of cross-lingual communication. The Dutch model (creating an elaborated variety of an existing local pidgin (Bazaar Malay)) was adopted in several places. Some of these planned languages were short-lived (e.g. simplified Tolai in German New Guinea) and others remained local (e.g. Dani Valley Police Talk in West Irian—Mühlhäusler et al. 1996b: 446). By contrast, Hiri Motu (Police Motu) gained considerable acceptance by the population of Papua and in 1975 was declared one of the official languages of Papua New Guinea. Despite its official recognition the number of its speakers in Papua is rapidly decreasing and Tok Pisin has become a strong competitor.

7.3.3. Military Pidgins The Japanese presence in Micronesia from 1914 till 1944 gave birth to a variety of military pidgins. Toki (1994) describes the remnants of Japanese amongst old people. See also Shibuya (2001) as to the remnants of Japanese in Palau. There is no future for these contact languages. The presence of large numbers of American military in the Second World War, the Korean War, and the Vietnam War brought into existence pidgins such as Bamboo English or Thai Pidgin English, none of which has been documented in detail and all of which have since disappeared from daily use.

7.3.4. Shopkeeper-pidgins Shopkeeper-pidgins are a common type of contact language but are poorly documented. Due to the high motivation to communicate they

126


tend to adapt rather quickly to the surrounding languages, but in some instances they become institutionalized, as in the case of Parau Tinito (Chinese speech) of Tahiti (Mühlhäusler et al. 1996b) or Nauruan Pidgin English.5 The contact languages of the Chinese shopkeepers in Queensland, Wallis and Futuna, or Samoa by contrast have been almost completely lost.

7.3.5. Trade Pidgins One can distinguish between pre-colonial and colonial trade pidgins. Commodities such as salt, food, or pottery were traded over long distances in pre-colonial societies. The linguistic means employed in such transactions included silent barter, bi-and multilingualism, or pidgins. The best-known Austronesian trade pidgins in New Guinea were Mekeo Pidgin (Jones 1985), the Siassi trading language (Harding 1967), and Trade Dobu. Remnants of Papuan-based trade pidgins in the Sepik river area (e.g. Pidgin Yimas) have been described by Foley (1988). Perhaps the best-known colonial trade pidgin is CPE which developed around Macau and Canton in the late eighteenth century. A Portuguese trade pidgin was once widely spoken all over South-East Asia, but when the Dutch, French, and English became the dominant trade nations, it rapidly declined. A small number of the inhabitants of Macau and Malacca used to speak it as a creole but changing lifestyles and resettlement led to its decline (Baxter 1990). Far less long-lived was Pidgin Hawaiian, which was used in the fur trade and in whaling (Drechsel and Kane 1982). Pidgin Hawaiian was supplanted by Pidgin English (Roberts 2000). Whaling was the rationale for English-based pidgins from the 1820s. However, as other commodities (sandalwood and trepang) became commercially more important, the domains of Pacific Pidgin English changed.

7.3.6. Mixed Types Frequently, contact languages are used in a range of domains and functions. This can strengthen the language under conditions of stable diglossia or polyglossia. Pre-colonial pidgins tended to be used in well-defined narrow domains. Colonial pidgins by contrast developed in conditions of rapid 5

Tryon and Charpentier (2004: 11).


127

social and economic change and their survival depended on constant adaptation to new functions. A pidgin such as Queensland Kanaka English, the Plantation Pidgin spoken by Melanesian indentured labourers (Dutton and Mühlhäusler 1984), was used as a working language on small farms where it was used primarily in vertical (master–servant) communication. It then became the (horizontal servant–servant) language of the large coastal sugar plantations of Queensland; its functions expanded when it became the principal language of proselytization by the Queensland Kanaka Mission. It fossilized, contracted, and eventually died out in Queensland following the discontinuation of the labour trade in 1904. The same pidgin expanded functionally and structurally in the Solomons and Vanuatu, whither it was transported by returning labourers.

7.3.7. Households and Work Situations Employment of domestics in European households was an important raison d’être for pidgins of the Pacific area. CPE was at its strongest when it was the language of regular intercommunication between Europeans and their Chinese servants. Similarly, the Pidgin German of New Guinea was almost exclusively a kitchen language on large mission stations which, like the Español de cocina of the Spanish Philippines, lost its usefulness when the colonizers departed. Two other languages that have originated in the domestic domain are Taj Boi, a contact language between French and Vietnamese, and Petjo, a Dutch–Malay creole in Batavia.6 The social context of their use (shared work in household and administrative areas) has disappeared since decolonization resulting in the extinction of the contact language.

7.3.8. Boarding Schools and Missions Boarding schools and missions were important catalysts of linguistic innovation in that they uprooted speakers from their traditional language ecologies and promoted new language contact systems. Depending on the number of students and the quality of teaching the learners could either get very close to the target language or end up speaking a pidgin (Mühlhäusler 1998).7

6 Petjo was still in use for some time after Indonesian independence; it has now been replaced by Bahasa Indonesia as a common language for the population (see Rheeden 1995). 7 For more on this point see § 7.4.

128


7.3.9. Varieties with Lexifier Languages other than English and French

7.3.9.1. Spanish The Spanish-based pidgins in the South Pacific have varied origins and some of the Philippines varieties may be relexifications of a pre-existing Portuguese creole. All of these creoles are declining, to various degrees. They are not protected through the education system.8 For more information on pidgins and creoles with Spanish as a lexifier see Lipski et al. (1996).

7.3.9.2. Pidgins Based on Local Languages (Especially Inter-village Trade Languages) The notion of pidgins being “based on local languages” refers to lexical affiliation, not to socio-historical events. Some such languages developed only in colonial times (e.g. Police Dani, or Pidgin Fijian) as new ways of life brought together people who previously did not need to intercommunicate. Some pidgins and creoles based on non-European languages were introduced to the Pacific from other parts of the world, e.g. pidgin varieties of Hindi spoken on the Fijian plantations and subsequent koineized Fiji Hindi (Moag 1977). The extent to which indigenous pidgins developed and were used and possibly creolized in the Pacific area remains a mystery as almost all of the known ones disappeared when colonization gained momentum.

7.3.10. Different Stages of Development It is important to note that the distinction between pidgins and creoles is not easy to make in the Pacific context as there are numerous situations where the creole of one section of the population is used as a pidgin by another group. This is the case for instance with Tok Pisin. The notion of a creole is becoming more generally applied to situations where a second language Pidgin becomes the primary language of a group of people, as with Solomon Pijin in Honiara, the capital of the Solomon Islands, which is taking over functions and domains from the numerous vernaculars spoken by urban immigrants. 8 In other regions of the world, there are thriving contact varieties between Spanish and local languages. Examples include the Amerindian–Spanish contact vernaculars of Middle and South America (Panama, Colombia) and the various Spanish–English contact languages in North America (Chicano, Tex Mex).


129

There is another terminological issue that needs to be mentioned. Whereas creolists have relatively well-defined labels for developmental stages of viable pidgins and creoles (e.g. jargon-stable pidgin—expanded pidgin—creole etc.), similarly clear characterizations for stages of decline are as yet not in existence.

7.3.11. Vitality The vitality of any language is determined by its ecological support system, which varies from language to language. Pidgins are not special (as was argued by Hall in his life-cycle paper) but merely a more visible case of dependency on external conditions. Neither the passing on of a pidgin between consecutive groups of L2 speakers nor the transmission of an L1 creole (or indeed the majority of the world’s languages) from one generation to the next can be taken for granted. In the case of Pacific pidgins, the main causes for extinction have been: (a) (b) (c) (d) (e) (f )

Loss of traditional economic behaviour and lifestyles; Introduction of asymmetrical power structures; Imposition of political boundaries; War; Major economic changes; Prohibition and persecution of languages.

For creoles, one of the principal reasons for language attrition has been the education system. Thus, the English-based creoles of Hawaii, Norfolk, the Solomon Islands, the Torres Straits, and to a lesser extent Bislama have been affected by systematic attempts by educators to eradicate any informal variety of English. Plantations count among the most important factors in the development of contact languages of the Pacific region. One can distinguish between plantations that relied on short-term indentured labourers (e.g. Samoa, Queensland, New Caledonia) and plantations that were staffed by permanent immigrant groups (e.g. Hawaii, Fiji). The Plantation Pidgin English of Samoa has been studied by Mühlhäusler (1979). The situation on the Fijian plantations was quite complex (Siegel 1992). At different times and at different stages in the social and economic development of Fiji, several contact varieties existed. Pidgin Fijian and, to a lesser extent, Pidgin Hindustani were the contact languages rather than Pidgin English. The most stable period lasted from 1865 to 1920, when it was the lingua franca of the European plantation owners and their workers. After

130


1920, the variation of the features was much more pronounced: “At the end of the plantation period, most Pacific Islanders were repatriated and those that remained married Fijians and spoke more standard Fijian. As more Fijians went to school, English was more commonly used for communication with Europeans. However, Pidginized Fijian continued to be used for communication mainly between Fijians and Indians and also Chinese” (Siegel 1992: 3). Up to 1904, Queensland Kanaka English was spoken on the coastal plantations of eastern Australia (Dutton and Mühlhäusler 1984). Its viability took a fatal blow when the government banned labour recruiting and insisted on the return of all Melanesians to their homelands. A small number of Melanesians were either allowed to stay or became illegal migrants. However, they soon shifted to standard English and little of the original Kanaka English is remembered by their descendants (Jourdan 1983). The Cattle Station Pidgin of Central Australia (Mühlhäusler 1997: 14) was the most important means of intercommunication between white cattle station workers and Aboriginal labourers. It also became the working language of black labourers from many different linguistic backgrounds. When in 1964 the Australian government introduced legislation that guaranteed equal pay to white and black workers in the cattle industry most black workers lost their jobs and a mode of living that had become institutionalized over about 100 years came to an abrupt end.

7.4. Creoles in the Context of Language Endangerment in the Pacific 7.4.1. Definition Creoles come into being when mixed pidgin-using populations become permanently relocated. In the Pacific this happened in Fiji and Hawaii, in a number of Spanish and Portuguese garrison settlements, various mission stations, and increasingly in the context of large-scale urbanization of indigenous people throughout the region. Most of the cases of contact can be described as gradual creolization in highly multilingual contexts. The Melanesian Pidgin that developed amongst the indentured Melanesian labourers of the Pacific plantation industry of Queensland, Samoa, New Caledonia, and Fiji belongs in this category: At the termination of their contracts these labourers returned to their highly


131

multilingual homelands where they spread knowledge of Melanesian Pidgin. The usefulness of this language over the years has led to considerable functional and structural expansion. Until the 1980s this expansion was not at the expense of local languages. More recently, an increasing number of urban dwellers have shifted to Melanesian Pidgin English as their first language and, even in remote areas, languages such as Tok Pisin have begun to replace local languages. During colonial days Melanesian pidgin Englishes enjoyed great prestige as they were associated with work opportunities, travel, and European possessions. Attempts by subsequent governments to eradicate them met with little success. Their decline coincides with the end of colonialism and the replacement of a strictly hierarchical society by an upwardly mobile one. In spite of the communicational usefulness of Melanesian pidgins, they lack the prestige of English and consequently have themselves begun to be eroded structurally.

7.4.2. Visualization of the Dynamics The combinations of different historical (§ 7.2) and social (§ 7.3) factors define the dynamics (life cycle) of a linguistic system; changes and variations lead either to the spread of a language or to its weakening and eventual extinction. Figure 7.2 shows the different domains of contact as presented in § 7.3. The main fields of contact are the home (with the contact languages of the beach communities), work (with the contact languages born for the needs of trade and the different types of agriculture) and law and order (with the contact languages used for military and for missionary reasons). There are also contact languages with a mixed-type origin: domestic pidgins between home and work, administrative languages between work and law and order, and the contact languages developing at school and especially at boarding schools can be situated between law and order and the home. Figure 7.3 shows the intensity of contact in the simplest case where speakers of two groups are entering into contact. The speakers of L1 have more power to influence the language choice. So the probability is high that their choice of a linguistic system will prevail. At the meeting point of the two coordinates, the communication is most dense and the mutual respect at its highest point, and bilingualism is possible. The intensity of communication is steadily decreasing according to the distance from the starting point and weaker contact varieties appear. It

132


Types of language contact: domains Domestic pidgins

Trade pidgins

Cattle and plantation contact languages

WORK

Administrative languages

HOME Beach communities

School creoles boarding schools

Military languages

LAW + ORDER Mission languages

Figure 7.2. Domains of contact

is important to distinguish between the targets of the two (groups of) participants in the interaction: does the speaker of L1 want to share his or her language completely with the exogenous group and does the L2 speaker

low contact

intensive contact

Access to L2 for L1 speaker

Resulting linguistic situation

Full

restrained

limited

small

very small

Bilingualism------Creoles--------Pidgins-------Jargons----Isolated linguistic elements

Figure 7.3. Intensity of contact


133

really wish to have a L1-like command of the foreign language? The target is of course closely related to the domain of exchange and the duration of the contact. Figure 7.4 shows the types of unbalanced language contact (i.e. without formation of bilingualism) according to the parameters of contact intensity and domain of contact. The different groups are: Type 1: The strongest case of contact beach communities where the language contact takes place at home between family members stemming from different language groups. This is the field where creoles can most easily arise. Type 2: This transitional type is situated at the connection between the different fields: home + work for domestic servants and home + law and order for the boarding schools, often under the authority of missionaries. Here pidgins are most likely to arise. Type 3: Plantation languages: they need a high amount of communication without specialization in the lexicon. Pidgins are likely to arise between the different levels of hierarchy and creoles in the domestic area. Type 4: This transitional type is situated at the connection between the different fields law and order + work, and concerns the mission languages and the languages used for administration. The type of language is most likely to be somewhere between a pidgin and a jargon.

L2 speaker’s target decreasing

Plantation contact languages

Domestic pidgins

Cattle contact languages

WORK

Trade pidgins

Administrative languages

Unstable Stable Beach communities

School creoles boarding schools

Military languages

LAW + ORDER Mission languages

Bilingualism--------------Creoles------Pidgins----------Jargons---------H I G H E R L O W E R

Figure 7.4. Domains & intensity of contact

L1 speaker’s target decreasing

134


Type 5: Unstable varieties: military languages and trade pidgins. They arise quickly and are linked with a vital need to survive, but they can disappear as quickly as they came. Type 6: The weakest point is extensive agriculture (cattle) that needs less communication than plantation work and very little exchange. Some words helped by gestures may be sufficient.

7.4.3. Examples of Creoles in the Pacific Context

7.4.3.1. Tok Pisin Tok Pisin has been recognized as a national language in Papua New Guinea and as yet remains strong, though its demise has been predicted many times. A recent study by Smith (2002) highlights a number of factors that could impact on its viability. They include (a) an increasing dependency on English for lexical and structural growth, (b) a lack of planning which prevents its use in more formal settings, and (c) growing regionalism which reduces its usefulness as a nationwide lingua franca.

7.4.3.2. Solomon Islands Pijin Solomon Islands Pijin is spoken by the majority of the population of the Solomon Islands. It is used by the media, the Christian missions, and more recently has also become one of the languages used by the government. English is still the language of education. The language is not immediately threatened by extinction, but Tryon and Charpentier (2004) mention the widening gap between urban and rural Pidgin which in the long term might lead to disintegration.

7.4.3.3. Bislama in Vanuatu Bislama in Vanuatu is one of the very few English-lexifier pidgins to have been accorded the status of national language. However, the reality is different from the administrative frame given by the official instructions. Thus, schools use English and French as teaching mediums and Bislama is excluded from these premises. Furthermore, there is no official spelling system either and Bislama is still mainly used only in oral communication. Because of its strategic position as the only common language of almost all the inhabitants of the country (there are about 120 different local languages and the population is split between English and French as access languages to either culture), Bislama is not threatened in the near future. Nevertheless, its long-term future may well not be so secure.


135

7.4.3.4. Boarding Schools and Mission Creoles The role of boarding schools and orphanages in the development of Pacific creoles remains poorly documented. Creolists often neglect the role of language creation within educational establishments. Hawaiian Pidgin, for instance, may well have derived its creolization from its use among children from numerous ethnic backgrounds in the school grounds (Reinecke 1969). Northern Australian Kriol became a first language of the children from a number of Aboriginal remnant groups who were brought together in the boarding school of Roper River mission (see Munro 2000). The continued existence and spread of this language is in part due to the fact that the mission generated a written standard variety and accepted Kriol as one of the classroom languages. Similarly, Torres Straits Broken became creolized at the Anglican St Paul’s boarding school on Moa Island. However, whilst Broken got recognition in the Torres Straits, the majority of islander families have moved to the Queensland mainland where the children are educated in monolingual English-medium schools. 7.4.3.4.1. tayo Tayo is a creole language spoken by people living in the Melanesian village of Saint-Louis in New Caledonia (Ehrhart 1993; Ehrhart and Corne 1996). It developed as a lingua franca at the Marist training centre for catechists after 1855. According to the documents, the very first Kanaks of the settlement spoke a variety close to French. With a bigger number of settlers and a change of objective—instead of flocking to the remote regions of New Caledonia, they decided to stay, to settle down, and to get married—several Melanesian languages were used for internal communication, especially the languages of Touho (with a large number of speakers) and those from the south (where Saint-Louis is situated). French in a more or less pidginized version was used in the contact with the Marists. In several families, the women were children of mixed couples (European-Melanesian) and were educated by the nuns also situated at Saint-Louis. Oral tradition in the tribe says that they taught French to their husbands and that the children of the first generation were still fluent in several Melanesian languages. The second generation of children born after 1920 is reputed to be the first creole-speaking one. Despite its location in the suburbs of Nouméa, Saint-Louis has maintained its creole language up to the present. The price for this is isolation and extreme violence in its surroundings. Since 1970, a village of Polynesian immigrants from Wallis and Futuna has sprung up in the close neighbourhood. This community developed very quickly from one person in 1970

136


until more than 1,000 people occupied the place. The relationship with the Melanesian village was always strained despite the common school for the children (where the Polynesians learned the creole as a L2). The tensions developed into a civil war. As a result, all Wallisians were relocated elsewhere. This experience was a traumatic one for the families, the last of whom left in mid 2003. It will be interesting to see whether they retain the creole language. 7.4.3.4.2. unserdeutsch Unserdeutsch is a German creole which developed at the Vunapope Boarding School for orphans in German New Guinea and became the first language of the mixed race community of Rabaul. It was still used daily as an in-group language until Papua New Guinea’s independence in 1975 when most community members dispersed to the Australian mainland (Volker 1982). It is no longer being passed on to the next generation. 7.4.3.4.3. australian kriol The large-scale dispersal and resettlement of Aboriginal people of Central Australia motivated government and missions to create new settlements where the remnant populations could be protected. One of these was the Roper River mission in the Northern Territory which commenced in 1908 and where by 1909 200 Aboriginals belonging to eight different language groups had gathered. According to Sandefur (1991) their children communicated in Pidgin English, which soon creolized and has since become the most important language for Aboriginal people in northern Australia. However, large speaker numbers, official support by the missions, and the education system have not removed the stigma which adheres to this language and its survival is not guaranteed.

7.4.4. Torres Straits Broken Pidgin English was introduced to the Torres Straits by Pacific Islanders working in the sea slug and pearling industries. One of the principal centres for the creolization in the 1890s of Torres Straits Broken was St Paul’s Anglican Mission. Over the last two decades the language has gained considerable status in education and public communication but the effects of this late recognition have been counterbalanced by two developments: the increasing outmigration to the Queensland mainland and globalization. Shnukal (1991: 193 ff.) discusses its growing role as a marker of identity but is uncertain whether the language will remain viable.


137

7.4.5. Cotabato Chavacao Spanish-based Cotabato Chavacao was once spoken around Cotabato City in the Philippines and it owes its existence to the Jesuit Order who around 1870 ransomed 100 Muslim children from the local slave market in order to educate and missionize them in their boarding school. A distinct Spanish-based creole developed there. After the American takeover of the colony in 1899 the children were taken to Zamboango where they mixed with the speakers of Zamboangeño and where their language lost its distinctiveness.

7.4.6. Creoles of Island Utopias

7.4.6.1. Pitkern and Norf ’k The mixed language that developed among the descendants of the Bounty mutineers has coexisted with standard English in a stable diglossia from its inception. Only about 20 per cent of the Pitcairners on Pitcairn (population 47) and Norfolk Island (population 900) still speak the language in its traditional form and it is not passed on to the next generation in many families, in spite of its recent revaluation as a language of identity. Recent external pressures on this community may reinforce the use of Pitkern as a language of identity or may lead to the abandonment of the island. The majority of Pitcairners live on Norfolk Island, where they constituted the majority until the 1950s. The island has become increasingly dominated by Australia. This factor together with the demands for fluency in English in tourism and employment constitutes a threat to its survival (Mühlhäusler 2002).

7.4.6.2. Palmerston English Palmerston English (PE) spoken on Palmerston Island (Cook Islands) combines elements from a variety of mid-nineteenth-century British English (with Gloucestershire and Yorkshire varieties being mentioned in the oral tradition) and Polynesian, especially in its variants of the islands of Manihiki and Rakahanga in the Northern Cook Islands. In the bilingual Marsters family, Polynesian speakers were compelled by the English-speaking family head to give up their mother tongue and to use English. The linguistic features of PE show evidence of the modification of English used as a L2 by L1 Polynesian speakers. Polynesian influence can be noticed in the phonology, whilst English and Maori are the main sources of the lexicon. The syntactic

138


structures of PE show dual constructions and inclusive/exclusive differentiation following the Polynesian system. Due to the high mobility of Palmerston Islanders in the Pacific region and the consequent dilution of their traditions, the number of fluent speakers of PE is rapidly decreasing, which makes the language a highly endangered one.9

7.4.6.3. Ngatikese Men’s Language (Ngatik Pidgin English) Ngatik Pidgin English is spoken on the island of Ngatik, the main island in the Sapwuahfik Atoll, situated about 150 km south-west of Ponape (Pohnpei) in the Caroline Islands of Micronesia (Tryon 2001: 345). The language developed after a violent fight between European sailors and beachcombers and local Micronesians. All male adults on the island of Ngatik were killed and a beachcomber community developed from the union of some of the Englishmen with the surviving local women. “Pidgin is commonly used by the male population of Ngatik today, especially when they are engaged in communal activities such as fishing, and house- and boat-building” (Tryon 2001: 346). Ngatikese Pidgin occupies an interesting position in the Pacific because of its similarities to early New South Wales Pidgin (Amery and Mühlhäusler 1996) and the fact that it is not intelligible to other Pacific pidgin speakers without knowledge of Ponapean because of its strong Micronesian influence. Ngatik Island has approximately 500 inhabitants. Its contact language is threatened, but not moribund. The special composition of this language with a stronger influence of the Pacific element in all linguistic domains than in the other pidgins of the region gives it a position apart which is of particular interest for language typology.

7.5. Conclusions The realization that pidgins lead a relatively precarious existence is not new. Indeed, it is one of the claims of Hall’s life-cycle theory (1962) that their continued use is dependent on their usefulness as second languages. We have schematized this life cycle in our diagrams. We have attempted to illustrate, 9 The core area of PE is situated on Palmerston Island with fifty-two fluent speakers, the entire population of the island. The diaspora is much larger, though still small in absolute terms, with about 400–500 more speakers living on other islands of the Cook group and a further c.400–500 persons in the Pacific region (New Zealand, Australia, French Polynesia) having a passive or an active knowledge of PE.


139

for instance, how languages of command and work cannot be maintained when the situation in which they emerged no longer exists. The more stable types of contact languages are those spoken in the home since they share a greater number of contact situations and a greater desire to communicate on both/all sides. In a small number of instances the addition of new domains and functions has increased the viability of a Pidgin. Tok Pisin, for instance has been greatly strengthened by its adoption as a language of missionization, administration, urban living and adult education in addition to its use in the plantation context. By contrast, Samoan Plantation Pidgin English, New Caledonian Pidgin English, and Fijian Pidgin English remained plantation languages and were eventually replaced by other media of intercommunication (cf. Mühlhäusler 1979). In Hall’s view creolization greatly enhances the viability of contact languages and turns them into “normal” languages, whose linguistic and social development does not exhibit special characteristics. From this point of view the disappearance of a small number of better-known creoles, such as Negerhollands of the former Danish Virgin Islands, is not a matter for concern. One of the claims of this chapter has been, however, that creoles are more vulnerable than most older and traditional vernaculars. Even subsequent to acquiring a community of native speakers many creoles continue to be transitional phenomena and the negative attitudes surrounding their predecessor pidgins are perpetuated. Put differently, creoles typically remain unvalued by both their speakers and the speakers of their lexifier languages. Negative associations are very difficult to overcome. Neither the interest of professional linguists nor the occasional official attempt to raise the status of pidgins/creoles is sufficient to arrest their decline. In our diagrams the likelihood of disappearance of a language is indicated by its distance from the crossover point. This reflects how a diminishing desire to communicate in a language among all parties weakens the language and eventually leads to its extinction. We do not wish to maintain that the well-being of a creole can be expressed in a simple formula. The number of factors that can trigger language decline is indefinitely large and we need to be aware of the rich range of components that constitute the post-colonial identity. As Mair (2002: 35) has put it: “The sociolinguistic evidence points to the existence of a continuum of gradual but by no means unsystematic transitions between more Creole-like and more English-like ways of speaking—with options for fine-tuning that go beyond anything that can be modelled against our current understanding of

140


bilingualism, code-switching or code-mixing.” The absence of accurate models is not an excuse to remain inactive, however. Creolists need to become more involved in the maintenance and preservation of the remaining pidgins and creoles of the Pacific area and address the question of how one can help create a greater appreciation of these languages among their speakers and among the dominant cultures of the region.

References Amery, R., and Mühlhäusler, P. (1996). “Pidgin English in New South Wales”, in Wurm et al. (eds.), 33–52. Baker, P. (2001). “No creolisation without prior pidginisation”, Te Reo, 44: 31–50. —— and Mühlhäusler, P. (1990). “From Business to Pidgin”, Journal of Asian Pacific Communication, 1: 87–115. —— —— (1996a). “The development and diffusion of pronouns in Pacific Pidgin English”, in Wurm et al. (eds.), 537–50. —— —— (1996b). “The origins and diffusion of Pidgin English in the Pacific”, in Wurm et al. (eds.), 551–94. Baxter, A. N. (1990). “Notes on the Creole Portuguese of Bidau, East Timor”, Journal of Pidgin and Creole Linguistics, 5: 1–38. Chaudenson, R. (1992). Des îles, des hommes, des langues. Paris: Harmattan. Drechsel, E. J. (1997). Mobilian Jargon: Linguistic and Sociohistorical Aspects of a Native American Pidgin. Oxford: Clarendon Press. Drechsel, E. J., and Kane, H. M. (1982). “Hawaiian loan words in two American pidgins”, International Journal of American Linguistics, 48/1: 460–7. Dutton, T. E., and Mühlhäusler, P. (1984). “Queensland Kanaka English”, English World-Wide, 4: 231–63. Ehrhart, S. (1993). Le Créole français de St-Louis (le tayo) en Nouvelle-Calédonie. Hamburg: Buske. —— (1996). “Palmerston English”, in Wurm et al. (eds.), 523–31. —— with Corne, C. (1996). “The Creole language Tayo and language contact in the ‘Far South’ region of New Caledonia”, in Wurm et al. (eds.), 265–70. Foley, W. (1988). “Language birth: the processes of pidginization and creolization”, in F. Newmeyer (ed.), Linguistics: The Cambridge Survey. Cambridge: Cambridge University Press, iv. 162–83. Grimes, B. F. (2001). “Global Language Viability”, in O. Sakiyama and F. Endo (eds.), Lectures on Endangered Languages, ii: From Kyote Conference 2000. ELPR C-002, 45–61 (www.sil.org/sociolx/ndg-lg-grimes_article.html). Hall, R. A., Jr. (1962). “The life cycle of pidgin languages”, Lingua, 11: 151–6. Hancock, I. (1971). “A survey of the pidgins and creoles of the world”, in D. Hymes, Pidginization and Creolization of Languages. Cambridge: Cambridge University Press, 509–23. Harding, T. G. (1967). Voyages of the Vitiaz Strait: A Study of a New Guinea Trade System, The American Ethnological Society, Monograph 44. Seattle: University of Washington Press.


141

Hockett, C. (1958). A Course in Modern Linguistics. New York: Macmillan. Hosokawa, K. (1987). “Malay Talk on Boat: an account of Broome Pearling Lugger Pidgin”, in D. C. Laycock and W. Winter (eds.), A World of Language. Canberra: Pacific Linguistics, 287–96. Jones, A. (1985). “Two Mekeo pidgins”, unpublished MS, Research School of Pacific Studies, Canberra: Australian National University. Jourdan, C. (1983). “Mort du Kanaka English à Mackay (Australie)”, Anthropologie et sociétés, 7/3: 77–9. Keesing, R. (1988). Melanesian Pidgin and the Oceanic Substrate. Stanford, Calif.: Stanford University Press. —— (1990). “Solomons Pijin: colonial ideologies”, in R. B. Baldauf and A. Luke, Language Planning and Education in Australasia and the South Pacific. Clevedon: Multilingual Matters, 149–65. Lipski, J. M., Mühlhäusler, P., and Duthin, F. (1996). “Spanish in the Pacific”, in Wurm et al. (eds.), 271–98. Long, D. (1999). “Evidence of an English contact language in the 19th century Bonin (Ogasawara) Islands”, English World-Wide, 20/2: 251–86. Mair, C. (2002). “Creolisms in an emerging standard: Written English in Jamaica”, English World-Wide, 23/1: 31–58. —— (2003). “The politics of English as a world language: new horizons in postcolonial cultural studies”, ASNEL Papers 7 Cross/Cultures 65, (Amsterdam), ix–xxi. Moag, R. F. (1977). Fiji Hindi. Canberra: Australian University Press. Mosel, U. (2002). “Description of the project ‘The Teop language of Bougainville, Papua New Guinea’”, financed by the VW foundation for endangered languages, Hanover: www.volkswagenstiftung.de/foerderung/foerderinitiativen/bewilligungen/ proldoku_d.ht Mühlhäusler, P. (1979). Growth and Structure of the Lexicon of New Guinea Pidgin. Canberra: Pacific Linguistics, C52. —— (1993). “The role of pidgin and creole languages in language progression and regression”, in K. Hyltenstam and Å.Viberg (eds.), Progression and Regression in Language. Cambridge: Cambridge University Press, 39–67. —— (1996a). “Pidgins and creoles of Queensland”, in Wurm et al. (eds.), 69–82. —— (1996b). “Dutch in the Pacific area”, in Wurm et al. (eds.), 339–44. —— (1996c). “German in the Pacific area”, in Wurm et al. (eds.), 345–52. —— (1996d). Language Ecology: Linguistic Imperialism and Language Change in the Pacific Region. London: Routledge. —— (1997). Pidgin and Creole Linguistics. London: University of Westminster Press. —— (1998). “Exploring the missionary position”, Journal of Pidgin and Creole Linguistics, 14/2: 339–46. —— (2002). “A language plan for Norfolk Island”, in D. Bradley and M. Bradley (eds.), Language Endangerment and Language Maintenance. London: RoutledgeCurzon, 167–82. —— (2004). “Norfolk Island-Pitcairn English (Pitkern-Norfolk): morphology and syntax”, in E. W. Schneider, K. Burridge, B. Kortmann, R. Mestrie, and C. Upton (eds.), A Handbook of Varieties of English, ii: Morphology and Syntax. Berlin: Mouton de Gruyter, 789–804. —— and McGregor, W. (1996). “Post-contact languages of Western Australia”, in Wurm et al. (eds.), 101–22.

142


Mühlhäusler, P., and Trew, R. (1996). “Japanese language in the Pacific”, in Wurm et al. (eds.), 373–400. —— Dutton, T., Hovdhaugen, E., Williams, J., and Wurm, S. A. (1996a). “Pre-colonial patterns of intercultural communication in the Pacific islands”, in Wurm, et al. (eds.), 401–38. —— Tryon, D. T., and Wurm, S. A. (1996b). “Post-contact pidgins, creoles and lingue franche, based on non-European and indigenous languages”, in Wurm et al. (eds.), 439–70. —— Lee-Smith, M. V., and Wurm, S. A. (1996c). “Preliminary thoughts on Chinese and Chinese contact languages in the Pacific area”, in Wurm et al. (eds.), 799–814. Munro, J. M. (2000). “Kriol on the move: a case of language spread and shift in northern Australia”, in Siegel 2000: 245–67. Pawley, A., Ross, M., and Tryon, D. (eds.) (2001). The Boy from Bundaberg: Studies in Melanesian Linguistics in Honour of Tom Dutton. Canberra: Pacific Linguistics. Reinecke, J. (1969). Language and Dialect in Hawaii. Honolulu: University of Hawaii Press. Rheeden, van H. (1995). Het Petjo van Batavia. Publicaties van het Instituut voor Algemene Taalwetenschap 64. Amsterdam: Universiteit van Amsterdam. Roberts, S. J. (2000). “Nativization and the genesis of Hawaiian Creole”, in J. McWhorther (ed.), Current Issues in Creole Studies. Amsterdam: Benjamins, 259–302. Sandefur, John (1991). “Kriol: the creation of a new language”, in S. Romaine (ed.), Language in Australia. Cambridge: Cambridge University Press, 195–203. Sapir, E. (1911). “Chinook”, in F. Boas, Handbook of American Indian Languages. Bureau of American Ethnology, Bulletin 40: 578–677. Shibuya, K. (2001). “The remnants of Japanese in Palau: introduction”, ELPR A4–001, 285–302. Shnukal, A. (1991). “Torres Strait Creole”, in S. Romaine (ed.), Language in Australia. Cambridge: Cambridge University Press, 181–94. Siegel, J. (1987). Language Contact in a Plantation Environment. Cambridge: Cambridge University Press. —— (1988). “Pidgin English in Nauru”, paper presented at the International Conference on Austronesian Linguistics, Auckland. —— (1992). “The transformation and spread of Pidgin Fijian”, Language Sciences, 14/3: 1–22. —— (ed.) (2000). Processes of Language Contact: Studies from Australia and the South Pacific. Quebec: Fides. Skutnabb-Kangas, T. (2000). Linguistic Genocide in Education: Or Worldwide Diversity and Human Rights? Mahwah, NJ: Lawrence Erlbaum Associates. Smith, G. P. (2002). Growing up with Tok Pisin. London: Battlebridge. Toki, S. (chief investigator) (1994). The Remnants of Japanese in Micronesia: Research on Japanese Language and Culture in Micronesia. Osaka: Osaka University, Faculty of Letters. Tryon, D. T. (2001). “Ngatikese men’s language”, in Pawley et al. 2001: 345–60. —— and Charpentier, J.-M. (2004). Pacific Pidgins and Creoles: Origins, Growth and Development. Berlin: Mouton de Gruyter. —— Mühlhäusler, P., and Baker, P. (1996). “English-derived contact languages in the Pacific in the 19th century (excluding Australia)”, in Wurm et al. (eds.), 471–96.


143

Volker, C. (1982). “An introduction to Rabaul Creole German (Unserdeutsch)”, MA dissertation, University of Queensland. Wurm, S. A. (2001). “Language endangerment in the insular Greater Pacific area, and the New Guinea area in particular”, in Pawley et al. (eds.), 383–97. —— Mühlhäusler, P., and Tryon, D. T. (eds.) (1996). Atlas of Languages of Intercultural Communication in the Pacific, Asia, and the Americas. Berlin: Mouton de Gruyter.

8

Linguistic Diversity in Decline: A Functional View Osahito Miyaoka

8.1. Prologue 8.2. Linguistic diversity and its backgrounds 8.3. Language in culture, culture in language 8.3.1. Adaptation strategies linked with cognition of environment 8.3.2. Primordial function of language 8.3.3. Culture as encapsulated in language 8.4. The less directly functional, the more possibility for diversity 8.4.1. Indirect function: language as referential system 8.4.2. Directly functional aspects of language 8.5. “Formhood” of words inevitable in human language: another background for linguistic diversity 8.5.1. A view on the “word” 8.5.2. Formhood responsible for diversity 8.6. An implication

144 145 147 148 149 150 150 151 152 154 154 157 158

8.1. Prologue Linguistic diversity is a theme with a long history of absorbing intellectual and academic enquiry, and interest in it has recently resurged partly through the growing realization of the value of biodiversity and partly through language endangerment as diversity becomes severely threatened. Concerning the current issue of linguistic diversity and its diminution, however, many different attitudes and views prevail—individual and collective—which largely depend upon how to see the functions of human language, besides historical and non-linguistic factors involved. The most commonly held view seems to be that a language is distinct from a culture and irrelevant to the process of thought. More specifically, most deeply seated seems to be the instrumental view that language is nothing

Special thanks should go to Sasha Aikhenvald, Mathias Brenzinger, Michael Fortescue, George W. Grace, Takashi Irimoto, Jeffrey Leer, Andrey Malchukov, Johanna Mattissen, Masayuki Onishi, Tamami Shimada Joseph Tomei, Honoré Watanabe, Takako Yamada, and Akira Yamamoto who read the earlier versions of this chapter and provided helpful suggestions. A recent version was read at the Workshop on Linguistic and Cultural Diversity—Dialogue of Cultures, an International Conference—held on 14 and 15 April 2005 in Reykjavik, in honour of Mme Vigdiís Finnbogatóttir.

linguistic diversity in decline

145

but a tool of communication/expression (and historical transmission) for human beings—a lever to get thoughts “across” or one way of “spitting it out”: so much so that it would be easy to harbour a pragmatic and negative view that multilingualism and linguistic diversity should be considered “poor economy” or costly nuisances (if not a Babel to be cursed) and, accordingly, should not be allowed to exist within a country, while monolingualism and diminution of linguistic diversity is in sober fact a welcome phenomenon, given in particular the globalized dominance of market economies coupled with build-up of information and communication infrastructures. For a better understanding of the theme, however, we would have to address it from a broader perspective.

8.2. Linguistic Diversity and its Backgrounds The diversity of language is very often taken as the best gauge of the diversity of (non-linguistic) culture, reasonably to some extent. With fundamental difference in function, however, language is interrelated with culture only partially (as on the lexical level sensitively tuned to the environment), implying its certain autonomy by which to operate at a more or less distant remove from culture and to be more stable (§ 8.3, § 8.4). Since the early days of American anthropology when “convergence” was a hot issue, it has been noticed that the fundamental forms of culture “lend themselves to a comparatively small number of types, which constantly recur as one passes from culture to culture”, partly owing to “the principle of limited possibilities in the development of culture” (Goldenweiser 1913: 273).1 In this trend, anthropological linguists, among others, have been aware that language, in terms of structure, shows a wider range of diversity than culture, although we have not identified the limits on the range of cross-linguistic diversity. In fact it is often the case that relatively greater diversity obtains in one and the same culture area, particularly among more or less isolated peoples in far-flung corners of the world not much spoiled and levelled out by encroaching civilization, who are found to speak

1 The principle, proposed in trying to formulate a theoretical justification of “convergence” of ethnological phenomena which was a contemporary theme that provoked lively discussion, is summarized as follows: “A limitation of possibilities checks variety. In relation to historical series of linked objects or features this means that wherever a wider range of variability in origins and developments coexists with a limitation of end results, there will be reduction in variability, decrease in dissimilarity, and increase in similarity or convergence” (Goldenweiser 1933: 45–6).

146

o s a h i t o m i ya o k a

languages with a surprising uniqueness and/or impressive complexity of grammatical—particularly morphological—structures; witness languages of the Pacific Rim, particularly in New Guinea (Ch. 13, this volume) and northwest North America (Chs. 21–23) as well as Amazonia, neighbouring the Pacific coast of South America (Ch. 10). As Edward Sapir (1884–1939) succinctly put it, “there seem to be no types of cultural patterns which vary more surprisingly and with a greater exuberance of detail than the morphologies of the known languages” (1933 [1951: 9]; italics mine). Syntax is by nature much less diverse. Linguistic diversity should be a combination of interwoven backgrounds or factors. The diversity of language as wider than culture is ascribed, first of all, to its being an “immensely ancient heritage” which Sapir is inclined to believe antedated any other cultural asset of man (1921: 22 [23], 1933 [1951]: 7). The diversity could also be traced to variously divergent processes (differentiation by inbuilt mechanism of linguistic change, creolization through language contacts, etc.) which reflect peoples’ sociocultural histories going back to antiquity. Recently resurged interest in diversity has yielded in-depth elucidations about lineage divergence “in space and time” by Nichols (1992) and Nettle (1999), for instance. My concern, however, is not in diachrony but in typological or structural diversity of human language, as contrasted with culture. The diversity goes side by side with the universality of principles, which, apparently a paradox, should nonetheless come from the nature of language itself or “something which is invariant in the human species” (Hale 1995: 137). Quite obviously, relative arbitrariness in linguistic categorization of experiences (§ 8.3.2) is admittedly responsible for the diversity, but a major attempt is made here to address the theme doubly in the same term called indirect functionality (§ 8.3; Miyaoka 1996b, 2002: 165–70) of (a) language (as contrasted with culture) in view of environmental adaptation (§ 8.4.1) and (b) the linguistic unit “word” (in parallel with the “syllable”) as the most essential “form” inevitable in human language at the first stage of what may be called as “articulation” (in the sense of “jointing together” but not “dividing”; § 8.5). Although the great variety among the world’s languages of packaging information into words is increasingly better known, it should remain an interesting question why linguistic morphology—both derivational and inflectional—shows dazzling diversity and daunting idiosyncrasy which obtains by far more widely than phonological, syntactic, or discourse/ pragmatic variations.


147

8.3. Language in Culture, Culture in Language In order to capture major functions of language (including communication/ expression as a “tool”), it would be important to see it in the broad context of “culture”, however notoriously ambiguous or elusive a term it is, namely in view of its fundamental constitution and overall workings. In order particularly to grasp that language and (non-linguistic) culture are not causally related with each other in any true sense and to see why linguistic (typological) diversity can be greater than cultural (§ 8.2), it would be crucial to consider the adaptive aspect of culture—adaptation to the environment which is the subjective and collective world of reality (realia) constructed by the speakers themselves.2 While development of culture and its diversity is reasonably related to the necessity of adapting to the ever-changing and unrelenting environment, linguistic diversity would be so only to a very limited extent. The ceaseless and dynamic interactions of a human group vis-à-vis the environment constitute the very process of ecological adaptation on the part of the people. Culture serves the serious (sometimes literally “life-and-death”) purpose of adapting to the environment, working as a buffer or cushion to alleviate the stress of multifarious pressures (or even strangulations) from it. So far as human culture serves this adaptive (versus exaptive) purpose, it is “functional in the immediate sense”,3 that is, directly functional. This functionality is generally most apparent in material culture (as of subsistence) but not irrelevant at all to the more symbolic aspects of (non-linguistic) culture (e.g. arts and religion) as well. By contrast, language as a symbolic system of reference stands aloof from non-linguistic reality, not acting upon nor being acted upon by it, aside from the portion of lexicon which is sensitively tuned to the environment. Language as such is functional only in the non-immediate sense, that is, (i) indirectly functional concerning environmental adaptation (§ 8.4.1). It indeed carries out the instrumental function, but it is still indirect (in contrast to such directly functional aspects as illustrated in § 8.4.2—with the caveat just below). This would imply that language is not such a tool in the common sense of the word (cf. Li 2005: 44)—be it an Eskimo toggle harpoon head (cf.n. 6) or a presentday computer—as can never cease vying for the improvement or evolution 2 The philosopher Kitaro Nishida (1870–1945) laconically remarked in a lecture that “a man makes up the environment and the environment makes up a man”, and “the environment is an environment only if it is subjective” (1940: 64, 65—translation mine). 3 “All activities may be thought of as either definitely functional in the immediate sense, or as symbolic, or as a blend of the two” (Sapir 1929 [1951: 163]; italics mine).

148


under unremitting pressure from the environment, which is something alien to language. This marked difference between language and culture in terms of functionality relative to the adaptation will turn out to be relevant to the present theme of linguistic diversity (§ 8.4.1). A caveat here, just so there is no confusion, is that language in actual use can also be (ii) directly functional in view of adaptation exactly like culture (§ 8.4.2), in a distinct sense from the language-internal functions as of syntax and phonology which have to serve directly for communication (not for environmental adaptation) (§ 8.4.2, end). An essential prerequisite to language being a symbolic system of reference, which eventually has the instrumental function, should be in tidying up of the environment. This would imply that the more fundamental or primordial function of language in relation to the environment would be (iii) cognitive, i.e. to construct reality for the (ethnolinguistic) group and the speaker (cf. Grace 1987), which is largely a matter of “linguistic classification” (Taylor 1995—§ 8.3.2).

8.3.1. Adaptation Strategies Linked with Cognition of Environment To grasp the fundamental role and unique status of language in human culture, a cognitive-ecological understanding of it would be an unavoidable necessity. The ever-changing environment, which a people constitutes a part of and is collectively in constant interaction with, should not remain a chaos (diffuse continuum) but be some cosmos (discretely classified world). Classification imposes an order on the world. It is not until the environment is classified and thereby “understood”4 that the people can stand ready to take any collective adaptation strategy—a manner in which it solves its impending problems with the environment. Given that adopting an appropriate adaptation strategy presupposes the group’s own way of cognition about the environs, human culture could be taken as a collectively inherited integration of: (i) cognition including here not only how the environment is classified by the group but how the classifications are manipulated in thinking and linguistic

4 The insight that to classify is to understand seems to be captured by the Japanese transitive/intransitive pair of verbs wakeru (“to divide, separate, classify”) vs. wakaru (“to divide [into] by itself, be understand[able]”), together with the nominal form wake “dividing, reason, cause” from the former, where the vowel alternation (a ~ e) is of a very common type in the morphophonology of the language (Miyaoka 2002: 160–61; cf. Li 2005: 44–6).


149

representation (by speech or writing), (ii) adaptation strategy, i.e. patterned activity in subsistence, society, education, arts, religion, and so on, with the time/period, space, and participant(s) agreed upon, and (iii) a plethora of associates connected with each category and strategy, including (a) manmade object(s), institution(s), and technology—i.e. material culture, (b) sentimental-emotional, metaphorical-symbolic, and magical-ceremonial values and implications, (c) belief and knowledge, myth and legend, etc. The intrinsic linkage of language with culture as such is implicit in Sapir’s laconic remark (1921: 233 [1951: 218]): “Culture may be defined as what a society does and thinks. Language is a particular how of thought.” Here we may see a harbinger of a rather general attitude fundamental to modern anthropological thinking about culture.

