Endangered Languages: An Introduction 1139033816, 9781139033817

"Most of the 7,000 languages spoken in the world today will vanish before the end of this century, taking with them

136 2 3MB

English Pages 246 Year 2015

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Half-title
Series page
Title page
Copyright information
Dedication
Table of contents
Preface
1 Introduction
1.1 How can we tell when a language is endangered?
1.2 Where are all the endangered languages?
1.3 A precondition for language endangerment: language contact
1.4 Summary and outline of the book’s contents
1.5 Sources and further readings
2 Why and how languages become endangered
2.1 Conquest
2.2 Economic pressures
2.3 Melting pots
2.4 Language politics
2.5 Attitudes
2.6 Loss of linguistic diversity via standardization
2.7 What doesn’t promote endangerment?
2.8 Summary
2.9 Sources and further readings
3 Sliding into dormancy: Social processes and linguistic effects
3.1 Five case studies
3.1.1 Case study #1: Eyak
3.1.2 Case study #2: Cornish
3.1.3 Case study #3: Egyptian
3.1.4 Case study #4: Yaaku
3.1.5 Case study #5: Mednyj Aleut
3.2 What do these case studies tell us?
3.3 Tip
3.4 Semi-speakers and rememberers
3.5 Attrition
3.6 An alternative route to language death: grammatical (and lexical) replacement
3.7 Summary and commentary
3.8 Sources and further readings
4 What a community loses: Language loss as cultural loss
4.1 Heritage languages and cultural identity
4.2 Language loss as loss of artistic expression
4.3 Loss of a language, loss of cultural knowledge
4.4 Would the world be better off without linguistic diversity?
4.5 Sources and further readings
5 What science loses: Language loss as a threat to our understanding of human history, human cognition, and the natural world
5.1 Endangered languages and human history
5.2 Endangered languages and human cognition
5.3 Endangered languages and knowledge of the natural world
5.4 Sources and further readings
6 Field research on endangered languages
6.1 Descriptive linguistics and documentary linguistics
6.2 On linguistic fieldwork in general (not just on endangered languages)
6.2.1 What is fieldwork?
6.2.2 What is a reasonable time span for a fieldwork project?
6.2.3 Dealing with psychological stress in the field
6.3 Investigating gravely endangered languages in the field
6.3.1 Working with consultants: access, collaboration, communication, selection, and data-collection techniques
6.3.2 Field sessions
6.3.3 Individual variation
6.3.4 An extended example: dictionary-making
6.4 Summary and commentary
6.5 Sources and further readings
7 Revitalizing endangered languages
7.1 Factors that contribute to successful revitalization efforts
7.2 Some types of revitalization programs
7.3 From the past to the future
7.3.1 Language revitalization in the digital age
7.3.2 The value of unsuccessful revitalization efforts
7.4 Sources and further readings
Glossary – languages and terms
Bibliography
Language index
Names index
Subject index
Recommend Papers

Endangered Languages: An Introduction
 1139033816, 9781139033817

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Endangered Languages Most of the 7000 languages spoken in the world today will vanish before the end of this century, taking with them cultural traditions from all over the world, as well as linguistic structures that would have improved our understanding of the universality and variability of human language. This book is an accessible introduction to the topic of language endangerment, answering questions such as: What is it? How and why does it happen? And why should we care? The book outlines the various causes of language endangerment, explaining what makes a language “safe” and highlighting the danger signs that threaten a minority language. Readers will learn about the consequences of losing a language, both for its former speech community and for our understanding of human language. Illustrated with case studies, it describes the various methods of documenting endangered languages, and shows how they can be revitalized. SARAH G. THOMASON is William J. Gedney Collegiate Professor of Linguistics in the Department of Linguistics at the University of Michigan. Her previous publications include Language Contact: An Introduction (2001).

CAMBRIDGE TEXTBOOKS IN LINGUISTICS General editors: P. A U S T I N , J . B R E S N A N , B . C O M R I E , S . C R A I N , W. D R E S S L E R , C . E W E N , R . L A S S , D . L I G H T F O O T , K . R I C E , I . R O B E R T S , S . R O M A I N E , N . V. S M I T H

Endangered Languages

In this series R . C A N N Formal Semantics J . L AV E R Principles of Phonetics F. R . PA L M E R Grammatical Roles and Relations M . A . J O N E S Foundations of French Syntax A . R A D F O R D Syntactic Theory and the Structure of English: A Minimalist Approach R . D . VA N VA L I N , J R, and R . J . L A P O L L A Syntax: Structure, Meaning and Function

A . D U R A N T I Linguistic Anthropology A . C RU T T E N D E N Intonation Second edition J . K . C H A M B E R S and P. T RU D G I L L Dialectology Second edition C . LYO N S Definiteness R . K AG E R Optimality Theory J . A . H O L M An Introduction to Pidgins and Creoles G . G . C O R B E T T Number C . J . E W E N and H . VA N D E R H U L S T The Phonological Structure of Words F. R . PA L M E R Mood and Modality Second edition B . J . B L A K E Case Second edition E . G U S S M A N Phonology: Analysis and Theory M . Y I P Tone W. C RO F T Typology and Universals Second edition F. C O U L M A S Writing Systems: An Introduction to their Linguistic Analysis P. J . H O P P E R and E . C . T R AU G O T T Grammaticalization Second edition L . W H I T E Second Language Acquisition and Universal Grammar I . P L AG Word-Formation in English W. C RO F T a n d A . C RU S E Cognitive Linguistics A . S I E W I E R S K A Person A . R A D F O R D Minimalist Syntax: Exploring the structure of English D . B U R I N G Binding Theory M . B U T T Theories of Case N . H O R N S T E I N , J . N U Ñ E S and K . G RO H M A N N Understanding Minimalism B . C . L U S T Child Language: Acquisition and Growth G . G . C O R B E T T Agreement J . C . L . I N G R A M Neurolinguistics: An Introduction to Spoken Language Processing and its Disorders

J . C L AC K S O N Indo-European Linguistics: An Introduction M . A R I E L Pragmatics and Grammar R . C A N N , R . K E M P S O N and E . G R E G O RO M I C H E L A K I Semantics: An Introduction to Meaning in Language

Y. M AT R A S Language Contact D . B I B E R and S . C O N R A D Register, Genre and Style L . J E F F R I E S and D . M C I N T Y R E Stylistics R . H U D S O N An Introduction to Word Grammar M . L . M U R P H Y Lexical Meaning J . M . M E I S E L First and Second Language Acquisition T. M C E N E RY and A . H A R D I E Corpus Linguistics: Method, Language and Practice J . S A K E L and D . L . E V E R E T T Linguistic Fieldwork: A Student Guide A . S P E N C E R and A . L U Í S Clitics: An Introduction G . C O R B E T T: Features A . M C M A H O N and R . M C M A H O N: Evolutionary Linguistics B . C L A R K: Relevance Theory B . L O N G P E N G Analyzing Sound Patterns B . DA N C Y G I E R and E . S W E E T S E R Figurative Language J . B Y B E E Language Change S . G . T H O M A S O N Endangered Languages: An Introduction Earlier issues not listed are also available

Endangered Languages An Introduction SARAH G. THOMASON University of Michigan

University Printing House, Cambridge CB2 8BS, United Kingdom Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9780521684538 c Sarah G. Thomason 2015  This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2015 Printed in the United Kingdom by Clays, St Ives plc A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication Data Thomason, Sarah Grey. Endangered languages : an introduction / Sarah G. Thomason, University of Michigan. pages cm. – (Cambridge textbooks in linguistics) Includes bibliographical references and index. ISBN 978-0-521-86573-9 (hardback) – ISBN 978-0-521-68453-8 (paperback) 1. Endangered languages–Case studies. 2. Language obsolescence–Case studies. 3. Languages in contact–Case studies. 4. Linguistic change–Case studies. I. Title. P40.5.E53T46 2015 306.44–dc23 2014045665 ISBN 978-0-521-86573-9 Hardback ISBN 978-0-521-68453-8 Paperback Additional resources for this publication at www.cambridge.org/thomason Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

for Rich

Contents

Preface

page xi

1 Introduction 1.1 1.2 1.3 1.4 1.5

How can we tell when a language is endangered? Where are all the endangered languages? A precondition for language endangerment: language contact Summary and outline of the book’s contents Sources and further readings

2 Why and how languages become endangered 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9

Conquest Economic pressures Melting pots Language politics Attitudes Loss of linguistic diversity via standardization What doesn’t promote endangerment? Summary Sources and further readings

3 Sliding into dormancy: Social processes and linguistic effects 3.1

3.2 3.3 3.4 3.5 3.6 3.7 3.8

Five case studies 3.1.1 Case study #1: Eyak 3.1.2 Case study #2: Cornish 3.1.3 Case study #3: Egyptian 3.1.4 Case study #4: Yaaku 3.1.5 Case study #5: Mednyj Aleut What do these case studies tell us? Tip Semi-speakers and rememberers Attrition An alternative route to language death: grammatical (and lexical) replacement Summary and commentary Sources and further readings

1 4 8 11 12 13

18 19 21 23 24 26 32 35 37 37

42 45 45 46 48 49 50 52 53 54 57 64 66 68

ix

x

Contents

4 What a community loses: Language loss as cultural loss 4.1 4.2 4.3 4.4 4.5

73

Heritage languages and cultural identity Language loss as loss of artistic expression Loss of a language, loss of cultural knowledge Would the world be better off without linguistic diversity? Sources and further readings

74 79 82 85 89

5 What science loses: Language loss as a threat to our understanding of human history, human cognition, and the natural world

94

5.1 5.2 5.3 5.4

Endangered languages and human history Endangered languages and human cognition Endangered languages and knowledge of the natural world Sources and further readings

6 Field research on endangered languages 6.1 6.2

6.3

6.4 6.5

Descriptive linguistics and documentary linguistics On linguistic fieldwork in general (not just on endangered languages) 6.2.1 What is fieldwork? 6.2.2 What is a reasonable time span for a fieldwork project? 6.2.3 Dealing with psychological stress in the field Investigating gravely endangered languages in the field 6.3.1 Working with consultants: access, collaboration, communication, selection, and data-collection techniques 6.3.2 Field sessions 6.3.3 Individual variation 6.3.4 An extended example: dictionary-making Summary and commentary Sources and further readings

7 Revitalizing endangered languages 7.1 7.2 7.3

7.4

Factors that contribute to successful revitalization efforts Some types of revitalization programs From the past to the future 7.3.1 Language revitalization in the digital age 7.3.2 The value of unsuccessful revitalization efforts Sources and further readings

Glossary – languages and terms Bibliography Language index Names index Subject index

94 100 105 107

111 112 115 117 120 122 125 126 135 141 142 145 146

153 155 163 167 167 170 171

175 199 214 219 223

Preface

Like most linguists of a certain age, I came to the topic of language endangerment accidentally. I first began to study the Salish-Pend d’Oreille language in northwestern Montana in 1981, not because it was endangered but because it would provide insights into the Pacific Northwest linguistic area – a large group of northwestern languages belonging primarily to three different Native American language families (Salishan, Wakashan, and Chimakuan) that have come to resemble each other as a result of widespread multilingualism. I expected that understanding one member of the Salishan language family well would eventually help me understand the whole family, including its history, and I could then move on to compare Salishan languages to the languages of the other families in the Pacific Northwest, so that I could discover how all the groups had influenced each other. I still believe all that; but now I also believe that it would take me another century or so to understand this one language thoroughly, and for many years now my scholarly motive for continuing the study has to do with a fascination with the language rather than with the larger intellectual goal of studying language contact phenomena in the region. Meanwhile, another motive has become more and more prominent as I have become more aware of the precarious state of the language: I want to contribute whatever I can to the Salish and Pend d’Oreille tribes’ efforts to preserve and revitalize their language, by documenting as much of it as I can while it is still possible to work with fluent speakers. I estimate that there are now fewer than twenty elderly tribal members who learned their heritage language in infancy and continued speaking it at least through the early decades of their lives. Time is short. Then when I began thinking about endangered languages as a general topic, I realized that my first sustained fieldwork experience, fifty years ago, was also an endangered-language project: my dissertation research was a dialect study based on fieldwork I conducted in the former Yugoslavia. It was an effort to document and analyze a set of word formation patterns in nonstandard dialects, dialects that even then were being eroded through contact with the language then known as Standard Serbo-Croatian. Endangered dialects, endangered languages – they have more in common than one might expect at first glance. In a real sense, my career has been bracketed by concern about the loss of linguistic diversity. Things are different nowadays. Research on endangered languages is a deliberate choice, not an accidental by-product of a project undertaken for other scientific reasons. Young scholars who engage in the research are most xi

xii

Preface

likely to have as their major goal the documentation of languages that might otherwise be lost to their communities and to science without leaving a trace. They often have other goals as well – the investigation of a particular set of grammatical features, for instance – but they also want to preserve knowledge of a unique linguistic system and culture, as well as unique ways of speaking and using language, for the benefit of future generations, both in the heritage community and in the scholarly world. Fieldworkers now also understand that it is unacceptable to visit a community, gather data, and then go away without a second thought. Giving back to the community – a dictionary, a grammar, help with the preparation of language lessons, copies of recordings and of old materials that are not locally available, videos, whatever the community wants and needs – is a standard feature of modern linguistic field research. This too is a major change in expectations for fieldworkers’ responsibilities. In Yugoslavia, back in 1965–66, the idea of giving back to the community never occurred to me, or (I am reasonably sure) to the hospitable dialect speakers I worked with in the villages; today, on the Flathead Reservation in Montana, I concentrate on those aspects of the language that are of the greatest interest to the community (primarily a dictionary, but also analyzed texts), and I also try in other ways to show my respect and gratitude for the priceless opportunity that the elders and the tribes’ Culture Committee have given me to study their language. I was editor of the journal Language when the late Ken Hale, one of the greatest linguists of the twentieth century, organized a symposium on endangered languages at the 1991 annual meeting of the Linguistic Society of America. By then it was clear that language endangerment is an urgent challenge from the viewpoint of speech communities and of linguists (who need to find out about the linguistic treasures in endangered languages). At my request, and following discussions between us that began in 1989, Ken collected papers from his symposium, edited them, and published them as the first article in the first 1992 issue of Language. Nancy Dorian’s earlier endangered-language research, on Scottish Gaelic, was already well known at the time, and her 1989 edited volume Investigating Obsolescence had made a significant impact. But in the United States, at least, it was primarily Ken Hale’s coauthored article collection ‘Endangered languages’ that led to the subsequent outpouring of articles, organizations, and funding initiatives devoted to endangered languages. I shepherded a sizable number of excellent articles to publication in Language during my seven years as editor, but commissioning and publishing ‘Endangered Languages’ is the editorial act I am proudest of. In spite of my long-standing interest in the subject, I have largely been a bystander in the important developments of the last twenty years in endangeredlanguage research; my main efforts in this area have been devoted to my continuing study of the Salish-Pend d’Oreille language. Other scholars and activists have shaped the field: catalogs of endangered languages, theoretical frameworks, technological advances in recording and archiving endangered-language data

Preface

and making it available on the internet, programs for revitalizing endangered languages, and outreach activities aimed at educating the general public about the impending catastrophic decline of global linguistic diversity. My goal in this textbook is to introduce the topic of language endangerment to interested students and other readers new to the subject, by presenting some of the results of the last few decades of activity (by endangered-language communities as well as by scholars). Given my training and career path, the presentation of the material necessarily reflects my perspective as a linguist. I have tried not to lose sight of cultural aspects of language endangerment, and readers should certainly keep them in mind; but coverage of endangered cultures here is unfortunately shallow. The book is meant to be accessible to readers with little or no background in linguistics. I have assumed only basic knowledge of grammatical terms, for instance, “noun”, “verb”, “transitive verb”, “intransitive verb”, “subject”, “object”, “direct object”, and “indirect object”; where other technical linguistic terms are unavoidable, they are defined on their first occurrence in the text and again in the glossary at the end of the book. There are no chapter-bychapter exercises. When I’ve taught courses in endangered languages, I’ve asked each student to choose one endangered language and investigate it from perspectives corresponding roughly to the chapters in this book: How did it become endangered, and why? What are some social processes and (if any) linguistic results of its decline? What has the threatened loss of their language meant to the speech community? What will science lose if this particular language disappears? How well documented is the language, and what preservation/revitalization programs have been undertaken by the speech community? Each chapter ends with a list of sources from which the specific pieces of information in the chapter are drawn, together with suggestions for further reading on the various topics. I acknowledge with gratitude the wonderfully helpful comments that several colleagues have given me while the book was being written and revised: Peter Austin, Nancy Dorian, Nick Emlen, Anna Fenyvesi, Kate Graber, and an anonymous reader for Cambridge University Press. Their suggestions have led to much improvement in the text, but of course none of these people is to blame for remaining errors of fact or interpretation.

xiii

1

Introduction

The post-contact history of Martuthunira is one that has led to their almost complete extinction in little more than a hundred years. Their decline is part of a general pattern which has seen the people of the coastal Pilbara and Ashburton River districts almost completely wiped out while inland groups such as the Panyjima and Yindjibarndi continue to boast thriving communities. The demise of the coastal groups can be attributed both to introduced disease and, perhaps, to a general despair following the complete breakdown of social structure following European settlement. There were only three remaining speakers of Martuthunira [in 1981]...While this grammar of Martuthunira allows the interpretation of the literal meaning of narrative texts and, to a lesser extent, the texts of songs, a full understanding and evaluation will never be possible. Too much of the cultural context which gives them their deeper meaning has been lost. (Alan Dench, 1995) About 97% of the world’s people speak about 4% of the world’s languages; and conversely, about 96% of the world’s languages are spoken by about 3% of the world’s people. (‘Language Vitality and Endangerment’, a report by the UNESCO Ad Hoc Expert Group on Endangered Languages, 2003, citing H. Russell Bernard, 1996:142)

The world’s seven billion people speak about seven thousand languages today. No one knows how many languages have existed over the entire history of humankind, but the number is certainly far higher than seven thousand. For most of human history we have no way of knowing what languages were spoken where or when; direct evidence becomes available only with the invention of writing, or rather with the earliest preserved writings. These come from ancient Sumer, whose language – once the language of a major civilization and empire – has been dead for more than four thousand years. The ancient Egyptians were close behind the Sumerians in developing true writing, but their language too is long dead (although its last descendant, Coptic, lost its final native speakers only a few hundred years ago). Why did Sumerian and Egyptian die? In these two cases the primary reason was that their speakers’ empires were taken over by alien (Semitic-speaking) cultures – Sumer by the Akkadians during the third millennium BCE, Egypt by Arabic speakers in the seventh century CE. Sumerians and Egyptians weren’t massacred wholesale; instead, their speakers eventually found it expedient to shift to other languages, primarily Akkadian and Arabic. It is certain that hundreds, perhaps thousands, of less famous languages have suffered the same fate over the past several thousand years. We have 1

2

introduction

information about a handful of these from ancient written documents, and we have information about many others from modern sources: language death is a worldwide phenomenon, a result of processes that continue at a greatly accelerated rate today. This book is designed to introduce the general topic of language endangerment – all too often a prelude to language death, or at least dormancy – with its potential cultural and scientific consequences, and to describe some methods designed to prevent endangerment from leading to the disappearance of a threatened language. According to most experts’ estimates, at least half of the world’s seven thousand languages will vanish before the end of this century. The loss of linguistic diversity is sometimes compared to the worldwide loss of biodiversity: Jonathan Loh and David Harmon argue in a recent paper, for instance, that ‘the world’s languages, as a group, are more severely threatened than three vertebrate taxa: mammals, birds or reptiles. Languages, globally, are at least as endangered as the most highly threatened vertebrate taxon, the amphibians.’ Pessimists predict the demise of about 90% of currently spoken languages by 2100; optimists say it will only be about 50%. Here are a few figures from the Catalogue of Endangered Languages (EL-Cat): there are currently 3,220 endangered languages (but as I will discuss later, this figure is elastic), 635 extinct or dormant languages, and 100 extinct language families – families of related languages that have all vanished. And a language dies – becomes dormant – every three months or so. In the following chapters I address the obvious questions: What is language endangerment? Why does it happen? How does it happen? Who cares? So what? And what, if anything, can be done about it? A complete map of endangered languages would have markers on every continent (except Antarctica, which has no settled population), as well as on islands scattered throughout the Pacific and other oceans. In fact, UNESCO’s Atlas of the world’s languages in danger fits this description, featuring an interactive map that can be searched by country and by language name. Examination of a global view of that map shows dense clusters of endangered languages in most areas; the only exceptions are places in which (as far as we know) there never were many languages, such as Greenland and parts of northern Asia. As we will see, the causes of language endangerment are diverse. They range from the death of all the speakers (by massacre, starvation, or devastating introduced diseases) to social, political, and economic pressures that lead people to give up their minority language in favor of a majority language. These causes are viewed from a macro-historical perspective; discussion of detailed case studies that sort out culture-specific causes of language shift and endangerment will be limited in this book. Before turning to more specific questions about language endangerment, I should make two general points – or, perhaps a better term, warnings – about the nature of the book. First, I have made no attempt to conceal my personal view that the loss of any language is a disaster. It is certainly a

Introduction

disaster to science. Language (it is often said) is a window into the human mind, and it’s clear that we’ll get more benefit from this language window if we come to understand seven thousand languages fully than if we have only seven hundred to three thousand five hundred languages to comprehend. Every language in the world differs significantly from every other language, so no language is redundant in scientists’ efforts to understand human cognition through understanding the workings of human language. Admittedly, the loss of a community’s language is not necessarily a disaster for that community: not all communities mourn the disappearance of their heritage languages (though their grandchildren often do). But when the heritage language is an indispensable part of the community’s identity, of community members’ sense of self, and an indispensable repository of the community’s culture, losing the language is a disaster indeed. In any case, the large number of language revitalization programs that have sprung up all over the world – programs driven by communities’ eagerness to save their languages – attests to the fact that a great many endangered-language communities do view the loss of their language as a very bad thing. My second point (or warning) is about my fairly frequent use in this book of terms like “language death”, “extinct language”, and “dying language” for the worst fate of an endangered language. I also use “salvage linguistics” to refer to linguistic fieldwork undertaken after the last fluent speakers of the target language have died. The use of such terms is controversial. Many endangeredlanguage activists, both community members and linguists, prefer more benign terms like “sleeping language” or “dormant language” to the grim (apparent) finality of “dead language”. As Nancy Dorian observed recently, as early as the 1970s ‘many language communities found it offensive to have the label “dying language” attached to their ancestral speech form by outsiders who had no direct connection to the ethnic community and often no personal experience of the language at all’. Dorian cites Bernard Perley (a member of the Maliseet First Nation in New Brunswick, Canada) for the argument that metaphors like “dying language” are ‘detrimental to small indigenous communities’ sense of vitality and restorative possibility’. Perhaps my habit of using terms like “dead language” stems from my professional persona as a historical linguist, a student of language change: historical linguists are used to thinking and talking about instances of (mostly permanent) language death through the ages, though in general we are referring to languages that were last spoken several millennia ago rather than just a few decades ago, or yesterday. In any case, both the grim terms and the benign terms are used in this book: I have tried to reserve the grim terms for general discussions and to use the euphemistic terms especially for languages that no longer have any speakers but that are, or are likely to be, targets for revival efforts. To me it makes good sense to consider a language that is being revived to have been dormant rather than absolutely dead. But sadly, some – probably most – languages that lose all their speakers are dead for all time. I hope that readers will not be offended by either set of terms.

3

4

1.1

introduction

How can we tell when a language is endangered?

A language is clearly endangered when it is at risk of vanishing within a generation or two – that is, when its last fluent speakers are elderly, when few or no children are learning it as a first language, and when no one is learning it as a second language. Some experts call a language moribund when it is no longer being learned as a first language: a language that is not being transmitted to younger generations cannot outlive the last generation of native speakers. This has been the fate of many Native languages of North America, for instance, and the same fate threatens most of the remaining indigenous languages of the United States and Canada. The same is true of most of the indigenous languages of Australia and of thousands of other minority languages in most countries all over the world. Some sources give a more elaborately fine-grained classification of language endangerment. Perhaps most prominently, UNESCO’s Atlas of the world’s languages in danger lists six categories of languages according to their status. Here is UNESCO’s Language Vitality and Endangerment Framework: • • • • • •

safe: ‘language is spoken by all generations; intergenerational transmission is uninterrupted’ vulnerable: ‘most children speak the language, but it may be restricted to certain domains (e.g., home)’ definitely endangered: ‘children no longer learn the language as mother tongue in the home’ severely endangered: ‘language is spoken by grandparents and older generation; while the parent generation may understand it, they do not speak it to children or among themselves’ critically endangered: ‘the youngest speakers are grandparents and older, and they speak the language partially and infrequently’ extinct: ‘there are no speakers left’

Other scales of language endangerment can be found in the literature. Lenore Grenoble and Lindsay Whaley, for instance, propose a six-way scale that is similar (though not identical) to the UNESCO framework: Safe; At Risk; Disappearing; Moribund; Nearly Extinct; and Extinct. A list of this type, though clearly very useful, is necessarily too condensed to cover the territory thoroughly, so let’s look more closely at ways of gauging a language’s status. The most obvious criterion for determining whether or not a language is endangered – the overall number of speakers – turns out to be useless as a predictor unless there is evidence of changes in speaker numbers over time. Most readers of this book will be most familiar with one or more major languages that have millions of native speakers and, in a few cases, many more millions of nonnative speakers. English, for instance. But languages of small communities are not necessarily endangered; some language communities with

How can we tell when a language is endangered?

just a few hundred speakers have held steady at that size for centuries or even millennia. The Pirahã language of Amazonia in Brazil is an example: in 1977, the community had, and as far as experts can tell had always had, about 200 members, all of whom spoke Pirahã and only Pirahã; since then the community has grown, so that there are now about 750 members, but most of them are still (or were until very recently) essentially monolingual. Until the last decade or so, some community members knew a few words and phrases of Portuguese, but none had any proficiency in Portuguese or any other language besides Pirahã. Pirahã is therefore in no danger of being lost in this community, although the community itself, under pressure from mainstream Brazilian society, is in a precarious situation. It is probably safe to say that the language will disappear only when and if the community disappears. Nor does a large speaker population necessarily guarantee safety from endangerment: see the discussion of Navajo immediately following. Speaker age is a more reliable indicator of endangerment: the more young speakers there are – the more children there are who are learning the language as a first language – the less likely the language is to disappear any time soon. Unfortunately, for many of the world’s languages we have no evidence of speakers’ ages, so it is often difficult or even impossible to apply this criterion. There are, as the various scales of endangerment indicate, degrees of endangerment. Even languages that have many thousands of speakers and are still acquired as first languages may reasonably be considered endangered if dwindling numbers of children learn them in each generation. The Navajo language of Arizona, New Mexico, and Utah is a case in point. In 1990, the United States Census Bureau reported 148,530 Navajo speakers (according to the Ethnologue), out of a tribal population of 219,198; 7,616 of these Navajo speakers were reported as monolingual in Navajo. This is an impressively large number of speakers for any Native language of North America, though the lack of information about level of fluency makes it difficult to interpret the significance of the figure. Still, especially compared with the many Native North American languages whose speaker numbers are in the double or even single digits, the large number of Navajo speakers makes the language look robust. Unfortunately, the evidence of child language acquisition tells a different story. A 1969 study found that 95% of Navajo six-year-olds spoke Navajo at that time; but in a 1992 study of 682 Navajo children in Head Start preschool programs, more than half of the children (54%) were reported to be monolingual in English, and as of 1998 only 30% of six-year-old children spoke Navajo as their first language. This sharp decline in first-language acquisition of Navajo is reflected in other studies too. The rapidly decreasing number of first-language learners of Navajo makes it clear that the language is in danger of disappearing within a few generations. And if even Navajo – which in 1990 had more speakers than any other indigenous North American language – isn’t safe, every other indigenous language of North America is also at risk.

5

6

introduction

Similarly, it is impossible to compile a definitive list of endangered languages for the world or any sizable part of it. Speaker numbers do not provide conclusive evidence, both because they tend to come from census reports (which have selfreport data and which rarely include questions about degree of fluency) and because they provide little or no information about the age of the speakers. Moreover, vigorous revitalization efforts have taken root in many communities, and some of them have achieved at least a partial reversal of language decline. The most spectacular success story is also the oldest one: Hebrew, dead for almost two thousand years as a language of everyday communication, was revived during the twentieth century to become the dominant language of modern Israel. It has now been learned by several generations of Israeli children as a first language, and although it is hardly the same as the Biblical Hebrew that formed its main foundation, it is nevertheless a recognizable offshoot of the ancient language. Among more limited recent success stories is that of Maori, a Polynesian language that is now one of the official languages of New Zealand. The Maori people have invested heavily in educational and other organizations designed to support their language, including preschool language nests and Maori-language immersion elementary schools; in recent decades, largely or entirely as a result of their activism, the government of New Zealand has also invested heavily in programs designed to reverse the sharp decline of Maori. The results are mixed. The perceived status of Maori is still relatively low, in part because of the urbanization of ethnic Maoris and especially because of the overwhelming dominance of English in New Zealand’s political, economic, and social life. The number of Maori speakers has increased since the 1960s and, crucially, that number now includes younger people. Bernard Spolsky, writing in 2009, said that ‘M¯aori looks like one of the few threatened languages that can be confident about its next hundred years.’ This optimistic view suggests that the increasing prominence of preservation and revitalization programs around the world will (with considerable luck) point to improvement in the condition of some languages that are now endangered, ideally to the point of long-term stability. Another complication in any effort to define ‘endangered language’ concerns the status of endangered dialects. The difficulty is most evident in the case of dialects that are spoken outside a language community’s traditional homeland. Hungarian, for instance, is obviously not an endangered language: it has millions of speakers in Hungary. But Hungarian immigrants in the United States are shifting to English, so that all varieties of American Hungarian are moribund. Even within a language’s home territory, dialects – especially nonstandard dialects – often disappear; in many countries with mass education in a dominant standard dialect (offically or unofficially acknowledged as the “correct” form of the language), nonstandard dialects have vanished or are vanishing rapidly. This is one of the primary effects of standardization: all dialects other than the newly standard one are likely to suffer. To most people, linguists as well as laymen, the loss of a dialect seems less important than the loss of an entire language.

How can we tell when a language is endangered?

