Language Diversity in the Pacific: Endangerment and Survival 9781853598685

Citation preview

Language Diversity in the Pacific

MULTILINGUAL MATTERS SERIES Series Editor: Professor John Edwards, St. Francis Xavier University, Antigonish, Nova Scotia, Canada Other Books in the Series Language Planning in Malawi, Mozambique and the Philippines Robert B. Kaplan and Richard B. Baldauf, Jr. (eds) Community and Communication: The Role of Language in Nation State Building and European Integration Sue Wright Language Planning in Nepal, Taiwan and Sweden Richard B. Baldauf, Jr. and Robert B. Kaplan (eds) Can Threatened Languages be Saved? Joshua Fishman (ed.) Language and Society in a Changing Italy Arturo Tosi The Other Languages of Europe Guus Extra and Durk Gorter (eds) Motivation in Language Planning and Language Policy Dennis Ager Multilingualism in Spain M. Teresa Turell (ed.) A Dynamic Model of Multilingualism Philip Herdina and Ulrike Jessner Beyond Boundaries: Language and Identity in Contemporary Europe Paul Gubbins and Mike Holt (eds) Bilingualism: Beyond Basic Principles Jean-Marc Dewaele, Alex Housen and Li Wei (eds) Ideology and Image: Britain and Language Dennis Ager Where East Looks West: Success in English in Goa and on the Konkan Coast Dennis Kurzon English in Africa: After the Cold War Alamin M. Mazrui Politeness in Europe Leo Hickey and Miranda Stewart (eds) Language in Jewish Society: Towards a New Understanding John Myhill Maintaining a Minority Language John Gibbons and Elizabeth Ramirez Urban Multilingualism in Europe Guus Extra and Kutlay Yagmur (eds) Cultural and Linguistic Policy Abroad: The Italian Experience Mariella Totaro-Genevois Language Decline and Death in Africa: Causes, Consequences and Challenges Herman M. Batibo Directions in Applied Linguistics Paul Bruthiaux, Dwight Atkinson, William G. Eggington, William Grabe and Vaidehi Ramanathan (eds)

For more details of these or any other of our publications, please contact: Multilingual Matters, Frankfurt Lodge, Clevedon Hall, Victoria Road, Clevedon, BS21 7HH, England http://www.multilingual-matters.com

MULTILINGUAL MATTERS 134 Series Editor: John Edwards

Language Diversity in the Pacific Endangerment and Survival Edited by

Denis Cunningham, D.E. Ingram and Kenneth Sumbuk

MULTILINGUAL MATTERS LTD Clevedon • Buffalo • Toronto

The Editors wish to acknowledge the support and encouragement of the UNESCO Centre of the Basque Country, Bilbao and the Fédération Internationale des Professeurs de Langues Vivantes (FIPLV) in producing this volume.

Library of Congress Cataloging in Publication Data Language Diversity in the Pacific: Endangerment and Survival/Edited by Denis Cunningham, D.E. Ingram, and Kenneth Sumbuk. Multilingual Matters: 134 Includes bibliographical references and index. 1. Pacific Area–Languages. I. Cunningham, Denis. II. Ingram, D.E. III. Sumbuk, Kenneth. IV. Series. P381.P3L36 2006 409'.1823–dc22 2005021285 British Library Cataloguing in Publication Data A catalogue entry for this book is available from the British Library. ISBN 1-85359-867-4 / EAN 978-1-85359-867-8 (hbk) Multilingual Matters Ltd UK: Frankfurt Lodge, Clevedon Hall, Victoria Road, Clevedon BS21 7HH. USA: UTP, 2250 Military Road, Tonawanda, NY 14150, USA. Canada: UTP, 5201 Dufferin Street, North York, Ontario M3H 5T8, Canada. Copyright © 2006 UNESCO Etxea (Centre of the Basque Country), FIPLV, the Editors and the authors of individual chapters. All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from the publisher. Typeset by Wordworks Ltd. Printed and bound in Great Britain by MPG Books Ltd.

Dedication Professor Stephen A. Wurm This volume is dedicated to the memory of our dear colleague, Stephen Wurm, who passed away on 24 October, 2001. Most of the contributors to this volume knew Stephen personally and several of us had worked with him on a range of projects concerned with the disappearance of the world’s linguistic diversity. It was Stephen’s recognition of the value of this diversity which characterised his vision. His life and academic work were a celebration of this diversity and he contributed an impressive range of publications, the best known being the Atlas of the World’s Languages in Danger of Disappearing (1996)1. He also helped find a solution by bringing this topic to the attention of the wider public, politicians and international bodies such as UNESCO. We trust that the contributions to this volume continue the task, begun by Stephen, of documenting and finding answers to the threat to the world’s linguistic diversity. Peter Mühlhäusler University of Adelaide 1. Wurm, Stephen A. (1996) Atlas of the World’s Languages in Danger of Disappearing. Paris/Canberra: UNESCO Publishing/Pacific Linguistics. 2nd edition, 2001.

v

Contents Dedication to Professor Stephen A. Wurm Peter Mühlhäusler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Foreword Fèlix Martí . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Editors’ Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii 1. Language Diversity in the Pacific: Endangerment and Survival: An Overview D.E. Ingram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2. World Languages Review: Some Data Andoni Barreña, Itziar Idiazabal, Patxi Juaristi, Carme Junyent, Paul Ortega and Belen Uranga. . . . . . . . . . . . . . . . . . . . . . . 15 3. Naming Languages, Drawing Language Boundaries and Maintaining Languages with Special Reference to the Linguistic Situation in Papua New Guinea Peter Mühlhäusler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 4. Obstacles to Creating an Inventory of Languages in Indonesia: A Dialectology Perspective Multamia R.M.T. Lauder. . . . . . . . . . . . . . . . . . . . . . . . . . 40 5. Keeping Track of Indigenous Language Endangerment in Australia Patrick McConvell and Nicholas Thieberger . . . . . . . . . . . . . . . . 54 6. Papua New Guinea’s Languages: Will They Survive? Kenneth M. Sumbuk . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 7. Language Endangerment and Globalisation in the Pacific Darrell Tryon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 8. Endangered Languages of China and South-East Asia David Bradley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 9. On the Edge of the Pacific: Indonesia and East Timor John Hajek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 10. The Future of the Languages of Vanuatu and New Caledonia Jean-Michel Charpentier . . . . . . . . . . . . . . . . . . . . . . . . . 131 vii

Language Diversity in the Pacific Contents

viii

Language Diversity in the Pacific

11. Trends and Shifts in Community Language Use in Australia, 1986–1996 Michael Clyne and Sandra Kipp . . . . . . . . . . . . . . . . . . . . . 137 12. Directions for Linguistic Research: Forging Partnerships in Language Development and Expansion of the Domains of Use of Australia’s Indigenous Languages Rob Amery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 13. The Contribution of Language Education to the Maintenance and Development of Australia’s Language Resources D.E. Ingram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 14. Globalisation, Languages and Technology: Some Recommendations Denis Cunningham . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 The Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212

Foreword Fèlix Martí Chairman, UNESCO Advisory Committee for Linguistic Pluralism and Multilingual Education At a meeting on linguistic policy held in the Basque Country (Spain) in March 1996, the Director General of UNESCO, Federico Mayor, proposed that a report be drawn up on languages from a global point of view. A text was needed that would help the governments of UNESCO member states to have objective information on hand regarding the world’s linguistic diversity, the threats facing the universal linguistic heritage and the directions to be taken by linguistic policies at national and international levels. At the same time, a text was needed that would be circulated beyond the exclusive circles of experts. The UNESCO report would become a reference for many of those involved in language policy – teachers, journalists, social leaders, religious authorities, non-governmental organisations, businesses and cultural agents of all sorts. The proposal by the Director General of UNESCO was determinedly and generously taken up by the Basque Government, who entrusted the drafting of the report to UNESCO Etxea, the UNESCO Centre of the Basque Country, based in Bilbao. A management team was set up for the project, as well as a scientific committee and a technical commission in charge of preparing the text of the report. These structures began serious work in 1997. Two lines of approach were agreed. Firstly, a worldwide survey of linguistic communities was to be carried out to allow experts as well as representatives of the communities themselves to provide meaningful data and opinions on the present and future of languages everywhere. Thousands of questionnaires were sent out to informants on every continent and a sufficient number of completed replies was received to enable collation with significant figures. The aim was not to obtain a statistically representative sample but to gather meaningful opinions on the problems affecting languages and on the range of circumstances conditioning languages’ existence. The UNESCO World Languages Report was to take into account the contributions provided through the survey and build upon this through further consultation. Secondly, meetings of experts were organised on different continents to study the points of view of specialists on the present and future of languages on each continent. Thus, meetings were held in Mons (Belgium) ix

Language Diversity in the Pacific Foreword

x

Language Diversity in the Pacific

in 1998, Cochabamba (Bolivia), Elista (Russia) and Ouagadougou (Burkina Faso) in 1999, and in Mysore (India) in 2000. For the area of Australia and the Pacific, renowned experts were invited to contribute papers for publication. The present publication exists in that wider context. As the Pacific Basin is probably the most linguistically diverse area of the planet, it harbours the greatest complexity of linguistic contacts. The reflections of experts, of their experiences, of problems and expectations of the languages of Australia and the Pacific, are of considerable value in the context of the whole world language situation. Some of the countries studied in this publication can be considered as genuine laboratories for the observation of relations between linguistic communities. The chapters in this volume are intended to reveal how those languages in contact are evolving and to predict the future of the prodigious linguistic diversity of these countries and of the region as a whole. Many of them also offer recommendations to government and non-governmental agencies for the preservation of linguistic diversity as a treasure for the whole of humanity. There are many universities and sociolinguistic research centres that publish descriptive works or linguistic maps but in the UNESCO report (as in this volume), we focus on an analysis of the changes affecting linguistic communities and especially on the recommendations that ought to be taken into account to safeguard the linguistic diversity of each country, of each continent and of the whole of humanity. Far from being a descriptive study of the languages of the world, the report and this present volume are designed to have an orientation towards the future. It may be useful to remember that languages are subject to many conditioning factors. Linguistic communities evolve according to their adaptation to their surroundings and the replies they have to make to the political, economic, media, technological, social and cultural challenges affecting them. Knowledge of the factors that determine how languages evolve from the sociolinguistic point of view allows linguistic management beyond merely looking on at the shrinking and death of many languages or at new forms of linguistic colonialism. The UNESCO study and report as a whole, and the present volume on the Pacific area, constitute texts on linguistic ethics, in the sense that their observations and recommendations can help linguistic communities everywhere to plan their future. In the same way that there is now a widespread ecological ethic that proposes the protection and furtherance of all living species, these publications are texts on linguistic ethics serving the protection and the esteem of all languages. In a context heavily marked by economic relations, it is worth pointing out that linguistic diversity has an unquestionable economic value. Linguistic ethics can reveal that economic progress and the promotion of linguistic diversity are perfectly compatible. In addition, a linguistic ethic can underline the importance of respecting linguistic and cultural diversity in the furtherance of understanding between peoples and of security and peace.

Foreword

xi

This is why UNESCO refers to its linguistic programs as Linguapax, that is, peace between languages and peace through languages. A linguistic ethic must reveal the cultural nature of languages. Languages are not just equivalent means of communication. Each language is an epistemological and axiological universe. Each language offers original metaphors, ideas and symbols for understanding and expressing the world, society and human experience. Each language is a vehicle for values, ethics and aesthetics expressing desires and utopias. All languages have shared territories and are enriched through interaction with one another. The defence of linguistic diversity and the other aims of the linguistic ethic are part of a future-oriented project through which humanity can gradually replace violence with an ethic of reason, harmony and peace. We would not like to see the dizzying increase in relations between cultures condemn them to a future of uniformity and of repression of diversity on behalf of economic or political interests. We want a future in which the gradual assertion of an agreed common ethic is compatible with esteem for the diversity of peoples, identities and cultures. The opening up of all cultures and all linguistic communities need not involve their dissolution in a supposedly universal culture or language. What needs to be universalised is the admiration of diversity, the spirit of dialogue and shared responsibility in the search for global replies to global challenges. This publication – as well as the UNESCO World Languages Report – hopes to speak out on behalf of those who believe in the possibility of a linguistic ethic and in the possibility of linguistic policies inspired by principles of respect and diversity, of non-violent relations and of interest in cross-cultural dialogue. Universal linguistic diversity will probably not be something natural. In the future, linguistic diversity will be the result of freely taken social and cultural decisions. Up to now, many communities have maintained their linguistic personalities as a result of isolation or lack of communication. Languages with few links to political, economic or media power will now continue to exist only if the linguistic communities exercise a specific wish, that is, if they freely decide to continue using their language and if, also, the more powerful languages abandon their colonially oriented policies and adopt a linguistic sympathy for minority or relatively less widespread languages. From this point of view it seems important to encourage a form of language learning that can open the doors of multilingualism to all citizens of the world. The citizens of small linguistic communities must have the chance to be multilingual if they are not to live condemned to a linguistic ghetto. The citizens of more powerful linguistic communities must have the chance to be multilingual so as to be able to get their cultural coordinates into perspective and establish fraternal relations with the speakers of weaker languages. Language learning is the basis of the linguistic empathy the speakers of all languages must share with one another.

Editors’ Note This volume should be seen in the context of the UNESCO-sponsored study of the world’s languages and the World Languages Report being undertaken by the UNESCO Centre of the Basque Country in Bilbao and the UNESCO Centre of Catalonia. The preparation of this volume brought together eminent linguists and applied linguists whose research has made them expert in the Asian-Pacific region, indubitably the linguistically richest and most diverse area of the globe. The sheer vastness and linguistic diversity of the Asia-Pacific region provide the greatest of challenges to linguists, researchers and educators both in documenting that language wealth and in finding ways to maintain and develop it in the face of heavy contrary pressures arising not least from the dominance of world languages such as English and French and the impact of the many other features of globalisation. It is fitting that a text be devoted to the challenges of linguistic diversity in the Pacific, the dangers confronting it and the strategies and actions that are being taken to try to ensure its survival. As the editors of this volume, we would like to acknowledge our gratitude to the many organisations and individuals who assisted in making this publication possible: · the UNESCO Centre of the Basque Country and the UNESCO Centre of Catalonia for their initiative and expertise and for their support in bringing together the contributors for this volume; · the Australian National Commission for UNESCO, for its support; · the members of the Scientific Committee of the World Languages Report for their persistence in the cause and, in particular, Professor Fèlix Martí for his invaluable assistance and advice; and · the Fédération Internationale des Professeurs de Langues Vivantes for its support and, especially, that of its Executive members and their liaison with UNESCO. For the preparation of this publication, we would like to thank all who contributed to the volume and provided assistance in editing the papers. Gratitude is also expressed to Multilingual Matters by both the editors and the other contributors, for their support and advice on the compilation of this book. Denis Cunningham, David Ingram and Kenneth Sumbuk xii

Language Diversity in the Pacific Editors’ Note

Color profile: Disabled Composite Default screen

Chapter 1

Language Diversity in the Pacific: Endangerment and Survival: An Overview D.E. INGRAM The Pacific Basin but, more specifically, the countries of the Western Pacific, the focus of this volume, constitutes, as Martí says in the Foreword to the present volume, ‘the most linguistically diverse area of the planet [and] harbours the greatest complexity in linguistic contacts’. This volume demonstrates the linguistic diversity of the Pacific countries and the enormous wealth of languages and cultures that that wealth entails. It also reveals in clear terms the danger that that linguistic and cultural diversity might be at least reduced, if not lost altogether, unless steps are taken to safeguard the diversity, to recognise the economic and cultural value of linguistic and cultural diversity, and to recognise, as Martí says, ‘the importance of cultural diversity in the furtherance of understanding between peoples and of security and peace’. However, language diversity and especially the minority languages within that diversity will continue to exist only if the communities of speakers themselves value the languages and continue to use them and if all societies recognise their value, accept multilingualism as the desirable norm, and adopt educational and social policies and practices that support and foster multilingualism. The chapters in this volume fall into three broad parts. If the languages of the Pacific are to continue to exist, the first step is to identify them, study them, and record their form and usage. Hence, the first four chapters (those by Barreña et al., Mühlhäusler, Lauder and McConvell and Thieberger) deal especially with issues of data collection, as does, in part, Sumbuk’s chapter in the next group. The next six chapters (Sumbuk, Tryon, Bradley, Hajek, Charpentier and Clyne) describe the state and use of languages in various parts of the West and South-West Pacific, focusing especially on particular language communities. Many of the languages of the Pacific, both those that are indigenous1 to the region and those that, while no less languages of the region, have been brought to the West and South-West Pacific by immigration. If the language diversity is to survive, it is fundamental that educa-

1

3 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

2

tional and other policies must support it and so the final group of chapters (those by Amery, Ingram and Cunningham) mainly discusses practical issues concerning language revival, maintenance and education.

Data Collection Issues Probably the most ambitious project, which highlights some of the difficulties in gathering comprehensive data on the world’s languages, is the UNESCO World Languages Review, described and discussed by Barreña and others in the first chapter. The aim of the project was to describe the problems affecting languages and, to this end, a survey of the world’s languages was undertaken by questionnaire. The chapter draws attention to the problems encountered in trying to obtain information on the state of the world’s languages but it also reveals some of the problems facing many languages in the context of what the authors call the ‘global tendency towards language homogenisation’. Though the number of languages actually reported upon by the time this chapter was written was only a small proportion of the world’s languages (at 725 fewer, in fact, than the languages of the small country of Papua New Guinea in the South-West Pacific), nevertheless the study identifies many of the problems confronting languages around the world, such problems as family transmission, the impact of the more prestigious languages, population movements, and mixed marriages. The survey highlighted the extent of the problem since almost half (some 42%) of the languages reported were in danger of disappearing with the dominant reason in most places being economic or cultural subordination or discrimination. Where languages were more secure, this was, according to the survey, generally because there were language policy and planning organisations that supported the languages and because they are used in education. However, other uses, for example, in the media, and the attitudes of the speakers and other communities towards any language were also influential. Mühlhäusler and Lauder focus on different aspects of the problems involved in studying languages, especially those in remote parts of the globe. A most basic problem for researchers exploring language diversity is, as Mühlhäusler argues, the very notion of what a language is and how the users of languages conceive of a language. Even the naming of languages is fraught with difficulty with most names having been imposed by outsiders using any of a great variety of non-linguistic or linguistic sources. This, in turn, leads to difficulties in classifying languages, in understanding their relationships and status, and even in recognising whether two languages are different or, in fact, the same language, i.e. in drawing boundaries between languages and what may or may not be dialects of a single language. Thus, Mühlhäusler concludes that vast prob-

4 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Endangerment and Survival: An Overview

3

lems remain in identifying and counting languages and dialects. He concludes by questioning whether the influence of linguists has always been good and has contributed to sustaining language diversity or whether their imposition of descriptions based on systems developed in other linguistic contexts may actually have been harmful. He also wonders whether the rate of change experienced by the languages of Papua New Guinea and Australia may have been too fast to allow adaptation to occur. Lauder is an Indonesian linguist who works amongst the languages in remote villages and islands of Indonesia. Indonesia, with 10% of the world’s languages, is said to be second only to Papua New Guinea in the number of languages in the one nation. Lauder draws on her experience of language mapping, especially in remote areas, to discuss the problems that confront linguists undertaking an inventory, description and the preservation of Indonesia’s linguistic diversity. All the data gathered is computerised, and each village in Indonesia is identified and linked to the linguistic data that is gathered and entered into the computerised database with 10% of the world’s languages. Lauder discusses the problem of selecting and training field workers and of identifying informants who, together with their immediate ancestors, have to be natives of the village being studied. A standard questionnaire is used to elicit local lexis for a standard set of words and phrases, while local folk tales are also recorded. Particular problems exist with isolated tribes, partly in identifying that they actually exist and partly in then accessing them. These problems are aggravated by the lack of secure data about them in official government records and by the fact that, because many of the houses that have been built for these people have been inappropriately designed, they are not occupied. The languages of the isolated tribes have been of particular interest for dialectology, describing boundaries between languages and dialects. Lauder also discusses the variety of viewpoints from which languages might be studied and she goes on to describe the array of difficulties that linguists encounter in trying to visit the isolated tribes in order to map their languages. These issues include gaining government permits, the reliability of the information gathered, problems in using guides and interpreters, the logistics of actually getting there, and how to cope within the customs of a tribe. McConvell and Thieberger mainly focus on how to keep track of the changes in the viability of indigenous languages in Australia. They emphasise how important it is to accurately describe the current state and trends in Australia’s languages if effective policies and programs for their maintenance are to be identified and maintained. At the same time, they provide a picture of the state of indigenous languages in Australia. Of the 250 indigenous languages spoken when the first Europeans settled at Sydney Cove in 1788, more than half are no longer spoken; of the remaining 100 or so, at least half are spoken by only a dwindling number of old people;

5 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

4

and the other 50 or so also show signs of language shift towards forms of English and are also endangered. McConvell and Thieberger consider a variety of means by which to identify the current state of languages but see the Australian census as the most practical tool even though its questions and data are limited and need expansion. They note that the inclusion of indigenous languages in the five-yearly Australian State of the Environment report is a ‘breakthrough’. They dispute the use of minimum absolute numbers of speakers as an indication of endangerment, especially since most of the figures proposed (from 10,000 to 100,000) would suggest that all Australian languages were ‘doomed’, even before European settlement. They refer to the geographical pattern of endangerment with the strongest language areas being those where more than 79% of the population speak the language at home. The languages have died out where European settlement occurred earliest and with most impact on the indigenous people. They note that, while the percentage of people who speak an indigenous language at home shows an accelerating decline, especially in the oldest age group, there is some indication that the decline may have slowed in the 15–24 age group. At the same time, they acknowledge that, because of the pyramidal age structure of the indigenous population, percentage figures may be disguising a growth in absolute numbers. It is also possible, perhaps as a result of more prestige being attached to being indigenous and speaking an indigenous language, that there is some over-reporting of indigenous language skills. In discussing regional patterns of language shift, they especially note the success of an active language revival movement around Kaurna, in Adelaide, a project reported on also by Amery in a later chapter of this volume. McConvell and Thieberger go on to consider various measures of endangerment, including the usefulness of the Australian census, the Canadian approaches, the use of age profiles of the indigenous population, and dedicated language surveys though the lack of consistency in the frameworks used for local and state surveys is a serious problem. Language endangerment can be classified in various ways. In particular, McConvell and Thieberger discuss approaches based around Wurm’s ranking of languages from ‘extinct’ to ‘full tribal use’ or ‘not in danger’ and suggest some modifications. They see value in tracking the state of, and changes in, the occurrence of indigenous languages in order to build an indigenous language ‘ecology’ that will enable help to be given where speakers wish to maintain their languages and so that people can know whether language maintenance programmes are actually working or not.

6 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Endangerment and Survival: An Overview

5

The State and Use of the Languages of the West and SouthWest Pacific The second group of chapters in this volume largely describe the state and use of languages in the West and South-West Pacific. Sumbuk’s chapter discusses the state of the languages in what is probably the world’s most linguistically diverse nation, Papua New Guinea (PNG), which has some 860 languages amongst a population of about 5 million. Sumbuk questions Krauss’s estimate of the need for 100,000 speakers if a language is to survive, and points out that many languages in PNG have co-existed harmoniously for a long time with as few as 50 speakers. Hard data on the linguistic diversity of PNG (the actual number of languages and the number of speakers) is not available because there has been no national linguistic survey. Thus, it is difficult to identify the actual linguistic trends, though studies of individual languages suggest that massive language extinction could already be occurring. Nevertheless, many languages with numbers well below 100,000 have survived for a long time. Sumbuk illustrates some of the issues by describing the specific cases of two neighbouring languages, Kaningara and Sare (the latter being Sumbuk’s own mother tongue). He also discusses the factors that will influence the survival of PNG languages, including such factors as technology, the need for documentation of the languages, globalisation, education (or the lack of it), and the social, political and economic rights of the people. He concludes by pointing to three priority needs if the languages of PNG are to survive: (1) a nationwide linguistic survey documenting the number of languages and their speakers and trans-generational language transmission; (2) education about the importance of the languages; (3) the need for the people and the government to decide whether they really want to preserve the nation’s linguistic diversity. Tryon discusses the state of the languages in the region known as Oceania, stretching from the island of New Guinea eastwards as far as Easter Island. This is probably the linguistically most diverse region of the world with approximately 1200 languages, generally spoken by very small communities for whom the language is seen as an ‘ethnic badge’ or ‘identity marker’. There are two major language families in the region (the Austronesian family of some 550 languages and the Papuan family of about 750 languages) together with a number of pidgins and creoles used as lingua francas and the main colonial languages of English and French. The Austronesian or Malayo-Polynesian family extends from Singapore in South-East Asia to Easter Island and is believed to originate in Taiwan. Languages of the Papuan family are mainly spoken on the island of New Guinea, but are also found on some of the Indonesian islands, Timor, the

7 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

6

Bismarck Archipelago, Bougainville, the Solomon Islands and east to Santa-Cruz. These languages are unrelated to the Austronesians, are more ancient, and have been dated back as much as 50,000 years at sites in Papua New Guinea. Tryon provides an extended discussion of the groupings within both the Austronesian and Papuan families of languages. A significant factor in the maintenance and preservation of these languages is the increasing diaspora of the peoples of this region with considerable movement within the region and with large numbers in the Pacific rim countries and as far away as France. The linguistic consequences of this diaspora are severe, with language loss or impoverishment occurring amongst the displaced peoples, not least because of the impact of English or of Englishbased pidgins and creoles such as Tok Pisin in Papua New Guinea or Bislama in Vanuatu. Despite the remaining multiplicity of languages within the region, considerable language loss has already occurred. The linguistic diversity has resulted from a variety of factors including geographic isolation, the rugged topography in New Guinea, inter-tribal hostility, and the contiguity of, and interaction between, the Austronesian and Papuan languages. Tryon draws attention to the urgent need for more documentation of these languages and notes that many are under threat from increasing urbanisation, the increasing use of pidgins and creoles, the impact of English and French, political decisions to use a language (generally a pidgin or creole) for nation-building purposes, evangelisation (commonly based around an English-based pidgin), and inter-marriage. Tryon concludes by pointing to how little is being done to arrest the loss of languages (except in Vanuatu where the local people are making efforts to record local languages, preserve the records, and transmit them to their children). Bradley briefly outlines the current policy framework for endangered languages and the distribution of endangered languages in China, especially in the south-west and in the neighbouring countries of South-East Asia. He says that China illustrates both the problems of, and the solutions to, language endangerment. Whereas in other parts of the West and SouthWest Pacific, the dimensions of the language endangerment are reasonably well understood, this is not the case in the region that he discusses. Part of the problem seems to lie in the difference between the officially designated national minorities (55), and the great linguistic diversity within any minority. Acontrast is drawn between such major groups as the Han, which share a common character-writing system and a sense of identity despite linguistic differences, and many of the ‘composite national minorities’ that lack such unitary feelings. Within the Yi of South-West China, for example, there are six official ‘dialects’, but more than 100 mutually unintelligible languages within them. While there has been considerable linguistic work on the major language recognised as the ‘standard’ for each minority, that is

8 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Endangerment and Survival: An Overview

7

not so for the other languages in each group. In Thailand, much useful linguistic work has been done, Malaysia has a policy of support for indigenous peoples, but there is no support for their languages and cultures and the use of Malay is spreading rapidly. The state of the minority languages in Laos, Vietnam, Cambodia and Burma is also difficult and, though some documentation has been carried out in Laos and Vietnam, throughout these countries priority is given to the national languages with a consequential adverse effect on the languages of the ethnic groups. In Burma, for instance, the policy of using only Burmese for official purposes and education has led to the situation where, in Bradley’s view, the languages of 135 recognised ethnic groups are endangered, with only Burmese being secure. Bradley reports on his surveys of languages in Yunnan Province of China, and suggests that there are many more unreported languages. The cause of this situation, he notes, lies in several facts: local authorities are not much interested in preserving minority languages, the small groups are classified within a larger national minority, and, as administration, education, communications and economic development have spread, putonghua, another local variety of Chinese, or the language of a larger local minority has displaced the traditional languages. He provides a case study of the Sanie people in and around Kunming, which, he says, presents ‘a typical example of language death in progress’. The exact situation differs according to the proximity of the village to the major urban centre and hence the impact of the Han Chinese. Only in the most remote villages are the young people fluent in the traditional language while the youngest children, in many instances, do not speak Sanie. Part of Bradley’s response has been to survey the languages and to devise a romanisation for Sanie at the request of the local authorities. He concludes by noting that, in addition to the 144 languages known to be endangered in the countries identified earlier, there are many others yet to be described and many more are endangered to some degree. This region is of particular importance because it has about 10% of the world’s linguistic diversity and, unlike many other parts of the world, the languages are not being replaced by Indo-European languages but by more or less closely related languages. Part of Bradley’s conclusion is worth quoting as an overall summary, not only of what he says about South-West China and neighbouring countries but of what this book is about: For more than ten years, UNESCO and related bodies ... have been supporting urgent work to document disappearing languages and, where possible, to reverse language shift ... It is of crucial importance that the momentum achieved should continue so that we may try, as linguists, to avert the ecolinguistic catastrophe widely predicted for

W:\Clients\MLM\Cunningham\Proof2\Cunningham Proof 2a 31 January 2006 20:38:05

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

8

this century: the disappearance of 90% or more of the world’s languages. (Bradley, this volume) Hajek reports on language vitality and endangerment in Indonesia and East Timor (in contrast to Lauder, who, in the chapter already discussed, mainly focuses on the difficulties of fieldwork and data gathering in Indonesia). East Timor was part of Indonesia until 1999 when it gained its independence but the two countries are in stark contrast: Indonesia has a population of some 206 million, East Timor approximately 800,000, Indonesia has 731 languages and East Timor just 20. Austronesian and Papuan (or non-Austronesian) languages occur in both countries. Hajek briefly outlines the turbulent history of East Timor, its long administration by Portugal and then by Indonesia for a further 23 years. In the circumstances, the indigenous languages have shown considerable resilience and few languages are seriously endangered or extinct. He describes at some length the diverse linguistic situation in Indonesia. Only three languages, Javanese, Sundanese and Madurese have extremely large numbers of speakers. Bahasa Indonesia (Indonesian) is promoted as the sole official language and as a unifying link among the Indonesian people. Its spread has been successful, growing from 40% of the population in 1970 to 67% by 1990 as a result of compulsory schooling, the mass media, and the government policy of transmigration in which large numbers of people from Java and Bali have been re-settled elsewhere. In addition, mother tongue teaching in the lower school grades is restricted to those languages with more than a million speakers. Religious influences have also contributed to the decline since the Catholic Church adopted a policy of using only Indonesian while, in some areas, conversion to Islam has encouraged a shift to Indonesian or Malay. Consequently, Hajek points to census figures that indicate a substantial decline in the number of young speakers of the larger languages. He provides a region-by-region report on the languages of Indonesia, noting that the diversity increases the further east one goes and, in addition to the factors just listed, identifies such influences as the expansion of the oil and logging industries, natural disasters, and sectarian turmoil as impacting on the survival of many of the languages. In some areas (Maluku and West Papua, for instance) the average size of the language communities is small with a quarter having fewer than 500 speakers and so especially vulnerable. Charpentier focuses on the two Melanesian island groups of Vanuatu and New Caledonia, in both of which the French were the colonial power, together with the British in the condominium of New Hebrides (now Vanuatu). The indigenous languages are Austronesian with 40 languages in New Caledonia (38 living and 2 extinct) and 106 in Vanuatu (of which 8 are nearly extinct and 17 moribund). In both areas, the languages vary in

10 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Endangerment and Survival: An Overview

9

size from as few as 10 speakers to Dehu in the New Caledonia group with 15,000 speakers and Lenakel in Vanuatu with 11,500 speakers. In Vanuatu, the indigenous languages were protected for a long time by its island structure and its remoteness from the rest of the world. Nowadays, however, the high mobility of the people, the cash economy, inter-marriage, the closeness and mutual intelligibility of the languages, and the people’s readiness to accept other languages (including the pidgin, Bislama) have reduced their security. Charpentier also notes other, especially cultural, factors that have impacted on the languages and political policies that seem to favour European languages. In New Caledonia, on the other hand, he notes that the former system of reserves has served to foster a diglossic situation and to protect the indigenous languages, with French and Western culture being separated from the indigenous languages and traditional culture with limited interference between them. Unlike the situation in other countries, French and English seem to have a smaller and, he says, decreasing influence. In Vanuatu, only Bislama seems to be expanding while, in New Caledonia, French is the only lingua franca and the only one permitted for education, governance, business and media. Contact between languages has different effects: Bislama in contact with English becomes Anglicised whereas, when Bislama interacts with the indigenous languages, it is the indigenous languages that are affected. In New Caledonia, French and the indigenous languages serve different purposes in different semio-cultural worlds with fewer adverse effects on the local languages. In Vanuatu, more efforts seem to have been made to secure the indigenous languages with numerous plans from religious, political and international organisations, though not all of these have been carried through. All the languages (106 indigenous languages plus English and French) have status under the constitution. Overall, it seems that French will continue to decline in Vanuatu to be replaced by Bislama, but a Bislama more and more influenced by English. In New Caledonia, the present diglossic situation is likely to continue for the foreseeable future. Unlike the other chapters in this section, the chapter by Clyne and Kipp focuses especially on the language diversity in Australia resulting from immigration. Their data is based on census findings of the Australian Bureau of Statistics, though they indicate that the extent of language use in Australia is undoubtedly under-estimated because of the nature of the question used, which asks what language is used at home. According to the 1996 census, 14.6% of Australians use a language other than English at home. The pattern of languages has changed with the pattern of migration,2 which has, itself, changed with different waves from different parts of Europe and Asia, the latter predominating in recent decades. Most recently, the nation’s language diversity has increased with substantial communities speaking such languages as Chinese (especially Mandarin), Vietnamese

11 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

10

and Macedonian. On the other hand, more traditional languages such as Greek, Italian and Maltese have declined significantly. Languages such as Tamil, Hindi and Korean, though small in actual totals, have increased considerably in percentage terms. Economic values have influenced the choice of languages in the education system, with priority given to Chinese (Mandarin), Indonesian, Japanese, and Korean. The maintenance pattern also differs across languages, with a relatively high percentage of Australianborn speakers of such long-established languages as Greek, Arabic, Italian, Macedonian and Turkish but with low maintenance rates for Dutch, Hungarian, and German. Geographically, most migration has been to urban areas, especially to Sydney and Melbourne, with the pattern of languages also differing according to the area most favoured for settlement by different ethnic groups at different times in the history of migration. Melbourne received a higher proportion of post-World War II European migration, Sydney receiving more from the newer (predominantly Asian) waves. Brisbane also has received more migration in recent decades and Queensland, being the most decentralised State, is the only one with significant numbers of community language speakers outside the main urban areas. More than 60% of the home users of Aboriginal languages are in the Northern Territory. Clyne and Kipp consider at some length the occurrence of language shift, i.e. the percentage of people born in a particular country who now speak only English at home (first generational language shift) or the percentage of people born in Australia with one or both parents born overseas who now speak English at home (second generational language shift). The rate of first generational language shift differs between languages, with cultural distance from English being a major factor in determining its extent. The closer the culture is to English, the more likely it is that language shift will occur – in contrast to Chinese-speaking cultures, for instance, which have a relatively low shift rate. In addition, languages are best maintained in those States where they are most represented, and Clyne and Kipp conclude that relative concentration of speakers is more important than absolute numbers of speakers in favouring language maintenance. They identify the birthplace of the parents as an important factor influencing language shift, which is greater where only one parent is from the country of origin of the language. Other factors considered include the time when migration occurred, gender, age and period of residence in Australia with males, for instance, tending to shift more to English in the home than females. Clyne and Kipp conclude that the data suggests ‘the ultimate inevitability of home language shift in the immigrant Australian context’. Looking to the future, they predict that Arabic, Cantonese and Vietnamese will gradually displace Italian and Greek as the most widely used languages in Australia (after English). They comment that the advent of more ‘pluralistic’ (i.e. ‘multicultural’) policies over recent decades may

12 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:26

Color profile: Disabled Composite Default screen

Endangerment and Survival: An Overview

11

be contributing to the relatively lower rate of language shift in the ‘newer’ community languages.