8.3.2. Primordial Function of Language Things come into being as the world is classified and named for fixity; witness the repeated occurrences of “divided (from)” and “called” in the beginning passages of Genesis 1 (cf. § 8.4.2). Language is destined to classify. It is inextricably interwoven with cognition—the way of viewing/classifying and organizing/representing the inner and outer world—playing a fundamental role therein. Classifications are conventionally fixed as linguistic symbols. Linguistically encoded classification (categorization) is multilayered. A simple design could be: (i) people classify the environment extensively and meticulously into discrete units (“conceptual categories”—Taylor 1995: 174), in paradigmatic/hierarchical/transverse ways, according to their particular modes of viewing the environment (primarily a matter of lexicon in which cognitive and linguistic anthropology—once under the name of folk taxonomy, ethno-X-ology, etc.—has had particular interest); (ii) upon the lexical categorization is laid a secondary categorization of (verbal) actions and states and of (nominal) referents, again possibly in different levels, either conformed to optionally or compulsorily (a part of grammar as it is, this is still a classification largely based upon (the aspects of the speakers’ view of) the non-linguistic world; and (iii) these categories are organized into the language structure by the purely linguistic (abstract or relational) categorizations and are manipulated and “articulated” (§ 8.5) into speech and thinking in accordance with certain channels or “grooves” based thereupon (Sapir 1921 passim).5

5 e.g. “Language and our thought-grooves are inextricably interwoven, are, in a sense, one and the same” (Sapir 1921: 232 [1949: 218–19]).

150


8.3.3. Culture as Encapsulated in Language Cognition as the primary function of language is closely integrated with adaptation strategies and their associates (§ 8.3.1). The whole of a culture as such—which has been cultivated by wisdom, ingenuity, and resourcefulness through the constant interactions between the group and its surrounding environment—permeates the language. In other words it is intrinsically encapsulated in language (cf. Dixon 1997: 135) from every aspect of it down to the finest details. Language, while being a subsystem of culture, would be the only one thereof that unites the whole of culture, as it were, into one capsule. It would be this “encapsulating” property of language, coupled with its “socializing” function below (§ 8.4.2), that leads to the “emblematic” function of a language—a function highly relevant to the current issue of language endangerment and revitalization. All this said about language vs. culture could be taken to mean that people do not live so much in a biological ecosystem per se as in a human-centred ecosystem, in which the former is largely integrated, and that it is the latter which is unique and meaningful to human beings (Miyaoka 2001: 3). Given that language makes up the very core of culture, one may even say that each group lives in a linguistic, i.e. linguistically constructed, ecosystem rather than a biological or even cultural one. So many languages, so many linguistic ecosystems—the integrated system of which may have something comparable to the ecosystem of linguistic diversity for which Krauss constructed the term “logosphere” (1996 and § 1.6.3, this volume).

8.4. The Less Directly Functional, the More Possibility for Diversity Any cultural activity as adaptation strategy (§ 8.3.1) which people carry out (either subsistence farming or a symbolic act in a ceremony) together with the associated objects (either a plough or an amulet) they use is supported by a certain purposiveness: It has a more or less defined purpose which, in the final analysis, resolves itself into ecological adaptation to the environment, however symbolic the device(s) employed may be. As such it can continue to exist only as long as it remains more or less functional and serves the purpose. Otherwise it becomes obsolete or replaced by another. (Non-linguistic) culture as a whole is obliged to be directly functional.


151

Because of the purpose, culture cannot but be largely conditioned or restricted by the environment (while retroactively working upon it). Given a serious (very often “life-and-death”) purpose to fulfil and the direct functionality relative to the environmental adaptation, culture elements may be apt to “converge” or conspire with each other and, cross-culturally, to be less diverse despite the appearances. As is most patently illustrated in the domain of material culture, they become largely limited in terms of “possibilities” or choices of shapes, patterns, materials, and technologies to be employed (§ 8.2; n. 1), affording little room for arbitrariness or irregularity and yielding greater uniformity and less diversity:6 The severer the restrictive pressure from the environment is, the heavier the purposiveness and the more direct the functionality is obliged to be. The more directly functional, the more limited possibility for diversity and the less stability (the greater transience).7

8.4.1. Indirect Function: Language as Referential System Language is not simply a self-contained tool for communication. Linguistic communication, while naturally attainable in close association with paraand non-linguistic factors, would presuppose collective sharing of not only the language code but also the whole culture (§ 8.3) by all the members and sectors of an group, even though in an approximate way with minor individual/regional/generational deviations and fluctuations. It is this collective sharing that enables speakers not only to properly communicate with each other but also to easily predict the addressee’s responses, often a difficult matter for people from other cultures, thereby facilitating group cooperation in environmental adaptation. Although language as a symbolic system of reference itself eventually serves for communication, it is, in contrast with (non-linguistic) culture, no more than indirectly functional in relation to the adaptation—apart from some portion of lexicon sensitive to the non-linguistic reality and from the directly functional aspects the language also has (§ 8.4.2). The less bound 6 Highly refined and sophisticated differences may result through the serious pursuit of efficiency under exigencies of environmental pressure, but only within a limited range of possibilities. Eskimo archaeology has disclosed the distribution of toggle harpoon heads in the Arctic with such exuberant differences in detail as to allow inferences about the chronology of Eskimo culture, but they cannot but be very much limited after all in the shape, structure, and material used. 7 Cf. “Purely functional explanations of language, if valid, would lead us to expect either a far greater uniformity in linguistic expression than we actually find, or should lead us to discover strict relations of a functional nature between a particular form of language and the culture of the people using it” (Sapir 1927 [1951: 550]). See § 8.5.2 also.

152


by functionality or purposiveness, the more possibility for diversity and the more stability. The diversity of language as wider than (non-linguistic) culture as a whole, however simplistic or superficial the contrast might appear to be, could be ascribed, for one thing, to this difference between the two in terms of functional directness. By the same token, the greater stability of language.

8.4.2. Directly Functional Aspects of Language All the above notwithstanding, language can also be (b) directly functional in relation to environmental adaptation, exactly as (non-linguistic) culture is. Language, though indirectly functional as communication, is actually overlaid with many different aspects of direct functionality. The earliest case with the greatest grandeur where language was directly functional is encountered again in Genesis: God called all things into being by his word, starting from “God said, ‘Let there be light,’ and there was light” (1: 3). All God had to do was to speak, and the word created day and night from the chaotic darkness of the void. Also, a magical spell (blessing or curse), once uttered, takes on a life of its own and influences the course of events as if the word and its signified thing (object or event) were an undivided whole (even with the arbitrariness of linguistic signs) and a shamanistic incantation achieves (or is at least believed to achieve) its desired effect by a kind of innate power. To that extent, these language uses are directly functional. The same functionality manifests itself exactly in the contexts relevant to language endangerment and revival/maintenance as well as in more common activities.8 One of the most relevant (and perhaps powerful) would be in its great force of “socialization” (Sapir 1933 [1951: 15–16]), which is considered responsible for emblematicity of language. Of whatever kind, size, and internal cohesion a group may be (whether it may be a family, a group of alumni, a generation, or a gang of organized criminals), the mere fact of sharing even a small bit of common vocabulary or stock phrases or of native naming of aboriginal land (as in Australia—§ 12.1.4) has the direct function of strengthening its members’ social bond and sense of identity, thereby effectively building up an emblematic barrier that outsiders could not be easily

8 Including the so-called “conative” and “poetic” functions, transient vogue words (just like dress fashion), words as “weapons”, mispronounced shibboleth, dictators’ rabble-rousing speeches, media commercials, judicial judgements and testimony, and other speech activities that bring about changes in the social environment. Again, “phatic communion” which is directly functional to the extent it serves as an interpersonal lubricant.


153

allowed to cross. It is true that there exist groups with a strong ethnic identity and pride but without having retained their heritage language and there exist communities where the knowledge of the heritage language is not regarded as essential for being a good member of the community (cf. Brenzinger et al. 1991: 35–6). But more often than not a language tends to be affiliated with a people (hence the very common identity of a language name with the people’s) and comes to be viewed inwardly as proof of its social or ethnic identity and pride but outwardly as a test for alienating or discriminating against a different people. If a people speaks one and the same language (beyond its barely retained lexical fragments), sharing the whole culture encapsulated therein, the social solidarity and identity, not surprisingly, may grow by far stronger than otherwise (unless, say, linguistic differences are too small for emblematicity as among closely related languages). Another instance of directly functional language in this context would be the forced name change very commonly executed upon native peoples. As an active attempt to directly affect the inter-ethnic environment,9 it was an important pillar of assimilation policy as toward Taiwanese and Koreans besides Hokkaido Ainu during Japan’s colonial rule (§ 18.1), with the aftermath lingering obstinately up to the present. Functionality—both indirect and direct—of language so far has been considered in relation to environmental adaptation, simply in order to compare it with (non-linguistic) culture (§§ 8.4.1, 8.4.2). Just so there is no confusion or mistake, there is another sense, of course, in which language is directly functional (the aforementioned caveat, § 8.3). Apart from functionality in relation to adaptation, language does have the direct purpose of communication to serve. Accordingly, many portions of linguistic structure cannot but be motivated and bound by the purpose—a great variety of languageinternal functions (thus functional load, unity, conspiracy, selection, and so on) in phonology and syntax and in such linguistic units that require functional definition (phonemes and morphemes), in marked contrast with morphology which has much less room for function and with words which could not be primarily defined other than as “forms”. Thus we have to move to the special status of words—in light of what I would venture to call “bilateral articulation”—among basic linguistic units, given that they would be responsible partly for the linguistic diversity as well as indirect functionality of language itself in view of environmental adaptation (§ 8.4.1).

9 Note by contrast that observance of taboo for certain words or names—either of a king, an animal, or a recently deceased—represents a passive stance of avoiding adverse effects from the environment.

154


8.5. “Formhood” of Words Inevitable in Human Language: Another Background for Linguistic Diversity There is widespread disregard for or suspicion of the word among linguists as it is too ambiguous a term for easy use in a technical literature or it may simply be language specific and not be a primitive construct. The tendency has strengthened as syntax became the mainstream of linguistic studies. It is encouraging to see, however, that many linguists are positive about wordhood: among others, Dixon and Aikhenvald (2002). The insight with which the late Rokuro Kono (1912–98) prefaced the monumental Encyclopaedia of Linguistics, that “the genuine object of linguistics is a word” (Kamei et al. 1988–96: i, p. ii; translation mine), should be duly appreciated. No one could fail to see my profound indebtedness here to Sapir (1921: 33, etc.), and beneath his passages we could not fail to hear the echoing words of Benedetto Croce (1866–1952), who claimed that linguistics is “nothing other than the Aesthetic” (aesthetics as the investigation of a fundamental capacity of human beings) and that “Aesthetic activity is … a matter of giving form, and nothing other than a matter of giving form” (1902 [1992: 156, 17]).

8.5.1. A View on the “Word” In my naive, possibly linguo-centric, perception (cf. Miyaoka 2000b; 2002; in preparation): (a) A word (語 go) would be an inevitable fundamental entity in human language in either an isolating language like Classical Chinese or a polysynthetic10 one like Eskimo (as illustrated in n. 15)—presumably a consequence of the very use of speech sounds (through air-stream mechanism) and brains, respectively having physiological and biological limitations and constraints (in articulatory and auditory apparatuses) and neuro-psychological and cognitive ones (in encoding and decoding, i.e. information processing). 10 Despite the renewed lively discussion of polysynthesis, linguists have not reached consensus on what type of languages it refers to (e.g. Fortescue 1994: 2601; Evans and Sasse 2002: 1–6; Mattissen 2003: 202). My stance on this is that it should not be sought so much in the morphological strategies employed (e.g. noun incorporations) or the kinds of semantic categories of the content (e.g. instrumentality, location) as in formal criteria—most notably of the Du Ponceau–Sapir–Greenberg tradition—closely linked to high morphological productivity and flexibility. See n. 17 also for P. S. Du Ponceau (1760–1844; American Philosophical Society president) who used the term polysynthesis most probably in 1816.


155

(b) In order to be free from oxymoronic or circular definition, a word could be defined as the primary, i.e. minimal, realized “form” of articulation (“jointing together”—結節 kes-setsu “connect-joint”; see below) at the content plane11 (with morphemes as the ultimate entities) which satisfyingly follows the morphological (grammatical) pattern of the language, but not in terms of the (semantic and functional) content packed therein. As a linguistic sign with two terms, sign-content and sign-expression, the morphological (grammatical) articulation is paired with the corresponding, i.e. phonological, articulation at the expression plane (with phonemes as the ultimate entities), again according to the satisfying pattern, thereby fortifying the formhood of the morphological articulation—hence (for some) a “phonological word” vs. a “morphological (grammatical) word”. The primary (minimal) “form” of articulation at the latter plane is a syllable (音節 on-setsu “sound-joint”)—see n. 11. (A single entity can be articulated by itself—a word may be mono- or polymorphemic, just as a syllable may be mono- or polyphonemic.) A word as such has necessarily both phonological and morphological (grammatical) relevance. This “form-assuming” process in the two planes may be called “bilateral articulation” which starts from the two formal units, as contrasted with the two functional units—a morpheme and a phoneme—abstracted by double articulation analysis.12 Double articulation and bilateral articulation are not simply the other sides of the same coin, despite the bivalency of the term articulation—“dividing” and “jointing together”. Seeing speech merely as the other side (i.e. spreading out) of the principle of double articulation might not sufficiently capture the status of a word as the essential feature of human language: single-track top-down analysis would blur the word. (c) Bilateral articulation proceeds simultaneously by jointing together respective entities in each plane according to the patterns of the language. Morphemes are hierarchically articulated (realized) into words, phrases,

11 A word is not taken as intermediate in status between the morpheme and the phrase. The “plane” here is in a bilateral implication as employed by Hjelmslev (1953), to whom “l’unité minimale essentielle est la syllabe” and the corresponding “syllabes de contenu” in the content plane are minimal “syntagmes” which practically coincide often with words (Hjelmslev 1966: 145–6)—a logical consequence from the glossematic isomorphism of expression and content. 12 Admittedly the very source for the economy and productivity of human language is “double articulation” without which language would be “a tool unusable for its purpose” (Martinet 1949; cf. Hjelmslev 1953: 29). Contrasted with the lucidity in defining this feature, Martinet does not seem to have assigned a clear place to the linguistic unit of word, though repeatedly discussing the disparity between a monème as the minimal sign and a “mot” (put in inverted commas) (1960: 112–13, 1965a, 1965b).

156

o s a h i t o m i ya o k a clauses, and so on,13 while phonemes are articulated into syllables and feet as they become simultaneously overlaid by a number of prosodic articulations (groupings characterized by pitch, stress, intonation, harmony, rhyming, aspiration/nasalization/glottalization, sandhi phenomena, etc.) as well as potential pause/insertability which is probably (yet controversial) the most essential feature for defining a linguistic articulation as a form.14

(d) The two terms of a linguistic sign, which necessarily presuppose each other, are symmetrically paired, but terms within one plane bear multifarious relations among themselves, independently of any within the other (cf. Hjelmslev [1953] and Saussurian “valeurs”). Owing to these plane-internal mutual relations of the terms as well as the pre-linguistic limiting conditions and constraints (above) in each articulation, the corresponding articulations on each plane do not necessarily go on in perfect correspondence. Hence, abundance of mismatches between the two articulations on different planes—a natural consequence of bilateral articulation as part and parcel of human speech. (e) For a linguistic “form” in itself, the content is secondary (cf. Croce 1902 [1992:17, 20]), just as an isosceles triangle, for instance, is taken as such so long as it has certain features (two of its sides being equal), however big a space (i.e. content) it may enclose. This means that a word may potentially be very diverse content-wise and that, without always being a static slot-fixed or ready-made unit established prior to a speaker and stored as a single concept, it can be a flexible process by which any speaker produces a new word as a matter of course (rather than nonce or ad hoc), though necessarily within certain limits.15 Thus a word can functionally be equivalent even to a complex sentence in other languages, with a variety of concrete, abstract, and grammatical concepts packed in, as typical in polysynthetic languages, while it can diametrically be an expletive devoid of content as in isolating languages. The nature of a word itself 13 Words include clitics. A clitic, which is inflectable in some languages, is taken here as a “bound” word (vs. “free” word) as defined by Hattori (1950), although I am fully aware that demarcation between a clitic and an affix (his “bound” form) is far from rigid (cf. Dixon and Aikhenvald 2002: 25–7). A “bound-word (clitic or non-clitic) phrase” is conceived between a word and a (syntactic) phrase. 14 Words, clitic phrases, and non-clitic (bound-word) phrases in Central Alaskan Yupik Eskimo (CAY), for instance, generally match with distinctive phonological units (with little asymmetry) which fortify the formhood of the articulations (Miyaoka 1971, 1996a, 2000). 15 In CAY, a “non-slot” type of polysynthetic language, we may readily encounter a very “heavy” word, e.g. angya-cuara-li-yu-llru-nrit-qapig-caaq-sugnarq-a-a-nga (boat-small-make- DESIR-NEGINTENS-but.actually-INFER-INDIC-he-me) “I thought he did not want to make me a small boat at all (but he actually made one)”, which is, clearly enough to speakers, a single word (consisting of a single stem for “kayak” followed by more than ten suffixes). Japanese is also a helpful language in getting a feel for articulation of the “non-slot” type, since, though not so polysynthetic as CAY, it is endowed with very productive “verbal complexes” (用言複合体 yougen-fukugou-tai; distinct from compound verbs)] which are actually non-enclitic phrases with verbal inflection (Miyaoka 2002: 102–18).


157

definable as nothing but a form should be responsible for its diversity, cross-linguistic as well as language internal. (f) As far as it is a mere form, a word (and a phrase) by itself is intrinsically liable to change, dilute, or reduce (close) to zero in content as a whole or in its constituent, with the original content being opacified or lost—a phenomenon which would be expected to hold of any human language. Hence expletives as well as (more or less opaque) lexicalizations at the word (and the bound-word phrase) level.16 (g) Intuitive knowledge of phonological and morphological properties (together with the patterns followed) backs up the formal unity, that is, the “form feeling” or psychological reality of an articulation. Thus the form feeling would be the very basis for the two kinds of writing systems, syllabic and lexigraphic (“word-writing”),17 corresponding to the two minimal forms in bilateral articulation which are intuitively the most salient of all the linguistic categories (cf. Taylor 1995: 176). Lexigraphism is the fundamental function of any kinds of writings, including alphabets (Kono 1977 [1994]).18

8.5.2. Formhood Responsible for Diversity After all is said, it would be this bilateral articulation of a word and a syllable as the minimal “form” which renders human language unique among any conceivable structures characterized by double articulation. “Forms” should be an absolute necessity for human speech controlled by brains and by speech organs. Back to the main theme now, it should be important to note that a word, being merely a “form”, is not so directly constrained by function—has little direct functionality—and that there is, again, an inverse relationship between functionality and diversity. It would be this “formhood” itself of a word that is partly responsible for the great diversity of linguistic morphology which can be full of sometimes daunting idiosyncrasy, arbitrariness, irregularity (or even “randomness” of association), and also for the limited possibility 16 It would be natural that, given the configurational structure and the very limited productivity at the word level, isolating languages like Chinese may tend to have expletives just for the sake of filling a syntactic slot and also to have the locus of lexical creativity at the bound-word phrase level which produces new lexical units in an innovative and extensive way. 17 The “lexigraphic” or “logographic”—instead of ideographic—nature of Chinese characters was first noticed again by the Americanist Du Ponceau (1838: pp. xxii–xxiv, etc. and 110; cf. n. 10). 18 As is most evident in English spellings—incoherency itself as a phonographic system. It is also to be noted that the Japanese syllabic writing of the running type (平仮名 hiragana-writing) makes full use of the characteristic of brush calligraphy in that its successional style (連綿体 renmen-style) has the function of representing words or enclitic phrases in one succession (Miyaoka 2002: 33).

158


for universals therein: Mutatis mutandis (cf. n. 7), if functional explanations of morphology (and definitions of words) are valid, we would expect such greater uniformity, regularity (even pseudo-universality), and less stability as found in those aspects of syntax in particular that have to be directly functional as regards communication (§ 8.4.2, end). To conclude, sources for linguistic diversity could be sought in the largely indirect functionality of language in relation to environmental adaptation (§ 8.4) and of words as “forms” inevitable in human language which makes it unique among all conceivable symbolic systems based on double articulation (§ 8.5), in addition to language as an “immensely ancient heritage” with variously divergent historical processes (§ 8.3) and the largely arbitrary nature of the three-layered linguistic classification (§ 8.3.2).

8.6. An Implication Admittedly the effect of diminution of linguistic diversity, now steadily in progress, is not so easy to decipher and the whole magnitude is difficult to gauge. Various reasons, however, why we now should care about diversity have been cogently argued by Crystal (2000: 32–6) and Krauss (2001: 28–32; § 1.6, this volume) in particular. As argued, each and every language has the whole content of a people’s unique culture—uniqueness in understanding and adapting to the environment—encapsulated in its entirety and in its finest details (§ 8.3.3). Extinction (or extermination as it often is) of a language as such, which leads not only to total collapse of the unique culture but also to loss of the people’s soul (pride and identity), could be likened to aphasia, in which the patient may have lost the language so inextricably linked to cognition of the outside world. Severe loss of language caused by brain injury works such profound damage that it looks as if the patient has had his or her soul taken away or lost their entire personality. By the same token, for a people to lose the language or to have it replaced by another may be to lose something deeply rooted in their very existence—a loss so profound that it would not be easy to see the significance. Loss of a single language from the earth would not affect the people alone who speak the language, however. For science, it would mean a crucial narrowing of our understanding of linguistic diversity and width (cf. “confirmatory function of linguistic diversity”, Hale 1996: 151), as Comrie (Ch. 2) argues, as well as an irretrievable loss of historical information about the language and its speakers, since it would


159

also mean an additional loss for their intellectual heritage. In other words, it would decrease the total amount of human understanding and knowledge about the world, and the “loss of any language is a loss for all humanity”, as concluded in the recommendations by the UNESCO Ad Hoc Expert Group on Endangered Languages (2003 at Paris). Each language, with the whole culture encapsulated, may be something to be fully entitled to the status of Masterpieces of the Intangible Heritage of Humanity. What is more, I would reiterate that structural diversity manifests itself in language to a greater extent than in (non-linguistic) culture (§ 8.3, 8.4) and to the greatest in linguistic morphology (§ § 8.5). If it is the case that the latter is “nothing more nor less than a collective art of thought” (Sapir 1921: 233 [1949: 218]), those threatened languages which still retain without much attrition the impressive and highly distinctive characteristics foreign to the major threatening languages could possibly reflect as many threatened ways of viewing the world. It would be none but human beings themselves, gifted with logos—language and faculty of reasoning—that can have intelligent control over the two currently urgent issues of maintaining the health of the ecosystem and the intellectual vitality of humanity. As Krauss (§ 1.6.1) argues, we have no option in intelligent planning for the future other than to depend upon this uniquely human gift. In exercising the gift, however, we should probably be alert to the chance that what is arrogantly claimed as “rationality” might possibly be biased by a particular language (and culture) and be in itself an extreme expression of a kind of folie (M. Foucault) which is liable to totally and forcibly exterminate what is regarded as “irrationality” of other peoples. This is not just a historical case such as was implied in B. de las Casas’s inner denunciation (1552) to defend the Indios “discovered” in the New World but may possibly be what we, quite simply, witness through recent world-shaking history. We students of linguistics are finally on the cusp of realizing that linguistic diversity, particularly with great typological difference and deep genealogical divergence, may testify to the rich and priceless variety of human thought, which could be the very source of intellectual vitality and thereby of the health and strength of the logosphere (§ 8.3.2). If the diversity may have the effect of effacing the folie inherent to the arrogance of “rationality”, it would be far from “poor economy” or something to be passively permitted, but should be something to be positively valued as the greatest asset of humanity and to be securely safeguarded as an indispensable requisite of it. Diminution of linguistic diversity, together with environmental

160


deterioration, would be the gravest issues that put human intelligence to trial today.

References Brenzinger, M., Heine, B., and Sommer, G. (1991). “Language Death in Africa”, in R. H. Robins and E. M. Uhlenbeck (eds.), Endangered Languages. Oxford: Berg, 19–44. Croce, B. (1902). The Aesthetic as the Science of Expression and of the Linguistic in General, trans. C. Lyas, from Estetica come science dell’espressione e linguistica generale, part I (1902), Cambridge: Cambridge University Press. Crystal, D. (2000). Language Death. Cambridge: Cambridge University Press. Dixon, R. M. W. (1997). The Rise and Fall of Languages. Cambridge: Cambridge University Press. —— and Aikhenvald, A. Y. (2002). “Word: a typological framework”, in Dixon and Aikhenvald (eds.), Word: A Cross-Linguistic Typology. Cambridge: Cambridge University Press, 1–41. Du Ponceau, P. S. (1838). A Dissertation on the Nature & Character of the Chinese System of Writing, in a Letter to John Vaughan, Esq. Transactions of the Historical and Literary Committee of the American Philosophical Society 2. Philadelphia. Evans, N., and Sasse, H.-J. (eds.) (2002). Problems of Polysynthesis. Studia Typologica, NS. Berlin: Akademie Verlag. Fortescue, M. (1994). “Polysynthetic morphology”, in R. E. Asher (ed.), The Encyclopedia of Language and Linguistics, vol. v. Oxford: Pergamon Press, 2600–2. Goldenweiser, A. A. (1913). “The principle of limited possibilities in the development of culture”, Journal of American Folk-Lore, 26: 259–90. —— (1933). History, Psychology, and Culture. New York: Alfred A. Knopf. Grace, G. W. (1987). The Linguistic Construction of Reality. London: Croom Helm. Hale, K. (1996). “Universal grammar and the roots of linguistic diversity”, MIT Working Papers in Linguistics, 28: 137–61. Hattori, S. (1950). “Bound words and bound forms” [付属語と付属形式], GengoKenkyuu,15:1–26. Hjelmslev, L. (1953). Prolegomena to a Theory of Language, trans. F. J. Whitfield. IJAL Memoir 7. (Omkring sprogteoriens grundlæggelse, 1943, University of Copenhagen.) —— (1966). Le Langage: une introduction, trans. M. Olsen. Paris: Les Éditions de Minuit. (Sproget–En Introduktion, 1963, Capenhagen: Berlingske Forlag.) Kamei, T., Kono, R., and Chino, E. (eds.) (1988–96). The Sanseido Encyclopaedia of Linguistics [言語学大辞典], 6 vols. Tokyo: Sanseido. Kono, R. (1977). “The true nature of writings [文字の本質]”, in Iwanami Series, Japanese 8 (Writings) [岩波講座日本語 8 (文字)] (reproduced in R. Kono, On Writings 文字論, 1–24, 1994, Tokyo: Sanseido). Krauss, M. E. (1992). “The world’s languages in crisis”, Language, 681: 4–10. —— (1996). “Linguistics and biology: threatened linguistic and biological diversity compared”, Papers from the Parasession on Theory and Data in Linguistics, CLS 32. Chicago: Chicago Linguistic Society, 69–75.


161

—— (2001). “Mass language extinction, and documentation: the race against time”, Lectures on Endangered Languages, ii: From Kyoto Conference 2000. ELPR C-002, 19–39. Li, Ch. (2005). “An fragment of language study [言語研究に寄せる断章]”, ∆ΥΝΑΜΙΣ Language and Culture (Kyoto University), 9: 39–53. Martinet, A. (1949). “La Double articulation linguistique”, TCLC 5: 30–7. —— (1960): Éléments de linguistique générale. Paris: Librairie Armand Colin. —— (1965a): ‘Le mot’, Diogènes, 48: 39–53. —— (1965b): La linguistique synchronique, 8th edn. 1975. Paris: Presses Universitaires de France. Mattissen, J. (2003). Dependent-Head Synthesis in Nivkh: A Contribution to a Typology of Polysynthesis. Typological Studies in Language 57. Amsterdam: John Benjamins. —— (2004). “Structural typology of polysynthesis”, Word, 55/2: 189–216. Miyaoka, O. (1971). “On syllable modification and quantity in Yuk phonology”, International Journal of American Linguistics, 37: 219–26. —— (1996a). “Sketch of Central Alaskan Yupik, an Eskimoan language”, in W. C. Sturtevant (gen. ed.), Handbook of North American Indians; I. Goddard (vol. ed.), xvii: Languages. Washington, DC: Smithsonian Institution, 325–63. —— (1996b). “The constitution of culture and functions of language [文化のしくみと言語のはたらき]”, in Miyaoka (ed.) (1996c: 3–41). —— (ed.) (1996c). An Introductory Reader on Linguistic Anthropology [言語人類学を学ぶ人のために]. Kyoto: Sekai-shisosha. —— (2000a). “Morphologie verbale en yupik alaskien central”, in N. Tersis and M. Therrien (eds.), Les langues eskaléoutes. Sciences du langage, Collection ed. Christian Hudelot. Paris: Centre National de la Recherche Scientifique, 225–48. —— (2000b). “A reflection on the ‘word’ : ‘Articulation’ in natural languages [についての断想自然言語における”, valedictory speech, Kyoto University, 11, Mar. 60. —— (2001). “Endangered languages: the crumbling of the ecosystem of language and culture”, in Lectures on Endangered Languages, ii: From Kyoto Conference 2000, ELPR C-002, 3–17. —— (2002). What is a ‘Word’? Japanese Viewed from Eskimo [とはなにかエスキモー語から日本語をみる]. Tokyo: Sanseido. —— (in preparation). “A preception of a ‘word’ ”. Nettle, D. (1999). Linguistic Diversity. Oxford: Oxford University Press. Nichols, J. (1992). Linguistic Diversity in Space and Time. Chicago: University of Chicago Press. Nishida, K. (1940). Problems of Japanese Culture [日本文化の問題]. Tokyo: Iwanamishoten. Sapir, E. (1921). Language: An Introduction to the Study of Speech. New York: Harcourt, Brace and Company. [A Harvest Book edition, 1949.] —— (1927). “The unconscious patterning of behavior in society”, in E. S. Dummer (ed.), The Unconscious: A Symposium. New York: Knopf, 114–42. [Selected Writings of Edward Sapir, ed. D. G. Mandelbaum, Berkeley and Los Angeles: University of California Press, 1951, 544–59.] Sapir, E. (1929). “The status of linguistics as a science”, Language, 5: 207–14. [Selected Writings of Edward Sapir, 160–66.]

162


Sapir, E. (1933) [1951]. “Language”, in Encyclopaedia of the Social Sciences, vol. ix. New York: Macmillan, 155–69. [Selected Writings of Edward Sapir, 7–32.] Taylor, J. R. (1995). Linguistic Categorization, 2nd edn. Oxford: Clarendon Press. UNESCO Intangible Cultural Heritage Unit’s Ad Hoc Expert Group on Endangered Languages (2003). Language Vitality and Endangerment. 10 Mar. 2003, Paris.

PART II Areal Surveys The South Pacific Rim


9

Languages of Middle America Yoshiho Yasugi

9.1. Current situation of Middle America 9.1.1. Typological characteristics 9.1.2. Threatened languages 9.2. Attempts to salvage endangered languages 9.2.1. Colonial documentation 9.2.2. Modern documentation 9.3. Maya revitalization movement 9.3.1. Standardized orthography 9.3.2. Language preservation

165 165 171 173 173 174 177 177 179

9.1. Current Situation of Middle America Many indigenous languages of Middle America (from Mexico to Panama) have disappeared since the conquest in the sixteenth century. During the last 500 years, 244 out of 365 languages, which include languages and dialects known only by name, were lost according to McQuown (1955), while B. Garza Cuarón and Y. Lastra (1991: 120–8) give a substantial list of 167 extinct languages in Mexico and Guatemala, and now only ninety-two languages are spoken by over nine million people (Yasugi 1995: 5–10).1 These languages can be classified into at least seventeen language families and isolates (as shown in Table 9.1) which show considerable typological diversity not only in phonemic systems but also in morphological and syntactic levels.

9.1.1. Typological Characteristics In the phonemic systems the number of consonants varies from eleven to thirty-five. Almost all languages share the plain consonants; p t c c k ’ s š h 1 The number of the languages is taken from Yasugi (1995). Yasugi counts fifty-seven languages in Mexico, taking in consideration the traditional standard, while Manrique Castañeda (1997) and Ethnologue count seventy-seven and 295 languages, respectively. The number varies with the definition of language. For example, some say Zapotecan is only one language, while others say Zapotecan contains eight languages and still others see it as consisting of ten or more (Manrique et al. 1994–5) and Ethnologue counts fifty-eight languages. On the other hand, Yaqui and Mayo are traditionally taken as two separate languages but they should be classified as only one language, Cahita. Traditionally, Zapotecan, Mixtecan, Chinantecan, and Mazatecan are treated as one language, but if we apply the criteria for the classification of Mayan to them, the number increases as is indicated above.

Figure 9.1. Distribution of the Native Middle Americian Languages

l a n g uag e s o f m i d d l e a m e r i c a

167

Table 9.1. Classification of Middle American Indian languages Language

Location

Number of speakers

I. Southern Uto-Aztecan A. Tepiman (Pimic) 1. Piman Pima Alto [1] 10,000 Papago [2] 15,000 Pima Bajo (Nevome, Ure, Yecora) [3] 740 2. Tepehuan (Odami/Odame) 17,900/18,470/25,550 Northern Tepehuan [4] Southern Tepehuan [5] *Tepecano D1 0 B. Taracaitan (Taracahitic) 1. Tarahumaran Tarahumara (Rarámuri) [6] 62,500/54,430/75,550 Guarijío (Varohío) [7] 1,670 2. Opatan *Opata(Teguima) D2 #12 *Jova D3 0 *Eudeve (Heve, Dohema) D4 0 3. Cahitan Yaqui (Cahita) [8] 9,300/10,990/13,320 Mayo (Cahita) [9] 56,400/37,410/31,510 4. *Tubar D5 0 C. Corachol Cora [10] 12,300/11,920/16,410 Huichol [11] 51,900/19,360/30,690 D. Nahuan 1. Azteca (General Aztec) 1,377,000/1,197,330/1448,940 Nahuatl [12] Nahual [13] Nahuat [14] Pipil [15] 20 2. *Pochutec D6 0 II. *Cuitlatec D7 0 III. Yuman (includes only Yuman languages of Mexico.) Paipai [16] 200 Cochimi (Kumyai, Kimiai) [17] 240 Kiliwa [18] 50 Cocopa (Cucapa) [19] 180 IV. Seri [20] 500/560/460 V. Tarasco (Purepecha) [21] 118,700/94,840/121,490 VI. Totonacan Totonac [22] 196,100/207,880/240,040 Tepehua [23] 8,500/8,700/9,440 VII. Otomanguean A. Chichimec (Meco, Jonaz) [24] 1,000?/1,620/1,640 B. Oto-Pamean 1. Pamean [25] 57,00/5,730/8,310 North Pame, Central Pame, South Pame 2. Matlatzincan Matlatzinca (Pirinda) [26] 18,00/1,450/1,300 (Continued)

168

yo s h i h o ya s u g i

Table 19.1. (Continued) Language

Location

Number of speakers

Ocuiltec (Tlahuica) [27] 400/760/460 3. Otomian (a) Otomí [28] 306,200/280,240/291,720 North-Western Otomí (Mesquital), Northeastern Otomí (Sierra) South-Western Otomí, Ixtenco Otomí (b) Mazahua [29] 194,200/127,830/133,430 C. Supanec 1. Tlapanec (Yope) [30] 55,100/68,480/99,390 2. *Subtiaba D8 0 (*Maribio El Salvador) D. Popolocan 1. Chochoan (a) Ixcatec [31] 200?/1,220/350 (b) Chochoan Popoloca [32] 6,800/1,730/16,470 Chocho [33] 12,400/12,550/990 2. Mazatec [34] 124,200/168,370/214,480 E. Amuzgo [35] 18,700/28,230/41,560 F. Mixtecan 1. Mixtecan Mixtec [36] 323,200/386,870/444,550 Cuicatec [37] 14,200/12,680/13,430 2. Trique [38] 8,500/14,980/20,710 G. Zapotecan 1. Zapotec [39] 423,000/403,460/452,890 (*Papabuco #5) 2. Chatino [40] 20,600/28,990/40,720 H. Chinantec [41] 77,100/109,100/133,370 I. Manguean (Chorotegan, Chiapanec-Mangue) 1. *Chiapanec D9 #180 2. *Mangue D10 0 (*Diria Nicaragua) (*Chorotega Honduras) (*Nicoya Costa Rica) VIII. Huave [42] 10,000/11,960/14,220 IX. Oaxaca Chontal (Tequistlatec) [43] 81,000/ 4,670/4,960 Lowland Chontal (Huamelultec) Highland Chontal (Tequistlatec) X. Mixe-Zoque (Zoquean, Mixean) 1. Zoquean [44] 31,000/43,160/51,460 (a) Oaxaca Zoque (San Miguel Chimalapa, Santa Maria Chimalapa) (b) Chiapas Zoque (c) Tabasco Zoque (Ayapa) (d) Veracruz Zoque (Zoque Popoluca) [45] Sierra Popoluca (Soteapan etc.) 23,800/ 29,030/38,140 Texistepec Popoluca 170 2. Mixean (a) Veracruz Mixe (Mixe Popoluca) [46] Sayula Popoluca 4000 Oluta Popoluca 3 (b) Mixe [47] 74,100/ 95,260/118,920 Eastern Mixe, Western Mixe (c) *Tapachultec D11 0

l a n g uag e s o f m i d d l e a m e r i c a Language

Location

169

Number of speakers

XI. Mayan A. Huastecan 1. Huastec [48] 66,100/103,800/120,740 2. *Chicomuceltec D12 #20 B. Northern Lowland Maya 1. Yucatecan (a) Yucatec [49] 665,400/713,520/800,290 (b) Lacandon [50] 300/100/40 (c) Itzaj [51] 650/1,835 (d) Mopan [52] 8,500/13,460 C. Southern Lowland Maya 1. Cholan (a) Chol [53] 96,800/128,240/161,770 (b) Chontal [54] 29,000/30,140/38,560 (c) Ch’orti’ [55] 27,097/76,782 (d) *Cholti D13 0 2. Tzeltalan (a) Tzotzil [56] 133,400/229,200/297,560 (b) Tzeltal [57] 215,200/261,080/284,830 (c) Tojolabal (Chaneabal) [58] 22,400/36,010/37,990 D. Western Highland Maya 1. Kanjobalan (a) Chuj [59] 50,000/87,489 (b) Jakalteko/Popti’ [60] 39,635/86,266 Q’anjob’al [61] 75,155/211,687 Akateko [62] 40,991/40,991 (c) Motocintlec (Mocho) [63] 170 Tuzantec [64] 300 2. Mamean (a) Tektiteko (Teko) [65] ?/ 4,895 Mam [66] 346,548/1,126,959 (b) Awakateko [67] 18,572/35,485 3. Ixil [68] 47,902/134,599 E. Eastern Highland Maya 1. Q’eqchi’ [69] 473,749/732,340 2. Poqom (a) Poqomchi’ [70] 94,714/266,750 (b) Poqomam [71] 46,515/130,928 3. K’ichean (a) Uspanteko [72] 12,402/22,025 (b) K’iche’ [73] 663,241/1,954,007 Sakapulteko [74] 3,033/43,439 Sipakapeño [75] 4,409/6,118 Kaqchikel [76] 343,038/1,032,128 Tz’utujil [77] 50,080/160,907 XII. Xinca [78] 107/306 XIII. Arawakan (includes only a Central American language.) Garífuna(Black Carib) [79] 70,000 XIV. Lencan *Lenca (Honduran Lenca) D14 0 *Chilanga (Salvadoran Lenca) D15 0 XV. Tol (Jicaque) [80] 300 (Continued)

170


Table 9.1. (Continued) Language

Location

Number of speakers

XVI. Misumalpan (Misuluan) A. Mískitu [81] B. Sumu (Ulwa=Southern Sumu) [82] Bawihka, Tawahka, Kukra, Panamaka C. Matagalpan *Cacaopera D16 *Matagalpa D17 XVII. Chibchan (includes only Central American Chibchan languages) A. Paya (Pech) [83] B. Rama [84] C. Guatuso (Malecu) [85] D. Boruca (Brunca) [86] E. *Huetar (Guetar) D18 F. Viceita Cabécar (Chiripó, Estrella) [87] Bribrí [88] G. Teribe/Téraba [89] H. Guaymí [90] I. Bocotá [91] J. Cuna [92]

67,000 4,900 0 0 300 650 300 5 0 6,000 5,000 1,100 56,500 15,000? 36,500

Notes: The population of Mexico is cited from the Census of 1980/1990/2000. The number includes speakers of more than 5 years old. Fractions have been rounded off to the nearest even number. The Guatemalan Mayan population is cited from Comisión de Oficialización de los Idiomas Indígenas de Guatemala (1999)/ Tay Coyoy (1996). The number of speakers in Central America is based mainly on Turpana (1987) and García Segura and Zúñiga Muñoz (1987). D before the identification number signifies that the language is extinct.

m n l r w y. Glottalized consonants are added to them in Mayan, Huehuetla Tepehua, Oaxaca Chontal, Xinca, and Tol. Prenasal stops are seen in Mixtecan and a fortis–lenis contrast is reported in Zapotecan. Tarascan, Terribe/Terraba, and Tol have aspirated stops. The number of vowels ranges from three to nine with elongation or nasalization. Otomanguean and Chibchan have tonal systems with two to five level tones and some combination of tones (Yasugi 1995: 11–57). Morphologically Tarascan has only suffixes while other languages have both prefixes and suffixes. Most Middle American languages are head marked. In terms of word order, the central part of Middle America, called Mesoamerica as a cultural area, is sandwiched by SOV and postpositional languages. In addition to SOV order and postpositions, Lower Central America languages share the traits of Genitive-Noun, Noun-Adjective, Noun-Numeral orders, while northern Mexican languages have modifierhead (Genitive-Noun, Adjective-Noun, Possessive-Noun, DeterminativeNoun, Numeral-Noun) orders. Almost all populous indigenous languages are concentrated in Mesoamerica; most Otomanguean are VSO and Mayan are VOS or VSO, head-modifier languages, while Mixe-Zoquean and


171

Nahuan permit SVO/VOS/VOS, PO/PR, GN/NG orders, though they seem to have been SOV, modifier-head languages. Some small families are also in Mesoamerica; Totonacan and Huave are SVO, modifier-head languages; Oaxaca Chontal is a VOS, modifier-head language (Yasugi 1995: 107–31).

9.1.2. Threatened Languages All Middle American languages are more or less in danger of extinction. The reason of endangerment varies according to the language in question, but the number of speakers is one of the most important factors to judge whether or not a language in question is “moribund”, i.e. on the verge of extinction. For example, in the case of Ixcatec, which is spoken in Oaxaca, Mexico, the number of speakers has been reduced to a number too small to maintain the language. Therefore, I will list all the languages whose speakers are less than 1,000. We have the 1990 census data of speakers under 5 years and that of bilingualism in Mexico. The number of speakers of the year 2000 is smaller than that of 1990 in many languages and the percentage of bilingual speakers is increasing (67.5 per cent in1950, 83.1 per cent in 2000). Although the sociolinguistic data are lacking, it seems that children do not inherit their native language. Almost all children speak Spanish and therefore it is natural to think that these are definitely threatened languages. However, it is very difficult to evaluate the degree of endangerment: Lacandon has only a few hundred speakers, a number too small to maintain the language, but people still continue to speak it proudly.2 Pipil is almost extinct, with about twenty old speakers out of almost 200,000 in that ethnic group. The numbers of Chichimec and Ixcatec are 200 and 119 according to Ethnologue, while the census of the year 2000 records that the numbers of these ethnic groups are 1,641 and 351, respectively. The languages of Opata, Chiapanec, and Chicomuceltec are extinct but some people still maintain their ethnic identity, with the number of 12, 181, and 24 respectively. Besides the small number of speakers, Spanish influence has been a serious factor in language endangerment. The Spanish language is dominant in Middle America except in Belize where English is dominant, and indigenous languages are subordinate. Subordinate languages are always in danger of extinction, even though they outnumber dominant language speakers (for example, Kaqchikel, one of the Mayan languages, is spoken around the Guatemalan capital city). They are strongly under Spanish 2 In the spelling of Mayan language names, I have followed the new orthography officially recognized in 1987 for languages spoken in Guatemala, but I have maintained the traditional spellings for languages spoken outside Guatemala.

172


Table 9.2. Languages whose speakers are less than 1,000 Census year

Paipai [16] Cochimi (kumyai) [17] Kiliwa [18] Cocopa [19] Pima Bajo [3] Seri [20] Ocuiltec [27] Ixcatec [31] Texistepec Popoluca [45] Oluta Popoluca [46] Lacandon [50] Motocintlec [63] Tuzantec [64]

1990

2000 Endangerment degree(ED)

Total

Speakers under 5 years

Speakers speaking Spanish

223 96 41 36 732 561 755 1220 172 3 104 235 300

48 34 6 17 183 68 216 280 50

48 33 6 16 176 64 204 262 47

201 243 52 178 741 458 466 351

26 98

25 94

40 174

3 3 3 3 4 3 3 3 3 1 4 3 3

Notes: Endangerment degree (ED): 4 = potentially endangered; 3 = endangered; 2 = seriously endangered; 1 = moribund.

influence. Although about a million people speak Kaqchikel, the language is also in danger of extinction due to the fact that the children prefer to speak Spanish instead of speaking their mother language. Here, we can observe gradual shift to the dominant language. The parents speak well, but their children do not, although they do understand what their parents speak. A policy of suppression also affects survival of a language. After the peasant uprising in 1932, the Salvadoran government suppressed the indigenous people. The result was that the Lenca and Cacaopera languages were abandoned and became extinct. The sudden language loss was due to a halt in speaking their native languages in order to avoid being identified as communist-inspired Indians (Campbell and Muntzel 1989: 183). During the civil war in Guatemala, which lasted over thirty years, there was a great danger of extinction. Maya people became more silent for fear of death. If genocide had escalated, speakers would have stopped speaking the language. Fortunately, this did not occur and the suppression finally came to an end in 1996, costing about 75,000 lives in the war (Warren 1997: 24). During that time, a number of young Maya people began to learn their own cultural heritage and found their values and pride. New movements toward indigenous liberation and identity are now becoming vigorous.