But dialects, like whole languages, are carriers of culture. A community that loses its unique dialect inevitably loses a part of its culture and its local identity within the larger community; and loss of a dialect, like the loss of a language, robs science of potential insights into the workings of the human mind. (This isn’t the place to enter into a detailed discussion of the distinction between two dialects of the same language and two separate but closely related languages. Briefly, the only viable linguistic criterion for making the distinction has to do with mutual intelligibility: if there is some, they’re dialects; if there is none, they’re separate languages. But often this common-sense criterion doesn’t give a straightforward answer because of various sociopolitical considerations that affect perceived intelligibility; and the saying ‘A language is a dialect with an army and a navy’, though definitely not universally applicable, is not wholly invalid.) Finally, the criterion that is most useful in deciding whether a language is endangered or not – the ages of the speakers – does not apply to all languages. Pidgin languages, which are by definition learned and used solely as second languages and for limited purposes (such as trade with outsiders), have no native speakers: no children learn them as first languages. It’s true that most known pidgins are now endangered, if they have survived at all. But the reason is unconnected with the age of learning; it has to do instead with changing worldwide economic and social conditions. The Age of Exploration, during which pidgins sprang up on African, Asian, and American coasts when Europeans began visiting them, ended several centuries ago, as did the earlier period of Arab expansion, which also gave rise to a number of pidgins. Pidgin languages have been reported from other contexts as well, such as the pidgin that was once used by (Native American) slaves of the Nez Perce tribe of the Columbia River plateau in northwestern North America. Populations are more settled nowadays, and most pidgins are being, or have been, replaced by multilingualism and/or shift to the languages of former colonial powers. The other fate of a pidgin language is development into a creole language – that is, the acquisition in a settled multiethnic community of a native-speaker population, until the language has so many native speakers that it can no longer be considered a pidgin. This is what happened to Tok Pisin, now one of the official languages of Papua New Guinea: it started life in the nineteenth century as a pidgin (the name literally translates as ‘talk pidgin’), with a vocabulary drawn primarily from English and a compromise grammar based mainly on indigenous Austronesian languages. It became so embedded in the life of the country, including increasing use as a home language when speakers of different indigenous languages married, that it is now a creole. By contrast, the seventeenth-century pidgin that was used between Delaware (Lenape) Indians and European missionaries and settlers from several different countries is long dead, and Lenape itself is gravely endangered. Other languages that fall outside the scope of endangered-language designation are ancient languages that survive only in limited domains of usage. Most of these are religious languages, dead for centuries or even millennia

7

8

introduction

as languages of everyday communication but preserved in religious contexts. Hebrew was one of these, but no longer. Latin, Classical Arabic, and Coptic – the last descendant of Ancient Egyptian, whose latest known use as an ordinary spoken language was sometime between the fourteenth and seventeenth centuries (see Chapter 3 for further discussion) – are examples. Sanskrit is a borderline case: it is the ancient sacred language of the Hindu religion, and as such it is used in religious contexts and codified in a form that was essentially fixed in the fourth century BCE. But the 1991 Census of India reported 49,736 fluent speakers of Sanskrit (in self-report data), and efforts are being made to revive it as a spoken language. There are several “Sanskrit villages” in India where Sanskrit is the everyday language of communication; for instance, in the state of Karnataka, in the village Mattur (also spelled Mathur or Mathoor), the entire population is said to be fluent in Sanskrit. Given the fact that several hundred other spoken languages compete with Sanskrit in India – almost thirty of them with more than a million speakers each – it is hardly likely that Sanskrit will ever become the country’s major spoken language; the conditions that favored the remarkable rise of Modern Hebrew (a new nation established by a beleaguered people, a linguistically mixed population uprooted from their homes and with no common language) do not exist in India. Still, if we exclude pidgins and ancient languages preserved in religious contexts, the age of the speakers is generally a good indicator of a language’s prospects for long-term viability. As a typical example, consider Veps, a Finnic language spoken in northwestern Russia, not far from Lake Ladoga. In 1926, 94.7% of ethnic Veps people spoke Veps as their first language; by the 1960s the percentage was down to about 60%, and it has continued to decline ever since. Nowadays no children are learning Veps as a first language, and there are few or no young people in the villages where Veps is still spoken. The language has no official status. Although legally it can be used in schools, it is mainly ‘taught as a hobby school subject or as a second foreign language twice a week’. Unless current revitalization efforts succeed in reversing its course, the language will soon disappear.

1.2

Where are all the endangered languages?

The short answer to this question is, everywhere. Endangered languages are found on every continent (except Antarctica). Counting endangered languages is as difficult as defining the category, mainly because far too little relevant information is available for many parts of the world. At best we have speaker figures from national census reports; but even these, as noted earlier in this chapter, are unreliable as indicators of a language’s status as endangered or not. The most comprehensive source that gives speaker numbers for all known languages is the Ethnologue (www.ethnologue.com), a product of the missionary

Where are all the endangered languages?

organization SIL International. The speaker numbers in the Ethnologue are often outdated, and any figures are bound to be approximate in any case; still, even with its inevitable limitations, this is an invaluable resource for anyone who wants to find out about the location(s) and general demographics of any language. One searchable category on the Ethnologue website is “nearly extinct languages”, which includes all languages that are spoken by just a few elderly people. The page lists about 516 of these in 58 different countries (the language count is approximate because the entries are organized by country and some of the languages are spoken in more than one country). In other words, according to the Ethnologue figures, more than 500 of the world’s approximately 7,000 languages are about to disappear; in fact, many have surely already disappeared, because some of the Ethnologue’s sources are twenty or thirty years old. Of the 516 nearly extinct languages, 46 are in Africa, 170 in the Americas, 78 in Asia, 12 in Europe, and 210 in the Pacific. The latter region includes both Australia (with 168 languages – but this is an exaggerated figure, as we will see) and Oceania (42 languages). These counts do not include nearly extinct dialects of languages in which one or more dialects are not nearly extinct. Australia is the country with by far the largest number of languages in the Ethnologue’s “nearly extinct” category, followed by the United States (68), Indonesia (31), and Brazil (30). The two countries in Africa with the largest numbers of nearly extinct languages are Nigeria (15) and Cameroon (13); in Europe only one country, Russia, has more than a single nearly extinct language, and Russia has only two. These figures of course do not include languages that were extinct before the Ethnologue was compiled, and we know that there were many of these in North America, Australia, and probably also Russia, as well as in most or all other parts of the world. Let’s look more closely at the status of Australian languages, using estimates for speaker numbers that are more recent than the Ethnologue figures, which date from 1970 to 1983. The starting point would be the number of Indigenous languages (including both Aboriginal and Torres Strait Islander languages) that are estimated to have been spoken before Whites arrived in Australia: most estimates give a figure between 200 and 300. By 1990, 64% of these languages were either extinct or reduced to a handful of elderly speakers; 28% were “severely threatened”, and a mere 8% were “relatively healthy”. As of 1990, 20 Indigenous Australian languages were considered “strong” (according to the criterion “spoken by all age groups regularly”); by 2001, three of those twenty were endangered. At the other end of the scale, 35 Indigenous languages were still spoken in 1990 by just a few elderly people each, and by the turn of the millennium all those languages had probably become extinct. The number of surviving Indigenous languages in 2000 would then be about 90. According to 1990 figures, 8 Indigenous languages still had 1,000 or more speakers each; the largest speaker counts were for Central Torres Strait (3,000–4,000 speakers), Arrernte (3,000+), Pitjantjatjara (3,000+), and Warlpiri (3,000+). By now the speaker numbers for most or all of these eight languages are likely to be lower,

9

10

introduction

perhaps much lower. In their 2001 survey of the state of Indigenous Australian languages, Patrick McConvell and Nicholas Thieberger emphasize the dramatic ‘decrease in the percentage of Indigenous people speaking Indigenous languages from 100% in 1800 to 13% in 1996’, a trend that was especially rapid in the decade from 1986 to 1996. They comment, ‘If these trends continue unchecked, by 2050 there will no longer be any Indigenous languages spoken in Australia.’ Even if the rate at which speakers and languages are lost diminishes (as McConvell and Thieberger believe it will), the prospects for the long-term survival of Australian Indigenous languages – which are not only the carriers of Indigenous cultures but also display some linguistic features that are totally unknown elsewhere in the world – are bleak. Another sobering aspect of the Ethnologue’s list of nearly extinct languages is that it includes nine of the forty languages classified by the Ethnologue as isolates – that is, languages that are not known to be related to any other language: Itonama and Leco (Bolivia), Muniche and Taushiro (Peru), Puelche (Argentina), Yámana (Chile), Kutenai (U.S. and Canada), Yuchi (U.S.), and Ainu (with fifteen speakers in Japan as of 1996; the last speaker of the Sakhalin dialect of Ainu died in 1994). To this list one might add the Lenca language of El Salvador, which is listed in the Ethnologue as ‘unclassified’, and some other unclassified languages are also likely to be isolates. If an isolate dies, its linguistic and cultural riches are lost for all time: it leaves no relatives that might continue some of its rarest and most interesting features. The Australian case is perhaps the most striking example of sweeping loss, both in terms of the overall number of extinct and nearly extinct languages and in terms of the potential loss to science and the broader study of human culture and society – of information vital to the construction of viable theories of universal aspects of language structure and of information about songs, narratives, kinship systems, and other products of individual and social creativity. Australia is hardly the only example, however; other groups of languages are in a comparably dismal situation. To give just one example, the Salishan language family of the U.S. and Canada has, or had, twenty-three known languages. Of these, eight are listed by the Ethnologue as nearly extinct (Bella Coola, Clallam, Coeur d’Alene, Lushootseed, Sechelt, Snohomish, Squamish, and Straits Salish), and seven may already be extinct (Lower Chehalis, Upper Chehalis, Cowlitz, Nooksack, Quinault, Tillamook, and Twana). That leaves just eight languages in the family that are in neither of these end-stage categories (Comox-Sliammon, Halkomelem, Shuswap, Lillooet, Thompson, Colville-Okanagan, Columbian, and Montana Salish-Kalispel-Spokane), all of which are “merely” gravely endangered. Three of the family’s five branches are in great danger of dying out entirely (the Bella Coola and Tillamook branches, each comprising just one language, and the four-language Tsamosan branch). The numbers and distributions of nearly extinct languages are of course just the tip of the iceberg: overall estimates of endangered languages are much higher. The only possible conclusion is that language endangerment is a pervasive

A precondition for language endangerment: language contact

worldwide phenomenon. This is a disastrous situation from the viewpoints of communities whose heritage languages are being lost and of science. Whether it is a disaster from the viewpoint of human society as a whole is of course a matter of opinion. But there is no doubt at all that the loss of hundreds and even thousands of languages, often including all representatives of ancient linguistic lineages, greatly reduces our chances of arriving at a more complete understanding of human language in all its complexity. The search for universals of human language has been pursued vigorously for decades now, but it has by no means ended; and, as already mentioned, many endangered languages have features that are rare or even unique. If, for instance, the handful of known click languages of southern Africa and Tanzania had disappeared before linguists found out about them, their various click sounds might well have been considered an impossible linguistic feature.

1.3

A precondition for language endangerment: language contact

With a very few known exceptions – cases where all the speakers of a language die in the absence of any event(s) triggered by interactions with another speech community – language endangerment is always preceded by language contact. That is, the usual route to language endangerment and death is shift by the members of one speech community to the language of another speech community; and even in the worst case, where all the speakers die and take their language with them, contact with other peoples typically precedes the final stage (such as when invaders introduce diseases to which an indigenous population lacks immunity). Where and when do languages come into contact? The short answer is everywhere and at all periods of documented and inferred human history. Probably most people in the world today are bilingual or multilingual, and there is no modern nation that harbors just one spoken language. Nor is this a modern phenomenon: language contact was a feature of all the ancient civilizations for which we have direct evidence in the form of written records. To take just one example, the Hittite empire (ca. 1750–1200 BCE) is known to have been a multilingual state in which subordinate peoples kept their own languages and maintained their own cultures. Language contact and the bilingualism/multilingualism that it typically produces are also pervasive and persistent features of modern societies around the world. When two speech communities come into contact and interact regularly, there are likely to be consequences for one or both languages. In particular, the more intense the contact is, the more likely people are to become bilingual; and while bilingualism is by no means a predictor of the future loss of a community’s heritage language, it is often a first step in that direction. This is especially obvious when the social, political, and economic relations between the two

11

12

introduction

communities in contact are markedly asymmetrical: in such cases bilingualism will also probably be asymmetrical, and a dominant group’s language very frequently replaces a subordinate group’s language. So, for instance, the arrival of Europeans on North American soil led to a severely asymmetrical power relationship, and all Native American languages (at least in the United States) eventually became endangered as a result, to the point where many have vanished in the last few centuries and most others are teetering on the edge. First their speakers learned English, sometimes under strong pressure from forced boarding-school and other government-promoted programs; then, as English became more and more important in their daily lives, Native American children ceased learning their heritage languages. With the obvious exception of English, the vast majority of immigrant languages in the U.S. have met a similar fate. Spanish is maintained in certain communities, but other immigrant languages have usually been abandoned after three or four generations. The classic three-generation rule predicts the following sequence: adult immigrants tend not to learn English well; their children are bilingual in English and their parents’ language; but the third generation – the original immigrants’ grandchildren – are monolingual English speakers. (Spanish in the U.S. is a special case, since many Hispanic U.S. citizens are neither immigrants nor descended from immigrants: they are instead descended from Mexicans who were living in what is now the U.S. Southwest when the region was taken by force from Mexico during the nineteenth century.) It must be emphasized, however, that in many other settings all over the world language contact is a steady-state phenomenon, with flourishing bilingualism and multilingualism and no tendency for one group to switch to another group’s language. Contact situations that feature stable bilingualism tend to lack extreme asymmetry in power relations between the two speech communities, and bilingualism is likely to be mutual rather than (as in the U.S.) one-sided. This book is not about language contact per se, but contact is nevertheless a constant theme that runs at least implicitly through every discussion of language endangerment. Finally, it’s worth keeping in mind that there are only two outcomes of language contact that don’t simply result in more language contact: elimination of the contact by migration or other separation, or language death. Only the latter outcome has direct relevance for the topic of this book.

1.4

Summary and outline of the book’s contents

This brief introduction has surveyed the what and the where of language endangerment, discussed the depth of the problem (without firm numbers because a definitive list cannot be compiled), and highlighted language contact, a (nearly?) universal precondition for endangerment. In the next chapter we will move on to the how and the why of endangerment – the complex networks

Sources and further readings

of reasons for it and how they cause the decline of a community’s language. Chapter 3 then focuses on the ultimate grim fate of most endangered languages: language death, or at best dormancy, in both its social and its linguistic aspects. In Chapter 4 I turn to a consideration of language loss as cultural loss, and Chapter 5 presents a scientist’s perspective, discussing language loss as a threat to the prospects for a comprehensive understanding of human history, human cognition, and information about the natural world. Chapter 6, the longest chapter, is devoted to the rewards and difficulties of documenting endangered languages, which offer special challenges to the fieldworker. In Chapter 7, turning finally to the brighter side of the general topic of language endangerment, I examine some methods of language revitalization and the (often limited) opportunities they offer for long-term success. The book closes with a glossary of technical terms that readers will encounter. A list of some important electronic resources for students and communities whose languages are threatened will be found on the publisher’s website for the book rather than in the printed book.

1.5

Sources and further readings

An important group of articles, collected and edited by the late Ken Hale and published as a set, entitled ‘Endangered Languages’, in Language in 1992, is often mentioned (at least in the United States) as the first highly visible general English-language publication that sounded the alarm about language endangerment. One article in this set is cited especially frequently: Michael Krauss’s ‘The world’s languages in crisis’. The authors of the other five articles in the set are Hale, Lucille J. Watahomigie and Akira Y. Yamamoto, Colette Craig, Laverne Masayesva Jeanne, and Nora C. England. Two short discussion pieces appeared in the same journal in 1992 and 1993, respectively, offering sharply differing views of the issues raised by Hale et al.: Peter Ladefoged’s ‘Another view of endangered languages’ and Nancy Dorian’s ‘A response to Ladefoged’s other view of endangered languages’. Nancy Dorian is the most prominent of the pioneering authors in both the linguistic and the sociolinguistic study of language obsolescence and death; her work has been extremely influential, from her earliest publications, such as her 1973 article ‘Grammatical change in a dying dialect’, to her 1981 book Language death: the life cycle of a Scottish Gaelic Dialect and important later articles and books. She also edited a 1989 book that may be the most widely cited work in the field of language endangerment, Investigating obsolescence: studies in language contraction and death. Another widely cited early book is Endangered languages (1991), edited by R.H. Robins and Eugenius M. Uhlenbeck. Since 1992, stimulated in part by Hale et al. 1992, the number of books about language endangerment and death has grown rapidly. A few of the most important of these – several of which have been translated into a number of other

13

14

introduction

languages – are Lenore A. Grenoble and Lindsay J. Whaley’s coedited 1998 volume Endangered languages: current issues and future prospects and their coauthored 2006 book Saving languages: an introduction to language revitalization; Daniel Nettle and Suzanne Romaine’s coauthored 2000 book Vanishing voices; David Harrison’s 2007 book When languages die: the extinction of the world’s languages and the erosion of human knowledge; Peter Austin’s 2008 book One thousand languages: living, endangered, and lost; Nicholas Evans’ 2010 book Dying words: endangered languages and what they have to tell us; Matthias Brenzinger’s 1992 edited volume Language death: factual and theoretical explorations with special reference to East Africa and his 1998 edited volume Endangered languages in Africa; David Bradley and Maya Bradley’s coedited volume Language endangerment and language maintenance (2002); Flores Farfán et al.’s coedited volume New perspectives on endangered languages (2010); and a book written for a nonspecialist audience, David Crystal’s Language death (2000). At least one handbook on endangered languages has also appeared, The Cambridge handbook of endangered languages, edited by Peter K. Austin and Julia Sallabank (2011). A bibliography, ‘Endangered Languages’, compiled by Chris Rogers and Lyle Campbell, appeared in the Oxford Bibliographies series in 2012. Among the many articles on the general topic, I would recommend Lyle Campbell’s 1996 article ‘Defense strategies for endangered languages’ as an excellent introduction. There is now an open-access electronic journal, Language documentation and conservation, with a primary focus on endangered languages; there is also an annual series of volumes on endangered languages edited by Peter K. Austin, Language documentation and description, published by the Hans Rausing Endangered Languages Project at the School of Oriental and African Studies, University of London. The Linguistic Society of America (LSA) established its Committee on Endangered Languages and their Preservation in 1992 and, in 2002, its annual Kenneth L. Hale Award (in honor of Ken Hale, who died in 2001) for ‘scholars who have done outstanding work on the documentation of a particular language or family of languages that is endangered or no longer spoken’. In 2004, as another response to the urgency of recording endangered languages before they disappear, the LSA endorsed a policy statement on ‘The need for the documentation of linguistic diversity’: www.linguisticsociety.org/files/lsa-stmt-documentation-linguistic-diversity.pdf Web sources on language endangerment have also increased dramatically over the past decade and more. One of the most important sources is UNESCO’s Atlas of the world’s languages in danger, formerly called the Red Book of endangered languages: www.unesco.org/culture/languages-atlas/. This is an attempt to list all of the world’s endangered languages; as I explain in this chapter, I don’t think that goal can be reached, but the effort, together with the resulting list, is immensely valuable for students, scholars, and laypeople interested in the general topic of language endangerment. Here are a few

Sources and further readings

other especially important websites: the Endangered Languages Project (www. endangeredlanguages.com), ‘an online resource to record, access, and share samples of and research on endangered languages as well as to share advice and best practices for those working to document or strengthen languages under threat’; DOBES (Documentation of Endangered Languages, dobes.mpi.nl), an archive that ‘contains language documentation data from a great variety of languages from around the world that are in danger of becoming extinct’; the Hans Rausing Endangered Languages Project at SOAS, University of London (www.hrelp. org), which ‘supports research, training, and archiving for endangered languages throughout the world’; and OLAC, the Open Languages Archives Community (www.language-archives.org), ‘an international partnership of institutions and individuals who are creating a worldwide virtual library of language resources by: (i) developing consensus on best current practice for the digital archiving of language resources, and (ii) developing a network of interoperating repositories and services for housing and accessing such resources’. See this book’s web page for a longer list of websites. The quotation about Martuthunira that begins this chapter is from pp. 17 and 21 of Alan Dench’s 1995 grammar, Martuthunira: a language of the Pilbara region of Western Australia. The second quotation at the beginning of the chapter is from the 2003 report of the Ad Hoc Expert Group on Endangered Languages; eleven Expert Group members contributed to the report (their names, in alphabetical order beginning with Matthias Brenzinger, are listed in this book’s Bibliography). They in turn are quoting percentages from p. 142 of H. Russell Bernard’s 1996 article ‘Language preservation and publishing’. The paper comparing the loss of biodiversity to the loss of languages is Jonathan Loh and David Harmon’s 2014 conference paper ‘Comparing status and trends in linguistic and biological diversity’; the quotation is from the paper’s abstract. A report on their work by John Vidal, entitled ‘As forests are cleared and species vanish, there’s one other loss: a world of languages’, was published in The Guardian on June 8, 2014: www.theguardian.com/environment/2014/jun/ 08/why-we-are-losing-a-world-of-languages (accessed 23 June 2014). The EL-Cat figures about endangered languages – 3,220 endangered, 635 extinct, and so on – are from a talk entitled ‘New knowledge: findings from the Catalogue of endangered languages’ (“EL-Cat”), presented at the annual meeting of the Linguistic Society of America in January 2014 by Lyle Campbell et al. Nancy Dorian’s observation about negative reactions to terms like ‘dying language’ and ‘language death’ are from the introduction to her 2014 book Smalllanguage fates and prospects: lessons of persistence and change from endangered languages: collected essays, and in this passage she points in particular to Jane Hill’s important 2002 article “‘Expert rhetorics” in advocacy for endangered languages: who is listening, and what do they hear?’ The quotation characterizing Bernard Perley’s argument in favor of avoiding ‘dying language’

15

16

introduction

and the rest is from Dorian’s introduction; the original source is Bernard C. Perley’s 2012 article ‘Zombie linguistics: experts, endangered languages and the curse of undead voices’: http://linguistics.berkeley/edu/fforum/readings/perley_ zombielinguistics_2012.pdf (accessed May 2014). Definitions of endangered languages are easy to find in the literature (see, for instance, the books listed on the previous page) and on the web. Here, to give just one of many examples, is Wikipedia’s definition (as of May 2014): ‘An endangered language is a language that is at risk of falling out of use as its speakers die out or shift to speaking another language. Language loss occurs when the language has no more native speakers, and becomes a “dead language”. If eventually no one speaks the language at all, it becomes an “extinct language”’. (I’d dispute the second sentence: there are non-dead languages, notably pidgins, that are spoken solely as second, or at least non-first, languages.) The pidgin language spoken by slaves of the Nez Perce was mentioned on p. 490 of the 1944 book Ka-Mi-Akin: Last Hero of the Yakimas (2nd edition), by A.J. Splawn. UNESCO’s Atlas of the world’s languages in danger, with its categories of language status and its interactive atlas, can be found at www.unesco.org/culture/en/endangeredlanguages/atlas. Grenoble and Whaley’s scale of language endangerment is on p. 18 of their 2006 book Saving languages: an introduction to language revitalization. The Pirahã community size and speaker numbers were provided by Daniel L. Everett (personal communication, 2009, 2012). The information about Navajo speaker numbers and first-language acquirers comes from several sources: Bernard’s Spolsky’s 1969 report The Navajo reading study; Paul R. Platero’s 1992 report Navajo Head Start language study, the Ethnologue; Tiffany R. Lee and Daniel McLaughlin’s 2001 article ‘Reversing Navajo language shift, revisited’; and Victor Golla’s 2007 endangered-languages encyclopedia article ‘North America’(which gives the approximate figure of 120,000 speakers). For further discussion of the Navajo situation, see also Spolsky’s 2009 article ‘Language management for endangered languages: the case of Navajo’. Lee and McLaughlin say that Navajo tribal members are shifting to English ‘with extraordinary speed’ (2001:24). Comments on the Maori revitalization program come from Rangi Nicholson’s 1997 article ‘Marketing the Maori language’, Spolsky’s 2009 article ‘Rescuing Maori: the last 40 years’ – the optimistic quotation about the future of Maori is from p. 32 of that article – and Jeanette King’s 2014 article ‘Revitalizing the M¯aori language?’ Two articles about Mattur (Mathur, Mathoor) as a “Sanskrit village” are Subha J. Rao’s ‘Keeping Sanskrit alive’ (The Hindu, 2 March 2008) and S. Kushala’s ‘This village speaks gods [sic] language’ (The Times of India, 13 August 2005). The information about the current status of Veps is from a talk given by Ulrikka Puura, ‘Searching for aspectuality in Finnic: how to interpret morphosyntactic change in Vepsian?’, presented at the Twelfth International Conference on Minority Languages, Tartu, Estonia, in May 2009. Veps use

Sources and further readings

in local schools is described by Evgenia Romanova in her 2007 University of Tromsø M.A. thesis, The process of revitalization of culture and indigenous ethnic identity: the case of the Vepsian people in Karelia. Her thesis can be found at www.ub.uit.no/munin/bitstream/10037/1156/4/thesis.pdf (accessed 7 September 2009). The information about the state of Indigenous Australian languages is from Patrick McConvell and Nicholas Thieberger’s report ‘State of Indigenous languages in Australia – 2001’; the quotation is from p. 2 of their report. The 1990 speaker estimates that McConvell and Thieberger cite are from Annette Schmidt’s 1990 report ‘The loss of Australia’s Aboriginal language heritage’. See also Graham McKay’s 2011 report on Australian language policies, with a brief update on the state of the languages. The “three-generation rule” was first described by Einar Haugen in his 1953 book The Norwegian language in America, a classic study of one immigrant community in the United States. A summary and update of Haugen’s 1953 results can be found in Haugen 1989.

17

2

Why and how languages become endangered

The Setos are linguistically close to the southeastern Estonians ... but are distinctly different in their customs, culture and Orthodox belief. In their territory, called Setomaa, the Russian language and culture had prevailed, with school education and administration in Russian, until 1920. Under the Tartu Peace Treaty (1920) the entire territory of the Setos became part of the Republic of Estonia. In the first years of independence, estonisation of Setos was initiated. ... During the Soviet occupation the position of the Seto language weakened even more, particularly after the 1960s. This was caused by weakening the village community, campaigning against the Seto language at schools, establishment of Estonian-language kindergartens, fast expansion of media in the Estonian language ... and out-migration of young people. (Estonian Bureau for Lesser-Used Languages, 2009)

As far as we can tell, the most obviously tragic routes to language death – genocide and fatal diseases introduced by foreigners – are rare in human history. Most languages become endangered through processes that are fatal only to the languages, not to their speakers. In this chapter we explore social, economic, and political causes of endangerment, among them the prestige of a victorious invader’s language; the expedience of pleasing conquerors by speaking their language; the need to know a particular language in order to get a better job, or any job; ethnically mixed school systems and workplaces that promote linguistic assimilation; and government efforts to force assimilation to a dominant culture, as seen for instance in the residential schools in the U.S., Canada, and Australia that actively suppressed indigenous languages. In the penultimate section I discuss the effects of standardizing a language on overall linguistic diversity – in particular the decline and loss of newly nonstandard dialects. Finally, I discuss factors that help to preserve a language from endangerment, most prominently official state support, economic power of the speakers, large numbers of speakers of all ages (and without noticeable reduction over time), and language loyalty. Two caveats are needed at the outset. First, endangerment never, or almost never, has a single cause. In discussing causal factors individually in this chapter I do not mean to suggest that any one of them has ever been the sole cause of endangerment in any contact situation; in almost all cases the factors contributing to endangerment are complex and interrelated. And second, as anthropologists discovered long ago, the macro-historical factors that cause language shift and endangerment (that is, the factors that I discuss in this chapter) are ‘ultimate, but not proximate explanations – the question of why people actually shift to another language is based on very local social and cultural interpretations of 18

Conquest

such changes’. So, for instance, the anthropologist Don Kulick, investigating causes and processes of language shift in the village of Gapun in Papua New Guinea, found only an indirect connection between the local values that drove the shift from Taiap to Tok Pisin (the major lingua franca in PNG) and the macro-economic, social, and political factors that obtained in the area.

2.1

Conquest

Conquest is a, or the, cause of language endangerment and death if the conquerors’ language replaces the language(s) of the conquered. This has happened in some colonial settings, among them the Americas and Australia: all indigenous North American languages, most indigenous Australian languages, and many indigenous languages of Central and South America are endangered, and presumably fewer of them would be if the regions hadn’t been conquered by Europeans. The vast cultural changes triggered by Anglo and Hispanic invasions of these continents made it useful and then necessary for indigenous peoples to learn the invaders’ languages and eventually shift to those languages, even where there were no government policies designed to force assimilation to the newly dominant culture and language. In speculating about the probable fates of indigenous languages if there had been no European invasions and settlements we need to be cautious, because of the untestable assumption that the Native languages wouldn’t have become endangered if the western hemisphere and Australia hadn’t been largely taken over by Europeans. A few Native empires emerged in MesoAmerica and South America before Europeans arrived on the scene, and at least one of these, the Aztec Empire in what is now Mexico, featured aggressive warfare against neighbors; it is possible that, given more time, the Aztecs’ language Nahuatl would have replaced other indigenous languages. (In fact, Quechua, the language of the Inca Empire, did completely replace numerous other indigenous languages of Peru and neighboring countries; but that effect was due to post-Conquest Spanish policy, not to pre-Conquest Inca policies.) Similarly, the expansion of Russian power into Siberia, which was more or less complete by the mid-seventeenth century, ultimately led to the endangerment of many or most of the indigenous languages of the region, including among others Chukchi, Central Siberian Yupik, Buryat, and Even, which are classified as ‘severely endangered’ on UNESCO’s Language Vitality and Endangerment scale. Conquest on a scale smaller than continent-wide invasion often has a similar effect. One well-known example, discussed in a bit more detail in Chapter 3, is Egyptian, which was eventually replaced by Arabic some centuries after the Muslim conquest of Egypt in 641 CE. Most examples from the past, however, are lost to history because both the languages that vanished and the languages of the conquerors were unwritten.

19

20

why and how languages become endangered

In this and most other cases of endangerment via conquest, a complex mix of political, social, and economic factors, together with hard-to-analyze attitudinal factors, contribute to language decline. It certainly isn’t just one factor that is causing indigenous languages of the Americas to disappear. Many indigenous people died from exposure to new diseases for which they had no inherited immunity. Most such deaths (one hopes) were tragic accidents, for instance when smallpox killed 90% of the speakers of Lower (Shoalwater) Chinook at the mouth of the Columbia River in the Northwest. There have been claims that Whites sometimes deliberately caused smallpox epidemics among Native Americans by giving them blankets infected with smallpox; some of these claims are certainly fabricated, but at least one genuine incident does appear to have been planned and perhaps implemented. A similarly distressing example is the series of events that led to the demise of Susquehannock, an Iroquoian language once spoken in eastern Pennsylvania and parts of neighboring states. First, the Susquehannock tribe was greatly diminished by smallpox and by wars with Native as well as European neighbors; then, in 1763, the remnant of the tribe was murdered by a lynch gang, and their language died with them. In spite of such grim examples, however, most causes of language endangerment in the Americas, as elsewhere, were not physical. Starting in the late nineteenth century, the U.S. government pursued aggressive assimilation policies, including forced attendance by Indian children at boarding schools in which they were sometimes beaten for speaking their heritage language. These beatings could be severe. One effect was that many children who suffered in the boarding schools refused to pass their language on to their own children when they grew up, as we’ll see later in this chapter; another reported effect was that some parents instructed their children to stop speaking their native language so that they could escape the danger of the beatings. It isn’t hard to see how boarding-school beatings could contribute to the endangerment and loss of a language. Children on the Flathead Reservation of northwestern Montana who spoke Salish-Pend d’Oreille were among the boarding-school children, and the suppression of their native language in those schools was surely one of the factors that led to the decline of the language. Another factor was perhaps even more important: the U.S. government concluded that assimilation to the dominant White culture would be speeded up if Indians switched from their traditional hunting and gathering culture to an agricultural economy, and to promote this switch they allotted 80-acre plots of land to Indian households and, in 1910, opened up the rest of the enormous Flathead Reservation to White settlement. (Needless to say, the government also had economic motives for pursuing such policies.) Prior to 1910, the few Whites on the reservation (which was established by the Hell Gate Treaty of 1855) tended to learn to “speak Indian” to communicate with the Indians at trading posts and elsewhere. After 1910, the flood of Whites entering the reservation shifted the linguistic burden to the Indians: they needed to learn English in order to talk to the White people they increasingly encountered in their daily lives. As time went on, the prestige of

Economic pressures

English – the language of education, the language of the media, the language of economically successful (mostly White) people, the language of the nearest cities outside the reservation’s boundaries – increased young tribal members’ preference for English. Today the handful of remaining fluent native speakers of Salish-Pend d’Oreille are not only fully bilingual in English, which they first learned when they went to school, but they speak English much more often than they speak their first language. They speak English with younger tribal members as well as with outsiders, and even with each other most of the time; English is now their dominant language. Of course the circumstances leading to language endangerment are not identical in all relations between conquerors and conquered peoples: the SalishPend d’Oreille route to endangerment is just one of many possibilities. But the ingredients, when there are conquerors in the mix, are typical. The winners dictate the terms, and the terms often include forced cultural assimilation, including language shift; and even without coercion, the political, cultural, and economic climate established by the conquerors tends to dominate the lives of the conquered, whose languages suffer as a result.