Practical Issues concerning Language Revival, Maintenance and Education The final group of three chapters considers practical ways in which the survival of languages may be assisted. Amery, whose focus is on indigenous languages in Australia, starts by considering practical issues in the description and documentation of languages, before focusing on language development and usage. The lack of proficiency in indigenous languages by government personnel and others interacting with indigenous people inevitably reinforces the role of English and reduces the domains in which indigenous languages can be used. This has not been helped by the ‘strong tendency’ for linguists to focus on traditional domains so that, for instance, very few items relating to ‘introduced concepts’ or ‘post-contact items’ appear in word lists or dictionaries of indigenous languages. Where efforts have been made to introduce words to enable, for example, health workers to communicate in the language of their patients, these words do not always survive very long. However, if the languages are to survive, there is need for them to be used in all the domains in which the people use language and for them to be seen and treated by linguists as living languages responding to the needs of the people today. Consequently, there is need for intensive language development, analogous to the development that has occurred in Maori in New Zealand, where thousands of new terms needed for use in such domains as science, technology, government, or economics have been introduced. Amery emphasises, however, that language development cannot be considered in isolation from the massive health and social problems that exist in many indigenous communities in Australia. He illustrates his discussion of language development with reference to the revival work he has been involved in with Kaurna, an indigenous language of Adelaide and the Adelaide Plains, which had not been spoken on a daily basis since the 19th century. He describes how the project identified the language and developed it further, its use as an ‘auxiliary language’ in the public domain, and its introduction to all levels of education. Amery concludes by emphasising that language revival is not just a matter of examining a language as an historical relic, rather it needs to be transformed to meet present day needs. Ingram’s focus is on some of the ways in which language education can contribute to the maintenance and development of Australia’s language resources and enable the languages to contribute to the social and economic development of the nation. He makes three basic assumptions in the chapter: that languages are intrinsically and extrinsically valuable; that a

13 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

12

language will survive only so long as it serves some useful purpose; and that the language resources of a community will survive only to the extent that the community values them and that all members of the community respect and value the cultures and languages of other people in the community. Fundamental to this is the importance of all people’s manifesting cross-cultural attitudes that are conducive to harmonious and rewarding lives in a culturally and linguistically diverse society. These issues are considered in the context of multicultural Australia and its language education policies under which three broad categories of languages are taught: ‘traditional’ European languages such as French and German, community languages including both indigenous languages and those resulting from immigration, and languages of economic or international political significance such as the priority Asian languages (Chinese, Indonesian, Japanese and Korean). A number of corollaries flow from these basic assumptions: languages will survive only if they serve some useful purposes which, in a multicultural society dealing with a linguistically diverse world, are not hard to find. Not least, there is a need for Australian industry to recognise the value to it of the nation’s language resources. If industry is to make use of the languages available, industry’s language needs must be identified through, for example, language audits and needs analyses, the available skills must be identified and certified, and language courses must be designed and taught, and be long enough to produce real practical proficiency at a level at which they can contribute to the needs of industry. If education is to contribute people with real and usable skills, there are profound implications for teacher quality and supply, an area that received insufficient attention in the national and State language policies that were developed in the 1980s and 1990s. All of this is contingent on attitudes being favourable, especially attitudes towards other cultures and their languages. Most language policies and syllabuses endorse the fostering of positive cross-cultural attitudes as one of their central goals but the mere endorsement of such a goal does not ensure its attainment. It is necessary for the design of courses and, especially, the teaching methodology to encourage interaction between the learners and native or other speakers of the language, to encourage students to consider issues of race and inter-cultural relationships, and to encourage them to introspect and modify their own attitudes. In addition, if the richness of the language resources of a nation such as Australia is to survive, the speakers of the languages must take pride in their language skills and value their languages. For this, the education system has an obligation to educate students to understand the nature of language and why languages are intrinsically and extrinsically valuable. Society at large also needs to value languages, to recognise the right of people to use and maintain their languages, and to respect the languages and cultures of the community. At

14 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Endangerment and Survival: An Overview

13

the same time, English has an essential role in Australia, anyone without English in Australia inevitably must suffer disadvantages and consequently all people need the opportunity to develop proficiency in English to a level that satisfies their needs. Rather than viewing itself as monolingual, the Australian population needs to regard bilingualism or multilingualism as the norm and to support language diversity through, in particular, the education system and other services. If these goals are to be achieved and the values that underlie them are to be supported, comprehensive language policy-making at the national and State levels must be firmly in place. This was progressively so during the 1980s and the first half of the 1990s, but has been largely neglected since 1996 when a change of Federal government occurred. In the final chapter, Cunningham, who, at the time of writing, has been President of the Fédération Internationale des Professeurs de Langues Vivantes for some years, considers especially the role of technology in supporting language maintenance and development. He also reviews the global state of languages, the factors that are endangering language diversity, and the important role that UNESCO can play in exploiting the technology in ways that can support rather than reduce language diversity. He refers to the potential role of languages in promoting communication, equity, access and peace, and discusses languages and technology in the context of the need to foster multilingualism. Perhaps the greatest threat to global linguistic wealth is the dominant role of English, which means that English is the most frequent choice for learning a second language and its dominance demotivates English speakers from learning another language. Cunningham discusses the issue of language death and cites Crystal’s estimate that between 50 and 90% of the world’s languages could disappear this century. Cunningham’s principal focus is on the role of technology in either hastening the decline of languages or in assisting in their survival. He discusses the range of technology that is increasingly available but notes that at least half of the world’s population has no access even to anything as basic as electricity and the telephone. The Internet is of major importance, partly because of the easy access it provides to a wealth of information. Though English predominates on the net, other languages (such as Chinese and Spanish) are rapidly catching up and it is estimated that, by 2007, Chinese content will exceed that of English. In developing an agenda for action, the first need is to document the state of the world’s languages and UNESCO, Cunningham asserts, is well placed to undertake such a task and to work towards greater equity of access to technology. In considering the needs implied by the gap that exists between the technological ‘haves’ and ‘have-nots’ across the globe, and in considering a possible role for UNESCO, Cunningham makes a variety of recommendations for action in such areas as the Internet, communication and information, hardware and

15 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

14

software, and web search engines. His recommendations focus especially on the steps needed to support the world’s linguistic diversity and to strengthen languages at risk. In conclusion, he notes that the world’s languages are under siege as minority languages disappear and as English assumes a more and more dominant role as an international lingua franca. Technology, he says, threatens to diminish language diversity but this can be counteracted if more equitable access to technology is encouraged. Almost in summary of this volume, Cunningham concludes: We must make a decided effort to retain the linguistic wealth currently enjoyed by the globe, by impacting on the issues of policy and planning, teacher training and development, and student learning ... What we as leaders, as policy-makers, and as language educators can do is to take all reasonable steps to ensure that the linguistic diversity of our multilingual globe is ensured and enhanced for future generations. (Cunningham, this volume) Notes 1. For consistency throughout this volume, ‘indigenous’ is written without capitalisation though, nowadays, in Australian texts, it is commonly capitalised. 2. In Australia, the term ‘migration’ is commonly used instead of ‘immigration’.

16 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Chapter 2

The World Languages Review: Some Data ANDONI BARREÑA, ITZIAR IDIAZABAL, PATXI JUARISTI, CARME JUNYENT, PAUL ORTEGA AND BELEN URANGA

Introduction The World Languages Review dates back to the late 1990s, and the signing of a Memorandum of Understanding between the Basque Government and UNESCO in Paris in July 1997 to undertake the project. The most important objectives of the project were to survey the situation of languages and the problems that affect them across the world, and to facilitate decision-making on policies to recover languages in danger of extinction. The aim, in other words, was to describe the problems affecting languages in order to make people aware of the importance of maintaining the world’s linguistic heritage. Recommendations were also to be made to protect living languages. The present chapter provides some of the data gathered in the Review.

The World Languages Review The World Languages Review aimed to obtain first-hand information by gathering the opinions of language communities on the situations of their languages. A questionnaire of 40 questions, many of which were openended, was drawn up. The answers to these questions provided data about the characteristics of languages and of their linguistic communities: their denomination, uses, representations, attitudes and the linguistic expectations shown by the speakers of different languages. The questionnaire was drawn up according to criteria now classical in sociolinguistics (see Haugen, 1972). One of the virtues of this questionnaire was that it allowed informants great freedom in their answers. They were able to supply the facts that they felt to be most relevant, regardless of whether or not specific questions elicited them. The fact that we used open-ended questions and that the informants were very diverse (organisations, linguists, members of the community, etc.) created some difficulties for the analysis of the objective data and for the representation of the language reality as given by the

15

17 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

16

informants. By 2001, more than 1000 questionnaires had been returned. The total number of different languages to which the project had access was 725. [For more information about the questionnaire and the languages, the reader can visit the following site: http://amarauna-languages.com.] The analysis, however, was carried out using a sample of 525 languages – the ones identified in the questionnaires returned by July 2001. Initially, different samples had been taken from, first, 100 languages and, subsequently, 400 languages and it was confirmed that the general tendencies were the same throughout the analysis. The World Languages Review Technical Committee also participated in international meetings with scholars and language researchers all over the world. Five important meetings took place to obtain information about the situation of languages and the problems that affect them in various regions of the world: • • • • •

Bolivia (Cochabamba) on March 3–6, 1999; Kalmukia (Elista) on May 10–16, 1999; Burkina Faso (Ouagadougou) on June 14–17, 1999; India (Mysore) on March 6–11, 2000; Australia (Melbourne) on April 26–28, 2001.

In these meetings, direct relations were established with different researchers, institutions and organisations so as to contrast and increase the information received.The creation of a worldwide network of reliable informants prepared to answer the questionnaire was of fundamental importance to the research. This network was established in various ways, using a bibliography on relevant aspects of the project and using databases with addresses of persons and institutions, religious groups, organisations of indigenous peoples, etc.

Selected Data from the 525 Questionnaires Analysis of the data obtained from the 525 sample questionnaires allowed the project participants to comment on some remarkable aspects. The informants The Technical Committee tried, often successfully, to obtain first-hand information, that is, to ensure that the information came from informants who were either members of the relevant linguistic communities or were closely related to them. Thus, more than half the informants (approximately 60%) declared themselves members of the linguistic community on which they reported. Almost 40% said they were not members of the community, but they were generally researchers or were working for the community in some way. Some researchers, however, identified themselves as members of the community precisely because of their work, or because they had learned the language.

18 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

The World Languages Review: Some Data

17

The proportion of informants claiming to belong to the linguistic community for which they supplied information was, we believe, an important factor since, even though it added subjectivity to the information obtained, nobody knows a linguistic community as well as its own members. Thus, the involvement of the community members, far from detracting from the Review’s validity, enriched it. In addition, since one aim of the project was to create awareness and to help reverse the trend towards global linguistic uniformity, the participation of members of a language community in the project made them more aware of, and determined to do something about, their own language’s situation. Official status One important aspect to take into account when analysing the situation of a language is whether or not it has been accorded any kind of official or co-official status, either on a national or a regional basis. Of the languages about which information was received, 28% were official or co-official, 37% did not have any official status, even though they may have had some recognition, and 28% did not have any recognition at all. From the initial data, it is noteworthy that a relationship exists between the official status of a language and the informants’ appreciation of the vitality of their language. As a result, it is to be noted that most of the languages that had declined were languages that did not have any official recognition. Similarly, of the languages that increased their number of speakers, 59% were official and 40% were co-official. Transmission Familial transmission, which, by definition, occurs when children acquire the language from their parents, is an important factor in considering the possibility of maintenance of the community language. There is no doubt that a transmission break generates linguistic substitution (i.e. the parental language is replaced by another) and, if such transmission breaks are generalised, they can produce linguistic substitution across the whole community. The data about family transmission of the language in question was gathered by direct questioning. According to the informants participating in the study: • 53% of languages were transmitted within the family; • 43% of languages were not transmitted on a regular basis; • 4% of the informants did not answer. According to this data, only 53% of the languages analysed were widely and normally transmitted. However, among those languages that were not transmitted on a regular basis (43%), some differences can be highlighted.

19 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

18

In 23% of the communities, transmission existed but the influence of more prestigious languages was already evident and transmission was maintained only in the more impermeable nuclei. In 8% of the languages analysed, transmission was practically interrupted and language shift imminent. Finally, 12% of the languages were not transmitted at all, and it is foreseeable that such languages will disappear along with their last remaining speakers. The fact that transmission had been altered in 43% of the cases, although in varying degrees, is a matter of serious concern. The most frequent reasons for the interruption of intergenerational transmission were pressure by other cultures or languages, population movements (emigration and immigration) and mixed marriages. A further group of reasons can be added to these: negative linguistic attitudes arising both from surrounding community members and from among speakers of the languages themselves. Together with the data about the intergenerational transmission of languages, it is useful to consider intergenerational usage. It can be predicted that, in communities where language transmission is complete, use across generations does not differ. However, in linguistic communities where the transmission break is beginning to take place, the use of the language increasingly declines from the oldest generations to the youngest and this illustrates the tendency towards language substitution in such communities. In fact, in the languages examined, the older generations spoke the language of the community in 65% of cases, adults spoke it in 54%, young people in 36% and children in 38%. In other words, the percentage of language use is higher among the elderly and adults than among young people and children. However, when data of intergenerational transmission and use are compared, the situation appears difficult. Although languages seemed to be transmitted normally in 53% of cases and with some difficulties in 23%, it is possible that the young generations will reduce transmission considerably in future since only 36% of community languages are normally used by the younger generations. The analysis did not show significant differences between men and women in the intergenerational use of the language. The data reported here seem to show the same tendencies of reduction of linguistic diversity that were predicted by such authors as Krauss (1992), Comrie et al. (1996), Junyent (1999) or Hagège (2000). Danger of extinction and death of languages The informants reported that 42% of the analysed languages are in danger in most parts of the globe surveyed and 2% are in danger in certain parts. However, according to the same data, 45% of the analysed languages are not in danger.

20 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

The World Languages Review: Some Data

19

Table 2.1 Reasons that suggest a language is in danger Frequency Percentage (%) The life of the people is in danger (genocide, wars, action of the army, etc.)

11

2.8

Movement of the population (lack of ownership of the territory, deportation, emigration, migration, etc.)

36

9.5

Economic or cultural subordination (influence of other languages and cultures, colonisation, economic subordination, influence of religion, economic changes, cultural discrimination, assimilation, etc.)

152

39.7

Direct linguistic discrimination (absence of recognition of linguistic rights, absence of the language at school, repression of the school, etc.)

34

8.9

Negative attitudes (few speakers, negative attitudes of the speakers toward the language, etc.)

42

11

Other reasons

106

27.8

Although the informants, in general, are not as precise as specialists such as Kincade (1991), Krauss (1992), Wurm (1996, 2001) or Crystal (2000) when calling attention to the level of risk of extinction, they did refer to processes or tendencies that lead to extinction. In their responses to the questionnaires, the informants indicated the following as some of the causes of extinction (Table 2.1): • • • • • •

the life of the people is in danger; movements of the population; economic or cultural subordination; direct linguistic discrimination; negative attitudes; other reasons.

The most important reasons given were economic or cultural discrimination, together with negative attitudes. Nevertheless, it is important to take note of the other reasons given, including the safety of the people’s lives, movements of the population, direct linguistic discrimination, and others. If the main reasons for danger identified in the questionnaires are studied continent by continent, the following classification can be made: • America: economic or cultural subordination and negative attitudes; • Africa: economic or cultural subordination and direct linguistic discrimination;

21 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

20

• Europe: economic or cultural subordination; • Asia: economic or cultural subordination; • the Pacific: movements of the population and economic or cultural discrimination. Attitudes of speakers Linguistic attitudes refer to the whole spectrum of representations, sensations, feelings, perceptions, etc., arising from the relationship between people or communities, their languages and other languages. The questionnaire asked two explicit questions on linguistic attitudes – one on attitudes of community members towards their language, and the other on attitudes of the surrounding community members towards the language being analysed. According to this, a scale of attitudes towards the language of the informant’s linguistic community and the languages of surrounding communities has been established: • • • • •

positive attitudes; indifference or neutral attitudes; negative attitudes; ambiguous attitudes; other types of attitudes.

According to the data gathered, 61% of the speakers within a language community had positive attitudes towards their language, but only 33% of the speakers in the surrounding communities were positive. Further, 6% of the speakers in a language community had linguistic attitudes of indifference, while 17% of those in surrounding communities had indifferent or neutral attitudes. Answers in which respondents mentioned indifference or apathy and answers that simply reflected the fact that a community speaks its language, without any other specification, were included in the ‘neutral attitude’ category. Negative attitudes were expressed as follows: young people preferred other languages, were not very sympathetic to their own, and felt shame or fear towards their own language. In the sample, speakers from 10% of language communities and 20% from surrounding communities had negative attitudes. There are communities whose behaviour does not correspond with their reported attitudes. Other data from the questionnaire suggest that this happens very often, but the results presented here make reference only to the answers that reported it explicitly. According to the informants, 10% of the language communities and 4% of surrounding communities showed these ambiguous attitudes.

22 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

The World Languages Review: Some Data

21

Language policy and planning organisations Many linguistic communities whose data were analysed for this investigation have organisations engaged in language planning and policy. In 61% of the communities, there is some linguistic organisation, which could indicate that, in those communities, there is a certain level of awareness of the need for language planning. The relationship between the existence of these organisations and the perceptions of informants on the expansion and decline of their languages was also examined: it appears that the percentage of decline is lower in those languages that have such organisations than in those that do not. Some 40% of the communities that do not have any planning have reduced their number of speakers, but this figure drops to 30% amongst communities that do have language planning organisations. The existence of such organisations does not guarantee the continuity of those languages but the danger that the languages will decline rises considerably if the organisations do not exist. It is clearly important that such language planning organisations are promoted and sustained. The use of language in education The responses of the informants indicate that almost 60% of the languages are used somehow in primary education. Obviously, the level of use varies: • 1% are present in education but the use is not specified; • 7% are used orally, as an instrument for teaching given in another language; • 8% are taught as a specific subject (as a second language); • 26% are present only in pre-school or the early school years; • 13% are extensively present in primary education and part of secondary, though with some difficulties; • 12% are present throughout the school system but not across the whole population; and • 33% of the languages were not used in teaching at all. When language use in primary education is compared with language expansion or decline, it can be seen that most languages used in primary education are also languages whose situations are perceived to be better. It is important to note, however, that school cannot by itself guarantee language revitalisation. In fact, according to our informants, 28% of languages used in primary education have declined and 27% of languages with no educative use have improved their situation. Consequently, education is important for the future development of a language, but is not the only area in which measures to assist in the survival of the languages are needed.

23 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

22

With regard to the use of the languages in the press and media, the data indicate that it is in radio that the languages are most used (43.6%), followed by the press (26.5%), and then by television (22.9%), where the languages are least used.

Conclusion In general terms, the data gathered in the sample of the World Languages Review reported here show a global tendency towards language homogenisation as well as an incipient reaction against this by communities. This reaction takes two forms: some communities react by trying to recover and revitalise their language, while others react by resisting all kinds of pressures, as is the case for communities whose language is still being transmitted. The World Languages Review must consider the real threat of language death and the efforts made by communities to resist that threat. Information gleaned from the questionnaires demonstrates many examples of creativity in the fight for language maintenance. Spreading information about the situation can help many communities by making the world aware of the existence of both the communities themselves and their languages, and by disseminating models that can be used by communities to support their languages. As a general principle, it is necessary to understand language dynamics as a global and multi-factorial phenomenon. Actions intended to revitalise weak languages are not, therefore, restricted to certain spheres (for instance, education), even in isolated communities. Another general rule suggested by this study is that, beyond each community’s circumstances, linguistic attitudes are an important part of the process of preservation and revitalisation of language diversity. In many cases, before any specific actions to recover a language are taken, it is important to get all the necessary information to permit a change of attitudes towards the language within the linguistic community. It is also important to gather information about the surrounding communities – especially where those communities represent speakers of powerful languages – so as to promote a change of attitudes among them. The recovery and revitalisation of threatened languages is a process involving every aspect of community life, including relationships with other communities, both surrounding and non-surrounding ones. The use of language in high spheres (i.e. in spheres of communication involving certain formal use, especially those that can be influenced by the administration and by other public bodies) can have a strong influence on the change of attitudes since it increases the prestige of the language. The information that has been obtained, however, demonstrates that the

24 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

The World Languages Review: Some Data

23

process of language preservation must include coordinated action, both in high spheres and throughout the community. References Comrie, B., Matthews, S. and Polinsky, M. (1996) The Atlas of Languages. New York: Facts On File Inc. Crystal, D. (2000) Language Death. Cambridge: Cambridge University Press. Hagège, C. (2000) Halte à la mort des langues. Paris: Odile Jacob. Haugen, E. (1972) The Ecology of Language. California: Standford University Press. Junyent, C. (1999) La diversidad Lingüística. Barcelona: Octaedro. Kinkade, M.D. (1991) The decline of native languages in Canada. In R.H. Robins and E.M. Uhlenbeck (eds) Endangered Languages. Oxford: Berg. Krauss, M. (1992) The world’s languages in crisis. In Language 68 (1), 4–10. Wurm, S.A. (1996) Atlas of the World’s Languages in Danger of Disappearing. Paris/ Canberra: UNESCO Publishing/Pacific Linguistics. Wurm, S.A. (2001) Atlas of the World’s Languages in Danger of Disappearing (2nd edn). Paris/Canberra: UNESCO Publishing/Pacific Linguistics..

25 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Chapter 3

Naming Languages, Drawing Language Boundaries and Maintaining Languages with Special Reference to the Linguistic Situation in Papua New Guinea PETER MÜHLHÄUSLER

Introduction In considering the languages of the Pacific region, both the definite article and the term ‘languages’ are problematic. I am reminded of a former colleague, the Professor of Romance Languages at the University of Oxford, who, when asked how many there were, was at a loss for an answer. Now, if you were to set up a syllabus for teaching the history of the Romance languages, a definite answer would be of considerable benefit. Given 8 languages and 10 teaching weeks you could have an introductory lecture, one lecture for each of the languages and a concluding lecture. If there were 27 Romance languages, the situation would be slightly different. A definite answer as to the number of Romance languages would also be of use to policy-makers in the European Union – its policy is to support small endangered languages and the allocation of funding would look very different if Valencian, Asturian, Limousin, and other Romance forms of speech were recognised as Romance languages. If it is difficult to apply a definite article to Romance languages, it is even more difficult for the languages of Papua New Guinea (PNG), which is famous for its diversity of peoples and languages. The web-based language database Ethnologue (www.sil.org/ethnologue/countries/Papu.html) lists 826 languages, which is higher than the 700 which was the figure when I began to study languages in PNG in the early 1970s. Nekitel (1998) meanwhile lists 846. It is equally difficult to justify the label ‘language’. This chapter explores what the term ‘language’ means in Papua New Guinea and in Melanesian linguistics and arrives at a conclusion that may not please many linguists or sociolinguists or, indeed, language maintenance policy-makers.

24

26 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

25

Whilst most linguists have become habituated through their experience of Western Nation States to the notion that there are such things as languages, a closer look at the notion reveals many problems. There is no universal concept of ‘a language’ in the sense encountered in introductory text books: ‘a language is a set of elements and a system of rules for combining those elements to form patterned sentences that can be used to do specific jobs’ (Finegan et al., 1992: 8). Indeed, most people who have studied names and looked at languages in PNG have never bothered to identify the set of elements or a system of rules that might be used to define them. The Finegan definition is not far removed from Sapir’s (1933: 155) characterisation of a language as ‘an arbitrary system of symbolism’ and numerous similar ones. That the traditional inhabitants of PNG had a similar definition of language seems quite unlikely. How different groups conceive of a language, if at all, remains ill understood. Linguists have largely ignored indigenous metalinguistic views and studied and described languages like PNG ways of speaking ‘as if they were standard languages’ (Coulmas, 1994: 175). The danger of such an approach has been highlighted by Harris who compares the question ‘What is a language’ with the question ‘What is electricity’, and argues: It might be supposed that there is a higher level of inquiry at which the question ‘What is language?’ may be raised, where the essential nature of languages is scientifically defined, and where the ordinary languageuser’s mundane language-using is, if not quite irrelevant, at least left far behind. If there is a temptation to suppose something like this, it is a temptation best resisted; at least until it has been shown that there is anything interesting to be gained by yielding to it. Languages are not like electricity. The language-user already has the only concept of a language worth having. Whereas the electricity-user does not automatically have the only concept of electricity worth having, or indeed any concept of electricity at all. Belief in some superior concept of a language, to which only the language expert has access, is probably the result of confusing the concept of a language with knowledge of what there is to know about languages. (Harris, 1980: 3) Given this state of affairs, it would seem useful to follow Harris’ suggestions as to how to proceed in this matter: To say that a language-user already has the only concept of a language worth having is not to claim that all languages-users have the same concept of a language; nor that such a concept is simply a reflection of

27 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

26

man’s intuitive awareness of his own linguistic capacities. We may properly ask where a concept of a language comes from, and what purpose it serves. Unfortunately, so familiar is our concept of a language and so easily taken for granted that such questions may at first sight appear just as idle or perverse as ‘What is a language?’ But they turn out not to be. To start off, it might be asked how far one’s concept of a language is based on what one has been taught about languages. (Harris, 1980: 4) If we wish to do something to preserve the ways in which people communicate in a country like Papua New Guinea or indeed anywhere else, it would seem appropriate to find out more about indigenous metalinguistic views. It would also seem important to find out how the views of the traditional inhabitants of PNG differ from those put forward by expatriate missionaries, administrators and linguists, and a new generation of post-independence PNG linguists. Before doing this, a few words need to be said about language names.