173

9.2. Attempts to Salvage Endangered Languages 9.2.1. Colonial Documentation Linguistic description and analysis of indigenous Middle American languages began soon after the Spanish conquest in the early sixteenth century. When Spanish missionaries encountered indigenous people, they must have been perplexed at the strange sounds they had never heard. Spanish missionaries studied indigenous languages and established orthographies for them, based on the Latin or Spanish alphabet. Sometimes they invented new letters such as for /tz’/ and ε for /q’/ etc. Based on the new orthography, they wrote grammars, dictionaries, catechisms, etc., and taught native elites how to write and read their languages. Indigenous people wrote myths, legends, letters, wills, etc. using the alphabet they had just learned. Among indigenous languages, Nahuatl and Yucatec have the most extensive documentary sources. Some of them were written by Spanish missionaries and others were recorded by indigenous people. Some other dominant languages in those days, such as Zapotec, Mixtec, Tarasco, K’iche’, and Kaqchikel, also have documents, dictionaries, and grammars, although they are fewer than Nahuatl and Yucatec. Other minor languages were poorly documented. Nevertheless, languages in Mexico and Guatemala are more documented than Central American languages (Contreras García 1985; Campbell et al. 1978). During the colonial period, fortunately compared with, say, North America, indigenous languages continued to attract interest, although the colonial government went ahead with a policy of Castilianization. There were professors of Nahuatl and Otomi in the University of Mexico as early as 1640 (Ligorred 1992: 17). In Guatemala Kaqchikel, a Mayan language, was taught in the university as a metropolitan language from the years 1678 to 1822 (Brinton 1884: 347). After independence, the word “indio” was prohibited to be used in any documents under this liberal thought, and the class of “indio” became non-existent (Favre 1996: 25). Consequently indigenous languages were ignored and indigenous people who had had a literacy tradition did not inherit it. In the twentieth century, indigenous languages came to the attention of North American and Mexican anthropologists. After the Public Education Secretariat (Secretaría de Educación Pública; SEP) was founded in 1921, alphabetization of indigenous languages started in order to Hispanicize indigenous peoples. In 1935, the Independent Department of Indian Affairs was created. In that same year the Summer Institute of Linguistics (SIL) c

174


began to study indigenous languages to diffuse the Bible among indigenous people. The SIL carried out descriptive studies for the vernacular languages and prepared didactic materials (cartillas) for the alphabetization of indigenous languages. The Institute for Alphabetization of Indigenous Languages was established in 1945 and the national campaign against analphabetism was organized (Bravo Ahuja 1977: 112). However, the purpose of alphabetization was to facilitate the learning of the national language for ethnic groups. The Mexican education system starting in 1921 operated under the thesis of incorporation and assimilation that denies the development of indigenous cultures. Bilingual-bicultural education became official in 1955 when the General Bureau of Indigenous Education (Dirección General de Educación Indígena) was created within the Public Education Secretariat, but the meaning of bilingual-biculture was different from that of our days. It did not mean teaching in both Spanish and indigenous languages or respecting the cultures of indigenous peoples. To teach indigenous languages was to facilitate the learning of Spanish. Even now, bilingual education tends to be considered as a learning method to facilitate the transition from the indigenous language to Spanish. Indigenous people were given an opportunity for education, but in practice their languages were disrespected and documentation of the indigenous languages was not a desideratum.

9.2.2. Modern Documentation Modern documentation of Middle American languages has been done by linguists. Major languages have at least some descriptions of phonology. Some languages have dictionaries and grammars. Especially SIL has contributed to documentation, and its linguists have published many useful dictionaries and grammars, but most of the early dictionaries are too short and lack helpful examples to demonstrate the significance of the terms. In the early 1970s documentation of the languages became popular among North American linguists. They studied indigenous languages with indigenous people working as informants and in exchange they taught them how to write and read. Some indigenous people working with North American linguists learned basic descriptive linguistics and dictionary-making procedures. Now they themselves produce dictionaries and grammars, especially in Guatemala (see Yasugi 2003: pp. xi–xvi).


175

To document a language adequately, three parts are required: a dictionary, grammar, and morphemically analysed texts. Looking at language documentation from this point of view, there are very few projects meeting the requirements of this standard. An example of an almost ideal documentation is the Itzaj case, where Hofling published a grammar, dictionary—with more than 20,000 entries—and texts (Hofling 1991, 2000; Hofling and Fernando Tesucún 1997). Recently the most extensive dictionary is the Great Tzotzil Dictionary of San Lorenzo Zinacantan—with 35,000 entries—by Laughlin (1975). Most other dictionaries contain fewer entries. I will now survey language documentation in each country. In Mexico linguists at Universidad Nacional Autónoma de México and Instituto Nacional de Antropología e Historia have played a central role in documentation. Besides their activities, two other projects should be mentioned. Jorge Suárez and others organized the Archive of Indigenous Languages in Mexico (Archivo de Lenguas Indígenas de México; formerly Archivo de Lenguas Indígenas del Estado de Oaxaca). The first publication was the data of Isthmus Zapotec (Pickett and Embry 1975). The book consists of 594 sentences responding to a questionnaire for structural analysis and short vocabulary. They had originally intended to describe indigenous languages in the state of Oaxaca, but, amplifying their scope later and applying the same standard to other languages, they had published data for twenty-five languages of Mexico by the year 2003. The series covers languages like Matlatzinca and Chocho, in addition to currently nonendangered (safe or stable yet threatened) languages like Zapotec, Mixtec, Otomi, etc. In 2003 the data of sixteen Guatemalan Mayan languages were added by Yasugi (2003). In 1993 the Project for the Documentation of the Languages of Mesoamerica, which was organized by T. Kaufman and J. Justeson, began research on Mixe-Zoquean languages, and then extended to Zapotecan, Oto-Pamean, Totonacan, and Nahuan. Now the Project has been studying four Mixean, six Zoquean, ten Zapotecan, two Chatino, ten Nahuan, three Totonacan, and two Oto-Pamean consisting of Matlatzinca and Ocuiltec. The preparation of a dictionary for each language is undertaken by a different linguist in collaboration with native speakers. A major feature of the Project is that a set of specialists in each language family is trained in the context of regular and long-term interaction, helping to generate a body of lore that is tested through discussion and comparison of the results of individual investigation. The dictionaries they are producing are medium in size,

176


ranging between 5,000 and 10,500 lexical items. Part of their results can be accessed at the website—www/ablany.edu/ anthro/maldp/index/html. In Guatemala, T. Kaufman founded the Proyecto Lingüístico Francisco Marroquín, where about eighty indigenous people learned how to write their language. USA linguists (such as N. C. England) analysed Mayan languages in collaboration with native speakers. Kaufman published the basic spelling and phonology of Guatemalan languages (Kaufman 1970). Mayan people organized the Academy of Mayan Languages (Academia de Lenguas Mayas de Guatemala) and established a “unified” alphabet for all Mayan languages spoken in Guatemala in 1987. The Academy of Mayan Languages set up branches in the head towns of each language group. The branches function as a regional centre for making up bilingual education programmes, surveying dialectal variation, and making dictionaries, grammars, etc. In 1990 N. C. England founded OKMA (Oxlajuuj Keej Maya Ajtz’iib’) where she trained native Mayan linguists. Now these Mayan linguists analyse their native languages and publish grammars, dictionaries, and dialectal descriptions. In Honduras and El Salvador, there are moribund languages, such as Pipil, Tol (Jicaque), and Pech (Paya). Tol (300 speakers) and Pech (300 speakers (Ethnologue at www.ethnologue.com/show_country.asp?name=HN) may be too close to death to be salvaged. As for Garifuna, whose territory extends to Belize and Nicaragua, documentation is also needed, although it does have a dictionary and a grammar (Hadel 1975; Suazo 1990). Pipil (nearly extinct) was documented by Campbell (1985). Other languages such as Lenca and Chilanga are extinct with little documentation (see Lehmann 1920; Andrews 1970; Campbell 1976). In Nicaragua, the Center for Research and Documentation of the Atlantic Coast (CIDCA) is responsible for the description of indigenous languages in that country (Craig 1992). The Nicaraguan linguists along with North American linguists are producing dictionaries of Ulwa, Miskito, and Rama (Marx and Heath 1992; Anonymous 1989; Rigby and Schneider 1989). After the Sandinista revolution was over, however, the revitalizing movement seems to have become less active. In Costa Rica, linguists at the University of Costa Rica are rescuing indigenous languages of the country. Since 1982 they have yearly published the Journal of Studies of Chibchan Linguistics (Estudios de lingüística chibcha) where they report the data of indigenous languages like Boruca (nearly extinct), Terraba, Bribri, Cabecar, Guaymi, etc. They have also published dictionaries and grammars (Constenla Umaña 1998; Constenla Umaña et al. 1998; Margery 1989).


177

Concerning the languages of Panama, we have good descriptions thanks to Lenguas de Panama (vols. i–viii) published by Instituto Nacional de Cultura and Instituto Lingüístico de Verano. For more information, see Ethnologue (www.ethnologue.com/show_country.asp?name=Panama).

9.3. Maya Revitalization Movement Indigenous people have come to recognize their ethnic distinctiveness and are beginning to promote their cultural heritage. The Maya revitalization movement that began in Guatemala in the late 1980s is the most active one among Middle American indigenous language groups. I can give some reasons why documentation is the most active. The most important contribution is made by North American linguists such as T. Kaufman, N. C. England, etc. who have been educating Maya people since the mid 1970s. These Maya people have grown up to be linguists who describe their own native languages. Other reasons may be that Guatemala has proportionately the greatest population of indigenous people and that the Maya people are descendants of the ancient Maya civilization about which they can boast. Worldwide movements have also influenced Maya revitalization activities. Although revitalization is very active, there are still problems; so I will discuss Guatemalan activities with regard to language preservation.

9.3.1. Standardized Orthography In Guatemala, in 1950 a new orthography was proposed for four major Guatemalan languages, Kaqchikel, K’iche’, Q’eqchi’, and Mam, and in 1952 SIL was established. They adopted the orthographical practices of Spanish such as ca, que, qui, co, cu for the sounds [ka], [ke], [ki], [ko], [ku] to ease the transition to Spanish for monolingual children. In 1970, however, the members of the Instituto Lingüístico Francisco Marroquín proposed new orthography. This orthography influenced later linguistic policy. In 1987 the Guatemalan government legalized a standardized orthography (Anonymous 1988). Now scholars use this orthography because it was legalized. The old orthography was rejected and the approach of earlier scholars was later regarded as assimilationist.

178


A standardized writing system is indeed important for literacy and cultural stability, but the orthography adopted by the Maya people is not applied to their own names and the names of their towns. They are familiar with the older still prevailing colonial spelling of place names and surnames. It is difficult to change traditional names into new ones, because legal documents must be conserved and the spelling in documents must inevitably be sustained. Because of the use of traditional names in daily life, two different spellings coexist. Then for what purpose did indigenous people change orthography and what advantage do they get? The reason they have changed traditional orthography and adopted a new one was to throw off the yoke of Spanish domination and to strengthen their ethnic identity. Since indigenous people were despised and indigenous languages were regarded as imperfect and inferior for many years, it is most important to take pride in their languages for salvaging and maintaining them by using the new orthography as a symbol of new movements. Spanish as the dominant language is indispensable for everyday life, while reading and writing in their own languages is not. If asked what advantage they have when they learn this new orthography, I must confess that it is useful only among the language groups and their knowledge does not spread to the outside world. However, if they do not know how to read and write Spanish, they can suffer some obstacles to living in their society. Rigoberta Menchú, a Nobel Peace Prize winner, wrote about many of the losses and absurd treatment she suffered from Ladinos (i.e. non-Indias) because she did not know Spanish (Rigoberta Menchú 1987). It is more economical to teach only the dominant language of Spanish in school. Teaching in both Spanish and an indigenous language at the same time may theoretically take twice as much time as teaching only in Spanish. For many years teachers tried to teach the writing and reading of indigenous languages in the first and second grades at elementary school. But even if children learn to write the indigenous language for two years, they cannot write and read their language fluently, because the education system is organized to serve for easy transition to Spanish and they have no need to write their language. More than ten years ago indigenous people unsuccessfully demanded that the law should be written in every indigenous language. If we consider an outcome proportionate to the cost, it is natural to think that it is less costly to make them understand the law in Spanish rather than to write the law in every indigenous language, because there are more than twenty languages and each language has far fewer speakers than Spanish-speaking people.


179

9.3.2. Language Preservation However, we must not treat language merely from the point of economy or of convenience. Now the Maya people promote bilingual education in schools called escuelas mayas (Maya schools), even though tremendous time needs to be devoted to developing proper curricula. In one Maya school at Palin I visited in 2001, children were taught in Poqoman. There are more children speaking Poqoman than there were some twenty years ago when the language was in danger of extinction, because Palin is near the Guatemalan capital. At that time many Poqoman people preferred to speak Spanish, partly because they were looking for a job in Guatemala City and partly because indigenous languages were looked down upon. Now, however, Poqoman may possibly have escaped from the danger of extinction, thanks to cultural revival. Ladinos say that there is no official discrimination, but actually there is a gap between them and indigenous people: the latter distrust the former, who despise the latter. Although we hardly know how to achieve this, it is important for Ladinos to come to respect indigenous languages and culture. However, the mobility and liberation of society is progressing in Guatemala after a hard time in the early 1980s. Years ago, once indigenous people learned Spanish, they abandoned speaking their native tongue and transferred their ethnic identity to Ladino. The change of identity went with the change of language usage. Especially the more highly educated indigenous people had abandoned their Indio-ness. But now, even people who cannot speak a Mayan language begin to declare that they are Maya. In Middle America there are many indigenous languages and generally each community has its own dialect. When indigenous people want to communicate with other groups, they have to speak in Spanish as a common language. Therefore, in order to prevent the extinction of these languages, it is vital, needless to say, that people speak to children in their language, thereby promoting bilingualism in their communities. To popularize bilingualism, normalized texts of a standard dialect should be published, though it has a great risk of diminishing dialect diversity. A standard dialect may be the dialect of regional centres where the Academy of Mayan Languages set up the branches. The Academy staffs are now studying their dialects and trying to write normative grammar and texts, and so the Academy should take a leading role in training teachers who can read and write indigenous languages. If these teachers start to teach in rural schools, there may be some hope that bilingual education will be fortified to some extent. Their operations are now active, but they depend on financial aids: if these are cut, then

180


their movements may deteriorate easily. Under many negative and very difficult situations, the most important thing should be for indigenous people to have a real passion and will for maintaining their own language. It entirely depends upon their awareness of the issue.

References Andrews V., and Wyllys, E. (1970). “Correspondencias fonológicas entre el lenca y una lengua mayance”, Estudios de Cultura Maya, 10: 341–87. Anonymous (1988). Lenguas mayas de Guatemala: documento de referencia para la pronunciación de los nuevos alfabetos oficiales. Guatemala: Instituto Indigenista Nacional y Ministerio de Cultura y Deportes. —— (1989). Diccionario elemental del ulwa (sumu meridional). Nicaragua: Karawala Región Autónoma Atlántico Sur and Centro de Investigaciones y Documentación de la Costa Altántica. Bravo Ahuja, G. (1977). La enseñanza del español a los indígenas mexicanos. México: El Colegio de México. Brinton, D. G. (1884). “A grammar of the Cakchiquel language of Guatemala”, Proceedings of the American Philosophical Society, 21: 345–412. Brown, R. M. (1996). “The Maya language loyalty movement in Guatemala”, in E. F. Fischer and R. M. Brown (eds.), Maya Cultural Activism in Guatemala. Austin: University of Texas Press, 165–77. Campbell, L. (1976). “The Last Lenca”, IJAL 42: 73–8. —— (1985). The Pipil Language of El Salvador. Berlin: Mouton. —— and Muntzel, M. C. (1989). “The structural consequences of language death”, in N. C. Dorian (ed.), Investigating Obsolescence: Studies in Language Contraction and Death. Cambridge: Cambridge University Press, 181–96. —— Ventur, P., Stewart, R., and Gardner, B. (1978). Bibliography of Mayan Languages and Linguistics. Albany: State University of New York Press. Comisión de Oficialización de los Idiomas Indígenas de Guatemala (1999). Oficialización de los idiomas indígenas de Guatemala: propuesta de modalidad. Guatemala: Proyecto Q’anil B. Constenla Umaña, A. (1998). Gramática de la lengua guatusa. San José: Universidad Nacional. —— Elizondo Figueroa, F., and Pereira Mora, F. (1998). Curso básico de bribri. San José: Universidad de Costa Rica. Contreras García, I. (1985). Bibliografía sobre la castellanización de los grupos indígenas de la República Mexicana (siglos XVI al XX). México: Universidad Nacional Autónoma de México. Craig, C. (1992). “A constitutional response to language endangerment: the case of Nicaragua”, Language, 68: 17–24. England, N. C. (1995). “Linguistics and indigenous American languages: Mayan examples”, Journal of Latin American Anthropology, 1/1: 122–49. —— (1996). “The role of language standardization in revitalization”, in E. F. Fischer and R. M. Brown (eds.), Maya Cultural Activism in Guatemala. Austin: University of Texas Press, 178–94.


181

—— (1998). “Mayan efforts toward language preservation”, in L. A. Grenoble and L. J. Whaley (eds.), Endangered Languages: Language Loss and Community Response. Cambridge: Cambridge University Press, 99–116. Favre, H. (1996). L’Indigénisme. Que sais-je? 3088. Paris: Presses Universitaires de France. García Segura, G., and Zúñiga Muñoz, Z. (1987). “Costa Rica: acciones educativas para la revitalización lingüística”, América Indígena, 47/3: 489–517. Garza Cuarón, B., and Lastra, Y. (1991). “Endangered languages in Mexico”, in R. H. Robins and E. M. Uhlenbeck (eds.), Endangered Languages. Oxford: Berg, 93–134. Hadel, R. E. (1975). A Dictionary of Central America Carib. 3 vols. Belmopan: BISRA. Hofling, C. A. (1991). Itza Maya Texts with a Grammatical Overview. Salt Lake City: University of Utah Press. —— (2000). Itzaj Maya Grammar. Salt Lake City: University of Utah Press. —— and Fernando Tesucún, F. (1997). Itzaj Maya Spanish English Dictionary. Salt Lake City: University of Utah Press. Horcasitas de Barros, M. L., and María Crespo, A. (1979). Hablantes de lengua indígena en México. México: Instituto Nacional de Antropología e Historia. INEGI (1992). Estados Unidos Mexicanos: resumen general, XI censo general de población y vivienda, 1990. México: Instituto Nacional de Estadística, Geografía e Informática. —— (2001) Población hablante de lenguas indígenas. www.inegi.gob.mx/est/default. asp?c=2397. Instituto Nacional de Cultura, and Instituto Lingüístico de Verano (1974–80). Lenguas de Panama. 8 vols. Panama. Instituto Nacional de Estadística (1996). X censo nacional de población y V de habitación. Guatemala. Kaufman, T. (1970). Proyecto de alfabetos y ortografías para escribir las lenguas mayances. Guatemala: José de Pineda Ibarra. Laughlin, R. M. (1975). The Great Tzotzil Dictionary of San Lorenzo Zinacantán. Washington: Smithsonian Institution. Lehmann, W. (1920). Zentral-Amerika. Berlin: Dietrich Reiner. Ligorred, F. (1992). Lenguas indígenas de México y Centroamérica. Madrid: Mapfre. McQuown, N. A. (1955). “The indigenous languages of Latin America”, American Anthropologist, 57: 501–70. Manrique Cantañeda, L. (1996). “Lenguas mexicanas en peligro de extinctión”, in Z. Estrada Fernández, M. Figueroa Esteva, and G. López Cruz (eds.), III encuentro de lingüística en el noroeste, i: Lenguas indígenas, vol. i. Sonora: Editorial Unison, 13–38. —— (1997). “Clasificaciones de las lenguas indígenas de México y sus resultados en el censo de 1990”, in B. Garza Cuarón (ed.), Políticas lingüísticas en México. México: Centro de Investigaciones Interdisciplinarias en Ciencia y Humanidades, UNAM, 39–65. Manrique, Leonardo, Lastra, Y., and Bartholomew, D. (eds.) (1994–5). Panorama de los estudios de las lenguas indígenas de México, 2 vols. Quito: Ediciones Abya-Yala. Margery, E. (1989). Diccionario cabécar–español, español–cabécar. San José: Universidad de Costa Rica. Marx, W. G., and Heath, G. R. (1992). Diccionario: miskito–español, español–miskito. Tegucigalpa: The Moravian Church.

182


Pickett, V., and Embry, V. (1975). Zapoteco del Istmo. Oaxaca: Instituto de Investigación e Integración Social del Estado de Oaxaca. Rigby, N., and Schneider, R. (1989). Dictionary of the Rama Language. Berlin: Dietrich Reimer. Rigoberta Menchú, I. (1987). Rigoberta Menchú: An Indian Woman in Guatemala. London: Verso. Suazo, S. (1990). Conversemos en garífuna: gramática y manual de conversación. Tegucigalpa: COPRODEIM. Tay Coyoy, A. (ed.) (1996). Análisis de situación de la educación maya en Guatemala. Guatemala: Cholsamaj. Turpana, A. (1987). “Panamá: lenguas indígenas”, América Indígena, 47/4: 615–25. Tzian, L. (1994). Kajlab’aliil maya’iib’ xuq mu’siib’: ri ub’antajiik iximuleew—Mayas y ladinos en cifras: el caso de Guatemala. Guatemala: Cholsamaj. Warren, K. B. (1997). “The indigenous role in Guatemalan peace”, Cultural Survival Quarterly, 21/2: 24–7. Yasugi, Y. (1995). Native Middle American Languages: An Areal-Typological Perspective. Osaka: National Museum of Ethnology. —— (2001). “Endangerment in Mayan languages”, in N. Shibata and T. Shionoya (eds.), Languages of the South Pacific Rim, vol. i. ELPR A1–001, 191–200. —— (2003). Materiales de lenguas mayas de Guatemala. 4 vols. ELPR A1–007abcd.

10

Languages of the Pacific Coast of South America Alexandra Y. Aikhenvald

10.1. Linguistic diversity in South America 10.2. Languages of the Pacific coast of South America: past and present 10.2.1. Choco family 10.2.2. Barbacoan family 10.2.3. Extinct languages 10.2.4. Quechua and Aymara 10.3. Typological features of the languages of the Pacific coast 10.3.1. Amazonian versus Andine linguistic types 10.3.2. Pacific coast languages 10.3.2.1. Phonology 10.3.2.2. Morphological profile 10.3.2.3. Evidentiality systems 10.3.2.4. Noun categorization devices and person marking 10.4. Summary

183 185 185 188 189 191 192 192 194 194 195 196 197 200

10.1. Linguistic Diversity in South America The degree of linguistic diversity in South America is comparable only to that in New Guinea (see Ch. 13, this volume). This genetic diversity is reflected in the sheer number of language families and isolates, many of which are now extinct or on the verge of extinction. In his classification, Loukotka (1968) established 117 languages families besides a large number of unclassified isolates, while Tovar and Tovar (1984) postulated over 170 groups. Many of the linguistic families are discontinuous. For instance, Arawak languages are spoken in over ten areas north of the river Amazon, and over ten to the south. Even though there is no doubt as to the genetic relationship between Arawak languages, the geographic diversity, and the extensive and prolonged language contact with genetically unrelated groups, have by and large created a situation whereby languages became restructured to various extents, and their subgrouping within the family—using conventional My deepest gratitude goes to R. M. W. Dixon, Willem Adelaar, Daniel Aguirre Licht, Willem de Reuse, Perihan Avdi, and Anthony Grant, for their insightful comments. Special thanks are due to Daniel Aguirre Licht for facilitating the book exchange between the Research Centre for Linguistic Typology and Centro Colombiano de Estudios de Lenguas Aborígenes.

184

a l e x a n d r a y. a i k h e n va l d

historical-comparative methodology—is problematic (see Aikhenvald 2001). The impact of long-term areal diffusion often makes it next to impossible to unequivocally state whether languages are related or not. This makes any attempts at establishing long-distance genetic relationships especially dubious.1 Since Columbus’s first explorations of Venezuela’s coast in 1498, the numbers of languages and of their speakers have been drastically reduced (see Adelaar 1991, 2004; Rodrigues 1986; Dixon and Aikhenvald 1999, and references therein). Many languages have disappeared without leaving more than a name, or perhaps a few place names. This is the case for Taino, the language of the first native American peoples encountered by Columbus (in 1492) in the Bahamas, Hispaniola, and Puerto Rico, who became extinct within the first hundred years of invasion (Rouse 1992; Hill and Santos-Granero 2002; Payne 1991); its Arawak affiliation is based on a mere handful of lexical comparisons. Other Arawak languages suffered a similar fate. Just under twenty words were recorded from Caquetio, once spoken on two islands near the Venezuelan coast, extinct since the mid-sixteenth century (Loukotka 1968: 128; Oliver 1989: 54–5); just fifteen words are known from Shebayo, once spoken on the island of Trinidad. Our ideas concerning the genetic affiliation of these and many other languages are bound to remain mere hypotheses. With its large number of discontinuous families and isolates, the linguistic map of indigenous South America resembles a patchwork quilt. The Amazon basin comprises over three hundred languages which include about fifteen well-established families as well as numerous isolates. (The six major linguistic families of the Amazon basin are Arawak, Tupí, Carib, Panoan, Tucanoan, and Macro-Jê; smaller families include Makú, Guahibo, Yanomami, BoraWitoto, Harakmbet, Arawá, Tacana, and Chapacura. Macro-groupings or “stocks” suggested by Greenberg 1987, Kaufman 1994, and others are almost without exception illusory and otiose.) The degree of genetic diversity on the continent is, however, not even. The linguistic map of the coastal areas of the Pacific, on the one hand, and the Atlantic, on the other, look impoverished. No indigenous languages survived on the east coast of Brazil (hence the term “linguistic wilderness” 1 The so-called “Arawakan” phylum is a prime example of a dubious grouping of demonstrably unrelated languages which includes Arawak proper (also known as Maipuran), Arawá, Chapacura, Harakmbet, and a few more (see Aikhenvald 1999; Payne 1991). We urge scholars who wish to adhere to proper comparative procedures in South American linguistics to employ the term Arawak to refer to a well-established genetic family.

t h e pac i f i c c oa s t o f s o u t h a m e r i c a

185

coined by Rodrigues 1986), as a result of the rapid European invasion and mass extinction of indigenous population. The relatively poor degree of linguistic diversity of the west coast compared to the neighbouring area of Amazonia—including the eastern slopes of the Andes—is due to a combination of factors. Apart from the effects of the European invasion—which resulted in language extinction and continuous language endangerment— the arid coastal area was not conducive to any large-scale migrations of population. That is, even prior to the invasion, the degree of genetic diversity in the Pacific coastal area appears to have been lower than that on the eastern slopes of the Andes. In § 10.2, I discuss the known languages in the Pacific coast area of Colombia, Ecuador, and Peru, their documentation and endangerment (languages spoken further south are discussed in Ch. 11, this volume). The typological diversity of South American languages is also noteworthy. Unusual, often unique, features include the obligatory expression of information source known as evidentiality, a variety of classifier types, and fascinating patterns of marking grammatical relations. How the extant languages of the Pacific coast compare with the two broad linguistic types recognized for South America—Lowland Amazonian and Andine—is addressed in § 10.3. The last section, § 10.4, contains a brief summary.

10.2. Languages of the Pacific Coast of South America: Past and Present The two major language families extending from inland areas to the Pacific coast of Colombia and Ecuador are Choco and Barbacoan (see Figure 10.1).2

10.2.1. Choco Family Languages of the Choco family are spoken by over 60,000 people in the northern parts of the Department of Choco, Colombia, on the tributaries of the Atrato river and the rivers flowing directly into the Pacific ocean, and along the Pacific coast from the Darien jungle in Panama to the southern Department of Nariño. Recently, the Choco-speaking peoples have expanded 2 The Kuna people speak a Chibchan language known as Kuna, Cuna, or Tule (see Aguirre Licht 1993: 311; Mejía Fonnegra 2000a: 56). The majority of the Kuna (about 43,000 people) live in Panama. About 700 live in Colombia, in the area of the Gulf of Uraba on the Atlantic Coast. In the Colombian linguistic tradition, the Cuna-Tule language is geographically grouped together with the languages of the Pacific coast (see Mejía Fonnegra 2000a; Llerena Villalobos 2000).

186


Figure 10.1 Languages of the Pacific Coast of South America

eastwards, into the departments of Antioquia, Córdoba, and Risaralda in Colombia. The family consists of Waunana3 on the one hand and the Embera dialect continuum with five distinct dialect areas, on the other (also see Adelaar 2004: 56–61). The latter fall into two subgroups. Southern Embera dialects include Chamí or Embera (in Alto San Juan), the epera, or Saija (on Costa Sur Buenaventura), and the epera or Baudó (in Bajo Baudó). Northern Embera dialects are Embera-Katío (north of Antioquia and Córdoba) and the Embera Proper (in Atrato, Alto Baudó, and in Panamá) (Pardo Rojas and Aguirre 1993; Aguirre, p.c.).4 The Southern Embera varieties can be viewed 3 Also known as Waunán, Wounaan (autodenomination: literally, “people”), Waunméu, or Waun Meu “people’s language” (Mortensen 1999) and Noanamá (term employed by other groups) (Mejía Fonnegra 2000b: 85). 4 Llerena Villalobos (1995b) divides the Choco family into Waunana, Eastern and Western Embera e˜pe˜ra. A somewhat different classification is given by Mortensen (1999). Incomplete and not always accurate lists of Embera languages and dialects (erroneously listed under Emberá) appear in Campbell (1997: 172–3) and Migliazza and Campbell (1988: 183).


187

as a continuum of dialects spoken on San Jorge, Verde, and Sucio rivers (Harms 1994). Most Waunana live in Colombia, with a few settlements extending into the Pacific coast of Panama. The current estimate for the Waunana in Colombia is of about 4,000 people. The Embera languages and Waunana are not mutually intelligible (according to Mejía Fonnegra 2000b: 85, they share about 50 per cent cognates); while most Embera varieties are (Pardo Rojas 1997: 337). The present consensus is that Choco languages form an independent family which cannot be demonstrated to be related to any other family (see Costenla-Umaña and Margery-Peña 1991: 139–40; Aguirre Licht 1999a). Various attempts at grouping Choco languages with Chibchan or Chibchan-Paezan (Greenberg 1987: 106–22; Ruhlen 1987; repeated in Migliazza and Campbell 1988: 183) and Carib (Loewen 1963a: 244) have not been substantiated. The first attempts at documenting the Choco languages go back to the diary of Father Joseph Palacios de la Vega, the head of the mission at San Cipriano on the San Jorge river around 1787. A brief analysis of this and other sources and a detailed bibliography were published by Loewen (1963a), who also completed a grammar of the Sambú dialect of Embera (1958). The Embera languages are now quite well documented. Recently, three grammars of Embera languages have been produced: Harms (1994) of Embera Saija; Mortensen (1999) of two Northern Embera varieties; and Aguirre Licht (1998, 1999a) of Embera Chamí. The bulk of recent research has been undertaken under the auspices of Centro Colombiano de Estudios de Lenguas Aborígenes (Universidad de los Andes, Bogotá). A useful overview of the sociolinguistic situation of the Choco languages is in Pardo Rojas (1997) and Mejía Fonnegra (2000a, 2000b).5 A major study of the phonology of Embera languages is in Llerena Villalobos (1995a) which contains both synchronic and comparative descriptions. Apart from a short description by Loewen (1954), there is no comprehensive grammar of Waunana (see Mejía Fonnegra 1995, 2000b). Materials include papers by Binder (1977), and Binder and Binder (1974), based on varieties spoken in Panama, and work by Mejía Fonnegra, based on Colombian varieties. The most up-to-date summary of Waunana grammar is Mejía Fonnegra (2000b), which also contains a history of studies of the language.

5 Landaburu (1999: vol. iv) presents a variety of 19th- and early 20th-century sources on Embera languages (including a large anonymous grammar of Katío).

188


Choco languages are not in danger of immediate disappearance. All the languages are still being learnt by children. According to Pardo Rojas (1997), most Choco-speaking peoples maintain their traditional lifestyle and their languages. However, the lexical knowledge of people under 70 appears somewhat impoverished, with a large number of Spanish loans; younger people tend to simplify the case and number marking systems. Among the Waunana, the only monolinguals are children under 7 and elderly women. It appears that, with the expansion of Spanish-based schools, this is rapidly changing (Mejía Fonnegra 2000b: 85). With increasing bilingualism in Spanish, the language is under threat. It is, however, a positive sign that cultural and language workshops continue to be organized by linguists and anthropologists (under the auspices of Universidad de Los Andes: see, for instance, Mejía Fonnegra’s 1995 discussion of a successful language and culture workshop conducted among the Waunana).

10.2.2. Barbacoan Family The Barbacoan languages are spoken, by and large, in the Pacific region of Ecuador and the adjacent area of southern Colombia (mostly the Municipio of Ricaurte). Guambiano and Totoró are spoken in the Andes in the department of Cauca, Colombia. The family consists of five languages divided into two major groups (Curnow and Liddicoat 1998; Curnow 1998; see Adelaar 2004: 392–4 for a broader approach): North Barbacoan consisting of Guambiano and Totoró, on the one hand, and Awa Pit (Cuaiquier), on the other hand; and South Barbacoan formed by Cha’palaachi (Chachi, Cayapa) and Tsafiki (or Colorado). See Figure 10.1. The earliest classification of Barbacoan languages is by Brinton (1891: 194–9); for an updated history of classification and language studies see Curnow (2002b, 1998).6 Cha’palaachi is spoken by about 3,000–5,000 people (Stark 1985: 162) in Esmeraldas province, in the north-west area of Ecuador. Most are monolingual (with only 20 per cent of the population bilingual in Spanish: Stark 1985: 162). Tsafiki is spoken by about 1,000–1,800 people (Grimes 2000; Stark 1985: 160) in the western lowlands of Ecuador. In the 1970s, about 50 per cent of the Tsafiki (who call themselves the Tsachi people) were bilingual in Spanish. A partial grammar of Tsafiki is in Dickinson (2002). At present, Tsafiki and Cha’palaachi appear to be relatively healthy. However, in

6 The lists of Barbacoan languages given by Campbell (1997: 174) and Migliazza and Campbell (1988: 182) are incomplete: Guambiano-Totoró and Cha’palaachi have been omitted.


189

view of their small numbers and the current political situation in Ecuador, they are threatened. Estimates for Guambiano vary from 8,000–9,000 (Adelaar 1991: 66) to 18,000 (Huber and Reed 1992: p. xiv), while Totoró is spoken by just a few people (Pabon Triana 1995b) and is now close to extinction.7 While Guambiano appears to be relatively healthy, Totoró is endangered. Pabón Triana (1995b) reports a high degree of bilingualism, negative attitudes towards the language, and a high degree of language loss. Efforts are being directed towards cultural and linguistic revival of Totoró. The estimated number of Awa Pit (Cuaiquer) in Colombia is around 20,000, and in Ecuador 1,000 (Tim Curnow, p.c.). Grammatical descriptions are available for Awa Pit (Calvache Dueñas 1989; Obando Ordóñez 1992; Curnow 1997, the latter based on work with obsolescent speakers in Colombia). Awa Pit is endangered, with the majority of people monolingual in Spanish and no evidence in favour of any monolingualism in Awa Pit (Curnow 1997: 18–20; also see Mejía Fonnegra 2000a: 57 on the high degree of acculturation among the Awa Pit in Colombia). Attempts to group Barbacoan languages with either Chibcha or Paez have proved unsatisfactory (see discussion in Curnow 1997, 1998; Landaburu 1993; Curnow and Liddicoat 1998, pace Campbell 1997: 173).

10.2.3. Extinct Languages Little is known about most of the extinct languages in the coastal area. The most important language on the Pacific coast before the Inca conquest was Mochica, or Yunga,8 spoken along the north-east coast of contemporary Peru. On the basis of toponyms, some scholars believe that Mochica extended southwards as far as northern Chile and Argentina (see discussion by Stark 1972: 119). Quingnam (also known as Chimú, or Yunga) was spoken around the present-day city of Trujillo, and possibly as far south as Lima. Hardly anything is known about this language, apart from a few place-names. Quingnam-Chimú was the major language of the Chimú culture, a major pre-Incan civilization (known for its distinctive pottery) (also see Harrington 1945). The Chimú capital Chan Chan was located in the Moche 7 According to Rivet (1941: 3), the number of the Guambiano in the 1930s was 5,623, while Totoró was spoken by about 1,200 people. According to him, the two were mutually intelligible with each other and with a third dialect, Polindara (then spoken by 695 people). 8 The name Yunka or Yunga is of Quechua origin; in Quechua, yunka means “jungle” (Harrington 1945: 24).

190


valley near the modern city of Trujillo, about 500 km north of Lima (hence the confusion of Quingnam-Chimú and Mochica). The Chimu empire state is believed to have begun to take shape in the first half of the fourteenth century. The major expansion of the state occurred under Ñancen Pinco (whose reign began in 1370). The Chimú were conquered by the Incas in 1465–70. They passed their social stratification, as well as technological achievements such as irrigation and road engineering, on to the Inca. The Mochica language was originally spoken in Eten and Monsefú near Chiclayo, up until the 1940s (when Huber collected a word list published in 1953). A few speakers of the Eten variety survived until the 1960s (Stark 1972; also mentioned by Loukotka 1968: 261). Two traditional grammars of this language are Carrera (1644),9 and Middendorf (1892), who complements Carrera. Both grammars describe the variety of Eten.10 Also see discussion in Adelaar (2004: 319–50). Other extinct languages on the Pacific coast of north-eastern Ecuador include Sechura and Tallan or Atalán (with dialects Colán and Catacaos). According to Tovar and Tovar (1984: 170), at least the dialect of Atalán was spoken until the end of the nineteenth century. The materials for these languages are limited to short word lists, insufficient for postulating any linguistic relationships, other than striking similarities between Sechura and Atalán. Schmidt (1926: 214) attempted to put together Mochica-Yunga and what he called Tallan (subsuming Sechura, Colán, and Catacaos). This relationship remains purely conjectural (just like Mason’s 1950: 195–6 hypothetical Yunga-Puruhã family consisting of Yunga, Puruhã-Cañari, and Atalán). The question of the wider genetic relationships of Yunga remains open. Various scholars have tried to link it with Chibcha, and with Barbacoan, without, however, much success. Stark (1972) detected typological similarities between Yunga, Maya, and Uru-Chipaya. The few lexical and morphological look-alikes she listed are mainly monosyllabic and are likely to be coincidental. There may have been more extinct languages in the Pacific coastal area. A short vocabulary of Yurumangui,11 once spoken on Yurumangui River in the Department of Valle de Cauca, was collected by Father Christoval Romero in 9 Carrera was born in the area of Trujillo and may have been a native speaker of the language (Stark 1972). 10 According to Loukotka (1968: 261–2), the Chimu family consisted of two subgroups: Southern (including Chimú, Eten, and Mochica) and Northern (including Puruhá, Cañari, Huancavilca, of which only four words are known, and Manabita or Manta, with only a few patronyms surviving). 11 The alternative term Yurimangui was erroneously used by Loukotka 1968, and then repeated by Kaufman (1994).


191

1768 (see Ortiz 1946, on other sources for this language). On the basis of it, Rivet (1942) attempted to establish a genetic relationship between Yurumangui and the putative Hokan group. However, the materials are too scanty for any positive statement. Loukotka (1968: 259–60) groups together Yurumangui with a number of languages in the same area (Timba, Lili, Yolo, Jamundi, and Puscajae), on which there is no information at all. Extinct languages on the Ecuadorian coast include Esmeralda, with some materials put together by Seler (1902), and Caraque, on which we know nothing (see Loukotka 1968: 133–4; Tovar and Tovar 1984: 185; Adelaar 2004).

10.2.4. Quechua and Aymara The Andean area is dominated by Quechua and Aymara. Quechua, with over seven million speakers, is spoken in the Andean highlands of Ecuador, Peru, and Bolivia (Hornberger and King 2001; Grimes 2000; Adelaar 1991; Cerrón-Palomino 1987); while Aymara, with over two million speakers, is spoken around Lake Titicaca and from Lake Titicaca towards the Pacific (Grimes 2000). The question of whether Quechua and Aymara are related or not remains open (see Campbell 1995; Adelaar 2004: 168–315). The Quechua dialects spoken on the Pacific coast included varieties known as Quechua IIb, extinct since the seventeenth century (Adelaar 1992, 2004). Quechua dialects I come close to the Pacific coast in the areas of Cajambo and Yauyos (Willem Adelaar, p.c.). Quechua is a well-documented language (see Adelaar 1991, 1992, 2004; Campbell 1997; Cerrón-Palomino 1987, for an overview of Quechua studies which go back over 400 years). It was legally recognized as an official language of Peru under Velasco Alvarado’s presidency in the 1970s. In its turn, Ecuador ‘‘saw the rise of a powerful national political organization in the early 1980s’’ (Hornberger and King 2001: 190, and further discussion in Hornberger 2000). There is nevertheless little doubt that with the encroaching dominance of Spanish, and the gradual spread of bilingualism, the language is under threat. As Hornberger and King (2001: 167) put it, “the Quechua language and Quechua speakers generally remain powerless and marginalised within their national contexts. Quechua continues to be strongly linked with the rural, uneducated and poor, while Spanish remains the primary language of national and international communication, literacy and education, and professional and academic success.” King’s (1999) investigation of the status of Quechua in the Ecuadorian Andes revealed patterns of slow but steady decline of Quechua: even remote and isolated Quechua-speaking

192


communities in highland Saraguro, Ecuador, can no longer be considered bilingual. Quechua is still used at certain traditional community events and as a means of communication between elders, but “Spanish has made in-roads into seemingly every speech situation” (King 1999: 25). The hope for the future of the language—so far uncertain, given the extent of cultural, sociolinguistic, and geographic differentiation of Quechua varieties—lies perhaps in the establishment of regional international institutions, like the Andean Program in Bilingual Intercultural Education (Programa de Formación en Educación Bilingüe para los Paises Andinos, PROEIB-Andes), a Master’s programme sponsored by a five-nation consortium with coordination between Ministries of Education, universities, and indigenous organizations spanning Bolivia, Chile, Colombia, Ecuador, and Peru, established at the University of San Simón, Cochabamba, Bolivia (Hornberger and King 2001: 188–9). The maintenance of Quechua is being enhanced by its growing role in the media (Luykx 2001).

10.3. Typological Features of the Languages of the Pacific Coast 10.3.1. Amazonian versus Andine Linguistic Types Two broad linguistic types have been recognized in South America: Amazonian and Andine. Amazonia can be recognized as a linguistic area in terms of a number of features which are shared by all (or most) languages in the area.12 A comparison between the typological characteristics of the Amazonian linguistic area in Lowland South America with those of the Andean linguistic area in the adjacent mountains, which comprises the Quechua and Aymara families, shows that these Andean languages are clearly different. A few representative features of Lowland Amazonian languages are summarized in the first column of Table 10.1, with Andine features in the second column. Other features, exceptions to pan-Amazonian tendencies, and regional areal characteristics are discussed in Dixon and Aikhenvald (1999). There is no sharp boundary between the Andean and Amazonian linguistic areas—they tend to flow into each other. (For instance, Andean features such as lack of prefixes and an accusative technique for marking 12 Whether Quechua and Aymara belong to the same linguistic type as other languages spoken outside the Amazonian Lowlands (such as Mapuche, Leko, Cholón and Uru-Chipaya) is an open question.

Table 10.1. Lowland Amazonian and Andean languages: a comparison Lowland Amazonian

Andean

(a) The majority of languages are polysynthetic and head marking; agglutinating with little fusion.

Andean languages are synthetic, and combine head and dependent marking; basically agglutinating with some fusion (subject, object, and

tense suffixes to the verb may be fused). (b) Typically one liquid phoneme, which is frequently a flap; usually more affricates than fricatives. The high unrounded central vowel is frequent. A typical Amazonian vowel system has five members: i, e, a, , u/o. There is typically contrastive nasalization of vowels.

Two or three liquids; fricatives rather than affricates; and a three-vowel system i, a, and u, with no contrastive nasalization.

(c) Many languages have extensive classifier and/or gender systems. Gender assignment is often semantically transparent, and is not overtly marked on the head noun.

No genders or classifiers.

(d) There are few oblique cases (often a locative and an instrumental/comitative), but hardly any core cases.

Extensive set of core and oblique case markers.

(e) Just one core argument is typically cross-referenced on the verb. There may be different bound pronominal paradigms depending on which core argument is being cross-referenced in each particular instance.

Two core arguments are marked on the verb.

(f) The rules for which core argument is cross-referenced can be complex (relating to the meaning of the verb, clause type, etc.) often giving rise to a “split-ergative” system. Fully accusative systems of marking for predicate arguments are rarely encountered.

Fully nominative/accusative systems.

(g) Most (although not all) languages have prefixes; there are typically fewer prefix than suffix positions.

No prefixes.

(h) There is generally only a small class of lexical numbers.

Full set of lexical numbers

194


syntactic function are found in languages of the Tucanoan family, which are in Amazonia but fairly close to the Andes.)

10.3.2. Pacific Coast Languages A brief look at the language families of the Pacific coast shows that, typologically, they differ from both Amazonian and Andean profiles. Most languages have rich consonantal and vocalic systems, and no tones; a few have contrastive vowel nasalization. Mochica was mostly dependent marking with a high degree of synthesis. Choco and Barbacoan languages are synthetic with head and dependent-marking properties. All are agglutinating.

10.3.2.1. Phonology The phonology of Choco languages is very complex: there is typically a series of three stops (voiceless aspirated, voiceless unaspirated, and voiced lax, as in Waunana and Saija), or voiceless plain (with elements of aspiration), voiced tense, and voiced implosive, as in the northen dialects of Embera), besides two affricates, and two vibrants (a trill and an alveolar sonorant). Embera languages have six vowels (i, e, ɯ, a, u, o); Waunana also has an open back vowel γ. Every vowel has a nasal counterpart (some authors consider nasalization a prosodic feature of the word or of a larger unit: Aguirre Licht 1999a: 20; Adelaar 2004: 56–61). Stress is not distinctive. Barbacoan languages typically have five or four vowels (cf. Moore 1962 and papers in Elson 1962; and Curnow and Liddicoat 1998). Awa Pit is unusual in distinguishing voiced and voiceless vowels, while Cha’palaachi appears to have phonemic vowel length and Tsafiki has phonemic nasal vowels. The consonantal system contains voiced and voiceless stops, up to four fricatives (f, s, h, ʃ), and one affricate, ts (see Curnow and Liddicoat 1998: 401, for a summary of reconstructed phonemes). Mochica had complicated phonology, with a six-vowel system reminiscent of the “Amazonian” type (i, u, e, o, a, and central or ə) with phonemic long vowels, a series of seven fricatives (labial, dental, alveopalatal, palatal, and velar, and two lateral), three affricates (dental, alveopalatal, and palatal), distinctive dental and alveopalatal nasals, laterals and vibrants, and a velar nasal (Stark 1972: 120; a somewhat different analysis of Carrera and Middendorf is given by Tovar and Tovar 1984: 168–9; also see Adelaar 2004: 321–8). It appears that the language had contrastive stress.