2.2

Economic pressures

Conquest is not necessary to provide a context for economic pressure, however. A dominant culture can say to a subordinate linguistic group, in effect, “You want a job? Speak my language!” This economic pressure can and does apply not only to conquered peoples but also to voluntary immigrants and to contact situations in which, although nobody has moved anywhere, one group comes to dominate the other(s). For instance, Spanish speakers in the U.S., both immigrants and descendants of people who lived in the Southwest when it still belonged to Mexico, have always been under economic as well as social pressure to learn English. Spanish is not endangered in the U.S. (primarily because its speaker numbers are maintained by a steady influx of Spanishspeaking immigrants), but it could become so if the English Only movement succeeds: in at least a few states that have passed Official English laws, including Florida and California, people have been fired from their jobs for speaking a language other than English at work. And in some of the twenty-seven states with exclusive Official English laws (that is, where English is named the only official state language), state employees have been forbidden to speak to state residents in any language other than English, even if their job requires communicating with non-English speakers. None of these laws is federal: the United States still has no official language. But the state laws can still have significant effects on speakers of minority languages, including Native American languages. Not all economic pressure comes from governments, and it is not always coercive in any political sense. In many or most countries, speakers of minority languages must become bilingual in order to survive economically, even if the

21

22

why and how languages become endangered

governments in question exert no explicit pressure toward linguistic assimilation. In Morocco, for instance, male speakers of minority Berber languages have traditionally been bilingual in Arabic because they earn their living among Arabic speakers; the economic factor is mutiplied by the school system, which until recently excluded both teaching in Berber and the teaching of Berber. Similarly, Greek speakers in what is now Turkey were first surrounded by Turkish speakers and then pressured by economic necessity to speak Turkish. Again, it was mainly men who traveled to work from Greek villages to Turkish towns, and therefore mainly Greek men who were bilingual in Turkish. Gradually, Turkish influence became pervasive in the local Greek dialects, to the point where one early observer said that Asia Minor Greek is a language in which ‘the body has remained Greek, but the soul has become Turkish’. Many Greeks simply shifted to Turkish as part of a general cultural change that, by the early twentieth century, included a shift from Christianity to Islam. Asia Minor dialects of Greek were still fairly widely spoken at the end of World War I, but in 1923 most of the remaining Greek population in Turkey was resettled to Greece; so these dialects may be said to have become seriously endangered only when their speakers came into intensive contact with the very different varieties of Greek that were spoken in Greece. Even in the U.S., the lack of an official national language hardly conceals the fact that economic viability, with the possible exception of a very few Spanish-speaking areas, requires fluency in the dominant language, English. These economic pressures, in combination with social factors of various kinds, account for the rapid shift of most immigrant communities in the U.S. to English. Here again, full participation in American society, in particular the need to know English in order to get a good job, requires that immigrants learn English. It also usually results in loss of the heritage language, because the number of domains in which the community’s former language is used shrinks rapidly once the community’s children (often the original immigrants’ grandchildren) grow up with at best passive knowledge of it. The majority of the heritage languages that are lost as a result of a three-generation shift are offshoots of the “old country’s” language – Polish, Czech, Slovak, Hungarian, Italian, Chinese, Vietnamese, Lao, or whatever – and the languages are thriving in the immigrants’ original homelands. Endangerment in such cases affects only the immigrants’ dialects, which in many cases are strikingly different from the varieties still spoken in the original countries. Pennsylvania German (also known as Pennsylvania Dutch, but it is German, not Dutch) is a prime example: this dialect of German has been the focus of much scholarly attention because of a wide range of linguistic features that distinguish it from the German currently spoken in Germany, and because its speakers have resisted pressures to shift to English for three hundred years. As the traditional isolation of most rural Pennsylvania German speech communities has eroded, their language has become endangered. Only with isolation from the dominant culture can a minority speech community hope to preserve its language and

Melting pots

culture; once the isolation is lost – and significant cultural isolation is difficult to maintain in modern America – the language and culture are probably doomed. The exceptions, in the case of Pennsylvania German, are communities in which the religion (Old Order Amish and Old Order Mennonite) contributes to cultural isolation and the preservation of varieties of the language: in these communities the language is still strong. The localized endangerment of immigrant dialects (rather than of whole languages) that is typical of immigrant communities in the U.S. is not universal, however, and this is especially true of speech communities which, though not conquered, nevertheless fall under the influence of a more powerful community. Southern Saami (Sámi), for instance, which is spoken in Norway and Sweden, is gravely endangered: few children are now learning it as a first language, and the Saami communities are shifting to Norwegian and Swedish. (We will learn more about Southern Saami, as well as other varieties of Saami, below and in later chapters; I have regularized the spelling of the language name to Saami throughout this book to avoid confusing readers with alternate spellings, primarily Sámi. Some readers will be familiar with the language under its former name, Lapp.)

2.3

Melting pots

The dominant cultural ideology in the United States is that the country is a melting pot in which people from a wide range of cultural and linguistic backgrounds come together to form a single united population. To many Americans, it seems obvious that merging into a single homogeneous culture means shifting to English and giving up other languages. The shift is seen as a positive good: the more homogeneous our society is, the argument goes, the more unified and harmonious it will be. (This belief flies in the face of a lot of evidence to the contrary, as the all-English-speaking American Civil War and many other conflicts around the globe attest, but never mind. We’ll return to this general issue in Chapter 4, in the section ‘Would the world be better off without linguistic diversity?’.) The views of linguistic minorities – unlike, so far, the views of most native speakers of American English – have not always embraced this ideology, especially in recent years, which have seen the rights of linguistic minorities become a matter of worldwide concern. In the U.S. there is no real controversy over the value of learning English; instead, controversies arise mainly when the question is whether or not to encourage “English plus”, where linguistic minorities learn English while maintaining their heritage languages for use in certain daily contexts. In many American industrial cities that once saw major immigration from central and eastern Europe, the melting-pot metaphor described the situation quite precisely, at least with respect to workplaces and schoolchildren. In

23

24

why and how languages become endangered

Pittsburgh, Pennsylvania, for example, Europeans came to work in the steel mills and related industries, and many of them settled in ethnically homogeneous neighborhoods – Greeks here, Poles there, Italians from one region in a different neighborhood from Italians who came from a different part of Italy. But they mingled in their work at the steel mills, where English was (or became) their shared language. Meanwhile, their children went to the same Catholic and public schools and with children of different ethnic backgrounds, so that their linguistic assimilation to English was rapid: it followed the three-generation shift pattern, and as they grew up the children retained little or no knowledge of their immigrant grandparents’ language. The Greek community deviated from the usual pattern because they had their own separate Greek Orthodox churches, so that the immigrants’ children and grandchildren had less exposure to other immigrants’ offspring and an additional cultural institution, the church, that promoted retention of their heritage language and culture – especially as the church taught the children Standard Greek, a prestigious literary language. In addition, Greek parents arranged marriages for their offspring with Greeks in Greece, so that new monolingual Greek-speaking immigrants regularly joined Pittsburgh’s Greek community. These differences between the Greek and other immigrant communities in Pittsburgh did not in the end prevent language shift; the Greek community also shifted to English, but the shift took about one generation longer than with the Roman Catholic immigrant communities. As we’ve already seen, no entire language was lost when the immigrants’ grandchildren grew up without learning their heritage languages. But here too immigrant dialects of Greek, Polish, Italian, and other languages – dialects that had diverged, thanks in part to English influence, from those spoken by the original immigrants – vanished when the communities assimilated to the dominant Anglo culture and shifted to English. The melting pot as a more or less concrete phenomenon is less obvious in other contexts. Aside from the mixing of ethnically diverse groups in schools and in factories and other workplaces, the path to language endangerment is more a case of pressure (economic, social, educational, political) from a dominant group on one linguistic minority at a time.

2.4

Language politics

It is all too easy to find countries all over the world that attempt to suppress minority languages, although the recent promotion of language rights on the international scene has begun to discourage discriminatory laws. Even France, which has gone so far as to make the use of English loanwords a job-losing offense for civil servants, has shown signs of changing its stance toward minority languages within France. In 1999, then-President Jacques Chirac rejected an effort by the prime minister to ‘modify the constitution to

Language politics

give even limited official recognition to France’s seven regional languages’. According to the French constitution’s Article 2, ‘The language of the republic is French’; one of Chirac’s aides said, ‘The President does not wish to take the initiative for a revision of the constitution that would violate the basic principles of our republic.’ His decision also accorded with a verdict of France’s highest court, which ruled that ratifying the European Charter on Regional and Minority Languages would threaten ‘the unity of the French people and the indivisibility of the Republic’. That was then. Almost exactly nine years later, on 9 June 2008, former President Jacques Chirac announced the new Chirac Foundation, whose mission is to promote peace, ‘to encourage cultural respect and diversity and to promote a truly sustainable model for development’ – including, according to materials delivered to the press, ‘protecting threatened languages and cultures’. In his speech given at the launch of his foundation, Chirac said that ‘we must place culture and cultural diversity at the heart of the human adventure’. Indeed. But what a difference from his 1999 stance! He even mentioned the fact that, of the 7,000 languages spoken in the world today, 90% could disappear during this century, and he promised that his foundation would ‘pay particular attention to languages and cultures threatened with extinction’. He then added, Is that what we want? Do we want an impoverished world, one that only preserves what is profitable in the short term? Personally, I reject that idea. ... [UNESCO should] devise ways of avoiding the disappearance of humanity’s priceless linguistic heritage.

Chirac’s change of heart may or may not be too late for the minority languages spoken in France, even assuming that his new view is shared by the current French government. To take just one example, Breton, the Celtic language spoken in Brittany, France, is severely endangered, with the majority of its remaining speakers over sixty-five years of age. Slovakia offers another example of official promotion of one language at the expense of others. In 1995, shortly after Czechoslovakia dissolved into the Czech Republic and the Slovak Republic, Slovak became the sole official language of the Slovak Republic. As Dennis Baron notes, ‘official language laws are put in place not simply to celebrate unity but also to enforce it, and that’s when legislating language can become not just a symbol but a weapon as well’. Slovakia’s language law was designed to promote assimilation to the Slovak majority. Hungarians comprise about 10% of Slovakia’s population, and they are under pressure to shift from Hungarian to Slovak – perhaps under special pressure because of the history of suppression of the Slovak language when Slovakia was part of Hungary. Such a shift would of course have no effect on the Hungarian language as a whole, because millions of Hungarians speak it in Hungary. But at least one minority language of Slovakia, Rusyn (or Ruthenian), is an endangered Slavic language that is spoken in small speech islands in several countries (Ukraine, Slovakia, Serbia, Croatia); eliminating it in Slovakia would damage its chances of long-term survival.

25

26

why and how languages become endangered

Like Slovakia’s language law, the residential schools in the U.S., Canada, Australia, and elsewhere aimed at promoting national unity by tossing everyone into the linguistic melting pot. In Norway, for instance, Southern Saami children were formerly put into boarding schools where they were forbidden to speak Saami, and – as with indigenous languages in boarding schools in the U.S., Canada, and Australia – they were sometimes beaten for speaking it. There are no longer any villages where Southern Saami speakers are in the majority, so it’s not even a village language; it’s a family language. To a considerable extent, it must be noted, the decline of Southern Saami has roots in the economy: the language and culture have traditionally been closely linked to the reindeerherding industry, and the language has faded along with the industry. Reindeer herding, as with the isolationist rural culture of the Pennsylvania German in the U.S., had the effect of isolating the Saami from the larger Norwegian and Swedish society; if reindeer herding loses its status as the main economic activity of the Saami people, their isolation is also lost. As we will see in the next section, attitudinal factors are also contributing to its decline. The title of this section, ‘Language politics’, might be more precisely (if clumsily) phrased as ‘Politics that affects languages, directly or indirectly’. Some government actions cause language decline without being specifically aimed at language use at all. In Botswana, for example, the language of the San (formerly known as Bushmen, a term that is now considered insulting) was preserved by isolation, as the San led a life of hunting and gathering quite apart from the dominant groups in the country. But starting in the mid-1990s the Botswana government, as part of an effort to force assimilation of the San – to ‘bring development to southern Africa’s most traditional people’ – forcibly removed the San from their villages in the Kalahari Desert and put them in government camps where they were provided with water and other government services. The move ended the San’s isolation from the larger Botswana society, and endangered their language along with other aspects of their culture. (A 2006 court order, the result of a lawsuit they had brought against the government, guaranteed them the right to return to their ancestral homeland in the Kalahari.) In Estonia, the Seto language – or dialect of Estonian, as it is sometimes classified – has gone through two different periods of policies that encouraged its decline over the past hundred years. First, as the quotation at the beginning of this chapter indicates, Seto suffered from the dominance of Russian in the educational administrative systems of the Setomaa region; then, after 1920, the new Estonian government pushed assimilation to the Estonian language and culture, and that pressure continued as Estonian gained ascendancy.

2.5

Attitudes

What people think about their language – its value, its usefulness, its importance to their culture – can play a decisive role in the language’s fate. If a

Attitudes

language is on the brink of endangerment or extinction, its speakers’ opinions about it may speed, hinder, or even reverse its decline. In many cases a dominant culture encourages negative attitudes within a minority community toward the community’s heritage language; in other cases, even without direct pressure from a dominant speech community, the minority community itself comes to view its own language with indifference or contempt. Attitudes like this do not bode well for the language’s chances for long-term survival. Here are a few typical examples. Children who attended North American residential schools where speaking their heritage languages was strongly discouraged often absorbed their teachers’ negative view of those languages. The children’s experiences stayed with them in later life; as one Tlingit elder (southeastern Alaska) said, ‘Whenever I speak Tlingit, I can still taste the soap.’ And here is another elder’s recollection of her boarding-school years: Sometimes I could just kick myself for not teaching my children the language ... When I was in school we were beaten for speaking our language. They wanted to make us ashamed. ... I have 17- and 18-year-old kids coming to me crying because the elders in their tribes will not teach them their own language.

This comment is especially poignant because the speaker was Marie Smith Jones, the last speaker of the Eyak language of Alaska. She was eighty years old, the chief of the Eyak tribe, when she said that; she died in 2008 at the age of eightynine, and her beloved language died with her – although, as we will see in Chapter 7, a program designed to revive Eyak is now under way. (Eyak is the sole member of a branch of the language family whose other branch is the large Athabaskan subfamily; see Chapter 3 for a fuller account of the loss of all its traditional speakers.) In Russian Karelia there is a saying, ‘You are as stupid as a Veps’, reflecting the common view there that a Veps is a ‘stupid, uneducated peasant’. Given this reputation, it isn’t surprising that Veps is gravely endangered (“severely endangered” on UNESCO’s Language Vitality and Endangerment scale): few or no children are learning it as a first language, and it is only ‘taught as a hobby school subject or as a second foreign language twice a week’. Other factors contributing to the language’s precarious situation are the fact that most ethnic Veps people live in large Russian-dominant cities and the related fact that interethnic marriages are frequent. Also in Russia and also a Uralic language, Udmurt – formerly called Votyak, a name favored by the former Soviet government but considered offensive by Udmurt people – is “definitely endangered” on UNESCO’s scale. A recent paper reports on several myths about the Udmurt language, among them some that are potential contributors to the language’s endangerment: ‘The Russian language is more beautiful than Udmurt’; ‘People can express themselves well only in Russian’; ‘The Udmurt language half consists of Russian loans’; and ‘Udmurt people speak neither their mother tongue nor Russian well’.

27

28

why and how languages become endangered

Sometimes negative attitudes toward minority languages are actively promoted by dominant cultures. In southeastern Alaska, some Christian churches denounced Native American languages, condemning Tlingit, for instance, as “demonic”; not surprisingly, many devout members of the Tlingit tribe then concluded that God does not like Tlingit. In one case, a striking contrast in the attitudes of two communities appears to be determining the different fates of their closely related languages (or dialects of the same language). Eda Derhemi, investigating the status of two endangered varieties of Albanian, found recently that Arbëresh, spoken in Sicily by the descendants of fifteenth-century immigrants, ‘continues an optimistic struggle for survival’, while Arvanitika, spoken in Greece by the descendants of thirteenth-century settlers from Albania, ‘is headed towards certain death in the next 10 to 15 years’. Each of these two varieties is surrounded by speakers of dominant national languages – Arbëresh by Italian, Arvanitika by Greek – and each has been losing speakers to the dominant language for a long time. The crucial difference lies in their attitudes toward their heritage language: In Arbëresh communities, ‘positive linguistic attitudes and language loyalty persist’, whereas most Arvanitika speakers and semi-speakers are ashamed of their language and wish to distance themselves from Albanian-ness. A different kind of attitudinal problem is hampering efforts to revitalize Southern Saami in Norway. The highest value is placed on speaking “correct” – that is, standard – Southern Saami. Speaking Norwegian (or, in Sweden, Swedish) with embedded Southern Saami elements is considered to have an intermediate value, followed by “normal” speech (Norwegian or Swedish) and then, at the lowest end of the linguistic scale, “incorrect” Southern Saami. Many community members have studied Southern Saami and want to speak it; but they are second-language learners who study the language in school and at Sámi University College, and their opportunities for practice in speaking are (in effect) too little and too late. Therefore, although they have quite a bit of knowledge of the language, they don’t dare use it, because speaking Southern Saami incorrectly is considered unacceptable. They therefore stay safely in the middle, speaking Norwegian (or Swedish) with embedded Southern Saami expressions, in an emerging ethnolect. This is not a good situation for those who wish to preserve and revitalize Southern Saami itself, and the goal now is to encourage people to use the language even if they do make mistakes – to upgrade “incorrect Southern Saami” to a higher level of valuation. A somewhat similar problem, though even more complex in its ramifications, has arisen with efforts to revitalize Dakota, an endangered Siouan language spoken in the northern Great Plains region of North America (North Dakota, South Dakota, Minnesota, Nebraska, and neighboring parts of Canada). John Hunt Peacock commented recently on the intolerance shown by speakers of some dialects of the language toward other dialects, and on the effects of that intolerance:

Attitudes

As counterproductive as using technology only one way is an insistence on the part of some fluent-speaking traditionalists that indigenous languages can or should be spoken only one way. Dakota speakers whom I have met in learning the language myself are sometimes less than tolerant of regional dialect differences ... a Dakota spiritual leader I know in Minnesota told me he wished that the Dakota teacher that his community hired from Canada would stop teaching the local youngsters to speak the way they do up there. ... In another Dakota community than mine, two competing Dakota day care centers were started by tribal factions; so much bad feeling was generated that now at that community there is no day care in the endangered language at all. ... Of absolutely no survival value to my ancestors, the European romantic ideal of oral language purity has been adopted by contemporary indigenous language purists, just as tribal governments adopted the European racist notion of blood quantum.

The ‘romantic ideal of oral language purity’ is closely related to an issue that troubles many language revitalization programs. If the last fully fluent elders are viewed as keepers of the language, and if they (and linguists who are documenting the language) scorn the “incorrect” forms used by younger community members who have learned the elders’ language imperfectly, efforts to create a new viable speaker base for the language are all too likely to fail. In other words, if the community can’t be brought to accept an altered version of their heritage language, the language may be doomed: there are cases in which elders’ laughter at youngsters’ mistakes have killed languages. One Chinookan language that used to be spoken along the Columbia River, which divides Washington from Oregon, was famous (or notorious) for being extraordinarily difficult to learn. Late in its life, at a time when young tribal members were already English-dominant, tribal elders used to make fun of the mistakes made by younger speakers of Chinook. Wishing to avoid ridicule, and having another language they could use just as easily, the young people simply switched to English and stopped trying to speak Chinook. Exit Chinook. The same thing happened to a variety of Swedish that is, or was, spoken in the Great Plains region of the United States. Swedish immigrants brought a wide range of dialects with them to the U.S., and different dialects carried more or less prestige. One woman explained to a fieldworker why she and other speakers of her dialect gave up Swedish entirely and switched to English: Father came from the island of Gotland and mother from Östergötland. Thus they had different dialects. When we people spoke, the other people laughed at us. So we went over to English. The north Sweden (Gästrikland) people – they thought they knew how to speak, but we didn’t. Theirs was right, but ours was not.

As these last two cases suggest, it isn’t only adults whose attitudes are relevant to a language’s fate. It’s well known that bilingual children make decisions about their languages, often deciding to speak just one of them – typically the language

29

30

why and how languages become endangered

of a dominant culture – exclusively. Here’s a story which, though it doesn’t concern endangerment (because Spanish is one of the world’s lucky “safe” languages), illustrates the complexity of such cases. Some years ago, in a class on endangered languages, one of my students described her family’s language situation. Her parents were immigrants from Mexico, and their home language was Spanish. My student was the second of three children, and like her brother and sister she was bilingual in Spanish and English. Her sister, the oldest sibling, gave up Spanish completely when she grew up and moved away from home, raising her own children to be monolingual English speakers. The youngest of the three children, the only boy, was also English-dominant and uninterested in maintaining his Spanish-language skills. Only the middle child cared to maintain Spanish in her life, and she cared passionately: she deplored her sister’s decision, and planned to raise her own children to be bilingual. She didn’t blame her brother so much for his lack of interest in Spanish; as is common with younger siblings in minority-language families, he was exposed to much more English than Spanish, and he would have needed a powerful motive for preserving his Spanish. My student’s motive for doing so couldn’t have been predicted; some people, including young people, feel a stronger attraction to their heritage language and culture than others do. A final example illustrates the importance of attitudinal factors in preserving an endangered language – or not preserving it. In Oaxaca, Mexico, two neighboring Zapotecan languages have followed very different paths in the face of pressure from Spanish. The Zapotec variety spoken in the large town of Juchitán, which is situated on the Pan-American Highway, is spoken by thousands of people who are surrounded by Spanish speakers. Juchitán Zapotec is preserved quite well, with little influence from Spanish, in spite of the fact that education was mostly in Spanish until quite recently, and in spite of its urban location, which brings Zapotec speakers into contact with Spanish on a daily basis. In sharp contrast, a closely related Zapotecan language that is spoken in a rural area, two hours from Juchitán by bus, has adopted many words from Spanish, and is gravely endangered: a linguist who visited the village found that only men over forty still spoke Zapotec regularly. The reason, he was told, was that the village women had decided that their children should speak only Spanish so that they could eventually get jobs at a nearby cement factory. They therefore decreed that everyone in the village should speak Spanish rather than Zapotec – thus hastening the demise of that variety of Zapotec. Before ending this section, I should emphasize that attitudinal factors can work the other way too, promoting preservation and revitalization of an endangered language instead of hastening its demise. Sometimes powerful cultural forces dictate language-preserving attitudes. In the Vaupès River basin in Colombia and Brazil, for instance, extensive multilingualism is the norm, and languages are preserved in part because of institutionalized exogamy – a culturally enforced requirement that one must marry someone from outside one’s own ethnic group. In this case, the requirement is language-based: you must marry someone from

Attitudes

a different language community. Maintaining extreme multilingualism is thus dictated by the need to have available marriage partners; as one speaker of Tukano (Brazil and Colombia) said, ‘If we were all Tukano speakers, where would we get our women?’ The most common favorable factor by far, however, is the perceived importance of a community’s language as a symbol of the group’s identity. An eloquent statement of this view is found on a sign in the Warradjan Aboriginal Cultural Centre, Kakadu National Park, Northern Territory, Australia: Language is fundamental to cultural identity. This is so for all people everywhere. For Bininj, their unique world is expressed in their language. For this reason, it is important that people keep their own language alive. For Bininj, language and land are linked. ... Each clan has its own name and territory ... a child’s language may not be his or her first language, because a family may spend time in the mother’s country and mainly speak her language. In this [Cultural] Centre, Bininj use Balanda [non-Aboriginal] words to talk about their country, but as with the translation of any language, some words cannot convey the original meaning exactly.

It is not that cultural identity is always bound tightly to language; sometimes it isn’t. In some formerly Chantyal-speaking villages in Nepal, for instance, the language was lost in part because it was ‘not seen as a key feature of their Chantyal identity’. (Chantyal is not extinct, but it is “definitely endangered” on the UNESCO scale.) And in parts of Africa, especially West Africa, the linkage between language attitudes on the one hand and language and identity on the other is considerably looser. Nevertheless, the equation of cultural identity with the heritage language is extremely widespread, and – as the quotations from Marie Smith Jones and from the Bininj sign indicate – a great many speakers are all too aware of the vital importance of their language to their culture, and to the preservation of their community’s cultural knowledge. Some things don’t translate easily. Another type of attitude that can help motivate speakers to maintain their ancestors’ language is a high cultural importance placed on multilingualism – that is, on the prestige of being multilingual. In many parts of the world multilingualism is valued so highly that it is normal to learn another group’s language, even a conqueror’s language, without giving up a heritage language. This attitude is still common in the Pacific Northwest of the U.S., for instance, and in interior British Columbia as well. One linguist living with the Carrier Nation in British Columbia was scolded by an elder for speaking only Carrier and one or two other nearby Athabaskan languages; he ought to learn at least one or two more, the elder said. And here is a quotation about the prevalence of multilingualism in the region, referring specifically to the period starting around 1750: As befitted their position as traders, Wasco, Wishram, and Cascade Indians often spoke more than one dialect of Chinookan and one or more of

31

32

why and how languages become endangered

Sahaptin. ... By 1812, at least one [Wishram] man knew some English, and the number increased slowly during the period. ... The multilingualism ... continued during [1858–1920]. ... A few Wascos and Wishrams learned even more distant languages, such as Nez Perce, which were useful for visiting and trading. There was prestige attached to knowing languages and perhaps aesthetic satisfaction; while serving as an army scout, one Wasco learned a certain amount of Delaware from a fellow scout. While this was not a “practical” accomplishment, it provided a basis for pride and prestige.

A similar situation obtains in several parts of the world where “asymmetrical bilingualism/multilingualism” has been stable over the long term: in these contexts, which have been reported from Australia, Sarawak, the Chaco in Argentina, and the Vaupés region of Brazil, each person chooses one of a community’s languages to speak and is spoken to in the community’s other language(s). In the Argentinian village Misión La Paz, for instance, Lyle Campbell and Verónica Grondona found in their fieldwork that ‘interlocutors in conversations usually do not speak the same language to each other’; instead, different community members make different choices among the three community languages: Chorote, Nivaclé, and Wichí. On the other hand, multilingualism is not universally viewed as prestigious. An intriguing comment appears in a field journal in the collected papers of Melville Jacobs, who conducted extensive fieldwork in Oregon in the 1920s and 1930s. According to Jacobs’s notes, Victoria Howard, a native speaker of Clackamas Chinook, gave this as her opinion of people who spoke many languages in the old (pre-contact) days: ‘You talk like a slave’—they were so often foreigners. Clackamas had no Clackamas slaves. Only slaves could talk all kinds of languages long ago, and a person who went thus from tribe to tribe and knew so many languages was likely to be derided for being a slave, homeless like a tramp or hobo.

Victoria Howard’s comment is in direct opposition to the quotation just above it; we cannot know why they disagree—whether it was because different tribes were involved, or different periods of time, or for some other reason. But it does indicate the futility of making hard-and-fast predictions about attitudinal factors, in this and other instances.

2.6

Loss of linguistic diversity via standardization

Most readers of this book have probably never thought about how a language gets standardized—how a particular dialect or combination of dialects is chosen to form the literary standard that will be considered “correct”, taught in schools, and used instead of local dialects in broader public discourse. English and other major languages around the world have such well-established standard forms that modern speakers have had no experience with the process by which

Loss of linguistic diversity via standardization

the standard was developed. But many or most minority languages, including (or in particular) those that are endangered, have no standard form; all dialects are roughly equal in prestige and spheres of usage, there is typically no written form, and the language is not taught in schools. Among the issues that must be addressed in standardizing a language are these: how choices are to be made, either of a single dialect as the standard or of different words and forms from different dialects; how the selected variants are to be codified as a single norm that will be taught in schools and promoted as the formal written and/or oral variety of the language; and how the selected variants will become conventionalized and established throughout the speech community. Standard English, to take the most obvious example (since this book is written in English), has been in existence for hundreds of years; grammarians with a passion for linguistic tidiness have been fine-tuning grammatical rules and denouncing sloppiness (a.k.a. linguistic change) for almost as long. Pundits write about grammar for newspapers and magazines and, more recently, in blogs, but the main target of grammarians has long been schoolchildren. What most people don’t realize is that you already knew most of your native language’s grammar before you set foot in a schoolroom. The grammatical rules taught by English teachers to grade-schoolers are merely the icing on the cake, forms and constructions that are either mostly confined to formal written prose or set apart by social stigmatization. In the latter category are words like ain’t and sentence constructions like He don’t have no money (myth: double negatives like this one make a positive); in the former category are disputes about whether the verb should be is or are in a sentence like The number of men and women with treatable diseases is/are immense. And then there’s spelling: writing She would of invited them if she’d known their in town can make a pundit faint with horror, although would of and their are phonetically identical to inoffensive (if colloquial) would’ve and they’re. But these are frills. No schoolteacher will bother telling native-English-speaking children to say Johnny threw Mary an orange rather than Threw to Mary by Johnny an orange because no nativeEnglish-speaking child comes to school saying such things – and the same is true for other basic grammatical constructions. Things are very different for a language that is in the process of being standardized. The larger the speech community, the more diversity there will be in the language as it’s spoken by children, who are in the front lines of the standardization process – they are the people who will be taught to use the new standardized form. Here too children will come to school at the age of five or six with a very good command of their native language; but they won’t all speak it the same way, and the standardizers will have to make choices. Take English as a hypothetical example of a language that is being standardized for the first time: Should ain’t and I don’t have no money be considered standard or nonstandard? With English, there would be no contest: ain’t and double negatives of this type are used primarily by people who have less power and less money than people who don’t use these forms. Where there’s a significant power asymmetry,

33

34

why and how languages become endangered

the people with the power will see their dialect triumph in the standardization market. Spelling decisions may be more arbitrary – if there, their, and they’re all sound alike, why not spell them alike? – but word structure often wins out over phonetic identity, as with these three words. That doesn’t explain why beat and beet have to be spelled differently, of course. There the difference is rooted in the history of the words: it’s part of the historical baggage of the language, left over from an era when the words were phonetically different. Standardization has become an issue with many threatened or endangered languages because of a new worldwide consciousness of minority language rights. As with UNESCO’s Atlas of the world’s languages in danger and the Chirac Foundation’s goal of ‘protecting threatened languages and cultures’, more and more international organizations and national ministries are concerning themselves with the establishment and implementation of language rights for minority groups. Granting minority language rights typically has several consequences: minority languages are introduced into the educational system, used in TV, radio, and other media and perhaps in government documents as well, and sometimes even given status as an official national language. For all this to happen, especially if a minority language has not previously been written, new materials will be needed: an alphabet and an orthographic (spelling) system, grammar books, schoolbooks written in the language for teaching other subjects (if the language is to be used as a medium of instruction rather than merely being taught as a subject), dictionaries, and maybe also new specialized vocabulary. Teachers will need to be trained to teach the language and/or to teach in the language. To achieve all this, the language will have to be standardized. The cost of supporting minority languages in this way is very high, and it would be prohibitively expensive to produce all these materials in multiple varieties of the language. So the only practical option is to select or create a single standard variety to be codified and written in a single unified orthography. It is then all too likely that other varieties, all of which are now nonstandard, will come to be seen as inferior – just as nonstandard varieties of English are seen as inferior to Standard English. And even if this doesn’t happen, a probable consequence is a partial loss of the sense of identity in a community with a newly nonstandard dialect. To the extent that the community’s cohesion and perceived connection to past generations is linked to their language, their ties will be weakened if their children are taught to speak and write the standard dialect rather than their parents’ dialect. Recall the speaker of a Swedish dialect that was perceived as incorrect by speakers of more prestigious Swedish dialects: she and others speakers of her dialect switched to English rather than endure the ridicule brought on by their speech patterns. This is one possibility with people whose dialects have suddenly become nonstandard. And with a minority language that is already under pressure from a majority culture, that could be fatal to the whole language. In parts of Scotland, for instance, the standardization of the gravely endangered Scottish Gaelic language has led to tensions within the speech

What doesn’t promote endangerment?

community; some speakers of nonstandard dialects are less than enthusiastic about pressures to have their children learn the standard dialect instead of the parents’ dialect. Vanessa Will reports that ‘older speakers of Gaelic often claim not to understand Gaelic-medium-educated children and actually refuse to speak Gaelic with them’. The situation isn’t hopeless – there are ways of relieving the tensions, ranging from educating the elders about the value of standardization to teaching the children how to negotiate legitimacy for the dialect they’re learning – but it’s not an easy problem to solve. The other possible result, when children are exposed to a standard dialect in schools, is a shift to the standard dialect. At first glance this result isn’t so drastic: the language seems more likely to survive in this one form than in many nonstandard forms, and the sacrifice of nonstandard dialects may seem a reasonable price to pay. If this leaves community elders alienated and angry, however, the speech community’s overall unity may suffer, and one long-term effect might be, again, a switch to a majority language. In some cases what is lost might even be a different language – a variety that is not mutually intelligible with the standard dialect and is therefore, by this measure, a separate language. The clearest example is Chinese. Victor Mair has commented on linguistic diversity within Chinese: not only are the major Chinese “dialects” mutually unintelligible – that is, they are actually separate Chinese languages – but even within the Mandarin “dialect” area (which includes Standard Mandarin, known as Putonghua), mutual unintelligibility is rampant. TV news broadcasts must use subtitles when they feature speakers of nonstandard dialects; Mair’s particular example is from Laiwu, a Mandarin-speaking city just 430 kilometers from Beijing. In other words, when people shift to a standard dialect of a minority language – or, in the case of Chinese, even of a majority language – in some cases they are in fact shifting away from their native language. None of this means that endangered minority languages should not be standardized; in many or most cases, preservation and revitalization of an endangered language cannot happen without standardization. It does mean that the importance of nonstandard dialects to individuals and communities should be recognized, so that standardization can be pursued with careful attention to avoiding alienation and loss of culture in communities whose native speech has become nonstandard.