Naming Languages The names of most PNG languages were imposed by outsiders who followed the conventions outlined by Laycock and Voorhoeve: A word must be said on nomenclature. Names of languages are cited in the form considered by the authors to be most appropriate, or – in some cases – in the form used in the literature cited; but it must not be assumed that these names have any more validity than as convenient labelling devices. It is rare for speakers of Papuan languages to have a name for themselves, in their own language, as a linguistic unit; rather, they will use a word which simply means ‘the people’, in an ethnocentric sense, and this term may frequently be much narrower in its extent than the linguistic group. This deficiency of nomenclature has been overcome by European observers in a number of ways: (1) by using a locality name, which may be the indigenous name of a village, island, mountain, valley or other geographical feature; or it may be an introduced topographical name (‘Western Highlands language’, ‘Big Sepik language’, etc.); (2) by using the name given to the people by another tribal group; (3) by using a group name: clan, totem, dialect designation; (4) by using an arbitrary name based on the language’s form for some common word. Those that have been frequently chosen are words for ‘man’ (Tuo, Moando, Nor, Pondo); ‘water’ (Ok); ‘language’

28 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

27

(Kam, Pay, Pila); ‘no’ (Olo, Elkei, Au); ‘what’ (Ngaing); and ‘my child’ (Natk, Barok). (Laycock & Voorhoeve, 1971: 509) Particularly explicit remarks on glossonyms are given by Laycock, who admits that his classification of languages is not based on ‘detailed lexicostatistical analysis or comparison’ and that it is ‘therefore impressionistic’ (Laycock, 1973: 2–3). He nevertheless feels that ‘it is not likely to be far from the truth’. He argues further that ‘the problem of language naming remains a very vexed one.’ In the case of one of the best known languages of Papua New Guinea, Tolai, the indigenous name was replaced by a number of designators chosen by missionaries, administrators and linguists: The language of the Tolai people, which is nowadays simply called ‘Tolai’ in Austronesian linguistics, has been given several names. The Tolai people themselves call it A Tinata Tuna, literally the indigenous language or Kuanua, which is originally a word of the language of the Duke of York Islands meaning ‘over there’ and which was first used by the Methodist missionaries who started their mission in these islands (cf. Brown, 1908; Mosel, 1982; Threlfall, 1975). The Catholic missionaries introduced the names Tuna, literally ‘indigenous’ (Meyer, 1961), Gunantuna (Zwinge, 1953); other names used by Europeans are Blanche Bay Dialect, New Britain Dialect (Rickard, 1889), Nordgazellen Sprache (Bley, 1912), Neu-Pommerische Sprache (Constantini) and Raluana (Lanyon-Orgill, 1960). (Mosel, 1984: 4) Examples of expatriates giving names abound, and only a few more can be mentioned. Manam is an Austronesian language spoken on the Pacific north coast of Papua New Guinea. There is no native name for the language (unnamed vernaculars seem to be the rule in New Guinea). Following scholarly tradition in this area, I have retained the geographical name. (Gregerson, 1976: 95) Occasionally, names stick but, more often, there is considerable fluctuation in name use. Reesink justifies his choice of the label ‘Usan’ as follows: How does a language get its name? Quite a few languages in Papua New Guinea are known by their word for ‘what’, ‘man’, ‘hand’, or simply by a variety of names, given by various explorers, missionaries, government officials, or others. When Z’Graggen did his comparative work on the languages in the Madang province during the late sixties (cf. Z’Graggen, 1971), he found many for which no name was available. In those cases, he named the language after the largest village. One such language he named Wanuma, a member of the Numugenan

29 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:27

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

28

family. Language families Z’Graggen named after a major river in the area of a family of languages. The Numugen is known by the Usan as Ibirin. Wanuma is derived from the Usan name Wonum, which designates one of the eleven territorial groups. The name Wanuma now stands for a government patrol post, Lutheran church station and small airstrip, originally built by the Lutheran mission. The administrative centre is situated on one of the lower Southern ridges of the Adelbert Range in the Madang province. There are three territorial groups known that use the name Wonum: the Wonum yag (yag = water) has nearly died out. Only a few people, recognised as their descendants, are living among the Wonum qurum (qurum = flat/low land), who live in five hamlets near Wanuma. The Wonum saut (sau = piece) live on a higher ridge to the North, most of them in a village called Qamamite. A number of other territorial groups form one linguistically homogenous group together with the Wonum. Although the name Wanuma has been used by Z’Graggen in various publications (1971, 1975), the name Usan seems more proper to refer to the language under consideration. It is the name by which a number of social groups identify themselves as a unity in distinction from surrounding groups, such as Qumin, Itimai, Yemei. (Reesink, 1987: 4–5) On the other hand, Z’Graggen’s name for Mugil, established in the 1970s, was subsequently changed by Summer Institute of Linguistics (SIL) missionaries. Hepner writes: Bargam is a non-Austronesian language of the Trans-New Guinea phylum spoken by approximately 3500 people located about 65 kilometres north of Madang town along the north coast of Madang Province. Z’Graggen (1975) refers to it as Mugil, and classifies it as a stocklevel isolate within the Madang-Adelbert Range subphylum. The data and conclusions in this chapter are based on over six years of active involvement in linguistic, literacy and translation work among the Bargam people. (Hepner, 1995: 1) Yet, another name suggested by Z’Graggen continues to be used by linguists working in the area: The Mari language has been called by various names – the early Lutheran missionaries and some explorers who first contacted the people called them Garamari, which is a version of the name Garam Mari given to them by their Adzera neighbours. In Hooley (1970) and Hooley and McElhanon (1970) they were called Hop, presumably because the Mari for ‘speech’ is hop. Hooley (1971; 1976) and

30 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

29

Z’Graggen (1975) subsequently referred to the language as Mari. (Holzknecht, 1989: 22) A look at Ethnologue reveals that many languages have been named several times by different people and that there is at least some confusion among those who count the languages of PNG as to whether two names should stand for one, two or more languages. For the language name Pinai the Ethnologue provides the following information: PINAI (PINAYE, PINAI-HAGAHAI, HAGAHAI, WAPI, ARAMO, MIAMIA) [PNN] 600 (1993 M. Mellinger, SIL). Mainly in Enga Province, north-west corner, Aramo River area, East Sepik Province, west of Haruai. Some in Madang Province, on either side of the Yuat River. Sepik-Ramu, Ramu, Yuat-Waibuk, Piawi. Dialects: MENGAU, BULUBAU (NANGENUWETAN), PINAI 1, PINAI 2, HAGAHAI (ARAMO, ARAMAUE), MENGAMANAO, LUYA, GINAM. Speakers in Enga Province use ‘Pinai’ or ‘Miamia’ to refer to the entire language group. Those in Madang Province use ‘Hagahai’ to refer to themselves, suggested to them by a medical anthropologist. ‘Wapi’ is sometimes used by the Enga, ‘Aramo’ by Haruai speakers. Dialects have 65% to 94% lexical similarity with each other. Hagahai has 63% to 70% lexical similarity with the three. Haruai has 29% to 37% lexical similarity with the four; Wapi has 8%. Hagahai has 19% lexical similarity with Kobon, some speak Harai, some Tok Pisin (especially boys). Medical workers report widespread health problems. Mountain slope. Hunter-gatherers. Altitude: 100 to 1300 meters. Work in progress. (http:// ethnologue.com/) Nekitel (1998) in his list of PNG languages has an entry Pinai (1500 speakers), Wapi (1000 speakers) and Hagahai (300 speakers). Not only do we have three languages instead of one, but there is a considerable discrepancy in their speaker numbers. This is not a particularly unusual example. Seiler recounts a similar problem with languages referred to as Punda, Imonda and Sowanda and his solution differs from that found in the Ethnologue: The languages assigned by Laycock to the Waris family are as follows: Manem, Senggi, Waris, Waina-Sawonda, Daonda, Simog and Amanab ... Waina-Sowanda is spoken to the south of Imonda and has villages on either side of the border. As far as the Papua New Guinea side is concerned, this language must be split in two. One is spoken at the villages of Umeda and Punda and I have (arbitrarily) opted for the name Punda in this grammar. I have chosen Sowanda as the name for the rest of the original Waina-Sowanda language area. (Seiler, 1985: 3)

31 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

30

Comparing documents that are temporally more distant increases these problems. Ray (1929), for instance, lists 15 language names for the Melanesian languages of the Central division of Papua. Only nine can be equated with a name or variant name in the Ethnologue and the match with Map 9 of Wurm and Hattori (1982, map compiled by D.T. Dutton) is equally difficult. The disappearance of six names over seventy years does not mean the decline of six languages but, in all probability, changes in naming practices, boundary drawing, etc. Laycock discusses the problems that investigators experience when consulting earlier records, in this instance going back to the days of German New Guinea: As many of the items in the short lists of Neuhauss and Friederici are identical in Sera and Sissano, it is not possible to be sure just where they were elicited; both, however, give the Sera word for moon (Neuhauss: wul; Friederici: bul) instead of the now current Sissano word sanar ... This suggests either a dialect change in the Sissano language in the last sixty years, or else non-homogenous lists, with some items elicited in Sarai, or from Sera speakers. (Laycock, 1976: 76) The problems surrounding language names are compounded when it comes to dialect names, as the following example illustrates. Cochran (1968: 134) confidently states that ‘the Yoliapi dialect is a Western dialect of Hewa’. No mention of such a dialect is made in the Ethnologue in the entry on Hewa and its four dialects.

Drawing Language Boundaries Drawing boundaries is closely associated with power – it is the boundaries drawn by the powerful that will prevail, be they boundaries between political units, boundaries between legal and illegal, acceptable and unacceptable, or languages and dialects. This is not to say that there are no natural boundaries – there are boundaries between species in biology, boundaries between chemical elements, and quite a few more. Cultural boundaries tend to be contingent on political, social and historical forces. They are manmade in both the literal and metaphorical sense of the expression. Anyone who has practised dialectology will be keenly aware of the great difficulties experienced when trying to characterise dialects as objects bounded by a bundle of (carefully selected and unselected) isoglosses (i.e. lines on a map, within which a particular linguistic feature is employed). The problem of drawing boundaries between languages and dialects in Melanesia is partly the result of the prevalence of extended dialect chains (see Wurm & Laycock, 1961) or language chains (Tryon, 1979) which make non-arbitrary cuts very difficult. In order to apply criteria such as shared

32 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

31

cognates or intelligibility one first needs to establish arbitrary reference points on the continuum. We note that, depending on the reference points chosen, what counts as a language and what counts as a dialect can be affected with an ultimate effect on the total number of languages recognised. Wurm and Laycock acknowledge this: Application of the suggested new classification method to New Guinea languages could bring about a considerable reduction in the number of what from the point of view of the Swadesh method would be classified as distinct languages. (Wurm & Laycock, 1961: 129) Ironically, the number of languages distinguished by linguists has grown significantly since, though not all additions have found their way into the Ethnologue. Voorhoeve (1980: 14), for instance, corrects Drabbe’s earlier claim that there was only one Asmat language by postulating four: Otenep, Pirimapun, Aorkel and Tareo. None of these names is found in the Ethnologue. The already-mentioned Pinai language illustrates this observation. Regarding the so-called Piawi family of languages, Comrie writes: The family consists (probably) of three languages: Haruai, Hagahai and Pinai ... Information from anthropologists working in the region ... suggests that Hagahai and Pinai are to a large extent mutually intelligible, with at least a possibility that they constitute dialect varieties of a single language. Other languages that have occasionally been posited in literature, such as Wapi, are almost certainly varieties of Pinai. (Comrie, 1992: 11) That different analysts come up with quite different boundaries is also illustrated by Franklin (1968), who is aware that the same languages or dialects may be classified quite differently, depending on the technique used. He proposes to overcome this by developing a more complex test combining lexicostatistics with intelligibility criteria, without addressing the question of how many Kewa dialects and languages there might be. The major problem with drawing boundaries is that the units identified by outsiders are typically not those recognised by indigenous users. Laycock in his fieldwork consulted indigenous opinion but in many instances saw fit to override it, as indeed is common practice. Holzknecht, on the other hand, appears to have ignored indigenous opinion in her studies of Adzera: It is important to mention here perceptions of differences between the dialects. The folk perceptions of dialectal difference do not always coincide with those of the linguist. The linguistic differences perceived by the speakers were frequently those of intonation, stress, or speed of

33 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

32

delivery. They were also sometimes based on a single item of vocabulary, for example the word for ‘no’. The boundaries of the ‘in-group’, that is those who speak ‘the same way’, were frequently drawn for me by speakers according to the boundaries of ancient political alliances rather than strictly linguistic criteria. Clearly minor linguistic differences are exaggerated in order to express ‘in-group’ affiliation, and to exclude the ‘out-group’ population. The ‘in-group’ is as inclusive or exclusive as an individual or group wants it to be in any given context. (Holzknecht, 1989: 26) Not only is there a discrepancy between indigenous and expatriate views on language names and boundaries, the solutions proposed by outsiders often do not take into account that people communicate (an activity) rather than being owners of languages (objects). Cooper, in his analysis of Coastal Suau, has highlighted the discrepancy between units identified by linguists and patterns of communication: COASTAL SUAU shall be defined to include those varieties of Suau generally spoken within the Suau Administrative Area of the Milne Bay District. Coastal Suau, thus defined, is a somewhat arbitrary label for a major part of the Suau speech continuum. Its western boundary is fairly distinct, for the neighbouring languages, Mailu and Oima, are nonAustronesian; its southern boundary is the sea itself, but across the mountains to the north and into the islands to the east there appear no sudden breaks. Kwato and Fife Bay circuits of the United Church have made Suau a lingua franca in some areas along Milne Bay and inland. Bearing these indeterminate factors in mind, it is probable that between eight and ten thousand persons use some form of Suau as a first or second language ... (Cooper, 1975:230) Cooper also admits: To designate these other forms of the Suau speech continuum whose boundaries are still indeterminate and provide a label for the entire domain of Suau communities, the terms SUAUIC and SOUTHERN MASSIM have been applied; e.g. the Logea and Sariba communities are clearly Suauaic though published data includes little more than short word lists. Wari and TubeTube speech also seem to be Suauic, but again data is limited. (Cooper, 1975: 230). The characterisation of Suau in the Ethnologue fails to do justice to the complexity of the situation despite distinguishing 9 dialects (with 12 names but no mention of Wari or TubeTube). Members of the same speech commu-

34 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

33

nity may hold a number of views as to similarities, boundaries and groupings. Thus, Davis comments: I have no consensus from informants on general dialect groupings or boundaries beyond the basic Miohana/Moitana (Waima/Roro) differences. While agreeing that the Paitana villages form a unit, informants differ over inclusion of Tsiria, Delena, Pinupaka, and Poukama within the Paitana dialect. Similarly, some informants link Kisiu and Nabuapaka, but claim that Bereina, Kivori and Waima are all different. (Davis, 1974: 6) When linguists speak of dialects, they imply (unlike politicians who often apply this label to any unwritten form of speech) that it is a dialect of a language. In Papua New Guinea it is often not clear what languages particular forms of speech are dialects of: Laycock, in his survey of Sepik languages, lists one language, Mayo, with a possible dialect, Yau, and a second language, Pasi, that is ‘very closely related to Mayo, and may in fact form part of that language’ (Laycock & Voorhoeve, 1975: 72) which differs ‘from other Mayo dialects except that of Yessan’. The Ethnologue lists Yessan-Mayo as one language, Yau as another and Pasi as either a separate language or a dialect of Yessan-Mayo. To a Pasi speaker, it will presumably make a difference whether their form of speech is a dialect of Yessan-Mayo or a dialect of Yau or, indeed, a language.

Some Preliminary Conclusions This chapter has, thus far, tried to demonstrate the following: (1) There remain vast problems in establishing and counting languages, dialects and subdialects even if linguists could agree on the criteria they apply. (2) Whereas many names exist, it is clear that they do not refer to comparable entities and it is far from clear what they actually refer to. Sutton has arrived at a very similar conclusion for Australia: Perhaps the greatest problem in this area has been not so much arriving at an accurate and standardised form of names as working out just what such names refer to and the question of what language ‘names’ or labels actually refer to, and how they function in Aboriginal societies, has in general been rather neglected and some languages, originally perhaps many, have no names. Many languages have several alternative names, and usually a different one used by each or several of the different linguistic communities in contact with them. Sometimes two grossly different dialects may be subsumed under a single term, while another two which differ by little more

35 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

34

than a single common word may be distinguished by a reference to that word. A complicating factor, overriding all the above, is that perhaps all defining terms (such as language names, territorial group name, etc.) may have different meanings depending on their immediate context and function. (Sutton, 1979: 89–90) This would seem to confirm that the names and boundaries of languages reflect expatriate practices, not local knowledge.

On Indigenous Views of Language Differences What is known about the traditional indigenous views on language in PNG is disappointingly little. Most outside observers did not bother to enquire or (in the case of professional linguists) held the view that a science of linguistics should overcome animistic views and ignorance and that secondary and tertiary responses to language should not be confused with knowledge of language (views strongly stated in Bloomfield, 1944). Studies on metalinguistic knowledge among Papua New Guineans are very few and much information has been lost. Those New Guineans who have received a Western education in linguistics tend to apply the categories of Western metalanguage. The available literature affords a few glimpses on indigenous views, however. Cooper writes on the Suau category of ‘alina’: The term alina is generally used to designate language, dialect, or speech in general. Any degree of difference from a small bundle of lexical and phonological distinctions to mutual unintelligibility may be designated by this term followed by a name, most often a place name. (Cooper, 1975: 231) Another example is given by Davis (in the quote cited earlier): I have no consensus for informants on general dialect groupings or boundaries beyond the basic Moihana/Moitana (Waima/Roro) differences. While agreeing that the Paitana villages form a unit, informants differ over inclusion of Tsiria, Delena, Pinupaka, and Poukama within the Paitana dialect. Similarly, some informants link Hisiu and Nabuapaka, but claim that Bereina, Kivori and Waima are all different. (Davis, 1974: 6) This would seem to confirm a more general principle that humans are inclined to see differences and similarities when they want to see them (Davis, 1974: 6). It is not different from the linguistic situation in parts of Europe where Valencians insist that they speak a language separate from Catalan and where Bosnians (and, perhaps, Montenegrans) insist that their language differs from Serbian and Croatian. By contrast, Americans and

36 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

35

Australians continue to believe that they speak English and Bavarians that they speak a dialect of German (which, incidentally, is Romance in its grammatical core structure and not mutually intelligible with standard German). A definite article in the phrase ‘the languages of the Pacific’ is quite inappropriate because (1) their number remains indefinite and (2) ‘language’ refers to various non-comparable concepts. One could argue that all this reflects an interesting academic puzzle or, less charitably, an attempt to obtain a determinate answer to a question that can have no such answer in principle. As long as the ‘languages’ of Papua New Guinea were regarded as an obstacle to progress, it did not matter. Now that linguists have begun to get concerned about the disappearance of the world’s linguistic diversity, now that development aid is given to set up literacy and maintenance programmes and local languages are redefined as assets, it matters a great deal. There is a need to have a reliable basis on which to set up a language maintenance programme. There is no evidence that this basis exists or that conventional linguistics can provide it.

Maintaining Languages The main issue addressed in this volume is the state of the languages in Australia and the Pacific. All the contributors to this volume share the view that preserving linguistic and cultural diversity is a good thing, and some would agree that the survival of the world’s linguistic diversity is closely related to the survival of biological diversity. They all share a common aim but where I may differ is on the means to achieve it. My argument is: (1) Linguistics was set up to address a number of issues – optimising grammatical description, exploring history and development of grammar, looking for structural communalities across languages, and so forth. Such aims have little or nothing to do with language maintenance and preservation. (2) Linguists have resorted to abstractions and idealisation that have forced them to ignore a large number of parameters relevant to the task of language maintenance. Linguists have assumed that the metalanguage of linguistics is relatively unproblematic and neutral. This is ethnocentric and simplistic: our terminology of ‘language’, ‘dialect’, native speaker’, ‘grammar’, phoneme’, ’sound segment’, or whatever is highly culture-specific and suited best to situations involving established official languages, not the ways of speaking found in traditional small communities. (3) Linguists have confused the constructs of their discourses, terminologies and professional practices with realities out there. By imposing boundaries, names and standards, they have not been neutral

37 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

36

recorders but language makers. Their observations have had a major effect on what they observed. (4) Linguists have assumed that humans are essentially monolingual, speakers of a single mother tongue or at least that single languages can exist in isolation from other languages. They have ignored the complex fabric of vernaculars, local pidgins, regional lingua francas, dual-lingualism, bi- and multi-lingualism which may be essential in sustaining the independence of smaller vernaculars. Giving languages names, often names that derive from a placename, again can create an unfortunate package of language and territory (of the European nation state type) and can wrongly suggest that the absence of a name is indicative of the absence of a language or way of speaking. In the domain of conservation of natural kinds (animals and plants) it is often the unnamed species that are most endangered, and preserving only the named ones may do little to preserve the ecological conditions on which the named ones depend. The practices of boundary drawing and naming have affected the relative status of ways of speaking. Those that have recognised boundaries and names have more power and prestige and can out-compete other ways of speaking. The first option, carrying on as usual and setting up an enterprise of describing idealised grammars, entails many dangers, among them the chance that this activity could further weaken indigenous ways of speaking. Instead, the question, ‘How can we preserve languages?’ should be re-framed as, ‘How can we preserve the ecological conditions which make it possible for a diversity of ways of speaking to continue to exist?’ Past conditions that have sustained diversity include: (1) The already-mentioned complex fabric of small local vernaculars and languages of inter-communication. It is not clear whether introduced local pidgins, English and missionary lingua francas offer an adequate substitute. (2) A highly mobile, quasi-nomadic lifestyle in many parts of Australia and Melanesia. (3) Regulated intermarriages between groups speaking in different ways. (4) Complex regulations regarding the use of language in particular places. (5) Rapid language change (particularly lexical) because of taboo. Regeneration was possible because of dialectal diversity and regulated borrowing. (6) The area of language currency tended to coincide with ecologically defined natural habitats. Most of these sustaining factors have begun to be eroded and it is impor-

38 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

37

tant to seek to reverse this trend and/or look for appropriate substitutes. It is quite unclear how often suggested substitutes such as literacy, Bible translations, modern education and such like could function as support systems. It is not easy to see how this could be so. No objection is raised here to the principle that languages and cultures have to change and adapt to survive. However, the rate of change (from stone age to space age within a generation) that the languages of Papua New Guinea and Australia have had to face may have been far too rapid to make adaptation possible. Toffler’s discussion of the limits to adaptation and the confusions that arise from overchoice in his Future Shock (1974) make relevant background reading. The belief that speakers in Melanesia today are in a position to make rational language choices may have little substance. All this leaves us with the question what can or should linguists, or indeed anyone, do to maintain linguistic diversity? Reframing the question is only the beginning. References Bley, B. (1912) Praktisches Handbuch zur Erlernung der Nordgazellen-Sprache. Münster: Westfälische Vereinsdruckerei. Bloomfield, L. (1944) Secondary and tertiary responses to languages. Language 20, 45–55. Brown, G. (1908) George Brown: Pioneer-missionary and Explorer: An Autobiography. London: Hodder & Stoughton. Cochran, A. (1968) Notes on Yoliapi. Kivung, Journal of the Linguistic Society of Papua New Guinea 1 (3), 134. Comrie, B. (1992) Recognition of the Piawi language family. In R. Dutton, D. Tryon and M. Ross (eds) The Language Game: Papers in Memory of Donald C. Laycock (pp. 111–113). Pacific Linguistics Series C, No. 110. Canberra: Australian National University. Cooper, R.E. (1975) Coastal Suau: A preliminary study of internal relationships. In T.E. Dutton (ed.) Studies in Languages of Central and South-East Papua (pp. 227–78). Pacific Linguistics Series C, number 29. Canberra: Australian National University. Coulmas, F. (1994) Protestant ethics and profane language: Economic aspects of language standardization. In G. Lüdi (ed.) Sprachstandardisierung (pp. 161–78). Freiburg: Universitätsverlag. Davis, M.M. (1974) The dialects of the Roro language of Papua: A preliminary survey. Kivung: Journal of the Linguistic Society of Papua New Guinea 7 (1), 3–15. Dutton, T.E. (1975) A Koita grammar sketch and vocabulary. In T.E. Dutton (ed.) Studies in Languages of Central and South-East Papua (pp. 227–78). Pacific Linguistics Series C, No. 29. Canberra: Australian National University. Ethnologue. Online database at http://ethnologue.com/. Accessed 1.12.04 Finegan, E., Besmer, N., Blair, D. and Collins, P. (1992) Language: Its Structure and Use. London: Harcourt, Brace, Jovanovich. Franklin, K.J. (1968) The Dialects of Kewa. Pacific Linguistics Series B, No. 10. Canberra: Australian National University. Gregersen, E.A. (1976) A note on the Manam language of Papua New Guinea. Anthropological Linguistics 18 (3), 95–111. Harris, R. (1980) The Language Makers. London: Duckworth.

39 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:28

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

38

Hepner, M. (1995) Tense, aspect and modality in Bargam. Language and Linguistics in Melanesia 26 (1), 31. Holzknecht, S. (1989) The Markham Languages of Papua New Guinea. Pacific Linguistics Series C, No. 115. Canberra: Australian National University. Hooley, B.A. (1970) Mapos Buang: Territory of New Guinea. PhD thesis, University of Pennsylvania. Hooley, B.A. (1971) Austronesian language of the Morobe District, Papua New Guinea. Oceanic Linguistics 10 (2), 79–151. Hooley, B.A. (1979) Austronesian languages: Morobe Province. In S.A. Wurm (ed.) Australian Linguistic Studies. Canberra: Australian National University. Hooley, B.A and McElhanon, K.A. (1970) Languages of the Morobe District, New Guinea. In S.A. Wurm and D.C. Laycock (eds) Pacific Linguistic Studies in Honour of Arthur Capell (pp. 1065–94). Pacific Linguistics Series C, No. 13. Canberra: Australian National University. Lanyon-Orgill, P. (1960) A Dictionary of the Raluana Language (New Britain, SW Pacific). Victoria, BC: Published by the author. Laycock, D.C. (1979) History of research in Austronesian languages. Sepik Provinces. In S.A. Wurm (ed.) New Guinea and Neighboring Areas: A Sociolinguistic Laboratory (pp. 73–93). The Hague: Mouton. Laycock, D.C. and Voorhoeve, C.L. (1971) History of research in Papuan languages. In T.A. Sebeok (ed.) Current Trends in Linguistics, 8: Linguistics in Oceania (pp. 509– 40). The Hague: Mouton. Laycock, D.C. and Voorhoeve, C.L. (1973) Sepik Languages Checklist and Preliminary Classification. In Pacific Linguistics Series B, No. 25. Canberra: Australian National University. Laycock, D.C. and Voorhoeve, C.L. (1975) Languages of the Sepik Region, Papua New Guinea. Pacific Linguistics Series D, No. 26. Canberra: Australian National University. Meyer, O. (1961) Wörterbuch der Tuna-Sprache. MBA 34. St. Augustin bei Bonn: Anthropos Bibliothek. (Compiled 1921.) Mosel, U. (1982) The influence of the church missions on the development of Tolai. In R. Carle, M. Heinschke, P. Pink, C. Rost and K. Stadtlander (eds) Gava: Studies in Austronesian Languages and Culture (pp. 155–172). Berlin: Reimer. Mosel, U. (1984) Tolai Syntax and its Historical Development. Pacific Linguistics Series B, No. 92. Canberra: Australian National University. Nekitel, O.I.M. (1998) Voices of Yesterday, Today and Tomorrow. Port Moresby: UISS Publishers. Ray, S.H. (1929) The languages of the Central Division of Papua. Journal of the Royal Anthropological Institute 59, 65–96. Reesink, G.P. (1987) Structures and their Functions in Usan. Amsterdam: John Benjamins. Rickard, R.H. (1889) A Dictionary of the New Britain Dialect and English. Manuscript held in the Library of the Institut für Indonesische und Südseesprachen, Hamburg. Sapir, E. (1933) Language. Encyclopaedia of the Social Sciences (pp. 155–69). New York: Macmillan. Seiler, W. (1985) Imonda, A Papuan Language. Pacific Linguistics Series B, No. 93. Canberra: Australian National University. Sutton, P. (1979) Australian language names. In S.A. Wurm (ed.) Australian Linguistic Studies (pp. 89–90). Canberra: Australian National University. Toffler, A. (1974) Future Shock. London: Pan Books.

40 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Naming and Maintaining Languages in Papua New Guinea

39

Voorhoeve, C.L. (1980) The Asmat languages of Irian Jaya. Pacific Linguistics Series B, No. 64. Canberra: Australian National University. Wurm, S.A. (1979) New Guinea and Neighbouring Areas: A Sociolinguistic Laboratory. The Hague: Mouton. Wurm, S.A. and Laycock, D.C. (1961) The question of languages and dialect in New Guinea. Oceanic Linguistics 37, 128–143. Z’Graggen, J.A. (1971) Classificatory and Typological Studies in Languages of the Madang District. Pacific Linguistics Series C, No. 19. Canberra: Australian National University. Z’Graggen, J.A. (1975) The Languages of the Madang District, Papua New Guinea. Pacific Linguistics Series B, No. 41. Canberra: Australian National University.

41 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Chapter 4

Obstacles to Creating an Inventory of Languages in Indonesia: A Dialectology Perspective MULTAMIA R.M.T. LAUDER

Introduction Scope and purpose Indonesia is a large and diverse country and it is estimated that 10% of all the world’s languages are found there. After Papua New Guinea, Indonesia has the largest number of languages of any nation state, a total of 726 by one estimate (SIL, 2001). While the eight largest languages (Javanese, Sundanese, Madurese, Balinese, Buginese, Banjarese, Minangkabau and Batak) have received attention, there are still many very basic things that we do not know about the others. This chapter discusses some of the things that stand in the way of the inventory, description and preservation of Indonesia’s linguistic diversity. It draws on my own first-hand experience in performing languagemapping research to explore language diversity in various regions of Indonesia in recent years. Particular attention is given to work done on languages in the more remote areas of the country. Mapping the regional languages of Indonesia The project called Research on Cognates and Mapping of Regional Languages in Indonesia (Penelitian Kekerabatan dan Pemetaan Bahasa di Indonesia) has the goal of documenting diversity and comparing the regional languages throughout Indonesia using a single, consistent methodology. The project is being conducted throughout the whole of Indonesia over a 15-year period (1992–2006) though the project’s completion date has been put back with no data collected since 1998, a consequence of the country’s political and economic crisis. A number of publications should also appear in this period that will report on data, selected findings and any hypotheses generated. Among those that have already appeared are a preliminary mapping and linguistic analysis of languages in the provinces of East Nusa Tenggara, North

40

42 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Obstacles to Creating an Inventory of Languages in Indonesia

41

Sulawesi, Central Sulawesi, South Sulawesi, East Kalimantan and South Kalimantan. Publications of linguistic data that have already appeared cover the prefectures of Ketapang, Kapuas Hilir, Sambas, Berau, Samarinda, Balikpapan, Banjarmasin, Pontianak, Bulungan, Belu, Ngada, West Sumba, East Sumba, Pasir, Kutai, Sangau and Sintang.

Research Procedures Research team The research is being conducted under the auspices of The Indonesian National Language Centre (Pusat Bahasa) and is run by a small core team of researchers, mostly from the Faculty of the Humanities at the University of Indonesia (Fakultas Ilmu Pengetahuan Budaya Universitas Indonesia). The Indonesian National Language Centre has had responsibility for handling routine administrative concerns. The core team will analyse the data using techniques from dialectology and historical linguistics and generate a comprehensive series of language maps for the entirety of Indonesia. The team formulated the initial research concept and designed, produced, trialled and refined the main research instrument, a questionnaire. It set out principles for deciding which villages were to be investigated and also produced materials and procedures for selecting, screening and training local field researchers, and checking the reliability of collected data. The core team also trained and supervised the Indonesian National Language Centre staff whose job it has been to key in to the computer the linguistic data from the field. The data was coded so that the representation of phonetic information could be entered using the most basic of computer systems. The core team has also worked closely with computer specialists who have developed an efficient program for processing the field data for dialectometry, lexicostatistic calculations and computerised language mapping. This language-mapping project has tied its data to the same system of classification of geographical places that is used by the National Centre for Statistics (Biro Pusat Statistik). Each village in Indonesia, in this system, has a unique, multi-digit number which identifies it and its context and links it to its data in a computer database. The language mapping data for any village can thus be tied to all of the demographic data obtained by the National Centre for Statistics for broader understanding and better interpretation of the language situation for that place. Field researchers The core team have to identify people who would be capable of becoming field researchers. A screening process is set up to determine whether applicants meet the selection criteria. Applicants have to be high

43 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

42

school language teachers with a minimum education of a first degree in language studies or linguistics. Nobody older than 40 is accepted and no distinction is made between men and women. However, pregnant women or nursing mothers are not considered, as it is thought their ability to focus on the research might be compromised by their maternal duties. Finally, all applicants have to state their willingness to commit themselves to working in the field for a period of one month. Candidates from each province who get through the initial screening process are brought together as a group for training. Training is conducted intensively over a two-week period in a residential course by the core research team. The training consists of lectures, trial fieldwork using the questionnaire, and candidate assessment. Trainees who do not meet the standards set by the core team are not chosen to perform fieldwork. The number of field researchers required for a particular province is determined by the geographical size of the province and the number of regional languages estimated to be found there. Informants The field researchers are the ones who have to go to the designated villages and identify suitable informants from among the villagers there. Informants must be natives of the village in question, and born and raised there. In addition, both their parents and all their grandparents must have been born in the village and have lived there continuously. Informants can be men or women and should be approximately 40 years old. No distinction is made between single or married people, but married informants are screened and only people whose spouses are also natives of the same village are used. Mobility history is a factor. Only people who have spent most of their lives in the village and who have never or only rarely travelled to other places are accepted. No one who has studied beyond the academic equivalent of Primary six is accepted as an informant. This is to minimise the potential influence of Indonesian, which would have been used as a language of education. Those who have never attended school formally and are illiterate are accepted. Farming is the preferred occupation for becoming an informant. Other criteria also apply. People who have any noticeable speech defect are not accepted; nor are people with hearing difficulties. People who appear shy or introverted are not selected. Finally, only those people who express their willingness to be interviewed over a period of several days are accepted as informants. Data gathering The field researchers use a standard questionnaire, which forms the basis for their interviews to elicit the local lexis for the listed words and phrases. These are recorded using a standard, phonetic notation that was

44 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Obstacles to Creating an Inventory of Languages in Indonesia

43

designed to facilitate later entry of the data in a standard ASCII character format into even the most rudimentary of computers. As well as eliciting the lexis and basic propositional content in sentences, social, geographical and other key data concerning the village and the informant(s) are recorded. In addition to the written data, the field researcher makes an audio recording and a transliteration of a folk story concerning the origin of people in the village in question. This recording allows the core team in Jakarta to check for any inconsistencies with the field-researcher’s written notation of speech sounds. If the core team discovers discrepancies between the recording and the written records of the field researcher, the data for that village is discarded and the research there is repeated. The recording also provides data that can be used for the detection of syntactic constructions and other linguistic features beyond those that would be elicited by the list of words, phrases and sentences in the questionnaire. Apart from the linguistic uses of the recording, the compilation of folk tales about the villagers’ ancestors for each ethnic and linguistic grouping is likely to become a very valuable source for the study of oral literature. It can also allow cross-comparisons of the generic structure and other classificatory features of the folk story throughout Indonesia. Languages in remote areas The inventory and description of language diversity would be incomplete without including languages found in remote areas and mostly spoken by isolated tribes. These are interesting from a number of viewpoints and certainly provide a rich source of linguistic data. So, despite the logistic difficulties, the core team planned to include in the project the languages of a number of isolated tribes. The first problem was how to locate the isolated tribes throughout Indonesia.