195

10.3.2.2. Morphological Profile Choco languages are exclusively suffixing, similar to Andean languages, while some Barbacoan languages have suffixes and prefixes (as apparently did Yurumangui: Rivet 1942). Mochica was suffixing, with a few proclitics sometimes analysed as prefixes (e.g. Harrington 1945). Unlike Quechua, all Choco languages are morphologically ergative, with nominative-accusative features in cross-referencing. The ergative case (Northern Embera -pa, Chamí -ba, Waunana -au/ua) marks the subject of a transitive verb (A), and also instrument (in all languages), reason, and cause in Northern Embera, Epena Pedee, and Chamí (Mortensen 1999: 47–9; Aguirre Licht 1999a: 88–90; Harms 1994: 65), and ablative in Northern Embera. The emphatic ergative case marker cannot occur together with the ergative case on nouns. In contrast, it can be added to the ergative marker on personal pronouns. Absolutive case marking is fourfold, depending on the status of the noun referent in discourse. If the noun referent is non-activated, there is zero marking. Special suffixes mark an S/O participant which is non-focal, in introductory focus, or as given information. There is no participant cross-referencing on the verb except for plural marking, which indicates the number of S/A. Unlike Embera languages, Waunana has subject cross-referencing on the verb (on a nominative-accusative principle: Mejía Fonnegra 2000b: 91), with different sets for present and non-present tenses. Singular masculine suffixes distinguish first person versus non-first person forms, while feminine and plural do not distinguish person. This unusual first/non-first person marking in Waunana is reminiscent of some Tucanoan languages in Amazonia (see Aikhenvald 2002). Other nominal cases include dative, allative, locative, benefactive, prolative, and associative. Chamí, Epena Pedee, and Waunana (Mejía Fonnegra 2000b: 89–90; Aguirre Licht 1999a: 91) have a special ablative marker (formally distinct from ergative). In Chamí, this can occur together with locative in one word (also see Harms 1994: 76). Waunana has just dative/allative, associative, ablative, locative, and genitive/partitive. In Northern Embera (Mortensen 1999: 51), the locative in combination with a personal pronoun marks possession (e.g. ãči-de “they-LOC” “theirs, their own”), and the resulting combination may take markers of syntactic function, a pattern similar to that discussed by Dixon (2002: 147–52) for Australian languages, but hitherto unattested in South America. Barbacoan languages are predominantly suffixing (a few prefixes are found in Cha’palaachi and Tsafiki). All are nominative-accusative. In Awa Pit, Cha’palaachi, and Tsafiki the accusative and the locative are marked

196


with the same morpheme (which is reminiscent of the polysemy found for Quechua -ta “accusative; locative/goal”: Adelaar 1992). Nominative-accusative marking is also shared by Mochica, which had pronominal cross-referencing on the verb, on a nominative-accusative basis (with suffixes cross-referencing subject). It also had a genitive and an instrumental case, alongside a rich variety of verbal derivations including a causative and several passives.

10.3.2.3. Evidentiality Systems Evidentiality systems with four terms in some Barbacoan languages and in Northern Embera are richer than three-term systems in Quechua, and are reminiscent of multiple evidentials in Amazonia (see Aikhenvald and Dixon 1998). Consider the four-way evidentiality distinction in Tsafiki (Dickinson 2000: 407–9). If an event was “directly” witnessed, the verb is morphologically unmarked, as in (1). Evidentials are in bold. (1)

Manuel ano fi-e Manuel food eat-visual+declarative “Manuel ate” (the speaker saw him)

If information was obtained by inference from direct physical evidence, (2) would be used. (2)

Manuel ano fi-nu-e Manuel food eat-inferred-declarative “Manuel ate” (the speaker sees the dirty dishes)

A nominalization followed by the verb class marker is employed if the inference is made on the basis of general knowledge and assumption. (3)

Manuel ano fi-n-ki-e Manuel food eat-nom-verb.class:do-declarative13 “Manuel ate” (he always eats at eight o’clock and it’s now nine o’clock)

The reported evidential—marked with the suffix -ti—indicates that the information was obtained from someone else. In (4), the reported speech specifies the source of information the “reporter” had: the report was based on inference from physical evidence. Typologically this is highly unusual (Aikhenvald 2004). 13 ACC - accusative; FEM - feminine; INTER - interrogative; MASC - masculine; NEG - negation; NOM - nominalizer; PL - plural.

t h e pac i f i c c oa s t o f s o u t h a m e r i c a (4)

Manuel Manuel

197

ano food

fi-nu-ti-e eat-inference.physical. evidence-hearsay-declarative “He said/they say Manuel has eaten (they didn’t see him, but they have direct physical evidence, e.g. dirty dishes)” The reported evidential can be repeated several times, to indicate up to three sources “between the speaker and the original event”, another unusual feature of Tsafiki. There are two repetitions in: (5)

tsachi-=la person-PL

jo-la-jo-ti-e ti-e be-pl-verb.class.be-reported- say-declarative declarative “They say he said they were people”

Guambiano (Vásquez de Ruíz 1992; Triviño Garzon 1992, 1994) also appears to have grammatical evidentials; this, however, requires further study. Among Choco languages, Northern Embera varieties distinguish visual, conjectural, quotative, and reported evidentials (Mortensen 1999: 86–7), while Saija (Epena Pedee) appears to have just a reported evidential (Harms 1994: 117, 177).

10.3.2.4. Noun Categorization Devices and Person Marking Most Choco languages have no genders or classifiers. Aguirre Licht (1999a: 74) points out the existence of animate and inanimate forms of the existential verb in Embera-Chamí; nouns can take suffixes with such meanings as “liquid”, “flat surface”, “internal organs” (pp. 75–6), which are reminiscent of noun classifiers. Of the Barbacoan languages, Tsafiki has classifiers used with adjectives (Dickinson, p.c.; Curnow and Liddicoat 1998: 387), while Cha’palaachi has numeral classifiers. Mochica had several numeral classifiers (including one for humans and animals, one for fruit and vegetables, and one for other inanimate objects). These features are only somewhat reminiscent of Amazonian languages. Just like most Amazonian languages, Choco and Barbacoan have only a small class of lexical numbers (up to four or five). In contrast, Mochica has a full set of numbers in a decimal system (Harrington 1945: 29). A cross-linguistically rare salient feature of Barbacoan languages is the conjunct-disjunct person marking system (found in a few Tibeto-Burman languages).14 In these languages statements which contain a first person 14

This is also known as locutor/non-locutor person-marking system.

198


participant are marked differently from those which do not. In questions the second person is marked in the same way as the first person in statements. Consider the following examples from Tsafiki, a Barbacoan language spoken by the Tsachi (Dickinson 2000) which has two construction types. One, shown in (6), appears with first person in statements and indicates “conjunct” marking. A summary of Barbacoan person marking is found in Curnow (2002b). (6)

tse Tsachi 1.fem Tsachi “I am a Tsachi”

jo-yo-e be-conjunct-declarative

The other one occurs in all other contexts and indicates “disjunct” marking: (7)

ya/nu Tsachi 3rd/2nd Tsachi “He/you are a Tsachi”

jo-Ø-e be-disjunct-declarative

In questions, the conjunct marking indicates second person: (8)

nu seke you good “Did you dance well?”

tera dance

ki-yo-n? do-conjunct-inter

Disjunct and conjunct forms contrast in meaning if used with a “wrong” person. When disjunct forms are used with first person in statements, they may indicate that the speaker did something unintentionally, without being in control. In (9), the pig was killed intentionally, and the speaker uses conjunct marking. (9)

la kuchi=ka tote-yo-e 1masc pig=acc kill-conjunct-declarative “I killed the pig” (intentionally)

In (10), disjunct marking is used: the speaker did not mean to kill the pig. (10)

la kuchi=ka tote-i-e 1masc pig=acc kill-disjunct-declarative “I killed the pig” (unintentionally)

Disjunct forms with first person may indicate a speaker’s surprise, that is, they have mirative overtones: (11) is a simple statement of a fact; as expected, conjunct marking is used (Dickinson 2000: 388).

t h e pac i f i c c oa s t o f s o u t h a m e r i c a (11)

199

kala ta-yo-e money have-conjunct-declarative “I have money”

If the speaker suddenly discovers to his surprise he has some money which he did not think he had, the disjunct marker would be used: (12)

kala ta-i-e money have-disjunct-declarative “I have money!” (what a surprise!)

Conjunct markers can in turn be used with third person subject. This implies that the speaker is a knowing participant. Such constructions are used to refer to something the speaker knows first hand: in (13) the conjunct marker indicates that the speaker is a knowledgeable member of the group (despite the fact that the sentence contains a third person subject): (13)

amana now

tsachi=la Tsachi=pl

fi-tu-min=la eat-neg-nom=pl

jo-yo-e be-conjunctdeclarative “Nowadays, we, the Tsachi, do not eat snakes” (lit. the Tsachi do not eat snakes)

Person marking in this and other conjunct/disjunct systems code the degree of congruence of the information with the speaker’s general knowledge, and thus is indirectly connected with the way of obtaining information. The conjunct marking—especially when used with the third person—has an overtone of information integrated into the person’s knowledge. The disjunct marking when used with the first person produces the effect of unintentional action, surprise, or irony. That is, in a conjunct-disjunct system, the choice of person marking acquires an additional connotation related to information source. In summary, even a brief comparison with typical features of Amazonian and Andean languages shows that, in spite of a seemingly low degree of genetic diversity, the languages of the Pacific coast are typologically very diverse. There is hardly any recognizable “Pacific coast of South America” language type. Rather, each language family and individual languages of the Pacific coast exhibit unusual characteristics. This underlines the desirability of their maintenance and urgent documentation.

200


10.4. Summary There is no doubt that most indigenous languages of the Pacific coast are either endangered, like Awa Pit, or threatened. In spite of high numbers of speakers, even Quechua is a threatened language, quickly moving towards becoming endangered. Among the numerous factors which have contributed to this imminent language loss are the lack of juridical and political statuses for the vernacular languages (Triana y Antorveza 1997: 118–19), and the expansion of Spanish via socioeconomic pressures and the official education system. The current situation in Colombia appears to be now changing due to the activities of local Colombian linguists, and especially of the CCELA (Centro Colombiano de Estudios de Lenguas Aborígenes, created in 1988 at the University of Los Andes), strengthened by its international collaboration with CNRS in France (Pineda Camacho 1997: 166–70). An important change in attitude towards the indigenous population was reflected in the 1991 Constitution which now acknowledges and protects “the ethnic and cultural diversity of the Colombian nation” (Art. 7), in contrast to the previous constitution (Pineda Camacho 1997: 158–9; 170–1). This will no doubt enhance the preservation of the remaining linguistic diversity and the unusual languages of the Pacific side of South America. There is now hope for language maintenance and revitalization.

References Adelaar, W. F. H. (1991). “The endangered languages problem: South America”, in R. H. Robins and E. M. Uhlenbeck (eds.), Endangered Languages. Oxford: Berg, 45–91. —— (1992). “Quechuan languages”, in William Bright (ed.), The International Encyclopedia of Linguistics. Oxford: Oxford University Press, iii. 427–32. —— with the collaboration of Muysken, Pieter C. (2004). The Languages of the Andes. Cambridge: Cambridge University Press. Aguirre Licht, D. (1993). “Lenguas vernaculas sobrevivientes”, in Pablo Leyva (ed.), Colombia Pacífico. Bogotá: Fondo para la Protección del Medio Ambiente “José Celestino Mutis”, 310–24. —— (1995). “Fonología del e˜be˜ra-Chami de Cristianía (Departamento de Antioquia)”, in Llerena Villalobos 1995a: 9–86. —— (1998). Fundamentos morfosintácticos para una gramatica embera. Lenguas aborígenes de Colombia. Descripciones. Santa Fé de Bogotá: CCELA, Universidad de los Andes. —— (1999a). Embera. Munich: Lincom Europa.


201

—— (1999b). “Ergatividad, focalización, tematización y topicalización en embera”, in Congresso de lingüística amerindia y criolla. Lenguas aborígines de Colombia, Memorias 6. Santa Fé de Bogotá: CCELA, Universidad de los Andes, 317–30. Aikhenvald, A. Y. (1999). “The Arawak language family”, in R. M. W. Dixon and A. Y. Aikhenvald (eds.), The Amazonian Languages. Cambridge: Cambridge University Press, 65–105. —— (2001). “Areal diffusion, genetic inheritance and problems of subgrouping: a North Arawak case study”, in A. Y. Aikhenvald and R. M. W. Dixon (eds.), Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics. Oxford: Oxford University Press, 167–91. —— (2002). Language Contact in Amazonia. Oxford: Oxford University Press. —— (2004). Evidentiality. Oxford: Oxford University Press. —— and Dixon, R. M. W. (1998). “Evidentials and areal typology: a case study from Amazonia”, Language Sciences, 20: 241–57. Bedoya, O. L. (2001). “Los nombres compuestos en la lengua embera de oriente”, in Diferencias y similtudes en la estructura del léxico de lenguas aborígenes. Lenguas aborígines de Colombia, Memorias 7. Santa Fé de Bogotá: CCELA, Universidad de los Andes, 69–84. Binder, R. (1977). “Thematic linkage in Waunana discourse”, in R. E. Longacre and F. Woods (eds.), Discourse Grammar: Studies in Languages of Colombia, Panama and Ecuador, vol. ii. Dallas: Summer Institute of Linguistics, 159–90. —— and Binder, K. (1974). “Fonología waunana”, in Lenguas de Panamá I. Panamá: Instituto Lingüístico de Verano, 71–92. Brinton, D. G. (1891). The American Race: A Linguistic Classification and Ethnographic Description of the Native Tribes of North and South America. New York: N. D. C. Hodges. Calvache Dueñas, R. (1989). “Fonología e introducción a la morfosintaxis del Awa Pit”, master’s thesis, Universidad de los Andes, Santa Fé de Bogotá. —— (2000). “Fonología y aproximación a la morfosintaxis del awa-pit”, in M. S. González de Pérez and M. L. Rodríguez de Montes (eds.), Lenguas indígenas de Colombia: una visión descriptiva. Santa Fé de Bogotá: Instituto Caro y Cuervo, 97–116. Campbell, L. (1995). “The Quechumaran hypothesis and lessons for distant genetic comparison”, Diachronica, 12: 157–200. —— (1997). American Indian Languages: The Historical Linguistics of Native America. Oxford: Oxford University Press. —— and Muntzel, M. (1989). “The structural consequences of language death”, in N. Dorian (ed.), Investigating Obsolescence: Studies in Language Contraction and Death. Cambridge: Cambridge University Press, 181–96. Carrera, Fernando de la, (1644). Arte de la lengua Yunga de los valles del Obispado de Trujillo del Peru, con un Confessionario y todas las Oraciones Christianas, traducida en la lengua, y otras cosas. Lima. (Other editions: Lima, 1880; Universidad Nacional de Tucumán, No. 3, University Publication No. 256, 1939.) Cerrón-Palomino, R. (1987). Lingüística Quechua. Cuzco: Centro de Estudios Rurales Andinos Bartolomé de las Casas. Costenla Umaña, A., and Margery-Peña, E. (1991). “Elementos de fonología comparada chocó”, Filología y Lingüística, 17/1–2: 137–91. Curnow, T. J. (1997). “A grammar of Awa Pit”, Ph.D. thesis, Canberra: ANU.

202


Curnow, T. J. (1998). “Why Paez is not a Barbacoan language: the nonexistence of “Moguex” and the use of early sources”, International Journal of American Linguistics, 64: 338–51. —— (2002a). “Conjunct/disjunct marking in Awa Pit”, Linguistics, 40: 611–27. —— (2002b). “Conjunct/disjunct systems in Barbacoan languages”, Santa Barbara Papers in Linguistics, 11: 3–12. —— and Liddicoat, A. (1998). “The Barbacoan languages of Colombia and Ecuador”, Anthropological Linguistics, 40: 384–408. Dickinson, C. (2000). “Mirativity in Tsafiki”, Studies in Language, 24: 379–421. —— (2002). “Complex predicates in Tsafiki”, Ph.D. dissertation, University of Oregon. Dixon, R. M. W. (2002). Australian Languages: Their Origin and Development. Cambridge: Cambridge University Press. —— and Aikhenvald, A. Y. (1999). “Introduction”, in R. M. W. Dixon and A. Y. Aikhenvald (eds.), The Amazonian Languages. Cambridge: Cambridge University Press, 1–21. Elson, B. (ed.) (1962). Studies of Ecuadorian Indian Languages, vol. i. Norman, Okla. Summer Institute of Linguistics. Greenberg, J. H. (1987). Language in the Americas. Stanford, Calif.: Stanford University Press. Grimes, B. (ed.) (2000). Ethnologue: Languages of the World, 14th edn. Dallas: Wycliffe Bible Translators. Harms, P. L. (1994). Epena Pedee syntax. Arlington: A publication of SIL International and the University of Texas at Arlington. Harrington, J. P. (1945). “Yunka, language of the Peruvian coastal culture”, International Journal of American Linguistics, 11: 24–30. Hill, J. D., and Santos-Granero, F. (eds.) (2002). Comparative Arawakan Histories: Rethinking Language Family and Culture Area in Amazonia. Urbana: University of Illinois Press. Hornberger, N. H. (2000). “Bilingual education policy and practice in the Andes: ideological paradox and intercultural possibility”, Anthropology and Education Quarterly, 31: 173–201. —— and King, K. A. (2001). “Reversing Quechua language shift in South America”, in J. A. Fishman (ed.), Can Threatened Languages be Saved? Clevedon: Multilingual Matters, 166–94. Hoyos Benítez, M. E. (1995). “Estudios embera en torno a la comparación de algunas variantes”, in Llerena Villalobos 1995a: 87–208. —— (2000). “Informe sobre la lengua embera del río Napipí”, in M. S. González de Pérez and M. L. Rodríguez de Montes (eds.), Lenguas indígenas de Colombia: una visión descriptiva. Santa Fé de Bogotá: Instituto Caro y Cuervo, 73–84. Huber, K. (1953). “Contribution à la langue mučik”, Journal de la Société des Américanistes de Paris, 42: 127–34. Huber, R. Q. and Reed, R. B. (1992). Comparative Vocabulary: Selected Words in Indigenous Languages of Colombia. Santa Fé de Bogotá: Asociación Instituto Lingüístico de Verano. Kaufman, T. (1994). “The native languages of South America”, in C. Moseley and R. E. Asher (eds.), Atlas of the World’s Languages. London: Routledge, 46–76. King, K. A. (1999). “Language revitalisation processes and prospects: Quichua in the Ecuadorian Andes”, Language and Education, 13: 17–37. Landaburu, J. (1993). “Conclusiones del seminario sobre classificación de lenguas indígenas de Colombia”, in M. L. Rodriguez de Montes (ed.), Estado actual de la


203

classificación de las lenguas indígenas de Colombia, Santa Fé de Bogotá: Instituto Caro y Cuervo, 313–30. —— (ed.) (1999). Documentos sobre lenguas aborígenes de Colombia del archivo de Paul Rivet, iv: Lenguas del Occidente de Colombia; iii: Lenguas del sur de Colombia. Santa Fé de Bogotá: CCELA, Universidad de los Andes. Llerena Villalobos, R. (ed.) (1995a). Estudios fonológicos del grupo Choco. Lenguas aborígenes de Colombia. Descripciones. Santa Fé de Bogotá: CCELA, Universidad de los Andes. —— (1995b). “Fonología comparada de las lenguas e)pe)ra de Occidente (Jaidukama) y Oriente (Cristianía y Alto Andágueda)”, in Llerena Villalobos (ed.) (1995a) 209–323. —— (2000). “Elementos de gramática y fonologia de la lengua cuna”, in M. S. González de Pérez and M. L. Rodríguez de Montes (eds.), Lenguas indígenas de Colombia: una visión descriptiva, Santa Fé de Bogotá: Instituto Caro y Cuervo, 59–72. Loewen, J. A. (1954). “Waunana grammar: A descriptive analysis”, MA thesis, University of Washington. —— (1958). “An introduction to e¸pe¸ra Speech: Sambú dialect”, Ph.D. dissertation, University of Washington. —— (1960). “Spanish loanwords in Waunana”, International Journal of American Linguistics, 26: 330–44. —— (1963a). “Choco I: introduction and bibliography”, International Journal of American Linguistics, 29: 239–61. —— (1963b). “Choco II: Phonological problems”, International Journal of American Linguistics, 29: 357–71. Loukotka, C. (1968). Classification of South American Indian Languages. Los Angeles: Latin American Center, University of California. Luykx, A. (2001). “Across the Andean airwaves: satellite radio broadcasting in Quechua”, in C. Moseley, N. Ostler, and H. Ouzzate (eds.), Endangered Languages and the Media: Proceedings of the Fifth FEL Conference. Bath: Foundation for Endangered Languages, 115–19. Mason, J. A. (1950). “The languages of South American Indians”, in J. H. Steward (ed.), Handbook of South American Indians. Bureau of American Ethnology, Bulletin 143, vol. vi. Washington: Smithsonian Institution, 157–317. Mejía Fonnegra, G. (1995). “El taller etnoeducativo: un espacio para la revitalización cultural entre los waunans. Ejemplo: la numeración”, in Pabón Triana 1995: 157–65. —— (2000a). “Lenguas aborígenes de la Costa Pacífica de Colombia”, in M. S. González de Pérez and M. L. Rodríguez de Montes (eds.), Lenguas indígenas de Colombia: una visión descriptiva. Santa Fé de Bogotá: Instituto Caro y Cuervo, 55–8. —— (2000b). “Presentación y descripción fonológica y morfosintáctica del waunana”, in M. S. González de Pérez and M. L. Rodríguez de Montes (eds.), Lenguas indígenas de Colombia: una visión descriptiva, Santa Fé de Bogotá: Instituto Caro y Cuervo, 85–96. Middendorf, E. W. (1892). Die einheimischen Sprachen Perus, iv: Das Muchik oder die Chimusprache. Leipzig. Migliazza, E., and Campbell, L. (1988). Panorama general de las lenguas indígenas en América. Historia General de América 10. Caracas: Instituto Panamericano de Geografía y História. Moore, B. R. (1962). “Correspondences in South Barbacoan Chibcha”, in B. Elson (ed.)1962: 270–89. Mortensen, C. A. (1999). A Reference Grammar of Northern Embera Languages. Arlington: A Publication of SIL International and the University of Texas at Arlington.

204


Obando Ordóñez, P. (1992). “Awa-Kwaiker: an outline grammar of a Colombian/ Ecuadorian language, with a cultural sketch”, Ph.D. dissertation, University of Texas, Austin. Oliver, J. R. (1989). “The archaeological, linguistic and ethnohistorical evidence for the expansion of Arawakan into Northwestern Venezuela and Northeastern Colombia”, Ph.D. dissertation, University of Illinois. Ortiz, S. E. (1946). “Los indios yurumanguies”, Acta Americana, 4: 10–25. Pabón Triana, M. (ed.) (1995a). Simposio: la recuperación de lenguas nativas como busqueda de identidad étnica. VII Congresso de Antropologua, Universidad de Antioquia. Medellín, junio 1994. Lenguas aborígenes de Colombia, Memorias 3. Santa Fé de Bogotá: Universidad de los Andes, CCELA. —— (1995b). “La lengua de Totoró: historia de una causa”, in Pabón Triana (1995b): 91–110. Pardo Rojas, M. (1997). “Aspectos sociales de las lenguas Chocó”, in Elsa Benavides Gómez (ed.), Lenguas amerindias: condiciones socio-lingüísticas en Colombia, gen. eds. Ximana Pachón and François Correa. Santa Fé de Bogotá: Instituto Caro y Cuervo, 312–81. —— and Aguirre, D. (1993). “Dialectología chocó”, in M. L. Rodríguez de Montes (ed.), Estado actual de la clasificación de las lenguas indígenas de Colombia. Santa Fé de Bogotá: Instituto Caro y Cuervo, 270–312. Payne, D. L. (1991). “A classification of Maipuran (Arawakan) languages based on shared lexical retentions”, in D. C. Derbyshire and G. K. Pullum (eds.), Handbook of Amazonian Languages, vol. iii. Berlin: Mouton de Gruyter, 355–499. Pineda Camacho, R. (1997). “La política lingüística en Colombia”, in Elsa Benavides Gómez (ed.), Lenguas amerindias: condiciones socio-lingüísticas en Colombia, gen. eds. Ximana Pachón and François Correa. Santa Fé de Bogotá: Instituto Caro y Cuervo, 155–75. Rivet, P. (1941). “Le Groupe kokonuko”, Journal de la Société des Américanistes de Paris, 33: 1–61. —— (1942). “Un dialecte hoca colombien: le yurumangí”, Journal de la Société des Américanistes de Paris, 34: 1–59. Rodrigues, A. D. (1986). Línguas brasileiras: para o conhecimento das línguas indígenas. São Paulo: Edições Loyola. Rouse, I. (1992). The Tainos: Rise and Decline of the People who Greeted Columbus. New Haven: Yale University Press. Ruhlen, M. (1987). A Guide to the World’s Languages. Stanford, California: Stanford University Press. Schmidt, W. (1926). Die Sprachfamilien und Sprachkreisen der Erde. Heidelberg. Seler, E. (1902). Gesammelte Abhandlungen zur amerikanischen Sprach- und Alterthumskunde, vol. i. Berlin: A. Asher and Co. Stark, L. R. (1972). “Maya-Yunga-Chipayan: a new linguistic alignment”, International Journal of American Linguistics, 38: 109–35. —— (1985). “Indigenous languages of Lowland Ecuador: history and current status”, in H. E. M. Klein and L. R. Stark (eds.), South American Indian Languages: Retrospect and Prospect. Austin: University of Texas Press, 157–93. Tovar, A., and Tovar, Consuelo Larrucea De (1984). Catálogo de las lenguas de América del Sur. Madrid: Editorial Gredos. Triana y Antorveza, H. (1997). “Factores políticos y sociales que contribuyeron a la desaparición de lenguas indígenas (Colonia y Siglo XIX)”, in Elsa Benavides Gómez


205

(ed.), Lenguas amerindias: Condiciones socio-lingüísticas en Colombia, gen. eds. Ximana Pachón and François Correa. Santa Fé de Bogotá: Instituto Caro y Cuervo, 85–153. Triviño Garzon, L. (1992). “Sobre la modalidad en la lengua Guambiano”, in II Congreso del CCELA: temas fonologicos. Santa Fé de Bogotá: CCELA, Universidad de los Andes, 219–27. —— (1994). “Hacia una tipología de la predicación de la oración simple en la lengua guambiana”, Bulletin de l’Institut des Études Andines, 23: 601–18. Vásquez de Ruíz, B. (1992). “La modalidad epistemica en Guambiano”, in II Congreso del CCELA: temas fonologicos. Santa Fé de Bogotá: CCELA, Universidad de los Andes, 209–18.

11

Fuegian Languages Oscar E. Aguilera

11.1. Chilean languages in general 11.2. Kawesqar 11.2.1. Distribution and current status 11.2.2. Documentation 11.2.3. Structural features 11.3. Yahgan 11.3.1. Distribution and current status 11.3.2. Documentation 11.3.3. Structural features

206 207 208 209 211 213 213 214 216

11.1. Chilean Languages in General Chile, with the longest coast in the whole Pacific Rim, is a multilingual country, but only six native language groups survive today: Aymara and Quechuan (mentioned in § 10.2.4), Mapuche, two Fuegian languages— Kawesqar and Yahgan—and Rapanui (or Pascuense; a Malayo-Polynesian language, cf. Ch. 14) in Easter Island. At least five other languages are known to have existed, some of them with little documentation, that is, Chango, Atacameño (below), Diaguita, Selk’nam, and Chono. See n. 4 for Ona, an extinct Fuegian language. The Mapuche or Mapudungu(n) (also known as Araucanian) family is the most vital group of Amerindian languages in Chile (also spoken in Argentina). While some of the languages are still not endangered (though threatened), others are considered moribund or seriously endangered. Although Atacameño, also known as Kunza or Likan-antai, is generally believed to be extinct, there is a rumour that some speakers survive in the Atacama region in northern Chile. The last and only trustworthy report about this language can be found in Mostny et al. (1954: 86) where it is mentioned that only a few old persons could speak the language though mixed with Spanish, and that there were some ceremonial songs whose meaning nobody could understand. Kawesqar and Yahgan (or Yamana) are the only Fuegian languages spoken today in Chile. Both are seriously endangered, with a small number of speakers, i.e. Kawesqar approximately twenty and Yahgan only two speakers.

f u e g i a n l a n g uag e s 75 W

207

70 W

C

ARGENTINA

H

K

A W

I E S Q

L S E

A

L K

R

E

E

N

A

M

Figure 11.1. Fuegian languages

11.2. Kawesqar Kawesqar is also known as Alacaluf, Alakaluf, Alikhoolip. Robert Fitz-Roy (1805–65), captain of the British navy’s Beagle in the surveying voyage of the southern shores of South America, was the first to name as Alikhoolip a group of people (about four hundred) he found west of the Beagle Channel and in the central section of the Magellan Strait (1839, II: 132). Fitz-Roy found two other groups (about two hundred adults), whom he failed to recognize as members of the same “Alikhoolip”. One of the groups he found about the central parts of Magalhaens Strait he identifies as “Pecherays”,

208

oscar e. aguilera

which was the usual exclamation they uttered, “whence Bougainville and others called them Pecherais” (Fitz-Roy, ibid.). This utterance, reported first by Bougainville (1771: i. 276), is also mentioned by Gusinde (1924, 1974: 115–16), who attributes this denomination to a Selk’nam loan, but it is no other than the Kawesqar word pælsc’éwe “alien”, “foreigner”, which is used today to designate a non-Chilean person. It is composed of pæls “the bill of a bird”, and -c’éwe a deictic which means “vertical”, “verticality”, pointing out the form of a ship’s captain’s or officer’s hat, and the vertical element referring to the distance from the canoe to the deck of the ship: therefore the meaning of the word would be something like “beak-headed-like man up on board”. For a complete account concerning the names of this ethnic group, see Aguilera (1978: 25–31, 2001a: 3–8).

11.2.1. Distribution and Current Status Kawesqars inhabited an extensive territory known as western Patagonia from the southern section of the Gulf of Penas to the Magellan Strait, i.e. the fjords and islands north of the Strait, and both sides of the latter. Kawesqars have been divided in three groups geographically rather than linguistically: Northern, Central, and Southern. This classification was first proposed by Gusinde (1974: 119), establishing as natural boundaries the Gulf of Penas in the north, Nelson Strait in the central part of the territory, and Magellan Strait in the south. He also mentions dialects which coincide with the geographical boundaries, though without providing evidence, and rightly points out that the differences among the three dialects are lexical rather than grammatical (Gusinde 1924: 48). Kawesqars themselves have distinguished several groups inhabiting particular zones within their territory, but they recognize only a Southern group with linguistic differences. Some of the groups are mentioned in the oral tradition, and others have been documented in chronicles of Spanish navigators (e.g. García Martí 1889). The number of fully competent speakers is no more than fifteen or sixteen, who live in Puerto Edén where the pristine form of the language has been preserved. Some other speakers live in Puerto Natales and Punta Arenas. A small number of semi-speakers, born in the 1960s, who belong to the last generation to learn Kawesqar as a first language, have moved to the cities in the south or the north and stopped using the language, except sporadically when they visit their relatives in Puerto Edén. Descendants of this generation and others who had migrated to the cities in the 1940s and 1950s, marrying Huilliches (Araucanian) in Punta Arenas and Puerto Natales (originally from Chiloe), never learnt to speak the language. They constitute nowadays

f u e g i a n l a n g uag e s

209

the so-called “urban Kawesqars”. According to the Bureau of Indigenous Affairs (CONADI), there are 279 Kawesqars who have registered themselves in the bureau, although some do not belong to this ethnic group. Such is the case of some married couples with one member belonging to another ethnic group. Urban speakers and semi-speakers usually do not live in the same neighbourhood and do not have daily communicational interchanges, and when they happen to meet, the language employed is usually Spanish. Nowadays Kawesqar is in fact spoken only in Puerto Edén, since it is the only community where Kawesqar people live close together. It has been the home of Kawesqars for more than seventy years, after they gave up their nomadic life. While the number of non-Kawesqar settlers in Puerto Edén was small, the language used to be unrestrictedly in daily use. But, as the alien population grew and the place became a small village, Kawesqars were looked down on, sometimes with aversion. The newcomers used to call Kawesqars “the Indian colony”, without realizing that the real colonists were themselves. Most of these colonists were from the Chiloe island, i.e. Huilliches, who felt superior and did not consider themselves as Indians. Chilotans usually mocked Kawesqars when they heard the language spoken. As a consequence, the language was restricted to home use among Kawesqars in private but never in the presence of strangers or Chilotans as they felt ashamed to speak in an “uncivilized” language. When a school was built in 1970, Kawesqars sent their children to study there and to learn “to speak well in Spanish”. These children are today’s semi-speakers, with different degrees of interference from Spanish. The feeling of shame was manifest even among Kawesqars themselves: they might pretend not to understand the mother tongue when they had sporadic encounters with kin in the city.

11.2.2. Documentation Linguistic documentation of Kawesqar dates back to the seventeenth century. Two kinds of documentation can be distinguished: (a) “informal”, represented by data gathered by early explorers, geographers, naturalists, or missionaries after the discovery of the Magellan Strait, and (b) the modern linguistic one which started in the 1970s. The informal documentation is composed of numerous lists of words and some sentences (cf. Cooper 1967), but an important number of items lack accuracy. Comparison of these informal records confirms that just one language was spoken in western Patagonia, with at least two dialects. An account of Kawesqar linguistic research from the seventeenth century till 1975 can be found in O. Aguilera (1978: 32–46).

210

oscar e. aguilera

Modern linguistic research on Kawesqar is represented by work carried out by Christos Clairis, Adalberto Salas, and Oscar Aguilera. Clairis (1977a, 1977b, 1987) and Salas and Valencia (1990) have separately examined the phonology of the language. Clairis’s grammatical description (1987) lacks many aspects due—it seems—to his failure to have a sufficiently prolonged contact with the language to corroborate his data and to get a more exhaustive analysis. Now we have a rather complete description of Kawesqar in the form of O. Aguilera’s descriptions in phonology, morphology, syntax, and some aspects of discourse (2002a, 2002b, 2005). A small comparative vocabulary of Northern and Southern Kawesqar was published by O. Aguilera (1978: 127–8), who now concludes that some of the items are Northern synonyms rather than dialect variants. It was during the second half of the 1970s that there arose an increasing academic interest in this ethnic group. Anthropological as well as linguistic fieldwork was carried out in the zone, and Kawesqars started to view their language as a cultural asset they had not realized, with some people in Puerto Edén collaborating intensively as consultants. All this work was done with academic interest, focused towards the preservation of cultural data outside the group: if the language became extinct, at least all the information obtained would guarantee that a cultural salvage had been successfully performed. In this context the Kawesqar community was not conceived as a beneficiary of the results of the research. The 1980s saw a decrease in research and fieldwork, and the people in Puerto Edén complained that they never saw the results of the research programmes carried out with their cooperation. What they claimed were concrete research products: reports, publications, pictures, sound recordings, videos, etc. that might evidence their contribution. This was important to them because they could show their importance as individual Kawesqars as well as an ethnic group, despite the discrimination they had to undergo. The published materials about their language would demonstrate that in fact scientific work had been produced by real scientists but not by non-specialists who were just interested in observing the people as if part of the zone’s unique fauna. The 1990s saw a revival of cultural interest regarding Chilean indigenous peoples. In 1993 the so-called Indigenous Law was promulgated. For the first time the government recognized Chile as a multicultural country and proposed “to develop several policies which were aimed at the promotion and conservation of the idiosyncratic values of Chilean ancient peoples” (Tonko 2001: 238). The purpose of this law was to protect the interests of the different ethnic groups and to rescue, make known, and revitalize all their cultural expressions.


211

One of the tasks that the Chilean government had to carry out was the implementation of a programme of Bilingual Intercultural Education (BIE) for all indigenous communities, which would be directed by the Ministry of Education. Nevertheless, an appropriate linguistic policy to face the problem of a multilingual country was never adopted, and BIE had to be left in the hands of enthusiastic and good-willed native teachers, who had to create, without a proper training, linguistic materials in order to teach these languages.1 Since there had been no reliable information, it was thought that both Kawesqar and Yahgan were extinct. Later the Ministry of Education funded projects which were aimed at multiculturalism rather than bilingual education, since none of the few Kawesqar or Yahgan schoolchildren spoke the vernacular languages any longer. Within both Fuegian communities, however, some members expressed their desire to recover the knowledge of ancient traditions and language. Thus, CONADI started to finance a programme of revitalization of these languages. The first concrete action began in 1998, when alphabets for both languages were proposed, and later accepted as “official alphabets” (§§ 11.2.3, 11.3.3), followed by a Kawesqar reference grammar, a first primer, and an encyclopedic dictionary Kawesqar–Spanish (Aguilera and Tonko 2003). A Spanish–Kawesqar dictionary is in progress now, although, because of considerable fund reduction, some parallel tasks had to be abandoned and the long-cherished sound archive may not be carried out. This will affect the Kawesqar community in a significant manner as they will not be able to preserve the testimony of the last speakers. In recent years several older speakers have died, and with the rate increasing, we cannot be optimistic about the recovery of this language.

11.2.3. Structural Features2 Kawesqar has phonemically six vowels /i, e, æ, a, o, u/, two contiguous ones of which often show alternation in the lexicon (/e/~/i/, /e/~/æ/, /æ/~/a/, /o~u/, for example, -sekóna~-sikóna “presumptive”; asétæl~asætæl “tell”; áltæl~ által “fly”; lol~lul “steal”), and eighteen consonants, i.e. (occlusive) /p, t, k, q/, (ejective) /p’, t’, c’, k’/, glottal /h/, fricative /f, s/, affricate /č/, nasal /m, n/, (lateral) /l/, trill /r/ (variants tap r, trill r, and retroflex), and approximant /j, w/. The official alphabet employed and the corresponding phonemes are shown in Table 11.1. Accent is varied, sometimes a same word has a different accent even spoken 1 For a more complete account of these early stages of BIE in Chile, see Tonko (2001: 237–48), and for the situation regarding Fuegian languages from a governmental perspective, see N. Aguilera (2001: 223–36). 2 For a complete account of Kawesqar grammar, see Aguilera (2001a).

212

oscar e. aguilera

Table 11.1. Grapheme

Phoneme

Grapheme

Phoneme

Aa Ææ Čč C’ c’ Ee Ff Hh Ii Jj Kk K’ k’ Ll Mm

/a/ /æ/ /č/ /c’/ /e/ /f/ /h/ /i/ /j/ /k/ /k’/ /l/ /m/

Nn Oo Pp P’ p Qq Rr Rr rr3 Ss Tt T’ t’ Uu Ww Xx

/n/ /o/ /p/ /p’/ /q/ /r/ /r/ /s/ /t/ /t’/ /u/ /w/ /k/

by the same speaker; the only regularity observed is that when a suffix is added, accent is displaced to the next syllable, e.g. jéksor “see” – jeksór-fqat “saw”, however some speakers don’t follow this rule. Kawesqar is basically an SOV language. It is morphologically agglutinative and moderately synthetic. Morphological processes employed are compounding and suffixation. The former is responsible for derived nouns (nóus-kstái nose-channel “nostrils”), and the latter for derivation (kiúrrowálak dog-female “female dog”, akčáwe-jeké cholga-DIM “small cholga”, i.e. Aulocomy ater, “a mollusc similar to mussels”) and for marking case (jemmá-s white man-genitive), plural verbs (-atál), and tense-aspect (áltqar-pas get.up-immediate.past “got up”). One of the most important features in the language is abundant deictics by which spatial location can be expressed very precisely. For instance, thirty-two forms to express “here” are possible by adding different spatial descriptors to the pronominal æs “ego” such as -asé “in this sound, bay, or any sound-like spot”, -k’ólaf “on this beach or beach-like flat surface”, -qálaks “on this pampa or steppe-like place, barren place or swampy spot”, -kar “on this isle, hill, or promontory”, -tqal “in this bay or place with a flat surface and rounded contour, of undetermined extension, external, frontal sight”, and -c’éwe “on/in a cliff-like terrain or ravine with an abrupt or smooth progressively ascending/ descending surface; verticality”. Spatial descriptors can be recursive for more precision, e.g. háute-kas-k’ólaf-kstái at the other side [of a channel or bay]-over there-beach-channel “over there at the other side of the channel, on the beach”. The most distinctive feature of Kawesqar verbs is its gradient past system, and its aspectual distinctions. 3

The grapheme corresponds to trill /r/, like Spanish “perro”.


213

Past distinguishes different moments that can point to: (a) a moment which occupies part of the present as a consequence of an action, process, or fact that has just occurred: this is the punctual or immediate past, suffix -pas; (b) a moment that occurred recently: recent past, suffix -(a)fqat; (c) a moment that occurred long before: remote past, suffix -horaras; (d) a moment that occurred in an undefined past: mythical or narrative past, suffix -hójok.

Aspect in Kawesqar is varied and exhibits different forms: aspectual suffixes or a particle that can be added to verb roots/themes and act as aspect.

11.3. Yahgan The people were first called Yahgan by the Anglican missionary Thomas Bridges (1842–98; below) from Yahga, the native name for the Murray Narrows region, a locality much frequented by some of them (Cooper 1963: 82). The original self-designation Yamana, which means “man”, is not accepted today by the members of the group, especially by women who say that they cannot call themselves “men”.

11.3.1. Distribution and Current Status Yahgan is spoken by a very much reduced number of speakers who live in Ukika, which nowadays has become a borough of the Chilean navy port Puerto Williams on Isla Navarino in the Beagle Channel. In earlier times they occupied a broader territory. According to Cooper (1963: 81), “the Yahgan regularly occupied the southern coast of Tierra del Fuego Island from about the eastern end of Beagle Channel to Brecknock Peninsula, and the islands south of this line to Cape Horn. But they evidently wandered more widely.” There were five dialects, spoken in different sections of their territory (Gusinde 1974; Koppers 1927): southern, eastern, central, western, and south-western. Today there are only two old women who can be considered completely competent speakers, Cristina Calderón and Emelinda Acuña. Their descendants as well as the descendants of other speakers now dead are semi-speakers. What they speak might be called “Ukika Yahgan”, which is probably a mixture of several dialects because of the settlement in Ukika of the last survivors.

214

oscar e. aguilera

The Yahgan population has been decreasing very rapidly since the nineteenth century. According to Cooper (1963: 83), “in the third quarter of the 19th century [it] totalled between 2,500 and 3,000 souls. In 1881 a sharp decline set in. By 1884 the number had dwindled to about 1,000, to 400 by 1886, to 200 by 1899, to 130 by 1902, to less than 110 by 1913, and to 40 by 1933. The immediate causes of the sudden drop in the eighties were the respiratory diseases and a severe outbreak of measles in 1884, followed by epidemics of typhoid, whooping-cough, and smallpox. Syphilis does not appear to have played an important role. Contributory, or rather basic predisposing, factors were, it seems, the then introduced European ways of life, especially the clothing but also the food, alcoholic beverages, and type of shelter and work.” Today’s Yahgan population, registered by CONADI, is seventy-eight, although the last official census inexplicably reports as many as 5,000. In 1946, according to Lipschütz and Mostny (1950: 24),4 the whole Yahgan community both adults and children spoke the vernacular language as well as Spanish, and some of them English, which was the first European language brought by British missionaries in the middle of the nineteenth century.5 In 1973, according to Ortiz (1973: 100), the situation had changed and the language was dying,6 as confirmed by Miyaoka (1992).7

11.3.2. Documentation T. Bridges, who succeeded to the evangelization of the Yahgan started by his foster-father (George Packenham Despard), lived among the people for thirty-six years and left a manuscript of an exceptionally rich dictionary for 4 “From the total of 19 adults, 13 spoke two languages: Yamana and Spanish; 5 spoke three languages: Yamana, Spanish, and English, or Ona; a 75 year old woman spoke four languages: Yamana, Spanish, English, and Ona” (translated by O. Aguilera). Ona, also known as Selknam, is an extinct language formerly spoken in Tierra del Fuego (in both Argentina and Chile). 5 The contact, much earlier than with Spanish, left a trace in a considerable number of English loan words, e.g. for “fat”, “salt”, “hat”, “sheep”, “pig”, “money”, “window”, “table”, “devil”, “nothing” (> “unimportant, nevertheless”), “to cook”, “to count” (> “to teach”), and “dirty” (Miyaoka 1992: 540). 6 “Since the recent generations have an imperfect command [of the language] and show no interest in broadening their knowledge of it, this is a completely understandable attitude, for the group is extremely reduced, and the language is the only remnant of a cultural past towards which they have no special esteem. In the future the knowledge or the lack of it will not change their point of view…. Yamana is relegated to the communication of elders—not even permanently—and can be partially understood by some youngsters who are far from being able to communicate correctly in it” (translated by O. Aguilera). 7 During a few weeks’ stay at Ukika in 1972–3, Miyaoka (1992: 539) found that there were no Yamana monolinguals (all speaking Spanish), about ten bilingual speakers all above the thirties (the two most fluent of whom had passed away by 1975—Clairis 1983), and a few who has some passive knowledge or would not speak in the presence of outsiders (the youngest being 19 years of age), and was told that a few Yamana speakers lived among Chilotes (of mixed Araucanian and Spanish parentage) on Isla Navarino and its surrounding islands as well as one old man at Ushuaia in Argentina (the southern coast of Tierra del Fuego).


215

the day and for any aboriginal languages of the New World. Posthumously published (Bridges 1933), it contains more than 32,000 entries, including abundant derivatives by the productive techniques of the language (compounding and affixation) as well as rather detailed semantic explanations and ethnological information.8 Bridges also published a grammatical note of less than thirty pages (1893), a primary but very fragmentary piece of information of rather limited value. Several restatements and vocabularies, which successively followed in the 1880s (L. Adam, A. J. Ellis, K. J. Platzmann, and C. Spegazzini), all relied upon Bridges’s material or instructions, occasionally with some distortions. Academic work did not start until the latter part of the twentieth century. An important work on Yahgan was published by Goodbar (1977, 1978, 1985) by supplementing Bridges’s material with his own research and following the descriptive schema in C. L. Hockett’s A Course in Modern Linguistics (1958), and Guerra explored various aspects of Yahgan grammar (1982 to 2000). Salas and Valencia (1990) published an article on the sound system with some notes on linguistic salvage, reporting a sporadic and declining use of the language, followed by Mendoza et al. (1997). From the 1990s till today, an undetermined number of recordings and notes have been made by aficionados, anthropologists, and sociologists, with no linguistic training. Most of these materials are lists of words gathered “to illustrate the Yahgan dialect”,9 but contain some recorded texts and part of the oral Yahgan literature that the Calderón sisters (see below) repeated. Though with a number of erroneous or misleading data, they constitute an important although scarce textual corpus that provides more linguistic information than uncontextualized sentences or word lists. In O. Aguilera’s first fieldwork carried out in 1999, he found only two speakers, Ursula and Cristina Calderón, whose use of the language had been reduced to a minimum, for most of the time they communicated in Spanish. A third speaker, Emiliana Acuña, I could not interview because (I was informed) her husband did not allow her to participate in any enquiry concerning the language, saying that she was too old and sick to be disturbed. Aguilera found out 8 The manuscript had been lost for a few decades after his death until it fell into the hands of the editorship (it is now to be archived in the British Library) finally after some unhappy circumstances as described in his son E. Lucas Bridges’s memoirs (1949). In the original manuscript Th. Bridges employed representations in the Ellis Phonotype (with modifications) which his foster-father had started for practical use among the natives, but this was transliterated by the editors into the Wilhelm Schmidt Anthropos alphabet. 9 Many people, including the press, call indigenous languages “dialects” in Chile, for they believe that these “dialects” have not the status of “language”, implying that they are “imperfect” or “primitive” forms of speech.