2.7

What doesn’t promote endangerment?

In this final section we’ll look at the other side of the picture. After inspecting factors that cause languages to decline to the point of endangerment, we will now turn to a survey of factors that tend to promote long-term viability. Some of these are simply the reverse of the factors that promote endangerment: lack of conquest and of subsequent pressure to assimilate to

35

36

why and how languages become endangered

a dominant culture, lack of economic incentives to shift to another language, lack of a melting-pot ideology, lack of political moves that deliberately or accidentally cause languages to lose speakers, and lack of negative attitudes toward vulnerable languages. But of course this list doesn’t tell the whole story. A language that is spoken by community members of all ages, that is learned by all of the community’s children as a first language, and that is spoken in all domains of everyday life – at home, at work, in the schools, and elsewhere – is not endangered. Whether it will become endangered or not is hard to predict: in principle, any language could become endangered in the future. But we can list a few specific factors that contribute to making a language safe from endangerment. Official state support for a language is one of these. If the language has official national-language status, and if a speech community has the right to have children taught in the community’s language (and not just taught about the language in special classes), those official measures may help to prevent or reverse a minority language’s decline. Economic power of the speech community can also help protect a language, both because it helps the speakers maintain their independence and, if they wish, separateness from other languages, and because rich people who speak a minority language are less likely than poor people to come to associate their language with a low status in society. Large numbers of speakers in both rural and urban areas may also help. As we saw in Chapter 1, speaker numbers alone do not guarantee safety: Navajo has declined rapidly from almost 150,000 speakers twenty years ago and is now classified as vulnerable on UNESCO’s scale. But a language with hundreds of thousands of speakers is nevertheless more likely to survive than a language under similar social conditions but with only a few dozen speakers. For minority-language communities, relative isolation contributes to maintenance. Pennsylvania German, Southern Saami speakers in Norway and Sweden, and the San language of Botswana were all preserved as long as their speakers remained isolated from the larger society around them. But once that isolation was lost – for the first two groups by cultural encroachment and by the San because of their forced move away from their Kalahari home to government camps – the languages became endangered. The general movement in many countries from rural settings to cities also eliminates the isolation that helps to preserve languages. What languages are safe, then? Many (but not all) of those that have the status of official languages in one or more countries and whose speakers are economically strong and numerous. Minority languages that are not actively discriminated against by their neighbors, that enjoy high prestige among their speakers, and that provide opportunities within the community for making a living. Among the most robust languages are those with many millions of speakers and official status in more than one nation. The number of nations and of official languages within nations changes frequently, so any list is sure to be out of date, but here are some numbers that show the relative position of a

Sources and further readings

few major languages: of the 170 nations that existed in 1990, English was a, or the, official language in 45 countries, French in 30, Spanish in 20, Arabic in 20, and Portuguese in 6. Only about 50 nations had none of these languages in their list of official languages. Standard Chinese (Putonghua) is exceedingly strong in the People’s Republic of China. German, Japanese, and Polish have no serious rivals in their respective countries (two countries, for German). This list could of course be expanded – Swahili is strong in several African countries, Latvian in Latvia, Vietnamese in Vietnam, Thai in Thailand, and so on – but vulnerable and endangered languages of the world nevertheless far outnumber the (apparently) invulnerable languages.

2.8

Summary

In this chapter we’ve considered several factors that contribute to language endangerment – genocide (rare) and fatal diseases, conquest, economic pressures, a melting-pot ideology, language politics, and negative or even just indifferent attitudes toward a community’s vulnerable language. We also explored the loss of linguistic diversity, including dialect diversity, as a result of standardization and the concomitant pressure on minority-language speakers to adopt a new standard variety. The final section was a brief look at the other side of the coin: factors that tend to make a language safe from decline. The most important point to keep in mind, when we assess factors that contribute to endangerment, is this: no single factor (except brutally efficient genocide or disease) will ever be responsible for the decline and death of a language. The residential schools, though they certainly affected speakers’ attitudes toward their native languages and the likelihood that they’d pass those languages on to their children, operated in a climate that also included other aspects of linguistic and cultural suppression. On the Flathead Reservation of Montana, for instance, the boarding schools alone probably wouldn’t have caused so drastic a decline in the use of Salish-Pend d’Oreille (Montana Salish); it was the combination of the schools and the opening of the reservation to massive White settlement, together with other government measures designed to force assimilation on the Indians, that led to the language’s current critically endangered state. The same is true, of course with differences in the details, of other endangered languages around the world.

2.9

Sources and further readings

The quotation at the beginning of the chapter is from the 2009 brochure of the Estonian Bureau for Lesser-Used Languages.

37

38

why and how languages become endangered

I am grateful to Nick Emlen (personal communication, 2014, including the quotation) for alerting me to the importance of anthropologists’ findings about macro-historical factors as ultimate rather than proximate causes of language shift and therefore language endangerment. For Don Kulick’s analysis of language shift in Gapun, Papua New Guinea, see his 1992 book Language shift and cultural reproduction: socialization, self, and syncretism in a Papua New Guinean village. See also the classic work on this subject, Susan Gal’s 1979 book Language shift: social determinants of linguistic change in bilingual Austria. A good account of Quechua before and after the Incas were conquered by Spain is given in Bruce Mannheim’s 1991 book The language of the Inka since the European invasion. The often-told story of Lord Jeffrey Amherst and smallpox blankets can be found in an online report by Peter d’Errico, ‘Jeffrey Amherst and smallpox blankets’: www.nativeweb.org/pages/legal/amherst/lord_ jeff.html (accessed 29 May 2014). And for an example of mythical (or fraudulent) smallpox blanket stories, see Thomas Brown’s 2006 article ‘Did the U.S. army distribute smallpox blankets to Indians? Fabrication and falsification in Ward Churchill’s genocide rhetoric’: http://quod.lib.umich.edu/p/plag/5240451.0001.009/--did-theus-army-distribute-smallpox-blankets-to-indians?rgn = main;view = fulltext#N2 (accessed 29 May 2014). The sad fate of the Susquehannocks is described on p. 2 of Marianne Mithun’s 1981 article ‘Stalking the Susquehannocks’. An excellent source on economics and language endangerment is Wayne Harbert’s 2011 article ‘Endangered languages and economic development’. See also Chapter 7 for comments on economics in relation to the prospects for success of a revitalization project. Wikipedia has a good general article, with good references, on the English-Only movement in the United States: en.wikipedia.org/wiki/Englishonly_movement (accessed 22 May 2014). The figure I give for states with exclusive English-Only laws – twenty-seven – may not be accurate by the time you read this book. First, laws can change; second, different sources differ on the actual number (I’ve found more or less contemporaneous figures on the internet ranging from twenty-six states to twenty-nine); and some especially restrictive state laws have been declared unconstitutional by federal courts. The quotation about the soul of Asia Minor Greek is from R. M. Dawkins’s book Modern Greek in Asia Minor, 1916, p. 198. There still are Greek speakers in Turkey, and their dialects are being intensively studied by modern linguists. But the dialects are endangered. Einar Haugen, for example in his 1989 article ‘The rise and fall of an immigrant language: Norwegian in America’, has discussed ways in which ethnically mixed school systems and the loss of isolation lead to linguistic assimilation. The account of the four-generation shift by Greek immigrants in Pittsburgh (contrasting with the more common three-generation shift pattern) is

Sources and further readings

from pp. 15–16 of Christina Bratt Paulston’s 1994 book Linguistic minorities in multilingual settings: implications for language policies. For more information about the various fates of different Pennsylvania German speech communities, see Charles E. Hurst and David L. McConnell’s 2010 book An Amish paradox and Steven Hartman Keiser’s 2012 book Pennsylvania German in the American Midwest. Another good source, an analysis of a Pennsylvania German community in Michigan, is Johanna Jansson’s 2014 conference paper ‘Saving threatened languages – findings from an Old Order Amish setting’. The story about Jacques Chirac’s earlier stance toward the minority languages of France, including the quotations, is from Jon Henley’s article ‘Chirac defends pure French tongue against regional Tower of Babel’, which appeared in The Guardian’s forum The Observer on 27 June 1999: see www.guardian.co.uk/ world/1999/jun/27/jonhenley.theobserver (accessed September 2009). The story about Chirac’s later position on the subject of minority languages (even in France) is from LINGUIST List vol.-19-1967 (21 June 2008), which in turn got it from Cultural Diversity: Newsletter: The Diversity of Cultural Expressions (Quebec), vol. 8, no. 21. Information about the status of Breton can be found in the UNESCO Red Book on Endangered Languages: Europe, compiled by Tapani Salminen (website: www.helsinki.fi/~tasalmin/europe_report.html) (accessed October 2009); see also the website of the International Committee for the Defense of the Breton Language (www.breizh.net/icdbl/saozg/endangered.htm, accessed October 2009) and the website of the U.S. branch of the Committee (www.icdbl.org, accessed 24 May 2014). The story of language politics in Slovakia is told by Dennis Baron in an article posted on his Web of Language website on 13 July 2009, ‘Supporters of official English in the United States can learn from Slovakia’ (see http://illinois.edu/db/ view/25). The fate of Saami in Norway (and Sweden) and the prospects for revitalization of the language were discussed by Inger Johansen of Sámi University College in a paper on ‘Changes in a small language community’, presented at the 12th International Conference on Minority Languages in Tartu, Estonia, in May 2009. The comment about children being beaten in those schools for speaking Saami was made at the same conference by Rolf Olsen Samediggi of the Sámi Parliament, Norway (personal communication, May 2009). The relocation of San from the Kalahari Desert to a government camp in Botswana is reported in a Washington Post article by Craig Timberg, ‘A culture vanishes in the Kalahari Dust’ (2 June 2005). The 2006 court order that permitted them to return to the Kalahari and to their cultural practices is from the Wikipedia article ‘San people’ (accessed 6 May 2014). The Tlingit elder’s comment about having his mouth washed out with soap when he spoke Tlingit in school is from p. 65 of a 1998 article by Nora Marks Dauenhauer and Richard Dauenhauer, ‘Technical, emotional, and ideological

39

40

why and how languages become endangered

issues in reversing language shift: examples from Southeast Alaska’; the characterization of Tlingit as demonic by some Christian churches is reported on p. 64 of the same article. The quotation from Marie Smith Jones, the last speaker of Eyak, is from Joan Raymond’s 1998 article ‘Say what? Preserving endangered languages’. The information about attitudes toward the Veps language (and ethnic Veps people) is from Evgenia Romanova’s 2007 University of Tromsø M.A. thesis; see the Sources and Further Readings section of Chapter 1 for details. For further information about Veps revitalization efforts, see the website of the Vepsän Seura-Vepsian Society: www.veps.de. The report of language myths that tend to downgrade the Udmurt language in the eyes of its speakers (and others) is from Larisa Shirobokova’s 2009 talk on ‘The role of sociolinguistics in progress and revitalization of minority FinnoUgric languages’, which was presented at the 12th International Conference on Minority Languages, Tartu, Estonia. Eda Derhemi described the differential results of different attitudes in two endangered Albanian-speaking diasporic communities in her 2014 conference talk ‘The puzzle of assessing linguistic vitality: Arbëresh (Sicily) and Arvanitika (Greece)’. The quotations from her report are taken from the abstract for this talk. John Hunt Peacock’s comments on intolerance among Dakota tribal members toward non-local Dakota dialects are from pp. 134–135 of his 2010 article ‘The good news and the bad news about reviving my Spirit Lake Dakota Nation’s endangered language’. I am grateful to Nancy Dorian for calling my attention to Peacock’s article and in particular to this passage. The story about why a non-prestigious southern dialect of Swedish was abandoned by a group of immigrants in the U.S. was told by Folke Hedblom in a 1980 book Languages in conflict: linguistic acculturation on the Great Plains (which focuses almost exclusively on immigrant languages rather than on Native American languages). I am grateful to Paul Kilpatrick (personal communication, ca. 1996) for relating his experience with Juchitán Zapotec and the dying nearby Zapotecan variety in Mexico. I am grateful to Nick Emlen (personal communication, 2014) for bringing to my attention the Vaupès example of exogamy promoting language maintenance. The source of the quotation about needing other languages in order to have available marriage partners is from Jean Jackson’s 1974 article ‘Language identity of the Colombian Vaupés Indians’; see also her 1983 book The Fish People: linguistic exogamy and Tukanoan identity in Northwest Amazonia. I saw the sign that described the Bininj view of the inextricable link between their language and their culture on a visit to Kakadu National Park in 1998. The observation that Chantyal cultural identity is (or was) not tightly connected to the Chantyal language is from Michael Noonan’s 1999 article ‘The rise and fall and rise of the Chantyal language’. I owe the observation about

Sources and further readings

the common lack, in parts of Africa, of an attitude that language equals cultural identity to Peter Austin (personal communication, 2014). The story of the Carrier elder who pushed for increased multilingualism in neighboring languages was reported by William Poser, and the quotation about the prestige of multilingualism in the Pacific Northwest is from pp. 370 and 396 of Edward Spicer’s 1961 book Perspectives in American Indian Culture Change; this passage was brought to my attention by David Robertson in a post on the CHINOOK List, an electronic discussion list devoted to the pidgin language Chinook Jargon. Thanks are due to Peter Austin for the sample list of parts of the world where asymmetrical bilingualism/multilingualism is found (personal communication, 2014); the example of asymmetrical multilingualism in Argentina is from Lyle Campbell and Verónica Grondona’s 2010 article ‘Who speaks what to whom? Multilingualism and language choice in Misión La Paz’. And I thank Henry Zenk (personal communication, 2005) for giving me the Victoria Howard quotation from Melville Jacobs’s fieldnotes; in citing this quotation, I have de-abbreviated Jacobs’s abbreviations. The problem of standardized vs. newly nonstandard varieties of Scottish Gaelic has been explored by Vanessa Will, for example in her 2004 paper ‘The semiotics of socialization: Gaelic-medium education and language revival in Scotland’. She argues that, in the past twenty years, ‘Gaelic-medium education in primary school has come to bear the brunt of the language revitalization project in Scotland.’ Victor Mair’s report on “dialect” diversity in Chinese is from his Language Log post ‘Shandong dialect intelligibility’ (24 May 2009).

41

3

Sliding into dormancy Social processes and linguistic effects

‘Tip’ is ‘a gradual accretion of negative feeling toward the subordinate group and its language, often accompanied by legal as well as social pressure, until a critical moment arrives and the subordinate group appears abruptly to abandon its original mother tongue and switch over to exclusive use of the dominant language’. (Nancy Dorian, 1986:75)

This chapter focuses on the most common outcome of language endangerment: language dormancy – that is, language death, unless and until the language is brought back to life. In the end stages of a slow decline that leads to endangerment and disappearance, children no longer learn their heritage language as a first language, so that the number of fluent speakers shrinks over the next few generations until none are left. Meanwhile, the domains in which the dying language is used contract, until it is likely to be strictly a home language; often, at the very end, it is used only by, and when talking to, grandparents. In the Salish–Pend d’Oreille community in the U.S., for instance, the elders who work with me on our dictionary project have told me that the only time they now speak their language is during our weekly sessions in the summer and, during the rest of the year, at monthly elders’ meetings. A few of them also speak bits of the language with their great-grandchildren, but otherwise it is apparently no longer used even as a home language; the last couple who spoke it at home was broken when the husband died in 2001 at the age of ninety-two. (Recently, however, revitalization efforts have made limited progress in reversing the decline of the heritage language, in this as in numerous other endangeredlanguage communities.) A growing number of case studies detail the kinds of linguistic changes that occur in language death, and I will discuss some of the most salient changes in this chapter, in particular attrition, the net loss of linguistic structure and stylistic richness. I will also discuss the much rarer phenomenon of complete grammatical replacement in one or more linguistic systems, which has been claimed – somewhat controversially – for a few mixed languages around the world. Before we start examining details of cases and processes of dying, however, we need to be clear on what the exact topic is. The term “dead language” refers to two quite different things. When most people think of dead languages, Latin and Ancient Greek come to mind first, followed (perhaps) by other ancient languages, Sanskrit and Ancient Egyptian and less-well-known languages that are no longer spoken: Sumerian, Hittite, Akkadian, Old Church Slavic, Old Irish, 42

Sliding into dormancy

Old English, and others. Some of these languages, notably Sumerian, Hittite, and Akkadian, haven’t been spoken for several thousand years; others, like Old Irish and Old English, have been dead for less than a thousand years. But not all of these languages died in the sense intended in this book. Latin didn’t vanish by losing native speakers and domains of usage; instead, it developed, step by gradual step, into the modern Romance languages: French and Italian and Spanish and Portuguese and Romanian and others. All these languages diverged independently from late Latin, first as dialects of Latin and then as separate languages, splitting off from their shared parent language. So although it is true that no one now speaks Latin natively, Latin evolved into its daughter languages rather than simply vanishing without issue. The same is true, with variations, of Greek: Ancient Greek changed gradually into Modern Greek, with no break in the transmission of the language from generation to generation. Some Ancient Greek dialects did die in the sense of losing all their native speakers and domains of usage: the dialects of Homer, Herodotus, and Sappho, for instance, died out completely. With one small and now endangered exception, the only Ancient Greek dialect that survived to develop into Modern Greek was Attic, the dialect of Athens and of most Greek authors whose writings have enriched Western culture for more than two thousand years – Plato, Aristotle, Aristophanes, Sophocles, and the rest. Of the ancient languages just listed, most have left descendants, and their cultures have passed down to their descendants along with their languages (although both languages and cultures have undergone many changes). Besides Latin and Greek, these are Sanskrit, Old Church Slavic, Old Irish, and Old English. The others – Ancient Egyptian, Sumerian, Hittite, and Akkadian – have not: they disappeared along with their cultures long ago. Speakers of Sumerian and Akkadian established their empires in Mesopotamia, and the Hittite Empire later flourished in Anatolia, but all three languages died long before the birth of Christ. The Sumerian language had no known relatives at all. Akkadian belonged to a branch of Semitic that died, but its Semitic cousins, most prominently Arabic and Hebrew, appeared later in the historical record; Arabic is still extant, and Hebrew, though dead as an everyday spoken language for roughly two thousand years, lives again in modern Israel. Hittite belonged to the Anatolian branch of the Indo-European language family, and it disappeared along with its Anatolian relatives; but again, many related languages survive in the other nine branches of the family. Egyptian, distantly related to Semitic languages, did change gradually over many centuries, but eventually it too died (see the discussion in the next section). In other words, languages like Latin and Ancient Greek and Old English are dead in the sense that they themselves are no longer spoken and are no longer learned as first languages by children. But they live on in their descendants, which (unlike human ancestors and their descendants) are changed later forms of the ancient parent language. In sharp contrast, Sumerian, Akkadian, Hittite, and Egyptian are dead indeed, with no descendants, no changed later

43

44

sliding into dormancy

forms. And the same is true of the languages that are the topic of this chapter. The main difference between the recently dead languages and the ancient dead languages like Sumerian is that Sumerian is well documented and is therefore (fairly) well known to scholars. Many or most languages that have died in recent decades and that are dying today did not, or will not, leave extensive records. This circumstance, as I emphasize in Chapter 7, makes efforts to revive the languages very difficult indeed. In this chapter I illustrate the social circumstances surrounding language death by describing a few cases. (No language dies for linguistic reasons; language death is always caused by social factors in the broadest sense.) We’ll then turn to a consideration of tip, the switch from a slow, gradual decline to the sudden collapse of a language; the problem of semi-speakers; and attrition, the loss of grammatical and stylistic resources in a dying language. The final main section of the chapter describes an alternative route to language death – “death by borrowing”, in which a language’s lexicon and structures are swamped by borrowed words and structures from a dominant language. The chapter closes with a summary of the chapter’s contents, a brief discussion of some categorizations of routes to language death that have been proposed, and a comment on the unfortunate implications of terms like “language suicide”. I do not devote much space in this chapter to a third obvious route to language death: the death of a language via the death of all of its speakers. As observed in Chapters 1 and 2, this phenomenon is apparently rare in human history (thank goodness!), although the massacre of a group’s enemies has sometimes been made easier when they can be identified by the fact that they speak another language. The identification of enemies by their speech is a rather familiar theme in (at least) Western culture, most notably in the Old Testament story of the Gileadites’ massacre of the Ephraimites: And the Gileadites took the passages of Jordan before the Ephraimites: and it was so, that when those Ephraimites which were escaped said, Let me go over; that the men of Gilead said unto him, Art thou an Ephraimite? If he said, Nay; [t]hen said they unto him, Say now Shibboleth: and he said Sibboleth: for he could not frame to pronounce it right. Then they took him, and slew him at the passages of Jordan: and there fell at that time of the Ephraimites forty and two thousand. (Judges 12:5-6)

This Bible story does not signal the massacre of all the Ephraimites, because presumably only men would have been present at the battle; the Ephraimite dialect of Hebrew would therefore have survived among the women and children and men who had not fought. Still, 42,000 deaths could well have led to the demise of the dialect, and then, if that was the only dialect in which people said Sibboleth, that would leave only those Hebrew dialects which, unlike the Ephraimites’ dialect, included š (a sound like the first consonant in English shop) in their sound system. Other massacres have certainly had this effect. The government of El Salvador responded to a 1932 peasant revolt by having soldiers murder 25,000 people

Five case studies

identified by cultural or physical features as Indians; several years later, the media were still advocating that all of El Salvador’s Indian people be killed to prevent renewed revolt. As a result, Salvadoran Indians stopped speaking two languages, Lenca (an isolate, with no known relatives) and Cacaopera (a member of a small four-language family), so that they wouldn’t be easily identified as Indians. By 1970, almost no one spoke either language fluently; the Ethnologue now classifies Lenca as “nearly extinct” and Cacaopera as “extinct”, and UNESCO’s Atlas of the world’s languages in danger lists both as extinct. As with (I assume) the ancient Israelite Ephraimite tribe, the Salvadoran government’s efforts didn’t eliminate the entire Indian population; but for these two languages, at least, the vicious official response to the uprising led directly to an abrupt and ultimately catastrophic decline in their use. Another source of language death through the demise of the speech community is natural disaster. One example is the death of Tamboran, a non-Austronesian language once spoken on the Indonesian island Sumbawa. Tamboran died when all its speakers were killed by the eruption of Mount Tambora in 1815. In spite of these and a few other examples, however, languages rarely die because all their speakers die. Let’s turn, therefore, to a consideration of five case studies that illustrate the most common cause of language death: the shift by all of a language’s speakers to another language.

3.1

Five case studies

3.1.1

Case study #1: Eyak

On 21 January 2008, Marie Smith Jones died in her sleep at her home in Anchorage, Alaska. Smith Jones was the last full-blooded Eyak and (as we saw in Chapter 2) the last fluent native speaker of the Eyak language, which was indigenous to Alaska and a fairly distant relative of the large family of Athabaskan languages. For the last fifteen years of her life she was the only fluent Eyak speaker. What brought Eyak to this final state? A (or the) major factor, of course, was contact with encroaching outside cultures. One of these was the Tlingit Nation, which, before contact with Whites, spread northward along the coast of British Columbia to Alaska. Some Eyak-speaking groups they encountered along the way shifted to Tlingit, reducing the range of the Eyak language and culture. The first Whites to encounter the Eyaks, as far as is known, were Russians who tried to set up trading posts in Eyak territory toward the end of the eighteenth century. English-speaking Americans came later, during the nineteenth century, and in 1867 the United States bought Alaska from the Russian Empire. In the course of the nineteenth century Eyak territory shrank until, by the 1870s, only about 200 Eyaks remained, mostly around Cordova. Besides the general impact of the languages and cultures introduced by Whites, the loss of Eyak was apparently accelerated by an epidemic, probably smallpox

45

46

sliding into dormancy

or measles, that eliminated several Eyak villages. Then, in the last years of the nineteenth century, Whites built salmon canneries near Cordova, and the activities associated with the factories – which included using dynamite to catch fish – severely damaged the Eyaks’ traditional fishing economy. By 1933 there were only thirty-eight Eyaks left. When the linguist Michael Krauss started to study Eyak in 1963, there were six speakers, among them Marie Smith Jones and one of her sisters. But although some of the six (including Smith Jones) were fluent in Eyak, none of them used the language in their everyday life; none of the elders had taught their own children to speak Eyak because, as Smith Jones’s daughter later said, when she and her siblings were growing up, ‘it was considered wrong to speak anything but English’. They reported that the last person whose main language was Eyak rather than English or Tlingit was Smith Jones’s mother, Minnie Stevens, who died in 1961. Krauss’s goal, therefore, was to document the language as fully as possible with the help of the remaining handful of speakers: his efforts will make it possible for the descendants of Marie Smith Jones and her fellow tribespeople to revive the language, as they are now trying to do. In 1972 Krauss recorded ‘Lament for Eyak’, which was printed in a 2005 New Yorker article about Smith Jones and her language. The narrator ‘describes what it is like to be a member of a vanishing people’: K’aadih ulah uuch’ q’e’iiłi’ee. Sitinhgayuudik six.a’iinsdi’ahł. Sitinhgayuu six.a’lisłi’ahłch’aht q’al ahnuu si’ahtgayuu q’uh yaan’ q’e’ disłiqahqł al iisinh. Aan, deelehtdal dlagaxuu, ts’it dlagaxuu atxsłilahł? Atgaxłalaał.

Useless to go back there. My uncles too have all died out on me. After my uncles all died out my aunts next fell, to die. Yes, why is it I alone, Just I alone have managed to survive? I survive.

As we will see in Chapter 7, a revitalization program has been initiated for Eyak, as with three of the other four cases of language death – or dormancy – in this chapter. 3.1.2

Case study #2: Cornish

Dolly Pentreath died in 1777 at the age of 91 or 102, depending on which source you believe. Her death is noteworthy because she is claimed to have been the last monolingual speaker of Cornish, a Celtic language closely related to Welsh. The legend that she was the last fluent speaker of the language is almost certainly false; after that claim was published by an Englishman who interviewed her in her old age, several Cornish-speaking Cornishmen wrote to protest it. The story that her last words were ‘Me ne vidn cewsel Sawznek!’