Isolated Tribes in Indonesia One of the major problems that has to be faced when doing research among the isolated tribes is a lack of hard information of even the most basic kind. This section describes some official and published sources of information that were consulted during this project. Problems related to the completeness, reliability and accessibility of these sources are also discussed. Among the baseline facts that the core team was looking for initially was information on how many isolated tribes there were, what they were called, and where they were located. The team turned first to official sources for this information, hypothesising that government officials working in the areas in question or government bodies charged with improving the

45 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

44

welfare of these people would be able to provide ready facts and figures. It turned out to be more complicated than that. Local government officials When we approached local government officials and asked them where the isolated tribes were, the standard answer from them was invariably that ‘there are no longer any isolated tribes in our province’. This may have been because the basic defining criteria for isolated tribes are not clear, but we can perhaps also explain such a uniform answer as being a kind of defence mechanism. Perhaps for government officials to acknowledge the existence of isolated tribes was to admit publicly the failure of major government initiatives to reduce poverty or backwardness. Erring on the side of caution would be less likely to affect the incumbent’s promotion prospects. The only way for us to obtain the necessary information was to take a step-bystep and personal approach, reassuring the officials that the results of the research would be used only for the purpose of advancing academic or scientific knowledge, and assuring them that their names would not appear anywhere in the research data. In brief, information obtained by visiting provinces is likely to yield misleading information if taken at face value. The data obtained can be useful, but will be in oral and not written form. It is probably best to consider government officials as a source of last resort. Underdeveloped villages Even though underdeveloped village are not necessarily the places where isolated tribes are to be found, they are a potential starting point for locating them. Development efforts targeted at people who are at risk of being left behind come under the authorisation of a Presidential Decree on Underdeveloped Villages (Instruksi Presiden tentang Desa Tertinggal). The first government census to inventory underdeveloped villages in each province provided some welcome news. The census claimed that every province had reported that only a very few underdeveloped villages remained in their respective provinces. Encouraged, the government subsequently decided to provide development funding of Rp 20 million (equivalent to A$2,630) for each of the remaining problem villages. Shortly afterwards, the number of villages throughout Indonesia eligible for the new funding had swollen to 20,633 classified as underdeveloped and 200 classified as extremely underdeveloped (Dikun, 1994: 1–2). However, the information about which villages are receiving government assistance under the Underdeveloped Villages programme cannot form the primary database for detecting isolated tribes because, in reality, not all isolated tribes are willing to receive financial support from outside

46 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Obstacles to Creating an Inventory of Languages in Indonesia

45

and are, therefore, not all indicated in the database. For example, the Baduy people, who are considered an isolated tribe in West Java, refused government assistance. Their refusal came in a letter written in Kanekes village, West Java, dated 14 April 1994, and signed by two authorities in cultural matters (ketua adat) and the head of the village. They explained that their refusal was because the Underdeveloped Villages programme was in conflict with their philosophy and way of life, and had strings attached. The National Centre For Statistics Members of isolated tribes typically have low education, low income, high levels of illiteracy, and do not yet have facilities such as electricity, roads, bridges, schools, shops or hospitals. Consequently, it was thought possible to find potential locations for isolated tribes by searching the appropriate demographic variables from The National Centre for Statistics (Biro Pusat Statistik). However, the data published by The National Centre for Statistics go down only to the provincial level of detail. Data at the village level exists but its cost and format make it, in practice, inaccessible to individuals. Financial and technical constraints, then, have made it impossible to use the National Centre for Statistics, at least for the time being. The Department of Social Affairs The Directorate for Isolated Tribes (Direktorat Bina Masyaraka Terasing) in the Department of Social Affairs (Departemen Sosial) is the government body authorised to oversee and improve the welfare of isolated tribes but it has, in fact, been active in only 18 of the 33 provinces of Indonesia. The provinces where the Directorate for Isolated Tribes has done no work turn out to be the same ones whose governors deny the tribes’ existence. The conclusion is that the Directorate’s information on the location of these tribes for the whole of Indonesia cannot be accurate (Marzali, 1995: 22). Other sources of information Other problems exist. First of all, not all the tribes are dealt with in the literature. The first ones likely to be encountered are those reported on during the Dutch colonial period. Next, information on any one tribe is scattered over a number of unrelated sources. Another problem relates to the names of the tribes. The same tribe may be known as one thing in one province but, to people in another province, it will be known as something different. For example, the Sakai people of Riau are called Kubu elsewhere. Finally, the depth of analysis and general framework of description can be quite different from one publication to another, making comparisons difficult. Consequently, at this point, there is no comprehensive and reliable source of information about all the isolated tribes in Indonesia.

47 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

46

Learning about the Isolated Tribes The definition created by the Department of Social Affairs for an isolated tribe is fuzzy. The Department’s definition provides four criteria which distinguish an isolated tribe from other peoples in Indonesia: (1) (2) (3) (4)

they possess similar physical features such as bodily posture or hair; they share a similar social structure and cultural values; they live in a geographically isolated or inaccessible area; their standard of life is low.

We can dispense with the first criterion because, for example, the Dani people belong to the Melanesoid, the Mentawai people to the MalayoPolynesoid, the Sakai to the Veddoid (an ancient race of southern and southeastern Asia and northern Australia, characterised by dark brown skin, slim build, and wavy hair), while the Punan belong to the Mongoloid racial groups (Koentjaraningrat & Simorangkir, 1993). The second criterion is also fairly unhelpful because every tribe has its own social and cultural values. The third criterion is misleading because not all ‘isolated tribes’ are geographically isolated, living in remote and inaccessible regions or nomadic; some of them live close to or are interspersed with accessible towns but are socially isolated. The Baduy, for example, are isolated by choice from others because they want to preserve their own cultural, social and spiritual identity. The fourth criterion, of course, depends on one’s definition of standard of life. The problem is that such inventories are ethnocentric and miss completely the kinds of positive qualities that are enjoyed in the total framework of the people’s lives. Government viewpoints Until 1994, the Directorate for Isolated Tribes still believed that its most important goal for these people was to resettle, educate and modernise them. The programme was run through the Resettlement Programme (Program Pemukiman Kembali). Some critics claim that the programme has not produced any credible results (Marzali, 1995: 26), as the following examples from the Dani, Kenyah, Sakai and Baduy people partially illustrate. The Dani

One illustration of the problem comes from personal experience when I was undertaking research into the languages of the Dani tribe in the Baliem Valley of Irian Jaya (Lauder, 1992). Some dozens of dilapidated and unoccupied dwellings there had been built using the funds from the Directorate of Isolated Tribes. They were very basic, but resembled the kind of dwellings seen in Jakarta. Some Dani people stated that their traditional round huts were a great deal warmer and more comfortable than the ones the government had put up. The open windows, in the cold, rarefied air of the Baliem

48 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Obstacles to Creating an Inventory of Languages in Indonesia

47

Valley, were seen as a major design flaw. In addition, in the Dani culture, people live in round dwellings and pigs live in square ones. The government had apparently not taken these things into consideration. The Kenyah

Similar errors occurred in East Kalimantan. Dozens of houses for these people, built with government money, sit empty, never having been occupied. A team of researchers from the University of Indonesia who interviewed people from the Kenyah Dayak tribe learned that an essential feature of Kenyah Dayak longhouses is that they should be built lengthwise directly on the riverbank. The river remains a potent symbol of life for them. The longhouse, which is a single communal dwelling for the tribe, should also back onto the forest whose resources and riches provide the people with all that they need to survive. The government built separate houses for nuclear families and built them away from any river, thus making impossible the kinds of interactions essential to Dayak culture. The Dayaks also believe that living in a house built away from the river will bring a life full of disasters and difficulties. The Sakai

The Sakai people live in Riau province and were scheduled by the government for relocation. The resettlement programme took place in several stages. Starting in 1978 and continuing up to 1994, 391 families were moved from their traditional environment mostly to Bengkalis Prefecture (Direktorat BMT, 1995: 32). The government provided agricultural tools and basic foodstuffs for the first six months. In their new rectangular wooden houses, the people had to adjust to their new way of life and learn rudimentary agriculture in their gardens. Planting in one place was new to them because, originally, they were hunter-gatherers. The Sakai people began to use money to purchase manufactured items. However, their naïveté about the workings of a modern money economy made them vulnerable to unscrupulous people selling them things on credit, and many of them fell into debt. The Sakai have come to conclude that the modern world is in conflict with their philosophy of life and have decided to forego its pleasures for a more traditional and peaceful lifestyle in the forest. The Baduy

The Baduy live in the Leuwidamar district, a mere five hours’ drive southwest of Jakarta but they are, nonetheless, an isolated group. One of the first things that an outsider notices is all the prohibitions the Baduy have. The prohibitions are not thought of as repressive by the Baduy, but rather form an explicit moral and ethical code to preserve their culture and the environment. Some of the prohibitions are fairly universal: do not kill, do not hurt

49 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

48

another, do not steal, do not lie. Others go further: do not go to school, do not wear clothing of any colour other than black, white and blue; do not clean your body with toothpaste, soap, or shampoo; do not use or wear any item made of glass or porcelain; do not perform wet rice-field plantation; do not raise four-legged animals; do not make roads; and do not accept assistance when tied with conditions. There are many other such prohibitions (see Semat, 1994; Danasasmita & Djatisunda, 1986). The Baduy never lock their houses because there is no one who would steal their belongings. There is no drunkenness or violence. The Baduy are a hard-working, self-sufficient people who do not want to burden others. They choose not to use nails or other metal objects, glass or porcelain items, or other products of the modern world because, once discarded, these turn to garbage that is a problem to dispose of. The Baduy do not use commercially produced toothpaste, soap or shampoo because these things pollute the water; instead, they prefer to use herbal-based products made from guava leaves. The Baduy people do not perform any wet-plantation because it could cause erosion in their land, 70% of which is hilly with slopes of 30o to 70o. They do not raise four-legged animals partly because the animals could cause erosion on the slopes and partly because animals eat the plants. The Baduy’s diet is predominantly vegetarian and the adat chiefs are likely to be exclusively so. Baduy people are famous for their physical fitness. With no roads to their villages and prohibition against using any vehicle, they develop great cardio-vascular stamina and overall fitness with all their walking. They feel their life is stable, happy and self-sufficient. The Baduy feel that the benefits of modernisation are overrated. They prefer a life based on respect for nature, a sense of oneness with the Creator and a sense of accountability for their actions. Linguistic viewpoints Linguists are interested, among other things, in the diversity and variation within languages, and this has led to work in dialectology, which attempts to describe the boundaries between languages and dialects. Another area of interest is the idea of there being ‘genetic’ relationships between languages that appear to be different, i.e. the idea that, at some point in time, two languages originated from a single ‘ancestor’ that may no longer exist. This is the field of historical linguistics, which attempts to reconstruct protolanguages. The isolated tribes are of interest here especially for the generation of theories or hypotheses or for modifying existing theory based on European, African or American Indian languages. The Baduy speak an old form of Sundanese. Their social isolation has led to the preservation of many archaic Sundanese words in their daily speech. The vocabulary often relates to social concepts no longer in existence elsewhere. Their lifestyle still reflects precepts about being a morally righteous

50 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Obstacles to Creating an Inventory of Languages in Indonesia

49

person found in the Kropak 630 and Kropak 632 manuscripts from the preIslamic Sundanese Kingdom in Indonesia, around the 13th century (Danasamita & Djatisunda, 1986: 1–23). When applied to these languages, the disciplines of dialectology and historical linguistics are likely to throw light not only on the languages of the tribes themselves, but also on the fabric of the whole of the larger language groupings and on the historical contexts in which they exist. The languages can also be studied from other standpoints. Anthropological and sociological viewpoints The isolated tribes and their social and cultural behaviour are significant to anthropologists because they constitute a kind of living museum of human development. Different tribes exhibit different stages of that development, one tribe seeming to be close to stone-age culture, another illustrating hunter gathering, and another, shifting cultivation. The beliefs, rituals, social relationships, technology and artefacts of such peoples are available for investigation, and it is possible to ask the people themselves what the meaning is. A sense of urgency pervades such work, as the lifestyle of most tribal peoples is under threat from encroaching development. Apart from anthropologists and sociologists, historians can interview the members of these tribes to trace their history through the orally transmitted stories, poems, or songs. Another fact that the anthropologist has to contend with is that there is no one-to-one correspondence between a particular culture and a language. While it is true for the larger ethnic groups (such as the Javanese, Batak or Balinese) that the group has a single language, this is not always the case. For example, the small tribes found in Irian Jaya (such as the Bgu and the Mawes) share the same culture but use different languages (Koentjaraningrat & Simorangkir, 1993: 3–4). This makes for an extremely rich and diverse situation.

Language Mapping of Isolated Tribes Mapping the languages of the isolated tribes of Indonesia is no easy matter. If you begin with the assumption of using standard research procedures for dialectology, you will inevitably run into difficulties. Insights are required from disciplines other than linguistics. Some of the problems encountered by the researchers and their responses to them are documented here. Permits It is not easy, even for Indonesian researchers, to get the necessary permits from the local authorities to perform research. The whole affair will

51 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

50

usually consume between two weeks and a month. In addition to the usual bureaucratic tangle, travel to some areas where there is conflict is now restricted. Basic information The most reliable data for identifying and locating isolated tribes at the moment seems to be from the field researchers taking part in the project. Their first-hand knowledge of their village and the surrounding countryside makes them the ones most likely to know where the isolated tribes can be found. Because there is no reliable and complete source of information other than the field researchers themselves and because the core team does not meet the researchers until the training sessions, it is not possible to form in advance a comprehensive research plan for the isolated tribes in any province. Both the number of isolated tribes that need mapping and where they are to be found have to be decided through discussion between the core team and the field researchers. Location Getting to an isolated tribe’s settlement normally involves leaving any form of transport and making the final part of the journey on foot. While doing research in Bekasi, to the east of Jakarta, I made a journey to an isolated Javanese group of people who still use archaic Javanese vocabulary in their daily life. These people are to be found on a small island located in the middle of a lake in Sukatani prefecture. To get there, I used a motorbike, which was transported to the island on a small boat. These island dwellers said that they were the descendants of Javanese military forces commanded by Sultan Agung, the Javanese King of Mataram Kingdom, who reigned from 1613 to 1645, and who had attacked the Dutch VOC forts in 1627 and 1628. The Verenigde Oost-Indische Compagnie (VOC) translates from Dutch into English as the Dutch East India Company, which was founded in 1602 and sought to monopolise trade in Indonesia until the company wound up in 1799. The people explained that their ancestors had been defeated in battle and felt too ashamed to return home to Java, so they searched for a suitable place to stay, deciding on the island because of its inaccessibility. Their ancestors considered the island ideal as a place to defend in battle, should the VOC return to attack (Tawangsih, 1987). Many of the groups that fall under the category of ‘isolated tribes’ do not live in a permanent location. This phenomenon was evident to me when I carried out language mapping among the Punan in East Kalimantan in 1994, the Sakai people in Riau in 1995, the Kubu people in Jambi in 1995, and the Morunene people in South-East Sulawesi in 1996. These people have a hunter-gatherer lifestyle, shifting cultivation or swidden agriculture (a

52 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Obstacles to Creating an Inventory of Languages in Indonesia

51

swidden is a plot of land cleared for farming by burning away the vegetation). This shifting phenomenon also causes difficulties for the National Centre for Statistics in performing their census and for the language researchers who use their published information. One of our field researchers, for example, using the National Centre for Statistics map, made his way to a village, only to find the place long since abandoned. Guide and translator Most members of isolated tribes are monolingual, so one difficulty that field researchers face is mutual understanding with the informants. Often it is necessary to hire the services of a guide to get to these places, one who can also act as an interpreter. Ideally this guide should be a member of that tribe, an insider, but someone who has travelled out of the group and had contacts with the outside world. Alternatively, the guide and interpreter can be an outsider who has frequent contact with the tribe and become a sort of liaison person with the outside world. Finding this kind of person is not easy. The problem is further exacerbated by uncertainty over the reliability of the information received and then translated. Casley and Lury (1981) stated that: Ordinary people in most developing countries are extremely tolerant of requests for information and often too courteous or puzzled to voice their misgivings. (Casley & Lury, 1981: 131) However, this cannot be taken for granted when one is working with people from isolated tribes, as outsiders may be viewed with suspicion. Customs and beliefs Each isolated tribe has its own set of beliefs, customs and taboos. While the problem of the language barrier can often be overcome with a translator-guide, the constant potential for unwittingly breaking one of the group’s taboos is another matter. Local guides, however, are likely to be familiar enough with the patterns of behaviour and belief held by the people to provide the field worker with a means to contact and communicate with them. The guide can advise the field researcher which actions accord with the people’s sense of propriety and which should be avoided so as not to cause offence. Even when you get all of these things right, you may still be unlucky if you unknowingly arrive at the wrong time. Time plays a notable part in the regulation of activities in the culture of isolated tribes. On holy days or when ceremonies are taking place, other actions may not be permitted, including the presence of outsiders. Needless to say, it is difficult to obtain such information in advance.

53 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:29

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

52

Conclusions Despite the difficulties mentioned in this chapter, the benefits of making the effort to inventory, describe and also work towards preserving linguistic diversity in Indonesia are real. Any language inventory that does not include the languages of isolated tribes would be incomplete. Creating an inventory of the names, locations, and basic features of the isolated tribes throughout Indonesia would allow us in the future to more easily pursue multi-disciplinary research on individual tribes. A number of points of possible interest can be mentioned. The languages of these peoples together are of enormous diversity and variety. Among them we can find new speech sounds not yet described anywhere else. The languages, isolated as they have been up to now, provide an opportunity for examining pre-modern language and for generating hypotheses. There are some languages among this group whose sound system appears to be extremely restricted and these may offer clues about the origins of language. The geographical distribution of language features, once mapped, may also give clues to the spread and development of languages found in Indonesia. Just creating an inventory of the languages in Indonesia is hard enough without considering attempts to preserve them. One of the factors that needs to be taken into consideration is the political situation in Indonesia, which is still highly fluid. Democratic processes are definitely in motion, but predicting outcomes is difficult. Conflict among socio-political, ethnic or religious groupings may break out and the increasing demand for autonomy from the provinces and independence movements threaten national stability and, in some cases, cause the displacement of peoples. While the demise of many languages within a generation is not likely, it is likely that, as time goes by, people in the regions and their cultures and languages will be threatened by influences from outside. To what extent this is true of the isolated tribes is still difficult to say. The devolution of power to the regions may turn out to be a boon for the preservation of the regional languages, as this would allow a resurgence of pride in regional cultures and languages, and local people might be more likely to care about local cultural values and the local languages than do people in the capital. Thus, for the issue of linguistic diversity, the turmoil of change in Indonesia can be seen in a somewhat positive light. References Casley, D.J. and Lury, D.A. (1981) Data Collection in Developing Countries. Oxford: Oxford University Press. Direktorat Bina Masyarakat Terasing, Departemen Sosial Dirjen Bina Kesejahteraan Sosial (1995) Data dan Informasi Pembinaan Kesejahteraan Masyarakat Terasing. Jakarta: Proyek PKSMT Pusat, Departemen Sosial.

54 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:30

Color profile: Disabled Composite Default screen

Obstacles to Creating an Inventory of Languages in Indonesia

53

Danasasmita, S. and Djatisunda, A. (1986) Kehidupan Masyarakat Kanekes. Bandung: Proyek Penelitian Sundanologi, Dirjen Kebudayaan, Departemen Pendidikan dan Kebudayaan. Dikun, S. (1994) Sebuah Potret untuk Kemiskinan di Kabupaten Aceh Utara. Profil Desa Tertinggal Indonesia. Yogyakarta: Aditya Media. Koentjaraningrat and Simorangkir, V. (1993) Masyarakat Terasing di Indonesia. Jakarta: PT Gramedia. Lauder, M.R.M.T. (1992) Persebaran Kosa Kata di Lembah Baliem. Paper presented at Seminar Sosiolinguistik V, Depok. Marzali, A. (1995) Pembinaan Masyarakat Terasing di Indonesia. In Kebudayaan Lain-lain dalam Masyarakat Indonesia. Special edition of Majalah Ilmu-Ilmu Sosial Indonesia XXII (1), 19–27. Jakarta: Lembaga Ilmu Pengetahuan Indonesia. Semat, S. (1994) Masyarakat Baduy Masih Enggan Menerima Program IDT. Profil Desa Tertinggal Indonesia. Yogyakarta: Aditya Media. SIL International, Indonesian Branch (2001) Languages of Indonesia. Jakarta: SIL International. Tawangsih, M.R.M. (1987) Bahasa-bahasa di Daerah Bekasi. Jakarta: Yayasan Pancamitra.

55 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:30

Color profile: Disabled Composite Default screen

Chapter 5

Keeping Track of Indigenous Language Endangerment in Australia PATRICK MCCONVELL AND NICHOLAS THIEBERGER

Introduction Australia had around 250 languages when Europeans first settled at Sydney in 1788. Over half of those languages are not spoken any more – some are remembered in a few words and phrases by the descendants of the original speakers, some not at all. Of the remaining 100 or so languages that are spoken, at least half are spoken by only dwindling numbers of old people, and as spoken languages are likely to disappear with those old people unless something drastic happens to reverse the process. The other 50 or so languages are in a slightly better state of health, being spoken by younger people to a varying extent. But, even with these, signs of language shift, mainly to forms of English, are evident in most of them, and they must be accounted endangered. Only a smaller number are still spoken fluently and frequently by all age groups. Annette Schmidt, reviewing the situation in 1990, put 20 languages in this category of currently un-endangered languages and, 11 years later, we (McConvell & Thieberger, 2001) regard at least three of these as having moved into the endangered category. This does not mean, however, that we can regard the remaining 17 languages as safe. There are heavy pressures on all of them, and it has to be reckoned possible that they may all be lost as living, spoken languages by the end of this century. The last few years have seen important support for these strong languages, in the form of bilingual education programmes (recently withdrawn by the Northern Territory with as yet untold consequences for their long-term vitality) (Nicholls, c. 2001; Hoogenraad, 2001). A change of government in the Northern Territory in 2001 promised to herald a reversal of this policy direction but, at the time of writing, this promise remains unfulfilled, despite the publication of a new curriculum including indigenous languages and cultures. There are a number of hopeful signs that initiatives of indigenous people to nurture and revive the languages are beginning to bear fruit, and not all State Education Departments are going in the retrograde direction chosen by the previous Northern Territory government.

54

56 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:30

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

55

The focus of this chapter is mainly on how we can keep track of the changes in the viability of indigenous languages in Australia as a whole. Being able to describe the current situation and trends within it in a wellunderstood framework is vital if we are to launch policies and programmes that will be effective in maintaining languages. As Crawford notes (2000: 80–81), ‘Successful strategies of Reversing Language Shift demand an understanding of the stage we are currently in’. In this examination of how language endangerment can be monitored and assessed, we draw on work that we have carried out for Environment Australia, through the Australian Institute of Aboriginal and Torres Strait Islander Studies, as a separate report (McConvell & Thieberger, 2001) and as a contribution to the Cultural and Natural Heritage chapter of the Australian State of the Environment 2001 report (Lennon et al., 2001). While other issues and sources of information are occasionally mentioned, this chapter concentrates on what can be ascertained by the use of national census and other survey data on the public record. The census measures only a few superficial aspects of languages but at least it is carried out regularly with a strict methodology, budgeted for by the Federal government, and the data is available (for a price). We can criticise the Australian census, but it does include questions on language (unlike the census of some countries) and, since 1996, it asks people which specific indigenous language they speak (rather than just whether they speak an indigenous language or not). In this chapter, we compare the Australian census with the approach of the Canadian census to Canadian indigenous language data and the indicators that have been proposed to analyse that data. This is because the Canadian census design is in some respects better than the Australian census in this regard, but also because the situation of indigenous languages in North America (and Canada, in particular) is quite similar to that of indigenous languages in Australia. From hundreds of languages throughout the continent, the numbers have dwindled and continue to drop, leading observers to fear the total extinction of these languages this century. Indigenous people in North America and Australia are both ‘fourth world’ peoples, encapsulated and marginalised as a small minority in relatively rich ‘first world’ countries. Both have been subject to intense assimilationist pressures for over a century, including deliberate efforts on the part of missionaries and governments to eradicate their languages. Yet for both North American and Australian indigenous people, their languages are a key part of their identity and heritage which they are struggling to keep. The identity functions served by indigenous languages (IL) can continue to operate for some time after the IL is no longer spoken as an everyday medium of communication and, indeed, the determination of when a language is no longer spoken is not unproblematic. Definitions of what

57 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:30

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

56

constitutes a speaker will vary in this context and, as the number of speakers declines, so too may the qualifications for being the ‘best’ speaker. Evans (2001) discusses this in an article appropriately titled ‘The last speaker is dead – long live the last speaker!’ Being fluent in a language is not necessarily a prerequisite to identifying oneself as being a speaker of that language. Language programmes supporting IL-use may well have this identity function as their target rather than fluency in the heritage language (Thieberger, 2001).

The State of the Environment Report and the State of Indigenous Languages ‘Indigenous languages’ was included as an area to be monitored in the 5yearly Australian State of the Environment (SoE) reporting round for the first time in 1996. It is a breakthrough that indigenous languages have been included in this major review, which largely considers environmental conditions and trends. Languages do not fit easily into bureaucratic categories (as their history in UN agencies also shows). Nor are they as sympathetically received by those in education (to which they are often assigned) as we might expect, especially in a country like Australia, which has a strong (and perhaps resurgent) monolingualist and assimilationist background. Being counted as part of heritage and environment is a positive move since these are considered important issues by the public and by politicians, and people in general are more likely to pay heed to indigenous languages if they see them raised in this context. There are also strong links between the fields of language endangerment and biological endangerment, including the use of similar terminology and strategies to deal with both (e.g. ‘language ecology’) and the fact that indigenous languages in particular are a treasure-house of environmental knowledge that both complements, and often exceeds, that of scientific investigators (Maffi, 1998, 2001; Fil & Mühlhäusler, 2001). The SoE gathers data to measure a relatively small number of indicators that are intended to provide a broad but generally accurate picture of the conditions, pressures and responses in the area under study. Constraints on finances and time available meant that indicators needed to be selected also with an eye to the feasibility and cost-effectiveness of collecting the requisite data. Henderson and Nash (1997) proposed the indigenous language (IL) indicators shown in Figure 5.1, which the 2001 researchers (McConvell & Thieberger) were charged with applying in the conduct of the survey. Some modifications have been made, some of which are discussed in this chapter (for further details, see McConvell & Thieberger, 2001). The McConvell–Thieberger report is described by Harmon and Loh as: ... in many ways a model of its kind, especially in terms of its compre-

58 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:30

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

57

Issue 1: Condition of indigenous languages IL1

Number of people who identify as knowing each indigenous language.

IL.2 Number of people in age group who identify as knowing each indigenous language; proportion of total identifying as indigenous. IL.3 Number of traditional languages at each recognised stage of intergenerational dislocation. Issue 2: State of documentation of languages IL.4 The number of indigenous languages for which (a) documentation is (i) good (ii) adequate (iii) inadequate (b) documentation is close to complete (given the state of the language) Issue 3: The wider use of indigenous languages IL.5 The number of/proportion of traditional language used in (a) broadcast media: radio, TV, published books, magazines, cinema, WWW, distinguishing: (i) programmes aimed at speakers; (ii) programmes aimed at a general audience; (iii) signage in public places (streets, parks), advertisements IL.6 Number of approvals of geographic names, including map sheet names, using indigenous place names. Issue 4: Funding, research and education IL.7 Amount (in $A) of funding provided for language programmes through government departments and agencies, including ATSIC, DEETYA, ARC and AIATSIS; distinguishing allocations to (a) research; (b) language maintenance; (c) education and training; and (d) information dissemination and public education (e.g. translation of notices of government programmes) IL.8 The number of projects that document knowledge of traditional languages, by type of project IL9

The number and type of indigenous language programmes undertaken in language centres, schools and other institutions.

Figure 5.1 Indigenous language indicators Source: Henderson and Nash (1997)

59 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:30

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

58

hensive treatment of the factors that produce moribundity (and vitality) in small languages as they struggle to coexist with larger, sociopolitically dominant languages. (Harmon & Loh, 2002) After boiling down huge amounts of results, the executive summary of the chapter ‘Natural and Cultural Heritage: Key threats to sustainability’ of the report Australia State of the Environment 1996 has just this to say about Australian indigenous languages: Detail: Some traditional languages are declining, despite language maintenance programmes. Few languages are used as primary means of communication. Comment: Many languages have been lost; others are declining. (Environment Advisory Council, 1996) The extreme compression of statements like this tends to produce ambiguity. The first sentence under ‘Detail’, for instance, could be read as meaning that language maintenance programmes are having little effect where they are in place. This interpretation could potentially have unfortunate policy repercussions. This, we are sure, was not the intention, rather that there are a few language maintenance programmes, rather poorly resourced, but that this is having little overall effect on the overall situation of languages. Some approaches to language endangerment attempt to give minimum absolute figures (ranging from 10,000 to 100,000) for populations speaking a language if it is to have a chance of survival (Krauss, 1992). Under this approach, all Australian ILs are doomed, since the largest languages have only a few thousand speakers. Even with a lower cut-off point of 500, as used for Mexican indigenous languages by Garza-Cuaron and Lastra (1991: 97), all but a few Australian languages are marked for extinction. However we, along with Henderson and Nash and a number of other authors (Crystal, 2000: 12–13; Nettle & Romaine, 2000: 9–10), are not convinced of the validity of this indicator. For one thing, in traditional times before white settlement, languages had only a hundred or so to a few thousand speakers and did not die out, at least not on a regular basis. Of course, we are talking about a very different situation in Australia and the world now but it has not been convincingly proved that these absolute numbers indicators are of any great value. Consequently we have looked at other ways of measuring language endangerment more suited to the Australian situation.