216

oscar e. aguilera

later, however, that this was not true; there was a sort of “conspiracy” against Emiliana, for her husband had died long ago. In 2003, during a very short visit to Puerto Williams, he was able to interview Emiliana, who was very helpful and eager to cooperate in the rescue of her language. Ursula and Cristina’s attitude towards their language was positive, since both felt that there was an urgent need to preserve it for future generations. The sisters felt that it was their duty to collaborate with persons interested in the Yahgan language and culture, since their sons and daughters and other relatives had only a passive knowledge or a partial command of the language, hardly ever using it. The Yahgan community welcomed the project proposed by CONADI, because they thought the revitalization of the language would contribute to strengthen their identity as an authentic Fuegian ethnic group whose cultural heritage was lost. However, the fund reductions and later the death of Ursula Calderón have prevented the continuation of this project. On the other hand, members of the Yahgan community seem to have lost their interest in the ancestral language, and Cristina Calderón would entrust the knowledge of her tongue to her granddaughter rather than to strangers. Her granddaughter Cristina Zárraga (called “Cristina the Younger”) is a young poetess, at present writing a biography of “Cristina the Elder”, but has a negative perception of linguists, unless they pay high fees for the interview. Thus, the two last extant Fuegian languages, Kawasqer and Yahgan, are in their last stage of existence, with no signs of recovery being envisioned, although the linguistic data gathered so far will at least ensure their presence among the documented languages of the world.

11.3.3. Structural Features10 Yahgan has phonemically seven vowels /i, e, æ, a, u, o, ә/ and seventeen consonants, i.e. occlusive /p, t, k/, fricative /f, s, š, č, x/, nasal /m, n/, lateral /l/, trill /r, ɹ, tr/. Vowels can occur lengthened, and consonants occur geminated. The official alphabet employed and the corresponding phonemes are shown in Table 11.2. Yahgan is basically an SOV/AN language (head final), with a predicative A following an NP. It is characterized by mostly dependent marking (case 10 Phonological data presented here are from Aguilera (2000). Grammatical data are based mainly on Bridges’s dictionary and have been processed by Jess Tauber ([email protected]) who has done a thorough examination of all Yahgan documents from the earliest sources till the present. I am very grateful for his help. Transcriptions of Yahgan examples are given in the standard alphabet, and rendered in the present Ukika Yahgan when possible, according to equivalents in our data.


217

Table 11.2. Grapheme

Phoneme

Grapheme

Phoneme

Aa Ææ Čč Ee Öö Ff Hh Ii Jj Kk Ll Mm

/a/ /æ/ /č/ /e/ /ə/ /f/ /x/ /i/ /j/ /k/ /l/ /m/

Nn Oo Pp Rr Řř Ss Šš Tt Uu Ww Xx

/n/ /o/ /p/ /r/ /ř/ /s/ /š/ /t/ /u/ /w/ /x/

on object nominals, when higher animacy), subject co-reference preverbally, and nominative/accusative alignment. Morphologically it is agglutinative and moderately synthetic. Morphological processes employed are suffixation, prefixation, and stem compounding. Suffixes are used for nominal case and number (but no gender marking), tense/aspect, adverbial subordination, nominalization, verbalization, etc., while prefixes (but largely proclitics) are verbal (subject) person, nominal (possessor) person, valency (causative, passive), etc. Instrument/body part prefixes and locational (spatio-temporal) suffixes frequently flank verb stems as more abundantly attested in Western North America than South America. Compounding includes verb stems with adverbs but is rare with nominals, and verb stem serialization is very productive. Tense-aspect markers developed from posture or movement verbs—e.g. “to stand” > “habitual, ready to, potential”, “to lie” > “definitively, extending to past”. Reduplication is found in some animal names, in emphasized adverbs, and “cognate object”-like constructions (e.g. uku-t-uku “materials to make a basket base (uku)”). The language has a small class of lexical numbers up to three or four (paucal rather than plural), which distinction obtains also with personal pronouns. This class of “numbered” contrasts with the “numberless”, i.e. “uncountable” beyond the paucal. Suppletive verb stems, which abound in the language, seem to be based upon this contrast in relation to intransitive S (“to die”, “to be afloat”, “to sit”, “to stand”, “to go”, “to go on board”, “to spread (blanket)”, “to lie”, etc.) or transitive O (“to give”, “to eat”, “pile up”, “to carry (in container)”, “to bear”, “to collect”, “to go on/off board”). Another characteristic—and complex—feature is the large-scale demonstrative system with stems the same as person pronominal roots (h-, s-, k-),

218

oscar e. aguilera

which enables very fine spatial/orientational distinction relative to the environment (“up”, “toward/away from the shore”, “right”, “left”, “strait”, the opposite shore’, etc.) as well as to the speaker/listener.

References Adelaar, W. F. M., with Muysken, P. C. (2004). The Languages of the Andes. Cambridge: Cambridge University Press. Aguilera, O. E. (1978). “Léxico Kawésqar–Español, Español–Kawésqar (Alacalufe septentrional)”, Boletín de filología (Universidad de Chile, Facultad de Filosofía y Letras, Departamento de Lingüística y Filología), 29: 7–149. —— (2000). “En torno a la estructura fonológica del yagán”, Fonología de la palabra, Onomazein (Santiago: Onomazein), 5, 233–41. —— (2001a). Gramática de la lengua Kawésqar. Temuco: Corporación Nacional de Desarrollo Indígena. —— (2001b). “A commentary on Dr. Wurm’s paper and revitalization of Fueguian languages”, in Lectures on Endangered Languages, 2: From Kyoto Conference 2000. ELPR C-002, 141–6. —— (2002a). “El discurso narrativo Kawesqar”, Onomazein. (Instituto de Letras de la Pontificia Universidad Católica de Chile), 7: 295–337. —— (2002b). “El tema en el discurso Kawésqar”, Revista chilena de antropología (Santiago: Facultad de Ciencias Sociales Universidad de Chile), 16: 167–86. —— (2004). “Basic Yaghan–English lexicon” (with CD), in N. Shibata and T. Shionoya (eds.), Languages of the Pacific Rim, vol. iii. ELPR A1-008, 339–84. —— (2005) El discurso narrativo kawésqar. Textos narrativos no míticos, Part 1. Onomazein 11. Instituto de Letras de la Pontificia Universidad Católica de Chile (in press). —— and Tonko, J. (2003). Diccioario enciclopédico kawésqar-español (unpublished MS). Aguilera A., N. 2001. “Revitalization of Kawésqar and Yaghan in Magallanes (Chile): governmental actions towards the rescue of two endangered Fueguian languages”, in Lectures on Endangered Languages, 2: From Kyoto Conference 2000. ELPR C-002, 223–36. Bougainville, L.-A. de (1771). Voyage autour du monde par la frégate la Boudeuse et la flûte l’Étoile; en 1766, 1767, 1768 et 1679. Paris: Chez Saillant et Nyon, libraires, rue S. Jean-de-Beauvais. (Many new editions, e.g. Paris: Club des libraires de France, 1958; Paris: ed. La Découverte, 1997.) Bridges, E. L. (1949). Uttermost Part of the Earth. New York: E. P. Dutton. Bridges, T. (1893). “A few notes on the structure of Yahgan”, Journal of the Royal Anthropological Institute, 23. —— (1933). Yamana–English: A Dictionary of the Speech of Tierra del Fuego, ed. F. Hestermann and M. Gusinde. Mödling: Verlag St Gabriel. Clairis, C. (1977a). “Première Approche du qawasqar: identification et phonologie”, La Linguistique (París), 13/1: 145–52. —— (1977b): “Lingüística fueguina”, Boletín de filología (Santiago), 28: 29–48.


219

—— (1983). Las lenguas de la Patagonia: América latina en sus lenguas indígenas, ed. Bernard Pottier. Caracas: Monte Avilla Editores, UNESCO. —— (1987). El qawasqar: lingüística fueguina. Teoría y descripción. Anejo de Estudios Filológicos 12. Valdivia: Universidad Austral de Chile. Cooper, J. M. (1963). “The Yahgan”, in Steward (ed.), 81–106. —— (1967). Analytical and Critical Bibliography of the Tribes of Tierra del Fuego and Adjacent Territory. Costerhout NB: Anthropological Publications (photomechanic reprint after the edn. of 1917). Fitz-Roy, R. (1839). Narrative of the Surveying Voyages of H.M.S. Adventure and Beagle, ii: Proceeding of the Second Expedition 1831–1836. London. García Martí, José (1889). “Diario de vida i navegación hechos por el padre José García de la Compañia de Jesús desde su mision de Cailin, en Chiloe, hacia el sur en los años 1776 i 1767”. Anuario hidrográfico de la marina de Chile (Santiago de Chile), 14: 3–47. González, S. (1990). El aymara de la provincia de Iquique-Chile y la educación nacional. Iquique-Chile: Taller de Estudios Regionales TER, Proyecto Educativo Bilingüe Intercultural PEBL. Goodbar, P. G. de. (1977). “Yagán I: las partes de la oración, Vicus (Amsterdam), 1: 5–60. —— (1978). “Yagán II: morfología nominal”, Vicus, 2: 87–101. —— (1985). “Hacia una morfología verbal de Yagán”, International Journal of American Linguistics, 51/4: 421–4. Guerra E., A. M. (1982). “Le Yahgan”, in Recueil de documents linguistiques. París: Université René Descartes - Paris V. —— (1985). “El caso del yahgan y de las lenguas de la Patagonia”, in Ciclo de conferencias sobre lenguas vernáculas contemporáneas de Chile. Santiago, 16–19. —— (1988). “Sobre algunos fenómenos de fonología de la frase en el yagán”, Nueva revista del Pacífico (Valparaíso), 33–6: 137–53. —— (1989). “Fonología del yagán”, MA thesis, Valparaíso, UPLACED. —— (1990a). “Esbozo fonológico del yagán”, in Actas del octavo seminario nacional de investigación y enseñanza de la lingüística. Santiago: Universidad de Chile y Sociedad Chilena de Lingüística, 5: 88–93. —— (1990b). “Yagán: aspectos culturales de un grupo étnico en vías de extinción”, Documentos lingüísticos y literarios (Valdivia), 16: 11–6. —— (1991). “Una visión etnolingüística de los yagán”; in Sobre culturas indígenas: lenguaje e identidad. Temuco: Universidad de la Frontera, 85–92. —— (1992). “Las fluctuaciones de fonemas en el yagán”, Revista de lingüística teórica y aplicada (Concepción), 30: 171–82. —— (1994). “Acento y sílaba en un corpus yagán”, Nueva revista del Pacífico, (Valparaíso), 38–9: 271–6. —— (1995). “Los sintemas del yagán”, Jornadas de lingüística aborigen (Buenos Aires: Universidad de Buenos Aires), 2: 271–6. —— (1997). “Adverbios en un corpus yagán”, Nueva revista del Pacífico (Valparaíso), 41–2: 59–70. —— (2000). “Yagan wordlist”, in M. R. Key (ed.), Intercontinental Dictionary Series i: South American Indian Languages. Irvine: University of California. CD-ROM. Gusinde, M. (1924). “Cuarta expedición a la Tierra del Fuego “informe”, Publicaciones del Museo de Etnología y Antropología de Chile, iv. Santiago de Chile. —— (1974). Die Feuerland-Indianer, iii/1: Die Halakwulup. Mödling bei Wien: Verlag St Gabriel.

220

oscar e. aguilera

Koppers, P. W. (1927). “Das Lautsystem der fuerländischn Sprachen”, Anthropos, 21: 1000–24. Lehnert, R. (1981–2). “Presencia del Runa-simi en el sector atacameño”, Cuadernos de filología, (Universidad de Antofagasta) 15–16: 29–47. Lipschütz, A., and Mostny, G. (1950). “Cuatro conferencias sobre los indios fueguinos”, Dictadas en la Universidad de Chile: revista geográfica de Chile (Santiago), 15: 86. Mendoza, P., Teresa, M., and Salas, A. (1997). “Fonemas yámana (yagán): estructura fonológica de la palabra”, Revista de lingüística teórica y aplicada (Concepción), 35: 125–33. Miyaoka, O. [宮岡伯人] (1992). Yahgan [ヤーガン語]. In The Sanseido Encyclopaedia of Linguistics [言語学大辞典]. Tokyo: Sanseido, iii: 539–43. Mostny, G. et al. (1954). Peine, un pueblo atacameño. Instituto de Geografía. Santiago de Chile. Ortiz, O. (1973). “Los Yámana, veinticinco años después de la misión Lipschütz’’, Apartado de Anales del Instituto de la Patagonia, 4/1–3: 77–105. Salas A. y A. Valencia (1990). El fonetismo del yámana o yagán. Una nota en lingüística de salvataje. Revista de Lingüística Teórica y Aplicada, 28: 147–169. Concepción. Steward, J. H. (ed.). (1963). Handbook of South American Indians, vol. 1, The Marginal Tribes. Smithonian Institution. Buereau of American Ethnology, Bulletin 143. Tonko, J. S. (2001). ‘‘Bilingual Intercultural Education in Chile: An Unfinished Project’’, in Lectures on Endangered Languages 2: From Kyoto Conference 2000. ELPR C-002, 237–248. —— (2004). ‘‘The View from Within. Revitalization of Kawésqar Language Achievements and Perspectives,’’ in Lectures on Endangered Languages 4: From Kyoto Conference 2001. ELPR C-004, 209–214.

12

Indigenous Languages of Australia Michael Walsh

12.1. The indigenous language situation in Australia 12.1.1. How many languages? 12.1.2. Indigenous languages of Australia 12.1.3. Contact varieties 12.1.4. Surveys of the indigenous language situation 12.1.5. Genetic relations and typological characteristics 12.2. Factors contributing to language loss 12.2.1. Stages of language loss 12.2.2. Language documentation 12.3. Addressing language loss 12.3.1. Language maintenance and revival 12.3.2. A range of terminology 12.3.3. Government funding 12.3.4. What can be achieved? 12.3.5. Renaissance in the south-east: Gumbaynggir, Kaurna, and Yorta Yorta 12.3.6. Educational initiatives 12.3.7. Training 12.3.8. Media 12.3.9. Cooperation among indigenous peoples 12.4. Factors contributing to regaining languages

221 221 222 223 223 225 226 226 227 228 228 229 229 230 231 232 233 234 234 234

12.1. The Indigenous Language Situation in Australia 12.1.1. How Many Languages? It is generally agreed that there were about 250 indigenous languages spoken in Australia when the first significant contact was made with outsiders (McConvell and Thieberger 2001: 16). Sustained contact began in earnest in 1788 with the European settlement of Australia and there is a strong correlation between the degree of contact and the relative viability of the languages today. Languages in the south-east of Australia bore the brunt of sustained contact and many of them have been referred to as “extinct”. On the other hand, the languages with the greatest degree of intergenerational transmission are to be found in the remoter parts of northern Australia where contact was much later and sparser.

222

m i c h a e l wa l s h

Figure 12.1. Indigenous language vitality

12.1.2. Indigenous Languages of Australia The expression “indigenous languages of Australia” has been used in the title because there are in fact two categories of indigenous languages in Australia: the languages of the Torres Strait and all the rest, the languages of Australian Aborigines. Traditionally there were just two distinct languages spoken in the Torres Strait: the Papuan language, Meriam, mainly on Murray Island, and an Australian language which includes varieties like Kalaw Kawaw Ya spoken on Saibai (see also McKay 1996: 69–84). Collectively Torres Strait Islanders and Australian Aborigines see themselves as indigenous people of Australia but they are keen to maintain a separate identity so that, for instance, the national government agency which dealt with these groups used to be called

i n d i g e n o u s l a n g ua g e s o f au s t r a l i a

223

the Department of Aboriginal Affairs (with the fine print indicating that this included Torres Strait Islanders) but has now become the Aboriginal and Torres Strait Islander Commission (henceforward ATSIC). Pama-Nyungan covers about five-sixths of the land mass of Australia so it is not surprising that most of the still healthy languages are among the PamaNyungan group. The healthy languages are all in northern Australia as various languages maps of Australia have shown.

12.1.3. Contact Varieties Contact with outsiders has spawned a range of contact varieties over the centuries. Most of these have been English based with some input from local indigenous languages. Pidgins developed in many areas and in turn influenced each other when they came into contact (see Wurm et al. 1996). With regard to endangerment the situation is complex: in some areas the creoles are spreading and in other areas the number of speakers is declining. Some pidgins have retreated because the conditions which created them have now disappeared. For instance, many Aboriginal people, particularly in northern Australia, were formerly employed in the pastoral industry—often receiving little more than some shelter, clothing, and rations. When full wages for Aboriginal people were introduced, for the most part the pastoral industry could no longer afford to employ them so the conditions under which the pidgin had developed were lost.

12.1.4. Surveys of the Indigenous Language Situation There have been many surveys of the indigenous language situation in Australia (for instance, Dixon 1980; Schmidt 1990; Dixon 1991; House of Representatives Standing Committee on Aboriginal and Torres Strait Islander Affairs 1992; Pearson et al. 1998; McConvell and Thieberger 2001; Dixon 2002) and all are agreed that there has been a decline in the number of languages spoken on a day-to-day basis. McConvell and Thieberger (2001: 2) conclude that ‘There has been a decrease of 90 per cent in the number of Indigenous languages spoken fluently and regularly by all age groups in Australia since 1800.’ However, this is not to say that only 10 per cent of the languages are spoken at all. There is a range of linguistic vitality and it is relatively uncommon to find an indigenous person in Australia who knows not a single word of their ancestral language(s). Indeed an indigenous person who declares that their language has gone will nevertheless know at least a few words and some stock phrases. Such a person

224


may be at once proud and ashamed of this knowledge: proud because it is a marker of their linguistic identity but ashamed that they do not know as much as their grandparents. The number of speakers of indigenous languages is therefore a problematic issue. There are degrees of fluency to be taken into account, for instance. Particularly in parts of northern Australia many indigenous people are multilingual and the survey instruments need to take account of such matters as what a person regards as their primary linguistic affiliation. However the overall trend in Australia is towards language shift, so that the Australian Census presents disturbing figures for the percentage of indigenous people over 5 years old who speak an indigenous language: 18 per cent in 1986 but 12 per cent in 2001. This means that a minority of indigenous people use an Indigenous language: in 2001, around 50,000 out of around 400,000 (within a total Australian population of around 17,500,000). Even taking into account the range of factors that might make one query these figures (McConvell and Thieberger 2001: 16–27, 33–45) the picture is disturbing and some have predicted accelerated loss in the years to come: “If these trends continue unchecked, by 2050 there will no longer be any Indigenous languages spoken in Australia” (McConvell and Thieberger 2001: 2). Schmidt’s survey (1990) asserted that of the 250 or so languages at first significant outside contact the current situation could be summarized as: twenty languages – healthy; seventy – weak or dying; 160 – extinct. Of the healthy languages just ten were said to have more than 1,000 speakers. For most of these languages children are learning the language from the earliest age – however in some cases the language of the children is significantly different from that of their grandparents (e.g. Lee 1987):

• • • • • • • • • •

Kala Lagaw Ya (Torres Strait): 3,000–4,000—including Kalaw Kawaw Ya; Warlpiri (Western Northern Territory): 3,000+; Arrandic languages (Central Australia): 3,000+—including Western, Eastern Arrernte, Anmatyerre, Alyawarre, Kaytitj; Western Desert languages, eastern (Central Australia): 3,000+— including Pitjantjatjara, Yankunytjatjara, Luritja, Pintupi; Dhuwal-Dhuwala (Yolngu) (Arnhem Land): 1,700–2,000—including Gupapuyngu, Djambarrpuyngu, Gumatj dialects; Tiwi (Bathurst and Melville Islands): 1,400; Western Desert languages, western (Central Australia): 1,000+ — including Manjiljarra, Yulparija, Kukatja, Ngaanyatjara, Ngaatjatjarra; Anindilyakwa (Groote Island): 1,000+; Walmajarri (Kimberleys) : 1,000; Wik Mungkan (Cape York): 900–1,000.


225

Of the seventy languages described as ‘weak or dying’ by Schmidt (above), fifty have less than 100 speakers. While there has been some controversy about these figures and some revisions, they reflect an overall situation in which most languages are either said to be extinct or in imminent danger of becoming so, to be classified as either ‘endangered’ (three), ‘seriously endangered’ (two), or ‘moribund’ (one) or ‘extinct’ (none). There is a wealth of information available on the vitality of Australian languages up to about 2000 at www.deh.gov.au/soe/techpapers/languages/indicator1a.html. And some of the information is provided in maps, e.g. www. deh.gov.au/soe/techpapers/languages/indicator1g.html#map5. See also the map in Wurm (2001) which gives a basic idea of the Australian situation, although it only shows about ninety languages.

12.1.5. Genetic Relations and Typological Characteristics Because Australian languages have been on the Australian continent for 40,000 years or more, genetic classification is controversial. At one time Dixon (1980: 228) asserted ‘that nearly all the languages of Australia form one genetic family’. The qualification, ‘nearly all’, was necessary because some languages were so poorly documented that it was impossible to say one way or the other whether a particular language was related. The most famous example of this concerns the Tasmanian languages where a systematic slaughter of the Tasmanians contributed to rapid language loss. There were also some languages like Djingili and Tiwi which resisted classification: Later it was shown that Djingili was a member of the Mindi subgroup but Tiwi remains an isolate. Despite his earlier assertion Dixon (2002: 38) now states: ‘there is no certainty that the modern languages of mainland Australia do go back—even in a long and indirect fashion—to a single ancestor language.’ What is clearer is the lack of genetic relations of mainland Australian languages with any languages outside Australia. There is no solid evidence of genetic links to any language outside. Typologically Australian languages have been classified into two broad categories: prefixing and suffixing (with the majority of non-Pama-Nyungan languages being of the prefixing type and the majority of the Pama-Nyungan languages being of the suffixing type). This is something of a misnomer as the so-called ‘prefixing’ languages have prefixes and suffixes whereas ‘suffixing’ languages are pretty much non-prefixing. The prefixing languages are confined to a small region (around one-tenth of the landmass) in northern Australia. Many languages follow a mixed pattern of accusative and ergative marking. Phonologically Australian languages are remarkably similar so that

226


a canonical consonant system can be set out (Dixon 2002: 63) with individual languages ‘selecting’ a particular consonantal inventory from this template mostly in a predictable way. Around two-thirds of the languages have the simple vowel pattern: i, a, u (Dixon 2002: 64).

12.2. Factors Contributing to Language Loss Many have commented on factors contributing to language loss but there is not always agreement. Dixon (1991: 236) proposes four current factors (since in the past there was also e.g. genocide): (a) white insistence, (b) Aboriginal choice, (c) shift of cultural emphasis; and (d) media pressure. By ‘white insistence’ Dixon means the insistence from non-Aboriginal people that English be used instead of indigenous languages. At times this was enshrined in official policy but more broadly Aboriginal languages were discouraged with varying degrees of explicitness. ‘Aboriginal choice’ includes the decision by parents to speak to their children only in English, thinking that this will give them a ‘better’ chance in life. At one time indigenous people may have used English for some purposes such as work and an ancestral language for traditional activities like hunting and social gatherings with indigenous people. As less time was devoted to traditional activities there was a ‘shift of cultural emphasis’ and English came to predominate. Finally, ‘media pressure’ refers to the barrage of English encountered by children from radio, TV, popular songs, print media, and educational instruction. Later we will see (§ 12.4) that these four factors have now been reshaped as factors promoting language revitalization.

12.2.1. Stages of Language Loss It is not feasible here to enter into all the details for stages of language loss. To start with there are many different schemes, one of the best known being the Graded Intergenerational Disruption Scale (GIDS) proposed by Fishman (1991). There have been critiques concerning the applicability of GIDS to Australian indigenous languages (for example, McConvell 1992, Lo Bianco and Rhydwen 2001: esp. 392–4). A fundamental problem with applying any of the schemes is the lack of detailed knowledge about levels of language proficiency across so many languages. A useful survey of the issues is provided by McConvell and Thieberger (2001: 52–69). Dixon (1991: 237–8) proposes five stages of language loss and then attempts to match the quality of linguistic documentation against these stages. Given the history of research on Australian Indigenous languages in


227

which the bulk of detailed work has been carried out in the last fifty years it is not surprising to find that the most poorly documented languages are also the ones which have experienced the greatest degree of loss.

12.2.2. Language Documentation As just mentioned there is a considerable range in the amount of documentation on Australian indigenous languages. Tragically some languages have virtually nothing. For example, one of the languages of the south-east of Australia where contact was earliest and most sustained has just 200 items of vocabulary and almost no grammar (Blake 2002: 164). Even that sort of minimal material will be subject to the vagaries of amateur recordings with no audiotapes to assist with phonetic accuracy. As one would expect, the best-documented languages are to be found in remoter parts of the country, especially northern Australia. Work has commenced on the compilation of a database which, amongst other things, will indicate the level of documentation for each indigenous language (McConvell and Thieberger 2001: 69–70). Fairly recently (based on Dixon 1991: 238–9) this was the general picture in terms of materials available : Good

c.50 languages

Fair

c.110 languages

Poor Minimal

c.50 languages c.40 languages

full grammars, sizeable dictionaries, text materials partial grammars, partial dictionaries, limited text materials minimal grammar, some word lists nil grammar, a few short word lists

More recently Dixon (2002: 2) gives: Good or fairly good

c.95

In preparation Lower quality Word lists only

5 c.110 c.25

reasonable grammar, together with a dictionary or word list descriptions

This array reflects—to a considerable extent—the history of research on Australian languages. Documentation of the languages began in 1770 when Captain James Cook and the botanist Joseph Banks took down word lists in a language of north Queensland, Guugu Yimidhirr. This was a by-product of an exploration party; it was another 200 years before a trained linguist, John Haviland, began a detailed study. Almost all the materials in the first two categories above (‘good or fairly good’ and ‘in preparation’) have been compiled

228


by professional linguists from the 1960s onwards. Most of the materials from the next category, ‘lower quality’, were compiled by amateurs often in the nineteenth century. Tragically, very little material was collected for any language for much of the twentieth century: from about 1910 until the late 1950s—amazingly similar to Alaska (Ch. 21, this volume) in general. One notable exception was Arthur Capell of the University of Sydney who pioneered work on a number of languages in northern Australia but whose major focus was on the Pacific. Another significant exception is Gerhardt Laves (Nash 1993) who undertook fieldwork in the period 1928–31—however, his valuable materials only became available from the 1980s. In 1959 Kenneth Hale (Green 2001; Hale 2001) began a remarkable trip which took him across vast tracts of Australia until 1961. During this period he recorded a prodigious amount of material from numerous languages. With the establishment of the first fully-fledged Department of Linguistics at Monash University in Melbourne in 1965 the discipline of linguistics entered a strong growth period with new departments being established, and graduate students carrying out fieldwork based Ph.D.s on Australian languages. With the publication of detailed descriptions by R. M. W. Dixon of the north Queensland languages Dyirbal (1972) and Yidiny (1977) during the 1970s, Australian languages became better known on the international scene. Dixon was a pivotal figure in fostering the documentation of Australian languages through his students (for example, Nick Evans’s detailed description of Kayardild (1995) and Bininj Gun-Wok (2003) as well as through the series co-edited with Barry Blake: Handbook of Australian Languages, 1979– ).

12.3. Addressing Language Loss 12.3.1. Language Maintenance and Revival Given the diversity of language situations in indigenous Australia when it comes to language intervention it is clearly not a matter of one size fits all (McKay 1996, which presents the range of situations including case studies of different types of language intervention; see also McConvell 2001; Ash et al. 2001: 19–35 esp. 20–5). In parts of northern Australia there are still languages being spoken by all generations. In parts of the more settled southeast there are languages for which there is so little knowledge that it is hard to see that much can be done.


229

12.3.2. A Range of Terminology Language revival can take a number of forms and this is reflected in a range of terminology. Basically the range of terms tries to capture the range of ambition and possibility that underpins the language revival process. For a language with just 200 items of vocabulary and little or no recorded syntax, it is hard to see how full language reclamation could be possible. The community that identifies with such a language may need to choose a language from somewhere else which is better documented and this choice would entail negotiations with the indigenous owners of that language as well as debate within the local Aboriginal community as to whether this is an appropriate choice. If, however, the ‘depleted’ language is their choice there are still meaningful options which will lead to a heightened sense of group identity. These include naming activities and songs. A school, for instance, might be assigned a name from the local language. A translation of the wellknown song ‘Heads, shoulders, knees, and toes’ basically needs just four words from the ancestral language. To give a concrete example, the Anewan (Nganyaywana) language of New South Wales is known only from a word list of 271 items. The city of Armidale, which sits within traditional Anewan territory, has adopted a policy of naming some places after words from the Anewan language. This involved assistance from a linguist and appropriate negotiations with the Aboriginal community. The result is a heightened acknowledgement towards those who claim an Anewan identity (Reid 2002). Some of the terms are language renewal, language revitalization, language reclamation, language awareness, and language resurrection (McKay 1996: 19–21). At this stage there is no consistency in the terminology although it would be useful if consistency could be achieved (McConvell and Thieberger 2001: 13–15). Whatever the term used in each case it is a matter for the Aboriginal community to decide how much should be done and how the process should be managed.

12.3.3. Government Funding Over the last fifteen years governments have made fairly modest investments in language maintenance and revival. Given the number of languages, the size of the country, and the dispersed nature of indigenous communities, huge amounts of funding would be needed to cater to all needs. Much of the funding has come from the national government (e.g. Dixon 1991: 239; Baldauf 1996; Hosking et al. 2000; McConvell and

230


Thieberger 2001: 77–8). It is not always easy to get access to up-to-date figures but a selection will give an idea of the scale of funding. A National Aboriginal Languages Program distributed funds of somewhat more than $A500,000 in 1987–8 and $A1,000,000 in 1988–9. A comparable national programme distributed just over $A4.5 million in 1997–8 and just under $A7.25 million in 1999–2000. There have been numerous reviews about the effectiveness of these programmes in terms of both delivery and outcomes. One difficulty that has been identified has been a lack of suitably qualified indigenous people who can provide the long-term commitment needed to ensure that a programme, once started, will be able to continue effectively. There are also programmes funded through the state and territory governments of Australia. In New South Wales there has recently been a major commitment from the state government with greatly increased funding for language programmes and the development of an Aboriginal languages policy. In particular the Northern Territory has invested in bilingual education programmes since 1973—but these efforts have not been without their problems (Hoogenraad 2001, 2002).

12.3.4. What Can be Achieved? Whenever there is an intervention governments need to have realistic expectations about what can be achieved. Unfortunately these issues have not usually been thought through sufficiently. Thieberger (2002) argues that decisions about the appropriate target for language maintenance programmes are too often driven by structural linguistics where the supposed ideal is intergenerational transmission of the language with all its complexity retained. Such a position is often in conflict with the actual preferences of the group of people who identify with the language in question and in any case may not be achievable. As Thieberger puts it: Fluency in a language may not be the achievable outcome of a language course for a number of reasons, not least of them being the enormity of the task perceived by learners of the language. For languages with few or no speakers we should be able to construct language programmes in which the use of a small number of terms in the target language, for purposes of identity, is a sufficient and realistic outcome. (Thieberger 2002: 310)

Having said this Thieberger warns that there are problems with this less ambitious aim: for instance, members of the community may be dissatisfied, preferring an all-or-nothing approach (2002: 324).


231

In addition, people outside the community may have stereotypes about the supposed simplicity of Aboriginal language reinforced. Others may regard the initiative as fundamentally wrong-headed. Dixon (1991: 239), for instance, complains of scarce resources from a national government programme for indigenous language maintenance being misdirected: “Many grants have been made to communities where a language ceased to be spoken many years in the past, with the aim of ‘reviving’ it (an impossible task)”. So there is a point of cleavage between those who believe in ‘full’ language maintenance/reclamation and those who accept other options, like Blake (2002: 165–6): Whether people manage to speak a language is not the only criterion by which the enterprise of language reclamation should be judged. Language can be a powerful marker of identity and even some knowledge of a language can serve that function and be a source of pride.

12.3.5. Renaissance in the South-East: Gumbaynggir, Kaurna, and Yorta Yorta In an earlier survey the situation for the south-eastern part of Australia had been described in the bleakest of terms: There are today many people of part-Aboriginal descent in the highly settled parts of Australia (Victoria, New South Wales and the southern parts of Queensland, South Australia and Western Australia) but virtually no full-bloods. Only one language is still spoken in this area (Bandjalang, in northern New South Wales), and that only by a handful of old people; over 100 languages have already become extinct. (Dixon 1991: 235)

In fact a number of languages in the south-east have been engaged in language revival (McKay 1996: 134 ff.). One of the better-documented cases of language revival concerns Kaurna, the language of the Adelaide Plains. Like many indigenous languages which had borne the brunt of early and sustained contact Kaurna had not been in everyday use since some time in the nineteenth century. During the 1980s Kaurna people began reinstating Kaurna place names. By 1997 a hundred public speeches in Kaurna were being delivered in a year. An initial focus for this language revival was the Aboriginal-controlled Kaurna Plains School. At first younger children were instructed in the language. Eventually instruction extended to senior high school and at times to university level (Amery 2000, 2001; McKay 1996: 135–6).

232


In New South Wales the Gumbaynggir language had fallen into disuse so that by 1994 there were just ten known speakers. However, from the mid 1980s, the Gumbaynggir community has made determined efforts to reverse the trend. The result is a growing number of younger adult speakers who are now passing on the language to their children, reinforcing the instruction they are receiving at a number of schools in the area (McKay 1996: 45–54; Walsh 2001). In north-east Victoria the Yorta Yorta language has also made a comeback—injecting life into a supposedly ‘dead’ language. Much of this effort has been focused on teaching Aboriginal students at senior high school level (McKay 1996: 145–6; see also www.fatsil.org/LOTM/oct99.htm). The result has been a marked lift in self-esteem among students who, prior to this language instruction, had often been quite troubled.

12.3.6. Educational Initiatives Whatever the intended target there is a range of options for delivering teaching and learning outcomes. Some of these are explored in Hartman and Henderson (1994). Apart from teaching indigenous people their own languages some of the wider community has come to learn or learn about indigenous languages. An important initiative in this regard is the development of an Australian Indigenous Languages Framework (SSABSA 1996a) which describes the range of language situations and offers a valuable approach to introducing indigenous languages to the wider community (see also SSABSA 1996b, 1996c). Another significant initiative is the recent development of an Aboriginal Languages Syllabus K-10 (from the earliest schooling to the latter years of high school). When teaching indigenous people their own languages, a range of issues must be addressed: who should teach the language; who can learn it; which language should it be; where should the language be taught? Some indigenous people (like the Gumbaynggir; §12.3.5) have very definite views. Some believe that only an indigenous person should teach the language; only an indigenous person can learn the language; only the local language should be taught or learnt and it should not be taught in a non-indigenous institution (see also Walsh 2003). If Aboriginal people are to teach the language, do they need to be taught how to teach? Again there is a range of opinion. Some Aboriginal people claim that one’s Aboriginality is sufficient qualification, while others concede that training is not just useful but necessary. For example, in New South


233

Wales one Aboriginal man lamented his lack of professional training: for years he had been trying to teach his language in a school situation and now regretted how much wasted effort had been expended because of his lack of expertise (Kevin Lowe, p.c.). Apart from training in language teaching some Aboriginal people have queried the need for expertise in linguistics. As with many of the issues so far discussed there is a wide range of opinion, from those who see no value whatsoever in expertise in linguistics to those who see it as an essential ingredient in the process of reclaiming one’s language. For those in the latter camp the idea of having Aboriginal people trained in relevant areas of linguistics has considerable appeal. A few NSW Aboriginal people already have some expertise in linguistics and more would like to gain the relevant training. Preliminary discussions have commenced about mounting a training programme for indigenous people modelled in part on the American Indian Languages Development Institute (McCarty et al. 1997).

12.3.7. Training Such a programme would complement training that has been in progress for many years. For instance, the School of Australian Linguistics trained “some two thousand speakers of about a hundred Australian languages and dialects between 1974 and 1989” (Black and Breen 2001: 161). Since then the College of Australian Languages and Linguistics has taken on this role (see also McKay 1996). Both institutions have been based in the Northern Territory but have an Australia-wide brief: to give indigenous people the necessary skills to engage in linguistic studies of their own languages and produce materials appropriate to community needs. A small number of indigenous people have pursued similar training through the Summer Institute of Linguistics and universities. Some have produced the kind of insightful studies only possible from a native speaker perspective (for example, Granites and Laughren 2001). Once indigenous people have acquired the relevant skills, two difficulties can arise: ‘retirement’ and information delivery. In small indigenous populations there will always be a need for talented individuals who can form a bridge between their own and the wider community’s culture, for instance, to handle government and other services. All too often an indigenous person who has gained training in linguistics and allied studies is called upon to fill such a role and in so doing enters into ‘retirement’ from the specialized tasks

234


of language maintenance or revitalization. The other difficulty is information delivery: often the speakers and owners of an indigenous language will be dispersed across a large area in small communities. Part of the solution to this problem is through the media.

12.3.8. Media Initially, print media were used to distribute information (McKay 1996: 110–12; Molnar and Meadows 2001), but more recently other media have supplemented these: “Indigenous radio and television help to sustain language and culture; they provide a vital channel of news and information for Indigenous people; and they have the potential to provide a means for better communication between Indigenous and other Australian.” (Productivity Commission 2000: 28, quoted in Molnar and Meadows 2001: 19–20). Some see broadcasting as a two-edged sword, with one commentator, the Aboriginal linguist Dr Eve Fesl, describing satellite TV as ‘cultural nerve gas’ (Molnar and Meadows 2001: 47). With the widespread advent of IT, people are looking to computers, CDROMs, and the Internet as a way to assist in the distribution of information. This can also contribute to cooperation among indigenous people.

12.3.9. Cooperation Among Indigenous Peoples One example of cooperation among indigenous people concerns the First Voices project based in Canada (www.fpcf.ca/resources/First%20Voices/ default.htm). Australian indigenous people have already been in contact with this project and are considering ways of adapting the Canadian experience to local needs. Within Australia the important task of fostering cooperation among indigenous people has been taken on by the Federation of Aboriginal and Torres Strait Islander Languages (FATSIL; www.fatsil.org). Although language maintenance efforts will continue to be focused at a local level there is a place for a national strategic approach to retaining and regaining language.

12.4. Factors Contributing to Regaining Languages Today the former insistence by whites that indigenous languages not be used has been muted. It is an increasingly common occurrence for public gatherings to acknowledge Aboriginal ownership of the place in which


235

the gathering is taking place. On such occasions indigenous people will often use some Aboriginal words and sometimes give short speeches in an Aboriginal language. This was extremely rare a mere twenty years ago and the move can be seen as an indication that Aboriginal people are now choosing to make their languages more public. In Dixon’s account (§ 12.2) it was Aboriginal choice which led to a shift away from Aboriginal languages. Now Aboriginal people are choosing to embrace their own languages in a variety of ways: naming activities; public speeches; songs; reclaiming and relearning. In Dixon’s terms it was a shift of cultural emphasis that led to language shift: indigenous languages became confined to specialized cultural domains and as those domains petered out so did the languages. In recent times indigenous peoples are creating new domains in which indigenous languages can be used: a shift of cultural emphasis—but one which fosters indigenous language use. To an extent, the media that have privileged English over indigenous languages can also be used to give new and greater prominence to indigenous languages. Indigenous people in central Australia, for instance, have said that they cried—with pride and joy—when they first heard their languages being broadcast.

References Amery, R. (1993). “Encoding new concepts in old languages: a case study of Kaurna, the language of the Adelaide Plains”, Australian Aboriginal Studies, 1: 33–47. —— (2000). Warrabarna Kaurna! Reclaiming an Australian Language. Lisse: Swets & Zeitlinger. —— (2001). “Language planning and language revival”, Current Issues in Language Planning, 2/2–3: 137–221. —— (2002). “Indigenous language programs in South Australian schools: issues, dilemmas and solutions” (www.boardofstudies.sw.edu.au/aboriginal_research/ #language_c). Ash, A., Fermino, J., and Hale, K. (2001). “Diversity in language maintenance and restoration: a reason for optimism”, in L. Hinton and K. Hale (eds.), The Green Book of Language Revitalization in Practice. San Diego: Academic Press, 19–35. ATSIC (1999). Digital Dreaming: A National Review of Indigenous Media and Communications, Executive Summary (based on a 500-page report). Canberra: ATSIC. Baldauf, R. B. (1996). “Back from the brink? Revival, restoration, and maintenance of Aboriginal and Torres Strait Islander languages”, Southwest Journal of Linguistics, 15/1–2: 1–22. Black, P., and Breen, G. (2001). “The School of Australian Linguistics”, in Simpson et al. 2001: 161–78. Blake, B. (2002). “Reclaiming languages in Aboriginal Victoria”, in D. Bradley and M. Bradley (eds.), Language Endangerment and Language Maintenance. London: RoutledgeCurzon, 156–66.

236


Dixon, R. M. W. (1972). The Dyirbal Language of North Queensland. Cambridge: Cambridge University Press. —— (1977). A Grammar of Yidin. Cambridge: Cambridge University Press. —— (1980). The Languages of Australia. Cambridge: Cambridge University Press. —— (1991). “The endangered languages of Australia, Indonesia and Oceania”, in R. H. Robins and E. M. Uhlenbeck (eds.), Endangered Languages. Oxford: Berg, 229–55. —— (2002). Australian Languages: Their Nature and Development. Cambridge: Cambridge University Press. Evans, N. (1995). A Grammar of Kayardild, with Historical-Comparative Notes on Tangkic. Berlin: Mouton de Gruyter. —— (2003). Bininj Gun-Wok: A Pan-Dialectal Grammar of Mayali, Kunwinjku and Kune. Canberra: Pacific Linguistics. Fishman, J. (1991). Reversing Language Shift: Theoretical and Empirical Foundations of Assistance to Threatened Languages. Clevedon: Multilingual Matters. Granites, R. J., and Laughren, M. (2001). “Semantic contrasts in Warlpiri verbal morphology: a Warlpiri’s verbal view”, in Simpson et al. 2001: 151–9. Green, J. (2001). “ ‘Both sides of the bitumen’: Ken Hale remembering 1959–1961”, in Simpson et al. 2001: 29–43. Hale, S. W. (2001). “Reminiscences of the trip to Australia 1959”, in Simpson et al. 2001: 19–28. Hartman, D., and Henderson, J. (eds.) (1994). Aboriginal Languages in Education. Alice Springs: IAD Press. Henderson, J., and Nash, D. (1997) (appeared May 1998). Culture and Heritage: Indigenous Languages. [Abstract] State of Environment Report Technical Paper Series (Culture and Heritage). Canberra: Environment Australia, part of the Department of Environment. Accessible as pdf file (view with Adobe Acrobat) via www.ea.gov. au/soe/ techpapers/series1/pubs/ indigeno.pdf. Hoogenraad, R. (2001). “Critical reflections on the history of bilingual education in Central Australia”, in Simpson et al. (eds.) 2001: 123–50. —— (2002). “Language and culture programs in NSW: the view from Central Australia” (www.boardofstudies.nsw.edu.au/ aboriginal_research/#language_d). Hosking, D. F., Lonsdale, T. J., Troy, J. F., and Walsh, M. J. (2000). Strong Language Strong Culture: New South Wales Strategic Language Study. Final Report and Strategy Action Plan. Canberra: AIATSIS. House of Representatives Standing Committee on Aboriginal and Torres Strait Islander Affairs (1992). Language and Culture: A Matter of Survival. Canberra: Australian Government Printing Service. Laughren, M. (2000). “Australian Aboriginal languages: their contemporary status and functions”, in R. M. W. Dixon and B. Blake (eds.), The Handbook of Australian Languages, vol. v. Oxford: Oxford University Press, 1–32. Lee, Jennifer (1987). Tiwi Today: A Study of Language Change in a Contact Situation. Canberra: Pacific Linguistics. Lo Bianco, J., and Rhydwen, M. (2001). “Is the extinction of Australia’s Indigenous Languages inevitable?”, in J. A. Fishman (ed.), Can Threatened Languages be Saved? Reversing Language Shift, Revisited: A 21st Century Perspective. Clevedon: Multilingual Matters, 391–422. McCarty, T. L., Watahomigie, L. J., Yamamoto, A. Y., and Zepeda, O. (1997). “School– community–university collaborations: the American Indian Language Development


237

Institute”, in J. Reyhner (ed.), Teaching Indigenous Languages. Flagstaff: Northern Arizona University, 85–104. McConvell, P. (1992). “Review of Fishman (1991)”, Australian Journal of Linguistics, 12/1: 209–20. —— (2000). “Two-way research resources for Indigenous languages: positioning resources in the Garma”. Paper to Linguistic Exploration: Workshop on Web-Based Language Documentation and Description. —— (2001). “Looking for the two-way street: Indigenous Australians battle to keep their languages strong”, Cultural Survival Quarterly, 25/2: 18–25. —— (2002). “Linking resources, linking communities: an Australian Indigenous Languages database, multimedia projects, and the role of metadata”. Paper for workshop on Resources and Tools in Field Linguistics, LREC Conference, Las Palmas, May 2002 (accessible through www.mpi.nl/lrec). —— and Thieberger, N. (2001). State of Indigenous Languages in Australia: 2001. Canberra: Department of the Environment and Heritage (www.ea.gov.au/soe/ techpapers/languages/). —— Amery, R., Gale, M.-A., Nicholls, C., Nicholls, J., Rigney, L. I., and Tur, S. U. (2002). ‘Keep that Language Going!’ A Needs-Based Review of the Status of Indigenous Languages in South Australia. Canberra: AIATSIS. McKay, G. (1996). The Land Still Speaks: Review of Aboriginal and Torres Strait Islander Maintenance and Development Needs and Activities. National Board of Employment, Education and Training, Commissioned Report 44. Canberra: AGPS. Mercurio, A., and Amery, R. (1996). “Can senior secondary studies help to maintain and strengthen Australia’s Indigenous languages?”, in J. Bobaljik, R. Pensalfini, and L. Storto (eds.), Papers on Language Endangerment and the Maintenance of Linguistic Diversity. MIT Working Papers in Linguistics 28. Cambridge, Mass.: MIT Press, 25–57. Molnar, H., and Meadows, M. (2001). Songlines to Satellites: Indigenous Communication in Australia, the South Pacific and Canada. Annandale: Pluto Press. Nash, D. (1993). “Gerhardt Laves [obituary]”, Australian Aboriginal Studies, 1/1993 [1994]: 101–2. —— (1998). “Indigenous Languages component” (esp. pp.19–20, 77–86), in M. Pearson et al., State of the Environment: Environmental Indicators for National State of Environment Reporting—Natural and Cultural Heritage, by State of Environment (Environmental Indicator Reports). Canberra: Environment Australia, Department of the Environment. Pearson, M., Johnston, D., Lennon, J., McBryde, I., Marshall, D., Nash, D., and Wellington, B. (1998). State of the Environment: Environmental Indicators for National State of Environment Reporting—Natural and Cultural Heritage, by State of Environment (Environmental Indicator Reports). Canberra: Environment Australia, Department of the Environment. Productivity Commission (2000). Broadcasting: Inquiry Report. Report No. 11. Canberra: Ausinfo, 3 Mar. Reid, N. (2002). “Creating Aboriginal placenames: applied philology in Armidale City”, in L. Hercus, F. Hodges, and J. Simpson (eds.), The Land is a Map: Placenames of Indigenous Origin in Australia. Canberra: Pandanus Books, 241–54. Schmidt, A. (1990). The Loss of Australia’s Aboriginal Language Heritage. Canberra: Aboriginal Studies Press.