Five case studies

(‘I don’t want to speak English!’) is also most likely false, but it may well represent her attitude accurately. The last native speaker may have been John Mann, who grew up in Cornwall speaking Cornish with other children and was eighty in 1914. Still, Dolly Pentreath is the most famous of latter-day Cornish speakers, in part because of the carved inscription on the tall gravestone erected by Prince Louis Bonaparte in June 1860 (punctuation, capitalization, and line breaks as in the original): Here Lieth Interred

Dorothy Pentreath who Died in 1777 said to have been the last person who conversed in the ancient Cornish the peculiar language of this county from the earliest records till it expired in the eighteenth century, in this Parish of Saint Paul This stone is erected by

the Prince Louis Lucien Bonaparte in Union with

the Revd . John Garrett Vicar of St Paul June 1860 Below this inscription on the gravestone is a Biblical text from Exodus XX,12, first in English – ‘Honour thy father and thy mother, that thy days may be long upon the land which the Lord thy God giveth thee’ – and then in Cornish. What happened to Cornish? Why, given the isolation of Cornwall on a peninsula in the southwestern corner of the island otherwise inhabited mainly by English speakers, did Cornish dwindle and finally become extinct? A major reason, analogous to the reasons in so many other cases, lies in the economic, social, and political pressure resulting from the overwhelming dominance of Anglo culture to the east and north. But Welsh speakers experienced these same pressures and yet their language is still alive, though endangered (it’s classed as “vulnerable”, the mildest level of endangerment, by UNESCO’s Atlas of the world’s languages in danger). It is said that, until about 1600, ‘almost everyone accepted that the Cornish were a separate ethnic group’, but ‘by 1700, practically no one did’. At first Cornish lost ground gradually, as more and more people shifted to English because it was the language of most landowners and was

47

48

sliding into dormancy

considered more ‘refined’. Then, in the sixteenth century, came Henry VIII’s Reformation. In its aftermath, under Henry’s son King Edward VI, the protestant government decided to replace the Latin liturgy with a new prayer book in an English translation. This caused Cornishmen to rise in rebellion, declaring that ‘we the Cornyshe men (wherof certen of us understande no Englysh) utterly refuse thys newe Englysh’. Several thousand Cornish rebels were killed in the suppression of the revolt. Although the Book of Common Prayer was eventually translated into Welsh, it was never translated into Cornish, perhaps because the rebellion linked the Cornish language with sedition in the minds of England’s rulers. The lack of a Cornish translation in turn ‘did much to hasten the subsequent decline of the language’. In the seventeenth century, during the Civil War, Cornwall strongly supported King Charles I against his opponents. This led to a new disaster for Cornwall and “Cornishness” after the king was defeated in 1646, and both the Cornish language and the sense of a separate Cornish identity declined rapidly; by 1700 the few thousand remaining Cornish speakers lived on the coast at the extreme western tip of the peninsula. The language became extinct as an everyday community language during the next century and a half. The extinction of Cornish has turned out to be temporary, however: efforts to revive the language have been under way since early in the twentieth century, though there is a significant temporal and linguistic gap between the revived language and the language as it was spoken by Dolly Pentreath and her neighbors. As discussed in Chapter 7, revitalization – “reawakening” – efforts have accelerated recently, so much so that on 24 April 2014 the government of the United Kingdom issued a press release with the headline ‘Cornish [that is, the Cornish people] granted minority status within the UK’, and in 2010 UNESCO’s Atlas of the world’s languages in danger changed the status of Cornish from “extinct” to “critically endangered” with an R indicating “revitalization”. 3.1.3

Case study #3: Egyptian

Of all the world’s languages, Egyptian has the longest recorded history: more than four thousand years of unbroken documentation, from ca. 3000 BCE to ca. 1400 CE. There are two other main contenders for the longest-written-language distinction. The first is Sumerian, which is, as we saw in Chapter 1, the earliest attested (i.e., documented) human language, with writings dating from ca. 3100 BCE to ca. 100 BCE, for a 3,000-year total; the second is Chinese, which is continuously documented from the late Shang dynasty, between 1400 and 1100 BCE, for a maximum total of 3,400 years so far. But until Chinese has been written for another millennium or so, Egyptian will remain the champion. Of course the language did not remain static throughout its lifetime; like all living languages, it underwent many changes as the centuries rolled by. Experts divide it into several stages: Old Egyptian (3000–2000 BCE), Middle Egyptian (2000–1300 BCE), Late Egyptian (1300–700 BCE), Demotic

Five case studies

(seventh century BCE to the fifth century CE), and finally Coptic, which lived from the fourth century to sometime between the fourteenth century and 1700 CE. Coptic, the last stage of the Egyptian language, is known from mainly religious texts written after the Egyptians were converted to Christianity in the third and fourth centuries CE. Nothing is known of dialect variation in ancient Egyptian, and if the language ever split into two or more daughter languages, the split is unrecorded and therefore unknown to modern scholars. What we have, then, is a more or less straight-line development from a single ancient language to a single relatively modern language. The various stages of Egyptian are written in several different scripts, from primarily hieroglyphic writing before the beginning of the Christian era to the Greek-based alphabet in which Coptic was written. What brought this language, with its illustrious history stretching over four millennia, to extinction? As with Eyak and Cornish, the ultimate cause was the encroachment of an alien culture (or rather, in the case of Eyak, two different alien cultures). In 641 CE, Egypt was conquered by Arabs and their Muslim allies, and over the next few centuries the country’s population abandoned their religion, which by then was Christian, and adopted Islam. In the same period Arabic replaced Coptic in public life, though Coptic presumably remained as a home language for some time afterward. By the eleventh century CE, Coptic was no longer spoken as an everyday language, though it survived for use in the Coptic church. Later, in the thirteenth century, Coptic grammars and other materials appeared, written in Arabic, indicating an interest in the country’s preIslam linguistic heritage; but so far no revival efforts have had significant success. 3.1.4

Case study #4: Yaaku

Early in the 1930s (according to a later report), the Yaaku people of Kenya held a public meeting at which they decided to abandon their heritage language and shift to Maasai. The two languages are unrelated: Yaaku belongs to the East Cushitic branch of the Afro-Asiatic language family, while Maasai is a member of the Nilotic branch of the proposed Nilo-Saharan family. By 1959, a visiting anthropologist found that the Yaaku people – who had changed their name to Mukogodo when they adopted the Maasai language – were mostly speaking Maasai, but still used Yaaku as a home language; only a few old men in the community lacked fluency in Maasai. In 1971, a visiting linguist found 64% monolingual Maasai speakers but no monolingual Yaaku speakers; 21% of the remaining population were bilingual in Maasai and Yaaku. Today UNESCO’s Atlas of the world’s languages in danger lists Yaaku as extinct, though this assessment may be premature: a few years ago, in another public meeting, the Yaakus decided to revive their heritage language, and the linguist they invited to help them found three fluent speakers of Yaaku in the region in 2005. The (near-)death of Yaaku can be traced more or less directly to a combination of attitudinal and economic factors. The Yaakus originally lived in forests,

49

50

sliding into dormancy

mainly in caves, and were hunter-gatherers and beekeepers. They and other hunter-gatherers in the region were called “Dorobo” by the Maasai, a pejorative term that implied poverty (because they had no cattle) and general inferiority. But the term also indicated indigenous status with accompanying land rights, in contrast to the intruding Maasai, who were numerous, rich in cattle, and dominant, but who lacked land rights. A plausible theory about why the Yaakus adopted Maasai cattle-herding culture is this: when marriages between Yaaku girls and Maasai men increased in frequency, the girls’ parents increasingly received cattle rather than the traditional beehives as bride payment. Moreover, with many Yaaku girls marrying Maasai men, young Yaaku men were also obliged to marry outside their tribe because of the paucity of available Yaaku girls; and to marry Maasai girls, the Yaaku bridegrooms had to pay with cattle. In this way, and no doubt for other reasons as well (many people are unenthusiastic about being viewed as inferior), the Yaakus adopted cattle-herding as a way of life, and they adopted other aspects of Maasai culture too – including the intruders’ language. Shifting to Maasai language and culture meant attaining greater prestige; at the same time, their original Yaaku language, ‘with its semantic emphasis on hunting, was considered unfit for a cattle-breeding society’. In addition, the language of mixed Yaaku-Maasai marriages was invariably Maasai, and the Yaaku language ceased to be a symbol of ethnic-group identity. At the end of all these processes, the Yaakus decided that their children should have Maasai as their first language, and so Yaaku died as an everyday community language. (This is of course a greatly oversimplified view of the causal relationship between the adult Yaakus’ decision and the children’s acquisition of Maasai as their first language. As anthropologists have discovered, language shift is often, if not typically, carried out by young children, and not as a direct result of a conscious decision by adults. The proximate causes are invariably complex, and in this case they are unknown, at least to me.) Nor was Yaaku the only case of a similar sort in this part of Kenya: several other Dorobo tribes also shifted to Maasai, for essentially the same set of reasons. Some Yaakus, however, are eager to revive their tribe’s original language and culture, partly for reasons of ethnic pride and partly to help establish their rights to their traditional lands. As all three of the remaining fluent speakers are about 100 years old, they are unlikely to be able to participate actively in a revival process, though at least one of them has provided data for the effort. Some of the semi-speakers – that is, people who were exposed to the language in childhood but did not learn it to full fluency – are much more active, and they will likely be the main players in revival efforts. (See Chapter 7 for further discussion of the Yaakus’ revitalization efforts.) 3.1.5

Case study #5: Mednyj Aleut

One of the world’s most remarkable languages is, or was, Mednyj Aleut. This language arose in the nineteenth century on Mednyj (Copper) Island,

Five case studies

one of the two Commander Islands in the Bering Sea off the coast of Russia’s Far East. If it has died, or if it dies within the next few years, its entire existence will have spanned less than two hundred years. Mednyj and the other Commander island, Bering, were uninhabited until 1826, when the Russian-American Company, established for the purpose of fur seal trading, resettled some Aleuts there from Aleutian Islands that lie farther to the east, closer to Alaska: the Aleuts’ job was to do the actual hunting of fur seals. From the beginning, the Russians required Aleuts under their control to learn Russian, so it is reasonable to suppose that the Aleuts on the Commander Islands were bilingual in Aleut and Russian during the nineteenth century; there is no information about how many Russians became fluent in Aleut, but it seems likely that relatively few learned the language well. Somewhat later, the Russian population on the two islands grew, but on Mednyj the Aleuts continued to outnumber Russian employees of the Company significantly. Children born to Russian men and Aleut women formed a mixed-blood (“creole”) population, and the creoles enjoyed a legal status superior to that of the pure-blood Aleuts, though still inferior to that of the Russians. Socially, the creoles were viewed with contempt by both Russians and Aleuts because they were all illegitimate in the early days of the Mednyj and Bering colonies. On Mednyj, at least, the creole population increased rapidly, but the pure-blood Aleut population did not – both because of venereal diseases and smallpox introduced by Russians and because many Aleuts were killed by the Russians in battles. Meanwhile, the fur seal trade diminished drastically during the second half of the nineteenth century, and most Russians left the Commanders after the United States bought the Aleutians from Russia in 1867. From then until the 1940s there was little Russian administrative activity on Mednyj or Bering. The legal and social distinctness of the mid-nineteenth-century creole population on Mednyj was surely the main contributor to the linguistic outcome of this contact situation. Sometime between 1826 and 1867 a new mixed language emerged, consisting of an Aleut foundation onto which large chunks of Russian verbal morphology (word structure) were grafted. The Russian pieces of a Mednyj Aleut verb were affixes (word pieces with grammatical meaning added to form new words or stems – in this case the affixes were primarily suffixes, added to the end of the stem) that expressed such grammatical features as tense (present, future, past) and person (first, second, and third person, singular and plural – ‘I’, ‘you’, ‘she’, ‘they’, and so forth); these Russian affixes completely replaced the corresponding chunks of the enormously elaborate Aleut verbal structure. Nouns were unaffected – they kept their elaborate Aleut structure. Many words were also borrowed from Russian, and some features of Russian sentence structure may also have entered Mednyj Aleut during this period, but the only sweeping replacement was in the system of finite verb inflection (that is, the set of forms used for different syntactic functions in verbs of main clauses). The result, though it is known as Mednyj Aleut (or, sometimes, as Copper Island Aleut), is not Aleut at all; it would not be intelligible to a

51

52

sliding into dormancy

monolingual Aleut speaker or, for that matter, to a monolingual Russian speaker. The creators of this new language must have been fluent AleutRussian bilinguals, because there is little distortion in either the Aleut or the Russian component of Mednyj Aleut. This places the language in the tiny set of fairly well-documented bilingual mixed languages around the world. Its particular linguistic mix – roughly, verb inflection vs. everything else – is unique even within the set of bilingual mixed languages. This makes it an irreplaceable linguistic treasure. But Mednyj Aleut is dying or maybe, by now, already dead. UNESCO lists it (under the alternate name Copper Island Aleut) as “critically endangered”; the Ethnologue lists it as “nearly extinct”, with an estimate of ten speakers as of 1995. (Bering Aleut – which, thanks at least in part to a larger proportion of Aleuts by comparison to Russians and creoles, is still an Aleut dialect, not a mixed language – was also nearly extinct as of 1995, with an estimated five speakers.) What happened to Mednyj Aleut? In many respects its fate resembles those of other languages that have slid toward extinction, including boarding schools and other educational programs that pushed people to shift to Russian, but there is one noteworthy difference: in the 1960s, the remaining speakers of the language were moved to Bering Island. Separated from their own territory and brought into close contact with Bering Aleuts as well as with Russian speakers, speakers of Mednyj Aleut rapidly abandoned their heritage mixed language.

3.2

What do these case studies tell us?

First, a brief summary of the five dead, dormant, or dying languages, the languages that replaced them, and some major (though not exclusive) factors that brought them to extinction: Eyak: Cornish: Egyptian: Yaaku: Mednyj Aleut:

Tlingit, English English Arabic Maasai Russian

encroachment; smallpox; economics rebellion: Cornish = sedition? conquest attitudes + politics, economics smallpox, resettlement

We can draw some conclusions from the routes to language death followed by Eyak, Cornish, Egyptian, Yaaku, and Mednyj Aleut. One is that language death and language shift are not strictly modern phenomena, and are not tied to the rise of nation-states; language extinction must have happened frequently in the unrecorded past as well. Yaaku was not replaced by the language of European invaders but by another indigenous African language, and Eyak might well have been doomed by Tlingit encroachment even if Anglos hadn’t taken over its speakers’ territory. Armed conquest is sometimes the beginning of the end for a language, as when Egyptian speakers were conquered by Arab armies. Cornish,

Tip

too, suffered from political reverses, first because of an unsuccessful rebellion by its speakers and then by their choice of the wrong side in the Civil War: these events caused the ruling powers to suspect Cornishmen of seditious tendencies, which in turn made it expedient for the Cornish to adapt to the majority English culture and language. This last is essentially an attitudinal factor. Attitudes helped hasten the disappearance of Yaaku as well, because adopting the language and the cattle-herding culture of the Maasais led to improved living conditions for the Yaakus. Still another recurring factor in these cases is a lack of built-up immunity to introduced foreigners’ diseases: smallpox contributed to the decline and ultimate demise of both Eyak and Mednyj Aleut. Finally, the (near?-)death of Mednyj Aleut illustrates the effect of moving a speech community to a new setting where the old ways of living and speaking lose their roots. A less gloomy conclusion to be drawn from these five case studies, however, is discussed further in Chapter 7: four of these languages, all but Mednyj Aleut, are now targets of language revitalization – or, probably more accurately, revival – efforts. In this respect they resemble dozens or hundreds of (merely) endangered languages all over the world.

3.3

Tip

So far in this chapter we have mostly examined sociolinguistic aspects of language death – combinations of social, economic, political, and attitudinal factors that have led particular languages to vanish. Before turning to the kinds of linguistic changes that characterize many (not all!) cases of gradual language death, we need to consider two other important aspects of the dying process: tip and (in the next section) semi-speakers and rememberers. The concept of tip, which we have already seen in the quotation that begins the chapter, was introduced by Nancy Dorian. In contact situations to which the term is applied, a minority language that appears to be demographically stable within its speech community suddenly and rapidly declines in response to sociopolitical change, and then vanishes: the tip is from apparent robust life to rather sudden death. Dorian’s 1981 book Language death: the life cycle of a Scottish Gaelic dialect (which was mentioned in Chapter 1) is a classic in this field, and her forty years of fieldwork on East Sutherland Gaelic have given her an unparalleled knowledge of the processes. In eastern Sutherland, she explains, ‘the end of protective isolation precipitated this tip locally, exposing Gaelic speakers to the forces which had greatly favored English nationally for several hundred years’. That is, as long as Gaelic-speaking communities had only limited contact with the dominant national language, they maintained their heritage language; but once English expanded dramatically into these Gaelic-speaking communities, in the spheres of education, economic activity, and communications, Gaelic declined very rapidly. Dorian emphasizes that the tip was not in fact as sudden as it appeared from the

53

54

sliding into dormancy

outside, noting that ‘the climate which led to [the] rapid adoption [of English] had been centuries in the making’. Nevertheless, East Sutherland Gaelic had been ‘demographically stable’ locally for centuries in spite of the growing dominance of English on the national scene, so the appearance of a sudden tip was not wholly misleading. Another example of tip is reported from Cape Breton Island, Nova Scotia, where until the mid-twentieth century Scottish Gaelic was widely spoken. During the 1930s and 1940s, however, the entire island ‘experienced a rapid linguistic tip toward English monolingualism’ – even though Cape Breton is separated into two distinct regions by geography, religion, length of Scots settlement, presence vs. absence of elite monolingual English speakers, Englishlanguage education, and other sociolinguistic features. The tip mechanism was apparently the decision by Gaelic-speaking parents to stop speaking Gaelic to their children; as one language consultant commented, ‘Well, I guess when I got near school age they started talking English.’ Gaelic became confined to adult conversation, used sometimes to talk about things they wanted to keep secret from the children. The parents’ decision was grounded in features of their linguistic ideology, which came to favor English monolingualism to the point that they in effect imposed it on their children.

3.4

Semi-speakers and rememberers

Most linguists who have conducted fieldwork on dying languages have found quite significant differences in the language skills of the last few speakers who comprise their pool of consultants. The signs of limited fluency are often easy to identify – halting, hesitating speech patterns, difficulty in offering a translation of a simple sentence, inability to provide a coherent narration, gaps in vocabulary that designates items of material cultural and natural environment and other everyday concepts, grammatical simplifications by comparison to the language as spoken by fully fluent speakers, clear evidence of interference from a dominant language. (This last clue may not be detectable unless the linguist has independent sources of knowledge of the dying language’s grammar, for instance closely related sister languages or old documentation dating from the days when the dying language was the main everyday language of the speech community.) Sometimes, however, identifying a semi-speaker can be extremely difficult – maybe even impossible – in the absence of independent sources of information on the dying language. This is especially true because many semi-speakers conceal their lack of fluency by their skill in using fixed phrases appropriately, so that their speech is neither halting nor hesitant. Nancy Dorian commented recently that the Scottish Gaelic semi-speakers that she worked with ‘possessed a remarkable degree of sociolinguistic skill in the deployment of their linguistic resources’, and that they ‘tended to be short-burst speakers rather than sustained conversationalists’. When there is a choice, fieldworkers will choose the most

Semi-speakers and rememberers

fluent speakers for purposes of primary documentation; but sometimes there is no choice – there may be only one or two speakers left, and with limited fluency. Even before the final stages of an endangered language’s decline, degrees of fluency among its speakers may vary widely. In her work with the Scottish Gaelic speech community of coastal East Sutherland, Dorian found a continuum of speakers, from Gaelic-dominant at the “most fluent” end of the continuum to skilled Gaelic-English bilinguals and finally to the “least fluent” speakers, English-dominant and able to speak only very imperfect Gaelic. She labeled this last category “semi-speakers”, and the term has become standard as a label for an end-stage speaker whose language skills are demonstrably not sufficient for true fluency. In semi-speakers, as numerous studies have now shown, the markedly non-fluent manner of speaking is accompanied by linguistic features that distinguish the linguistic structure from that of fluent speakers. In particular, in the Scottish Gaelic variety studied by Dorian, changes that eliminated irregularities were the most salient grammatical marker of semi-speaker status. In other words, ‘reduction in structure accompanying reduction in use’ characterized the semi-speaker. Of course what is true of East Sutherland Scottish Gaelic is not necessarily true of every case of language death, or of every group of semi-speakers of a dying language. Here, as in other contact situations and indeed in every language situation, deterministic predictions of language behavior are wildly unlikely to be universally valid. Here is an example to illustrate the problem. It is often claimed – and this is surely true for some dying languages – that the more loanwords and structural features an endangered language has received from a dominant language, the closer the receiving language is to death. This seems to imply that the most fluent remaining speakers can be counted on to speak a relatively “pure” form of the language, unadulterated by foreign material. But Jane Hill and Kenneth Hill, in their long-term study of Nahuatl (the direct descendant of the language of the Aztec Empire in pre-contact Mexico), found that the most fluent speakers of dying varieties of Tlaxcalan Nahuatl used more borrowed Spanish elements than less fluent speakers did. The Hills observed that Spanish loanwords comprised up to 40% of all words used in everyday conversation by fluent speakers, a huge percentage. The distribution of speakers along a fluency continuum is also not a universal characteristic of dying languages. In some speech communities there is a sharp division between the elders who are the last fluent speakers of the language as it was spoken before it became endangered and younger people who are learning it as a second language. This can happen if the language’s decline is relatively sudden, rather than a slow, gradual process that stretches over many decades and many generations. A case in point is the Salish-Pend d’Oreille language of western Montana. Cultural pressure from English speakers began in about the mid-nineteenth century, but it was two twentieth-century events that produced the sudden collapse of a viable speech community: as described in Chapter 2, the rapid shift to English was caused by a combination of the government-decreed

55

56

sliding into dormancy

opening of the reservation to White settlement and forced attendance at boarding schools in which the tribes’ children were prevented from speaking their native language. Because the effect of the schools and the greatly increased presence of monolingual English speakers on the reservation was to produce Englishdominant bilinguals, and because parents stopped passing Salish-Pend d’Oreille on to their children, the linguistic divide in many families was total. As a result, the number of speakers of the language shrank steadily during the last half of the twentieth century as older generations passed from the scene and were replaced by generations of monolingual English-speaking tribal members. Vigorous efforts to revitalize the language are now under way, but the fluent elders who act as advisors are too few to enable the language teachers to achieve full fluency in the language as it is spoken by the remaining fluent native speakers, so that child learners also cannot achieve full fluency in the traditional version of the language. The language teachers and the child learners may now count as semispeakers, but they are not at the end of a continuum of fluency, as suggested by the usual notion of semi-speaker. There is no continuum; there is a handful of fluent elders, and there are younger speakers – teachers and learners – who are in effect creating a new Salish-Pend d’Oreille language that is based on, but is not a direct descendant of, their ancestors’ language. We’ll return to this topic in Chapter 7, where we’ll see that many or most revitalization programs do not, and in many cases cannot, aim to revitalize the version of a language that is or was spoken by the last traditional fluent speakers. It may, as we have seen, be difficult to distinguish a semi-speaker of a dying language from a fluent speaker. The distinction can be of crucial importance for the success of a documentation project, because the data provided by semispeakers will not reflect the language as it was spoken when it was the usual language of everyday communication in the speech community. But if there is no choice – if the most skilled users of the language are in fact semi-speakers – then the linguist must work with them, doing what is sometimes called salvage linguistics (linguistic fieldwork undertaken after the last fully fluent speakers are gone), collecting and analyzing what can be recovered from language users with severely limited proficiency. Alan Dench, in his fieldwork in Western Australia, found himself in this situation when he undertook an analysis of the Aboriginal language Yingkarta, and warned the readers of his published grammar accordingly: The competence of the speakers I interviewed ranged from semi-speaker abilities through to limited fluency ... the quality of the data is not comparable with that which may be collected from the fluent speakers of fully viable languages, and the current analyses must be considered in this light.

From the viewpoint of discovering and analyzing a dying language’s lexicon and structure, there is an even more problematic category of endstage non-fluent speakers of a dying language. These are people who are, so to speak, below the lowest end of the fluency continuum. They are called

Attrition

rememberers – community members who can provide some words and phrases of the dying language, but who cannot use it in conversation. This last characteristic may be hard to judge, because often a rememberer is isolated from the remaining fluent speakers and/or semi-speakers, if there are any; and if so, they have no opportunity to use the language. In any case, the task of trying to analyze a dying or (more likely) an effectively dead or dormant language solely from the utterances of rememberers is a desperate enterprise, sure to be met with limited success at best. Even this situation is not necessarily completely hopeless for prospects of revitalization, however, given the fact that efforts to revive and revitalize even extinct languages like Cornish have enjoyed some success; another such example is the Algonquian language Miami, which we will see in Chapter 7. But in these and other (so far partially) successful cases of revival, the revivers were able to draw on substantial documentation of the languages and/or of their very close relatives. A similar example is the Australian Aboriginal language Gamilaraay, whose revival has been based in part on the contributions of rememberers and semi-speakers, but also in part on the results of fieldwork conducted by Stephen Wurm in 1955 with the last fluent native speaker of the language; Gamilaraay also has two close relatives. The difficulty of revitalizing a language is greatly increased if the language has no close relatives and is known only from a few remembered words and phrases.

3.5

Attrition

Attrition in a dying language is the loss of words and structural features, with no replacement features taking their place. Imperfect learning is responsible for the loss of linguistic material: as fewer and fewer community members learn the language and use it in fewer and fewer areas of daily life, new learners have too little access to the language, and often too little motivation, to achieve fluency at the level of earlier speakers. The attrition process is gradual in the sense that it is likely to happen over several to many generations. But it apparently does not usually occur gradually over a single speaker’s lifespan: people who learned the dying language as their first or only language in early childhood, when it was still in regular everyday use in the community, and who spoke it regularly for some years before switching almost entirely to another language, do not show the same linguistic loss late in life as do semi-speakers, who never learned the language thoroughly in the first place. Elderly speakers may be rusty, certainly, if they haven’t spoken the language for many years; but it is likely to come back if they find themselves in a situation that encourages its use with other (once-)fluent speakers. This may be more true of structural features than of lexicon, however. When attrition occurs, the most obvious loss is in specialized lexical domains (a topic we’ll return to in Chapter 4) – specifically, in vocabulary pertaining to aspects of traditional culture that are being lost along with the language. In

57

58

sliding into dormancy

a hunting-and-gathering tribe, for instance, the hunting and gathering traditions must die when tribal members are fenced out of their former hunting grounds and a big-box Costco store is built on top of their age-old favorite spot for collecting edible roots. If the tribal economy shifts to farming, even the remaining hunting grounds and gathering locales may be neglected, and spending less time in wild areas means paying less attention to the creatures that live there and the plants that grow there. When the people begin to go to dominant-culture doctors instead of consulting traditional healers, knowledge of medicinal plants as well as food plants will fade. The effect of these and other cultural changes erodes the language’s lexicon until the few remaining speakers no longer know (if they ever did know) the old names of animals and plants and other features of the natural world around them. Here’s a small example that illustrates the kind of loss of specialized vocabulary that is characteristic of attrition. In this case the lexical loss is a case of forgetting rather than never learning – the speaker had almost surely forgotten many words he knew in his youth, when his community’s language was in wider use. The example comes from Alan Dench’s fieldwork on the Australian Aboriginal language Martuthunira, whose path to near-extinction was outlined in the quotation at the beginning of Chapter 1. Years ago, when he was doing salvage linguistics on the language, Dench worked with the last two speakers, both of them elderly, eliciting vocabulary items. One of the words on his list was ‘wild onion’, and speaker #1 translated this as ngal.yu. But later, when Dench asked speaker #2 how to say ‘wild onion’, he was told that it was partunya. Dench returned to the first speaker and asked him whether indeed partunya was the regular Martuthunira word for ‘wild onion’. ‘Oh, right, yes, it is,’ said speaker #1. ‘But then why did you say ‘wild onion’ was ngal.yu?’ Dench asked. ‘Well’, the old man said, ‘Panyjima speakers say ngarlku, and Yindjibarndi has ngarku, and Kurrama has ngartku; so it ought to be ngal.yu in Martuthunira!’ This example illustrates at least three points. First, if speaker #1 had been the very last speaker of Martuthunira instead of one of the last two, Dench would have recorded a plausible but erroneous word for ‘wild onion’, and he would not have discovered the language’s original word for this plant. Second, speaker #1 typifies the extensive multilingualism – in his case, in three Aboriginal languages closely related to his own, in addition to English – that characterizes Aboriginal Australia and many other parts of the world. Third, the speaker had enough knowledge of which sounds in his language corresponded to particular sounds in three other Aboriginal languages that he could construct a new word based on the pronunciation of words for ‘wild onion’ in the other languages. To give an idea of the kinds of culturally specific words that are lost in a dying language, here is a list of the English meanings of some words connected with traditional culture that have been forgotten by current tribal elders who are among the last native speakers of Salish-Pend d’Oreille. All the items on this list came up in my fieldwork sessions when elders said, on being asked for a word for an item or concept, that they knew there had been a word for

Attrition

that, but they’d forgotten it: avalanche, wind coming up when you’re in the mountains, snow blows off the trees, impassable mountainous terrain, a kind of moss found on rocks (used as baby diapers), it’s getting narrow (said of a trail that’s becoming too narrow to pass on horseback), cottonwood sap, a baby mountain goat, a male bighorn sheep, a female bighorn sheep (they do still know a general word meaning ‘bighorn sheep’), a bighorn sheep’s horn, prairie dog, minnow, the bugling of a bull elk, the sound of a fawn bleating, the sound of a ruffed grouse drumming, a bird molts, a raft lying on the water, eat a bit of raw meat of a newly killed deer or elk. Specialized vocabulary is of course not the only type of word that disappears when a dying language undergoes attrition; ordinary words also vanish from the language as its use declines. Modern elders who are fluent in Salish-Pend d’Oreille haven’t been able to think of words for squint, stutter, lisp, wheeze, funny, the pupil of the eye, spinal cord, a splint, weak eyes, an animal skull, make ice form, beckon with the head (summoning someone), a baby beginning to take notice of its surroundings. Again, they know there were old words for all these things, but they can’t think of them. And they often comment, when a word is called to mind by mention of its occurrence in the big nineteenth-century dictionary of the language, that they ‘haven’t heard that word in a looong time’, not since the old people died. A typical example is kw ’ëxw ét’cn ‘her hair is cut short’ (something their ancestors considered inappropriate except for widows and widowers, who cut their hair off straight to show they were in mourning). Loss of vocabulary is by no means the only kind of attrition, but it is the only kind found in Salish-Pend d’Oreille. As noted earlier, there seem to be no semispeakers at the lower end of a continuum of gradually decreasing fluency in this language. There are fluent elders, some of them admittedly rusty because they now speak Salish-Pend d’Oreille only occasionally, and there are new learners whose access to the language as spoken by fully fluent native speakers is limited. This means that Salish-Pend d’Oreille is dying (unless and until the revitalization program is successful) “with its morphological boots on” – that is, without significant simplification in its structure. It is almost certain that some of the fluent elders never learned some of the old words for the concepts listed here (and others as well); in that sense, their lack of knowledge is a matter of incomplete learning rather than forgetting. In other instances, though, elders have forgotten words that they once heard from older speakers but haven’t heard in many years and maybe never did use themselves. Before turning to a consideration of structural changes in attrition, I should emphasize that the loss of specialized and even some nonspecialized vocabulary as a result of cultural changes is hardly confined to dying languages. To give just a few examples, the names of most parts of a horse’s harness are obsolete in today’s English, or at least unknown to the vast majority of English speakers, and most of my undergraduate students have no idea what carbon paper or a ditto machine or a mimeograph might be. Since English is one of the most richly documented languages in the world, recovering the vocabulary of

59

60

sliding into dormancy

(for instance) horse harnesses would be an easy matter; anyone who wants to know the words can simply google them. (I did this: googling ‘horse harness parts’ took me quickly to a Wikipedia page with a long list of parts.) If for some reason horses were to reassume their formerly vital role as beasts of transport in mainstream Anglo culture, learning this vocabulary would be no problem. Similarly, it would be possible to recover specialized vocabulary in the process of reviving an extinct (or dormant) language, provided that the language either had a written tradition or had been well enough documented to include the specialized vocabulary items. When we examine structural changes in dying languages that are undergoing significant attrition, we find a great deal of variability in the linguistic effects. This is unsurprising, both because language change is notoriously unpredictable and because of the large amount of variability among individual speakers of dying languages; but it makes it difficult to draw useful generalizations from the cases reported in the literature. In the phonology (the sound system), the most-cited prediction is that a dying language will lose items and processes from its sound system, yielding a simpler overall system. A second widely accepted prediction is that phonemes (distinctive speech sounds, those that are capable of distinguishing words) that are shared with the dominant language will not be lost, even if they are quite exotic from a crosslinguistic viewpoint. And a third is that robust phonemic distinctions, those that distinguish many words, are likely to be maintained at least until a late stage in the decline of the language. A related proposal is that when two sounds merge in a dying language, the one that is more marked universally (less common in the world’s languages, harder to learn) will probably be the one that disappears – unless the marked member of a two-member distinction has become a salient symbol of the speech community’s identity, in which case it may be not only preserved but extended well beyond its previous domains of occurrence. Here are a few typical examples to illustrate these predictions. The Vilela language of Argentina’s Chaco region – which is classified by the Ethnologue as “nearly extinct” and by UNESCO as “extinct”, but which has two remaining elderly speakers who are isolated from each other – is a member of the twomember family Lule-Vilela. (Its only sister language, Lule, was last reported as still spoken in 1981; it is now considered extinct and is not mentioned in either the Ethnologue or the UNESCO listing.) Both of the last two speakers of Vilela showed evidence of weakening of the velar/uvular distinction (that is, the distinction between velar consonants, pronounced with the back of the tongue near or at the soft palate, the velum, and uvular consonants, pronounced with the back of the tongue near or at the uvula at the back of the oral cavity). Uvular consonants are much, much less common than velar consonants in the languages of the world, so the uvular consonant is considered to be the universally marked member of a velar/uvular distinction. In accordance with the prediction that the marked member of a pair of consonants will probably be the one to disappear, the uvular stop /q/ was being replaced by the velar stop /k/ in Vilela, as in joko

Attrition

‘giant armadillo’ vs. the older form joqo that was reported for the language by an earlier fieldworker. This same change is reported for other indigenous languages of Latin America, for instance in Tuxtla Chico Mam (a member of the Mayan language family), where the /q/ of viable Mam varieties has been replaced by /k/. For both languages the fieldworkers note that this change may be at least partly due to the influence of Spanish, the dominant language for both Vilela and Tuxtla Chico Mam, since Spanish has only a velar /k/, not a uvular /q/. At least one of the last Vilela speakers was also forgetting the voiceless (pronounced without vocal cord vibration) lateral fricative /ł/, replacing it with /š/ – which is far more common crosslinguistically, and which (unlike /ł/) is found in Spanish. Significantly, however, he recovered the /ł/ fricative after he had worked for some time with a visiting linguist; for instance, he pronounced the word for ‘woman’ as [kiše] in late 2003, but by the end of 2004 he was pronouncing the word consistently as [kiłe], the form that is attested in earlier materials collected on this language. A slightly different kind of simplification is exemplified by the degemination of double (geminate) consonants between vowels in McKeesport Hungarian, a dying variety of Hungarian spoken in an immigrant community near Pittsburgh, Pennsylvania. This is not the loss of a segment, but a phonological assimilation to English, which has no double consonants within a morpheme (that is, the smallest unit of grammatical analysis, including, in English, words like run and photo as well as grammatical elements like un- and -er in unhappier). (English does have geminate consonants, pronounced as long consonants, in compound words, e.g., a long k in bookkeeper and a long s in housesitter.) Attrition in phonological processes can also be exemplified from McKeesport Hungarian: both voicing assimilation (making the pronunciation of two or more sounds more similar to each other) in consonant clusters and palatal assimilations have been lost. In addition, the dying dialect displays decay of the vowel harmony system, according to which, in other varieties of Hungarian, the vowels in a word change to match (harmonize with) each other in certain phonetic features. An example of a phonological process that has been lost with a somewhat unusual effect is found in a language of El Salvador, Teotepeque Pipil, a dying variety of the Uto-Aztecan language Pipil. A phonological rule devoicing word-final non-nasal resonant consonants has been lost through generalization of one or the other allophone: /l/ is now voiceless in all positions in the word, not just at the end of the word, while /w/ and /y/ are now voiced in all positions. The latter effect is the expected one, since it is the marked allophone – the voiceless one – that is lost when the other allophone is generalized; the generalization of the voiceless allophone of /l/ is surprising, given the rarity of this sound in the world’s languages. And finally, an example in which a marked sound – or, in this instance, a whole series of marked sounds – has been retained and extended far beyond its earlier distribution is found in Xinca, a Guatemalan language variously

61

62

sliding into dormancy

identified as “critically endangered” (UNESCO) or “extinct” (the Ethnologue). In the Jumaytepeque variety of this language, glottalization of consonants was extended until all possible consonants were glottalized, not just those consonants that inherited glottalization from Proto-Xincan. (A proto-language is the parent language of a language family, that is, the language from which all the other languages in the family are descended.) This overextension of glottalization was almost certainly deliberate, a way of emphasizing the distinctness of the Xinca language from the last speakers’ dominant Spanish language. Attrition in the morphology and syntax of a dying language, as in the phonology, results in overall reduction of the morphosyntactic systems: elimination of categories, elimination of distinctions, and analogic leveling of paradigms, a process that does away with alternations within a given morpheme. In Vilela, for instance, the distinction between inclusive ‘we’ (including the hearer as well as the speaker and perhaps others as well) and exclusive ‘we’ (including the speaker and one or more others, but excluding the hearer) – which does not exist in Spanish – has been lost; the same change is occurring in the Australian Aboriginal language Warlpiri, where the merger of pronominal forms for inclusive and exclusive ‘we’ has resulted in differing individual choices of the inclusive or exclusive form to express the merged category. In Vilela, verbs have lost their first- and second-person agreement markers (though one consultant partially recovered the old agreement markers in later elicitation sessions). In a partly similar set of developments, McKeesport Hungarian has lost the agreement system between quantifiers (words indicating quantity) and their head nouns, between predicative adjectives (such as clever in he is clever) and their nouns, between relative pronouns (an English example of a relative pronoun is who in the man who walked) and their antecedents, and between subjects and verbs. The aspectual system was also significantly eroded in the last two Vilela speakers’ speech, from six aspects in 1970 to a single aspect about thirty years later. Similar changes have affected the case system of the Australian language Dyirbal, where one case affix has been generalized to fulfill several different case functions, and McKeesport Hungarian, which has lost most of the Hungarian case system. Dyirbal has also undergone regularization of irregular verb paradigms through analogic leveling. In McKeesport Hungarian, preverb-verb constructions have been simplified, possessive marking (adding affixes to express a pronominal possessor like ‘my’, ‘your’, ‘their’, etc.) has been reduced or lost in certain constructions, and imperative suffixes have been lost. Complex sentence types tend to be largely or entirely missing in semispeakers’ version of a dying language. This is probably what accounts for Leonard Bloomfield’s reaction, in 1927, to the ‘few threadbare [syntactic] models’ part of the utterances of a semi-speaker of the Algonquian language Menomini: White Thunder, a man round forty, speaks less English than Menomini, and that is a strong indictment, for his Menomini is atrocious. His vocabulary is small; his inflections are often barbarous; he constructs sentences on a

Attrition

few threadbare models. He may be said to speak no language tolerably. His case is not uncommon among younger men, even when they speak but little English. Perhaps it is due, in some indirect way, to the impact of the conquering language.