The Geography of Language Endangerment in Australia Figure 5.2 shows the distribution of strong and weak languages in Australia using as a criterion the percentage of the indigenous population

60 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:30

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia 0

59

1000

Darwin Kilometres

Cairns

Derby

Townsville

Dampier

Alice Springs Rockhampton

Brisbane

Geraldton Kalgoorlie

Broken Hill

Esperance

Perth

Sydney

Albany

Canberra

Adelaide 79% or more 30% to 79% 7% to 30% Less than 7%

Melbourne

Based on indigenous locations

Hobart

Figure 5.2 Indigenous language speakers as proportion of indigenous population (based on indigenous locations) Source: ABS website (2205): Map S1.5

who claim to speak an IL at home. The strongest traditional language speaking areas, where over 79% speak a traditional language at home, are in Arnhem Land, Central Australia and the Western Desert, and a small part of Cape York Peninsula and the Torres Strait in Queensland. Elsewhere in the Northern Territory and the north of Western Australia, some languages are fairly strong but threatened and others are fading out, leading to overall figures of between 30% and 79%. The next category down, with figures between 7% and 30%, accounts for a few areas of Western Australia, a large part of South Australia, and most of Cape York Peninsula, where the languages are generally seriously endangered, with only elderly speakers. In the remainder of the country, throughout all of Victoria, Tasmania, New South Wales, most of Queensland and the southwest of Western Australia, the percentage registers at less than 7%. For the most part, no traditional indigenous languages are used, although identifi-

61 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:31

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

60 0

1000 Kilometres

Early European settlement

Figure 5.3 White settlement of Australia, 1860 Adapted from Ross, 1999

cation with a language remains important for most indigenous people and many groups are keen to reclaim their heritage languages if they can. The figures here are drawn from the 1996 census, based on the indigenous population of regions compared with the number of people who claim to speak an indigenous language at home. This is not broken down by language, and cannot be since there is no question in the census that asks people which language group they identify with, separately from the question of the language spoken at home. This issue is further discussed below. Figure 5.3 shows the primary reason for the distribution of languages – languages have died out where white settlement has had the earliest and most powerful impact: in southern and eastern areas of the continent in the first half of the 19th century. Although indigenous resistance seriously challenged the further spread of settlement through Queensland from the 1850s on, the deployment of police and settlers with greater fire-power turned the tide in favour of the settlers and led to the decimation or annihilation of

62 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:31

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

61

some language groups. In Queensland in particular, from the turn of the century draconian policies of removing indigenous people from their home country into institutions far away exacerbated the decline of languages in the interior. Elsewhere, in northern and central Australia, traditional languages tended to be spoken until the 1950s on cattle and sheep stations and some missions. On other missions, however, use of languages was deliberately suppressed and people were often dislocated from their tribal groups and mixed up, not sharing a common language except for a variety of English or ‘pidgin English’ which became the first language of the succeeding generations. After World War II, the pidgin that had become a lingua franca for many groups across the ‘cattle station’ belt of the northern savanna also became the first language of many children on the stations (today often called Kriol).

Decline of Percentage of Speakers Figures for those who speak an indigenous language at home as a percentage of indigenous population can be combined for the whole of Australia. We can then compare the figures from the 1986 census with those from the 1996 census as shown in Figure 5.4. The graph shows the continuing and accelerating decline in IL speakers as a proportion of the indigenous population over those 10 years and there is little reason to believe this trend has ceased since then. The major drop in percentage in the oldest age group between 1986 and 1996 would be in large part due to the death of speakers of languages who have not passed the language on to younger people (or at least not their household use). Figure 5.4 also appears to show attrition of use of languages over the lifespan as well as failure of transmission to succeeding generations. The similarity of figures of the 5–14 and 15–24 age groups in 1986 indicate some levelling off of a downward trend in the 1970s and 80s. There is a parallel trend in Canada in this period and, in both countries, it may stem from the scaling down of assimilationist policies, leading to greater willingness to report speaking of indigenous languages, as well as perhaps a return to higher levels of transmission. This translates into a similarity between the 15–24 and 25–43 age groups in 1996, 10 years later. However, the percentage of speakers in the age groups has declined. While the 1986 15–24 group cannot be validly equated with the 25–43 group of 1996 because the age range is different, the 1986 5–14 age group can be equated with the 1996 15– 24 group. Here we find a drop in the percentage speaking a traditional language at home from 16% to 14.5%, which seems to be due to a lessening of use of the languages, rather than failure to receive transmission of them.

63 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:31

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

62 35%

1986 30%

28.48%

1996 25%

19.70%

20% 16.96%

16.80%

16.00% 14.46%

15%

14.21%

10.88%

10%

5%

0% 5–14 yrs

15–24 yrs

25–44 yrs

45+ yrs

Age

Figure 5.4 Percentage of Indigenous people speaking an indigenous language, comparing age groups in 1986 and 1996 Based on census data for indigenous people over 5 years old who speak an indigenous language

Attrition of indigenous language use in the young adult age group, when people marry, leave home, go to work etc. is also reported from Canada (Norris, 1998). The youngest age group in 1996 in Australia has also plummeted to a low level of speaking traditional languages, no doubt due to the attrition of language use in the older groups negatively affecting transmission. These percentage figures, while disturbing, do not mean that there are now fewer indigenous people who speak languages at home than in the past. This is due to the fact that there is a highly pyramidal age structure in most of the Australian indigenous population, with many more children and young people than middle-aged and old people (McLennan & Madden, 1999). So, if the language is being transmitted to a similar percentage of the younger generation as speak it in the older generation, it tends to be transmitted to more people than are speakers in the older generation. This contrasts starkly with the age structure of the non-indigenous population of Australia, where numbers in age groups tend to be about the

64 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:32

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

63

same from young to old. High fertility in the indigenous population is not curbed now as it was for many years by high infant mortality, which has largely been brought under control by aggressive use of Western medicine and hospitalisation. On the other hand, ‘lifestyle’ diseases and trauma are claiming many more lives in middle age and even younger. Another relevant factor in interpreting the trends in Figure 5.4 is that there has been an increase in people identifying themselves as indigenous over the same period, concentrated in the late 80s and early 90s. The people newly identifying themselves as indigenous tend to be mixed-race people, mainly in southern Australia, whose families may not have broadcast their indigeneity in the past or had it recorded for official purposes because of the negative consequences of doing so. Nowadays, however, the situation has changed and people are much more likely not only to admit to, but to be proud of, their indigenous origin. Since most of the people in this category are non-IL speakers, we might attribute at least part of the decline in speaker percentages to this addition of non-speakers. This is likely to be a one-off historical event, not to be repeated on that scale at least, so future census figures will more accurately represent the real balance of IL speakers and non-speakers. On the other hand, another factor to be weighed on the other side of the balance is that there is some evidence of over-reporting of IL speaking. This would tend to skew the figures in a more positive direction for language maintenance. This is discussed further below.

Regional Patterns of Language Shift A nationwide survey of indicators of indigenous demographics and life patterns was carried out by ATSIC (the Aboriginal and Torres Strait Islander Commission) in 1994. Unlike the Australian Census, this 1994 ATSIC survey included more than one question on language. This was the only survey of such scale that compared a general ‘speaking’ criterion against ‘speaking at home’ (only the second is used in the census). In this way it was more like the Canadian census, which asks about ‘mother tongue’ (MT, generally interpreted as ‘first language’) and separately about ‘home language’ (HL). In what follows, the criterion other than HL which is used in the 1994 census is referred to as MT, although it must be borne in mind that this may not correspond exactly to MT definitions used in other census work. Unlike the 1996 census, the 1994 survey did not ask which specific languages people spoke, but only whether they spoke a traditional indigenous language. The 1994 ATSIC survey was carried out using ATSIC regions, which are also used in classifying census data. There are discrepancies in a few regions between what we might expect from observation of language use

65 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:32

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

64

and the results of this survey, pointing to possible problems in the implementation of methods in some areas. However, generally the results are in line with what we know from observation of language use in regions and census data and may be considered reliable. Figure 5.5 presents those results: clearly the levels of both ‘speaking’ and ‘speaking at home’ vary greatly between regions, and so does the discrepancy between the two. In some regions (like Jabiru and Nhulunbuy in Arnhem Land, and Aputula and Warburton in the Western Desert) there is little difference; in other regions there is markedly less reporting of ‘speaking at home’ than just speaking or knowing a language. Figure 5.6 shows some of the same ATSIC regions as they appear in the 1996 census data, broken down by age group, using just the HL criterion. As with Figure 5.5, there is a high level reported and little decline from old to young in Arnhem Land (Nhulunbuy) and the Western Desert (Warburton). In other regions there are various starting levels among old people and degrees of rapidity of decline through the age groups. If we look mainly only at those cases where there is reporting of significant IL speaking, two measures, emerging from the 1994 survey and the 1996 census respectively, tend to line up: • the level of difference between HL and MT in the 1994 survey; and • the pattern of maintenance/decline of languages across age groups. Three broad groups can be identified, using the above criteria. These roughly line up with strong, endangered and (near) extinct categories, which have been proposed for language endangerment and are discussed further below. • Strong: Nearly all speak IL at home; little difference between age groups. • Endangered: Many fewer speak IL at home than claim to know it; sharply declining use of IL among young. • ‘Nearly extinct’ or critical/terminal: Very few speak IL at home; proportion of speakers low in all age groups and declining. Figure 5.7 shows how these patterns are distributed for a sample of regions given on the charts, overlaid on the map of percentages of HL speakers (Figure 5.2). Generally speaking the distribution is similar to that discussed for Figure 5.2. The age data on some of the ‘endangered’ language regions shows a dip in the middle of the graph in Figure 5.4 with particularly the 30–39 group showing slightly less HL than younger groups. This contradicts the general trend of continuous decline and, in the case of Kununurra, the dip is quite significant. This may be related to the fact that these people were language learners in the late 1950s and early 1960s when indigenous affairs were still

66 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:32

W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:33 67

Source: 1994 ATSI Survey (ABS, 1996). Relative proportions of those who can speak an IL and those who speak an IL as the main language at home

Speak IL at home

Can speak IL

Figure 5.5 ‘Speaking’ and ‘speaking at home’ by TSIC region

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Ro ma Qu Hob e an art be ya T a mw n ort Br is h b an Wa Ba e ll gg a arat W Co ag ffs ga H W arbo a u Ro ngar r at c k h am ta pto Sy n dn ey Pe Na rth rog i Bo n To urke w S o uth nsvi He lle d Ge land ral dto A n d ela ide C Ka airns lgo orl Ce ie du n Da a M rwi o un n t Is a B To room rre sS e Ku trait Po nun ur rt A ug a us ta D A l i c e erby Sp r Co ings ok t Ka own T the e n n an rine tC W ree a rbu k rto n Ja N b h ulu iru nb u Ap y utu la

Indigenous Language Endangerment in Australia

65

Color profile: Disabled Composite Default screen

IL speakers 60 and over

IL speakers 50-59

IL speakers 40-49

IL speakers 30-39

IL speakers 20-29

a

Source: ABS, 1996

Figure 5.6 ‘Speaking at home’ across some ATSIC regions, by age group

0%

10%

20%

30%

40%

50%

60%

70%

80%

rth

Pe

IL speakers 0-19

me

Br oo

ille sv

To wn

ide ela Ad

s Ca

irn Po

rt A u

gu

sta ut So

nd d la hH e

rw in Da

90%

Ku n

un

u rr A

k r ee

s ng p ri

lice S

rby De Te n

S

tC na n

ea Ar tra it es To rr

on urt Wa rb

wn k to Co o

ne N

eri Ka th

uy nb hu lu

u b ir Ja

W:\Clients\MLM\Cunningham\Proof2\Cunningham Proof 2a 31 January 2006 20:38:07

la u tu

66

Ap

100%

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

67

Pattern of language shift 0

1000

Nearly all speak IL at home; not much difference between age groups. Many fewer speak IL at home than claim to know it; sharply declining use of IL among young.

Darwin

Kilometres

Cairns

Very few speak IL at home; proportion of speakers low in all age groups and declining.

Brisbane

Perth Sydney

Albany

Canberra

Adelaide 79% or more 30% to 79% 7% to 30% Less than 7%

Melbourne

Based on indigenous locations

Hobart

Figure 5.7 Stages of language shift in the grip of assimilations. But the next cohort, growing up in the late 1960s to 1970s, were slightly more inclined to be exposed to ILs (see discussion above, in relation to Figure 5.4, of the reducing decline in IL speaking at the same period). On the other hand, more local explanations may play a role, such as younger IL speakers moving into the region from outside. Adelaide has an unusual pattern, which should be commented on. In 1994, 10% of the indigenous population reported knowing an IL, but virtually none spoke it at home. In 1996, the HL figures were low but increasing in the younger age groups. There are a couple of factors here that need to be disentangled by further research. Adelaide has a fairly high transient and resident population of IL speakers from the north of the state, where Western Desert dialects are strongly maintained. These people know an IL, but may not use it much while in the city. A second factor is the presence of an active language revival movement surrounding Kaurna, the original language of the Adelaide region (Amery, 2000). This is also reflected in high prestige accorded to other ILs of the region, which have not been generally spoken for some decades but which indigenous people are seeking to reclaim. The difference in the response to the HL question in 1994 (when almost no people claimed to speak an IL at home) and 1996 (when a small proportion, and – contrary to trends everywhere else – more young than old, did

69 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

68

claim to speak an IL at home) requires discussion. It is unlikely that over two years the language revival movement has created new full speakers. But it is true that Kaurna words and phrases are being used, interspersed within English, more and more, in public contexts and at home (Rob Amery, personal communication). Respondents may choose to label this behaviour ‘speaking an IL at home’, and this could be seen as a variety of over-reporting. Canadian work suggests that speaking an IL gained a lot more prestige in the 1980s, causing higher reporting (when previously this activity was more hidden from authorities), and even over-reporting. Crawford (2000: 64) writes of a similar phenomenon in the USA. It is quite likely that such over-reporting is also occurring at least in some situations in Australia, and Adelaide may be one of them. Sophisticated methods may be able to use degree of over-reporting as an index of the ‘covert’ prestige of a language (cf. Trudgill, 1983: 173ff). From a methodological point of view, though, it is interesting that the possible ‘over-reporting’ of HL occurred only where there was a single question about language in the census. Here it is more likely that people will take the opportunity to assert their linguistic identity and pride, irrespective of the detail of the question. Where two distinct questions were asked (MT and HL) in 1994, however, people distinguished more accurately between linguistic knowledge that they had, or were gaining, on the one hand, and what they actually spoke at home. This provides another reason for preferring a survey instrument that has two questions about language, not just one, apart from the ability to gauge language shift from this, which has been touched upon above and will be discussed further below.

Measuring Language Endangerment One of the key questions in measuring language endangerment and language shift is the extent to which people who identify themselves with a particular linguistic group actually speak the language of that group. In Australia, most indigenous people tend to identify themselves primarily as members of a language group or ‘tribe’, a term still widely in use among indigenous people, although disdained by anthropologists nowadays. Some who have lost touch with, or find little regular use for, these aspects of their identity may instead describe themselves just as ‘coloured’, ‘Aboriginal’ or one of the regional terms for this (‘Koori’, ‘Murri’, ‘Nunga’, ‘Nyungar’, ‘Yamatji’, etc.) and as associated with a locality or former mission settlement. In recent years, especially in the context of Native Title, however, language group identity has become much more salient for a wide range of indigenous people. A different but related question is the extent to which a person actually

70 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

69

uses his or her mother tongue or first language. This is different from the issue of ‘identity’ language above because, even in traditional times and certainly today, people might for various contingent reasons have grown up speaking, as a first language, a different language from their primary identity language. One can, of course, multiply such complications, but what is important is to choose criteria that can provide feasible and costeffective indicators of language shift and language endangerment. The Australian census asks only one question about language, the HL question, and there is no other easy way to find out about tribal identity or mother tongue with which to compare the HL data. The Canadian census has a number of advantages over the Australian census for recording of language situations and trends; the provision of separate MT and HL questions is one. The Australian census tends to follow the pattern of the United States census. The United States census can, however, be used in conjunction with records of tribal membership (of the kind not kept in Australia) to track language shift by the extent that numbers speaking a specific language diverge from the recorded tribal membership numbers. Canadian measures of language vitality/language shift In Canada, several ways of measuring language shift using national census data have been suggested, most of them based on the distinction between numbers of MT and HL speakers of languages. These include: (1) difference between MT and HL, Language shift (Cook, 1998): MT speakers – HL speakers MT speakers Language vitality index (Norris, 1998; Drapeau, 1998): HL speakers MT speakers Cook suggests a major difference between endangered and non-endangered languages: 30% (or 70% vitality) could be the threshold for endangerment. (2) percentage of speakers in younger (5–14) age group: less than 30% ‘very poor condition’ (Drapeau, 1998); (3) gap between percentage of speakers in oldest group (55+) and youngest (5–14) (‘language transmission index’, Drapeau, 1998); (4) average age of speakers of IL (Norris, 1998): 40–50 years old indicates endangerment; 20s–30s viable.

71 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

70

(1) to (3) are all quite reasonable proposals, but all rely on having data on both MT and HL numbers or being able to calculate the percentage of speakers in a whole group or age group within it. These data are not available through the Australian census, nor elsewhere except in a few localities where a systematic linguistic survey has been carried out that includes the appropriate questions. Hence, no comprehensive national survey of individual languages in Australia is possible at this stage. Even where two questions have been included in local surveys, they have tended to be neither the pair used in Canada (MT and HL) nor one seeking to know language identity as well as ability to speak a language. In the Maningrida survey of a number of languages, for instance, carried out as part of the ATSIC 1997 survey, the questions were about the languages known and the language most used, not including primary language identity. Index (4) is feasible with Australian IL data, and we implemented such a measure (McConvell & Thieberger, 2001). The ordering of languages as to degree of endangerment using this measure in Australia turned out to be wildly at variance with what we know to be the case from observation and reliable reports, whether using average or median age. The error appears to arise from at least two sources: (1) As mentioned above, indigenous populations in Australia have a highly pyramidal age structure and such populations, with very large numbers of children and young people and very few old people, have a low average age of speakers, supposedly indicating low language endangerment, even if they have relatively few young speakers; and (2) compounding this effect, a number of language communities appear to be over-reporting the speaking of an IL at home among children and young people. The reasons for this are further discussed below.

An Endangerment Index using Age Profiles? The Canadian endangerment index (4) above, while not particularly helpful in its present form, does, however, suggest that we look more closely at the possibility of drawing a valid index from just the age profile of HL speakers, without needing to know what percentage these are of MT speakers or the whole ‘tribal’ group. The advantages of this approach are reinforced when we look at how closely the age profiles of HL speakers in regions (in Figure 5.6) line up with other measures and other sources about the regions. In discussing Figure 5.6, we pinpointed three broad demographic patterns in the age group data, corresponding to strong, endangered, and (near) extinct languages, without rigorously defining these. The graph is, in fact, plotted on the basis of percentages of the total population of the particular age group in a region that speak a language – a method which, as already noted, cannot be used for individual language groups.

72 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

71

However it seems feasible that raw numbers of HL speakers of a particular language in age groups, which are available in census data, could be used to generate similar profiles. The fundamental idea is that strong languages will have high numbers of HL speakers across all age groups, and (near) extinct languages low numbers across the board. Between these, and the main focus of our interest, are the endangered languages that generally show a pattern of sharp decline in language use from old to young. Although language shift is obviously made up of a complex set of behaviour patterns that could be studied in detail, we are stressing here what can be obtained from census data only, i.e. answers to the HL question. As the census doesn’t give us HL versus MT figures in Australia, we must look for other ways of measuring endangerment using census figures. One possible way is to capitalise on the difference in age profile of indigenous people in general and the age profile of speakers. The indigenous population profile in most regions is a pronounced pyramid with large numbers of children and a steep decline through early and middle adulthood. With endangered languages, the number of speakers in the younger age groups does not reflect this profile. The gap between the two youngest age groups (of speakers and in the general population) can be used to give a language endangerment index as in the following formula. Scores below 1 are said to be endangered; over 1 are not. Age-profile index of language endangerment of IL: IL speakers in 20–39 age group total number of IL speakers

divided by

IL speakers in 0–19 age group total number of IL speakers

This is a simplified index based only on the two youngest age groups. In some situations of rapid language shift, the ‘tip’ might have occurred between say the middle-aged and older groups, with the two younger groups both being at a low level of speaking the IL. If this index were used, such languages would not, therefore, show a low score indicating endangerment. The index could be modified to the relationship between, say, the youngest and oldest instead, as in the Canadian ‘language transmission index’ (3) above, but this would run into the problem of exceptionally low figures in the older groups due to mortality. For the moment, we will continue with this index to test it against actual figures for ILs in Australia. It turns out that there are other problems in the data that throw doubt on the use of this index with census data.

Age Profiles for Australian Endangered Languages In the 1996 census of Australia, data on HL speaking of individual

73 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

72

languages was gathered for the first time. Table 5.1 shows age group data for a number of languages that had been classified by Schmidt (1990) as strong (above the heavy line in the table) and endangered (using information, usually rough estimates, supplied by linguists, not census data). Most of the languages formerly classified as ‘strong’ did yield an index above 1. Kuku Yalanji gave a much lower figure and other reports suggest that this language is indeed endangered. Yindjiparnti and Yulparija are marginally now endangered and this also seems an accurate reflection of their situation. The low index for Yankunytjatjara does not seem to fit this strong language category and is probably due to incomplete compliance with the census. Some languages in the previously endangered group (shown in bold in Table 5.1) show age profile endangerment indices below 1, as expected for endangered languages. Others, however, show figures well above 1 – examples discussed below are marked with asterisks. One of the present writers (McConvell) is personally familiar with three of the asterisked language situations with the ‘anomalous’ high indices, through living and working on language and education in or close to the relevant communities. Miriwoong is in severe decline: children do not Table 5.1 Age data on some endangered languages in Australia Language

ABS Schmidt Endang Census (1990) erment 1996 index

0–19 (%)

20–39 (%)

40–59 (%)

60 and over (%)

Anindilykwa

1224

1000+

1.24

45.97

37.10

14.11

2.82

Arrernte total

6493

3000+

1.27

43.83

34.46

14.77

6.93

8203 Arrente (Aranda)

3817

1.24

43.57

35.13

15.43

5.87

8202 Anmatyerr (Anmatyirra)

1224

1.23

41.58

33.82

15.60

8.99

8201 Alyawarr (Alyawarra)

1452

1.35

46.35

34.44

13.29

5.92

Burarra

702

400–600

1.49

49.14

33.05

13.79

4.02

Central Torres Strait (KKY)

929

3000– 4000

1.19

38.15

32.11

18.97

10.78

Dhuwaya / Dhuwala

3648

1700– 2000

1.4

49.19

35.12

13.36

2.33 11.05

Jaru (Djaru)

344

250

1.48

43.02

29.07

16.86

Kuku Yalanji

256

300

0.61

25.71

42.45

22.45

9.39

Kunwinjku

1400

900

1.28

45.55

35.52

13.52

5.41

Maung

234

200

1.08

39.75

36.82

19.67

3.77

Murrinhpatha

1434

900+

2

56.85

28.39

11.61

3.15

Nyangumarta

263

700–800

1.1

38.22

34.75

13.13

13.90

Tiwi

1822

1400

1.07

40.99

38.37

16.92

3.71

Warlpiri

2667

3000+

1.25

42.99

34.43

14.63

7.95

74 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

73

Table 5.1 – continued Western desert Eastern total

2584

8218 Pitjantjatjara 8217 Pintupi 8225 Yankuntjatjara

1000+

1.02

36.20

35.64

19.84

8.31

2121

1.04

38.90

37.29

17.35

6.46

390

1.2

42.31

35.38

16.15

6.15

73

0.8

27.40

34.25

26.03

12.33 10.17

Western Desert Western

1669

1.12

39.81

35.49

14.52

8215 Ngaatjatjara

993

3000+

1.19

41.59

34.84

16.41

7.15

8224 Yulparija

96

0.94

35.42

37.50

12.50

14.58

8212 Kukatha (Gugaja)

580

1.24

42.41

34.14

14.66

8.79

1.14

39.64

34.91

18.46

6.98

Wik Mungkan

845

900– 1000

Yindjibarndi

324

500–600

0.97

34.94

35.84

21.08

8.13

8118 Yanyuwa (Anula)

51

70–100

0.62

25.49

41.18

19.61

13.73

8115 Rembarrnga

69

150

0.94

42.03

44.93

13.04

0.00

8104 Dhay’yi

70

1.32

47.14

35.71

8.57

8.57

8116 Ritharrngu

94

300

1.32

35.11

26.60

24.47

13.83

8107 Karrwa (Garrwa, Garawa)

110

200+

0.59

20.91

35.45

27.27

16.36

*8213 Miriwoong

111

20-Oct

1.39

47.75

34.23

15.32

2.70

8214 Mutpurra (Mudburra)

115

50

0.67

26.96

40.00

20.87

12.17

8106 Djinang

120

200–300

1.09

41.67

38.33

15.00

5.00

8601 Adnymathanha (Yura Ngawarla)

127

20+

0.68

29.92

44.09

21.26

4.72

8205 Bunuba (Bunaba)

165

50–100

0.83

33.33

40.00

12.73

13.94

*8502 Nyungar (Noongar)

167

20

1.28

46.11

35.93

11.38

6.59

8113 Ngangkikurungurr

223

100-

1.37

47.98

34.98

10.76

6.28

8402 Meryam Mir

317

100+

0.54

17.03

31.55

37.54

13.88

8114 Nunggubuyu

356

300–400

1.22

39.61

32.58

18.82

8.99

8204 Bardi

380

100–200

0.66

23.42

35.26

25.53

15.79

* 8207 Kija (Gidya)

408

300

1.39

48.53

34.80

8.33

8.33

8222 Warumungu (Warumunga)

518

200

1.01

37.26

36.87

17.95

7.92

* 8208 Kuurinji (Gurindji)

545

250

1.43

47.52

33.21

14.50

4.77

8302 Guugu Yimidhirr

739

400

1.3

42.35

32.61

16.78

8.25

8221 Walmajarri (Walmadjari)

858

1000

0.87

31.70

36.36

22.03

9.91

Note: Those languages above the bold dividing line are considered ‘strong’ in Schmidt’s 1990 report. Those marked with an asterisk are discussed in the text.

75 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

74

speak it and probably understand little. On the other hand, there is a strong pride in the language and a school language programme. Kija is similar (McConvell, 1986). Gurindji children speak a mixed language based mainly on Kriol (an English-based creole), but regard this as Gurindji (Dalton et al., 1995; McConvell, 2001a, 2001b). It is well known that traditional Nyungar of the south-west of Western Australia has not been spoken for at least 40 years, but some people in that area use many Nyungar words and phrases interspersed in local English (Douglas, 1968; Von Brandenstein, 1988) and this is evidently classed as Nyungar in census responses. It would not reflect reality if these languages were to be classified as ‘not endangered’ on the basis of this age-profile test and we are, therefore, not in a position to recommend it as an index, given the nature of census data presently collected. There appears to be over-reporting of indigenous language use in the home in some communities, but not in others. It is difficult to be certain of the reasons without further detailed research. The methods used in the census may play some role: individuals in a household or even a neighbourhood may simply classify large groups of people as ‘speaking X language in the home’ when they probably have in mind ‘have an X language identity’, and this is accepted by the census in some places. Local pride in a language and language programmes that increase its profile may also add to the tendency to classify young people as speakers when they rarely, if ever, speak the IL at home. Situations of rapid change in the language and use of mixed language by the children and young people present a problem since the census does not have a category for this. Particularly if the young people themselves think what they speak is a form of the old language, this is what will be recorded. We will return to these problems at the end of the chapter.

Dedicated Language Surveys Despite the criticisms of census data above, we still feel that use of the census to collect data on IL use has a lot of advantages. There needs to be room for regular review and occasional modification of methods and perhaps an addition of a question about mother tongue (MT) or language identity. The alternative is to carry out surveys that are more directly dedicated to finding out important information about ILs. The State of Indigenous Languages report, which we have been working on (McConvell & Thieberger, 2001), could be such a survey. At present, however, it is not funded to the extent that it could carry out field surveys, only to collate and analyse other sources, and that only to a limited extent circumscribed by a small budget. We have produced an ‘Indigenous Languages Database’ as

76 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

75

part of the report, but there is no on-going support for this to be maintained and updated. It is unlikely that the Australian Federal Government would fund a major sociolinguistic survey of ILs of the kind undertaken by the New Zealand Government in 2001 following on from a previous survey in 1995 (Te Puni Kokiri et al., 1998). Australian ILs do not have the political importance that Maori does in Aotearoa (New Zealand), and there is no recognition of Australian ILs in the way that the Treaty of Waitangi and the Maori Language Act recognise the Maori language. We have drawn on a survey of indigenous indicators carried out by ATSIC in 1994 and referred to above, which in one respect contained slightly more information than the census. Amore thorough-going national ‘Needs Survey’ on ILs was embarked on by the languages section of ATSIC in 1996–7. Although driven mainly by concerns about reforming budgetary procedures in the languages area, it included questions about language speaker numbers and use of languages. A large questionnaire was circulated to bodies such as regional language centres. The quantity and quality of the responses was very mixed and, while a report summarising results was written, it was judged by ATSIC to be of poor quality and unreliable and was never released, even to investigators such as ourselves. While this experience may have disposed some people not to repeat a national survey, there is a continuing need for reliable information on language situations, and this has motivated some local and state-wide surveys. Hoogenraad’s (1992) survey of language populations in Central Australia to find out about educational needs is widely respected as a methodical piece of research carried out by visiting many isolated communities and talking to key indigenous people to assemble a total picture of languages. The survey revealed that there were around 15% more indigenous people and IL speakers than recorded by the census. A survey currently being carried out by the Diwurru-wurru Jaru Corporation (Katherine Regional Aboriginal Language Centre) is using the methods pioneered by Hoogenraad. In New South Wales, another ‘Needs Survey’ was funded by ATSIC and produced by a team from the Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS) (Hosking et al., 2000). It was again mainly concerned with developing a funding model, with an appendix giving some notes on language programmes and situation but not in a systematic framework. Some of the indicators developed in the McConvell–Thieberger report were also used and modified in a survey of the status of indigenous languages in South Australia carried out by AIATSIS for ATSIC in 2001–2 (McConvell et al., 2002). While local and state initiatives are to be applauded, they are not being carried out using a consistent framework that would yield comparable data

77 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

76

across Australia. There is a need for consistent guidelines for gathering data. In the remainder of this chapter we focus on just one aspect of the problem: how to find an indicator for language endangerment that can be applied by different surveys to yield comparable results.

A Language Endangerment Indicator There are a number of different proposals for ways of classifying language endangerment. One of the most influential internationally has been that of Kinkade (1991), which divides language situations into five stages. This has been taken up, particularly by Canadian researchers and, below, we recommend a quite similar approach. Another influential scheme is that of Fishman (1991: 395) on ‘reversing language shift’, in which eight stages are recognised. The terminology of this scheme was largely adopted by Henderson and Nash (1997) in their report, although the immediate source was McKay (1996) which followed Fishman to some extent, but was also critical. Henderson and Nash provided the indicator (see above IL 3) for use in the current SoE round. We consider the merging of assessment of situations and of language maintenance intervention strategies, which was quite deliberately incorporated by Fishman into his scheme (1991), to be a bad move, and prefer to deal with the two issues separately (McConvell & Thieberger, 2001; see also McConvell, 1992; Lo Bianco & Rhydwen, 2001). Fishman’s scheme also incorporates assumptions from his theory of the relationship between diglossia and language maintenance, which are both questionable in themselves and unnecessary for the purpose of assessing language endangerment. We have, therefore, not used the scheme provided by Henderson and Nash but are recommending a scheme of five stages of endangerment to be measured by degree of proficiency in the IL in five age groups. This we regard as a refinement of Kinkade’s proposal. It also follows in the footsteps of schemes proposed by Stephen Wurm. Wurm, to whom this book is dedicated, was early on the scene, and proposed a five-fold classification scheme for the state of health of Australian indigenous languages, which he applied to the languages of New South Wales and South Australia (Wurm, 1963). His ‘ranking of languages’ was as follows: (1) extinct; (2) some, usually very old, individuals remember a little of the language, usually vocabulary; (3) a few, mostly very old, individuals can speak the language more or less fluently; (4) the language is still spoken but no longer in full tribal use; (5) the language is still in full tribal use.