238


Simpson, J., Nash, D., Laughren, M., Austin, P., and Alpher, B. (eds.) (2001). Forty Years On: Ken Hale and Australian Languages. Canberra: Pacific Linguistics. SSABSA (1996a). Australia’s Indigenous Languages Framework. Wayville: Senior Secondary Assessment Board of South Australia. —— (1996b). Australia’s Indigenous Languages in Practice. Wayville: Senior Secondary Assessment Board of South Australia. —— (1996c). Australia’s Indigenous Languages, ed. D. Nathan. Wayville, SA: Senior Secondary Assessment Board of South Australia, and Wakefield Press. With CD-ROM. Thieberger, N. (2002). “Extinction in whose terms? Which parts of a language constitute a target for language maintenance programmes?”, in D. Bradley and M. Bradley (eds.), Language Endangerment and Language Maintenance. London: Routledge Curzon, 310–28. Walsh, M. (2001). “A case of language revitalisation in ‘settled’ Australia”, Current Issues in Language Planning, 2/2–3: 251–8. —— (2002). “Teaching NSW’s Indigenous languages: lessons from elsewhere” (www. boardofstudies.nsw.edu.au/ aboriginal_research/#language_e) (www.boardofstudies.nsw.edu.au/aboriginal_research/pdf_do/teach_indig_lang_nsw_walsh.doc). —— (2003). “Raising Babel: language revitalization in New South Wales, Australia”, in R. M. Brown (ed.), Endangered Language and Identity. Proceedings of the FEL Conference, Broome, Western Australia, 22–4 Sept. 2003. Bath: Foundation for Endangered Languages, 113–17. Wurm, S. A. (ed.) (2001). Atlas of the World’s Languages in Danger of Disappearing, 2nd edn., revised, enlarged, and updated. Paris: UNESCO. —— Mühlhäusler, P., and Tryon, D. T. (eds.) (1996). Atlas of Languages of Intercultural Communication in the Pacific, Asia, and the Americas. Berlin: Mouton de Gruyter.

13

Languages of New Guinea Alexandra Y. Aikhenvald and Tonya N. Stebbins

13.1. Linguistic situation in New Guinea area 13.2. Genetic diversity in the New Guinea area 13.2.1. The Austronesian languages in New Guinea 13.2.2. The Papuan languages in New Guinea 13.3. Typological diversity among Papuan languages 13.3.1. Phonology 13.3.2. Morphology 13.3.3. Syntax 13.3.4. Documentation 13.4. Linguistic diversity and language endangerment in Papua New Guinea

239 243 243 246 250 251 252 255 256 256

13.1. Linguistic Situation in New Guinea Area The New Guinea region (as defined below) is one of the most linguistically diverse and complex areas in the world, with over 1,000 languages spoken in an area of about 900,000 square km. About three to four hundred languages spoken there belong to the Austronesian family. Other, non-Austronesian, languages are often referred to as “Papuan” (see Foley 1986: 1–3, 8; 1997a; Dixon 1991: 245). The term “Papuan” is a rough denomination subsuming over sixty language families, which are not demonstrably related, and a fair number of isolates in the area. This term is used for convenience (similarly, perhaps, to terms like “Paleo-Siberian” (§ 20.1, this volume) or “Amazonian” (§ 10.1, this volume)). The island of New Guinea includes the state of Papua New Guinea and the Indonesian province of Papua (formerly known as Irian Jaya). The language area centred on New Guinea stretches from the Halmahera and Timor islands in the west to the Solomon Islands in the east. Due to the limitations of space, we will be able to mention only a fraction of languages in the area, and a small portion of the varieties of genetic groupings that have been put Our deepest gratitude goes to Les Bruce, Wiem Burung, Cathy Easton, Karl Franklin, Claudia Gerstner-Link, John Hajek, Anthony Grant, Gerd Jendraschek, Joan Kale, Ger Reesink, René van den Berg, Sheena Van der Mark, Adam Bowles, and Perihan Avdi and especially R. M. W. Dixon, for their insightful comments. Endless thanks go to Cindi and Jim Farr, Steve Parker, and René van den Berg for allowing Aikhenvald to undertake research in the SIL archives in Ukarumpa in 1997 and 2004. Abbreviations: A–subject of transitive verb; AN–Austronesian; DS–different subject; du–dual; O–object; sg–singular.

240

a l e x a n d r a y. a i k h e n va l d a n d t o n ya n . s t e b b i n s

Figure 13.1. Papuan languages of the New Guinea area and their approximate locations

forward. Figure 13.1 indicates the approximate locations of some of the major Papuan families which are referred to in this chapter. For an overview of Austronesian languages spoken in the area, see Chapter 14 (this volume), and discussion in Lynch et al. (2002). The state of Papua New Guinea (independent since 1975) features over 830 languages (Grimes 2000; Nekitel 1998; Ford n.d.; Landweer 2000; Whitehead 1994 gives a figure of 870), with the number of Papuan languages exceeding 600 (see Foley 1986: 1–3; Dixon 1991: 245). Only about 20 per cent of the population speak Austronesian languages. The approximate numbers of languages and their speakers are given in Table 13.1 (based on Ford n.d., Nekitel 1998: 90–116, and Grimes 2000). The only language spoken by over 100,000 people is Enga (Engan family, EHP),

Table 13.1. The languages of Papua New Guinea and approximate numbers of speakers Number of speakers

Austronesian

Non-Austronesian

Over 10,000–100,000 1,000–10,000 Over 500–under 1,000 Between 100 and 500 Fewer than 100

about 12 about 100 about 30 under 50 about 10

about 60 about 230 about 105 about 170 about 50

l a n g uag e s o f n ew g u i n e a

241

whose speakers number between 164,750 (Grimes 2000: 755) and 200,000 (Foley 2000b: 359). Indigenous languages traditionally used as lingua francas include the Austronesian languages Suau (Milne Bay), spoken over the south-west end of Papua New Guinea (now with only about 6,795 L1 users and about 14,000 L2 users), and Dobu (Milne Bay), spoken off the islands of Eastern Papua (now with 8,000 L1 and 100,000 L2 speakers). Kuanua (or Tolai), an Austronesian language with about 100,000 speakers, is still used as a lingua franca in East New Britain (Fry 1977). (See Wurm 1977a and Foley 1986 on missionary lingua francas.) Three post-contact lingua francas dominate the linguistic scene in contemporary Papua New Guinea. Tok Pisin (Melanesian Pidgin) is currently the most important language spoken in most provinces (especially in the northern parts of New Guinea: see Ch. 7, this volume). The estimated numbers are 50,000 first language speakers, and over 2,000,000 second language users (Grimes 2000). (The total population of the country is 4,600,000 people.) English was the official language of the colonial administration in the southern half of the country (formerly known as Papua) from 1875 and of the northern half (formerly called New Guinea) from 1919. It is now rapidly gaining ground as a means of schooling and communication, especially in East Sepik province, Western province, and a number of other coastal provinces (see Sankoff 1980: 126–70). Hiri Motu is a Creole based on the Austronesian language Motu (still spoken by about 14,000 people in Central province).1 It developed around 1900 as a contact language for speakers from different language backgrounds in the Motu-speaking environment around Port Moresby (Dutton 1997), especially members of the indigenous police (hence its alternative name Police Motu). Hiri Motu is still widely used in the southern part of country (roughly corresponding to the old administrative division of Papua, covering Central, Oro, Gulf provinces, and parts of Milne Bay, as well as of Western province), but appears to be receding under the pressure from Tok Pisin. It has hardly any mother tongue speakers. The number of second language learners reported for Hiri Motu varies from 120,000 (Grimes 2000: 276) to 200,000 (Foley 1986: 31). 1 The origin of Hiri Motu remains uncertain (Dutton 1997). Until recently, it was assumed that Hiri Motu was a continuation of a trade language used by the Motu on annual trading voyages, or hiri, to the Gulf of Papua, where they traded with other linguistic groups. However, present-day Hiri Motu most likely developed from a special variety of simplified “foreigner talk” the Motu used with those who came to visit or trade in their own area.

242


The 1975 constitution of Papua New Guinea gives recognition to all languages, specifying that every citizen has the right to literacy in English, Tok Pisin, Hiri Motu, or a vernacular (Whitehead 1994). Traditional multilingualism, however, tends to be replaced by new diglossic and triglossic patterns with Tok Pisin and English. This, typically unstable, relationship often results in the dominance of the two lingua francas and the loss of the vernacular. According to materials in Sankoff (1980: 129–30), in 1971 the percentage of Papua New Guineans aged 10 and over who were unable to speak any of the official languages ranged from 5.7 per cent in New Ireland to 82.9 per cent in the Southern Highlands. Now the number of people with no knowledge of at least one official language is negligible. We return to the problems of language endangerment in Papua New Guinea in § 13.4. The Indonesian province of Papua has about 260 languages (Grimes 2000: 469; Wurm 2001), many of them non-Austronesian. Over 50 Papuan languages are also spoken in Northern Halmahera, Alor and Pantar, west of Timor, and the coastal areas and mountainous interior of eastern Timor. According to Wurm (2001: 394), at least forty-three languages in Papua are highly endangered or already extinct, and over seventy have fewer than 400 speakers. Languages in the Bird’s Head Peninsula spoken by over 10,000 people include Maybrat (Dol 1999), Hatam, and Meyah (Reesink 2002a, 2002b). The largest language in Papua is Western Dani in the Central Highlands area, with 180,000 speakers. Due to the current political situation in the Indonesian-controlled province of Papua and the compulsory spread of Indonesian and other languages (resulting from a settlement policy known as “transmigration”: Dixon 1991: 242–3), all minority languages in the area are endangered. According to Wurm (2001: 393–4), “the slide from a healthy state of a language to endangerment and extinction tends to be more rapid in Irian Jaya than is usual in Papua New Guinea” (despite the state’s obligation to support the maintenance of “regional languages” proclaimed in the constitution: Jones 1994). The main causes for language endangerment in Papua include migration to the centres and the impact of Indonesian-language electronic media and, especially, of compulsory Indonesian-language education (the latter contributing towards negative attitudes to the minority languages). The activities of the Summer Institute of Linguistics in the domains of language documentation, literacy, and education in local languages are nowadays restricted, to the detriment of language maintenance (Wiem Burung, p.c.).


243

When the Solomon Islands became independent in 1978, English was proclaimed its official national language. The main lingua franca in the country is Solomons Islands Pidgin (or Pijin) (French 1994). Of sixty-nine living languages in the state of Solomon Islands, seven are non-Austronesian. These are: Bilua, on Vella LaVella Island, with 8,000 to 9,000 speakers; Lavukaleve, on Russel Islands, with 2,000 speakers; Baniata, on South Rendova Island, with 1,480 speakers; Savosavo, on Savo Island, with 2,200 speakers; and Nanggu, with 450 speakers; Ayiwo, with 7,100, and Santa Cruz, with 5,000, on the Santa Cruz Islands. Terrill (2003: 11–12) reports that, in spite of the current number of 2,000 speakers, Lavukaleve is seriously endangered: many children grow up learning just Pijin. The same fate appears to face Bilua (Obata 2003), in spite of a higher number of speakers. Santa Cruz languages have undergone drastic simplification and restructuring under the influence of Pijin, and are just as threatened as all the rest (see Wurm 1992). With the current political turmoil in the area the future for these languages appears grim. The remainder of the chapter is organized as follows. In § 13.2, we consider the genetic make-up of the New Guinea languages and their major families. Some distinctive typological characteristics of Papuan languages are discussed in § 13.3. We briefly discuss the problems of language endangerment in Papua New Guinea in § 13.4.

13.2. Genetic Diversity in the New Guinea Area With over sixty language families and dozens of isolates, the genetic diversity of the New Guinea area is without parallel. First, we briefly look at Austronesian languages in New Guinea (§ 13.2.1). Then, in § 13.2.2, we focus on Papuan (non-Austronesian) languages.

13.2.1. The Austronesian Languages in New Guinea The largest family of languages spoken in New Guinea is the Austronesian family. It is also one of the largest well-established families in the world, with 1,000–1,200 member languages spread from Taiwan and South-East Asia to Hawaii and Madagascar. The migration of Austronesian peoples into New Guinea is considered to be fairly recent (going back to approximately 4,000– 5,000 years ago: Foley 1994, 2000b).

244


The ancestors of Austronesian languages are believed to have arrived from the west, spreading through the Philippines and Indonesia to the north coast of the island of New Guinea. This hypothesis is corroborated by the distribution pattern of Austronesian languages: they are spoken along the northern and eastern coasts of the New Guinea island and neighbouring islands. Hardly any Austronesian languages are spoken inland, or along the south coast of New Guinea, from around 200 km west of Port Moresby to the western tip of the island. The expansion of Austronesian languages in the islands close to New Guinea (including the Bismarck Archipelago, Bougainville, and Solomon Islands) resulted in the contraction and extinction of non-Austronesian languages spoken by earlier inhabitants. Nowadays, Kuot is the only nonAustronesian language spoken in New Ireland (Lindström 2002). Just two non-Austronesian languages are left in West New Britain: Anêm (Thurston 1987, 1992), on the north-west coast, and Wasi (Ata, Pele-Ata) in the Nakanai mountains. It is likely that many more non-Austronesian languages were originally spoken in these areas (see Thurston 1987, 1992: 125; Chowning 1996: 8–9); a number of non-Austronesian languages have become extinct during the past hundred years (for instance, Butam and Makolkol in East New Britain: Stebbins forthcoming). For possible effects of a nonAustronesian substratum in the Oceanic subgroup of Austronesian, see Lynch et al. (2002: 15–19). The Austronesian languages of New Guinea belong to the CentralEastern Malayo-Polynesian subgroup—see Fig. 13.2 (Foley 2000b; Lynch et al. 2002: 4). Oceanic languages in New Guinea (over 250) are spoken to the east of Cenderawasih Bay in Papua. A few Austronesian languages in Papua belong to the South Halmahera-West New Guinea branch. A small group

Central-Eastern Malayo-Polynesian

Central Malayo-Polynesian

Eastern Malayo-Polynesian

South Halmahera/West New Guinea

Oceanic

Figure 13.2. Austronesian languages and their subgroups in the New Guinea area


245

of languages in the Bomberai Peninsula in Papua are Central MalayoPolynesian. Oceanic is one of the best-known subgroups of Austronesian, with fairly detailed reconstructions of its proto-language (see Lynch 1998; Lynch et al. 2002; Pawley and Ross 1995). Considerably less is known about the two other Austronesian groups in the New Guinea area (see Blust 1993; and a summary in Foley 2000b: 362). A state-of-the art description and study of the subgrouping, and reconstruction of Oceanic languages, is in Lynch et al. (2002). The Oceanic languages spoken in New Guinea display unusual features, due to their interaction with non-Austronesian languages (see Ross 1994, on the influence of the non-Austronesian substratum on the phonologies of Oceanic languages in New Ireland). Many Oceanic languages of New Guinea have lost the typical Austronesian decimal counting system (Lynch 1998: 244, 245), replacing it with quinary systems common among Papuan languages. Prolonged contact between Austronesian and non-Austronesian languages often resulted in further structural changes. As an outcome of the impact of the non-Austronesian language Waskia (Ross 2001), Takia developed postpositions and verbal enclitics as clause-linking devices, and the orders Noun-Adjective and Noun-Determiner. All Oceanic languages of the Bel subfamily in Madang province—to which Takia belongs—have adopted verb-final order and developed medial verb forms; one of these languages, Dami, has also developed a same-subject/different-subject switch reference system, typical for Papuan languages (Roberts 1997: 123, 192; Ross 1987). In a comprehensive case study of a language contact situation in New Guinea, Thurston (1982, 1987) presents ample evidence of lexical and grammatical borrowing between Anêm, the only Papuan language of the north-west of New Britain, and the coastal Austronesian languages Kabana, Amara, Kove, and Lusi. In these languages, reciprocal is marked by a suffix on the verb, rather than by a prefix, as is usual in Oceanic languages, just as in Anêm (Thurston 1987: 79–80). Tense/aspect, negation, and modality are marked at the end of the verb phrase rather than with prefixes or preverbal particles, as would be the case in other Oceanic languages outside the contact area with Anêm. Due to the influence of Anêm, or other, now extinct, Papuan language(s), the four Oceanic languages have lost prenominal articles, another typically Oceanic feature, which survive only as part of a nominal stem. Anêm, in its turn, acquired a number of typically Oceanic features,

246


such as the inclusive-exclusive distinction in pronouns, and the division of nouns into inalienably possessed and two types of alienably possessed (“edible” and “neutral”) (Thurston 1987: 91). Semantic convergence between Anêm and its Oceanic neighbours resulted in a large number of shared idioms and semantic patterns (e.g. “my ankle” is literally “neck of my foot”, “my toenail” is “claw of my foot”: Thurston 1987: 80–7). Contacts with Papuan languages have contributed to the typological diversity of Oceanic languages, resulting in the development of unusual features. A striking example is the multiple classifier system in Kilivila (Senft 1996), somewhat reminiscent of multiple classifiers in Papuan languages elsewhere in New Guinea (see § 13.3.2).

13.2.2. The Papuan Languages in New Guinea The extreme genetic diversity among the non-Austronesian languages in New Guinea, with numerous families interspersed with isolates, remains a puzzle for comparative linguists. We first discuss well-established genetic units, and then turn to proposals on a larger scale. A. The Sepik river basin (which includes East Sepik and Sandaun provinces) is the most complex linguistic area within New Guinea. It has about 200 languages, a language density unparalleled anywhere else in the world. The Sepik river basin displays cultural as well as linguistic diversity and fragmentation, perhaps more so than other areas of New Guinea. Reasons for this include geographic diversity, inaccessible terrain, patterns of language contact, and language attitudes (see Foley 1986, 1988: 167–8; Aikhenvald 2004b). The average size of language communities is significantly lower than in the Highlands. Established families in the Sepik area include (see code labels on Figure 13.1): The Lower Sepik-Ramu family. This consists of: S1. The Lower Sepik family with six languages spoken by about 12,000 people, in the Lower Sepik and the adjoining riverine and coastal regions, and S2. The Lower Ramu family with eight languages spoken by about 11,000 people in the lower reaches of the Ramu river. The genetic relationship of S1 and S2 was demonstrated by Foley (2000a). S3. The Ndu family with eight languages spoken by over 100,000 people along the course of the middle Sepik river and to its north (Laycock 1965; Aikhenvald 2004b, forthcoming);


247

S4. The Sepik Hill family with sixteen languages spoken by about 8,000 people in the foothills between the central range and the flood plain of the Sepik river, with the southernmost members extending across the central range in the Southern Highlands and the Enga Provinces (Bruce 1984: 296); S5. The Ram family with three languages spoken by about 1,200 people in the southern foothills of the Torricelli Range in the Sandaun Province (Feldman 1986); S6. The Tama family with four or five languages spoken by about 10,000 people in the Ambunti district of the East Sepik province and the Sandaun province, stretching into the lower foothills of the Torricelli mountains (see Foreman 1974).

An attempt has been made to establish genetic links between S3–S6, and a few other languages of the Sepik area, such as Wogamusin-Chenapian, Kwoma-Kwanga, and Abau (Foley 2000a, 2000b), grouping them into a larger Sepik family. Though some similarities in pronouns appear suggestive, much more comparative analysis is required, including low-level reconstructions for individual language families, before any definitive conclusion can be reached. S7. The Skou family with eight quite closely related languages, spoken by about 6,000 people along the northern coast of New Guinea and the adjacent areas of Papua (Donohue 2002a, 2002b, forthcoming). S8. The Torricelli family with about fifty languages spoken by over 80,000 people in the area of the Torricelli mountains, between the north coast and the Sepik river, spreading to the north into coastal areas of western Madang province.

The Torricelli languages share a number of highly unusual typological features, such as complex noun class systems with largely phonological assignment, irregular plurals for nouns, and complicated pronominal cross-referencing on the verb (see Fortune 1942; Nekitel 1985; Conrad and Wogiga 1991; Dobrin 1999; a brief comparative study of Arapesh languages is in Conrad 1978). Further comparative work is required to demonstrate convincingly that these languages form a family and not just an area (Foley 1986: 241–2). B. The New Guinea Highlands are home to over a dozen well-established genetic units, including: H1. Angan family, with twelve languages spoken by over 60,000 people in the mountains in the Eastern Highlands, Morobe, and Gulf provinces, and in the lowland areas of the Gulf province;

248


H2. Binanderean family, with fourteen languages spoken by over 50,000 people in Oro province; H3. Kalam-Kobon family, with three languages and about 15,000 speakers on the northern slopes of the central highlands in Madang and Western Highlands province; H4. Huon family, with fifteen languages spoken by over 80,000 people on the Huon Peninsula; H5. Engan family, with eight languages spoken by over 330,000 people in Enga province and the adjacent areas of East Sepik and Southern Highlands province; H6. Gorokan family, with eight languages spoken by over 180,000 people in the Eastern Highland province around Goroka; and H7. Kainantu family, with four languages spoken by about 40,000 people in the Eastern Highlands province.

The Gorokan and Kainantu family (Foley 1986: 245–63) appear to be genetically related. C. Within the southern part of Indonesian Papua and adjacent Papua New Guinea areas, the well-established units include: H8. Asmat-Kamoro-Sempat (or Asmat) family with four languages and about 50,000 speakers in southern parts of Papua; H9. Marind family, with six languages with about 20,000 speakers, in southern Papua and the adjacent areas of Papua New Guinea in the Lake Murray area of the Western province; and H10. Awyu (or Awyu-Dumut) family, with nine languages and about 20,000 speakers in the lowlands of Papua between the areas of the Asmat and Marind families. An attempt has been made to group H1–H10 and numerous other language families in the New Guinea Highlands, from the Huon Peninsula across the island (covering the Fly river area and extending probably as far as the languages of Timor and Alor), into a large genetic unit called the “Trans New Guinea” phylum.2 These languages share a number of typological features (see § 13.3), and are mostly spoken by mountain dwellers. This macrogrouping, however suggestive, has not yet been provided with systematic justification following the established principles of comparative linguistics

2 This genetic and areal unit was first proposed by Wurm (1975a), who did not provide much linguistic evidence to justify it. The original hypothesis has been revised in the recent studies by Pawley (1995, 1998, 2000) and Ross (1995, 2000). The exact language membership and subgrouping within the putative phylum is still a matter of debate. For instance, Engan languages are treated as bona fide Trans New Guinea languages by Pawley (2000), but only as its “likely” members by Foley (2000b).


249

(covering lexicon, phonology, and all aspects of the grammar including paradigmatic correspondences). The Trans-New Guinea features suggested so far include: (a) some person markers, for instance a first person in *n and second person in *k (this may well be a coincidence, given the frequency of both n and k in personal pronouns);3 (b) alternation of vowels showing number in independent pronouns with the initial consonant being the same, e.g. *a “singular”, *i “plural”, as in * na “first person singular”, *ka “second person singular”, *ni “first person plural”, *ki “second person plural” (Ross 1995) (which may well be an areally diffused typological feature); and (c) a small set of lexical cognates (most of which are either monosyllabic, or involve conjectural semantic and/or phonological correspondences). The Madang area includes over a hundred distinct languages forming five or six large families, among them Southern and Northern Adalbert, Mabuso, and Rai coast. These have been provisionally grouped into a larger unit under the name of Madang (this also includes the Kalam-Kobon family, under H3 above). This hypothesis is currently under investigation (see Ingram in press). Other established families within Papua (not suggested for incorporation into the Trans New Guinea phylum) include (see Foley 2000a): P1. The Border family, with over twenty languages spoken by about 13,000 people in the Sandaun province and the border areas of Papua New Guinea and the province of Papua (Seiler 1985, for the Waris subfamily; see discussion by Gerstner-Link 2004); P2. The Lakes Plain family, with twenty-four languages spoken by about 5,000 people, in the northern lowlands of Papua, in the basin of the Taritatu and Tariku tributaries of the Mamberamo river (Clouse 1997); P3. The Wissel Lakes family, with four languages spoken by over 80,000 people in the central highlands of Indonesian Papua; and P4. The Sentani family, with three languages spoken by about 10,000 people on the northern coast of Papua, to the west of Jayapura around Lake Sentani and to its west.

Reesink (1998, 2000) postulated a “West Papuan” linguistic area covering up to twenty-six non-Austronesian languages spoken in the area of Indonesian Papua, from Cenderawasih Bay westward to the islands of 3 Along similar lines, Campbell (1997: 345) criticized the proposed diagnostic trait of n “first person” and m “second person” for the putative “Amerind” as a chance coincidence.

250


Timor, Alor, and Pantar, including the Bird’s Head Peninsula and the island of Halmahera. These languages (spoken by a total number of under 200,000 people) share a number of typological features (see under W(est) P(apua), on Figure 13.1). D. The non-Austronesian languages of Bougainville, the Bismarck Archipelago, and the Solomons present a complex picture.4 The Bougainville family (B on Figure 13.1) consists of eight languages, four in the north and four in the south of the island of Bougainville (Onishi 1994), spoken by about 40,000 people. The question of the genetic relationships between the four nonAustronesian languages in the Solomons remains open: they may well be isolates, or relics of now extinct language families. Kol and Sulka, in Central New Britain, and Kuot, in New Ireland, are also isolates. Ross’s (2000) hypothesis that Anêm and Ata in West New Britain are related requires further substantiation. The only well-established language family in East New Britain is the Baining family (TB on Figure 13.1), consisting of six languages (probably including the extinct Makolkol). There are over 10,000 speakers of Baining languages today. Structural and formal similarities between Taulil (and the extinct Butam) and the Baining family are equally suggestive of a distant genetic relationship and of contact. In summary: in quite a few cases in New Guinea it is premature to go beyond establishing small families. Given the time depth of human occupation of the area (estimated as about 50,000 years: Bellwood 1998; Foley 2000a), the erstwhile genetic relationships tend to be obscured by borrowing and multilayered diffusion, “a major source of diversification in Papuan languages” (Foley 1986: 262). Much more comparative and descriptive work is required before genetically inherited features can be usefully distinguished from typological similarities or chance resemblances (see Foley 2000a).

13.3. Typological Diversity among Papuan Languages In addition to their extreme genetic diversity, Papuan languages are typologically diverse. A few Pan-Papuan tendencies are typical for the New Guinea 4 Typological similarities between these languages were investigated by Dunn et al. (2002), and Terrill (2002). For instance, in contrast to the neighbouring Austronesian languages, many have genders and multiple classifier systems. No linguistic justification has been provided for grouping them into an “East Papuan” phylum (as advocated by Wurm 1975b, 1982; see extensive criticism in Terrill 2003: 8–10).


251

Highlands languages within the putative Trans New Guinea areal grouping. Exceptions occur elsewhere.

13.3.1. Phonology Phonological systems of Papuan languages are relatively simple, usually with three places of articulation for consonants (labial, dental/alveolar, and velar; some languages, such as Yimas, also add a palatal). Languages of the Sepik area have the same number of distinctions in place of articulation for stops and nasals, while the Highlands languages have fewer distinctions for nasals than for stops. Some Ndu (S3) languages have a series of labialized stops. In contrast to neighbouring Austronesian languages, many Papuan languages tend to have only a few fricatives and one liquid (r/l). Voiced stops tend to become prenasalized in the syllable-initial position. Papuan languages tend to have a standard five vowel system /i, e, a, o, u/. Languages of the northern half of Papua New Guinea often have at least one non-low central vowel, for example Fore (Scott 1978) and Kewa (Franklin 1971). A “vertical” vowel system consisting of three central vowels -i , ә, and a has been reported for Iatmul, a Ndu language (Staalsen 1966) (however, other Ndu languages have up to seven vowels: see Wendel 1993: 36; Wilson 1980; Aikhenvald forthcoming). The most complex vowel systems are found in Skou (S7) and some Waris (P1) languages. Both Skou and Dumo have seven vowels, two front, three central, and two back. In Skou, all except the high central vowel have a nasalized counterpart. In Dumo, all the vowels exist in long and short form, and all have a nasal counterpart except for the front mid-rounded vowel (Ingram 2005). The Papuan languages of the Lower-Sepik Ramu (S1–2), Torricelli (S8), and numerous families in the Sepik–Ramu area have stress systems. No tonal languages have been found in the languages of the Bismarck Archipelago, Bougainville, and the Solomons. Tone is a prominent feature in numerous languages in the Trans New Guinea and West Papuan (WP) areas. In contrast, most Austronesian languages have stress (some Huon Gulf languages are an exception: Ross 1993). Tone appears to be a family feature of Lakes Plain (P2) and Skou (S7) languages. In the Skou family, Skou has three tones (high, low, and rising: Mark Donohue, p.c.), while Dumo has high and non-high. The most complex tone system is found in Iau (Lake Plains family), which is said to have eight contrastive tones and up to two tones per syllable (Edmondson et al. 1992; Donohue 1997: 356). The development of these tones in Iau is most

252


probably due to loss of vocalic and consonantal segments. Other languages of this family have at least high and low tone; some are reported to have four (Clouse 1997: 152). An overview of tone in New Guinea is in Donohue (1997); see a summary in Foley (2000b).

13.3.2. Morphology In their morphological type, Papuan languages vary. Mildly synthetic languages with little bound morphology include those in the West Papuan area (WP on Figure 13.1; see Reesink 1999, 2000), Lake Plains languages (P2; Clouse 1997: 151), and the Skou family (S7; Ingram forthcoming). Complex polysynthetic languages include Yimas, from the Lower Sepik subfamily (S1) of the Lower Sepik–Ramu family. Papuan languages often have complex patterns of pronominal cross-referencing. Marking transitive subject and object on the verb in Yimas is illustrated in (1) (Foley 1991: 291). Yimas (1) na-ŋa-tmi-kwalca-t 3sgA-1sgO-say-rise-PERFECTIVE “She woke me up” (by verbal action) In contrast, Watam, from the Lower-Ramu branch (S2) of the Lower Sepik–Ramu family, has no cross-referencing on the verb (Foley 1999). Most languages from the Trans New Guinea area employ a bound agreement prefix for object and a suffix for subject, for instance Fore (Scott 1978). So does Anamuxra, from Madang area. In Manambu, from the Ndu family (S3), suffixes cross-reference the subject and, optionally, another constituent, provided it is topical (this second suffix position is absent from its relatives Iatmul, Wosera, and Abelam). Skou languages encode the subject with partly predictable alternations of the first consonant of (a monosyllabic) verb root. The only other morphological mechanism is reduplication of the verb, to mark irrealis. Various aspectual and modal meanings are often expressed using verb serialization, a mechanism found in most Papuan languages, including Skou (S7), Lower Sepik–Ramu (S1–2), Sepik Hill (S4), and languages of the Trans New Guinea area (see Ingram 2005, in press; Foley 1991; Pawley 1993; Bruce 1984). Many heavily serializing languages have a closed class of verbs, with fairly generic meanings. The verbal lexicon in Kalam (H3; Pawley 1993) has under 100 monomorphemic lexemes. New lexemes consist of a generic verb preceded by several verb stems (together forming one serial verb), or by one


253

or more nominal or adverbial complements. For instance, the generic verb ag “make a sound, utter” enters into the following combinations: mnm ag(speech utter) “speak”, kmap ag- (song utter) “sing”, swk ag- (laughter utter) “laugh”, sb ag- (bowel utter) “fart” (Pawley 1993: 98). Serial verbs are absent from the languages of the south coast of New Guinea, Torricelli languages (S8; e.g. Arapesh), and most languages of the Bismarck Archipelago and Bougainville (B). Languages without extensive verb serialization tend to have complex verbal morphology (Stebbins forthcoming). There is typically no passive, but at least one causative. Grammatical relations are frequently expressed with verbal crossreferencing, as in Motuna, Anamuxra, Yimas, and numerous languages from the Trans New Guinea area. A few have core cases, e.g. Imonda (Waris: P1) and Ndu languages (S3). The marking of verbal arguments may be semantically rather than syntactically based. In Bauzi, from the Mamberamo river region of Papua (Briley 1997), a suffix marks an agentive argument with both transitive and intransitive verbs. Similar systems are attested in a few languages from the New Guinea Highlands (e.g. Ku Waru: Merlan and Rumsey 1991: 335; Folopa: Anderson and Wade 1988). In addition, Bauzi has a typologically unusual split marking of core participants conditioned by tense-aspect (the progressive/continuative aspect does not require the agentive suffix, while the completive aspect does), and by polarity (no agentive suffix is required in negative clauses). Papuan languages exhibit the world’s most sophisticated systems of noun classification. The Lower Sepik (S1) and Torricelli languages (S8) have extremely complex systems of noun classification and plural marking (Foley 1986: 222–9, 1991, 2000a; on Arapesh languages, see Fortune 1942; Nekitel 1985; Dobrin 1999). Yimas has eleven noun classes, four of which are assigned to nouns by their semantics: I—human males; II—human females; III—animals; class IV—culturally important plants. Classes V–XI are motivated phonologically: the agreeing constituent repeats the last consonant of the nominal root. Small gender systems are found in the West Papuan area (WP), Skou languages (S7) and numerous families in the Sepik basin, and in numerous language isolates, such as Burmeso, Sulka, Kuot, Bilua, and Lavukaleve of the Bismarck Archipelago. In Alamblak (S4) and Manambu (S3), from the Sepik area, feminine gender is associated with small size, and masculine gender with large size and long shape (Bruce 1984: 9; Aikhenvald 1998). In contrast, feminine gender assignment to inanimates is associated with large size, with masculine gender assignment being linked with slender elongated objects in

254


Yonggom (Ok, Western province: Christensen 1995: 9–10) and in Olo (S8: Torricelli: McGregor and McGregor 1982: 55). The Baining languages of East New Britain (TB) have two noun classification systems and a threeway number contrast (singular, dual, and plural). A simple masculine, feminine, and neuter (“everything else”) gender system is used for subject cross-referencing on the verb and in possessive constructions; a more complex, multi-term shape- and size-based system is realized through agreement within noun phrases of other types. Complex classifier systems in multiple environments are a feature of the languages of Bougainville (B). Nasioi (Hurd 1977) has over a hundred classifiers used in possessive constructions, with numerals, demonstratives, adjectives, and verbs. Motuna, from the same language family, has fifty-one classifiers used in the same environments as Nasioi (Onishi 1994), in addition to five noun classes (masculine, feminine, diminutive, local, and manner) realized through agreement with articles, demonstratives, and some adjectives in a noun phrase, and with a topical constituent on the verb. Anamuxra (Pomoikan, Madang province: Ingram 2003) has several dozen classifiers as suffixes on post-head modifiers (adjectives, numerals, demonstratives, articles, and the general question word), the possessive marker, the vocative, and most nouns. Unusual multiple classifier systems are attested in Awará and Wantoat (Davis n.d; Susan Quigley, p.c.), from the Finisterre family; Angan languages, and Santa Cruzan languages in the Solomons (Aikhenvald 2000: 218) Classificatory verbs which combine reference to the orientation of the noun and to its inherent properties are widespread in the languages of the New Guinea Highlands, especially Engan (H5), Chimbu, and East New Guinea Highlands families (Foley 1986: 89 ff.), and also languages of the Border family (P1: Brown 1981; Aikhenvald 2000: 170–1). Enga (Lang 1975; Foley 1986: 89–91) has seven classificatory verbs. Men and other referents judged to be tall, large, strong, and powerful (such as houses and trees) are classified as “standing”, while women, possums, and referents judged to be small, squat, horizontal, and weak “sit”. Protruding objects “hang”, subterranean creatures and objects (such as worms or sweet potatoes, as well as internal organs, such as the heart) “lie inside”. Liquids and gases “come”; crawling or aquatic creatures, as well as orifices and locations, “lie”, and genitalia “carry”. Ku Waru (Merlan et al. 1997: 75) has a similar, albeit smaller, system of classificatory existential verbs. Waris (Border family, P1: Brown 1981) has verbal classifiers which categorize the direct object in terms of its shape and size, in addition to “stance”


255

classificatory verbs used to categorize the intransitive subject as to its orientation in space. Prefixed verbal classifiers in Waris derive historically from compounded verbs (as they do in the closely related Imonda: Seiler 1985). A few languages have rather unusual evidentiality systems. Oksapmin, an isolate from Sandaun province (Lawrence 1987), distinguishes visual, nonvisual, and reported information sources. The systems in Foe and Fasu, from the Kutubuan family in the Southern Highlands, are among the most complex in the world, with at least six distinctions (Rule 1977: 71–4; May and Loeweke 1980: 71–4; Aikhenvald 2004a).

13.3.3. Syntax Table 13.2 features a comparison of typical syntactic features of Austronesian languages contrasted with those found in most Papuan languages (Foley 1998: 513–16). Clause chaining is a prominent syntactic property of the Papuan languages. It strongly correlates with their verb-final constituent order (atypical for Austronesian languages: see Foley 2000b: 385). A great number of languages distinguish between independent (“final”) and dependent (“medial”) verbs. Final verbs have a full range of inflectional possibilities, such as pronominal agreement, tense, aspect, and mood. Medial verbs typically occur before the final verbs, and have fewer inflectional possibilities. A medial verb carries a special morpheme indicating whether its subject is the same as that of the final verb, or different from it. This is known as switch reference. Medial verbs typically distinguish relative tense (sequential versus simultaneous forms).

Table 13.2. Some syntactic features of Austronesian and of Papuan languages: a comparison Syntactic feature

Austronesian

Papuan

Constituent order Prepositions or postpositions Determiners

Verb medial Prepositions Determiners precede noun within a noun phrase Follow the head noun

Predominantly verb final Postpositions No determiners

Position of modifiers Clause linking

Coordinating or subordinating conjunctions

Either follow or precede the head noun Clause chaining

256


Switch reference markers are typically suffixed to the verb (but in a few Angan languages in the Highlands some switch reference markers are suffixes, and others are prefixes). Switch reference is found throughout the Trans New Guinea area, in numerous languages of the Sepik area and of Madang, as well as in the languages of Bougainville (Onishi 1994). It is absent from Lakes Plain (P2), Skou (S7), Lower Sepik–Ramu (S1–2), and Torricelli languages (S8), as well as from most languages in the Bismarck Archipelago. And we have mentioned above that Austronesian languages from the Bel family, in Madang province, have developed the distinction between medial and final verbs, and even switch reference, under the impact of neighbouring Papuan languages. No other Austronesian languages have switch reference. For an upto-date overview of switch reference in Papuan languages, see Roberts (1997). This sample of the unusual features of Papuan languages illustrates, in a nutshell, their considerable typological diversity.

13.3.4. Documentation The major centres for the description and analysis of languages of the New Guinea area are the Summer Institute of Linguistics (Ukarumpa, Papua New Guinea), the Research School of Pacific and Asian Studies in Canberra (Departments of Linguistics and of Anthropology), the Department of Linguistics at the University of Sydney, the University of Leiden (with major concentration on the languages of Indonesian Papua and the islands off the coast of New Guinea), and the Research Centre for Linguistic Typology at La Trobe University in Melbourne. Pacific Linguistics continues to be the major publisher for grammars, dictionaries, and other materials on New Guinea languages. In spite of the large amount of materials produced (see the bibliography in Carrington 1996), only a few score non-Austronesian languages have been well documented. Among these are Alamblak, Amele, Arapesh, Dani, Hua, Kewa, Kilmeri, Korafe, Marind, Usan, Yimas, and a few others. Several scholars are active in the area of Skou languages, and Papuan languages of the Solomons. But the majority of languages are still in urgent need of documentation.

13.4. Linguistic Diversity and Language Endangerment in Papua New Guinea We will now briefly discuss problems of language endangerment in Papua New Guinea. There are few documented instances of the “sudden death”


257

of a language (when all the speakers are exterminated or die of a disease: Campbell and Muntzel 1989), or of “radical death” whereby a rapid language loss is associated with genocide or severe political repression.5 Sudden death of Mansim, due to disease and intertribal warfare, was discussed by Reesink (2002a). Most instances of language obsolescence are those of “gradual death”, that is, the loss of a language due to a shift to the dominant language with an intermediary stage of bilingualism, and subsequent contraction in the sphere of usage of the erstwhile vernacular. Numerous communities in the Sepik area with a small number of speakers have effectively undergone language shift. Children tend to acquire Tok Pisin rather than the vernacular as their first language, and full competence in the vernacular is only found among adults. A classic case of such shift is Taiap (an isolate spoken in Gapun village, ESP, by about 100 people) documented by Kulick (1992). Language socialization in Taiap involves the conceptualization of Tok Pisin as a symbol of modernity and sought-after prosperity, while the vernacular is associated with “backwardness”. A similar example is Yimas (Lower Sepik: Foley 1991: 4–6) spoken by about 250 people in two villages (and further examples in Loving 1980). Languages with fewer than fifty speakers have hardly any chance of survival (see Smith 1992, on the imminent extinction of Susuami, an Angan language from Morobe province). One of the reasons for the vitality of the Mehek language in Sandaun province (Bugenhagen 1980: 93) is that, with over 6,000 speakers, it is numerically the dominant language in the area. The political situation in the Indonesian province of Papua is hardly promising for the survival of the vernaculars. Processes of rapid language shift have been observed for languages with seemingly large numbers of speakers. Murik (1,200, Lower Sepik) is not being learnt by children, and neither is Abu’ Arapesh, an ethnic group of over 5,000 (S8; Nekitel 1985). A similar situation has been described for Cemaun Arapesh (Lise Dobrin, p.c.) and Makopin Arapesh (Nidue 1990: 65–6) (see Aikhenvald 2004b, for further examples). Language prestige is another factor. Smith (1992)—in his discussion of the apparently healthy situation of Musom, an Austronesian language of Morobe province—stresses the importance of pride in one’s language and identity, and “an ideology that newcomers must adapt to village norms of language and culture” (1992: 119). In contrast, speakers of Kuot, which is

5 Indications are that, before European contact, intertribal warfare did result in extinction of some tribes and languages. See Harrison (1993).

258


now endangered (Lindström 2002), do not attach any particular importance to their language. Even if the language is acquired by children, its acquisition may be incomplete. For instance, the knowledge of a ritual “pandanus” language used by a number of peoples of the Southern Highlands province during the harvest of pandanus nuts has decreased during the past thirty–forty years, as Franklin and Stefaniw (1992) report for Kewa and Imbongu (Kewa has over 40,000 speakers, and Imbongu has 16,000; neither of these languages is in immediate danger of becoming extinct). Stylistic reduction often pre-dates language obsolescence. For instance, the oratorical style sesade kwanif associated with the ritual food exchange in Abu’ Arapesh was dying out in the 1960s, long before children stopped acquiring the language (Nekitel 1985: 182). Similar observations on Ilahita Arapesh can be found in Tuzin (1976). Also see Foley (1991, 1997b) on Yimas and Watam, Ingram (2003, 2006) for Dumo and Anamuxra, and Aikhenvald (2004b) for Manambu. Balanced and stable di- or triglossia involves Tok Pisin and often also English as the languages of government, local council, missions, and schooling, with the vernacular used in day-to-day communication in other circumstances (including the domestic). The tendency towards a triglossic situation in Papua New Guinea was first identified by Sankoff (1980: 35). If the di- or triglossic situation is stable, the vernacular is not endangered, as appears to be the case with Kilmeri (Border family: Brown 1980 and Claudia Gerstner-Link, p.c.), and Kaki Ae (Eleman family, Gulf province: Clifton 1994). See further examples in Loving (1980). In a situation where prestige and economic opportunities come to be associated with proficiency in Tok Pisin, diglossic relationships between languages often become unstable. Wom (Torricelli family: Moeckel and Moeckel 1980) had 2,500 speakers spread over five villages. Besides diglossic relationships with Tok Pisin, the Wom-speaking villages preserved traditional patterns of bilingualism in Urat (Southern Arapesh) and Bumbita Arapesh. But most speakers regard Tok Pisin and now often English as the language of economic opportunity. As a result, numerous parents preferred speaking Tok Pisin to their children—this is indicative of a tendency towards destabilization, at least in the long run. Destabilized diglossia is becoming the norm in many areas of Papua New Guinea, including the Sepik. This is conducive to increasing endangerment of the vernaculars. What factors enhance the vitality of indigenous languages? A wellplanned and well-executed teaching programme at the primary school level is extremely important for language maintenance (Nidue 1990 considers this


259

crucial for the survival of Makopin Arapesh). Watam (Lower Ramu: Foley 1997b), spoken by 700 people in three villages, is used by all generations. The fact that all speakers are bilingual in Tok Pisin does not appear to affect Watam’s vitality. Indeed, the Watam community school has been opened to the children of a neighbouring Kopar-speaking village (Kopar, from the Lower Sepik family, no longer has any real speakers under 50). These children are now bilingual in Watam. Watam is thus expanding at the expense of Kopar. The Summer Institute of Linguistics is playing an important role in promoting literacy and enhancing the prestige of indigenous languages (see Cahill 1999), having produced literacy and other materials on over 100 languages. Dixon (1991: 246–7) reports how the Urat language (Torricelli family) in the East Sepik province underwent a strong revival as a result of the work of the SIL team, Robert and Dawn Barnes. The vitality of some indigenous languages is enhanced by the maintenance of traditional patterns of subsistence agriculture. In such cases, there is no motivation to drop the vernacular and just speak Tok Pisin and/or English (see Bugenhagen 1980: 93, on Mehek and Siliput). In contrast, communities which are dependent on economic transactions involving Tok Pisin are in danger of losing their language (cf. Landweer 2000: 15). Emblematicity of a language can enhance its chances for survival. Anêm is a case in point. This Papuan language, markedly different from Austronesian languages in the area, is a mark of ethnic distinctiveness, and is being maintained as such (Thurston 1992). Most Austronesian languages in the area are too similar to each other to serve as proper markers of ethnicity; as a result, people have a low degree of language loyalty, and the languages are in danger of disappearing. There is additional cause for optimism in the decision of the Papua New Guinea government to support “tok ples” (lit. “language (of) village”) programmes, or “vernacular based pre-primary education” (Dixon 1991: 247; Nekitel 1998: 179–80). (These programmes allow children to acquire basic literacy skills in their first language before moving on to English language education in later primary school.) By the end of 1991, the number of provincial vernacular schools amounted to ninety-four, with programmes in twenty-one languages (Wurm 2001: 389), scattered across most provinces. Since then, more active vernacular teaching programs have been implemented. (The latest figures, for mid 2005, are of over 4,000 schools: Joan Kale, p.c.). Tok Ples Elementary programmes for children in the first three

260


years of school are now available in densely populated parts of East New Britain and across Sandaun and Madang provinces. As funding allows, this programme is being rolled out into other areas. Communities across the country have been enthusiastic in their support for this programme. Funding limitations are hampering its expansion into less developed areas. Additional complications arise in the areas where schools have to service a number of different language communities, and are compelled to choose one indigenous language over another (as in the case of Kopar and Watam mentioned above), or Tok Pisin. Establishing a comprehensive language education programme across Papua New Guinea requires a dramatic increase in the number of schools. These education programmes, together with urgent documentation of hitherto undocumented or poorly described languages, are the most compelling tasks for those who have the future of New Guinean languages at heart.