Other examples are easy to find in the literature on language death. Semispeakers of Cupeño and Luiseño, for example, used fewer subordinate clauses by comparison to the usage of former speakers as recorded in old materials. In concluding this section on attrition in languages that are sliding into dormancy, I should emphasize again that the changes that characterize attrition do not differ in kind from changes in languages that are not dying or endangered. The parallels are especially striking in instances of contact-induced language change in viable languages. Of course, given my preferred definition of contactinduced change, all examples of attrition are in fact contact-induced changes: for me, a linguistic change is caused at least in part by contact with other languages if it is less likely to have occurred outside a particular contact situation. This definition is broad enough to include both direct influence from the language that is replacing the fading language and simplifications that have nothing to do with the influence of dominant-language structure. Here are a few typical examples of contact-induced changes in non-dying languages, starting with phonology: /l/ and palatalized /ly / (an l pronounced with the tongue body arched up toward the hard palate) are claimed to have merged in Czech as a result of German influence (German has no such distinction); Swahili, unlike almost all its sister Bantu languages, has lost phonemic tone distinctions (that is, pitch variations that distinguish words) under the influence of Arabic, a non tonal language; vowel harmony has been lost in urban varieties of the Turkic language Uzbek, together with front rounded vowels (tongue pushed forward, lips protruded) and the high back unrounded vowel (tongue pulled back and up, lips not protruded), due to influence from the Iranian language Tajik, which lacks these marked features. In morphosyntax, the Dravidian language Brahui of Pakistan, under the influence of the Iranian language Baluchi, has lost the characteristic Dravidian distinction between exclusive and inclusive ‘we’; some Semitic languages spoken in the highlands of Ethiopia have lost their dual number category (singular, just one; dual, two; plural, more than two) and replaced their original initial or medial verb placement with Subject-Object-Verb (SOV) word order under the influence of Cushitic languages; and English, which is about as far from endangered status as one can imagine, has lost almost its entire system of noun declension, including case and gender – mostly from internal causes, but possibly in part due to Norse and (less likely) French influence. What distinguishes these simplifying bits of language history from cases of attrition, then, is not the changes themselves taken one by one, but rather the fact that only attrition results in overall loss of linguistic structure. Ethiopic Semitic, for example, has also acquired features from Cushitic languages that complicate its structure – notably a series of labialized velar consonant phonemes

63

64

sliding into dormancy

(pronounced with protruded lips), e.g., /kw / (vs. a plain /k/, which was already a phoneme in Proto-Semitic) and a causative verb formation (one in which someone causes someone else to do something) that is marked by a double prefix modeled partly on the Cushitic double-suffix causative formation and partly on a single-prefix formation that is found elsewhere in Semitic. And even English, which certainly has a much simpler morphology than its Old English ancestor had, has such a complex syntactic system that it would be difficult to argue for overall morphosyntactic simplification.

3.6

An alternative route to language death: grammatical (and lexical) replacement

So far in this chapter we have seen two ways in which languages die: they die when all their speakers die, or they die by losing speakers to one or more dominant languages until they have no remaining speakers. There is one other route by which languages die – a somewhat controversial one, in part because the hypothesized processes have not (at least not yet) been observed as they occur. This is the rare phenomenon of language death by complete grammatical and large-scale lexical replacement: speakers of an endangered language borrow more and more linguistic material from a dominant language, structures as well as words, to the point where their heritage language has been entirely absorbed into the dominant language. But why would this happen? The answer seems to lie in a stubborn resistance to complete cultural assimilation to a dominant group. It presumably starts with ordinary, if extensive, borrowing – incorporation of many words and some structural features from the dominant language. Then, instead of simply shifting to the dominant language and bringing about language death through the usual process of losing all the speakers, the speakers of the endangered language continue borrowing material from the dominant language until little or nothing is left of their original language’s structure. Here is a description of such a process for Laha, an Austronesian language spoken in Central Maluku, Indonesia, where the dominant language is Ambonese Malay, an Austronesian language that is rather distantly related to Laha: All speakers of Laha are fluent speakers of Ambonese Malay ... Laha has maintained its indigenous language in the face of increasing pressure from Ambonese Malay but only at the expense of drastic revision of its grammar ... Bit by bit the grammar of Laha has become nearly interchangeable with Ambonese Malay grammar. This adaptability in the Laha language has contributed to its survival.

In other words, the vocabulary of Laha is still primarily native, but the grammar is borrowed; this means that Laha itself, as a whole language passed down from generation to generation, is dead. If the description is correct, only the lexicon is now transmitted to younger generations. In this instance, the fact that the two

An alternative route to language death: grammatical (and lexical) replacement

languages are related would have facilitated grammatical replacement, because there would be few or no structural barriers to transfer of structural features: most of the basic categories might have been similar before massive borrowing occurred. This is also true, only more so, of Votic, which belongs to the Finnic branch of the Uralic language family. It has been infiltrated by a closely related language, Ižora (also called Ingrian), through a long period of intimate contact: ‘Ižora words and grammatical features made their way into Votic almost imperceptibly, until they achieved preponderance. Then the language of these Vots was no longer Votic, but Ižora.’ This would put Votic one step beyond Laha in a continuum of borrowing leading to total replacement of the linguistic material of one language with the substance of another; Laha has retained its native vocabulary, but Votic has replaced its lexicon along with its structure. Even more than in the Laha-Ambonese Malay case, Votic and Ižora are such close relatives that they would have shared the great majority of their phonological and morphosyntactic structures before any borrowing occurred. The same cannot be said of the Ma’a (Mbugu) language of Tanzania, a bilingual mixed language with a lexicon inherited from the non-Bantu heritage language – most likely a Cushitic language, but in any case unrelated to Bantu languages – and the rest of the lexicon and virtually the entire grammar from Bantu. Structurally, Cushitic and Bantu languages are very different indeed: their morphosyntactic categories and phonological features differ in a great many ways. In Ma’a documents from the 1930s there are a few minimal traces of nonBantu structure, but in more recent data these traces have vanished. Ma’a is now essentially dead, retained only in the form of some vocabulary; the dominant language for its speakers is a Bantu language, Shambala. As with Laha, the original Ma’a language (whatever it was) no longer exists. One characteristic of the Ma’a people, at least until recently, was a determination not to assimilate to the dominant Bantu-speaking culture that surrounded them; they also resisted Arabic and Western influence, and were famous for keeping themselves separate from their Bantu-speaking neighbors. These three languages are the only ones I’ve found in the literature whose demise has apparently happened as the end stage of a gradual process of massive borrowing. It may be, of course, that other such deaths have occurred, and that we will never know about them because – like Votic – the languages have vanished by turning into other languages. This is certainly one way that dialect death occurs, so it wouldn’t be astonishing if it also happened fairly often with different, but closely related, languages. A case like Ma’a, involving unrelated and structurally very different languages, would be expected to be much rarer; but once we admit the possibility that a language can be borrowed to death, there is no principled way of ruling out such a process in a case like that of Ma’a. A powerful sociolinguistic motivation would be needed – an extremely stubborn resistance to total cultural assimilation, as with the Ma’a people (at least in former times) – because otherwise it would be hard to understand why the people didn’t simply shift to the dominant language.

65

66

3.7

sliding into dormancy

Summary and commentary

This chapter has focused primarily on gradual processes leading to language death: the “tip” that sometimes occurs toward the end of the process, the problem of semi-speakers and rememberers, and attrition in many, but not all, dying languages. I have also touched on the most tragic route to language death, when all the speakers of the language die, and I’ve described a third route, massive borrowing that ends in replacing all of a vanishing language’s linguistic structural material and much, most, or all of its lexicon as well. Two topics remain for this final section. First, we will look at three authors’ proposed models for the analysis of language death, considering them in light of the analyses and examples in this chapter. And second, I will assess some of the more dramatic labels for language death that have been used in the literature. In 1992, Hans-Jürgen Sasse proposed a theoretical model of language death. Drawing on his own long-term investigation of the variety of Albanian that is spoken in Arvanitika, Greece (a language we met in Chapter 2), and on Nancy Dorian’s work on East Sutherland Scottish Gaelic, he divides the general topic into three subtopics: external setting, which was exemplified in the five case studies described early in this chapter; the speech behavior of the dying language’s speakers – basically, in the words of the sociolinguist Joshua Fishman, ‘who speaks what language to whom and when’ (although Fishman was talking about sociolinguistic studies in general, not specifically about language endangerment); and structural changes in the dying language, primarily attrition. Two other scholars who have written about language death, Lyle Campbell and Martha Muntzel, have classified language death situations into four types according to their sociolinguistic characteristics and their linguistic correlates: (1) sudden language death, in which the language dies when all the speakers die; (2) radical language death, also a rapid process, in this case because ‘speakers stop speaking the language out of self-defense, a survival strategy’; (3) gradual language death, the usual case, which features gradual loss of speakers over several to many generations and a developing continuum of fluency in which semispeakers display extensive attrition; and (4) “bottom-to-top” language death, in which the language is lost as a vehicle for everyday communication but is retained for such formal purposes as religious ceremonies. Of the five case studies described in this chapter, at least two – Eyak and Cornish – are instances of gradual language death. Egyptian probably also belongs in this category, but it also fits into the “bottom-to-top” death class, because it was and is retained as a ritual language in the Coptic Church. In other words, the four classes are not all mutually exclusive. Yaaku and Mednyj Aleut appear to be candidates for class (2), radical language death – Yaaku because its speakers decided to abandon their heritage language, and Mednyj Aleut because the resettlement of the speech community to a Russian-intensive environment on a different island seems to have caused the language to disappear rapidly. Salish-Pend d’Oreille probably also belongs in this category, given the sharp

Summary and commentary

break in the transmission of the language and the resulting lack of a fluency continuum and semi-speakers. I mentioned one case that might perhaps go into category (1), the Ephraimites’ dialect of Hebrew, although as mentioned earlier, some speakers of the dialect surely survived the massacre; this case might best be split between categories (1) and (2). The cases of Tamboran and Susquehannock would fit in (1). Some less violent instances of dialect death would belong in category (2), if parents urge their children to abandon a nonstandard dialect in favor of a standard one. A third classification of language death phenomena focuses on the linguistic processes in cases of attrition. This is Anna Fenyvesi’s division of the processes into three categories: borrowing alone – namely, changes in the dying language that make it more similar to the dominant language but do not simplify it; attrition alone – changes that simplify the dying language but do not make it more similar to the dominant language; and both borrowing and attrition – changes that both simplify the dying language and make it more similar to the dominant language to which all its speakers are shifting. In her study of McKeesport Hungarian, the largest number of changes fell into the third category, with changes of the first type being a fairly close second. The smallest number of changes were those that could confidently be attributed to the second process, attrition alone. Almost all of the changes in McKeesport Hungarian cited earlier in the chapter in fact belong to Fenyvesi’s third category, because all of them both simplify the language and make it more like English: loss of voicing and palatal assimilations in consonant clusters (no such processes in English, at least when a syllable boundary intervenes), loss of most cases (English has almost no case system), simplification of preverb-verb constructions (no such constructions in English), loss of possessive marking (English lacks the type of possessive marking found in ordinary Hungarian), and loss of imperative endings (none in English). Just one change, degemination of consonant sequences, might possibly belong in her first category, borrowing only: English does permit geminate consonants, though only with a morpheme boundary between them, as in the bookkeeper and housesitter examples cited earlier. As with Campbell’s categories, Fenyvesi’s are not necessarily mutually exclusive; some changes arguably fit in two different categories. The final topic to be mentioned in this chapter (briefly) is the issue of labels for varieties of language death. The designations tend toward the lurid, and some of them have achieved quite general currency: googling them one day in 2009, I got 25,000 hits for “language suicide”, 26,800 for “language murder”, and 37,100 for “linguicide”, but a mere 2,000 for “linguistic genocide”. (A re-search in 2014 yielded, for some reason, much lower figures for the first three terms but a much higher figure for “linguistic genocide”.) Some conflicting definitions can be found in the literature. Language (or linguistic) suicide, for instance, is generally defined as a situation analogous to that of Yaaku, where the speech community decides to abandon its heritage language, but one author, April McMahon, defines it instead as a situation like that of Laha or Votic – where ‘the less prestigious of two closely-related languages ... progressively borrows

67

68

sliding into dormancy

words and constructions from the more prestigious language, until the two eventually become almost indistinguishable.’ (The ‘closely-related’ condition would rule Ma’a out of this category, but the process should be essentially the same regardless of whether the languages involved are related.) It seems to me that these terms are overblown, and that they contribute nothing of substance to discussions of language death. As we will see in the next two chapters, the death of any language is a sad event for a speech community that views its heritage language as an essential part of its cultural identity, and for linguists’ hope of using the widest possible range of linguistic data in their efforts to arrive at a better understanding of human cognition. But using labels of violence tends to distract attention from the real and complex sociolinguistic issues surrounding processes of language death. A term like “suicide” appears to blame the victim, the speech community that is losing its heritage language; the term “murder” suggests that dominant-language speakers are actively killing the language, as opposed to creating – often without intending any such result – conditions under which a minority speech community finds it impossible to maintain its own language. And “linguistic genocide” evokes the Holocaust, a comparison that is surely detrimental to efforts to encourage public interest in saving endangered languages: there is, after all, an enormous difference between murdering millions of people and merely urging or even forcing a community to shift to a dominant group’s language. Admittedly, the term “language death” itself is the inspiration for all the other lethal labels, and replacing it with a process term like “falling asleep” – explicitly leaving the door open for revival or “awakening” – might be preferable. But in spite of controversies about the label, “language death” has been an important topic in the field of linguistics for several decades now, and it is firmly entrenched in the literature; as I said in Chapter 1, it still seems suitable for use in discussing general and permanent outcomes of language endangerment, while “dormant” and “sleeping” are good terms to use when a language has lost all its speakers but is being, or might yet be, revived. The other labels, “language suicide” and the rest, are more recent; they could easily be abandoned with no loss to analytic clarity. I think they should be abandoned.

3.8

Sources and further readings

Among the rapidly growing number of books and articles on language dormancy, several stand out. Nancy Dorian’s work has already been mentioned in this chapter and in Chapter 1. David Crystal’s 2000 book Language death and K. David Harrison’s 2007 book When languages die: the extinction of the world’s languages and the erosion of human knowledge, both already mentioned at the end of Chapter 1, are also important contributions. For good general introductions to this topic, see Lyle Campbell’s 1994 encyclopedia article ‘Language death’ and Suzanne Romaine’s 2010 article ‘Contact and

Sources and further readings

language death’. There are also books with regional coverage, such as Matthias Brenzinger’s edited volume Language death: factual and theoretical explorations with special reference to East Africa (1992). The quotation at the beginning of the chapter is from p. 39 of Nancy Dorian’s 1986 article ‘Abrupt transmission failure in obsolescing languages: how sudden the “tip” to the dominant language in communities and families?’; the quotation from the Book of Judges about the pronunciation of shibboleth is from the King James translation of the Bible. The Wikipedia article on shibboleth tells other stories too of pronunciations and other linguistic and cultural features used to distinguish friends from enemies in wartime. The massacre of 25,000 members of El Salvador’s Indian population and its linguistic effects are described by Campbell and Muntzel in their 1989 article ‘The structural consequences of language death’. The fate of Tamboran is described by Romaine in ‘Contact and language death’ (2010). The story of the death of Eyak comes from obituaries of Marie Smith Jones (especially Pemberton 2008) and Elizabeth Kolbert’s 2005 New Yorker article ‘Last Words’, on Smith Jones and on Michael Krauss’s work with Eyak and other endangered languages. Most of the historical information about Cornish comes from Mark Stoyle’s article ‘The Cornish: a neglected nation?’, accessed 12 April 2014 on the BBC-History website: www.bbc.co.uk/history/british/empire_seapower/cornish_ nation_01.shtml. The quotation about Cornishmen’s rejection of the English translation of the Book of Common Prayer is from this source, as is the quotation about the effect of this rejection, as hastening the decline of the Cornish language. See also Philip J. Payton’s 2000 and 2006 encyclopedia articles on Cornish. A photograph of Dolly Pentreath’s gravestone can be found at www.geocities.com/teammanley/Cornwall/DollyGrave.htm (accessed 10 October 2009). This website also mentions that her age is disputed, but affirms that ‘popular history recognises the death of Dolly Pentreath as representing the death of the language’. Her supposed last words are reported by Wikipedia, which refers to that story as a legend. The UK government’s press release about the Cornish people being granted minority status within the UK is at www.gov.uk/government/news/cornish-granted-minority-status-within-the-uk. The approximate dates of attestation of Sumerian and Chinese, respectively, are from p. 271 of Graham Cunningham’s 2006 encyclopedia article ‘Sumerian’ and p. 58 of Jerry Norman’s 1988 book Chinese. The information about early and late stages of the Egyptian language is taken from pp. 5–8 of Antonio Loprieno’s 1995 book Ancient Egyptian: a linguistic introduction, and the rest of the passage on the death of Egyptian is based on pp. 3–5 of Chris Reintges’s 2004 book Coptic Egyptian (Sahidic dialect): a learner’s grammar. The description of the death of Yaaku is from Matthias Brenzinger’s 1992 article on ‘Lexical retention in language shift’. Brenzinger in turn cites several other sources, among them (prominently) Bernd Heine’s 1975 article ‘Notes on the Yaaku language (Kenya)’; the quotation about Yaaku being ‘unfit for a

69

70

sliding into dormancy

cuttle-breeding society’ is from p. 33 of Heine’s article (cited by Brenzinger on p. 224 of his article). The report about efforts to revive Yaaku is by Matthijs Blonk, Maarten Mous, and Hans Stoks in an article ‘The last speakers – Yaaku language saved from extinction’: www.matthijsblonk.nl/paginas/YaakuENG.htm (accessed 18 December 2013). The warning that the causes of the shift from Yaaku to Maasai were almost certainly more complex than the description of the adults’ decision to shift might suggest is an inference from anthropologists’ detailed studies of causes of shift; see, for instance, Don Kulick’s 1992 book on language shift (cited in Chapter 2). Most of the information about Mednyj Aleut is based on my 1997 article ‘Mednyj Aleut’, which contains references to nineteenth-century and modern research on the language (most notably Menovšˇcikov 1969). Linguistic fieldwork on Mednyj Aleut was conducted by G. A. Menovšˇcikov before 1963 and by Evgenij V. Golovko and Nikolai B. Vakhtin in 1982, 1985, and 1987. Note that although Mednyj Aleut is (or was) spoken by a creole – mixed-blood – population, it is not what linguists classify as a creole language: the mixture of linguistic features in Mednyj Aleut differs sharply from the kinds of structural mixture that characterize creole languages, for instance those which (like the Caribbean creoles) arose centuries ago as a result of the Atlantic slave trade. Unlike Mednyj Aleut, Caribbean and other creole languages apparently did not arise among fully bilingual speakers. Mednyj Aleut was a bilingual mixed language, a category that is discussed further in Chapter 5. Dorian’s early discussions of language “tip” are from p. 51 of her 1981 book Language death: the life cycle of a Scottish Gaelic dialect and pp. 74–75 of her 1986 article ‘Abrupt transmission failure in obsolescing languages: how sudden the “tip” to the dominant language in communities and families?’. The definition of ‘tip’ given earlier, based on Dorian’s description of the process, is from p. 356 of Kathryn Woolard’s 1989 article ‘Language convergence and language death as social processes’. The report of tip in the Scottish Gaelic spoken on Cape Breton Island is from Elizabeth Mertz’s 1989 article ‘Sociolinguistic creativity: Cape Breton Gaelic’s linguistic “tip”’. A third example of tip that is said to parallel the East Sutherland Gaelic situation is described on p. 58 of Bernard C. Perley’s 2011 book Defying Maliseet language death: emergent vitalities of language, culture, and identity in eastern Canada. Nancy Dorian has discussed semi-speakers in a number of her writings. The passage reporting on this aspect of her research is from her 1977 article ‘The problem of the semi-speaker in language death’, and a general discussion of semi-speakers’ linguistic skill is on pp. 106–110 of her 1981 book Language death: the life cycle of a Scottish Gaelic dialect. The quotation about semispeakers’ sociolinguistic skill is from p. 108 of her 2010 book Investigating variation: the effects of social organization and social setting. Jane Hill and Kenneth Hill discuss the use of Hispanisms by fluent speakers of the endangered language Tlaxcalan Nahuatl in their 1977 article ‘Language death and relexification in Tlaxcalan Nahuatl’.

Sources and further readings

The quotation from Alan Dench about working with semi-speakers and other speakers with limited fluency is on p. 9 of his 1998 book Yingkarta. The information about the revival of Gamilaraay is from Peter Austin’s 2006 encyclopedia article ‘Gamilaraay’. For more on the loss of culturally specific lexicon, see the passage in Chapter 5 about K. David Harrison’s discussion in his 2007 book When languages die: the extinction of the world’s languages and the erosion of human knowledge. The Martuthunira example of words for ‘wild onion’ is taken from pp. 117–118 of Alan Dench’s 2001 article ‘Descent and diffusion: the complexity of the Pilbara situation’. The comment about an endangered language dying with its morphological boots on is from p. 608 of Nancy Dorian’s 1978 article on aspects of the death of a variety of East Sutherland Gaelic: ‘The fate of morphological complexity in language death’. This is a much-cited comment, as other scholars have also discovered that some dying languages, even languages whose dying process is gradual, undergo little or no attrition. Arguments and evidence in support of the claim that language change is unpredictable are found in my 2000 article ‘On the unpredictability of contact effects’. Lyle Campbell’s 1996 encyclopedia article ‘Defense strategies for endangered languages’ contains a very useful brief survey of language death and its effects on a language’s structure. The examples of attrition in Vilela are from Lucía Golluscio and Hebe González’s 2008 article ‘Contact, attrition and shift in two Chaco languages: the cases of Tapiete and Vilela’ (see pp. 222ff). Earlier Vilela materials are found in Elena Lozano’s 1970 collection Textos vilelas. The report on the status of Lule, the only known relative of Vilela, is from p. 194 of Lyle Campbell’s 1997 book American Indian languages: the historical linguistics of Native America. The examples from a variety of American Hungarian are from Anny Fenyvesi’s 1995 M.A. thesis, Language contact and language death in an immigrant language: the case of Hungarian. Campbell and Muntzel, in the 1989 article cited earlier in this section, discuss the Tuxtla Chico Mam example of a uvular-to-velar change, the Teotepeque Pipil example of a voiceless /l/ allophone replacing its voiced counterpart, and the Jumaytepeque Xinca example of glottalization run rampant. The merger of Warlpiri exclusive and inclusive ‘we’ pronominal forms is described by Edith Bavin on pp. 282–283 of her 1989 article ‘Some lexical and morphological changes in Warlpiri’, and the Dyirbal examples of regularization and simplification are discussed by Annette Schmidt in her 1985 book Young people’s Dyirbal: an example of language death from Australia (pp. 229–231). Bloomfield’s White Thunder quotation is from p. 154 of his 1927 article ‘Literate and illiterate speech’. The reduction in use of subordinate clauses is reported by Jane Hill in her 1978 article ‘Language death, language contact, and language evolution’ (as cited in Campbell 1996, p. 661).

71

72

sliding into dormancy

For comparison between changes in dying languages and changes in viable languages, see Nancy Dorian’s 1993 article ‘Internally and externally motivated change in language contact settings: doubts about dichotomy’ and, for contactinduced changes, p. 230 of my 2001 textbook Language contact: an introduction. The definition of contact-induced change is also from my 2001 textbook (p. 62). The examples of contact-induced changes in non-endangered languages are from the 1988 book Language contact, creolization, and genetic linguistics, by me and Terrence Kaufman, where original sources for the various examples are cited. The section on language death by massive borrowing is based on pp. 232– 235 of my 2001 textbook. The quotation about extreme borrowing by Laha from Ambonese Malay is from pp. 13–14 of James Collins’ 1980 article ‘Laha, a language of the Central Moluccas’, and the quotation about Votic turning into Ižora comes from Paul Ariste’s 1970 article ‘Die Wege des Aussterbens zweier finnisch-ugrischer Sprachen’ [‘The routes to death of two Finno-Ugric languages’]; the quoted sentences are my translation of his German sentences (‘Ishorische Wörter und grammatische Züge sind ins Wotische fast unmerklich eingedrungen, bis sie das Übergewicht bekamen. Dann war die Sprache dieser Woten nicht mehr Wotisch, sondern schon ishorisch’). The analysis of the developmental history of Ma’a is my own hypothesis, as explained on pp. 209–211 of my 2001 language contact textbook. But the hypothesis is controversial, as also explained on pp. 209–211; for the major alternative view, see Maarten Mous’s 2003 book The making of a mixed language: the case of Ma’a/Mbugu, and see also further references in my 2001 textbook. Hans-Jürgen Sasse’s book on Arvanitika Albanian is Arvanitika: die albanischen Sprachreste in Griechenland (1991). Joshua Fishman’s saying about the speech behavior of speakers of all languages is the title of his 1965 article ‘Who speaks what language to whom and when’. Campbell and Muntzel discuss types of language death in their 1989 article (cited earlier); Campbell also discusses these types in his 1996 encyclopedia article ‘Language death’. Anna Fenyvesi’s classifications of change types in dying languages is from her 1995 M.A. thesis (also cited previously). The term “language suicide” in the sense of replacement of all of a language’s native lexicon (and maybe also the structures) by borrowed material is discussed by April McMahon in her 1994 book Understanding language change, pp. 287– 291 (and the quotation is from p. 287). Following Jean Aitchison’s lead (in her 1981 book Language change: progress or decay?), McMahon uses as her prime example cases of decreolization, where, by hypothesis, a creole language borrows extensively from the language from which its lexicon is almost entirely drawn (pp. 287–288). The term “language suicide” in the more common sense of voluntary abandonment of a language by all its speakers is discussed by numerous authors, among them David Crystal on p. 86 of his 2000 book Language death. McMahon also discusses language murder – language death in Campbell’s “gradual death” category – on pp. 291–307 of her 1994 book.

4

What a community loses Language loss as cultural loss

Language is our unique relationship to the Creator, our attitudes, beliefs, values, and fundamental notions of what is truth. Our languages are the cornerstone of who we are as a People. Without our languages, our cultures cannot survive. (Towards Linguistic Justice for First Nations, Principles for Revitalization of First Nations Languages, Canada, 1990) When another nation comes and destroys your language they are sinning on more fronts than they can ever imagine. (Pat Ingoldsby, 2005) Many Arapaho see the loss of their language as a kind of spiritual test. Without it, the tribe’s ceremonies can’t be conducted correctly. “You lose the language, you lose the soul”. (Sergio Maldonaldo, director of Northern Arapaho tribal education, Wind River Reservation, Wyoming, 2009) When you lose a language, you lose a culture, intellectual wealth, a work of art. It’s like dropping a bomb on ... the Louvre. (Ken Hale, quoted in his obituary, 1 November 2001) [T]he ancestral language connects a people to its heritage in ways that there is simply no substitute for ... There is something inexpressibly sad about watching the disappearance of a unique local language that will never again be heard flowing in its full magnificence from the tongue of a verbally gifted speaker. I conversed with a number of speakers of East Sutherland Gaelic who had just such exceptional verbal gifts. ... The sadness lies in the realization that the great-grandchildren of those magnificent speakers will never have the chance to hear the like of what I heard. (Nancy Dorian, 1999)

This is the first of two “So what?” chapters, which attempt to answer the question of why endangered languages are worth saving. All over the world, endangered-language communities are coming to the realization that their cultures cannot fully survive if their languages die. This chapter considers the implications of language loss for a community’s sense of cultural identity and also – more concretely – for its links to the natural world, especially in the form of ethnozoological and ethnobotanical knowledge. The loss of a medium for verbal artistic expression that is deeply embedded in a community’s cultural world also comes into sharp focus here. 73

74

what a community loses

But not everyone agrees that the impending worldwide loss of linguistic diversity is a bad thing, and it is vital to avoid simplistic assumptions in this domain. Linguists tend to assume that our diversity-is-good perspective on the value of linguistic and cultural diversity is so obviously correct that it needn’t be justified. That’s fine if we’re preaching only to the choir, but if we hope to convince a wider audience, we need to examine our assumptions critically. This chapter will therefore include a discussion of, and responses to, the opposing position – namely, the view that linguistic diversity is not only unnecessary but is in fact dangerous for any society, and for human society as a whole. The chapter is organized as follows. In the first three sections we examine aspects of language endangerment and language loss that concern cultural cohesion and identity; a traditional community language as a medium for artistic expression; and the cultural impact of losing knowledge about the natural world. The final main section of the chapter addresses the opposing view – that linguistic homogeneity is preferable to linguistic diversity, for a variety of psychological, social, and political reasons.