78 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

77

This ranking is based on the two parameters of age group and fluency/ proficiency. Phrases such as ‘full tribal use’ may be difficult to interpret and operationalise. In 1996, Wurm produced the following categorisation of language endangerment, which is quite similar to those of Kinkade and of Australianist researchers such as Dixon and Schmidt: (1) (2) (3) (4) (5)

extinct language; moribund language; seriously endangered language; endangered language; autochthonous language not in danger.

Although this scheme is not operationalised in detail, at least in that source, Wurm suggests a definition of ‘endangerment’ whereby any language that is not learnt by 30% of the children of the community should be considered endangered (Wurm, 1996: 1). This accords with Drapeau’s (1998) suggestion that languages spoken by less than 30% of the 5–14 age group are in a ‘very poor condition’. We propose extending this criterion to all age groups as a basis for a five-fold classification below (Table 5.2). That is to say, in this table, in the case of the youngest age group able to speak the language, ‘speak’ means that 30% or more of the age group are able to speak the language, and ‘don’t speak’ means that less than 30% are able to speak the language. It is assumed, in the normal course of language shift, that the percentage speaking is likely to go down to zero or close to zero in groups even younger than the oldest group classified as ‘not speaking’, but this is not recommended at this stage as a necessary criterion. One anomaly of terminology is that, under this classification, if the usual terminology of ‘extinct’ were to be used for the last stage, a language with less than 30% of speakers in the 60+ age group would be classified as ‘extinct’, whereas this term is normally thought to apply to languages with no speakers at all. ‘Extinct’ and ‘dead’ are terms that annoy many Australian indigenous people whose heritage languages are seen to be in that condition. In order to cater for both these problems we suggest that we add

Table 5.2 McConvell and Thieberger’s proposed endangerment index (2001) Endangered Seriously (early stage) endangered

Critical

Terminal

don’t speak

don’t speak

Age

Strong

5–19

speak

don’t speak

don’t speak

20–39

speak

speak

don’t speak

don’t speak

don’t speak

40–59

speak

speak

speak

don’t speak

don’t speak

60+

speak

speak

speak

speak

don’t speak

79 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:35

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

78

the terms ‘critical’ and ‘terminal’ as alternatives to ‘near-extinct’ and ‘extinct’ respectively. Strict quantification is not necessary for use of this indicator: the classification may be applied, in a provisional way at least, by someone with knowledge of the linguistic situation in a community without carrying out a survey. However, it is necessary to formulate ways in which the indicator can be more rigorously applied when feasible. Among issues that arise in operationalising such an indicator are: (1) Does ‘speaking’ refer to ability or actual use? If the latter, when or where should the use be displayed? (2) What quantity of speaking a particular language in a multilingual environment counts as ‘speaking’? (3) What quality or proficiency of language counts as ‘speaking’? Commenting on these questions in turn: (1) For Australia at present, it may be more practical to adopt the census criterion of ‘speaking at home’. The data referred to in Table 5.2 cannot be recovered from the census for the reason already stated several times: we do not have a count of the whole of an age group identifying with a language (speakers + non-speakers). However, we can recover numbers of HL speakers for each language and age group and it may be possible by combining other information available to get close to reproducing the percentages relevant to the indicator. (2) Many Australian indigenous communities and individuals are highly multilingual and a single conversation may be carried on in a number of different varieties including English, creoles and a number of traditional languages as well as code-switching and mixing (McConvell, 1988). Under these conditions, it could be that the target IL may be spoken for only part of the time at home. It is unlikely that any researcher is going to measure the percentage but, as a guideline, we might use 30% or above as counting as ‘speaking’ at home. (3) Many classifications of language shift and endangerment do refer to degrees of proficiency but often only using vague terms like ‘fluently’. In the following section, we broach rather briefly the question of how to proceed if we wanted to gauge proficiency in an IL more accurately. From the point of view of simply matching a language situation to the suggested indicator, this is superfluous. However, if a criterion for speaking is needed that can be assessed roughly from casual observation, we suggest (based on a similar criterion used in the New Zealand census) ‘able to follow a conversation of normal difficulty on everyday topics and contribute coherently to it in fully formed sentences’. In Table 5.2, it is assumed that language shift is proceeding in a single

80 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

79

direction through age groups. The scheme can be adapted to cater for revitalisation, where young people start re-learning the language, but this is not discussed here (see McConvell & Thieberger, 2001).

Language Ecology One of the main reasons for tracking the state of indigenous languages is to understand their situations better so that appropriate help can be given to language speakers to maintain their languages where they want to do so. Nationwide surveys can, for instance, spot situations where a language is not being lost as fast as we might have expected under the circumstances. For instance, Arrernte is being maintained in Alice Springs, whereas all other ILs whose territory has been the site of a large town for 100 years or more are now extinct, or virtually so. This will allow us then to centre research on that language situation to find out what it is that is having this effect, whether it be an intervention programme or a combination of social circumstances. Lessons learned can be translated into trying to replicate the conditions that favour language maintenance elsewhere. Research of the type that can pinpoint conditions favouring language maintenance or language shift goes beyond just counting speakers of languages in different age groups, although this is a good start. What we are talking about is not ‘saving’ a language in isolation but maintaining and building a ‘language ecology’ in which the language can thrive. In all current situations in Australia today (as in the past), such an ecology will involve community bilingualism or multilingualism; the difference today is that English in some form will be part of that ecology. So ‘language maintenance’ might better be expressed as ‘maintaining bilingualism’ or ‘maintaining multilingualism’ in communities in which ILs have a role (McConvell, 1991). Determining the role and function of languages in a linguistic ecology is not just a question of slotting languages into ‘domains’ – while this may be useful, some language ecologies defy this type of analysis. The social meaning and value of codes in multilingual speech must also be examined. Nor should it be assumed that ‘diglossia’ is necessarily a positive sign for language maintenance nor lack of it negative. As Myers-Scotton has pointed out (1993), code-switching itself may be a positive force in maintaining bilingualism and multilingualism. A second point that is difficult to capture in the survey approach to shift and maintenance discussed in most of this chapter, but essential to examine, is that some languages are changing fast and in a radical fashion. The younger generation may not be shifting to speaking English or another known indigenous language of another group (assumed to be the ‘normal’ cases in Australia), but something is happening that points to disruption in

81 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

80

transmission patterns. Moreover, the emergence of a radically new variety can, in turn, cause disquiet, even alarm, among the speakers of the language. While the emerging new language may be considered to be ‘the same’ language as the one spoken by the older generation or a variety of it, the pair of varieties may in fact fail the ‘objective’ tests for languagehood usually used. Modern and Traditional Tiwi, spoken by adjacent generations of the same group, are said to be scarcely mutually intelligible (Lee, 1987). The problem raised by such rapid and radical change has sometimes been elided by placing such phenomena within a ‘language death’ analytical framework. Radical changes in languages are attributed to the endpoint – ‘death’ – towards which they are heading. This approach is fatally flawed theoretically by the use of a predicted endpoint as the guiding principle of analysis. There is no denying that some languages that undergo radical change also eventually die out but not all do: Tiwi, for instance, in some form, is still actively spoken by all generations today and does not appear to be on the verge of dying out. One case with which McConvell is familiar because he has worked with the group for many years on and off is that of the Gurindji. When he first saw the situation in the mid-1970s children appeared to be speaking Kriol, not Gurindji, and he thought that language shift would be completed within that generation. Later, however, McConvell observed that the same age cohort, as teenagers, had a passive competence in Gurindji and, on particular occasions for purposes of privacy or expression of social identity in a multi-ethnic environment, actively used full Gurindji. By 1988, when a study was carried out by Gurindji college students with McConvell, patterns of adult code-switching had been largely regularised in certain ways in the young people’s speech, which could be characterised as a mixed language derived from Gurindji and Kriol (McConvell, 1994; Dalton et al., 1995). In 2001, teenagers from the cohort who had been 5–6 years old at the time of the 1988 study continued to speak this mixed language. They were also able to (and on occasion did) produce full Gurindji in natural conversation with the correct morphology, which they seemed not to control when they were young children (McConvell, 2001b). How widespread this pattern is, whether it is due to a ‘grandmother effect’ only with certain individuals and whether it is strong enough to enable transmission to the next generation are matters for further research. For this chapter, however, the main point is that this kind of detail is not captured by approaches to language endangerment that simply ask whether or not a particular age group speaks the traditional indigenous language. Alongside ecological approaches that take account of the interactive functions of different languages, it is possible to carry out more fine-grained

82 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

81

testing procedures for indigenous language proficiency in different age groups, but this has not really been attempted in Australia (McConvell, 1994). It would be a good thing to carry out this kind of research in selected cases to compare the results with census and survey data to find where divergences in the two kinds of results occur (for instance, in the cases where apparent over-reporting of children’s ability occur, as noted above). The reasons for these anomalies can be tracked down and ways of fixing the problem examined. Such assessment can also assist people in knowing whether language maintenance programmes are actually doing their job in maintaining language proficiency. It can also feed, for instance, into developing staged goals for education programmes that include indigenous languages, like the 2000 Northern Territory Curriculum Framework.

Conclusions We have concluded that analysis of census data on indigenous languages in Australia has yielded, and has the potential to yield further, valuable facts about language endangerment. At the level of the nation and regions, percentages of people who speak an IL at home can be tracked to show up trends over time, and age group and gender data can assist here in isolating particular patterns of language shift. The overall trend, except for a few ‘strong language’ areas, is towards language shift away from ILs through time. This might lead to predictions of loss of most of the remaining languages by around the middle of the century, with the remaining currently ‘strong’ languages at risk of following them by the end of the century. There are a number of more hopeful signs to be read in the data, especially if we go beyond the basic census figures. The numbers of people speaking ILs is being maintained and is actually rising slightly, despite the drop in percentages, because of the high fertility of many groups. However, if those same groups lose the link of language transmission with old people, language shift can happen quickly. Language maintenance and language revival programmes are having some impact, and this is beginning to be visible in census figures. If these are given the chance to continue (and not be cut off as in the Northern Territory) there should be more substantial results to report, especially if more fine-grained studies of proficiency and language ecology are carried out to back up census studies. A third source of optimism is studies like that on the Gurindji reported above. These show that children and young people actually have a lot more ability in the traditional IL than is obvious from superficial observation and language shift is occurring at a much slower rate than might have been expected. The 1996 census has provided, for the first time, data on these individual languages. However, the fact that only one question is included (on HL),

83 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

82

without the potential of comparing this to the overall population identifying with the language, reduces the value of this census information for studying language endangerment. We would recommend to the Australian Bureau of Statistics that they consider adding another census question to improve the situation. This chapter has looked at ways of measuring language endangerment other than from just HL data on age groups. This attempt, however, ran into problems with the dubious nature of the data collected on some endangered ILs, apparently involving over-reporting of children’s use of ILs. At least a sample of these cases needs to be checked (by fieldwork preferably), both for accuracy and for the reasons for overreporting if any. Such validating exercises are best set in the context of a research project on the language ecology and proficiency in all languages spoken in the community. References Amery, R. (2001) Warrabarna Kaurna! Reclaiming an Australian Language. Lisse: Swets and Zeitlinger. ABS (1996) National Aboriginal and Torres Strait Islander Survey 1994: Detailed findings (ABS, 4190.0). Canberra: Australian Bureau of Statistics. ABS Website (2005) At http://www.abs.gov.au/Ausstats/[email protected]/0/caef79080 a31b757ca2569de002139b7?OpenDocument. Accessed 10.10.05. Cook, E-D. (1998) Aboriginal languages: History. In J. Edwards (ed.) Language in Canada. Cambridge: Cambridge University Press. Crawford, J. (2000) At War with Diversity: US Language Policy in an Age of Anxiety. Clevedon: Multilingual Matters. Crystal, D. (2000) Language Death. Cambridge: Cambridge University Press. Dalton, L., Edwards, S., Farquharson, R., Oscar, S. and McConvell, P. (1995) Gurindji children’s language and language maintenance. International Journal of the Sociology of Language 113, 83–96 Douglas, W.H. (1968) The Aboriginal Languages of South-West Australia: Speech Forms in Current Use and a Technical Description of Njungar. Canberra: AIAS. Drapeau, L. (1998) Aboriginal languages: Current status. In John Edwards (ed.) Language in Canada (pp. 144–159). Cambridge: Cambridge University Press. Environment Advisory Council (1996) Australia State of the Environment 1996. An independent report presented to the Commonwealth Minister for the Environment. Mimeograph. Also available online at http://www.deh.gov.au/ soe/soe96/index.html. Accessed 10.10.2005. Evans, N. (2001) The last speaker is dead, long live the last speaker! In P. Newman and M. Ratcliff (eds) Linguistic Fieldwork. Cambridge: Cambridge University Press. Fil, A. and Mühlhäusler, P. (eds) (2001) The Ecolinguistics Reader: Language, Ecology and Environment. London: Continuum. Fishman, J. (1991) Reversing Language Shift. Clevedon: Multilingual Matters. Garza-Cuaron, B. and Yolnda L. (1991) Endangered languages in Mexico. R.H. Robins and E.M. Uhlenbeck (eds) Endangered Languages (pp. 93–134). Oxford: Berg. Harmon, D. and Loh, J. (2002) Draft Framework for an Index of Biocultural Diversity. Washington, DC: Terralingua.

84 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Indigenous Language Endangerment in Australia

83

Henderson, J. and Nash, D. (1997) Culture and Heritage: Indigenous Languages. Australia: State of the Environment Technical Paper series (Natural and cultural heritage). Canberra: Department of the Environment. Hoogenraad, R. (1992) Preliminary Report on the Language Census of the Barkly and Sandover Regions. Darwin: Northern Territory Education Department. Hoogenraad, R. (2001) Operations South Aboriginal schools in language areas. Unpublished manuscript. Hosking, D.F., Lonsdale, T.J., Palmer, K., Troy, J.F. and Walsh, M.J. (2000) New South Wales Strategic Language Study. Canberra: AIATSIS. Kinkade, M.D. (1991) The decline of native languages in Canada. In R.H.Robins and E.M. Uhlenbeck (eds) Endangered Languages (pp. 157–176). Oxford: Berg. Krauss, M. (1992) The world’s languages in crisis. Language 68, 4–10. Lee, J. (1987) Tiwi Today: A Study of Language Change in a Contact Situation. Canberra: Pacific Linguistics. Lennon, J. (with contributions from Pearson et al.) (2001) 2001 State of the Environment Report: Natural and Cultural Heritage. Canberra: Environment Australia. LoBianco, J. and Rhydwen, M. (2001) Is the extinction of Australia’s indigenous languages inevitable? In J. Fishman (ed.) Can Threatened Languages be Saved? Reversing Language Shift Revisited: A Twenty-First Century Perspective (pp. 391– 422). Clevedon: Multilingual Matters. Maffi, L. (1998) Language: A resource for nature. Nature and Resources: The UNESCO Journal on the Environment and Natural Resources Research 34 (4), 12–21. Maffi, L. (ed.) (2001) On Biocultural Diversity: Linking Language, Knowledge, and the Environment. Washington DC: Smithsonian Institution Press. McConvell, P. (1986) Aboriginal language programs and language maintenance in the Kimberleys. Australian Review of Applied Linguistics S3, 108–121. McConvell, P. (1988) Mix-im-up: Aboriginal codeswitching, old and new. In M. Heller (ed.) Codeswitching: Anthropological and Sociolinguistic Perspectives (pp. 97– 124). Berlin: Mouton de Gruyter. McConvell, P. (1991) Understanding language shift: A first step to language maintenance. In S. Romaine (ed.) Language in Australia (pp. 143–155). Cambridge: Cambridge University Press. McConvell, Patrick (1992) Review of Fishman 1991. Australian Journal of Linguistics 12 (1), 209–220. McConvell, P. (1994) Oral proficiency assessment for Aboriginal languages. In J. Henderson and D. Hartman (eds) Aboriginal languages in Education (pp. 301–315). Alice Springs: IAD Press. McConvell, P. (2001a) Looking for the two-way street: Indigenous Australians battle to keep their languages strong. Cultural Survival Quarterly Summer, 18–21. McConvell, P. (2001b) Mix-im-up speech and emergent mixed languages in Aboriginal Australia. Proceedings of SALSA 2001. Texas Linguistic Forum 44 (1– 2), 328 – 349. McConvell, P. and Thieberger, N. (2001) The state of indigenous languages in Australia. Report to Environment Australia (downloadable from WWW at www.deh.gov.au/soe/). McConvell, P., 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.

85 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

84

McKay, G. (1996) The Land Still Speaks: Review of Aboriginal and Torres Strait Islander Language Maintenance and Development Needs and Activities. National Board of Employment, Education and Training, No. 44. Canberra: Australian Government Publishing Service. Myers-Scotton, C. (1993) Duelling Languages: Grammatical Structure in Code Switching. Oxford/New York: Clarendon Press/Oxford University Press. Nettle, D. and Romaine, S. (2000) Vanishing Voices: The Extinction of the World’s Languages. Oxford: Oxford University Press. Nicholls, C. (c. 2001) The fate of bilingual education in the Northern Territory. AIATSIS Discussion Paper (unpublished manuscript). Norris, M.J. (1998) Canada’s aboriginal languages. Canadian Social Trends 51 (Winter), 8–16. Ross, K.E. (1999) Population Issues, Indigenous Australians,1996. Occasional paper (ABS 4708). Canberra: Australian Bureau of Statistics. Schmidt, A. (1990) The Loss of Australia’s Aboriginal Language Heritage. Canberra: AIATSIS. Te Puni Kokiri, Maori Language Commission, Statistics New Zealand (1998) The National Maori Language Survey:Te Mahi Rangahau Reo Maori. Wellington: Te Puni Kokiri, Maori Language Commission, Statistics New Zealand. Thieberger, N. (2001) Extinction in whose terms? Which parts of a language constitute a target for language maintenance programmes? In D. Bradley and M. Bradley (eds) Language Maintenance for Endangered Languages: An Active Approach. London: Curzon Press. Trudgill, P. (1983) On Dialect: Social and Geographical Perspectives. Oxford: Blackwell. Von Brandenstein, C.G. (1988) Nyungar Anew: Phonology, Text Samples and Etymological and Historical 1500-Word Vocabulary of an Artificially Re-created Aboriginal Language in the South-West of Australia. Canberra: Pacific Linguistics. Wurm, S.A. (1963) Aboriginal languages. In W.E.H. Stanner and H. Sheils (eds) Australian Aboriginal Studies (pp. 127–148). Melbourne: Oxford University Press. Wurm, S.A. (1996) Atlas of the World’s Languages in Danger of Disappearing. Paris/ Canberra: UNESCO Publishing/Pacific Linguistics.

86 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Chapter 6

Papua New Guinea’s Languages: Will They Survive? KENNETH M. SUMBUK

Introduction In the recent past, many scholars, both linguists and non-linguists, have widely commented on the survival and the non-survival of languages in general and specifically those of Papua New Guinea. It has been stated time and time again that many of the indigenous languages in Papua New Guinea will die out in similar fashion to those in Australia and in other linguistic regions. Many commenting on the chances of the Papua New Guinean languages surviving have been mostly influenced by the number of speakers the languages have since many have fairly small numbers of speakers, and this factor has been one of the main reasons, but not the only one, for the extinction of many small languages elsewhere. The recent substantial increase in scholarship on the fate of the world’s linguistic diversity demonstrates the concern that many people have over the linguistic resources of the world. Many have warned of large-scale extinction of languages in the near future. One such recent warning comes from Nettle and Romaine (2000: 2), who warn that ‘the trickle of (language) extinction of the last few centuries is now turning into a flood’. Others who have expressed similar sentiments include Crystal (2000) and Mühlhäusler (1999, see also his chapter in this volume). Crystal’s work takes more of a global perspective, while Mühlhäusler focuses largely on the Pacific region. Crowley (2000, 1999 and 1998) has also made telling observations about the future of the languages in the Pacific region. In the current concern over language death, the Pacific region is rightly receiving a lot of attention. That region boasts of being the most linguistically diverse region in the world with an estimated 1200 languages. A very large part of this diversity is in the Melanesian countries of Solomon Islands, Vanuatu and Papua New Guinea. In light of what has been observed in other linguistic regions, many of the linguists working in the Pacific region ask whether the 1200 or so indigenous languages spoken in the region will survive. This question is especially important because many of the Pacific languages have fairly small numbers of speakers in compar-

85

87 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

86

ison with the number of people who speak a single language in, say, Asia or Europe. If Krauss’ (1992) figure of 100,000 speakers is the safety figure for language survival, then most of the languages of the Pacific have no chance at all of surviving. Nevertheless, many of these small languages have coexisted harmoniously for a fairly long time. Thus, though small numbers of speakers may be seen as a sure sign of language death in other linguistic regions, this is not necessarily so in the Pacific region. In this chapter, the question of language survival focuses specifically on the languages of Papua New Guinea, where the number of speakers of a distinct language can be as low as 50. Papua New Guinea has long been known as the most linguistically diverse country in the world: with 860 or so distinct languages, Papua New Guinea boasts of having more languages spoken within its borders than any other single country in the world. This figure does not include the languages spoken in West Papua,1 the western half of the island of New Guinea, still under Indonesian rule.

The Present Linguistic Situation in Papua New Guinea According to the 2000 national census figures, the overall population of Papua New Guinea has rapidly increased from 3 million to 5 million. This increase was observed in both the rural and urban areas. Thus, generally speaking, the number of speakers of many of the 860 or so indigenous languages has increased. However, it is difficult to tell what is happening with individual languages because the census questionnaires have not included a question seeking to establish people’s first language. Attempts had been made to have a question to establish people’s first language included, but the census officials have refused. The absence of a national linguistic survey also contributes to this uncertainty about the state of the languages in Papua New Guinea. It is hard to tell the actual number of speakers for each language or even the actual number of languages spoken within the borders of Papua New Guinea. The census figure indicates only the number of people counted in the villages during the time of conducting the census, but does not indicate the number of people who live outside their ancestral homes. This, in itself, is a very poor way of estimating the actual number of speakers. For instance, a significant number of speakers of Sare2 (my own mother tongue) left the villages for towns or other rural areas before and during the census period and were simply not counted as speakers of Sare. Also, many village people simply do not show interest in such national events as the census, do not take part, and are not counted. Considering these factors, the figures that are used in the existing literature to forecast the future of languages in Papua New Guinea are not very accurate representations of the state of these languages. Data provided by individual researchers are more reliable

88 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Papua New Guinea Languages: Will They Survive?

87

when forecasting the future of these languages since the communities they work in are small and it is easy to work out the number of speakers of a language simply by counting them. With the inaccurate figures from the national census and in the absence of a national linguistic survey in Papua New Guinea, the true nature of the future of the country’s languages is not known. Most of what is asserted about Papua New Guinea’s linguistic resources is based on experiences from other countries such as Australia or from research carried out in the 1960s and the 1970s. The actual current linguistic trends in the country are yet to be established but this is not to say that the warnings of the possible extinction of many of these small languages should be ignored. On the contrary, research on individual languages shows that the linguistic prognosis warning of massive language extinction is already showing telltale signs in Papua New Guinea, and these warnings must be taken very seriously both by the government and by the speakers of these languages. If any language in Papua New Guinea is in imminent danger of dying out, then it is one of those in Table 6.1. These are languages with 100 or fewer speakers and, according to the UNESCO endangered language report (see Barreña et al. in this volume), any language that has 100 or fewer speakers is in danger of dying out. However, most of the languages listed in Table 6.1 have survived with about the same number of speakers for hundreds of years. Thus, the number of speakers of a language does not tell us much about the language’s future survival. The languages in the table that may have already died out or are in real danger of doing so are Guranalum, Lae, Laua, Malkolkol, Ouma, Sene and, possibly, Hermit and Kamasa. If we were to go by Krauss’ figure of 100,000 as the measure of safety, then more than 90% of Papua New Guinea’s languages are in danger or would be extinct by now. In the absence of any work on these small languages, it is very difficult to confidently comment on their state. Of these languages at risk, only three are known to the present writer to be being worked on currently: Taiap3 (by Don Kulick), Tench4 (by Dicks Thomas), and Makolkol5 (by Anton Neinkama). All three are reported to be close to extinction, with Tench being the most likely to be the first to disappear through language shift. According to Thomas (personal communication), the Tench speakers are abandoning the language for the more popular Musau language. Neinkama (personal communication) is currently carrying out research on Malkolkol, which he reports to have 111 speakers, which does not match the figure in Table 6.1 derived from other studies. Neinkama does, however, admit that only 16 speakers can be regarded as truly fluent while the rest are shifting to the neighbouring dominant Nakanai language.

89 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

88

Table 6.1 Languages with 100 or fewer speakers Language name

Number of speakers

Province

Ainba

100

West Sepik

Ak

83

West Sepik

Arawun

73

Madang

80–100

Western

Ari Bepour

57

Madang

Bikaru

100

East Sepik

Bilakura

34

Madang

Dorro

80

Western

Dumun

42

Madang

Faita

57

Madang

Gorovu

50

Guranalum

3 or 4

East Sepik New Ireland

Hermit

20

Iteri

90

Manus West Sepik

Kalamo

100

Western

Kamasa

20

Morobe

Kawucha

30

Morobe

Kowaki

31

Madang

Lae

1

Morobe

Laua

1

Central

Makolkol

7

East New Britain

Mawak

31

Madang

Mindiri

93

Madang

Moere

56

Madang

Mosimo

58

Madang

Musan

75

West Sepik

Ouma

4

Central

Papi

75

West Sepik

Piame

100

West Sepik

Pyu

100

Irian Jaya

Samosa

94

Madang

under 10

Morobe

Sene Sumariup

65

East Sepik

Taiap

89

East Sepik

Tench

49

New Island

Turaka

35

Milne Bay

Usu

93

Madang

Vehes

100

Morobe

Yapunda

69

West Sepik

Yarawata

98

Madang

Data from Grimes, 2000

90 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Papua New Guinea Languages: Will They Survive?

89

The Specific Case of the Kaningara and Sare Languages Kaningara and Sare are both Papuan languages spoken in the same Blackwater Lake area of the East Sepik Province. These languages are members of the Sepik Hill Family of the Sepik-SubPhylum. There are altogether 327 people in only one village of the same name that speak the Kaningara language, while 1339 people in seven villages speak Sare. Geographically, these languages pose an interesting scenario in that the Kaningara village is located in the midst of the Sare-speaking villages, in fact, between two Sare-speaking villages along the Blackwater River, a tributary of the Sepik River. With this bizarre location of two distinct indigenous languages, one would have expected some form of language shift, death or suicide. These languages have, however, survived side by side and have not given an inch to each other. The bigger Sare language has not absorbed the smaller Kaningara language, nor have the speakers of the smaller Kaningara language shifted to Sare, as we see in the cases of Tench and Malkolkol. Many older Kaningara speakers are bilingual in Sare, which is the medium of communication when speakers of these languages come into contact but the use of Sare is restricted to contact situations with Sare speakers and not among the Kaningara speakers themselves. Even in speech, despite some shared lexical items, very rarely do Kaningara speakers incorporate Sare lexical items. Table 6.2 shows the lexical items that indicate the differences between these languages. There are two main reasons why these languages have sustained themselves for hundreds of years. Firstly, the speakers of these languages maintain and practise basically the same culture. They have very similar beliefs, customs, initiations and ceremonies, and thus there is no need for one to adopt the other’s culture and language since, in most cases of language shift, speakers of the minority language normally adopt the language and Table 6.2 Some lexical differences between the Sare and Kaningara languages Sare Yima wuiagu Tugani Rari Motu Togu Tuwam Wombugedza

Kaningara nimar ramboromu tanmiatu yanomu yanomu ygat tunbu apugat

Glossary man husband wife son daughter head hair ear

91 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

90

culture of the dominant group. In this case, there was no need for the speakers of the smaller Kaningara language to adopt the culture, when it was basically the same. What we have is a situation where one culture is being practised using two distinct languages. Thus, one culture has managed to sustain two languages over a fairly long period of time. This, however, goes against the norm that linguistic diversity reflects cultural diversity, since what is observed here is linguistic diversity but not cultural diversity. However, what this case does support is that language death is symptomatic of cultural death, i.e. the disappearance of a way of life results in language death but, here, the culture has not disappeared and therefore the two languages have not disappeared. Secondly, these two languages continue to perform the cultural functions that they have been performing with no immediate sign of either one losing any ground to the other. Most importantly, these two languages are being transmitted continuously to their respective younger generations, a practice guaranteeing their continued coexistence. The only noticeable change that has been observed for these languages, like many of the Sepik and Madang languages, is the use of Tok Pisin. The case for Sare has been presented in Sumbuk (1992). All Sare and Kaningara speakers are bilingual in Tok Pisin and there is now a noticeable shift towards Tok Pisin. Children of both endogamous and exogamous marriages in the villages are acquiring Tok Pisin as well as the two indigenous languages. What is not quite clear is whether the acquisition of Tok Pisin and these two languages is simultaneous, or whether the acquisition of the local languages precedes that of Tok Pisin or vice versa. It is clear, however, that Tok Pisin has not replaced these languages in performing the village-based activities, even among the young people. Tok Pisin and the local languages are used interchangeably for general conversation but, when it comes to culturally oriented activities, the local languages are the medium of communication. One can only hope that this scenario lasts.

The Future of Papua New Guinea Languages Factors affecting the survival of languages Very many factors are cited as affecting the survival of languages in general. In some countries, only one or two major factors have been identified as the main cause while in others, a number of factors combine to cause the extinction of languages. Some of the most cited factors include: (1) technology (Cunningham, in this volume); (2) the lack of documentation of mostly indigenous languages (Crystal, 2000: 152–154);

92 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Papua New Guinea Languages: Will They Survive?

91

(3) economic globalisation (Nettle & Romaine, 2000: 126–149); (4) the lack of education in these languages (Ingram, in this volume); and (5) the lack of socio-political and economic rights amongst indigenous peoples (Crowley, 1998). The literature has also pointed to many other factors as further contributing to the demise of the world’s linguistic resources (see Crystal, 2000; Graddol, 1997; Krauss, 1992). In most cases, it is a combination of factors that has led to the extinction of languages, with one or two being the dominant causes. In the case of Australia, both Cunningham and Ingram (in this volume) refer to how English has dominated at the expense of indigenous Aboriginal languages and the immigrant minority languages such as Chinese and Italian. Cunningham sounds a warning that technological developments should be multilingual and not monolingual in English, as is currently the case. For more developed languages like German, French, or Japanese, technological development is already in sight. However, for indigenous languages, this is a task that may eventuate only if there is a change of government policy and attitude. As it stands now, technologically the future of indigenous languages looks bleak, while the lack of a National Language Institute in Australia also aggravates the situation (see Ingram, in this volume). Ingram also argues that language education can contribute to the maintenance and development of a nation’s language resources. More than half of Australia’s indigenous languages became extinct as a direct result of economic globalisation, to which Nettle and Romaine (2000) refer. This is because European colonisation in the Americas and Australia did not respect, nor even consider, the socio-political rights of the colonised peoples. This resulted in the mass eradication of hundreds of indigenous peoples, their languages and their cultures. The linguistic, cultural and demographic casualties of indigenous peoples as experienced in the Americas and Australia were not experienced in Papua New Guinea. This is because economic globalisation did not take place there on a scale similar to that of the Americas and Australia and, consequently, the historical events and the current linguistic observations in the Americas and Australia cannot possibly be compared to the linguistic situation in Papua New Guinea. This does not mean to say that there are no signs of languages dying in Papua New Guinea, but the factors and the reasons for their death are different from those observed in other linguistic regions such as the Americas and Australia. The next section will attempt to discuss the popularly-cited factors for language death, and consider them in the context of Papua New Guinean languages.