References Aikhenvald, A. Y. (1998). “Physical properties in a gender system: an example from Manambu, a Ndu language of New Guinea”, Language and Linguistics in Melanesia, 27: 175–87. —— (2000). Classifiers: A Typology of Noun Categorisation Devices. Oxford: Oxford University Press. —— (2004a). Evidentiality. Oxford: Oxford University Press. —— (2004b). “Language endangerment in the Sepik area of Papua New Guinea”, in O. Sakiyama and F. Endo (eds.), Lectures on Endangered Languages, v: From Tokyo and Kyoto Conferences 2002. ELPR C-005, 97–142. —— with Ala, Jacklyn Yuamali Benji, and Laki, Pauline L. (Forthcoming). The Manambu Language from East Sepik, Papua New Guinea. Anderson, N., and Wade, M. (1988). “Ergativity and control in Folopa”, Language and Linguistics in Melanesia, 19: 1–16. Bellwood, P. S. (1998). “From Bird’s Head to bird’s eye view; long term structures and trends in Indo-Pacific prehistory”, in J. Miedema et al. (eds.): 951–78. Blust, R. A. (1993). “Central and Central-Eastern Malayo-Polynesian”, Oceanic Linguistics, 32: 241–94. Briley, D. (1997). “Four grammatical marking systems in Bauzi”, in K. Franklin (ed.), Papers in Papuan Linguistics 2. Canberra: Pacific Linguistics, 1–131. Brown, R. (1980). “A sociolinguistic survey of Pagi and Kilmeri”, in R. Loving (ed.): 193–206. —— (1981). “Semantic aspects of some Waris predications”, in K. J. Franklin (ed.), Syntax and Semantics in Papua New Guinea Languages. Ukarumpa: Summer Instit0ute of Linguistics, 93–123. Bruce, L. (1984). The Alamblak Language of Papua New Guinea East Sepik. Canberra: Pacific Linguistics.


261

Bugenhagen, R. D. (1980). “A sociolinguistic survey of Mehek and Siliput”, in R. Loving (ed.): 79–108. Cahill, M. (1999). “From endangered to less endangered: case histories from Brazil and Papua New Guinea”, Notes on Sociolinguistics, 5: 23–34. Campbell, L. (1997). “Amerindian personal pronouns: a second opinion”, Language, 73: 339–51. —— and Muntzel, M. (1989). “The structural consequences of language death”, in N. Dorian (ed.), Investigating Obsolescence: Studies in Language Contraction and Death. Cambridge: Cambridge University Press, 187–96. Carrington, L. (1996). A Linguistic Bibliography of the New Guinea Area. Canberra: Pacific Linguistics. Chowning, A. (1996). “Relations among languages of West New Britain: an assessment of recent theories and evidence”, in M. D. Ross (ed.): 7–62. Christensen, S. (1995). Yonggom Grammar Essentials. Ukarumpa: Summer Institute of Linguistics. Clifton, J. (1994). “Stable multilingualism in a small language group: the case of Kaki Ae”, Language and Linguistics in Melanesia, 25: 107–24. Clouse, D. A. (1997). “Toward a reconstruction and reclassification of the Lakes Plain languages of Irian Jaya”, in K. Franklin (ed.), Papers in Papuan Linguistics 2. Canberra: Pacific Linguistics, 133–236. Conrad, R. (1978). “A survey of the Arapesh language family of Papua New Guinea”, Workpapers in Papua New Guinea Languages, 25: 57–77. —— with Wogiga, Kepas (1991). An Outline of Bukiyip Grammar. Canberra: Pacific Linguistics. Davis, D. R. (n.d.). Noun Class Markers in Wantoat. Ukarumpa: Summer Institute of Linguistics. Dixon, R. M. W. (1991). “The endangered languages of Australia, Indonesia and Oceania”, in R. H. Robins and E. M. Uhlenbeck (eds.), Endangered Languages. Oxford: Berg, 229–55. Dobrin, L. (1999). “Phonological form, morphological class and syntactic gender: the noun class systems of Papua New Guinea Arapeshan”, Ph.D. dissertation, University of Chicago. Dol, P. (1999). “A grammar of Maybrat: A language from the Bird’s Head, Irian Jaya, Indonesia”, Ph.D. thesis, University of Leiden. Donohue, M. (1997). “Tone systems in New Guinea”, Linguistic Typology, 1: 347–86. —— (2002a). “Which sounds change: descent and borrowing in the Skou family”, Oceanic Linguistics, 41: 171–221. —— (2002b). Papuan Languages. Materials for the course taught in July 2002 at ALI, Macquarie University. —— (forthcoming). A Grammar of the Skou Language of New Guinea. Drabbe, P. (1955). Spraakkunst van het Marind zuidkust Nederlands Nieuw-Guinea. Studia Instituti Anthropos 11. Vienna: Missiehuis St Gabriel. Dunn, M., Reesink, G., and Terrill, A. (2002). “The East Papuan languages: a preliminary typological appraisal”, Oceanic Linguistics, 41: 28–62. Dutton, T. E. (ed.) (1992). Culture Change, Language Change: Case Studies from Melanesia. Canberra: Pacific Linguistics. —— (1997). “Hiri Motu”, in S. G. Thomason (ed.), Contact Languages: A Wider Perspective. Amsterdam: John Benjamins, 9–41.

262


Edmondson, J. A., Bateman, J., and Miehle, H. (1992). “Tone contours and tone clusters in Iau”, BLS 18: 92–103. Feldman, H. (1986). A Grammar of Awtuw. Canberra: Pacific Linguistics. Foley, W. A. (1986). The Papuan Languages of New Guinea. Cambridge: Cambridge University Press. —— (1988). “Language birth: the processes of pidginisation and creolisation”, in F. J. Newmeyer (ed.), Linguistics: The Cambridge Survey, iv: Language: The SocioCultural Survey. Cambridge: Cambridge University Press, 162–84. —— (1991). The Yimas Language of New Guinea. Stanford, Calif.: Stanford University Press. —— (1994). “New Guinea languages”, in R. E. Asher (ed.), The Encyclopedia of Language and linguistics. Oxford: Pergamon Press, iii. 86–92. —— (1997a). “Papuan languages”, (www.arts.usyd.edu.au/departs/linguistics/ling/ research/papuan.html), 1, 2. —— (1997b). “Watam”, www.sultry.arts.usyd.edu.au/ling/research/watam.html1,2. —— (1998). “Toward understanding Papuan languages”, in J. Miedema et al. (eds.): 503–18. —— (1999). “Grammatical relations, information structure, and constituency in Watam”, Oceanic Linguistics, 38: 115–38. —— (2000a). “Linguistic prehistory in the Sepik–Ramu basin”, paper delivered at the conference Papuan Pasts, RSPAS, ANU, Canberra. —— (2000b). “The languages of New Guinea”, Annual Review of Anthropology, 29: 357–404. Ford, K. (n.d.). “The languages of Papua New Guinea”, MS, Department of Language and Literature, University of Papua New Guinea. Foreman, V. (1974). Grammar of Yessan-Mayo. Santa Ana, Calif.: Summer Institute of Linguistics. Fortune, R. (1942). Arapesh. Publications of the American Ethnological Society XIX. New York: J. J. Augustin. Franklin, K. J. (1971). A Grammar of Kewa, New Guinea. Canberra: Pacific Linguistics. —— and Stefaniw, R. (1992). “The ‘pandanus languages’ of the Southern Highlands Province, Papua New Guinea: a further report”, in T. E. Dutton (ed.): 1–6. French, M. A. (1994). “Solomon Islands: language situation”, in R. E. Asher (ed.), The Encyclopedia of Language and Linguistics. Oxford: Pergamon Press, viii. 4046–7. Fry, E. (1977). “Missionary lingue franche: Kuanua”, in S. A. Wurm (ed.): 865–74. Gerstner-Link, C. (2004). Das Kilmeri. Typologische Annäherung an eine Sprache Neuguineas. Puwani-river, Sandaun Province, Papua New Guinea. Habilitationsschrift, Lüdwig-Maximilan-Universität Munich: Institut für Allgemeine und Typologische Sprachwissenschaft. Gravelle, G. (2002). “Morphosyntactic properties of Meyah word classes”, in G. P. Reesink (ed.): 109–80. Grimes, B. (ed.) (2000). Ethnologue, i: Languages of the World. Dallas: Summer Institute of Linguistics International. Harrison, S. (1993). The Mask of War. Manchester: Manchester University Press. Hurd, C. (1977). “Nasioi projectives”, Oceanic Linguistics, 16: 111–78. Ingram, A. (2003). “Classifiers in Anamuxra”, Anthropological Linguistics, 45: 129–68. —— (2006). “Serial verb constructions in Dumo”, in A. Y. Aikhenvald and R. M. W. Dixon (eds.), Serial Verbs: A Cross-linguistic Typology. Oxford: Oxford University Press, 202–22.


263

—— (In press). A Grammar of Anamuxra. Canberra: Pacific Linguistics. —— (forthcoming). A grammar of Dumo. Jones, L. K. (1994). “Indonesia: language situation”, in R. E. Asher (ed.), The Encyclopedia of Language and Linguistics. Oxford: Pergamon Press, iii. 1667–9. Kulick, D. (1992). Language Shift and Cultural Reproduction: Socialisation, Self, and Syncretism in a Papua New Guinean Village. Cambridge: Cambridge University Press. Landweer, M. L. (2000). “Endangered languages: indicators of ethnolinguistic vitality”, Notes on Sociolinguistics, 5: 5–22. Lang, A. (1975). The Semantics of Classificatory Verbs in Enga and Other Papua New Guinea Languages. Canberra: Pacific Linguistics. Lawrence, M. (1987). “Viewpoint in Oksapmin”, Language and Linguistics in Melanesia, 16: 54–70. Laycock, D. (1965). The Ndu Language Family (Sepik District, New Guinea). Canberra: Linguistic Circle of Canberra Publications. Lindström, E. (2002). “Topics in the grammar of Kuot, a non-Austronesian language of New Ireland, Papua New Guinea” Ph.D. dissertation, University of Stockholm. Loving, R. (ed.) (1980). Sociolinguistic Survey of Sepik Languages. Workpapers in Papua New Guinea Languages 29. Ukarumpa: Summer Institute of Linguistics. Lynch, J. (1998). Pacific Languages. An Introduction. Honolulu: University of Hawaii Press. —— Ross, M. D., and Crowley, T. (2002). The Oceanic Languages. London: Curzon Press. McGregor, D. E., and McGregor, A. R. F. (1982). Olo Language Materials. Canberra: Pacific Linguistics. May, J., and Loeweke, E. (1980). General Grammar of Faso (Namo Me). Reprinted from Workpapers in Papua New Guinea Languages 27. Ukarumpa: Summer Institute of Linguistics. Merlan, F., and Rumsey, A. (1991). Ku Waru: Language and Segmentary Politics in the Western Nebilyer Valley, Papua New Guinea. Cambridge: Cambridge University Press. —— Roberts, S. P., and Rumsey, A. (1997). “New Guinea ‘classificatory verbs’ and Australian noun classification: a typological comparison” in M. Harvey and N. Reid (eds.), Nominal Classification in Aboriginal Australia. Amsterdam: John Benjamins, 63–103. Miedema, J., Odé, C., and Dam, R. A. C. (eds.) (1998). Perspectives on the Bird’s Head of Irian Jaya, Indonesia. Amsterdam: Rodopi. Moeckel, B., and Moeckel, B. (1980). “A sociolinguistic survey of Wom”, in R. Loving (ed.): 43–62. Nekitel, O. (1985). “Sociolinguistic aspects of Abu’: a Papuan language of the Sepik area, Papua New Guinea”, Ph.D. thesis, Australian National University. —— (1998). Voices of Yesterday, Today and Tomorrow: Language, Culture and Identity. New Delhi: UBS, Publisher Distributors Ltd. Nidue, J. A. (1990). “Language use in a New Guinea village: a triglossic profile of Makopin I”, Language and Linguistics in Melanesia, 21: 47–69. Obata, K. (2003). A Grammar of Bilua, a Papuan Language of the Solomon Islands. Canberra: Pacific Linguistics. Onishi, M. (1994). “A grammar of Motuna (Bougainville, Papua New Guinea)”, Ph.D. thesis, Australian National University.

264


Pawley, A. K. (1993). “A language which defies description by ordinary means”, in W. A. Foley (ed.), The Role of Theory in Language Description. Berlin: Mouton de Gruyter, 87–129. —— (1995). “C. L. Voorhoeve and the Trans New Guinea phylum hypothesis”, in C. Baak, M. Bakker and D. van der Meij (eds.), Tales from a Concave World: Liber amicorum Bert Voorhoeve. Leiden: Leiden University, 83–123. —— (1998). “The Trans New Guinea phylum hypothesis: a reassessment”, in J. Miedema et al. (eds.): 655–90. —— (2000). “The Trans New Guinea phylum: recent historical research and its implications”, paper presented at the conference Papuan Pasts: Investigations into the Cultural, Linguistic and Biological History of the Papuan Speaking Peoples, RSPAS, ANU, Canberra. —— and Ross, M. D. (1995). “The prehistory of Oceanic languages: a current view”, in P. Bellwood, J. Fox and D. Tryon (eds.), The Austronesians: Historical and Comparative Perspectives. Canberra: Australian National University, RSPAS, 39–74. Reesink, G. P. (1998). “The Bird’s Head as a Sprachbund”, in J. Miedema et al. (eds.): 603–42. —— (1999). A Grammar of Hatam: Bird’s Head Peninsula, Irian Jaya. Canberra: Pacific Linguistics. —— (2000). “West Papuan: roots and development”, paper presented at the conference Papuan Pasts: Investigations into the Cultural, Linguistic and Biological History of the Papuan Speaking Peoples, RSPAS, ANU, Canberra. —— (2002a). “Mansim, a lost language of the Bird’s Head”, in G. P. Reesink (ed.): 277–340. —— (2002b). “The eastern Bird’s Head languages compared”, in G. P. Reesink (ed.): 1–44. —— (ed.) (2002c). Languages of the Eastern Bird’s Head. Canberra: Pacific Linguistics. Roberts, J. R. (1997). “Switch-reference in Papua New Guinea: a preliminary survey”, in Papers in Papuan Linguistics, iii. Canberra: Pacific Linguistics, 101–241. Ross, M. D. (1987). “A contact induced morphosyntactic change in the Bel languages of Papua New Guinea”, in D. C. Laycock and W. Winter (eds.), The World of Languages: Papers Presented to Professor S. A. Wurm on his 65th Birthday. Canberra: Pacific Linguistics, 583–607. —— (1993). “Tonogenesis in the North Huon Gulf Chain”, in J. A. Edmondson and K. J. Gregerson (eds.), Tonality in Austronesian Languages. Honolulu: University of Hawaii Press, 133–53. —— (1994). “Areal phonological features in north central New Ireland”, in T. Dutton and D. T. Tryon (eds.), Language Contact and Change in the Austronesian World. Berlin: Mouton de Gruyter, 523–72. —— (1995). “The great Papuan pronoun hunt: recalibrating our sights”, in C. Baak, M. Bakker and D. van der Meij (eds.), Tales from a Concave World: Liber amicorum Bert Voorhoeve. Leiden: Leiden University, 139–68. —— (ed.) (1996). Studies in Languages of New Britain and New Ireland, i: Austronesian Languages of the North New Guinea Cluster in Northwestern New Britain. Canberra: Pacific Linguistics. —— (2000). “Using pronouns as a preliminary diagnostic for grouping Papuan languages”, paper presented at the conference Papuan Pasts: Investigations into the Cultural, Linguistic and Biological History of the Papuan Speaking Peoples, RSPAS, ANU, Canberra.


265

—— (2001). “Contact-induced change in Oceanic languages in north-west Melanesia”, in A. Y. Aikhenvald and R. M. W. Dixon (eds.), Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics. Oxford: Oxford University Press, 137–66. Rule, W. M. (1977). A Comparative Study of the Foe, Huli and Pole Languages of Papua New Guinea. Oceania Linguistic Monographs 20. Sydney: University of Sydney. Sankoff, G. (1977). “Multilingualism in Papua New Guinea”, in S. A. Wurm (ed.): 265– 307. (Also in Sankoff 1980: 95–132). —— (1980). The Social Life of Language. Philadelphia: University of Pennsylvania Press. Scott, G. (1978). The Fore Language of Papua New Guinea. Canberra: Pacific Linguistics. Seiler, W. (1985). Imonda, a Papuan Language. Canberra: ANU, Pacific Linguistics. Senft, G. (1996). Classificatory Particles in Kilivila. New York: Oxford University Press. Smith, G. P. (1992). “Language obsolescence in Morobe province, Papua New Guinea: two contrasting case studies”, in T. E. Dutton (ed.): 115–21. Staalsen, P. (1966). “The phonemes of Iatmul”, in Papers in New Guinea Linguistics 5. Canberra: Pacific Linguistics, 69–76. Stebbins, T. N. (2005). “Noun classification in Mali”, Anthropological Linguistics, 47: 77–131. —— (Forthcoming). A Grammar of Mali (Baining). Canberra: Pacific Linguistics. Terrill, A. (2002). “Systems of nominal classification in East Papuan languages”, Oceanic Linguistics, 41: 63–88. —— (2003). A Grammar of Lavukaleve, a Papuan Language of the Solomon Islands. Berlin: Mouton de Gruyter. Thurston, W. R. (1982). A Comparative Study in Anêm and Lusi. Canberra: Pacific Linguistics. —— (1987). Processes of Change in the Languages of Northwestern Britain. Canberra: Pacific Linguistics. —— (1992). “Sociolinguistic typology and other factors effecting change in northwestern New Britain, Papua New Guinea”, in T. E. Dutton (ed.): 123–38. Tuzin, D. (1976). The Ilahita Arapesh: Dimensions of Unity. Berkeley and Los Angeles: University of California Press. Wendel, T. D. (1993). “A preliminary grammar of Hanga Hundi”, MA thesis, University of Texas at Arlington. Whitehead, C. R. (1994). “Papua New Guinea: language situation”, in R. E. Asher (ed.), The Encyclopedia of Language and Linguistics, vol. vi. Oxford: Pergamon Press, 2924–6. Wilson, P. (1980). Ambulas Grammar. Workpapers in Papua New Guinea Languages 26. Ukarumpa: Summer Institute of Linguistics. Wurm, S. A. (1975a). “The application of the comparative method to Papuan languages: general and highlands”, in S. A. Wurm (ed.), New Guinea Area Languages and Language Study, i: Papuan Languages and the New Guinea Linguistic Scene. Canberra: Pacific Linguistics, 237–62. —— (1975b). “The East Papuan phylum in general”, in S. A. Wurm (ed.), New Guinea Area Languages and Language Study, i: Papuan Languages and the New Guinea Linguistic Scene. Canberra: Pacific Linguistics, 783–807. —— (1977a). “Missionary lingue franche: Kiwai”, in S. A. Wurm (ed.): 893–906. —— (ed.) (1977b). New Guinea Area Languages and Language Study, iii: Language, Culture, Society, and the Modern World. Fascicle 2. Canberra: Pacific Linguistics. —— (1982). Pacific Languages of Oceania. Tübingen: Gunter Narr Verlag.

266


Wurm, S. A. (1992). “Change of language structure and typology in a Pacific language as a result of culture change”, in T. E. Dutton (ed.): 149–65. —— (2001). “Language endangerment in the insular Greater Pacific area, and the New Guinea area in particular”, in A. Pawley, M. Ross and D. Tryon (eds.), The Boy from Bundaberg: Studies in Melanesian Linguistics in Honour of Tom Dutton. Canberra: Pacific Linguistics, 383–97.

14

Languages of the Pacific Region: Malayo-Polynesian Osamu Sakiyama

14.1. Linguistic background 14.2. Linguistic strata 14.3. Minority languages and common languages in the Pacific region 14.3.1. China 14.3.2. Malaysia 14.3.3. Indonesia 14.3.4. Belau (Palau), Micronesia 14.3.5. Yap, Micronesia 14.3.6. Propinsi Papua and Papua New Guinea, Melanesia 14.3.7. Solomon Islands, Melanesia 14.3.8. Vanuatu, Melanesia 14.3.9. New Caledonia, Melanesia 14.3.10. Polynesia 14.4. Conclusion

267 270 271 271 272 273 274 275 275 277 278 278 279 279

14.1. Linguistic Background The languages that are currently spoken in the Pacific region can be divided broadly into three groups: the Australian and New Guinean languages were formed by people who participated in the region’s earliest migrations over a period of 20,000–30,000 years starting several tens of thousands of years ago, and the Malayo-Polynesian (Austronesian) languages were spoken by Mongoloid people who migrated from the Asian continent around 3000 bc. The Pacific region has numerous languages, including 250 Aboriginal languages in Australia and 750 Papuan languages on the island of New Guinea (including Irian Jaya of the Indonesian territory, now called Propinsi Papua) and neighbouring areas. There are also 350 Malayo-Polynesian languages in Melanesia including New Guinea, twenty in Polynesia, twelve in Micronesia, and 100 in New Guinea (Comrie et al. 1996). There is wide variation not only among language groups, but also among the families of languages. Few language families have been identified among the languages of Australia and New Guinea using the methods of comparative linguistics. In the north-west Pacific, Malayo-Polynesian languages are spoken not only in Taiwan (see Ch. 15, this volume) but also over an area ranging

268

o s a m u s a k i ya m a

Figure 14.1. Malayo-Polynesian languages

from the Philippines to Indonesia, Malaysia, Indochina, and as far even as Madagascar. Nowadays the Austronesian family is used as a general term to include the indigenous languages of Taiwan, but when referring to the so-called extra-Formosan languages it is usual to use the term MalayoPolynesian. It is estimated that there are currently about 1,000 languages belonging to the Austronesian family (Blust 1990), some 6.6 times as many as the number of Indo-European languages. There is very little documentary material concerning these languages from the period before European colonialization, apart from a few exceptions such as epigraphic documents. Even today, perhaps no more than one-third of the 1,000 Austronesian languages have all three of grammars, dictionaries, and texts available. For many of them we do not even have simple word lists. Malay inscriptions have been found dating as far back as the seventh century ad, and its role as an interisland lingua franca was of great importance. It is thought that there are well over 100 million speakers of the language, mainly in Malaysia and Indonesia (Gordon 2005). Malay was an important medium of communication among various peoples for

m a l ay o - p o l y n e s i a n l a n g u a g e s

269

over a millennium, and it did not come by this role by virtue of being the language of a colonizing nation or of an ethnic majority, but as the result of very peaceful development. Ironically, the present “globalization” of this language is contributing to increasing the number of languages in danger of extinction. It is unfortunate, moreover, that “problems of bilingualism or rather multilingualism in Indonesia have not attracted much interest from linguists” (Uhlenbeck 1971). There has not been much change in the state of research in this field since then. The situation seen in Malaysia and Indonesia has also been observed in recent years in Oceania, where the growth of Melanesian Pidgin has driven languages with small speaking populations to the brink of extinction. In the People’s Republic of China and in Taiwan, too, Chinese (the language of the Han ethnic majority) is similarly endangering the survival of the languages of ethnic minorities, apart from fifty-five designated languages in the PRC that enjoy various privileges (see also § 17.4). Languages in the Pacific region are also characterized by the small size of speaker populations and by the absence of dominant languages. However, there are usually bilingual people who can speak or at least understand the languages of neighbouring peoples, and it is believed that this situation has existed for a long time. In terms of cultural factors, it appears that linguistic peculiarities in the Pacific region were emphasized for the sake of identity, and over the millennia the emblematic function of language has generated differences (Comrie et al. 1996). The languages of New Guinea and the region around it show diverse linkages and wide variations between languages. The Malayo-Polynesian languages of the Pacific region are mostly classified as Oceanic languages, while the Chamorro and Palauan of Micronesia are classified into the languages of Western Malayo-Polynesian (WMP, previously called the Indonesian family), and the indigenous languages of Propinsi Maluku and Propinsi Papua in Eastern Indonesia into the Central Malayo-Polynesian (CMP) and the South Halmahera-West New Guinea (SHWNG) subgroups. In particular, there are strong similarities between the linguistic characteristics of the CMP and SHWNG languages and those of the Melanesian family of the Oceanic languages. These linguistic conditions and characteristics are attributable to ethnic migrations within the region over a long period of time, accompanied by contacts and linguistic interference with indigenous non-Austronesian or Papuan languages. Papuan languages are still found in parts of Indonesia, including Northern Halmahera and the islands of Pantar

270


and Alor and central and eastern Timor in Propinsi Nusa Tenggara. In New Guinea, contact with Papuan languages has caused some Malayo-Polynesian languages to exhibit a word order change from Subject-Verb-Object to Subject-Object-Verb (see also § 13.2.1).

14.2. Linguistic Strata With the start of colonization by the European powers in the nineteenth century, a new set of linguistic circumstances developed in the region. First, pidgin languages based on European and Melanesian languages gradually emerged as common languages. The establishment of plantations in Samoa and in Queensland, Australia, which had concentrations of people who spoke Melanesian languages, was important in providing breeding grounds for pidgin languages. A pidgin language is formed from elements of the grammar of both contributing languages, though the pidgin languages tend to be looked down upon from the perspective of the more dominant of the two parent languages (see Ch. 7, this volume). The region’s newly formed common languages, including Tok Pisin, Bislama, and Solomon Pijin (Pidgin), flourished after they were taken back to the homelands of the various speakers. This was possible because Vanuatu, the Solomon Islands, and Papua New Guinea were all multilingual societies without dominant languages. The number of speakers of pidgin languages increased rapidly in this environment. At the same time, the continuing existence of ethnic minority languages came under threat. Examples of pidgins that were creolized (adopted as mother languages in their own right) include Solomon Pijin, which eventually had over 1,000 speakers aged 5 and over (1976) in the Solomon Islands. Bislama, overspreading about 100 indigenous languages and the former official languages of English and French, is now spoken by almost the entire population of Vanuatu (170,000 in 1996) and is partially creolized. Of particular interest is the fact that a group of more than 1,000 people who emigrated to New Caledonia have adopted Bislama as their first language. The situation in Papua New Guinea, which has a population of 4.3 million (1996), is even more dramatic. Although exact figures are difficult to establish, by 1982 the number of people using Tok Pisin as their first language had reached 50,000, while another four million used it as a second language (Gordon 2005).


271

14.3. Minority Languages and Common Languages in the Pacific Region The revised edition of the Atlas of the World’s Languages in Danger of Disappearing published by UNESCO (Wurm 2001) provides a brief overview of the current situation in the Greater Pacific Area including Japan, Taiwan, Philippines, Malaysia, Indonesia, Papua New Guinea, Solomon Islands, Vanuatu, New Caledonia (Loyalty Islands), Fiji (Rotuma), Micronesia, Polynesia, and Australia. It is a pity that language names in this atlas are missing on the page opposite to the Pacific language map (p. 65); no meaningful discussion of the map is possible. According to Lynch, over 160 languages in Papua New Guinea, the Solomon Islands, and Vanuatu, with fewer than 200 speakers each, are on the edge of extinction (1998). The following reports cover areas and endangered languages that I have researched.

14.3.1. China Ten languages are spoken on the island of Hainan off the south-eastern coast of China’s mainland. The languages with the largest speaker populations are Hainanese (the common language of the island’s inhabitants) and the Li language (which has the longest history and belongs to the Kadai stock of the Tai-Kadai phylum) with over 1.1 million speakers (see also § 16.2.2). With one exception all the other languages have speaker populations numbering tens of thousands. The exception, called Huihui-hua in Chinese, is the only Malayo-Polynesian language on the island and has about 5,000 speakers living in two villages. The language is tonal, with a monosyllabic structure, and was once classified as belonging to the Malayo-Sino family (Ni Dabai 1988). However, it seems certain that it belongs to the Chamic branch of Western Malayo-Polynesian (see also § 16.2.1). We may conclude that its tonal feature is the result of contact with and linguistic interference from the surrounding monosyllabic languages of Hainan following the tenth-century Cham migration to Hainan from the Indo-Chinese coast facing the island. “Huihui” is not really an appropriate name for the language since it is the name of only one of the two villages where it is spoken. In English it is simply called “Tsat”. Pronouced /caanʔ/, the word derives from the Chinese zhan, denoting the Cham (Champa), so this is also not an appropriate name (Gordon 2005). The speakers themselves call it “Poi Tsat” (“speech/speak” + Cham, “the Cham language”), and the term they use to refer to themselves is “U Tsat”

272


(Proto Malayo-Polynesian *uRang “people” + Cham). In 1951 the population had dropped to between 1,000 and 1,500 people, but it has been significantly revitalized since then. Poi Tsat speakers live in a society characterized by multilevel language use and their language, as a vernacular spoken only within their villages, is at the bottom of the hierarchy, with Chinese (the official language), Hainanese (the language of the marketplace), and Arabic (the language of their religion) ranked above. They are thus forced to constantly switch linguistic codes and their attitude to Poi Tsat is neutral or even negative, perhaps partly because they have no script to represent their language (Sakiyama 1997). The language has great importance as an example of an Austronesian language that has undergone extreme changes, situated as it is on the periphery of the region. We must not be complacent about its future. It is to be welcomed that there is one report on Poi Tsat, with an account of the grammar, a comparison with the Chamic languages, basic word list, and sample phrases, although it is of limited accessibility since it is in Chinese (Zheng 1997).

14.3.2. Malaysia It is estimated that the eighteen indigenous ethnic groups in Malaysia number about 93,000, constituting 0.5 per cent of the population. The government agency responsible for the affairs of indigenous peoples (the Department of Orang Asli Affairs) looks after the education and welfare of this indigenous population, and addresses any problems that may arise. However, the peoples designated as indigenous are restricted to those located within the Malay Peninsula and exclude those in East Malaysia. As regards the ethnic groups making up the indigenous population, the oldest stratum is made up of a number of Negrito communities. These are followed by several Austroasiatic Senoi communities, and finally the Jakun, Temuan, Seletar, Orang Kuala, and Orang Kanaq peoples (see also § 16.3.1). The latter peoples are Malayo-Polynesian in terms of language and of Proto Malay stock, the ancestors of modern Malays. The Semelai, alone of the Proto Malay communities, are linguistically not Malayo-Polynesian but Austroasiatic. According to official figures, the Orang Kanaq and Orang Kuala (with populations of 65 and a little over 2,000 respectively) are numerically in the most critical state among the Malayo-Polynesian speakers (JHEOA 2003). The authorities have as their priority the social welfare of the indigenous population and their integration into Malay society, and although some measures such as the building of a museum to preserve their cultures are being taken, no particular steps are being taken to record and


273

preserve their languages. The situation differs from that in China, where for designated ethnic minorities local education is conducted bilingually in Chinese and their mother tongue (Okamoto 1999).

14.3.3. Indonesia Indonesia is a multilingual state. Article 36 of its constitution stipulates that the national language is Bahasa Indonesia, but that where inhabitants of an area have their own language and it is preserved in good condition, this language will be respected and maintained by the state. Currently, however, only major languages like Javanese, Sundanese, Madurese, and Balinese, which have speaker populations in the millions and tens of millions, enjoy such benefits as having elementary schools where education is carried out in the local language as well as Bahasa Indonesia and having numerous publications in the language. The over 500 languages said to be extant in Indonesia, including Propinsi Papua, include languages on the point of extinction. Among these are Lom, spoken on the island of Bangka near Sumatra (2–10 speakers, Wurm 2000) and Dampal, spoken on Sulawesi (1–2 speakers, Wurm 2000). No special action is being taken in connection with these critically endangered languages. The Education and Culture Ministry’s Center for Language Construction and Development is, however, undertaking surveys of regional languages as well as working for the diffusion of Bahasa Indonesia. In this connection, the country was divided into ten regions in 1974 under the Second Five Year Plan, with a further five regions designated in 1983. A considerable number of reports have been published by the Center. (Unfortunately these are not for sale, so they are difficult to acquire by official means.) Many of these are on the major languages with large speaker populations, although it should be noted that they include reports on regional aspects of Malay, which make a useful contribution to the study of language dynamics. Also included is a report on Bajau, with around 3,000 to 3,500 speakers on the island of Laut to the south-east of Kalimantan, which is a contribution to research on endangered languages (Djebar Hapip et al. 1979). The Bajau live on the sea (they are sometimes referred to as “sea nomads”), although some have started to establish themselves on land. It is therefore difficult to form an accurate picture of the state of their language. According to the report, almost all Bajau adults speak Bahasa Indonesia and Banjarese, the dominant language around them. Bajau children only speak Bajau at home. Outside, they use Banjarese even among themselves.

274


Fortunately data on two endangered languages with less than 1,000 speakers in the southern part of Sulawesi (Propinsi Sulawesi Selatan) have been published based on recent surveys: Topoiyo (400–500 speakers, Wurm 2000) and Panasuan (750 speakers, Wurm 2000). The data include basic vocabularies with English translations and simple accounts of their grammars (Nurhayati et al. 2003; Manda et al. 2003). As regards the languages of southeastern Sulawesi, basic vocabularies of 226 words each have been published for twenty-seven languages (Mead 1999). These include a number of languages with small speaker populations, such as Taloki (500 speakers), Koroni (500), Bahonsuai (200), Tomadino (66), and Waru (350). The area in Indonesia with the next largest number of languages after Propinsi Papua is Propinsi Maluku. The Atlas Bahasa Tanah Maluku (Taber et al. 1996) covers 117 vernaculars (Malayo-Polynesian and Papuan), including numbers of speakers for each language, areas of habitation and migration, access routes, simple cultural information, and basic numbers and expressions. This work is especially valuable since it corrects inaccuracies and errors in the 1977 Classification and Index by Voegelin and Voegelin. It also distinguishes languages and dialects according to their a priori mutual intelligibility. Fifteen languages are listed as having fewer than 1,000 speakers. They include the Nakaela language of Seram, which has only five speakers, the Amahai and Paulohi languages, also of Seram, which are spoken by fifty people each, and the South Nuaulu and Fatamanue (hitherto called Yalahatan) languages, which have 1,000 speakers each on Seram Island. The data, however, are not complete. For example, Bajau languages are not included, presumably because of the difficulty of accessing the various solitary islands where the Bajau people live. The author researched the Fatamanue language spoken in two villages, Yalahatan and Haruru, in 1997 and in 1998 (Sakiyama 1999), and the Bajau language (2,000 speakers) on Sangkuwang Island to the east of Bacan Island in 1997.

14.3.4. Belau (Palau), Micronesia According to Belau (Palau) government statistics (1990), the total population of 15,122 people includes sixty-one people living on outlying islands in Sonsorol State, and thirty-three in Hatohobei (Tochobei) State. Apart from the Sonsorol Islands, Sonsorol State also includes the islands of Fanah, Meril, and Pulo An. In addition to the Hatohobei language, the languages scattered on these isolated islands also include Nuclear Micronesian (Chuukic or Trukic) languages, which are mainly spoken in the Carolines. They differ


275

from Palauan, which is a branch of WMP. To lump these languages together as the Sonsorol languages with a total of 600 speakers (Tryon 2000) is as inaccurate as combining the Miyako dialects of Okinawa into a single dialect group. The number of Chuukic speakers has declined steadily since these figures were compiled. Starting in the German colonial period of the early twentieth century, people have been relocated from these outlying islands to Echang village on Arakabesan Island in Belau. Today there are several hundred of these people. Many of those born in the new location only speak Palauan. A study by Oda (1977) estimated that there were fifty speakers of Pulo Annian. The language of Merir continued to decline and has now become extinct.

14.3.5. Yap, Micronesia Ngulu Atoll is situated between the Yap Islands and the Belau Islands. The Nguluwan language is a mixture of Yapese and Ulithian, and belongs to the Chuukic family. It has inherited the Ulithian phonemic system and a partial version of Yapese grammar (Sakiyama 1982). Nguluwan appears to have evolved through bilingualism between Yapese and Ulithian. In 1980 there were twenty-eight speakers. Even with the inclusion of people who had migrated to Guror village, which is the traditional trading counterpart on the Yap Islands, the number of speakers was fewer than fifty. Speakers are being assimilated rapidly into the Yapese language and culture.

14.3.6. Propinsi Papua and Papua New Guinea, Melanesia Detailed information about the names, numbers of speakers, and research data for over 800 Malayo-Polynesian and Papuan languages spoken in New Guinea and its coastal regions can be found in the works of Barr and Barr (1978), Voorhoeve (1975), and Wurm (1982). In Papua New Guinea, Nekitel (1998) reckoned languages with less than 1,000 speakers constitute more than half the number (417), including six languages (MalayoPolynesian: Getmate, Kaniet, Karore, Ahi/Lae, and Papuan: Karami, Mulaha) extinct since 1950. These six languages have either vanished leaving us with virtually no information about them, or else they are in the process of vanishing (Carrington 1996). For Lae, speaker counts were taken twice: No. 6 pegged Ahi/Lae and Marobe (sic) at no speakers; No. 16, meanwhile, listed Lae/Marobe as having ten speakers (Nekitel 1998). No. 6 is probably incorrect. Ahi, spoken at the eastern end of the Markham Valley,

276


is also called Lahe or Aribwatsa. Only one speaker of this language remains, but there are said to be concerted efforts by the direct descendants of this last speaker to at least partially revive it (Wurm 1997). It seems likely that the number of speakers reported in count No. 16 takes this circumstance into account. The sudden death of Professor Otto I. M. S. Nekitel is all the more regrettable as he was planning a survey of Lae in the Aribwatsaspeaking area deep in Morobe province. For the present, not only the minority languages but even the majority languages, other than a few, have yet to be surveyed and researched adequately. There are many languages for which vocabulary collection has yet to be undertaken. It appears that dictionaries or grammars have been published for less than one-tenth of the region’s languages. Many speakers of various languages in Papua New Guinea—although their numbers are perhaps not yet so reduced as to be critical—are switching to Tok Pisin, i.e. are becoming creolized (Dixon 1991). Ever fewer young people speak these languages as they find they have to intermarry with other ethnic groups and work in districts away from home in order to acquire an education and an occupation. Papuan languages like Koiari, Murik, and Kâte are examples of this trend (see also § 13.4). Information about the languages of Propinsi Papua in western New Guinea is in many cases scant or inaccurate. A case in point is the language of the Tobati people. The Tobati live in pile houses over the sea along the coast of Yotefa Bay in Jayapura. When I conducted fieldwork there in 1982 they numbered around 700, but information about their current numbers is confused. While one report (Tryon 2000) puts the number of speakers at 2,460, classifying “Yotafa” (sic) as “endangered”, another puts the number at 100 and considers the language (“Tobati”) to be “seriously endangered” (Wurm 2000). If the plans the Indonesian government had in 1982 to resettle such people on land for reasons of hygiene and public order have been implemented then we may surmise that their languages and cultures are being lost (Sakiyama 1989b). A grammar sketch on Tobati has been made also by Donohue (Lynch et al. 2002). There is also confusion in the information we have about the Serui people based in Yapen Island in Cenderawasih Bay. The Serui live in the east and the west of the island, the central part being occupied by a people speaking the Papuan language Yava. In addition, many Serui have moved to the Hamadi and Dok 9 districts in the environs of Jayapura, where they live in seaside communities (Sakiyama 1989a). Serui is made up of two dialects, Ambai in east Yapen and Ansus in west Yapen, but the differences between the two are slight. The Serui themselves call their language Bahasa Serui. The number


277

of speakers is estimated to be in the range of 10,000 to 15,000. The Serui are bilingual in their vernacular and Bahasa Indonesia, but children only speak Serui in the home. “Serui-Laut” (“Sea-Serui”) is what the Ansus dialect is called in some of the settlements where it is spoken. It is not necessary, therefore, to classify Serui speakers into Ambai, Ansus, and Serui-Laut communities (Voorhoeve 1975; Barr and Barr 1978). The state of the languages spoken in the Raja Ampat Islands and the size of their speaker populations was unknown (Voorhoeve 1975; Barr and Barr 1978). It has been reported, however, that the Ma’ya dialect of Salawati is tonal and has about 4,000 speakers. It has also been suggested that it should not be considered as belonging to the above-mentioned SHWNG Subgroup of Malayo-Polynesian, which exemplifies the languages of this region (Leeden 1993).

14.3.7. Solomon Islands, Melanesia The total population of the Solomon Islands is 390,000 (1996). There are sixty-seven Papuan, Melanesian, and Polynesian indigenous languages, of which only thirty-three are spoken by over 3,000 people (Gordon 2005). The Kazukuru languages (Guliguli, Doriri) of New Georgia Island, which were known to be endangered as early as 1931 and were referred to Papuan, have become extinct already, leaving behind just some scant linguistic information. The Melanesian language Laghu of Santa Isabel Island was considered to have become extinct but five speakers were subsequently discovered. Zazao (100 speakers) and Vano (Vanikolo Island) are also considered to be seriously endangered. Other Melanesian languages are in the endangered category, with around 100 speakers each, including Ririo, Kokota, Asumboa, Tanimbili, Tanema, Oroha, and Buma (Tryon 2000). Of these, the number of speakers of Ririo on Choiseul Island had at one time sunk to eighteen (Tryon 1979), but has since risen to seventy-nine (Gordon 2005). And Buma, also known Teanu, spoken on the island of Vanikoro, is sketched by Tryon (Lynch et al. 2002). But because of bilingualism with or even a shift toward Mbambatana (the regional common language), the size of the speaker population does not necessarily correspond to the state of the language. The large number of languages in Melanesia that have small speaker populations has been attributed to the fact that multilingualism has been historically common in the region for social reasons such as trade and intermarriage (Comrie et al. 1996). A survey of the Solomon Islands found that 60 per cent of men

278


and 42 per cent of women spoke at least three languages. But the fact that 83 per cent of the population speak Solomon Pijin (Solomon Islands 1992; the figure for English is 26 per cent) gives grounds to fear that minority languages with small speaker populations are being driven into a critical state.

14.3.8. Vanuatu, Melanesia The situation in Vanuatu is very similar to that in the Solomon Islands. The official view, written in Bislama, is as follows: “I gat sam ples long 110 lanwis evriwan so i gat bigfala lanwis difrens long Vanuatu. Pipol blong wan velej ol i toktok long olgeta bakegen evridei nomo long lanwis be i no Bislama, Inglis o Franis” (Vanuatu currently has 110 indigenous languages, which are all very different linguistically. On an everyday basis people in villages speak only their local languages, not Bislama, English, or French) (Institute of Pacific Studies 1980). Among the Melanesian and Polynesian indigenous languages spoken by 170,000 people, or 93 per cent of the total population (1996), there are many small minority tongues. These include Aore, which has only a single speaker (Tryon 2000; extinct, Lynch and Crowley 2001); Maragus (Tape) and Ura (with ten speakers each); Nasarian and Sowa (Seke) (with twenty); and Dixon Reef, Lorediakarkar, Mafea, and Tambotalo (with fifty). If languages with around 100 speakers are included, this category accounts for about one-half of the total number of languages. The spread of Bislama has had the effect of putting these languages in jeopardy except to make dominant vernaculars such as Uripiv and Hano (Raga) survive, although languages with no information available, or not well known at all, amount to two-thirds or more (79) of a total of 106 languages (Lynch and Crowley 2001). Among potentially endangered languages, only Vinmavis (Neve‘ei), which is spoken probably by about 500 people on the northern part of Malakula, is outlined by Crowley (Lynch et al. 2002).

14.3.9. New Caledonia, Melanesia New Caledonia has a total population of 145,000 people, of whom 62,000 are indigenous. As of 1981, there were twenty-eight languages, all Melanesian except for the one Polynesian language, Uvean. The only languages with over 2,000 speakers are Cemuhi, Paici, Ajie, and Xaracuu, along with Dehu and Nengone, which are spoken on the Loyalty Islands. Dumbea (Paita), which is spoken by several hundred people, has been described by Shintani and Païta (1983). And Osumi (1995) described Tinrin,


279

which has an estimated 400 speakers. Speakers of Tinrin are bilingual in Xaracuu or Ajie. Nere has twenty speakers and Arho ten, while Waamwang, which had three speakers in 1946, is now reported to be extinct (Gordon 2005). Descendants of Javanese, who began to migrate to New Caledonia in the early part of the twentieth century, now number several thousand. The Javanese language spoken by these people, which has developed in isolation from the Javanese homeland, has attracted attention as a new creole language.

14.3.10. Polynesia Penrhyn, a language spoken on the Northern Cook Islands with around 1,500 speakers according to a recent survey, has recently had a word list published (Shibata 2003). In the terms of the Atlas of the World’s Languages in Danger of Disappearing, its degree of endangerment corresponds to an endangered language whose “youngest speakers are young adults”. At the same time, judging by this atlas, Hawaiian (the Hawaiian Islands) and Maori (New Zealand) share this degree of endangerment, yet both appear to have escaped the danger of extinction. The revival of these languages is not something that could have been accomplished by positioning them as passive cultural indicators. At work here was the political factor that associated speaking these languages with prestige (Lynch 1998). There is food for thought here when considering the revitalization of the Ainu language in Japan.

14.4. Conclusion The Pacific has been heavily crisscrossed by human migration from ancient to modern times. All Pacific countries except the kingdom of Tonga have experienced colonization. This historical background is reflected in the existence of multilevel diglossia in all regions of the Pacific. Depending on the generation, the top level of language in Micronesia is either English (the official language) or Pidgin Japanese (used as a lingua franca among islands). The next level is made up of the languages of major islands that exist as political units, such as Palauan, Yapese, and Ponapean. On the lowest level are the various vernaculars spoken mainly on solitary islands. In the Maluku Islands of Indonesia, local Malay languages, such as Ambonese Malay, North Maluku Malay, and Bacanese Malay, form a layer beneath the official language, Bahasa Indonesia. Under them are the dominant local languages, such as Hitu, which is spoken by 15,000 people on Ambon Island, and Ternate and

280


Tidore, which are spoken in the Halmahera region. These are important as market languages. On the lowest level are the various vernaculars. In Papua New Guinea, standard English forms the top level, followed by Papua New Guinean English. Tok Pisin and Hiri Motu are used as common languages among the various ethnic groups. Beneath these layers are the regional or occupational common languages. For example, Hiri Motu is used as the law enforcement lingua franca in coastal areas around the Gulf of Papua, Yabêm as a missionary language along the coast of the Huon Gulf, and Malay as a trade language in areas along the border with Indonesia. On the next level are the ethnic and tribal languages used on a day-to-day basis. An example of a similar pattern in Polynesia can be found in Hawaii, where English and Hawaiian English rank above Da Kine Talk (also called Pidgin To Da Max for fun!—Simonson et al. 1981), which are grammatically mixtures of English and Oceanic languages and are used as common languages among the various Asian migrants who have settled in Hawaii. Beneath these are ethnic languages, including Hawaiian and the various immigrant languages, such as a common Japanese based on the Hiroshima dialect, as well as Cantonese, Korean, and Tagalog. All of the threatened languages are in danger because of their status as indigenous minority languages positioned at the lowest level of the linguistic hierarchy. Reports to date have included little discussion of the multilevel classification of linguistic strata from a formal linguistic perspective. It will be necessary in the future to examine these phenomena from the perspectives of sociolinguistics or linguistic anthropology.