4.1

Heritage languages and cultural identity

Let’s begin with a quotation from the late great South African leader Nelson Mandela: If you talk to a man in a language he understands, that goes to his head. If you talk to him in his own language, that goes to his heart.

Mandela’s comment was apparently intended as advice about negotiation and related sorts of interactions; that is, it was not about cultural identity. But its message nevertheless relates to questions of identity: for many, perhaps most (though not all) people, their ethnic identity is part of who they are. When they speak a language other than their own, their link to it is likely to be intellectual, not emotional. And the difference touches people’s sense of themselves. A young Blackfoot man, speaking about his native language, echoes Mandela’s comment: I want to pray to my ancestors in my own language. It has to come from the heart. The English version is not from the heart.

The deep emotional tie to one’s own language and the culture it expresses might be easier to understand through a thought experiment. Imagine that you are obliged – by force, or economic necessity, or social pressure – to abandon your native language entirely and speak another language instead, for the rest of your life. Such a situation implies cultural disruption to accompany the language shift, because language is a part of culture and language shift is almost certain to involve a significant degree of culture shift as well. So there you are, with a new

Heritage languages and cultural identity

language and partly or wholly new cultural patterns to go with it. Being uprooted is at best an unsettling experience, and many of us have experienced it, at least temporarily. During the year I spent as a student in Germany, speaking only German for months at a time and cut off from American culture, I felt comfortable enough in my new milieu, and pleased with my increasing fluency in German. But I also felt as if I weren’t quite the same person as when I was speaking English and living at home in the United States. The feeling wasn’t unpleasant – it seemed almost like acting in a long-running theater production – but I did feel separated from my fundamental nature. (I once described it as having a slight personality transplant.) That feeling, greatly intensified and much less pleasant, must be what people experience when they are permanently alienated from the culture and the language they grew up with. It is also akin to the point Nelson Mandela was making in the quotation that opened this section: millions of people all over the world do business in languages other than their own (often English, in the modern world), but they still feel most at home, and most comfortable, when speaking their own language. My experience with my German-speaking self was far from unique. Michèle Koven says that bilinguals ‘often report that they feel like a different person in their two languages’, and her observation is illustrated in a comment made by one of her French/Portuguese bilingual consultants: ‘When I speak Portuguese, automatically, I’m in a different world ... it’s a different color.’ Reactions like these are trivial when compared to what so many people suffer when they lose their language and culture entirely. In my case, I regained my own identity as soon as I returned home, and the bilinguals Koven studied were keeping both of their languages active. What hurts is having one’s fundamental cultural and linguistic identity taken away forever. The last lines of a poem by the Irish poet Pat Ingoldsby were quoted at the beginning of this chapter; the whole poem gives a poignant glimpse of the deep sadness and bitterness of language loss: Micilín Buachaill Beag Cróga It was the single saddest thing that I have heard in my life. Sad, heartbreaking and desolate. A little boy called Micilín was in hospital. Sixty or seventy years ago. The life was going out of him with tetanus. His brother, a good age now, was remembering this terrible thing. He was speaking on the television. He was recounting the awful way that it happened. The little boy had got a raging thirst. It comes with the tetanus. The little boy had got a raging thirst. He was asking for a drink. Over and over.

75

76

what a community loses

“Deoch!! Deoch!!” Nobody could understand him. Little boy dying. Asking in his native language for a drink. Nobody could understand him. With the sound of the word they thought he was asking for his duck. That was how he died. Amongst strangers. Frightened and confused. Asking for a drink. When another nation comes and destroys your language they are sinning on more fronts than they can ever imagine.

The connection between language loss and culture loss is obvious. As Anthony Woodbury has argued in an article about language systems (linguistic codes) and culture, [C]odes are really not interchangeable: individual codes, and the ways they are practiced in individual communities, are linked, indirectly or directly, to essential cultural content. Language preservation is therefore a crucial part of the maintenance of cultural diversity.

This view of language and culture as inextricably intertwined is widely shared, both by those who wish to promote linguistic diversity and by those who would suppress it. An especially striking example in the latter category is found in old Japanese colonial policies in Korea. After annexing Korea in 1910, the Japanese wanted, according to David Halberstam in his book about the Korean War, to obliterate Korean culture, starting with the language. The official language of Korea was proclaimed to be Japanese; in schools, lessons were to be taught in Japanese. The Japanese language text book was called The MotherTongue Reader. Koreans were to take Japanese names. The Korean language was to become a regional dialect, nothing more.

Now, the goal of destroying Korean language and culture might seem to have been doomed from the outset. Not only does Korean have many millions of speakers, but it also has a venerable tradition as a written language, with more than a thousand years of documentation in Chinese characters before the native Korean alphabet, Hangul, was invented in 1446. It is hardly an endangered language. But then, the Japanese occupation lasted only thirty-five years, until the defeat of Japan in 1945 at the end of World War II. If Japanese colonial rule had

Heritage languages and cultural identity

lasted as long in Korea as the colonial rule of some European nations in Africa and elsewhere, perhaps their harsh anti-Korean language policies would have succeeded: any language in the world can become endangered, no matter how many speakers it has, given the right (or wrong) combination of sociopolitical and socioeconomic circumstances. Halberstam goes on to emphasize the fact that Koreans vigorously resisted total linguistic and cultural assimilation to Japanese. But it is a sad fact that not all cultures can mount an effective resistance against a conqueror’s assimilation policies. The Irish have not (at least not yet) been able to do so; neither have Native Americans in the United States. Almost twenty years ago, after listening to a Montana Salish elder’s stories about her boarding school experiences – when she and other children were beaten for speaking their own language instead of English – I asked whether she had brought up any of her own five children to speak Salish. “No,” she replied. “I didn’t want my kids to go through what I went through.” The U.S. government’s harsh language suppression policies have long since been abandoned, but they were still in place when her children were in school. I asked her if she would make a different choice if she had it to do all over again, knowing that her language would be on the point of vanishing by the time her children reached middle age. “Yes, of course,” she answered. “But it’s too late now.” That is the sadness evoked in Pat Ingoldsby’s poem: this elder cared passionately about her heritage language and her heritage culture, but she had lost the one – she had almost no one left to speak to in her own language – and she had also lost most of the other, because the culture could not thrive without the language. Unlike the child in Ingoldsby’s poem, the Salish elder had native fluency in English (which she learned in the boarding school as a young child) as well as in her first language, so she had no difficulty communicating with everyone around her. Indeed, English had become her dominant language. But she could no longer live fully in her native culture, because it had been replaced by mainstream American culture in her commmunity. Her boarding-school experience was far from unique. A Pomo woman recalled the beginning of her time in a California boarding school this way: I was eleven years old, and every night I cried and then I’d lay awake and think and think and think. I’d think to myself, ‘If I ever get married and have children I’ll never teach my children the language or all the Indian things I know. I’ll never teach them that, I don’t want my children to be treated like they treated me.’ That’s the way I raised my children.

Another perspective on the close ties between language and cultural identity is provided by some of the ways in which kinship systems can break down in an endangered language. In Salish-Pend d’Oreille, for example, the complex inherited kinship system, along with the terms that express the various kin categories, has largely been replaced by the much simpler English kinship system. Current tribal elders who speak their heritage language fluently, and some of

77

78

what a community loses

their younger family members too, still remember and use (even when speaking English) a few of the most important terms. They still use both q’éneP ‘father’s mother’ and yayáP ‘mother’s mother’, for instance (there is no general term for ‘grandmother’). In church, they find it amusing that women have trouble reciting the Lord’s Prayer: the standard beginning that they all use in praying is qeP lPéw ‘our Father’, but in all other contexts lPéw would be said only by men and boys – it means ‘father of a male’. A woman’s father is m’estm; as with the grandmother words, there is no general term for ‘father’. But the prayers, including this one, were translated and prescribed by the missionaries and their successors, so women have to pray in words that they could not possibly use in everyday speech. (There is also no general term for ‘mother’, but since the only mother who appears often in the prayers is Mary, no difficulty arises: as the mother of Jesus, she is of course skw ’uy ‘mother of a male’, not tum’ ‘mother of a female’.) Many of the other kin terms have been forgotten by the last fluent speakers of the language, although the elders do remember some of them when their memory is jogged. One elder mentioned her smamáP, her father’s brother (‘mother’s brother’ is, or was, nunúmeP). She said she was shocked when, after her father died, she had to refer to his brother as ëwéstn, which means ‘aunt or uncle after the death of the connecting relative’. The old kinship system had other terms too for relatives after the death of the connecting relative. And they had reciprocal kin terms: smamáP in fact means both ‘paternal uncle (father’s brother)’ and ‘a man’s brother’s child’. Similarly, another elder said that he called his greatgrandmother t’ot’ó and she called him her t’otó (‘great-grandchild’). The lack of general terms for kin categories that are expressed by a single term in English is not confined to ‘uncle’; there is also no general term for ‘brother’—in the inherited system one must distinguish between qews ‘older brother’ and sínceP ‘younger brother’. The examples in these two paragraphs represent all the different parameters of kin categorization, although there are numerous additional kin terms that vary along these parameters. Like other complicated systems of kinship terminology, this one reflects, or reflected, the cultural cohesion of the speech community: kinship ties were among the most important, if not the most important, aspects of cultural identity. But in endangered speech communities all over the world such complexities in kinship terminology are being lost, and losing the terms promotes loosening of the cultural ties and, ultimately, loss of the culture. The inevitable linkage between language and culture is accepted as fact by a great many communities whose heritage languages are in danger of disappearing. We saw several examples earlier in this book, among them the Arapaho and Blackfoot comments in this chapter. We also saw it in the Bininj view expressed on a sign in Kakadu National Park, Australia: ‘Language is fundamental to cultural identity’, the sign begins, and it continues with an eloquent expression of what their language means to the Bininj people (see Chapter 2 for the entire quotation). Another example: in Chapter 3 we read the “Lament for Eyak”

Language loss as loss of artistic expression

recorded by Michael Krauss in 1972, thirty-six years before the language died (but perhaps only temporarily) with its last speaker. It is all too easy to multiply these examples. In the next two sections, communities’ sense of loss is reported mainly at second hand – through a linguist’s observations. This is a reflection of the fact that most of what linguists (and other people as well) know about the languages and cultures of endangered-language communities comes to us in linguists’ and anthropologists’ scholarly writings.

4.2

Language loss as loss of artistic expression

One way to understand what a speech community loses when its language falls silent is to focus on verbal art, an especially salient medium of cultural expression. Not all of the ways in which people engage in artistic and cultural activities are directly connected to language, of course: visual arts (including things like body painting and painting symbols on one’s horse), dance, and even wordless song may be connected only indirectly, if at all, to the artist’s language. But some artistic forms are inextricably tied to the language they are expressed in. For endangered languages especially, this is likely to be most evident in oral performance, both storytelling and nonnarrative poetry. To be sure, stories and poems can be translated into whatever encroaching language is dominant in an endangered-language speech community; notoriously, however, some aspects of verbal art cannot be translated in such a way as to preserve their artistic and cultural impact. As the late Ken Hale observed (1992:36), Some forms of verbal art – verse, song, or chant – depend crucially on morphological and phonological, even syntactic, properties of the language in which it is formed. In such cases the art could not exist without the language, quite literally. Even where the dependency is not so organic as this, an intellectual tradition may be so thoroughly a part of a people’s linguistic ethnography as to be, in effect, inseparable from the language.

One widespread and generally untranslatable feature of many poetic traditions is phonological or morphological distortion for expressive (emotionally loaded) purposes. Dell Hymes’s famous article ‘How to talk like a bear in Takelma’, for instance, explores the use of the prefix ë- (a lateral fricative) in Grizzly Bear Woman’s speech to indicate ‘diminutive meanings that have to do with condescension, perhaps, but also deprecation, disdain for coarseness and stupidity, and ... distance between actors’. A different expressive sense is connected with the insertion, in Navajo verbal art, of a phoneme /x/ (a velar fricative) to signal uncontrolled (often augmentative, indicating large size) actions or states. So, for instance, -tsaaz means ‘grow big’, while -tsxaaz means ‘grow very large’; compare also -si ‘make numb’ vs. -sxi ‘paralyze, deaden’ and -chah ‘cry’ vs. -chxah ‘scream’. To say that such features are untranslatable does not mean that

79

80

what a community loses

their referential sense can’t be given in translation. The English glosses of these Navajo /x/ forms – ‘grow very large’, ‘paralyze, deaden’, and ‘scream’ – convey the referential meaning without difficulty. But the insertion of /x/, yielding words that differ in just one sound from the non-affective (emotionally neutral) base form, conveys much more than the merely referential meaning. It contrasts the base form and the derived affective (emotionally charged) form directly, and it serves as a poetic device as well. In the short poem analyzed by the authors who describe Navajo /x/ insertion, for example, the /x/ appears in each line, contributing to the emotional force of the poem. Language endangerment puts affective features such as these at risk, because they are so closely tied to the language itself. Affective Takelma speech can no longer be studied with native speakers: the last fluent speaker of the language, Frances Johnson, died in 1934. Navajo may still have more than 100,000 speakers, but, as we saw in Chapter 1, only a small and dwindling percentage of children now learn Navajo as a first language, so it must be classified as endangered. If Navajo shares the fate of Takelma, its unusual expressive features too will vanish. Affective phonological and morphological markers do not, of course, exhaust the inventory of untranslatable artistic devices that become endangered when the artists’ languages become endangered. Another type of example is the use of morphosyntactic means of tracking characters in a story. Such devices are found in a number of languages, many of them endangered. In the Salish-Pend d’Oreille language of Montana, for instance, different characters can be tracked by means of two different third-person verbal suffixes, -es and -em. Verb forms that differ only in having one or the other of these two suffixes have exactly the same referential meaning (third-person subject ‘he’, ‘she’, ‘it’, or ‘they’) and exactly the same morphosyntactic properties in sentences. Typical examples ˇ ˇ would be the sentences cuPntés ëu Susép t Coní and cuPntém ëu Susép t Coní, both meaning ‘Johnny hit Joseph’ (literally, ‘he.hit.him Joseph Johnny’). Their meanings differ only in context: the verbal suffix -és indicates that the verb’s subject, Johnny, is the primary actor in the story, or at least in that part of the story; the -ém suffix, by contrast, indicates that the subject is of less importance in this context than the verb’s object, Joseph. In a free translation, the sentence with -ém might be rendered as ‘Joseph got hit by Johnny’, with an English passive construction. Here’s an example from a real narrative. In a two-story sequence told by Pete Beaverhead, one of the Pend d’Oreille tribe’s last master storytellers, the two main characters are Qeyqeyši and his sidekick One-Night. Qeyqeyši is clearly the more prominent member of the pair. In the first story he is also the main actor; when he is the subject of a transitive verb, the verb form always ends in -és. In the second story, however, One-Night is the main actor – he is the instigator of the two friends’ prank. Nevertheless, because Qeyqeyši is the main man in general, he is treated as primary in the second story too: when he is subject, the verb ends in -és. Throughout this second story, the more numerous

Language loss as loss of artistic expression

transitive verbs of which One-Night is the subject end in the suffix -ém. Many Pend d’Oreille verbs have only pronominal subject and object indicators, with the subject and object nouns understood; but it’s always possible, in stories like these, to tell which character is subject and which is object, in context, because of the two different third-person suffixes. The grammatical distinction, in other words, acts as a narrative device, part of the storyteller’s art. Comparable devices are found in other endangered languages. The best-known and perhaps the most complex examples are the forms that mark relative prominence of third-person actors in Algonquian languages. In Meskwaki, for instance, the most prominent third-person actor in a stretch of narrative – which, depending on the text, may extend over an entire story or over shorter segments, sometimes as short as a single sentence – is marked as proximate, while all other third-person actors receive obviative (less prominent) marking. An analysis of eight different versions of a single fragment of one story, the story of Lodge-Boy and Thrown-Away, illustrates the ways in which storytellers can exploit the flexibility in the grammatical marking to shape the same series of events (in this case, “a violent, dramatic series of incidents”), thus creating very different effects: ‘The narrators chose their proximates almost entirely by the characters’ importance, which resulted in a dry, detached narration; or almost entirely by the characters’ affect, which resulted in a vivid, engaged, appalled narration; or, most interestingly of all ... by importance punctuated at dramatic crises by a brief switch to affect, which greatly emphasized the dramatic peaks of the story.’ As with the affective examples discussed above, this feature of Meskwaki grammar is essentially untranslatable, and it has ‘a far-reaching effect on the poetics and rhetoric of the language’. So far the examples presented here have emphasized specific kinds of linguistic features in storytelling and poetry. In many or most instances, however, the totality of the artist’s performance is part of the artistic culture, and in order to appreciate that culture one needs a thorough understanding of the whole performance. One set of examples is found in Yokiri in southern Peru, where overlapping performance patterns are found in both of the community’s indigenous languages, Matsigenka (an Arawakan language of the upper Amazon) and Quechua, and also in the variety of Spanish spoken in Yokiri. Members of the community recognize the various aspects of a “good performance” – poetic features that include several kinds of sound symbolism (most of which are confined to the narratives), specific opening and closing formulas in myth performances, creative elaborations that may extend a performance to several hours, and a dialogic performance milieu in which other participants besides the storyteller enrich the story with questions and comments. One of the sound-symbolic devices is especially striking (and is also used by Quechua speakers elsewhere in southern Peru): to express “great intensity, quantity, size, or duration” of an action or thing, the storyteller lengthens the articulation of a voiceless stop (such as t or k) at the beginning of a syllable, and this long stop is “often accompanied by an increase in pitch, a squinting of the eyes, and a tilt of

81

82

what a community loses

the head”. All these performance features combine to make the event an elegant expression of Yokiri oral culture. But the traditional narrative performance is dying out in Yokiri in both Matsigenka and Quechua: only the oldest people still tell the traditional stories in the traditional way. And just as there are few remaining traditional storytellers, their audience is also nontraditional, so that the dialogic component of the traditional performances has vanished almost entirely. Some younger community members know the old stories, of course, and they even tell them sometimes; but the elders who know the stories better deplore the younger people’s strippeddown “just the facts” versions, which lack the creative elaborations and other features of the traditional versions. The languages themselves are endangered, especially Matsigenka. Quechuan languages, with millions of speakers, seem at first glance to be robust; but all the Quechuan languages are losing ground steadily to Spanish, so not even Quechua can be considered safe. It is clear, however, that the oral performance tradition will vanish long before either language disappears. Salish-Pend d’Oreille (as we saw in Chapters 2 and 3) and Meskwaki are gravely endangered, with perhaps 20 and 400 fluent speakers, respectively. The artistic devices in these languages that permit narrators to highlight or background different characters at different points of a narrative are at risk along with the languages. The same is true of other features of endangered languages that are manipulated by poets and storytellers for artistic effect, as in the performance tradition of Yokiri. If the languages are lost, these creative treasures will inevitably be lost as well, because they will not carry over into the dominant language. This is the kind of cultural tragedy lamented in the quotations at the beginning of this chapter.

4.3

Loss of a language, loss of cultural knowledge

The easiest way to understand how language loss leads to loss of cultural knowledge is to examine elaborated sets of culture-specific terms in different lexical domains. Earlier in this chapter we saw how the complicated Salish-Pend d’Oreille system of kin terms has been eroded; current elders remember only fragments of the old system, and even fewer terms are still in common use. Other examples reflect culturally prominent semantic domains. The most famous example by far – the only one that nonlinguists are likely to have heard of – is the semantic domain ‘snow/ice’ in Eskimoan languages. These languages have often been claimed to have dozens or even hundreds of different words for snow and ice, but the claims have been debunked, on the grounds that there are only a few different Eskimo roots (word bases, to which suffixes are added to form words) having to do with snow and ice and all the other words are derivatives based on those roots. An English analogue would be phrases like

Loss of a language, loss of cultural knowledge

‘slushy snow’, ‘powdery snow’, ‘crusty snow’, ‘hard-packed snow’, and the like; but Eskimoan languages have such complex morphological systems that each of these concepts would be expressed as a single complex word rather than as a phrase. Objections to the widespread claim that Eskimoan languages have lots of words for snow miss an important point about the lexical elaboration, however. Even if there are only a few separate word roots in this domain, the fact that there are so many words derived from those roots reflects the cultural significance of snow and ice in the speakers’ environment – and the ways in which speakers of Eskimoan languages have learned to adapt to their harsh physical environment. Among English speakers, by contrast, only skiers (and perhaps rural non– Native Alaskans) are apt to use more than a few different words and phrases for snow and ice. To be sure, snow and ice are part of many English speakers’ environment, but they are not a vital part of most English speakers’ economy. In sharp contrast, the Dogon languages of Mali do not have lots of words for snow, just as one would expect given the subtropical-to-arid climate in their neighborhood. They do, however, have up to thirty words for different grasshopper species – although any given Dogon speaker is likely to know only about twenty of the species names. Grasshoppers have various kinds of cultural significance in Mali and at least some other African countries: some species cause serious damage to crops, some species are prized as snacks (and as a valuable source of protein), and so forth. It’s probably safe to asssume that grasshoppers don’t play a major role in Eskimoan culture, any more than they do in Eskimoan vocabulary: dictionaries of Eskimoan languages give just one word for ‘grasshopper’ in each of the languages. The point is that lexical elaboration in different semantic domains provides insights into the cultural and environmental importance of different features of the natural world. Another example comes from Northern Saami, an endangered Uralic language whose speakers’ economic life focuses on reindeer. Not surprisingly, the language features a great many words having to do with a reindeer economy. Terminological differences depend on age and sex; body size, shape, and condition; color and nature of the animal’s coat (shaggy, short-haired, etc.); head (according to identifying marks); antlers; and feet (color). Some sets of terms are derived from the same roots, but numerous others have different roots. In the body size, shape, and condition category, for instance, we find (among other terms) gissor ‘small draft reindeer’, leamši ‘short, fat female reindeer’, njoalppas ‘reindeer with sloping hind quarters’, r˙avnnot ‘draft or pack reindeer which remains in good condition for a long time’, r˙avža ‘miserable, emaciated reindeer without a proper coat’, roašku ‘big thin reindeer’, silan ‘lean feeble reindeer that soon tires’, skoaldu ‘reindeer with a big head and a long nose’, spoairu ‘longlegged, thin reindeer’, and livat ‘draft reindeer that has worked so hard that it cannot be used for long journeys’. None of these examples, of course, has anything to do with a strong version of the linguistic relativity hypothesis (sometimes called the Sapir-Whorf

83

84

what a community loses

hypothesis), which holds that the language one speaks influences or even determines the way one thinks. All of them, and the innumerable other examples out there, do illustrate the tight connection between the speech community’s environment and the language’s vocabulary. Obviously, speakers of endangered languages who shift to a dominant group’s language could invent or borrow words to fill the resulting lexical gaps. Eskimoan and Dogon speakers would probably do so as long as their physical and cultural environments maintain the importance of snow and grasshoppers. The same is true of Northern Saami speakers; but loss of their language may well be accompanied by abandonment of their reindeer economy, and analogous cultural losses occur in many endangered-language communities. If that happens, the elaborate lexicon tied to that economy will also be lost, along with the cultural values and experience expressed by all those reindeer-related words. (This is especially easy to imagine when we remember the importance of the horse, up to about a hundred years ago, in the economies of the Western world. As noted in Chapter 3, for instance, most current English speakers do not know all the terms for parts of a horse’s harness and other aspects of a culture that once relied on horses for transportation of people and goods.) Ethnobotany is a particularly vulnerable area where traditional knowledge is lost when the community’s language vanishes. The names of medicinal plants accompany knowledge of their uses, and in many cases the names themselves have cultural power – the power to heal. Translated names are often considered to lack that power; names revealed to those outside the community are sometimes also believed to cause the medicines to lose their ability to heal. In such cases only transmission of the language from healer to healer can preserve the efficacy of the medicines, and the healers may be the only people who are able to use them effectively. Plants used for nonmedicinal traditional purposes may be less vulnerable in these respects, but they too depend on maintenance of the cultural tradition along with the words that designate the plants. In a prefatory note about the ethnobotany of the Thompson (Salish) First Nation of southern British Columbia, Nancy Turner and her colleagues reported that ‘[a]t least 350 species of native plants were recognized and named by the Thompson Indian people’ and traditionally used as ‘foods, medicines or materials’. But, they continued, ‘[o]nly a few native plant species are actively used by the Thompson people today ... With few exceptions, only members of the oldest generation are still using traditional medicines.’ In a recent article, Suzanne Romaine lists several communities in which knowledge of plants, their names, and their uses has declined or disappeared as a result of deforestation and other physical and cultural changes: some young people in eastern Indonesia have not learned either the names or the various traditional uses of native trees; in some Mexican highland communities, knowledge of medicinal plants has been lost, so that community members must depend on visiting health workers for ailments they once treated with their own medicines; Barí speakers in Venezuela are reported to be losing 40–60% of

Would the world be better off without linguistic diversity?

their traditional plant names from one generation to the next; similarly rapid rates of decline are reported for names of plants used in herbal medicines among Gaelic speakers in Ireland and Scotland, along with vocabulary for fastdisappearing traditional activities such as fishing and farming. Examples like these can easily be multiplied from endangered-language communities all over the world. The implications of this scale of loss are profound – primarily for the speech communities themselves, but also (as we will see in Chapter 5) for the world’s stock of knowledge. The loss of a community’s language brings with it loss of the cultural center. Even if, as has happened in a few cases, a vanished language is reawakened and revitalized, the cultural knowledge expressed by that language cannot be reconstituted if no one is left who remembers. The revived language can certainly serve as the carrier of a new culture, but it cannot revive the old culture. It is not difficult to imagine the grief of a speech community that sees its language and culture swept away by assimilatory pressures. But feelings of distress and a sense of loss are not universal. In the final main section of this chapter, we will explore an opposing view of linguistic and cultural diversity.

4.4

Would the world be better off without linguistic diversity?

There is a widespread view, held by many members of minoritylanguage communities as well as by many or (more probably) most speakers of major world languages, that the world would be a better place, a more peaceful and harmonious and prosperous place, if we all spoke the same language – or at least if we all spoke one of the handful of major world languages. On this view, if everyone spoke the same language, we could all communicate easily, and language-caused misunderstandings would disappear. With linguistic and cultural homogeneity as the norm, war would also disappear. Enormous amounts of money would be saved in translation costs, for both governments and business enterprises all over the world. Educational systems would no longer need to include either foreign-language teaching for majority-language students or transitional dominant-language instruction for minority-language students. On the individual level, a great many parents in minority communities today are eager for their children to have access to their country’s dominant language, as a tool for achieving economic and social success. Often these parents see their heritage language as a hindrance rather than as a valuable cultural resource, at least as far as their children’s education is concerned. And surely the desirability of preserving global linguistic diversity is not taken into account in minoritylanguage parents’ decisions about whether to pass on their heritage language to their children.

85

86

what a community loses

In this section we examine the most prominent beliefs and arguments in favor of global linguistic homogeneity. First, let’s consider the proposal that linguistic uniformity would eliminate war. This is not a new suggestion; one pre-modern example was articulated in a nineteenth-century U.S. document, the Report to the President by the Indian Peace Commission, January 7, 1868. Addressed to the President of the United States, the report begins as follows: The undersigned, commissioners appointed under the act of Congress approved July 20, 1867, “to establish peace with certain hostile Indian tribes,” were authorized by said act to call together the chiefs and headmen of such bands of Indians as were then waging war, for the purpose of ascertaining their reasons for hostility, and, if thought advisable, to make treaties with them, having in view the following objects, viz: 1st. To remove, if possible, the causes of war. 2d. To secure, as far as practicable, our frontier settlements and the safe building of our railroads looking to the Pacific; and 3d. To suggest or inaugurate some plan for the civilization of the Indians.

Among the commissioners’ recommendations for ending Native Americans’ violent resistance to the takeover of their lands and the destruction of their traditional livelihood is this passage, quoted at length here in order to convey the flavor of the language-related reasoning (the added emphasis in one phrase is mine): The wave of our population has been from the east to the west. The Indian was found on the Atlantic seaboard, and thence to the Rocky mountains lived numerous distinct tribes, each speaking a language as incomprehensible to the other as was our language to any of them. ... The white and Indian must mingle together and jointly occupy the country, or one of them must abandon it. If they could have lived together, the Indian by this contact would soon have become civilized and war would have been impossible. ... What prevented their living together? First. The antipathy of race. Second. The difference of customs and manners arising from their tribal or clannish organizations. Third. The difference in language, which in a great measure barred intercourse and a proper understanding each of the other’s motives and intentions. Now, by educating the children of these tribes in the English language these differences would have disappeared and civilization would have followed at once. Nothing then would have been left but the antipathy of race, and that too is always softened in the beams of a higher civilization. ... But suppose, when civilized, our pride had still rejected his association, we could at least have removed the causes of war by giving him a home to himself, where he might, with his own race, have cultivated the arts of peace. Through sameness of language is produced sameness of sentiment and thought, customs and habits are molded and assimilated in the same way, and thus in process of time the differences producing trouble wold [sic] have been gradually obliterated. By civilizing one tribe others wold [sic] have followed. Indians of different tribes associate with each other on terms of equality; they have not the Bible, but their religion, which we call

Would the world be better off without linguistic diversity?

superstition, teaches them that the Great Spirit made us all. In the difference of language to-day lies two-thirds of our troubles. But one thing then remains to be done with honor to the nation, and that is to select a district or districts of country, as indicated by Congress, on which all the tribes east of the Rocky mountains may be gathered ... agriculture and manufactures should be introduced among them as rapidly as possible; schools should be established which children should be required to attend; their barbarous dialects should be blotted out and the English language substituted. ... The object of greatest solicitude should be to break down the prejudices of tribe among the Indians; to blot out the boundary lines which divide them into distinct nations, and fuse them into one homogeneous mass. Uniformity of language will do this – nothing else will.

Fast-forward to the modern United States. Nowadays no one in the Anglo mainstream feels threatened by Native American languages or cultures, much less by any Native American potential for violent uprisings; indeed, one branch of the English Only movement welcomed into their group proponents of saving indigenous U.S. languages, on the grounds that Native American languages are, well, native to the U.S. and therefore have a right to a permanent existence here. Instead, the perceived threat to U.S. unity today is seen by English Only advocates to be the Spanish spoken by immigrants. In 2004, for instance, Samuel Huntington, the influential cofounder of Foreign Policy magazine and a prominent Harvard professor, published an article entitled “The Hispanic Challenge” in the magazine. The article, which expresses fears of cultural bifurcation as a result of Hispanic (primarily Mexican) immigration, ends with these sentences: Sosa ends his book, The Americano Dream, with encouragement for aspiring Hispanic entrepreneurs. “The Americano dream?” he asks. “It exists, it is realistic, and it is there for all of us to share.” Sosa is wrong. There is no Americano dream. There is only the American dream created by an AngloProtestant society. Mexican Americans will share in that dream and in that society only if they dream in English.