93 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

92

Technology In the age of technological evolution, international lingua francas such as English will continue to be the dominant threat to the global linguistic wealth. In the case of English, Graddol (1997: 58) has observed that the dominance of English will only increase and consolidate. Few languages have the chance of breaking the stranglehold that English has on technological evolution. The languages that have a better chance to do so are those spoken in technologically more advanced countries such as Japan, Germany, France and even China. In light of this observation, Papua New Guinean languages have very little chance of breaking the stranglehold of the English language in the evolution of technology for several reasons. Papua New Guinea is not a technologically advanced country, many Papua New Guinean languages have no alphabets, there are no written records and, most importantly, the country does not have the economic might to have all its 860 languages technologically developed. Thus, multilingual technological advances are not a realistic option in the fight to preserve the linguistic resources of Papua New Guinea. Lack of documentation For a language to have any chance at all to be considered for use in technology, education or economics, it must first be documented and described. So far, only about 10% of Papua New Guinea’s languages have been comprehensively analysed and documented. Most of those have been documented by members of the Papua New Guinea branch of the Summer Institute of Linguistics and the rest have been documented by individual researchers from Australian universities or from the Linguistics Department of the University of Papua New Guinea. It is now known that many languages in Papua New Guinea have become extinct without any form of documentation. No one will know what these languages were like, their grammatical systems, sound systems, and semantic and sociolinguistic patterns. We can only guess what they were like from the neighbouring languages and cannot know their own intrinsic systems. Documentation of many of the small languages of Papua New Guinea is essential as this is the most immediate way to preserve them, at least for a short while, and so that there will, at least, be something in writing about the languages if they should die out. Economic globalisation The wave of colonisation and the economic globalisation that accompanied it reached the shores of New Guinea much later than the Americas and Australia. When it did eventually reach New Guinea, it was not on a

94 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Papua New Guinea Languages: Will They Survive?

93

massive or a lasting scale: the colonial masters left the country after it gained independence from Great Britain through Australia. In the mid1970s, the indigenous people took charge of the governance and decisionmaking processes of their own country and are now creating their own destiny. Thus, the effects of economic globalisation did not directly affect the speakers of the many languages in Papua New Guinea. However, it has now become evident that speakers of many of these small languages, especially in the Sepik-Madang area, are beginning to abandon their languages for Tok Pisin, a language seen by many as the symbol of modernisation and a means of attaining economic success. This assumption stems from the fact that Tok Pisin is an English-based pidgin, which was first used by indented plantation labourers, who brought the language to the villages and were seen as more advanced and modernised. This assumption has stayed on and is still prevalent today in many of the indigenous Papua New Guinean societies. Lack of education in indigenous languages Officially, Papua New Guinea does not promote or discriminate against any language. The national constitution recognises all languages as equal and important, including English and the two pidgin languages of Tok Pisin and Hiri Motu. Even the National Literacy Policy adopted in 2000 recognises all indigenous languages as media of instruction at lower levels in the schools. However, in practice, it is English that is promoted as the language of education, governance, business and media. It is this language that receives government funding for teacher training, for production of teaching materials, and for the salaries of English teachers. In the last decade there was a wave of literacy awareness campaigns in the country. These led to the setting up of literacy programmes throughout the country using indigenous languages, which came under the brand name of ‘vernacular education’. When the idea was initially introduced, there was a lot of excitement, as is the case for many newly introduced ideas or concepts. This excitement has, however, since dwindled. These vernacular schools were community-based, and it was expected that support for them would come from the local community. However, with most local communities being cash-stricken, community support was always going to be a problem in the absence of government support and, consequently, most of these vernacular schools have ceased to exist. This current scenario in Papua New Guinea is unlikely to change, mostly for economic and political reasons. Economically, it is cheaper to use English since the materials are already produced in English and there are teachers trained in English. It would be too expensive to use one or a couple of the indigenous languages. For this to happen, the language or languages

95 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:36

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

94

must firstly be developed, materials produced, and teachers trained, a process that would be very costly for the government. Politically, it is unwise to develop any one indigenous language in a linguistically diverse country such as Papua New Guinea. Indeed, it would be political and linguistic suicide to do so. It is, therefore, economically cheap and politically safe to continue using a neutral and already developed language such as English. Thus, the possibility of indigenous Papua New Guinean languages being used in education is very slight and this is very much to the disadvantage of many of these small languages. The lack of socio-political and economic rights of the indigenous peoples The social, political and economic take-over that we know happened in other Pacific societies such as Australia, New Zealand and Hawaii did not take place in Papua New Guinea. In these societies, the indigenous cultures and languages were victims of white European economic and political dominance. The indigenous people lost their cultures and languages because the European settlers imposed theirs and, at the same time, deprived the indigenous people of their social, political and economic rights. Papua New Guineans can count themselves lucky in this sense since the European settlers were not as forceful in depriving the indigenous people of their socio-economic and political rights. This may be attributed to the fact that European settlement in Papua New Guinea was not as permanent as in other societies. However, what has been observed in other societies is now also being observed in a number of Papua New Guinean societies. These observed changes are not from domination but from forces within the Papua New Guinean societies themselves. The commonly cited example of Taiap (Kulick, 1992) and that reported by the author for Sare (Sumbuk, 1992) exemplifies a situation where the language shift is for reasons other than European dominance. The local people find Tok Pisin fascinating because they associate it with English, which they assume gives them socio-economic and political benefit. Tok Pisin for many people, then, is a symbol of socio-economic and political advancement. It is obvious to say that, once the indigenous people lose their social, political and economic rights, the chance of their also losing their culture and language is much increased. This is already an observable trend in Papua New Guinea and it is, therefore, important to safeguard the many small languages before it is too late.

96 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:37

Color profile: Disabled Composite Default screen

Papua New Guinea Languages: Will They Survive?

95

Recommendations In order to try to safeguard many of the small languages in Papua New Guinea from extinction, a number of actions should be taken, including the following: (1) There is need for a nationwide linguistic survey of Papua New Guinea or, better still, of the whole island of New Guinea. This survey should focus on specific linguistic information, including clearly establishing such information as: • the actual number of languages; • the number of speakers of each of the languages; and • whether the speakers are naturally transmitting their languages to the younger generations or abandoning them for other languages, and for what reasons. (2) Both the people and the government need to be educated about the importance of their linguistic resources: they must be made to understand that this rich resource is in danger of dying out if it is not protected. (3) The speakers themselves and their government must decide whether or not they want to preserve these unique linguistic resources. Imposing the idea of language preservation on the speakers from outside may not work if they do not see the importance of such an exercise, and it will simply be a waste of time.

Conclusion Due to the immense linguistic diversity of Papua New Guinea and the small number of speakers of many of these languages, it is a widely held view that it won’t be long before all the languages will die out. However, many of these small languages have survived for hundreds of years with about the same number of speakers. The causes of language death in other linguistic regions may not necessarily be the same for the languages in Papua New Guinea. What is needed, then, is a thorough linguistic survey of Papua New Guinean languages to clearly establish why these languages have survived with so few speakers. Such a survey must also try to establish the most likely causes of language death in such a diverse linguistic region as the island of New Guinea and the adjacent island groups. Notes 1. West Papua, formally also known as Irian Jaya or West Irian, was that western part of the Island of New Guinea that was under the colonial rule of the Dutch and was subsequently claimed by Indonesia upon its independence.

97 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:37

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

96

2. Refer to the following section for a brief comparative discussion of Sare with its neighbouring Kaningara language. 3. Taiap is a Sepik-Ramu language spoken only in the Gapun village, where it was thought, in 1987, to have 89 speakers. It was observed that there are no speakers younger than 10 years old speaking the language. It is very near extinction if the reported trend of language shift continues. 4. Tench, also known as Tenis, is a member of the St. Matthias subfamily of the New Guinea Oceanic Subgroup of Austronesian languages (Pawley). The language is spoken by 49 speakers on Tench Island in the New Ireland province. 5. Makolkol is an East Papuan language spoken in the Baining-Taulil area of the East New Britain province. The language is nearly extinct.

References Crowley, T. (1998) How many languages will survive in the Pacific? Te Reo (Journal of Linguistic Society of New Zealand) 41, 116–125. Crowley, T. (1999) Linguistic diversity in the Pacific. Journal of Sociolinguistics 3 (1), 81–103. Crowley, T. (2000) The language situation in Vanuatu. Current Issues in Language Planning 1 (1), 47–132. Crystal, D. (2000) Language Death. Cambridge: Cambridge University Press. Graddol, D. (1997) The Future of English? London: British Council. Grimes, B.F. (ed.) (2000) Ethnologue (Vol. 1): Languages of the World (14th edn). Dallas, TX: SIL International. Krauss, M. (1992) The world’s languages in crisis. Language 68, 4–10. Kulick, D. (1992) Language Shift and Cultural Reproduction: Socialisation, Self and Syncretism in a Papua New Guinean Village. Cambridge: Cambridge University Press. Mühlhäusler, P. (1999) Linguistic Ecology: Language Change and Linguistic Imperialism in the Pacific Region. London: Routledge. Nettle, D. and Romaine, S. (2000) Vanishing Voices: The Extinction of the World’s Languages. Oxford: Oxford University Press. Sumbuk, K. (1992) Is Tok Pisin a threat to Sare? In F. Byrne and J.A. Holm (eds) Atlantic Meets Pacific. A Global View of Pidginization and Creolization (pp. 309–317) Amsterdam: John Benjamins Publishing Company.

98 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:37

Color profile: Disabled Composite Default screen

Chapter 7

Language Endangerment and Globalisation in the Pacific DARRELL TRYON

Introduction The Pacific region encompassed in this chapter covers all of what is commonly known as Oceania, namely the area extending from the great island of New Guinea at its western extremity eastwards through Island Melanesia, Micronesia and Polynesia as far as Easter Island (Figure 7.1). The area includes all of the languages of the island of New Guinea (West Irian and Papua New Guinea), the Solomon Islands, Vanuatu, New Caledonia and Fiji, in Melanesia; the languages of the Marshall Islands, the Federated States of Micronesia (the Caroline and Mariana Islands, and Yap), together with Palau (Belau) and Guam; the languages of Polynesia (Tonga, Niue, Samoa, Wallis and Futuna, the Cook Islands, New Zealand, French Polynesia, Hawaii and Easter Island). The Pacific has perhaps the greatest linguistic diversity in the world today, with approximately 1200 distinct languages, spoken generally by very small communities, the great majority of which are egalitarian societies without hereditary chiefly structures, especially in Melanesia, where the vast majority of the Oceanic population resides. In these societies, communities are headed by a big-man or entrepreneur who has become the leader of his society, basically through his fellow tribesmen’s indebtedness to him as a result of gift-giving, largely pig-sacrifice and feast-giving. The Melanesian big-man generally rules over only a couple of thousand, very often only a few hundred, souls. This fact, together with the absence of any hereditary principle of transmission of power, has resulted in highly fragmented societies in this area, most speaking their own language or dialect, which is used as a kind of ethnic badge or identity marker. In Polynesia and most of Micronesia, there are relatively large political and social units, governed traditionally by a paramount chief or even a king or queen.

97

99 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:37

W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Su

m

ra

Java

Guam

Marianas

100

Figure 7.1 Languages of the Pacific

NEW ZEALAND

New Caledonia

Fiji

Rotuma

Vanuatu Niue

Samoa Co ok

Hawaii

Isl ds

an

Easter Island

Papuan language(s) in otherwise Austronesian areas

Overwhelmingly Papuan

Austronesian

Tuamoto Archipelago

1000 miles

1000km

(© Cartography ANU 04-006_T2)

Rapa

Tahiti

0 0

Marquesas FRENCH POLYNESIA

Raratonga

Tokelau

Tonga

Wallis & Futuna

Tuvalu

Kiribatu

Marshall Islands

OCEANIC AUSTRONESIA

Federated States of Micronesia PAPUA NEW Nauru GUINEA W. Papua Solomon Islands

Yap

AUSTRALIA

E.TIMOR

Moluccas

Sulawesi

Palau

PHILIPPINES

Taiwan

98

Madagascar

NON-OCEANIC AUSTRONESIA

at

INDONESIA

Borneo

CHINA

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

Color profile: Disabled Composite Default screen

Language Endangerment and Globalisation in the Pacific

99

The Languages of the Pacific There are two major and quite unrelated language families represented in the Pacific. These are: • the Austronesian family (550 languages in Oceania); • the Papuan family (750 languages). In addition to these, there are a number of pidgin and creole languages (Tok Pisin, Solomons Pijin, Bislama and Hawaiian Pidgin) used as lingua francas, along with the colonial languages English and French and, to a much lesser extent, Spanish. The Austronesian family, also known as Malayo-Polynesian, is a vast language family with almost 1000 languages, extending from Singapore in South-East Asia to Easter Island in the eastern Pacific. It includes all of the languages of the Philippines, nearly all of the languages of Indonesia, and the indigenous languages of Taiwan and Madagascar. In mainland SouthEast Asia, Austronesian languages are spoken in Singapore and Malaysia, as well as in parts of Vietnam and Cambodia. Austronesian languages are also distributed around the coastal areas of New Guinea and its offshore islands, right down the Melanesian chain and out into Polynesia and Micronesia. The Austronesian language family is well established (Tryon, 1995), and is considered to have its original homeland on the island of Taiwan, just off the coast of mainland China, where it must ultimately have originated. In terms of relative chronology, the Austronesians are believed to have settled in Taiwan about 6000 years ago (Bellwood, 1989), and to have moved south from there, into the Philippines and Indonesia, expanding and dispersing following the development of agricultural techniques. From south-eastern Indonesia, they moved along the north coast of New Guinea, settling in the New Britain–New Ireland area about 4000 years ago, before moving further south and east to colonise the whole of the South Pacific. The Austronesian family may be represented in a family tree (Table 7.1), showing the major subgroupings. All the Austronesian languages east of a north–south line drawn about 130o east longitude are members of a single lower-order subgroup known as the Oceanic subgroup, consisting of nearly 500 distinct languages (see Figure 7.1). Approximately 250 of these languages are spoken in the greater New Guinea area, nearly always in coastal areas and on offshore islands, in contrast with the Papuan languages, which are nearly all spoken in the mountainous interior of the island. The Oceanic subgroup of Austronesian, then, consists of three major groupings, the Admiralty Islands group, the Western Oceanic group and the Central-Eastern Oceanic group:

101 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

100

Table 7.1 The Austronesian languages Austronesian Taiwan (22 languages)

Malayo-Peninsula (approx. 1000 languages) Western MalayoPolynesian (530 languages) [Philippines] [Indonesia] [Malaysia] [Singapore] [Madagascar]

Central-Eastern Malayo-Polynesian

Central Malayo- Eastern MalayoPolynesian Polynesian (156 languages) [Eastern Indonesia] SouthHalmaheraWest New Guinea (47 languages)

Oceanic (470 languages)

Oceanic Admiralty Islands [Manus]

Western Oceanic [New Guinea] [Western Solomons]

Central-Eastern Oceanic [South-East Solomons] [Vanuatu] [New Caledonia] [Fiji] [Polynesia] [Micronesia]

• The Admiralties Island group consists of all of the languages of the Admiralty Islands, to the north of New Guinea, and the Western Isles. • The Western Oceanic group is a large group, consisting of three clusters: the North New Guinea cluster, the Papuan Tip cluster and the Meso-Melanesian cluster (Ross, 1988). • The Meso-Melanesian cluster consists of the languages of New Britain and New Ireland, Bougainville, and the western Solomon Islands as far south-east as Santa Isabel.

W:\Clients\MLM\Cunningham\Proof2\Cunningham Proof 2a 31 January 2006 20:38:07

Color profile: Disabled Composite Default screen

Language Endangerment and Globalisation in the Pacific

101

Table 7.2 The Central-Eastern Oceanic subgroup of languages Central-Eastern Oceanic South-East Solomonic [South-East Solomons] [Vanuatu] [New Caledonia] [Micronesia]

Remote Oceanic Central Pacific Rotuma West Fijian East Fijian

Tokalau Fijian Polynesian

The Central-Eastern Oceanic subgroup, which consists of the remainder of the indigenous languages of the south and central Pacific, may be represented as shown in Table 7.2. The Papuan family languages number approximately 750. They are spoken chiefly in the interior of the great island of New Guinea, along the great central mountain chain. To the west, they are also found on the Indonesian islands of Pantar and Alor, and in the northern half of Halmahera. They are also spoken in parts of West Timor and the newly-independent East Timor. To the east of the island of New Guinea, Papuan languages are also found in the Bismarck Archipelago (New Britain and New Ireland), on parts of Bougainville, and scattered through the Solomon Islands as far south-east as Santa Cruz. The Papuan languages are quite unrelated to the Austronesian family of languages. Indeed, the archaeological record (Golson, 1991; Swadling & Muke, 1998; Spriggs, 1997) shows that the Papuan-speaking populations are much more ancient than the Austronesians, with sites dated as far back as 50,000 years in Papua New Guinea. When the first Austronesians arrived in the region, they frequently encountered Papuan populations established for thousands of years before their arrival. In New Ireland, for example, there is only one Papuan language, Kuot, remaining today, surrounded by Austronesian languages. The earliest archaeological date recorded for New Ireland is 32,000 years (Spriggs, 1997), while the earliest date for Austronesian settlement on that island is only 4000 years before the present. The Papuan languages are not nearly as well known and described as the Austronesian languages. Their genetic relationships are only partially established at present. Wurm (1975) proposed the existence of a very large grouping of Papuan languages, some 500 languages, which he termed the Trans New Guinea Phylum. In his later study, Wurm (1982) also proposed a number of other phyla or broad subgroups, as follows: (1) Trans New Guinea phylum (507 languages) (these languages extend right across the island of New Guinea from east to west);. (2) West Papuan phylum (24 languages);

103 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

102

(3) (4) (5) (6) (7)

Sepik-Ramu phylum (98 languages); Torricelli phylum (48 languages); East Papuan phylum (27 languages); minor phyla (29 languages); phylum-level isolates (8 languages).

Wurm‘s classification of the Papuan languages was based largely on inspection rather than the rigorous application of the comparative method, which requires the establishment of sets of regular sound correspondences. For this reason, Wurm’s ideas did not win universal acceptance, although more recent studies ( see below) have to a large extent vindicated his claims, especially with reference to the Trans New Guinea phylum. Foley (1986) takes a much more conservative approach, not accepting any of the major groupings advanced by Wurm as proven. Rather, he suspends judgement on the Trans New Guinea phylum and limits himself to recognising some 60 smaller lower-level groupings of Papuan languages which do constitute demonstrable genetic units. He recognises the following major Papuan language families: (1) (3) (5) (7) (9) (11) (13) (15) (17) (19) (21) (23) (25)

Asmat family, Marind family, Suki-Gogodala family, Goilalan family, Binanderean family, Kainantu family, Chimbu family, Ok family, Wissel Lakes family, Sko family, Ndu family, Grass family, Huon family,

(2) (4) (6) (8) (10) (12) (14) (16) (18) (20) (22) (24) (26)

Awyu family, Kiwaiian family, Eleman family, Koiarian family, Angan family, Gorokan family, Engan family, Dani family, Sentani family, Torricelli family, Lower Sepik family, Kalam family, South Bougainville family.

In more recent years, Pawley (1998) has made a further attempt to come to grips with the classification and higher-level subgrouping of the Papuan languages, applying the traditional comparative method to the lowerorder groups with which he was most familiar in an attempt to reconstruct features of higher-level Papuan subgroups. To date, Pawley and his team have managed to reconstruct some 100 Trans New Guinea proto-forms, including the TNG pronominal system. They recognise the Trans New Guinea phylum, which they define in the following terms: The core of the group consists of many small subgroups spoken in the central mountain ranges of New Guinea, starting from the Bird’s Head and extending into South-East Papua, together with the Asmat-

104 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Endangerment and Globalisation in the Pacific

103

Kamoro and the Awyu-Dumut groups of the southwest lowlands and the two large groups of northeast New Guinea: the Madang group (about 100 languages) and the Finisterre-Huon group (about 70 languages) – some 400 languages in all. (Pawley, 1998: 683) As far as the major language subgroups outside the Trans New Guinea phylum are concerned, two main areas of New Guinea exhibit such internal genetic complexity that currently available data do not permit any firm conclusions about broader relationships. These areas are: • the North New Guinea area between the highlands and the north coast, from the western border of the Madang group in Papua New Guinea to the eastern border of the Geelvink Bay phylum in Irian Jaya; • the Gulf of Papua area, covering most of the Gulf Province and the adjacent coastal part of the Western Province of Papua New Guinea. Thus, while there are many established lower-order Papuan language groupings, higher-order groupings are much more problematic. However, as more detailed descriptive and comparative research progresses, the ultimate demonstration that Papuan languages are all genetically related appears more likely.

Pacific Demography The populations of the states and territories that comprise the Oceanic region are as shown in Table 7.3.

Pacific Diaspora The Pacific today is characterised by the phenomenon of ever-increasing diaspora, especially in Polynesia and Micronesia (see Ward, 1997), with significant consequences for language maintenance and preservation. Table 7.4 shows the number of Pacific Islanders in four Pacific Rim countries in 1991 according to Ward (1997). Hawaiians and New Zealand Maori are not included. There are a further 21,900 people born in the Pacific Islands resident in metropolitan France. This means that there are now well over 400,000 people of Pacific Island ethnicity (or who could claim to be Pacific Islanders by birth), living in Pacific Rim countries or outside their homelands. This is about the same number as the total of all Micronesians currently living in Micronesia or 75% of all Polynesians resident in Polynesia (excluding New Zealand and Hawaii). For example, Samoa has a resident population of some 250,000. However, almost the same number of Samoans live outside Samoa, in New Zealand, Hawaii, California and Australia. Samoans and other Pacific

105 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

104

Table 7.3 Populations of the Oceanic region State West Irian (Indonesia) Papua New Guinea Solomon Islands Vanuatu New Caledonia (France) Fiji Tonga Niue Wallis & Futuna (France) Samoa Tokelau Tuvalu Cook Islands New Zealand French Polynesia (France) Hawaii (USA) Easter Island (Chile) Kiribati Nauru Marshall Islands Federated States of Micronesia Northern Mirianas Palau (Belau) Guam

Population

Languages

2,000,000 4,300,000 4,000,000 193,000 196,000 (of whom 50,000 indigenous) 800,000 (of whom about 400,000 indigenous) 100,000 (+ population in NZ, diaspora) 2080 (+14,000 in New Zealand) 14,000 (+ 10,000 in New Caledonia) 250,000 (+ 160,000 in New Zealand) 1,500 10,900 10,020 (+ 50,000 in New Zealand) 4,000,000 (of whom 100,000 indigenous) 219,000 2,000,000 (of whom 30,000 indigenous) 50,000 77,560 11,400 60,000 105,500 58,800 17,225 145,000

250 750 62 113 28 3 1 1 2 1 1 1 3 1 6 1 1 1 1 1 7 1 1 1

Table 7.4 Pacific Islanders in Pacific Rim countries Country/State USA (excluding Hawaiians) (E) USA (excluding Hawaiians) (B) New Zealand (E) Australia (B) Canada (B & E)

Estimated Number of Pacific Island Residents 145,000 (88,000) 170,400 84,600 16,700

B = place of birth data, E = ethnicity data. Sources: 1990 Census for United States of America; 1991 censuses for New Zealand and Australia; 1990 Census for France; estimate based on place of birth of immigrant population and ethnic origin tables of 1991 census for Canada. (Ward, 1997: 185)

W:\Clients\MLM\Cunningham\Proof2\Cunningham Proof 2a 31 January 2006 20:38:07

Color profile: Disabled Composite Default screen

Language Endangerment and Globalisation in the Pacific

105

Islanders, especially Polynesians, have left their homelands in search of paid employment, the proceeds of which are remitted home on a regular basis, thereby forming a significant proportion of Polynesian economies. The Samoan language as spoken in New Zealand has changed, especially lexically, to such an extent that there is now a separate dictionary of New Zealand Samoan. The same situation obtains with respect to most other Polynesian languages: • Tongan has large expatriate communities both in New Zealand and Australia; • Niue has a resident population of only 1800 while some 14,000 Niueans live and work in New Zealand; • the Cook Islands has something approaching 50% of its population living in New Zealand too; • there are many more Wallisians (East Uvea) and Futunans living in French-speaking New Caledonia than in their home islands; • in French Polynesia more than half of the Marquesan population resides and works in Tahiti, as do a high proportion of speakers of languages spoken in the Austral Islands. Even the New Zealand Maori people have a large presence in Australia, lured away from their home communities by the promise of a better life. As far as the Micronesians are concerned, they were formerly nearly all part of the United States Trust Territory. As a consequence, they have a right of free entry to Hawaii and the mainland United States of America, a right that many Micronesians have exercised. The linguistic consequences of this diaspora are considerable, all of them negative in terms of language preservation. In nearly all cases, the displaced population has suffered language loss or language impoverishment. In the case of Niuean, for example, the New Zealand Niuean population reports that 65% of Niueans living there have lost the ability to speak Niuean fluently. It is widely reported in the Cook Islands also that Cook Islands Maori has been so affected by contact with English that many Cook Islanders consider that it is in serious danger of disappearing. As for New Zealand Maori, spoken fluently today by only about 30,000 people, there are grave doubts that it will long survive the depredations of the Englishspeaking environment that confronts it at every turn. The modern Pacific diaspora phenomenon has its roots in the 19th century, when Pacific Islanders, especially Melanesians, were engaged as plantation labour in Queensland (Australia), Fiji, Samoa and New Caledonia. Many of these Pacific Islanders remained outside their homelands when recruiting was abolished at the beginning of the 20th century, and their descendants have remained there right up until the present –

107 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

106

witness the number of ‘South Sea Islanders’ resident in Queensland today. In fact, the recruiting of Pacific Islander labour during the second half of the 19th century had a dramatic indirect effect on indigenous Pacific languages in that this displacement of Islanders was one of the principal drivers in the evolution and development of English-based Pacific Pidgins (Tok Pisin in Papua New Guinea, Pijin in the Solomon Islands, and Bislama in Vanuatu). These pidgins have today become major lingua francas throughout Melanesia and, as such, have begun to make deep inroads into the local vernacular languages. At the same time, the colonisation of the Pacific had an impact on language survival, even in the 19th century. In Fiji, for example, there are 300 different dialects spoken (Geraghty, 1994: 9). The British colonial government chose one of these from the island of Bau, near present-day Suva, as the language of communication for administrative purposes throughout the Fiji archipelago. Although many local dialects have survived and are relatively healthy today, a number have also disappeared as a result of the imposition of a dialect that was destined to become ‘Standard Fijian’. In Vanuatu and the Solomon Islands, where there are many very small language communities, 19th century recruiting of labour for the plantations also had grave consequences for a number of the smaller languages. Some, such as Aore in Vanuatu, have become extinct in recent decades, while Araki, a little to the west, off the south coast of Santo, is in grave danger of disappearing within a generation or so. In the Solomon Islands, Tanima and Vano, both formerly spoken on Vanikoro in the extreme southeast, have become extinct in the past 20 years. Communities and language group sizes in Island Melanesia are extremely small, often no more than 500 souls. In Papua New Guinea, many languages are known to have disappeared since European contact – as many as 100, especially in the Sepik region. The reasons for their disappearance are many and varied.

Language Multiplicity and Diversity in the Pacific A cursory glance at Table 7.3 above reveals an astounding multiplicity of languages spoken in the Pacific by very small populations, especially in Melanesia. In Papua New Guinea alone, for example, Nekitel (1998) identifies 282 languages with fewer than 500 speakers, and a total of 417 languages with fewer than 1000 speakers. In Vanuatu (Tryon, 1995) there are 113 languages for a total population of 193,000. Language:speaker ratios are similar in other parts of the Pacific, with the exception of Triangle Polynesia, where large political units and a pyramidal social structure make for much larger linguistic communities, sometimes more than 100,000.

108 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Endangerment and Globalisation in the Pacific

107

However, the languages of the Pacific are characterised not only by their vast number but also by their degree of diversity, often the result of contact between languages of the two genetically unrelated language families of the region, the Austronesian and the Papuan families (Lynch, 1981; Pawley, 1981). Austronesian and Papuan languages are frequently geographically contiguous, especially in the Papua New Guinea and Solomon Islands. The resulting contact has been a major factor in contributing to language diversity, both in terms of Papuan and Austronesian languages. Indeed, some of these languages have undergone such extensive contact that it is unclear whether they should be classified as Papuan or Austronesian, for example, Maisin in Papua New Guinea and Reefs/Santa Cruz in the Solomon Islands (Wurm, 1978; Lincoln, 1978). Other factors contributing to language diversity in Melanesia are physical isolation and inter-tribal hostility and rivalry, resulting from the traditional social organisation referred to above. In fact, the consequences of these factors are having a major political impact in the Solomon Islands at the present time. There is considerable urgency in recording and documenting representative languages in the Melanesian region in particular because, in spite of increasing efforts over the past 30 years, our record and knowledge of the languages of the Pacific is still poor with many languages just a name on a map. This applies especially to the Papuan languages of the New Guinea area, but also to the languages of the Solomon Islands and Vanuatu.

Threats to the Survival of Indigenous Pacific Languages The greatest threat to the survival of Pacific languages is rapidly increasing urbanisation, as Pacific Islanders are inevitably drawn into a cash economy and so are exposed to the growing effects of globalisation. Indeed, many Pacific Islanders, especially Polynesians and Micronesians, have left their homelands temporarily or permanently in search of better economic opportunities. On the other hand, languages spoken in areas remote from urban centres are not so much under threat, as traditional lifestyle is more likely to be maintained. Proximity to urban centres is also a threat even to languages with relatively greater numbers of speakers as they become impoverished under pressure from local trade and administrative languages, pidgins and creoles, and metropolitan languages, especially English and French, introduced by colonising powers in the 18th and 19th centuries. Increasing access to education, linked to urbanisation, is another important factor contributing to language endangerment as it involves the movement of young students away from their home areas. Nearly all students who undertake secondary schooling have to move away from their home

109 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

108

villages and/or islands. Those who advance further inevitably finish up in provincial or capital cities. Since these are the main areas in which paid employment can be found, those who have come to live in urban centres rarely return to their home communities and so have less and less contact with their home languages. As Pacific states have attained independence over the past forty years, governments have had to make use of language as a nation-building instrument, especially in areas such as Melanesia, where no single indigenous language covers more than a few kilometres. In Vanuatu, for example, at independence in 1980, there was no single language used or understood by the whole population, some students in the former Anglo–French condominium having been ‘educated’ in French and others in English. The nearest thing to a national language was Bislama, an English-based pidgin, commonly used among ni-Vanuatu (as the people of Vanuatu are called) of different language backgrounds. Since independence, Bislama has become the national language and very much the dominant language in urban agglomerations. Marriages between partners of different vernacular language backgrounds are increasingly common in Melanesia, predominantly but not exclusively in urban areas, as young people from different island groups come together for education and employment. The children resulting from these unions almost invariably grow up with either an English-based pidgin or a variety of English or French as their mother tongue. The consequences for the transmission of the parental mother tongues are not difficult to foresee. While this phenomenon is most acute in Melanesia, it is not uncommon in Polynesia and Micronesia, where one would imagine that larger political units and a single local vernacular would constitute a bastion against invading metropolitan languages. On the other hand, mixed-race Tahitians in their thirties and forties, for example, have confessed with embarrassment to the present writer that they could not express themselves in Tahitian but only in French in spite of having Tahitian mothers. In summary, the major point to be made here is that a small number of speakers of a language does not per se constitute a major threat to the survival of many of the Austronesian (or indeed non-Austronesian or Papuan) languages of Oceania. Much more significant are the ravages wrought by galloping urbanisation, the displacement of young students during the education process, and the exigencies of nation-building in states composed of micro-societies, each speaking a distinct, if related, language. Of course, there are other factors that impinge on language endangerment, quite apart from the threats posed by increasing globalisation and monoculturalism. In Melanesia especially, evangelisation is increasingly carried out in an English-based pidgin, rather than a local vernacular, especially since the

110 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:45

Color profile: Disabled Composite Default screen

Language Endangerment and Globalisation in the Pacific

109

beginning of the 1970s. This simply exacerbates the pressure on local vernaculars, particularly as their role shrinks in relevance to only the village level, while population mobility is on the increase. This gradual reduction in role and function of the indigenous languages of the Pacific is becoming more and more widespread despite some significant rearguard action in some countries.