References Barr, D., and Barr, S. (1978). Index of Irian Jaya Languages. Prepublication draft. Abepura: Universitas Cenderawashih and Summer Institute of Linguistics. Blust, R. A. (1990). “Summary report: linguistic change and reconstruction methodology in the Austronesian language family”, in P. Baldi (ed.), Linguistic Reconstruction and Reconstruction Methodology. Berlin: Mouton de Gruyter, 133–53. Carrington, L. (1996). A Linguistic Bibliography of the New Guinea Area. Canberra: Pacific Linguistics, D-90. Comrie, B., Matthews, S., and Polinsky, M. (1996). The Atlas of Languages. New York: Chackmark Books. Crowley, T. (2002). “Vinmavis”, in J. Lynch, M. Ross and T. Crowley (eds.), 638–49. Dixon, R. M. W. (1991). “The endangered languages of Australia, Indonesia and Oceania”, in R. Robins and E. M. Uhlenbeck (eds.), Endangered Languages. Oxford: Berg, 229–55.


281

Djebar Hapip, A., Darmansjah, and Basran Noor (1979). Bahasa Bajau. Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikan dan Kebudayaan. Donohue, M. (2002). “Tobati”, in J. Lynch, M. Ross, and T. Crowley (eds.), 186–202. Gordon, R. G. (2005). Ethnologue: Languages of the World, 15th edn. Dallas: International Academic Bookstore. Institute of Pacific Studies (1980). Vanuatu. Suva. JHEOA (2003). www.kempadu.gov.my/jheoa/orang%20asli/proto-malay.htm Leeden, A. C. van der (1993). Ma’ya: A Language Study. LIPI-RUL Series. Jakarta: Indonesian Institute of Sciences. Lynch, J. (1998). Pacific Languages: An Introduction. Honolulu: University of Hawaii Press. —— and Crowley, T. (2001). Languages of Vanuatu: A New Survey and Bibliography. Canberra: Pacific Linguistics, 517. —— Ross. M., and Crowley, T. (eds.) (2002). The Oceanic Languages. Richmond: Curzon Press. Manda, M., Yamaguchi, M., and Nakashima, H. (2003). Kosakata Dasar Bahasa Panasuan serta Tata Bahasa Ringkas Bahasa Panasuan dan Kosakata Dasar Bahasa Tangkou serta Tata Bahasa Ringkas Bahasa Tangkou. ELPR A3-010. Mead, D. (1999). The Bungku-Tolaki Languages of South-Eastern Sulawesi, Indonesia. Canberra: Pacific Linguistics, D-91. Nekitel, O. (1998). Voices of Yesterday, Today and Tomorrow. New Delhi: UBS Publishers’ Distributions Ltd. Ni Dabai (1988). “The genealogical affiliation of the language of the Hui people in Sanya Hainan”, Minzu Yuwen, 2: 18–25 (in Chinese). Nurhayati, Yamaguchi, M., and Nakashima, H. (2003). Kosakata Dasar Bahasa Topoiyo serta Tata Bahasa Ringkas Bahasa Topoiyo. ELPR A3-009. Oda, S. (1977). “The syntax of Pulo Annian”, Ph.D. dissertation, University of Hawaii. Okamoto, M. (1999). Chugoku no Shosu-minzoku-kyoiku to Gengo-seisaku [Ethnic Minority Education and Language Policy in China]. Tokyo: Shakai-hyoronsha. Osumi, M. (1995). Tinrin Grammar. Oceanic Linguistics Special Publication 25. Honolulu: University of Hawaii Press. Sakiyama, O. (1982). “The characteristics of Nguluwan from the viewpoint of language contact”, in M. Aoyagi (ed.), Islanders and their Outside World: A Report of the Cultural Anthropological Research in the Caroline Islands of Micronesia in 1980– 1981. Tokyo: St Paul’s (Rikkyo) University, 105–28. (Reprinted in Studies of Minority Languages in the Western Pacific Rim. ELPR C-006. 2002, 107–28.) —— (1989a). “Serui”, in T. Kamei, R. Kono and E. Chino (eds.), The Sanseido Encyclopedia of Linguistics, vol. ii. Tokyo: Sanseido, 472–74 (in Japanese). —— (1989b). “Tobati”, in T. Kamei, R. Kono and E. Chino (eds.), The Sanseido Encyclopedia of Linguistics, vol. ii. Tokyo: Sanseido, 1326–28 (in Japanese). —— (1997). “What had happened in the Austronesian rims?”, paper presented at the 115th Meeting of the Linguistic Society of Japan, University of Kyoto, 11–12 Oct. —— (1999). “Nominals of Fatamanue, Seram Maluku: a subgrouping argument in Central Malayo-Polynesian”, Bulletin of the National Museum of Ethnology (Osaka), 24/3: 467–84. (Reprinted in Studies of Minority Languages in the Western Pacific Rim. ELPR C-006. 2002, 129–47.) Shibata, N. (2003). Penrhyn–English Dictionary. ELPR A1-005.

282


Shintani, T., and Païta, Y. (1990). Grammaire de la langue de Païta. Nouméa: Société d’Études Historiques de la Nouvelle-Calédonie. Simonson, D., Sasaki, P., and Sakata, K. (1981). Pidgin To Da Max. Honolulu: Peppovision. Solomon Islands National Committee (1992). A Survey of Literacy and Language March–November 1991. Honiara. Taber, M., et al. (eds.) (1996). Atlas Bahasa Tanah Maluku (Maluku Languages Atlas). Ambon: Pusat Pengkajian dan Pengembangan Maluku, Universitas Pattimura and Summer Institute of Linguistics. Tryon, D. (1979). “Remarks on the language situation in the Solomon Islands”, in S. A. Wurm (ed.), New Guinea and Neighboring Areas: A Sociolinguistic Laboratory. The Hague: Mouton, 33–51. —— (2000). “The languages of the Pacific Region: the Austronesian languages of Oceania”, paper presented at the colloquium Language Endangerment, Research and Documentation-Setting Priorities for the 21st Century, Karl Arnold Academy, 12–17 Feb. —— (2002). “Buma”, in J. Lynch, M. Ross, and T. Crowley (eds.): 573–86. Uhlenbeck, E. M. (1971). “Indonesia and Malaysia”, in T. Sebeok (ed.), Current Trends in Linguistics, viii: Linguisitics in Oceania. The Hague: Mouton, 55–111. Voegelin, C. F., and Voegelin, F. M. (1977). Classification and Index of the World’s Languages. New York: Elsevier. Voorhoeve, C. L. (1975). Languages of Irian Jaya: Checklist, Preliminary Classification, Language Maps, Wordlists. Canberra: Pacific Linguistics, B-31. Wurm, S. (1982). Papuan Languages of Oceania. Tübingen: Gunter Narr Verlag. —— (ed.) (1997). Materials on Languages in Danger of Disappearing in the Asia-Pacific Region No. 1: Some Endangered Languages of Papua New Guinea: Kaki Ae, Musom, and Aribwatsa. Canberra: Pacific Linguistics, D-89. —— (2000). “Threatened languages in the western Pacific area from Taiwan to, and including, Papua New Guinea”, paper presented at the colloquium Language Endangerment, Research and Documentation-Setting Priorities for the 21st Century, Karl Arnold Academy, 12–17 Feb. —— (ed.) (2001). Atlas of the World’s Languages in Danger of Disappearing, 2nd edn., revised, enlarged, and updated. Paris: UNESCO. Zheng Yiqing (1997). Huihuihua Yanjiu (A Study of the Huihui Language). Shanghai: Shanghai-yuandong-chubanshe.

South-East Asia


15

Indigenous Languages of Formosa Naomi Tsukida and Shigeru Tsuchida

15.1. Introduction 15.2. Language endangerment in Taiwan 15.2.1. Speakers over 60 15.2.2. Fifties to thirties 15.2.3. Late teens to twenties 15.2.4. Below mid-teens 15.2.5. Situation of small ethnic groups 15.2.6. Indigenous peoples in cities 15.3. Efforts for revitalization 15.3.1. Early movements by Yuanquanhui and the establishment of the Council of Indigenous Peoples (CIP) 15.3.2. Education 15.3.3. Literacy and publication 15.3.4. Media 15.4. Documentation 15.4.1. Amis 15.4.2. Atayal 15.4.3. Bunun 15.4.4. Kanakanavu 15.4.5. Kavalan 15.4.6. Paiwan 15.4.7. Pazeh 15.4.8. Puyuma 15.4.9. Rukai 15.4.10. Saaroa 15.4.11. Saisiyat 15.4.12. Seediq 15.4.13. Thao 15.4.14. Tsou 15.4.15. Yami

285 288 288 288 289 289 290 290 290 290 291 292 293 293 294 294 295 295 295 295 295 296 296 296 296 296 297 297 297

15.1. Introduction Taiwan (Formosa) is an island of about 36,000 square km, with a population of about 22,000,000. The population of Austronesian ethnic groups is about 370,000 people, occupying only 1.8 per cent. The rest is all Han Chinese.

286

naomi tsukida and shigeru tsuchida

Figure 15.1. Formosan languages

Han Chinese are subgrouped into three: Holo (75 per cent), Hakka (13 per cent), and mainlanders (10 per cent). Their languages are not endangered. The ancestors of the Holo, which is one of their self-appelations, and the Hakka began to migrate to Taiwan in the seventeenth century. The Holo are from the Fujian province and speak a type of Southern Min dialect of Chinese, which is often called Taiwanese. The mainlanders came to Taiwan after the Second World War and held political power for almost half a century until 1988, when Li Teng-hui, a native Taiwanese, took over the presidency. According to Bellwood (1991), the Austronesian peoples came to Taiwan about 6,000 years ago. On the plains on the western side of Taiwan, after having contacted the Chinese people and culture, the Austronesian peoples became Sinicized over the centuries, and they almost lost their own languages and cultures by the end of the nineteenth century. They were called collectively “Peipo”, but are actually comprised of the following

i n d i g e n o u s l a n g uag e s o f f o r m o s a

287

twelve different ethnic groups (from north-eastern to north-western, and down to south): Trobiawan, Basay, Ketangalan, Kulon, Taokas, Babuza, Papora, Pazeh, Hoanya, Taivoan, Siraya, and Makatao. These Peipo languages, except Pazeh, have become totally extinct and are not dealt with here. Readers who are interested in them are referred to Tsuchida (1992). As for Pazeh, there is only one competent speaker left, who is 83 years old, as of 2003. In the central mountain area and on the eastern side of the island, Austronesian peoples still retain their own languages. There are thirteen of them (more or less from north to south, and from west to east): Atayal, Seediq, Saisiyat, Thao, Bunun, Tsou, Kanakanavu, Saaroa, Rukai, Paiwan, Puyuma, Kavalan, and Amis. Among them Kavalan and Thao were often counted as Peipo but the Thao was officially recognized as a major Austronesian group in Taiwan in 2001, and the Kavalan in 2002. Orchid Island also belongs to Taiwan, and Yami, which belongs to the Batanic languages of the Philippine group, is spoken there. Their population, as officially reported in 2000, is approximately as follows (alphabetically arranged). Administratively Seediq is included in Atayal, and Kanakanavu and Saaroa in Tsou, so an administrative census on Seediq, Kanakanavu, and Saaroa is not available. Amis Atayal Bunun Kavalan Paiwan Puyuma Rukai Saisiyat Thao Tsou Yami

150,000 90,000 40,000 1,000 70,000 10,000 12,000 5,000 300 6,000 4,000

(including Seediq) (as of 2002)

(as of 2002) (including Kanakanavu and Saaroa)

The number of people who can speak their own language is far less than these figures. Aborigines were generally regarded as vulgar, brutal, uncivilized, and retarded. Even they themselves thought so, and were losing their self-respect in their own culture and language. Because of such an inferiority complex elders do not speak their language to their children or grandchildren, which obstructs the succession of their traditional languages. Aboriginal students achieve lower

288


academic levels than Chinese students. The rates of delinquency, alcoholism, and unemployment are all higher than those of Han Chinese. Many aborigines have to engage in unstable jobs with low prestige, such as construction workers and truck drivers. Some women may work as bar hostesses.

15.2. Language Endangerment in Taiwan Fourteen Austronesian languages are still spoken in Taiwan, but all of them are on the verge of extinction due to the successive Japanization and Sinicization during the twentieth century. The Japanese occupation lasted for fifty years (1895–1945), during which period they tried to give a primary education to youngsters in Japanese. After the Second World War the Nationalist Party government came over from mainland China, and switched the official language to Mandarin. This situation lasted until Li Teng-hui eliminated the mainlanders’ control over languages and cultures in Taiwan in the 1980s. In contrast, Holo language is not endangered, though it was also suppressed in public in the same way as aboriginal languages were. Holo are the majority in the Taiwan population, with social and economical, and now political power, since the Democratic Progressive Party came to power.

15.2.1. Speakers over 60 People in their sixties or older can speak their own languages fluently. When elderly people of the same ethnic group talk to one another, they speak in their language. Old people still use Japanese as a lingua franca when they talk to people of other ethnic groups or Chinese of the same generation. They may sometimes code-switch between their languages and Japanese. In south-western Taiwan, they may be also able to speak Holo. Many old people can understand and speak Mandarin to some extent now because of the necessity in their daily life. They even try to talk to younger generations in Mandarin.

15.2.2. Fifties to Thirties People in their thirties, forties, or fifties are bilinguals of their traditional language and Mandarin, as education was conducted in Mandarin after the Second World War. They use Mandarin as their lingua franca. Code switching,


289

or more often code mixing, between their language and Mandarin is very often observed. Even when only people of the same ethnic group are involved in the conversation, they may code-switch to or mix Mandarin, according as the topic of the conversation changes. The more young people are involved, the higher the proportion of Mandarin is. The younger the speaker is, the poorer the command of their tribal language is. They tend not to make the distinctions that older people do.

15.2.3. Late Teens to Twenties Young people in their late teens or twenties usually speak Mandarin and cannot speak their traditional language or do so very poorly if ever. This is true of all tribes, regardless of the size. They may utter exclamational expressions in the language of their ethnic group, but cannot utter sentences with more complex structure. Even when they try to speak their own language, it is likely to be affected by Mandarin or to be much simpler. Many young people of indigenous populations drop out of junior high school and stay in the village, doing nothing. Some of the teenage girls, unmarried, get pregnant and have to raise their children in their villages. Ironically, however, it is such girls that have more chances to talk with the elderly people and to learn their own language than those who go to a nearby city to work every day, or to study among Mandarin or Holo speakers, not to mention those who stay in cities for higher education.

15.2.4. Below Mid-teens Those children in their early teens and younger do not speak their traditional languages. Young aboriginal parents do not talk to their children in their native language, but only in Mandarin. The grandparents also try to use Mandarin in speaking to them. As a consequence, the children speak only Mandarin, and at best can only understand the traditional language when they hear it. The family’s attitude is an important factor which affects the language competence of young children. There are few exceptional cases. Some families think that a language is an integral part of their traditional culture, and thus parents and grandparents try to talk to their children in their own language and, as a result, exceptionally, children come to understand and speak the native language.

290


15.2.5. Situation of Small Ethnic Groups Small groups often do not form a hamlet by themselves but tend to live together with other ethnic groups, and speak the language of the majority. The result is low proficiency in their own language. For example, the Kanakanavu and the Saaroa people have a very small population and are administratively classified into Tsou, following the tradition of anthropology. Geographically, however, they coexist in the same areas with the Bunun, and people in their thirties, forties, or fifties speak Bunun more fluently, and their own languages are seriously endangered. Those below thirty speak only Mandarin, as is true of other tribes. The Thao people, who live near Sun Moon Lake, a famous scenic spot, and earn their living by tourism, speak Holo by now better than their own language. There are only about ten old people who can speak the language fluently. The Kavalan people live in Ilan and Hualien counties. In Ilan, however, they live together with Holo and have been completely assimilated into them. In Hualien many live in the Amis-speaking area, and have almost lost their own language. The village Hsinshe in Hualien county is the only village where people speak Kavalan daily still today.

15.2.6. Indigenous Peoples in Cities Many indigenous peoples go to cities to seek jobs. In some areas aboriginals of the same group come to live in a neighbourhood, but in other cases, aboriginals in cities live alone surrounded by Han Chinese. In the latter case, their proficiency in their traditional language declines. Even people in their fifties or forties lose command of their languages, not to mention younger generations.

15.3. Efforts for Revitalization An effort for revitalization is being made in present-day Taiwan. In the background of this effort is the end of the closed-minded rule by the Nationalist Party, and the progress of democratization.

15.3.1. Early Movements by Yuanquanhui and the Establishment of the Council of Indigenous Peoples (CIP) Some indigenous students in National Taiwan University established Yuanquanhui (Association to Promote Aboriginal Rights) in 1984. Yuanquanhui made appeals


291

persistently to the government. This movement developed into the Council of Indigenous Peoples (CIP) which was established within the government in 1996 “for the purpose of organizing aborigine-related matters under one general organization” (Section 1 of the law concerned). The chairperson and two vice-chairpersons are aborigines and more than half of the members are mandatorily aborigines. It has been working on development programmes of Austronesian native peoples ever since, focusing on political participation, education, preservation of their cultures, social welfare, economics, and land affairs. Education includes rejuvenation of indigenous languages also. One thing Yuanquanhui gained in the early 1990s is the term yuanzhumin “aborigines”. Before 1994 the Nationalist Party government called aborigines shan-pao “mountain-fellows”. This term was misleading because not all the groups are inhabitants of the mountains; the Amis and Puyuma people live on the plains and the Yami people on Orchid Island. Aborigines claimed to be called Yuanzhumin “aborigines” (lit. “the people that originally inhabited [Taiwan] from the beginning”) and the government accepted it in response to their claim. The term yuanzhumin, however, is only an administrative label, and some peoples of Austronesian origin are not counted as Yuanzhumin. The Thao people were officially identified as Yuanzhumin only in 2001, and the Kavalan in 2002. Yuanzhumin can enjoy a specially arranged protection policy, such as tax exemption or favourable treatment in entrance examinations, and this encourages people of Austronesian origin to be identified as Yuanzhumin. They also made the government approve the use of their names in their own languages. Aborigines had to use Chinese names in public before. In 1995 there was a revision of an appropriate law enabling aborigines to register their names in their languages, though the names had to be transcribed in Chinese characters. In June 2000, the Full Name Registration Law passed and they are now able to register their names in Roman letters. In 1998 the Aboriginal Education Act (AEA) was enacted, and an implementation procedure was established in 1999, which will be dealt with more precisely in the following section.

15.3.2. Education In the mid 1990s, a class called “local culture” started in elementary schools. In this class aboriginal children learned their traditional language and culture, such as simple sentences and basic vocabulary, songs, and dances, although the class was optional and no grades were given.

292


The purpose of the aforementioned AEA is to “safeguard the educational rights of aborigines and upgrade the aborigine ethnic education and culture” (Section 1 of the said act). This act applies to all stages of education of all the aboriginal peoples, from preschool to graduate school, including adult education. It also covers many areas relevant to education, such as curriculum, grant/budget/subsidies, teaching resources and materials, representation in media such as TV, radio, and Internet, and so on. Following this act an implementation procedure was established in 1999, and in January 2000 it was decided that all students have to take “mother tongue classes” starting from 2001. In the “mother tongue classes” Holo children are to study Holo and aboriginal children are to study an aboriginal language of their own tribe. It is compulsory for all students and they are graded. At this point it differs from the previous “local culture” class, which was optional and for which students were not graded. In principle indigenous children who live in cities can also take classes in the language of their ethnic group at their schools. Schools have to hire teachers who can teach the language. At present most schools have only one class per week, however, which does not seem sufficient. All other classes are given only in Mandarin. No immersion programme is offered yet. To motivate indigenous students to study their own languages, the CIP set quite pragmatic goals for the children. At present, aboriginal children are accepted in higher schools with 25 per cent lower scores than nonaboriginal descendants and can apply for a scholarship. From 2005, however, a certificate of mother tongue proficiency will be required to enjoy such benefits. This also aims to motivate parents to talk to the children in their mother tongue so that children will become proficient in their language to get the certificate. Children who study aboriginal languages at school may change the view of adults. A teacher who supervises a children’s choir at an elementary school gives us a good example. When children learn to sing songs with lyrics in aboriginal languages, they get interested in the lyrics and then in the languages. It motivates children to talk with parents and grandparents about their native languages. Talking with children about their indigenous languages also helps lessen the inferiority complex of the grown-ups toward their languages.

15.3.3. Literacy and Publication Many textbooks on aboriginal languages have been published since the mid 1990s. Publication other than textbooks is gradually increasing, but no


293

magazines or newspapers are published in aboriginal languages yet. Most of the indigenous peoples adopted a Roman script which has been used by missionaries. There are Bible translations available in Amis, Seediq, Atayal, Bunun, Paiwan, Rukai, and Yami. Christian churches (mainly Presbyterian and Catholic) were the only organization that tried to invent a way of writing for the indigenous peoples except for Li (1991). They have still been playing an important role in compiling dictionaries and textbooks. The use of the Bible based on one of its dialects brought about the standardization of their language to some extent, especially among young people because their proficiency in their own language is not stable.

15.3.4. Media Until the 1980s the use of languages other than Mandarin was strictly prohibited on TV and radio. Holo and Hakka were not exceptions, either. This ban was lifted, and there have been radio programmes using aboriginal languages since the mid 1990s, although there are no TV programmes yet. The CIP recognizes the importance of mass media in language preservation. It is now working to establish a TV station which broadcasts programmes in aboriginal languages, as is reported in its newsletter (2003/1). This goal is based on the statement in the AEA. Aboriginal songs are being appreciated more and more, even among Han Chinese. As is well known, the German music group Enigma used a part of a traditional Amis song in their song titled “Return to Innocence”, which was broadcast all over the world as the official theme song of the Atlanta Olympic Games in 1996. Difang, an old Amis man who sang the original song, brought a lawsuit against the group over the copyright. This case made people in Taiwan appreciate the aboriginal music. The number of CDs of aboriginal music increased considerably after this. Recently, young aborigine singers are making pop/folk/ rock music in their languages, sometimes drawing on some elements of the traditional music of their tribes. This music has gained a measure of popularity in the Taiwanese market where most (98 per cent) of the listeners are Han Chinese.

15.4. Documentation Published before the Second World War, we cannot miss the great work by Ogawa and Asai (1935), which is a huge collection of texts in native languages written in broad transcription of IPA, with precise footnotes, a brief

294


description of grammar for most of the languages and their dialects, and a comparative lexicon. This book is comparable to F. Boas’s Handbook of American Indian Languages (1911, 1922) (though only with a minimum amount of text), and still remains an invaluable source in Formosan Austronesian studies. After the war, we had to wait until the 1970s when Paul Jen-kuei Li and Shigeru Tsuchida started publishing the results of their fieldwork. In the 1990s as the Taiwan economy began to enjoy its prosperous years, many young and competent researchers grew up in Taiwan, and they have been publishing many valuable papers. Because of the limited space here, it is not easy to cover all of the extant documents and studies, however. Most of the historical studies are omitted, as well as those on the languages which have become extinct by now, such as Siraya or languages of other Sinicized populations. Only some of the major published works will be listed in the following sections under each language. The listing is divided into four categories: dictionary (or vocabulary), grammar (phonological, morphological, and syntactic studies), text collection, and others (subgrouping, sociolinguistic studies, conversation, etc.). For the more detailed information, including that of historical studies and of extinct Peipo languages, see Li (1975b) and Huang (2000d). Readers are also referred to the following URLs: www.ling. sinica.edu.tw/formosan/; www.alcd.nccu.edu.tw/index_0.html

15.4.1. Amis Fey (1986) is a pretty good dictionary, but unfortunately it is very hard to find a copy in the market. The Sakizaya dialect, located at the northernmost tip of the Amis area, seems to preserve the most archaic features, but probably it might be a little too late to collect linguistic materials of this dialect. Dictionary: Fey (1986). Grammar: Cao (1991), He et al. (1986a), Wu (2000). Texts: Ogawa and Asai (1935). Others: Tsukida (1993).

15.4.2. Atayal Egerod (1999) is a very good dictionary with plenty of examples. However, the lexical items were taken only from the texts he collected, and it is a pity that some very basic words in the Austronesian languages, such as “the louse found in clothes, as against that in one’s hair” and “nit (egg of a louse)”, which happen not to appear in his texts, are not found in his dictionary.


295

Dictionary: Egerod (1999). Grammar: Huang (1993, 1995, 2000a), Rau (1992). Texts: Ogawa and Asai (1935). Others: Li (1980a, 1980b, 1981, 1986).

15.4.3. Bunun No good vocabulary or dictionary is available for Bunun. Grammar: He et al. (1986b), Jeng (1977), Zeitoun (2000a). Texts: He et al. (1986b), Sh.-Hs. Lin (1998), T. Lin et al. (1998), Ogawa and Asai (1935). Others: Li (1986, 1988), Moriguchi (1995).

15.4.4. Kanakanavu The population of the Kanakanavu people is not known, but most likely less than 200. The number of those who can still speak their own language is perhaps between five to ten. And yet a precise description of Kanakanavu has not been done. Vocabulary: Szakos (1999a). Grammar: Ho (1997), Szakos (1999a), Tsuchida (1976). Texts: Ogawa and Asai (1935), Szakos (1999a), Tsuchida (2003).

15.4.5. Kavalan A Kavalan dictionary is being prepared by Paul Jen-kuei Li and S. Tsuchida, and is expected to appear in 2006. Grammar: Y.-L. Chang (2000b). Texts: Shimizu (1998).

15.4.6. Paiwan Dictionary: Ferrell (1982). Grammar: H.-Ch. Chang (2000), Chen and Ma (1986), Egli (1990). Texts: Ogawa and Asai (1935).

15.4.7. Pazeh Dictionary: Li and Tsuchida (2001). Grammar: Li and Tsuchida (2001), Y.-J. Lin (2000). Texts: Li and Tsuchida (2002).

296


15.4.8. Puyuma Dictionary: Cauquelin (1991b). Grammar: Cauquelin (1991a), Huang (2000c). Texts: Ogawa and Asai (1935), Quack (1981, 1985), Schroeder and Quack (1979). Others: Ting (1978).

15.4.9. Rukai Grammar: Li (1973), Zeitoun (2000c). Texts: Li (1975a), Ogawa and Asai (1935). Others: Ho (1983), Li (1977).

15.4.10. Saaroa Just as is the case with Kanakanavu, the population of the Saaroa people is not known because both of these groups are politically counted under Tsou. A rough estimation is about 500, but again, those who can speak Saaroa are far less, perhaps under twenty. This is one of the least studied languages. Vocabulary: Szakos (1999b). Grammar: Li (1997b), Tsuchida (1976). Texts: Ogawa and Asai (1935).

15.4.11. Saisiyat Since there is no Saisiyat dictionary yet, collections of words are listed here. Vocabulary: Li (1978), Yeh (2000). Grammar: Yeh (2000). Texts: Ogawa and Asai (1935). Others: Li (1978).

15.4.12. Seediq Although Pecoraro (1977) is a mimeographed version, containing quite a number of confusions of k with q, and what is more, it is not generally available, this is the only dictionary available for this language. Dictionary: Pecoraro (1977). Grammar: Y.-L. Chang (2000a), Holmer (1996).


297

Texts: Catholic Association (1994), Ogawa and Asai (1935), Tsao (1997), Tsukida (1995). Others: Li (1986).

15.4.13. Thao Dictionary: Blust (2003). Grammar: Blust (2003), Huang (2000b). Texts: Blust (2003).

15.4.14. Tsou Dictionary or vocabulary: An et al. (1996), Nevskij (1993), Tung (1964). Grammar: Szakos (1994), Tsuchida (1976, 1990), Tung (1964), Zeitoun (2000b). Texts: Ogawa and Asai (1935), Szakos (1994), Tung (1964).

15.4.15. Yami Dictionary: Rau and Dong (2006), Li and Ho (1989), Tsuchida et al. (1987, 1989). Grammar: Hs.-H. Chang (2000). Texts: Benedek (1991), Ogawa and Asai (1935), Shi (1992), Tung (1995). Others: Numao (1996), Tsuchida (1984).

References Abbreviations used in the following references: BIHP The Bulletin of the Institute of History and Philology, Academia Sinica CIA Collectanea Instituti Anthropos. Augustin: Haus Voelker und Kulturen (later Berlin: Dietrich Reimer Verlag) IHPSP Institute of History and Philology, Special Publications PL Pacific Linguistic Series. Canberra: Department of Linguistics Research School of Pacific Studies, The Australian National University SBDMLC Series of Brief Descriptions of Minority Languages in China. Beijing: Minzu Publishing Co. (in Chinese) TALS Taiwan Austronesian Language Series. Taipei: Yuan-liu Publishing Co. (in Chinese) TULIP Tokyo University Linguitics Papers

298


An, Ch.-Ch., et al. (1996). Compact Dictionary of Tsou in Alishan County, Jiayi County. Tapang Elementary School (in Chinese). Bellwood, P. (1991). “The Austronesian dispersal and the origin of languages”, Scientific American, 265: 70-5. Benedek, D. (1991). The Songs of the Ancestors: A Comparative Study of Bashiic Folklore. Taipei: SMC Publishing Inc. Blust, R. (2003). Thao Dictionary. Language and Linguistics Monography Series A5. Taipei: Institute of Linguistics, Academia Sinica. Cao, S.-J. (1991). Grammar of Amis in Taiwan. Beijing: Chung-yang Minzu Yuyuan Pub. Co. (in Chinese) Catholic Association for Rural People Research and Service (1994). Kari Pnsltudan Seediq Cbeyo [Folk Tales of the Seediq People]. Catholic Association for Rural People Research and Service) (in Sediq, Chinese, and English). Cauquelin, J. (1991a). “The Puyuma language”, Bijdragen tot de Taal-, Land- en Volkenkunde, 147/1: 17-60. —— (1991b). Dictionnaire Puyuma–Français. Textes et Documents Nousantrariens IX. Paris, (Jakarta): École Française d’Extrême-Orient. Chang, H.-Ch. (2000). Paiwan Reference Grammar. TALS 11. Chang, Hs.-H. (2000). Yami Reference Grammar. TALS 6. Chang, Y.-L. (2000a). Seediq Reference Grammar. TALS 6. —— (2000b). Kavalan Reference Grammar. TALS 12. Chen, K., and Ma, R.-Sh. (1986). Outline of the Formosan Native Languages (Paiwan). SBDMLC (in Chinese). Egerød, S. (1999). Atayal–English Dictionary, ed. J. O. Petersen., 2nd edn. Historiskfilosofiske Skrifter 20, Royal Danish Academy of Sciences and Letters. Copenhagen: C. A. Reitzels Forlag. Egli, H. (1990). Paiwangrammatik. Wiesbaden: Otto Harrassowitz. Ferrell, R. (1982). Paiwan Dictionary. PL C-73. Fey, V. (ed.) (1986). Amis Dictionary. Taipei: The Bible Society. He, R.-F., et al. (1986a). Outline of the Formosan Native Languages (Amis). SBDMLC. —— et al. (1986b). Outline of the Formosan Native Languages (Bunun). SBDMLC. Ho, D.-A. (1983). “The position of Rukai in the Formosan languages”, BIHP 54/1: 121-68. —— (1997). “Kanakanavu”, in P. J.-K. Li (ed.): 228-71 (in Chinese). Holmer, A. (1996). A Parametric Grammar of Seediq. Travaux de l’Institut de Linguistique de Lund 30. Lund: Lund University Press. Huang, L. M.-J. (1993). A Study of Atayal Syntax. Taipei: The Crane Publishing Co. —— (1995). A Study of Mayrinax Syntax. Taipei: The Crane Publishing Co. —— (2000a). Atayal Reference Grammar. TALS 1. —— (2000b). Thao Reference Grammar. TALS 4. —— (2000c). Puyuma Reference Grammar. TALS 10. —— (2000d). “Formosan linguistics: past and future”, Chinese Studies, 18 Special Publication: 79-110 (in Chinese). Jeng, H.-Hs. (1977). Topic and Focus in Bunun. IHPSP No. 72. Li, P. J.-K. (1973). Rukal Structure. IHPSP No. 64. —— (1975a). Rukai Texts. IHPSP No. 64.2. —— (1975b). “Taiwan tuzhu yuyan de yanjiu ziliao yu wenti (Research materials and problems of Formosan languages)”, Bulletin of the Institute of Ethnology (Academia Sinica), 40: 51-84. —— (1977). “The internal relationships of Rukai”, BIHP 48/1: 1-92.


299

—— (1978). “A comparative vocabulary of Saisiyat dialects”, BIHP 49/2: 133-99. —— (1980a). “The phonological rules of Atayal dialects”, BIHP 51/2: 349-405. —— (1980b). “Men’s and women’s speech in Mayrinax”, in Papers in Honor of Professor Lin Yu-k’eng on her Seventieth Birthday. Taipei: Wen Shin Publishing Co., 9-17. —— (1981). “Reconstruction of Proto-Atayalic phonology”. BIHP 52/2: 235-301. —— (1986). “Linguistic variations of different age groups in some Formosan languages”, in P. Geraghty, L. Carrington and S. A. Wurm (eds.), FOCAL II: Papers from the Fourth International Conference on Austronesian Linguistics. PL C-94, 33-50. —— (1988). “A comparative study of Bunun dialects”, BIHP 59/2: 479-508. —— (1991). Taiwan Nantao-yuyuan-de Yu-yuan Fu-hao Hsi-tung [Orthographic Systems for Formosan Languages]. Taipei: Ministry of Education (in Chinese). —— (ed.) (1997a). Gaoxiong-xian nandao yuyan [Austronesian Languages in Kao-hsiung Province]. Research Series on Kao-hsiung Province 7. Kao-hsiung: Government of Kao-hsiung Province. (in Chinese). —— (1997b). “Saaroa”, in P. J.-K. Li (ed.): 272-97 (in Chinese). —— and Ho, Y.-L. (1989). “Wordlist of Yami (Izanumilek dialect)”, in P.-Hs. Liu (ed.), Preservation and Development of Yami and Yami Culture. Taipei: Society for Wildlife and Nature, 44-86 (in Chinese). —— and Tsuchida, Sh. (2001). Pazih Dictionary. Language and Linguistics Monograph Series A2. Taipei: Institute of Linguistics, Academia Sinica. —— (2002). Pazih Texts and Songs. Language and Linguistics Monograph Series A2-2. Taipei: Institute of Linguistics, Academia Sinica. Lin, Sh.-Hs. (1998). Shadows of Mountain Palm under the Moon. Taipei: Chen-Hsin Pub. Co. (in Bunun and Chinese). Lin, T., et al. (1998). Moon Which Passed Over Time and Space [Daingasbae Tus? a tu Buan]. Taipei: Chen-Hsin Pub. Co. (in Bunun and Chinese). Lin, Y.-J. (2000). Pazeh Reference Grammar (巴則海語). TALS 3. Moriguchi, T. (1995). “Linguistic and anthropological analysis of avoidance in Bunun, Taiwan”, TULIP 14: 23-47. Nevskij, N. A. (1993). Vocabulary of Tsou, Taiwan. Taipei: Tai-yuan Publishing Co. (in Chinese). Numao, T. (1996). “On the classification of fishes among the Yami (Iranomilek)”, Studies of Cultural Anthroplogy 14. Nagoya: Department of Cultural Anthropology, Nanzan University, 20-47 (in Japanese). Ogawa, N., and Asai, E. (1935). Gengo ni yoru Taiwan Takasago-zoku Densetsu-shu [The Myths and Traditions of the Formosan Native Tribes]. Taihoku (= Taipei): Taihoku Teikoku Daigaku Gengo-gaku Kenkyu-shitsu (in Japanese). Pecoraro, F. (1977). Essai de dictionnaire Taroko–Français. Cahier d’ Archipel 7. Paris: SECMI. Quack, A. (1981). Das Wort der Alten. Erzaehlungen zur Geschichte der Pujuma von Katipol (Taiwan). CIA 12. —— (1985). Priesterinnen, Heilerinnen, Schamaninnen? Die poringao der Puyuma von Katipol (Taiwan), dargestellt und analysiert nach Aufzeichnungen aus dem Nachlass von D. Schroeder. CIA 32. Rau, D.-H. V. (1992). A Grammar of Atayal. Taipei: The Crane Publishing Co. Rau, D. V., and Dong, M.-N. (2006). Yami Texts with Reference Grammar and Dictionary. Language and Linguistics Monograph Series A10. Taipei: Institute of Linguistics, Academia Sinica.

300


Schroeder, D., and Quack, A. (1979). Kopfjagdriten der Puyuma von Katipol (Taiwan), eine Textdokumentation. CIA 11. Shi, N.-L. (1992). Myth of Badaiwan: Myths and Stories from the Homeland of Flying Fish. Aboriginal Series 9. Taichung: Chenxing Publishing Company. Shimizu, J. (1998). Kvalan Myths and Traditions, as Recorded in the Original Language. Taipei: SMC Pub. Co. (in Japanese, English, and Chinese as well as Kavalan). Shun Ye Taiwan Aborigines Research Group in Japan (ed.) (2001). Taiwan Genjuumin kenkyuu gairan [General Survey of Taiwan Aborigines Studies]. Tokyo: Fukyosha. Szakos, J. (1994). “Die Sprache der Cou, Untersuchungen zur Synchronie einer austronesischen Sprache auf Taiwan”, dissertation submitted to Rheinischen FriedrichWilhelms-Universität zu Bonn. —— (1999a). Final Report of Kanakanavu Language and Culture. Kaohiung County, Sanmin Township, and Ching-yi University (with a CD). —— (1999b). Final Report of Saaroa Language: Vocabulary. CIP, Executive Yuan. 159 pp. Ting, P.-Hs. (1978). ‘‘Reconstruction of Proto-Puyuma phonology’’. BIHP 49/3: 321-92 (in Chinese). Tsao, R.-L. (Teminawi C. Tseng) (1997). A Study of Seediq Atayal’s Foods and their Stories, vol. i. Taipei: Tangshan Publishing Company (in Chinese and English). Tsuchida, Sh. (1976). Reconstruction of Proto-Tsouic Phonology. Study of Languages and Cultures of Asia and Africa Monograph Series 5. Tokyo: Institute of Languages and Cultures of Asia and Africa. —— (1984). “Fish names in Yami (Imorod dialect): second interim report”, TULIP 84: 11–90. —— (1990). “Classificatory prefixes of Tsou verbs”, TULIP 89: 17-52. —— (1992).「平埔族各語言研究瑣記（上・下）」[Miscellanies of languages of sinicized ethnic groups in Formosa (1, 2)], 『臺灣史田野研究通訊』 [Newsletter of Taiwan History Field Research], 22: 9-22; 23: 26-42. —— (2003). Kanakanavu texts (Austronesian Formosan), ELPR A3-014. —— Yamada, Y., and Moriguchi, T. (1987). Lists of Selected Words of Batanic Languages. Tokyo: Dept. of Linguistics, University of Tokyo. ——— and Constantino, E. (1989). Batanic Languages: Lists of Sentences for Grammatical Features. Tokyo: Dept. of Linguistics, University of Tokyo. Tsukida, N. (1993). “A brief sketch of the Sakizaya dialect of Amis”, TULIP 13: 372-413. —— (1995). “Texts in the Truku dialect of the Sediq language”, Asian and African Linguistics, 24: 103-48. Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa, University of Tokyo Foreign Affairs. Tung, M.-N. (1995). Praising Taros. Panchiao: Taohsiang Pub. Co. (in Yami and Chinese). Tung, T.-H. (1964). A Descriptive Study of Tsou Language, Formosa. IHPSP No. 48. Wu, C.-L. (2000). Amis Reference Grammar. TALS 11. Yeh, M.-L. (2000). Saisiyat Reference Grammar. TALS 2. Zeitoun, E. (2000a). Bunun Reference Grammar. TALS 5. —— (2000b). Tsou Reference Grammar. TALS 6. —— (2000c). Rukai Reference Grammar. TALS 8.

16

Languages of Mainland South-East Asia David Bradley

16.1. Introduction 16.2. Austro-Thai endangered languages 16.2.1. Austronesian 16.2.2. Thai-Kadai 16.2.3. Miao-Yao 16.3. Austroasiatic endangered languages 16.3.1. Aslian 16.3.2. Katuic-Bahnaric 16.3.3. Northern 16.3.4. Pakanic 16.3.5. Monic 16.3.6. Pearic 16.3.7. Vietic 16.4. Tibeto-Burman endangered languages 16.4.1. Digarish “Mishmi” 16.4.2. Nungish 16.4.3. Burmic 16.4.3.1. Burmish 16.4.3.2. Gong 16.4.3.3. Ngwi 16.4.4. Karenic 16.4.5. Sal 16.4.6. Kuki-Chin-Naga 16.5. Indo-European endangered languages 16.6. Conclusion

303 308 309 310 313 314 315 315 317 319 320 320 321 322 322 322 323 323 324 325 329 330 331 332 332

I am pleased to acknowledge the support of the UNESCO Endangered Languages programme and of the Australian Research Council (A59701122, A59803475, A00001357) and assistance from various institutions in China, especially the Yunnan Academy of Social Sciences; Thailand, especially Mahidol University; and France, especially the Centre National de la Recherche Scientifique. Information from a number of colleagues is included here: 2000 China census figures from Colin Mackerras; information on Aslian from Geoffrey Benjamin; on various Mon-Khmer languages from Ilia Peiros, Suwilai Premsrirat, Theraphan Luangthongkum, and Gérard Diffloth; on languages of Burma from Denise Bernot; on Miao-Yao languages from Barbara Niederer; on Kristang from Alan Baxter; and on various Kadai and other languages of Vietnam from Jerry Edmondson. Of course, all remaining shortcomings and errors and any misunderstanding of information from others are my sole responsibility.

302


Figure 16.1. Languages of mainland South-East Asia

l a n g uag e s o f m a i n l a n d s o u t h - e a st a s i a

303

16.1. Introduction This chapter discusses ninety-six endangered languages (ELs) and six extinct languages of mainland South-East Asia including Burma, Cambodia, Lao, Thailand, Vietnam, and the non-Austronesian languages of West or peninsular Malaysia, in the context of related languages and the dominant languages which are replacing them. The main generally recognized families in this area are Sino-Tibetan (ST) and Austroasiatic (AA). Both also extend into east and south Asia. The other languages are in families whose higher-level classification is a matter of controversy. The Austro-Thai (AT) superfamily proposed by Benedict (1942) includes the Thai-Kadai (TK) family (which includes Thai, and is found in southern China as well as South-East Asia and into the north-eastern part of south Asia) and the Austronesian (AN) family (mainly in the islands, but also on the coast). Benedict (1975) adds the Miao-Yao (MY) or Hmong-Mien family (mainly in southern China and extending into northern South-East Asia), and Benedict (1990) proposes to add Japanese as well. More recent proposals by Sagart (2002) based on more solid lexical, phonological, and morphological correspondences link AN and ST. An earlier proposed link was of AN and AA in Austric (Schmidt 1906), but this is not widely accepted nowadays. One endangered Indo-European (IE) creole is also spoken in the area. Languages of the longer-established politically dominant groups within these families have been replacing other related and unrelated languages in central areas for many millennia, leading to greatly reduced linguistic diversity in Cambodia, coastal areas of Vietnam and Malaysia, and the central plains of Thailand and Burma. Marginal upland areas both preserve preexisting diversity and accommodate migrations from adjacent areas where majority group pressure forces other groups away. This has led to the spread of numerous languages from south-western China into South-East Asia. Of at least 960 AN languages, only a few are spoken on the Asian mainland. This includes two endangered indigenous Malayic languages of southern Thailand and the extreme south of Burma. The other AN group here is Chamic, which includes four endangered languages and three other languages of southern Vietnam and north-eastern Cambodia, with one additional endangered language, Tsat, spoken on Hainan Island in China, as noted in the chapter by Sakiyama. Of more than eighty additional AT languages from subgroups other than AN, over fifty are TK, twenty-seven are MY, and a lexical component in Japanese may also fit here. Of the non-AN AT languages spoken in mainland

304


Table 16.1. List of languages No.

Name

Austronesian Malayic 1. Moken 2. Urak Lawoi Chamic 3. 4. 5. 6.

Moklen

Mon-Khmer Aslian 18. Batek 19. Besisi 20. Chewong 21. Jah Het 22. Jehai 23. Kenaboi 24. Kensiu 25. Kintaq 26. Lanoh 27. Maniq 28. Mendriq 29. Semai 30. Semaq Beri 31. Semelai 32. Temiar Katuic-Bahnaric Katuic 33. Chatong 34. Dakkang 35. Triw

Speakers

6,000 3,000

4,000 2,000

250,000 11,000 25,000 20,000

Endangerment location

B/T T/M

100,000 9,000 22,000 18,000

PE PE PE PE

V/C/T/M V V V

150,000 100,000 150,000 7,863 2,100 200 689

2,000 50 3,000 5,000 800 200 400

SE SE SE E PE PE E

Ch/V Ch/V Ch/V V V V V/Ch

20,000 100,000 95,000

10,000 80,000 70,000

E PE PE

L/T L L

48,700

25,000

PE

Ch/V

960 2,185 400 3,193 1,200

500 235 350 Tonga’/Mos/Ten’en 300 Menriq 145 Senoi/Sengoi 29,000 2,488 4,103 17,000

800 1,500 300 2,500 1,000 0 400 200 300 200 100 10,000 2,000 2,500 15,000

E SE SE E E Ex E E E E E E E E PE

M M M M M/T M M/T M M T M M M M M

580 1,198 1,328

500 1,000 1,200

E E E

L L L

Hagei/Hoki A-o/A-uo Tudu Lati Khla Phlao Ainh Pupeo/Laqua

Saek Tai Daeng Tai Neua

Miao-Yao 17. Baheng

Population

E E

Cham Chru N. Roglai S. Roglai

Thai-Kadai Kadai 7. Gelao (Green) 8. Gelao (Red) 9. Gelao (White) 10. Lachi 11. Laha 12. Nung Ven 13. Pubiao Tai 14. 15. 16.

Other names

Pathen

Mah Meri Jahai

l a n g uag e s o f m a i n l a n d s o u t h - e a st a s i a No.

Name

Other names

Bahnaric 36. 37. 38. 39.

Chrau Juk Kaco’ Swoeng

Choro Suai “Lamam” Lavi

Northern Angkuic 40. Mok

Population

Speakers

15,000 1,500 1,000 492

7,000 1,400 800 450

E E E E

V L C L

10,000?

6,000?

E

B/L/Ch/T

5,000

2,000

SE

B

Mae Rim Lawa Lawa

200 20,000

0? 12,000

M E

T T

Tai Hat/Hat Phi Tong Luang Kha Bit/Bit

200 150 2,232

30 100

The Vanishing Languages of the Pacific Rim

The Vanishing Languages of the Pacific Rim (Oxford Linguistics)

Food for Health in the Pacific Rim

Food for Health in the Pacific Rim

Vanishing Voices: The Extinction of the World's Languages

Vanishing Voices: The Extinction of the World's Languages

Rim Rim Of Space

Rim To Run the Rim

Rim To Run the Rim

The Rim of Space

The Rim of Space

The Vanishing

The Vanishing

Rim Rim Of Space

The Vanishing

The Vanishing

The Vanishing

The Vanishing

The Vanishing

The Vanishing

Rim o the World

The Rim Gods

Rim In the Box

Driving on the Rim

Driving on the Rim

Under the Tonto Rim

Under The Sweetwater Rim

The Vanishing of Katharina Linden

Rim o' the World

The Rim Gods

The Rim Gods

The Vanishing Languages of the Pacific Rim