Pacific Island Responses to Language Endangerment While it is widely believed that a considerable proportion of the languages of Oceania are doomed to disappear over the next century, if not before, there is little being done to arrest this slide, either by Pacific governments or by Pacific Islanders themselves, so preoccupied are they with issues of governance and economic development. There are some exceptions, however. In Vanuatu, for example, the people themselves are making considerable efforts to record, preserve and transmit the local vernaculars to their children, an investment for the future. At the Vanuatu National Museum (formerly the Vanuatu Cultural Centre), there was founded in 1980 a network of community cultural officers or ‘fieldworkers’. The network comprised a total of approximately 100 men and women, who have received linguistic and ethnographic training through a series of annual workshops, funded until recently by the Australian Government’s South Pacific Cultures Fund, and now funded predominantly by the Vanuatu Government. The Vanuatu ‘fieldworkers’ are chosen by their communities and work to preserve language and culture without salary, simply through pride in their own identities. The motivation for the setting-up of the Vanuatu Cultural Centre Fieldworkers’ Program was the concern expressed by the Malfatumauri or National Council of Chiefs that much traditional and linguistic knowledge was being lost as the result of the death of a number of Vanuatu elder statesmen. Concern was also expressed at the inroads being made by Bislama, the Melanesian Pidgin lingua franca, and the languages of education, English and French. The role of the fieldworkers is a dual one: the compilation of dictionaries of their mother tongues and the writing of their own ethnographies for the benefit of future generations. Because much of this information is either secret or sacred, some parts of it may not be communicated outside the language-owning community group. In order to address this situation and also because Vanuatu is in a zone where cyclones damage the country almost every year, a novel proposal was developed at the Vanuatu Cultural Centre, one which has received national recognition and support. What was established was a ‘Taboo Room’, a cyclone-proof air-conditioned room, rather like a bank vault, in the middle of the Vanuatu Cultural Centre

111 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

110

(or Vanuatu National Museum as it is also known today). The Vanuatu Cultural Centre fieldworkers and many other ni-Vanuatu deposit there copies of their dictionary files and ethnographic field note-books and taperecordings, much as one makes a bank deposit. Only the depositor or his/ her nominee may withdraw or consult the materials, managed by the Vanuatu Cultural Centre staff. The originals are usually kept at home in the village of the individual fieldworker. If by mishap they should be destroyed by cyclone, fire, rats or occasionally children, they can easily be replaced and new copies made. References Bellwood, P. (1989) The colonization of the Pacific: Some current hypotheses. In A.V.S. Hill and S.W. Seargentson (eds) The Colonization of the Pacific: A Genetic Trail (pp. 1–59). Oxford: Clarendon Press. Foley, W.A. (1986) The Papuan Languages of New Guinea. Cambridge: CUP. Geraghty, P. (1994) Fijian Phrasebook. Hawthorn: Lonely Planet Publications. Golson, J. (1991) Bulmer phase II: Early agriculture in the New Guinea Highlands. In A. Pawley (ed.) Man and a Half: Essays in Pacific Anthropology and Ethnobiology in Honour of Ralph Bulmer (pp. 484–491). Auckland: Polynesian Society. Lincoln, P.C. (1978) Reef-Santa Cruz as Austronesian. In S.A. Wurm and L. Carrington (eds) Second International Conference on Austronesian Linguistics Proceedings (pp. 929–967). Pacific Linguistics Series C, No 61. Canberra: Pacific Linguistics. Lynch, J.D. (1981) Melanesian diversity and Polynesian homogeneity: The other side of the coin. Oceanic Linguistics 20 (2), 95–129. Nekitel, O. (1998) Voices of Yesterday, Today and Tomorrow: Language, Culture and Identity. New Delhi: UBS Publishers’ Distributors Ltd. Pawley, A. (1981) Melanesian diversity and Polynesian homogeneity: A unified explanation for language. In K.J. Hollyman and A. Pawley (eds) Studies in Pacific Languages in Honour of Bruce Biggs (pp. 269–309). Auckland: Linguistic Society of New Zealand. Pawley, A. (1998) The Trans New Guinea phylum hypothesis: A reassessment. In J. Miedma, C. Odé and R.A.C. Dam (eds) Perspectives on the Bird’s Head of Irian Jaya, Indonesia (pp. 655–690). Proceedings of the Conference. Leiden: University of Leiden, Department of Languages and Cultures of South East Asia and Oceania Ross, M.D. (1988) Proto Oceanic and the Austronesian Languages of Western Melanesia. Pacific Linguistics Series C, No. 98. Canberra: Pacific Linguistics. South Pacific Commission (1997) Pacific Island Populations Data Sheet. Noumea: South Pacific Commission Spriggs, M. (1997) The Island Melanesians. Oxford: Blackwell. Swadling, P. and Muke, J. (eds) (1998) 9000 Years of Gardening: Kuk and the Archaeology of Agriculture in Papua New Guinea. Port Moresby: Papua New Guinea National Museum. Tryon, D.T. (ed.) (1995) Comparative Austronesian Dictionary (4 vols). Berlin: Mouton de Gruyter. Ward, R.G. (1997) Expanding worlds of Oceania: Implications of immigration. In K.Sudo and S. Yoshida (eds) Contemporary Migration in Oceania: Diaspora and Network (pp. 179–196). JCAS Symposium Series 3. Osaka: National Museum of Ethnology.

112 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Language Endangerment and Globalisation in the Pacific

111

Wurm, S.A. (1975) New Guinea Area Languages and Language Study (Vol.1): Papuan Languages and the New Guinea Linguistic Scene. Pacific Linguistics Series C, No. 38. Canberra: Pacific Linguistics. Wurm, S.A. (1978) Reefs-Santa Cruz: Austronesian, but...! In S.A. Wurm and L. Carrington (eds) Second International Conference on Austronesian Linguistics Proceedings (pp. 969-1010). Pacific Linguistics Series C, No. 61. Canberra: Pacific Linguistics Wurm, S.A. (1982) Papuan Languages of Oceania. Tübingen: Gunter Narr Verlag.

113 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Chapter 8

Endangered Languages of China and South-East Asia DAVID BRADLEY

Introduction While, in most parts of the world, the dimensions of the language endangerment problem are quite well-known, in China and some adjacent areas of mainland South-East Asia, this is not so. This is mainly because the official policies of the countries in this area normally amalgamate related and sometimes even unrelated small minorities and their languages into larger, recognised groups. For example, China has officially designated 55 national minorities but, within many of them, there is an enormous linguistic diversity, just as there is great linguistic diversity within the Han Chinese majority nationality. Some linguists say that the (Han) Chinese language comprises a large number of distinct related languages, including Mandarin or northern Chinese, Wu in the area around Shanghai, Min mainly in Taiwan and in Fujian Province, Yue or Cantonese in Hong Kong, much of Guangdong Province and some surrounding areas, and Kejia or Hakka scattered across the south-east of China. From the Chinese perspective, all Han Chinese share a common character writing system, a long history and a feeling of cultural unity; so it is not so strange that they prefer to be regarded as a single group (for more discussion on this, see Bradley, 1991). This is so despite the fact that direct oral communication between Han Chinese speaking different varieties is often difficult or impossible unless one or both resort to a lingua franca variety such as standard Beijing Mandarin, now known in China as putonghua, ‘common speech’. It is clear that the degree of diversity expected or allowed within a minority nationality in China is derived from and reflects this degree of diversity within the majority nationality. For many of the composite national minorities of China, such unitary feelings are absent though modern political processes are beginning to create an incipient shared identity. For an account of one such nationality, the Yi, see Bradley (2001a, 2001b) and Harrell (2001). Briefly, this group is divided by official Chinese studies into six ‘dialects’, within each of which

112

114 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Endangered Languages of China and South-East Asia

113

there is a variety of mutually unintelligible languages, probably more than 100 altogether, and many completely undescribed. The Yi are scattered across four south-western Chinese provinces, with a very small number of some of the southernmost groups extending into northern Vietnam and northern Laos. Unlike Han Chinese, for the Yi there are four fairly distinct literary traditions, within each of which there is further internal diversity. On the whole, Yi people do not understand or learn the languages of other Yi subgroups, and so Chinese must be used to communicate outside the local area. Chinese and other linguists have done very substantial work, especially since 1950, on the major language selected as the ‘standard’ for each recognised national minority. In many cases, this includes the development and dissemination of orthographies and very extensive publication in these languages as well as descriptive linguistic work. However, this provides no support for the other languages included within a nationality, whose speakers are expected to learn through the ‘standard’ language, even if it is not their own; nor does it provide support for people who live outside the ‘autonomous’ areas where each particular nationality is concentrated. In Thailand, a great deal of excellent work has been done by Thai and other linguists, reaching out to nearly all of the endangered language communities there. While the government policy promotes the use and spread of Thai, there is unofficial tolerance for local diversity. However, in many areas, people are spontaneously choosing to assimilate and speak Thai. Peninsular Malaysia, like the rest of Malaysia, has a policy of support for indigenous peoples including the Orang Asli (‘original people’) which, in that area, are or were mainly Mon-Khmer language speakers or small coastal groups of speakers of Austronesian languages close to Malay. However this does not extend to support for their languages and cultures and the use of Malay is spreading rapidly. Field research by non-Malaysians is now banned, and few Malaysians do any linguistic work on these languages. In Laos and Vietnam, as in China, there is an official inventory of minority categories: 47 in Laos and 54 in Vietnam. While the policy is, in principle, supportive of the rights of such groups, in both countries the overall priority is the national language rather than any other indigenous language. Some linguistic work has been carried out in Vietnam but rather less in Laos. Cambodia also promotes the national language, Khmer; the 1970s and 1980s were disastrous for its other indigenous Mon-Khmer languages and for their speakers, who have no official status as separate groups. Descriptive work is urgently needed to supplement limited materials collected mainly by French scholars during the colonial period. The situation in Burma is even more problematic. Fieldwork by

115 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

114

outsiders has been virtually impossible for nearly 40 years and, in the meantime, the official policy of using the Burmese language as the sole official language and medium of education has had the effect one might expect on many minority languages. The 135 recognised ethnic groups of Burma include many whose languages are endangered; in fact, most of the languages of Burma other than Burmese are endangered to some extent. For a survey which includes data on a number of languages, see Luce (1985). Basic facts about all of the known languages of this area have been documented in a number of previous studies (see Ramsey, 1987; Bradley, 1981, 1994a; and Shearer & Sun, 2002 for some examples). More extensive details on the situation of the endangered languages are given in Bradley (2005b). These papers list and provide information on 144 endangered languages that represent all of the genetic linguistic groupings found in the area. There is also further information on the sociolinguistic effects of official policy on such languages in Bradley (1987, 1994b, 1998) on the usage and spread of various lingua francas in the area in Bradley (1996a) and on other aspects of the sociolinguistics of language endangerment in Bradley (1983, 1996b) and Bradley and Bradley (2002) and on other aspects of linguistic and cultural contact in Harrell (1995). Given the limited linguistic surveys that have been carried out in China and Burma, it is clear that there are many other unreported endangered languages there. Because of this, we have been carrying out surveys in parts of Yunnan Province in south-western China for some years and, in the process, have located and started to describe a number of previouslyunknown languages, all of which are endangered (see Bradley et al., 1999 for details). Our surveys have mainly covered areas quite close to centres of government administration, which have also long been foci for minority assimilation; nevertheless, we have found undescribed languages spoken in villages within a short taxi ride of the centre of Kunming, the capital of Yunnan Province in south-western China. It is highly likely that the density of unreported languages is higher in more remote areas. Even were this not so, the density found so far (more than one unreported language per county) would suggest at least another 100 such languages in Yunnan Province alone, but this is almost certainly too low an estimate. In China, local authorities are not interested in preserving such languages; most are not recognised for any purpose and their speakers are classified as members of some larger national minority or just left unclassified. Since 1950, with the spread of direct administration and major improvements in education, communication and economic integration, putonghua, another local variety of Chinese, the language of one or more larger local minorities, or several of these are used increasingly, displacing the traditional languages of smaller groups and thus endangering their linguistic survival. Since such languages are considered to be part of the

116 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Endangered Languages of China and South-East Asia

115

language of the national minority to which they have been assigned, in some cases the ‘standard’ language selected for each national minority may also spread within that group. This is happening, for example, among the Yi nationality in Sichuan Province, just to the north of Yunnan Province. Of the 144 languages in this area recently reported as endangered in Bradley (2005a), 135 are in southern China and mainland South-East Asia. Of these, 56 are Tibeto-Burman languages, 51 are Mon-Khmer languages, 17 are Thai-Kadai languages, 5 are Austronesian, 4 are Miao-Yao, and 2 are Portuguese creoles with an Indo-European lexical source and acrolect. Of the 9 endangered languages of northern China, 5 are Manchu-Tungus, 3 are Mongol, and 1 is Turkic. The publication cited gives name, location, number of speakers and degree of endangerment, as well as relevant sociolinguistic information. More in-depth studies of a number of these languages and of the sociolinguistic situation of several languages of China and mainland South-East Asia are in Bradley (2005b).

Case Study: The Sanie of Kunming The Sanie are a group of just over 17,000 people within the Yi nationality of south-western China, who live in 76 villages immediately to the west of Kunming city in central Yunnan, in most areas of Xishan District and extending into the south-west corner of Fumin County and the north-west corner of Anning County. The Sanie are not the only Yi group in this area; immediately to their west, north and south, there are the Nasu (in Chinese, Hei Yi or ‘black Yi’) and, in a couple of villages, the Sanie and Nasu live together with limited bilingualism, mainly Sanie who can speak some Nasu. Most of the Nasu, a much larger group of several hundred thousand speakers, live further north in Fumin, Luquan, Wuding and surrounding counties. The Sanie language has, of course, been known to local authorities but never previously studied. Many Sanie are uncertain what to call themselves in Chinese; most now say that they are Bai Yi (‘white Yi’); before 1950, they called themselves Bai Luoluo (‘white Lolo’), but the term Lolo is not used in China now. Of course, the term Bai Yi is also applied to many other Yi groups. According to Sanie tradition, their original home was the area of the west gate of Kunming city itself, whence they moved westwards into the Xishan District. The Sanie have no books and no tradition of having had books; in this, they are unlike most surrounding Yi groups including the Nasu. Sanie presents a typical example of language death in progress. Firstly, there are doubtless many residents of the Kunming area with unremembered Sanie ancestry. In the villages closest to the city (Cjiabi, Shiju and Zhaozong at the foot of Xishan), the language is dying; only those over 60 years of age speak it well, some younger people are able to understand a bit,

117 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

116

but children do not know the language at all, even though they are often cared for by their grandparents, who may well be bilingual. These villages have long been in contact with the Han Chinese majority; the main road from Kunming to Dali used to pass right through the centre of Chejiabi, and the main road to the west is still next to the village. They are now completely surrounded and outnumbered by Han Chinese and are well-integrated into the urban economy. In the first villages in the mountains, such as Daxing and Huahongyuan, the youngest fluent speakers are young adults but, again, children are not learning the language. These villages have also been in close contact with the Han for many years. Moving outwards, in the next ring of villages, such as Baimei, Yuhua and those in north-east Gulu, closer contact with Han Chinese started after 1950 and Sanie is still used to a limited extent by young people though many are semi-speakers. The most distant villages, such as those in western Tuanjie, central Gulu, Anning and Fumin, are still relatively remote, but have also started to have extensive contact with Han Chinese and not all young people speak Sanie there, either. The provincial government has recognised the predominant position of the Yi and other minority nationalities in Tuanjie Township and Gulu Township by designating Tuanjie as a Yi and Bai township and Gulu as a Yi township. The other nearby nationalities are mainly Bai in eastern Tuanjie with a few Miao in western Gulu and Tuanjie. Most of the Yi in western Gulu are Nasu. The main dividing line is the Pudu River, but there are a few Nasu villages to its east in Gulu and Anning and a few Sanie villages to its west in north-western Gulu and south-western Fumin, where the two varieties are in the closest contact. Chinese linguists have classified the Yi of Kunming and Anning in the Kun-An vernacular of the Eastern dialect of Yi. This classification is correct, though the differences within the Kun-An vernacular are so great that the other nearby and closely-related Kun-An vernacular Yi languages, Samei and Samataw, are not mutually intelligible, much less Nasu which is also Eastern Yi but from another cluster of languages further north and northeast. The main consonant characteristic that distinguishes Sanie, Samei and Samataw from Nasu and other Eastern Yi varieties is the lack of distinct prenasalised or breathy-voiced initial stops; where most Nasu varieties have [mb], [mph] or [bh], these languages have merged them with the plain voiced stops. They also lack a distinct voiceless, creaky or breathy nasal series such as [hm], [m] or [mh] as often found in Nasu and, apart from the speech of one Sanie village (Zhaozong), also lack a voiceless lateral. Another difference from most types of Eastern Yi including ‘standard’ Nasu is the absence of retroflex initials. In Sanie, most of these are merged

118 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Endangered Languages of China and South-East Asia

117

with the corresponding alveolars or alveopalatals, as also happens in some southern varieties of Nasu spoken around Kunming adjacent to the Sanie. Sanie also shows typical Eastern Yi vowel and tone systems. For example, it has a high rising tone corresponding to the high level tone of most other nearby Eastern Yi languages, including Nasu, Samei and Samataw. This distinguishes these languages and the Northern Yi languages from the rest of the languages spoken by the Yi nationality, which have a low creaky tone in the same words. The degree of dialect differentiation within Sanie is very great, despite the small distances involved. In the south-eastern part of the Sanie area, speakers in five villages use a cluster of velar plus /w/ which corresponds to a palatal before front vowel rhymes and a simple velar before most rhymes elsewhere in Sanie and in Eastern Yi and Yi generally. These labiovelar clusters are of great comparative linguistic importance since they provide the only direct documentation for the original state of these clusters in any language of this group of Tibeto-Burman. A similar local difference is in the dropping of high vowels after nasals, leaving a syllabic nasal. For example, in the speech of Qitai in south central Tuanjie Township, syllabic [m] is particularly frequent as it replaces not just syllables with initial [m] followed by a high back vowel but also syllables with an initial velar nasal followed by a high back vowel. There are also major lexical differences in some very frequent words. For further linguistic details on Sanie, see Bradley et al. (1999). We are also preparing a volume on the outcomes of our surveys of the Kunming region, which will include extensive material on the various subvarieties of Sanie and several other newly described languages. At the request of local authorities in Qinghe Township, Fumin County, we devised a romanisation for Sanie to help them in their efforts at language maintenance and produced a version of the standard Yi literacy textbook (which uses reformed Yunnan Yi characters1) in Sanie. This romanisation for Sanie is based on the same principles as the standard pinyin romanisation of Chinese and is intended both to support the maintenance of Sanie and to assist in the learning of standard Chinese.

Conclusion For more than 10 years, UNESCO and related bodies such as the Comité International Permanent des Linguistes (CIPL) have been supporting urgent work to document disappearing languages and, where possible, to reverse language shift (Robins & Uhlenbeck, 1991). Over the same period, this has also become a central focus for a wide variety of linguists. A large number of language maintenance and revitalisation projects have been started in a wide range of communities, mainly among indigenous

119 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:46

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

118

minority groups in the developed world. It is of crucial importance that the momentum achieved should continue so that we may try, as linguists, to avert the ecolinguistic catastrophe widely predicted for this century: the disappearance of 90% or more of the world’s languages. In this part of the world, apart from the 144 endangered languages and a similarly-large number of unknown and undescribed endangered languages yet to be surveyed in China and Burma, there are many languages endangered to a lesser degree. They include all of the Mon-Khmer languages apart from Vietnamese, Khmer and Khasi, as well as most other Tibeto-Burman, ThaiKadai and Miao-Yao languages. Most of these languages are losing domains and are being learned by fewer and fewer children. Every Tungusic language, including those in Russia as well as in China, is more or less severely endangered. In South Asia, most Tibeto-Burman languages are also endangered to some degree, some more severely than others. All the Thai-Kadai languages still spoken in South Asia are endangered. This means that about another 100 Mon-Khmer languages, possibly as many as 200 or more Tibeto-Burman languages, and at least 50 Thai-Kadai, MiaoYao and other additional languages in this and surrounding areas are potentially endangered. The area under consideration is thus the locus of about a tenth of the world’s linguistic diversity and an even higher proportion of the language shift that may be expected in the future. Given this, it is grossly under-studied and under-resourced. If this situation existed in any developed nation, there would be hundreds or thousands of linguists already in the field, and communities would be clamouring for and receiving language maintenance and revitalisation assistance! Future studies, if they are done in time, can still provide details of the additional unreported languages but, if such surveys are not carried out soon, the many still-undescribed and unknown languages in southwestern China and Burma will disappear without a trace. This may thus be the area in the world most urgently in need of primary linguistic descriptive work, perhaps with as many languages as New Guinea. It is also a fascinating sociolinguistic laboratory where large numbers of languages are being influenced by, absorbed into or replaced by more or less closely related languages, rather than being replaced by an Indo-European language, as is the case in most other areas of the world. In addition to doing linguistic studies, we can fulfil our obligation to language communities, not just by documenting their languages, but by helping them to maintain them. Acknowledgement I wish to dedicate this paper to the memory of Professor Stephen A. Wurm, who passed away on 24 October 2001. He has been the central figure in worldwide research on a wide range of endangered languages for many

120 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:47

Color profile: Disabled Composite Default screen

Endangered Languages of China and South-East Asia

119

years and has acted as a mentor for a large number of projects, including ours. I also gratefully acknowledge the funding support of the UNESCO endangered languages programme, initiated by Professor Wurm, and of the Australian Research Council (A59803475, Language Maintenance for Endangered Languages), the assistance of the Yunnan Academy of Social Sciences, and the participation of numerous colleagues in China and elsewhere. Naturally, all remaining errors are my responsibility. Notes 1. This is a new system of Yi characters, based on a compromise between the four traditions of Yi characters. It has been implemented in Yi areas of Yunnan Province since 1989. It does not indicate pronunciation in any way and speakers of each variety of Yi are meant to pronounce words according to their own speech, not according to any external standard. For more details see Bradley, 2001a.

References Bradley, D. (1981) Mainland South-East Asia. In S.A. Wurm and S. Hattori (eds) Language Atlas: Pacific Area (maps 35–37). Pacific Linguistics Series C, No. 66 and 67. Canberra: Pacific Linguistics Bradley, D. (1983) Identity: The persistence of minority groups. In J. McKinnon and Wanat Bhruksasri (eds) Highlanders of Thailand (pp. 46-55). Kuala Lumpur: Oxford University Press. Bradley, D. (1987) Language planning for China’s minorities: The Yi branch. In D.C. Laycock and W. Winter (eds) A World of Language: Papers Presented to Professor S.A. Wurm on his 65th Birthday (pp. 81–89). Pacific Linguistics Series C, No. 100. Canberra: Pacific Linguistics. Bradley, D. (1991) Chinese as a pluricentric language. In M.G. Clyne (ed.) Pluricentric Languages (pp. 305–324). Berlin: Mouton de Gruyter. Bradley, D. (1994a) Mon-Khmer, Austro-Thai, Tibeto-Burman. In C. Moseley and R. Asher (eds) Atlas of the World’s Languages (pp. 159–182; maps 47–51).London: Routledge. Bradley, D. (1994b) Building identity and the modernisation of language: Minority language policy in Thailand and China. In A. Gomes (ed.) Modernity and Identity: Asian Illustrations (pp. 192–205). Bundoora: Institute of Asian Studies, La Trobe University for Asian Studies Association of Australia. Bradley, D. (1996a) Chapters 65-74 and 79–80, Maps 85–90 and 97. In S.A. Wurm, P. Mühlhäusler and D. Tryon (eds) Atlas of Languages for Intercultural Communication (pp. 745–786, 835–844). Berlin: Mouton de Gruyter. Bradley, D. (1996b) Language policy and the typology of scripts. In Suriya Ratanakul et al. (eds) Pan-Asiatic Linguistics. Proceedings of the Fourth International Symposium on Languages and Linguistics (pp. 1845–1856). Salaya: Institute of Language and Culture for Rural Development, Mahiodl University. Bradley, D. (1998) Minority language policy and endangered languages in China and Southeast Asia. In K. Matsumura (ed.) Studies in Endangered Languages (pp. 49–83). Tokyo: Hituzi Syobo. Bradley, D. (2001a) Language Policy for the Yi. In S. Harrell (ed.) Perspectives on the Yi of Southwest China (pp. 195–214). Berkeley: University of California Press.

121 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:47

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

120

Bradley, D. (2001b) Yi. In J. Garry and C. Rubino (eds) Facts about the World’s Major Languages (pp. 826–829). Chester, CT: H.W. Wilson. Bradley, D. (2005a) Endangered languages and language maintenance in China. International Journal of the Sociology of Language, 173. Bradley, D. (2005b) Language endangerment in China and South-East Asia. In C. Moseley (ed.) Encyclopedia of Endangered Languages. London: Routledge. Bradley, D. and Bradley, M. (eds) (2002) Language Endangerment and Language Maintenance: An Active Approach. London: RoutledgeCurzon. Bradley, D., Bradley, M. and Li, Y. (1999) Language maintenance for endangered languages of Central Yunnan, China. In N. Ostler (ed.) Endangered Languages and Education (pp. 13–20). Proceedings of the Third Foundation for Endangered Languages Conference. Maynooth: Foundation for Endangered Languages. Harrell, S. (ed.) (1995) Cultural Encounters on China’s Ethnic Frontiers. Seattle: University of Washington Press. Harrell, S. (2001) Ways of Being Ethnic in Southwest China. Seattle: University of Washington Press. Luce, G.H. (1985) Phases of Pre-Pagán Burma, Language and History (2 vols). Oxford: Oxford University Press. Ramsey, S.R. (1987) The Languages of China. Princeton: Princeton University Press. Robins, R.H. and Uhlenbeck, E.M. (eds) (1991) Endangered Languages. Oxford: Berg. Shearer, W. and Sun, H. (2002) Speakers of the Non-Han Languages of China. Lewiston, NT: Edwin Mellon Press.

122 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:47

Color profile: Disabled Composite Default screen

Chapter 9

On the Edge of the Pacific: Indonesia and East Timor JOHN HAJEK

Introduction A brief report is presented here on language vitality and endangerment in two neighbouring nations, Indonesia and newly-independent East Timor. The area covered by Indonesia is a large one and extends from the small Barrier Islands off the Sumatran coast at the extreme west to the western half of the island of New Guinea (today commonly referred to as West Papua) to the east. East Timor is geographically much smaller and occupies the eastern half of the island of Timor in the Nusa Tenggara region of Indonesia. Indonesia dwarfs East Timor not only physically but also in terms of population size, with more than 206 million inhabitants in contrast to East Timor’s estimated 800,000 citizens. The most recent edition of Ethnologue1 lists 731 languages for Indonesia (of which three are noted as extinct and two are sign languages) and 20 (one extinct) for East Timor. Indigenous languages in both countries fall into two major groupings: Austronesian and the so-called Non-Austronesian or Papuan languages. The Austronesian language family is, of course, enormous – both in number of languages and in physical dispersion, extending well beyond the boundaries of the Indonesian–East Timorese region into the Pacific region. Non-Austronesian languages are concentrated in and around the island of New Guinea, and fall into a number of different language phyla. They are found as far west as Timor (mostly East but also West), and parts of the Moluccas. The non-Austronesian languages of Timor and of the much smaller islands of Alor and Pantar are considered to be most closely related to languages spoken in West Papua.

East Timor East Timor also includes two off-shore islands: Atauro and Jaco, as well as a small enclave known as Oé-Cusse located in West Timor. As in Indonesia, most languages spoken in this region are Austronesian, the best known being Tetun. The Dili variety of this language is widely spoken as a

121

123 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:47

Color profile: Disabled Composite Default screen

Language Diversity in the Pacific

122

lingua franca across ethnic groups and has recently been declared, alongside Portuguese, the ‘official language’ of East Timor. Indonesian has now lost all the official status it held between 1975 and 1999. From the 1500s until 1975, East Timor was a Portuguese colony although full Portuguese control was not established until the early 1900s when the last major uprising was crushed. Until then, the Portuguese presence was strongest in the capital, Dili, along the northern coast, and in some highland areas. During World War II, the entry of opposing Allied and Japanese troops into officially neutral Portuguese Timor led to bloody conflict and the death of some 50,000 East Timorese. Bloody conflict returned in 1975 when Indonesian troops entered East Timor. Although East Timor was officially declared a province of Indonesia by Indonesian authorities in 1976, local resistance to incorporation into Indonesia continued until the final withdrawal of Indonesian forces and administration in September 1999. It is widely reported that some 200,000 East Timorese lost their lives during this period (see Hajek, 2000). The events before and after the United Nations-sanctioned referendum on self-determination in East Timor on 30 August, 1999 received global attention as a result of the activities of the withdrawing Indonesian military and their militia allies. In addition to the destruction of East Timor’s infrastructure, there was mass dislocation of the local population and some 80% of the population are thought to have fled their homes. Approximately 40,000 to 50,000 remain to this day in militia-controlled refugee camps in West Timor. Indigenous languages have shown remarkable resilience despite the stresses placed on them and their speakers by the events of the last sixty years. Massive loss of life, dislocation and social upheaval do not appear to have permanently affected the vitality of indigenous languages. The report by Carey (1997) that local languages were no longer spoken is now known to have no basis. Indeed, recent fieldwork in the region confirms strong vitality – with the exception of tiny Maku’a which may have as few as five elderly speakers (see Himmelmann & Hajek, 2001: 126). Language shift in the Maku’a community to predominant Fataluku has been under way for some time and is now almost complete. Although Ethnologue reports 19 living languages in East Timor (Table 9.1), the number is undoubtedly higher. Missing, for example, is Bekais, with some 2000 speakers on the East/West Timor border.

Indonesia The degree of language diversity in Indonesia is, not surprisingly, much greater than that found in East Timor. The evidence in favour of language maintenance and preservation is, however, extremely mixed in Indonesia.

124 W:\Clients\MLM\Cunningham\Proof2 22 January 2006 15:25:47

Color profile: Disabled Composite Default screen

On the Edge of the Pacific: Indonesia and East Timor

123

Table 9.1 Distribution of living languages in East Timor according to (sub)family and size West M-P Sundic

Central M-P

TNG

Other

Number of languages

%

>1,000,000 >100,000

1

>30,000

4

3

1

5.3

7

36.8

>10,000

1

1

5.3

>5,000

2

2

10.6

>2,500

1

1

5.3

>1,000

4

4

21.1

1

5.3

1

2

10.6

1

19

>500 1,000,000

1

7

6

14

9.3

>100,000

7

22

6

35

23.3

>30,000

19

13

32

21.3

>10,000

18

6

1

25

16.7

>5,000

8

2

10

6.7

>2,500

5

>1,000

14

>500

1

1,000,000 >100,000

South Sulawesi

SangirMinah

Mong.Goront

Muna- Number of Buton languages

2

%

2

1.8

3

17

15.2

19

17.0

3

3

24

21.4

3

1

2

11

9.8

1

1

4

6

2

2

>30,000

5

8

5

1

>10,000

12

4

2

>5,000

5

>2,500

4

>1,000

7

2

4

>500

2

1

2

5

4.5

1,000,000

1

>100,000

1

Central M-P (AN)

TNG

Number of languages

%

1

1.4

10

11

16.2

>30,000

8

8

11.8

>10,000

10

4

14

20.6

>5,000

2

4

6

8.8

>2,500

3

3

4.4

>1,000

3

3

4.4

>500

1

1

1.4

18

3

21

30.9

48

18

68

1,000,000 >100,000 >30,000

2

2

2

>10,000

10

1

2

4

6

4.6

3

20

15.3

>5,000

3

1

1

1

5

11

8.4

>2,500

14

1

1

1

8

25

19.1

>1,000

5

6

5

8

24

18.3

>500

1

1

9

11

8.4

1,000,000 >100,000

2

>30,000

2

>10,000

13

>5,000

12

3

3

>2,500

21

2

2

>1,000

27

8

2

4

>500

12

10

1

3

3

29

11.1