In other words, the idea that bilingualism is a threat to national (and, by extension, world) peace persists, and of course this view is not limited to the United States. The 1868 report’s assertion that ‘[t]hrough sameness of language is produced sameness of sentiment and thought’ is widely shared. But a very few moments’ consideration shows that this assumption is simply false. As noted in Chapter 2, the U.S. Civil War was fought primarily by armies whose dominant language was English on both sides. The same is true of the Revolutionary War. Hindu speakers of Hindi and Muslim speakers of Urdu speak dialects of the same Hindi-Urdu language and do not always live in harmony. The language formerly known as Serbo-Croatian has been split into four separate official languages – Serbian, Bosnian, Montenegrin, and Croatian – as one result of a bitter civil war that ultimately led to the dismantling of the former Yugoslavia into several different countries; efforts are under way to differentiate

87

88

what a community loses

the four official languages so that they will be linguistically as well as politically distinct. In Chapter 3 we saw the Biblical battle between the Gileadites and the Ephraimites, speakers of different dialects of Hebrew. Other examples of same-language wars are all too easy to find. So the argument that one language guarantees political harmony collapses under the weight of all the evidence to the contrary. The argument about the expense of multilingual societies is valid, as far as it goes: it is certainly very costly to run a country with several languages, where official documents, schoolbooks, and other materials need to be translated into different languages and educational systems need to be multilingual. There are obvious fixes for this problem, among them the position advocated by the English Plus movement in the United States: in response to the English Only movement, the English Plus movement was launched to promote an environment in which minority-language speakers would learn English while still maintaining their heritage languages. That is, the English Only monolingual movement was countered by the English Plus bilingual movement. The argument about eliminating the need for foreign-language instruction in a country’s educational system does not fare well. There is ample evidence that being bilingual has beneficial cognitive effects that go beyond language, so teaching schoolchildren other languages is good educational practice. It is true that money would be saved if all the foreign-language teachers were fired, but decisions about what to teach in schools are not taken entirely – ideally, not primarily – on the basis of cost. Nevertheless, some minority-language communities see assimilation to a dominant culture and its language as a positive good. This position was expressed eloquently by Peter Ladefoged, speaking of some countries in which he had conducted fieldwork projects. The Toda people of the Nilgiri Hills in southern India, he reports, ‘want to honor their ancestors, but also to be part of a modern India ... the cost of doing this is giving up the use of their language in their daily life’. Another of his examples is Dahalo, a “rapidly dying” Cushitic language of Kenya: a Dahalo consultant, explaining that his teenage sons spoke only Swahili and not Dahalo, was proud of the fact rather than distressed about his sons’ lack of Dahalo skills. These reactions are not rare, and they are also not surprising. As Ladefoged urges, linguists and anthropologists working with minority communities are most certainly obligated to respect the community’s views on such matters. Ladefoged’s arguments were answered, in part, by Nancy Dorian, who emphasized that ‘[i]f there is any one thing that speakers of seriously threatened languages have in common, it is likely to be low status within the region where they live’. In order to achieve a better socioeconomic position, they are often willing to give up their heritage language; but, Dorian points out, their children or grandchildren often bemoan the fact that they had no opportunity to learn that language. If their elders have been conditioned by dominant-language prestige, economics, and politics to abandon their language, the younger speakers look back and see the value of the lost language and culture.

Sources and further readings

The exchange between Ladefoged and Dorian is instructive for a variety of reasons (some of which won’t be touched on here). One lesson to be drawn from it is this: from a community perspective, there is no right or wrong answer to the question of whether an endangered language should be preserved, revitalized, and brought back into daily use. Some speech communities feel the loss of their language keenly; others don’t. Or, more accurately, members of some speech communities may have different views of the issue at different times and at different ages. The people Ladefoged spoke to were unconcerned about the loss of their traditional language and culture; they were focusing on the economic advantages of shifting to the dominant language and culture. But they, or their children or grandchildren, might feel differently later on. Years ago, while meeting with Salish-Pend d’Oreille elders in the tribes’ Culture Center in weekly summer sessions, I noticed that some of the teenage boys who were hired during the summer to do odd jobs around the Culture Center would slip into the back of the room to listen to us working on the language. It struck me that these were teenagers going out of their way, during work breaks, to attend to what to them would have been essentially a language class. I tried and failed to imagine Anglo teens spending their free time in a language class. I thought then, and I still think, that these young people were motivated by an eagerness to learn about an aspect of their tribes’ traditional culture that they had been unable to learn at home. This is one of many indications that minoritylanguage community members who have mastered the dominant language often regret the absence of their heritage language in their lives, and it is therefore evidence that community perspectives can also change. In fact, the dramatic recent increase in community-led language revitalization efforts around the world indicates a growing sense in minority-language communities that their linguistic heritage is worth maintaining. As I said in Chapter 1, this book reflects my own belief that the loss of any language, not to mention the worldwide loss of hundreds or even thousands of languages, is a very sad thing. Ladefoged was correct in saying that a minoritylanguage community has the right to give up its heritage language in exchange for socioeconomic benefits, or for any other reason; no linguist or anthropologist, no matter how passionate s/he is about an endangered language, can or should attempt to force a community to try to save their endangered language. But even if a community willingly lets its language go, the loss to the heritage culture is profound. The loss to science, as we will see in the next chapter, is similarly profound.

4.5

Sources and further readings

The quotation from the 1990 document Towards Linguistic Justice for First Nations is from Patricia A. Shaw’s 2004 article ‘Negotiating against loss: responsibility, reciprocity, and respect in endangered language research’.

89

90

what a community loses

The quotations from Sergio Maldonaldo and the young Blackfoot man are from a 2009 article by Emily Underwood, ‘The lost art of listening’, which focuses on efforts to save the Northern Arapaho language of Wyoming: www.hcn.org/ issues/41.20/the-lost-art-of-listening/article_view?b_start:int=0&-C= (accessed 26 May 2010). The quotation from Ken Hale, comparing language loss to the cultural effect of dropping a bomb on the Louvre, is from his obituary (he died on 8 October 2001), published in The Economist on 1 November 2001: www.economist.com/ node/842137 (accessed 26 May 2014). I came across this quotation in the 2014 article ‘Native tongue title: compensation for the loss of Aboriginal languages’, by Ghil’ad Zuckermann et al. The quotation from Nancy Dorian is from p. 39 of her 1999 article ‘Linguistic and ethnographic fieldwork’. Googling the Mandela quotation gets 518,000 hits (as of June 2014), and I only checked a fraction of them; but none of the ones I checked gave a date or a precise source for the quotation. Only 81,000 of the quotations include the word own; the rest lack that one word. The comment about the personality transplant is from my article ‘At a loss for words’ (Natural History magazine, 2007/2008). Michèle Koven’s quotation is from the abstract of her book Selves in two languages: bilinguals’ verbal enactments of identity in French and Portuguese, a study of some adult bilingual children of Portuguese immigrants in France. Pat Ingoldsby’s poem about the boy dying of tetanus is from his 2005 book i’mouthere (p. 18) and is reprinted here with the author’s permission. The Woodbury quotation is from his article ‘In defense of the proposition, “When a language dies, a culture dies” ’, and the Halberstam quotation is from p. 72 of his posthumously published 2007 book The coldest winter. According to the Ethnologue, there are 66 million speakers of Korean worldwide; Wikipedia’s article ‘Korean language’ gives a figure of 78 million. Either way, Korean has a lot of speakers. The Montana Salish elder who told me about raising her children to speak only English was Harriet (Alyé) Whitworth (1918–2008); this story is told in my 2007–2008 Natural History article ‘At a loss for words’. The same article is the source of the paragraphs about the loss of some Salish-Pend d’Oreille kinship terms. The Pomo woman who recalled her dismal days in the California boarding school was Elsie Allen; her story appears on pp. 175–177 of Leanne Hinton’s 1994 book Flutes of fire, in chapter 17, ‘Languages under attack’. Expressive affixes in Takelma are discussed by Dell Hymes in his 1979 article ‘How to talk like a bear in Takelma’, p. 105. The death of Mrs. Frances Johnson, together with her language, is reported by Marianne Mithun on p. 514 of her 1999 book The languages of Native North America. The Navajo expressive element /x/ is analyzed by Blackhorse Mitchell and Anthony K. Webster in their 2011 article “‘We don’t know what we become”: Navajo ethnopoetics and an expressive feature in a poem by Rex Lee Jim’.

Sources and further readings

The examples cited, which appear on p. 269 of their article, were drawn from Gladys Reichard’s Navaho grammar (1951), pp. 141–142. The description of variable proximate and obviative usage for artistic effect in Meskwaki (also known as Fox) is from Lucy Thomason’s 2003 Ph.D. dissertation, Proximate and obviative contrast in Meskwaki; the first quotation is from p. 162, the second and longest quotation is from p. 316, and the third quotation is from the dissertation abstract. (See also Lucy Thomason’s 1995 article on the same general topic.) Traditional Matsigenka and Quechua oral performance and its decline in Yokiri are discussed in Chapter 6 of Nicholas Emlen’s 2014 Ph.D. dissertation, Language and coffee in a trilingual Matsigenka-Quechua-Spanish frontier community on the Andean-Amazonian borderland of southern Peru, from which the discussion and quotations in this chapter are taken. A good source for exploring the topic of language loss linked to cultural loss is the 2001 book On biocultural diversity: linking language, knowledge, and the environment, edited by Luisa Maffi. The two main debunkers of the hundreds-of-Eskimoan-words-for-snow claims are Laura Martin (in a 1986 article, ‘Eskimo words for snow: a case study in the genesis and decay of an anthropological example’) and Geoffrey Pullum in The great Eskimo vocabulary hoax (1991). Dogon words for grasshopper are reported by Jeffrey Heath in a personal communication to Steven Moran and his coauthors in a 2012 article. Heath elaborated on the number of grasshopper species names that any given person is likely to know in a personal communication to me (18 November 2013), commenting that ‘[t]he maximum for individual speakers/villages is 20 ... Toro Tegu (not endangered but small) is perhaps the richest’. My statement that an Eskimoan language has only one word for ‘grasshopper’ is based on consultation of two dictionaries, Donald H. Webster and Wilfried Zibell’s Iñupiat Eskimo dictionary (1970) and Michael Fortescue et al.’s Comparative Eskimo dictionary (1994). The information about Saami words for reindeer is taken from Ole Henrik Magga’s 2006 article ‘Diversity in Saami terminology for reindeer, snow, and ice’; the examples are from p. 27 of the article. The Sapir-Whorf hypothesis of linguistic relativity, first introduced by Benjamin Lee Whorf and Edward Sapir, has been hotly debated over the past sixty years. The strongest version of the hypothesis would be that your language controls your thought (so that you can’t think about things that are not encoded in the language you speak). A weaker and more plausible version is that you will find it easier to think about things that your language encodes (so that I, as a non-skier and non-entomologist, tend to just talk about snow as a single kind of substance and grasshoppers as a single kind of insect). But these are kindergarten-level sketches of the hypothesis; interested readers should consult the literature on the subject. One modern scholar who has devoted much attention to the topic is John Lucy; see for instance his 1992 book Language diversity and

91

92

what a community loses

thought: a reformulation of the linguistic relativity hypothesis, his 1996 book Grammatical categories and cognition: a case study of the linguistic relativity hypothesis, or his 1997 article ‘Linguistic relativity’. See also Chapter 5 for further discussion of linguistic relativity. Here’s one example of a cultural belief that medicinal plants will lose their power if their names and medicinal properties are revealed to outsiders: the Salish and Pend d’Oreille elders that I work with in Montana report that publishing information about their traditional medicines would ensure that the medicines will no longer work. The quotations about Thompson Salish plant names and usage are from an unnumbered page at the front of the 1990 book Thompson ethnobotany: knowledge and usage of plants by the Thompson Indians of British Columbia, by Nancy J. Turner, Laurence C. Thompson, M. Terry Thompson, and Annie Z. York. Suzanne Romaine’s list of communities that are losing, or have lost, much ethnobotanical vocabulary is from p. 327 of her 2010 article ‘Contact and language death’. For the Barí example she cites Manuel Lizarralde 2001, and for the Gaelic example she cites Seosamh Watson (1989:52-53). The material from the Report to the President by the Indian Peace Commission is from a transcription by Carolyn Sims of Furman University’s Department of History, of the Annual Report of the Commissioner of Indian Affairs for the Year 1868 (Washington, DC, Government Printing Office, 1868, pp. 26–50). The first quotation from the report is on pp. 1–3 of the publication; the second quotation is from pp. 81–84 and 87. See http://alpha.furman.edu/~benson/docs/peace.htm (accessed 15 June 2014). (To find the report on the website, scroll down to the category “Post-Civil War Documents”.) Emily Underwood’s 2009 article about efforts to save the Northern Arapaho language, ‘The lost art of listening’, brought this Indian Peace Commission report to my attention. The quotation from ‘The Hispanic challenge’ (Foreign Policy, 2004, p. 45) is from the late Samuel P. Huntington’s last book, Who are we? The challenges to America’s national identity, published later in 2004. The full title of the book by Lionel Sosa that is cited in the quotation is The Americano dream: how Latinos can achieve success in business and in life (1998). For most of the past century there has been a lively debate about whether growing up bilingual is harmful or beneficial to a child’s cognitive development. In a 1983 review of the literature, Rafael Diaz shows that, up until about 1960, most of the evidence pointed to negative effects of bilingualism; but he also notes that serious methodological flaws (for instance in not controlling for the child subjects’ actual knowledge of their two languages) made the relevant studies suspect. After 1960, with improved methodologies, experimental studies regularly showed an opposite effect: ‘when compared to monolinguals, balanced bilingual children show a wide range of advantages in different cognitive tasks’ (Diaz, p. 34), both verbal and nonverbal. More recent research, for instance Ellen Bialystok’s 2009 article ‘Effects of bilingualism on cognitive and linguistic performance across the lifespan’, presents a more nuanced picture,

Sources and further readings

with bilinguals surpassing monolinguals in certain vitally important nonverbal cognitive tasks throughout their lives but falling short in some verbal tasks, for instance lexical retrieval (choosing the right word from one’s stored mental vocabulary). Peter Ladefoged’s 1992 article ‘Another view of endangered languages’ is a response to the earlier 1992 article by Ken Hale et al., ‘Endangered languages’. The quotations are from pp. 810 and 811, respectively, of Ladefoged’s article. In 1993, Nancy Dorian published ‘A response to Ladefoged’s other view of endangered languages’ in the same journal; the quotation from her article is on p. 576.

93

5

What science loses Language loss as a threat to our understanding of human history, human cognition, and the natural world

I am not very willing that any language should be totally extinguished. The similitude and derivation of languages afford the most indubitable proof of the traduction of nations, and the genealogy of mankind. They add often physical certainty to historical evidence; and often supply the only evidence of ancient migrations, and of the revolutions of ages which left no written monuments behind them. (Samuel Johnson, 1766) A knowledge of their several languages would be the most certain evidence of their derivation which could be produced. In fact, it is the best proof of the affinity of nations which ever can be referred to. ... It is to be lamented, then, very much to be lamented, that we have suffered so many of the Indian tribes already to extinguish, without our having previously collected and deposited in the records of literature, the general rudiments at least of the languages they spoke. (Thomas Jefferson, 1782) If English were the only language, we could learn a lot about the fundamental principles of grammar, but we could only guess at the nature of that which can vary, except to the extent that this is evident from the varieties of English itself. (Ken Hale, 1992:35) Baagi Letsapa, 54 ... is a leathery man who still remembers the teachings of his father. He knows that the boiled roots of the sage plant help ease indigestion, that papyrus makes the best sleeping mats and that the ukayi plant produces a poison that kills fish, but not people. This knowledge is disappearing as young people abandon the rivers, the canoes and the old ways. (Rachel L. Swarns, about a man of the Wayeyi people of Botswana, 2001)

This is the second of two “So what?” chapters – as in, “So what if languages die? Who cares?” Chapter 4 described the inevitable diminishment of a community’s cultural resources, and even its very identity, when its language is lost; in this chapter we focus on the loss to science, both in the quest to understand human history and human cognition and in valuable practical knowledge of the natural world.

5.1

Endangered languages and human history

Well before the emergence of historical linguistics, which is now a highly successful historical science, Samuel Johnson and Thomas Jefferson understood that human languages in all their diversity can provide evidence for historical connections among human populations, as well as evidence for ancient population movements. Johnson, at least, apparently did not value linguistic 94

Endangered languages and human history

diversity for its own sake. After the sentences quoted on page 94 he goes on to say, I would not unwillingly compound, by wishing the continuance of every language, however narrow in its extent, or however incommodious for common purposes, till it is reposited in some version of a known book, that it may be always hereafter examined and compared with other languages, and then permitting its disuse.

In other words, he doesn’t object to letting a language disappear as long as it is documented first. (Jefferson offers no opinion on whether the long-term survival of Native languages of North America would be a good or a bad thing, but he seems to expect that they will continue to vanish.) Johnson’s point about historical clues being found in unwritten languages has a surprisingly modern flavor. Beginning near the end of the eighteenth century (with a few earlier precursors), the field of historical linguistics – the study of language change – grew slowly during the first half of the nineteenth century and rapidly during the second half. By the last decades of that century, most of the standard methodologies now used for investigating language change and language diversification were in place, including the Comparative Method, which permits the reconstruction of unattested (undocumented) ancestral words and structures through systematic comparison of the vocabularies, sound systems, word structures, and sentence structures of a group of related languages. Both Johnson and Jefferson were right to believe that documenting languages before they vanish will make it possible to unravel some of their speakers’ history. One North American example is the establishment of the Algonquian language family. All the members of this family are now at least vulnerable and most are gravely endangered, except for the ones that are already temporarily or permanently extinct. The outlines of this far-flung family, with member languages from the Great Plains in the U.S. and Canada eastward to the Atlantic coast, have been known for well over a century. Indeed, the Algonquian languages that are or were spoken east of the Mississippi River were first grouped into the same family in the late eighteenth century, in a 1787 report to the Connecticut Society of Arts and Sciences by Jonathan Edwards, Jr. A son of the distinguished philosopher Jonathan Edwards the elder, Edwards grew up speaking the Algonquian language Mohegan (which he spelled Muhhekaneew) and became familiar with numerous other Algonquian languages later on. He did not fall into the trap of assuming that all Native American languages belong to the same family – that is, that all were descended from the same ancestral language; he was aware of the sharp distinction between Algonquian and Iroquoian languages, the two language families whose speakers once occupied most of the northeastern U.S. and neighboring parts of Canada. Like Samuel Johnson’s comments, Edwards’s treatise has a rather modern flavor. He compiled and compared lists of words, phrases, and grammatical features to arrive at his conclusion about the historical

95

96

what science loses

connections among these languages, just as modern historical linguists do in efforts to discover the same kinds of historical relationships. Because Algonquian languages were the first Native American languages that European settlers came into close contact with, there are missionary accounts of several of the languages, including word lists, from as early as the seventeenth and eighteenth centuries. For a related reason – too many settlers, too much conflict between the newcomers and the Native peoples – several of the languages disappeared quite early. Among these were the New England languages Etchemin (gone by the mid-seventeenth century), Loup (also gone during the colonial period), Nanticoke and Quiripi (dead by the late eighteenth century), Narragansett (last used in the early nineteenth century), and Massachusett (last used at the end of the nineteenth century), as well as Powhatan (spoken in Virginia, died around the 1790s), Mohegan, and Mahican (the tribe made famous by James Fenimore Cooper’s historical novel The last of the Mohicans). Some of these languages have been the target of revival efforts in recent years – most notably Wampanoag, a variety of Massachusett, whose recent reawakening has been the subject of a prize-winning documentary film, ‘We Still Live Here – Âs Nutayumeân’. A few of the Algonquian languages that were spoken farther west have also vanished. The last fluent speaker of Atsina (also known as Gros Ventre) died in 1981, the last fluent speaker of Miami-Illinois in 1989 (but see Chapter 7 for discussion of tribal members’ vigorous revival efforts), and the last speaker of Eastern Abenaki in 1993. Of the surviving Algonquian languages, all appear in the UNESCO Atlas of the world’s languages in danger, with different levels of endangerment: vulnerable (Micmac, Northwestern Ojibwe, and all the varieties of Cree); definitely endangered (Central Ojibwe, Cheyenne, and Blackfoot); severely endangered (Maliseet-Passamaquoddy, Shawnee, Eastern Ojibwe, and Arapaho); or critically endangered (Western Abenaki, Munsee Delaware, Unami Delaware, Sauk-Fox [Meskwaki], Potawatomi, and Menomini). The crucial point is this: in spite of the large number of languages in the Algonquian family, different accidents of history could easily have left us ignorant of their number and their geographical spread, and perhaps even their historical connections. Without such early gatherers of wordlists and other materials on these languages – mainly curious missionaries and Thomas Jefferson, but others as well – we would know nothing at all about the languages that vanished early; for that matter, there may well have been still more Algonquian languages that disappeared without being documented at all. And if more of the Algonquian languages spoken in the Great Plains had disappeared before modern linguists were able to study them, it might even be thought that the Algonquian languages and their speakers had always been confined to the eastern part of North America. This point can be made even more strikingly if we include the so-called Ritwan languages Wiyot and Yurok in the discussion. During the late(ish) twentieth century, these two Californian languages were firmly established as distant

Endangered languages and human history

relatives of the Algonquian language family; the larger family is called Algic. This means that the geographical range of this large family stretches from coast to coast in North America. The puzzle of how and when the languages arrived at their known traditional locations remains, but it appears that the ancestral (ancient and unattested) Algonquian language, Proto-Algonquian, was spoken in the west, so that the tribes and languages spread eastward from somewhere west of the Mississippi River. The homeland of the larger Algic family is still unknown. But scholars are lucky to have any evidence at all about Algic: the last speaker of Wiyot died in 1962, and by the year 2000 Yurok had only about twenty elderly fluent speakers. If these two languages had died before they were documented, the Algonquian family’s California connection could not have been discovered. The Algonquian/Algic example illustrates both the ability of scholars to use systematic comparative linguistic evidence – comparing the vocabularies and structures of a group of related languages by means of the Comparative Method – to reveal the historical connections and movements of diverse peoples, and the effect that language endangerment can have on the success of such endeavors. Wiyot and Yurok could very easily have vanished without leaving behind sufficient documentation to permit the establishment of their distant historical connection with Algonquian. Of course we have no way of knowing how many languages have in fact disappeared without leaving a trace, but especially over the past century or two the number must be large. These disappearances have deprived scholars of the opportunity to understand important parts of human history, and we are all the poorer for that loss. Worse still, the detailed historical lineages of numerous languages – especially in New Guinea, Africa, South America, and parts of Asia, which are linguistically diverse – are still unknown; and because most of those languages are small and vulnerable, there is great danger that many of them will disappear without being documented. If that happens, the result will be permanent ignorance of sizable chunks of human history, including the spread and diversification of peoples as well as the ways in which people entered and occupied new landscapes. A final example in this section illustrates the kinds of detailed information that can be revealed by comparing related languages – or rather, in this case, dialects of the same language. In the second half of the nineteenth century, historical linguists formulated the regularity hypothesis of sound change, a well-supported hypothesis that underpins the Comparative Method. According to the regularity hypothesis, if a sound x in a particular word of language A changes into a sound y in language A", a changed later form of A, then every instance of x in the same phonetic environment in words of A will change into y in A". The power of this hypothesis is what makes it possible to find systematic sound correspondences to help establish that two languages are related – that is, that they are descendants of the same ancestral language – and to reconstruct significant portions of an undocumented ancestral language, a proto-language. The hypothesis is supported by enormous quantities of data from dozens of language families all over the world.

97

98

what science loses

Table 5.1. Regular High German (South) voiceless stop changes gloss

North

South

pound sleep ship to water that make I

pund slapen schipp to water dat maken ik

Pfund schlafen Schiff zu ([tsu]) Wasser das machen ich

But there are exceptions, and they are instructive. The most famous set of exceptions to the regularity of sound change is found in dialects of German near the Rhine River. This is the so-called Rhenish Fan, which was discovered inadvertently by a nineteenth-century German dialect specialist who was trying to prove that all sound changes were (in accordance with the regularity hypothesis) exceptionless. East of the Rhine area, starting about 40 kilometers east of the river, the boundary between southern German dialects (called High German dialects because of the southern mountains) and northern German dialects (called Low German dialects because of the relatively flat terrain) is sharp. In the South, there are exceptionless sound changes in a particular set of consonants, the results of the High German Consonant Shift, in which (among other changes) the voiceless stops /p t k/ turned into affricates (pf, ts) or fricatives ( f, s, x), depending on their position in the word. (The symbol x represents a velar fricative, spelled ch in German; and the German spelling sch represents a sound [š], which is the sound of sh in English ship.) North of the boundary there are unchanged voiceless stops like those in the North column of Table 5.1; south of the boundary we find changed forms like those in the South column of Table 5.1: Close to the Rhine, however, there is clear evidence of a gradual south-tonorth spread of the individual High German stop changes, with different results in different places. The boundaries between these differential results correspond to old political (national) borders that were abolished in 1789. The boundaries were presumably divisive enough that communication from one state to another was diminished; the fact that people were less likely to talk to each other across the boundaries would help account for the differential spread of the consonantal innovations. Table 5.2 (somewhat simplified) shows, with illustrative words, how the spreading changes faded away differentially in the general south-to-north spread. The first change to stop spreading was t to ts, s; the second change to stop was p to pf, f. The changes in k never did regularize completely in the Rhineland: some words changed and some didn’t. North of the ik/ich line, however, there is no sign of any of the changes.

ik ix

Treves Trever

B r u s s e l s en mak

Antwerp p dor dorf

en

Amsterdam Amilardam

FRISIAN

S NI

Coblens

Cologne

ldorf Dusse

dat das

A

E

ROMANCE

H

maken maxen

Figure 5.1. Map of the Rhenish Fan (from Bloomfield 1933:344)

M

C

Basle

Rhin e

mak

DA

River

99

R

O

N

S L A V I C

S L AV I C

Berlin

M A G YA R

S L A V I C

xen ma

maken

B A LT IC

100

what science loses

Table 5.2. Irregularities of the High German voiceless stop changes near the Rhine River, south to north. South:

North:

das, Schiff, machen, ich (all three stops changed) the das/dat line (no t’s changed north of this line) the Schiff/schipp line (no p’s changed north of this line) the machen/maken line (some k’s changed north of this line) the ich/ik line dat, schipp, maken, ik (none of the stops changed)

Various explanations have been offered for the lack of a neat dividing line between High German and Low German in the Rhineland, in addition to the effects of the national borders in that area. For the present context, the important point is that we have in the Rhenish Fan an excellent example, precisely because of the irregularities, of how a set of sound changes comprising many individual pieces might spread – not all at once as a complete package, but piecemeal and even, in the farthest reach of the spread, word by word (where k changed in some words but not in others). Now, Standard German is a High German dialect, so it of course displays all the regular components of the High German Consonant Shift: all three voiceless stops, p t k, changed in all words. Standard German is the official language of Germany and is used throughout the country as the language of education, and all nonstandard dialects are threatened with extinction, including Low German dialects as well as the dialects in the Rhenish Fan. If Standard German had replaced the nonstandard dialects of the Rhenish Fan before they were documented, we would not have this outstanding record of the gradual spread of a complex set of sound changes. We saw in Chapter 2 that loss of nonstandard dialects has effects very much like those resulting from the loss of languages; the example of the Rhenish Fan shows that the loss of dialect diversity also has consequences for efforts to understand language history and, through it, human history. As Leonard Bloomfield observed, the Rhenish Fan ‘shows that the spread of linguistic features depends upon social conditions. ... Important social boundaries will in time attract isogloss-lines’. (An isogloss is a boundary line indicating where a particular dialect feature is found.)

5.2

Endangered languages and human cognition

If a language dies without being documented, its disappearance robs science of a unique window into the human mind (to repeat the cliché used in Chapter 1), even aside from the loss of the kinds of specialized knowledge that so often disappear along with a language. Linguists believe that certain features are universal to human language, and therefore either innate – hard-wired in the human brain – or arising from universal human learning strategies; linguists also

Endangered languages and human cognition

believe (necessarily!) that all other features vary from language to language. The problem is to distinguish the universal from the nonuniversal features. As Ken Hale said in a quotation at the beginning of this chapter, linguistic diversity is necessary for our hope of understanding the range and limits of variation in human languages. In this section we will see how language endangerment has the potential to diminish linguists’ chances of understanding human language thoroughly, and I give examples to show what could be, and indeed has been, lost to science. In Chapter 4 we examined several examples that fall into the category of linguistic elements that we would know nothing of if the languages in which they occur had disappeared without being documented. The risk of loss is greater, of course, for features that are primarily drawn from less-thoroughly studied languages, because these are predominantly small minority languages that are at greatest risk of disappearing. Most of the world’s big languages have been well documented and are also not endangered: English, French, German, Spanish, Portuguese, Danish, Norwegian, Russian, Czech, Polish, Hungarian, Finnish, and many other European languages; Arabic, Hebrew, Turkish, and a few other Middle and Near Eastern languages; Hindi, Urdu, Marathi, Gujarati, Tamil, Kannad.a, and other major languages of the Indian subcontinent; Chinese, Japanese, Korean, Vietnamese, Thai, and other major languages of East and Southeast Asia; Indonesian, Malay, and a few other Oceanic languages; and, among the indigenous languages of the New World, Quechuan languages, Aymara, and a few Mayan languages. Outside this charmed circle of “safe” languages (safe for the time being, at least), there are several thousand languages that are all too likely to vanish within the next century and take irreplaceable linguistic riches into oblivion. Two of the examples we saw in Chapter 4 had to do with adding affective (emotionally charged) meanings to words: Grizzly Bear Woman’s expressive diminutive ë- prefix in Takelma and the Navajo inserted element -x- that signals uncontrolled action. Two other examples were features of word structure, with partly similar functions: the two third-person (‘he/she/it/they’) subject suffixes in Salish-Pend d’Oreille that distinguish between a more important actor and a less important actor in a narrative, and the obviative (“less prominent in the context”) vs. proximate (“more prominent in the context”) distinction in Meskwaki (and other Algonquian) verbs. Three of the examples illustrated lexical elaboration in culturally important semantic domains – snow and ice in Eskimoan languages, grasshoppers in Dogon languages, and reindeer terminology in Saami. (These vocabulary examples, admittedly, could survive even if all three languages fell into dormancy; but although the terms could be translated, the translations would almost certainly not be single complex words, and the linguistic cohesion of the sets of words would likely be lost.) Examples of these and other types could easily be multiplied many times over from a catalog of the world’s endangered languages. I’ll give just a few additional examples here. The click sounds of the Khoisan languages of southern Africa

101

102

what science loses

are so rare as to be virtually nonexistent elsewhere in the world, though they have diffused from Khoisan languages (a convenient grouping comprising all click languages, which fall into several distinct language families) into some neighboring Bantu languages. Few Khoisan languages have many speakers; most are gravely endangered. Other endangered languages too have sounds that are almost unknown elsewhere in the world; among them are Saami, which has three length distinctions in both consonants and vowels: short, e.g. /l/ or /a/; long, e.g. /l:/ or /a:/; and extra-long, e.g. /l::/ or /a::/. All six of these segments are separate phonemes in Saami, capable of distinguishing words from each other. The ritual language Damin, once spoken on Australia’s Mornington Island by initiated Aboriginal men of the Lardil community, boasted a lexical structure unlike the vocabulary of any other known language: its words are ‘abstract names for logically cohesive families of concepts’. The entire vocabulary consists of less than 200 words in all, thanks to this remarkable principle of lexical construction. So, for instance, the pronoun n!aa refers to any set of people that includes the speaker (either the speaker plus other(s), like English we, or the speaker alone, ‘I’), and the contrasting pronoun n!uu refers to any set that does not include the speaker (‘you’, ‘he’, ‘she’, ‘it’, ‘them’). Damin also has sounds unknown to its speakers’ native language, Lardil (including one or two sounds that may not be found in any other language anywhere); its morphology and syntax, however, are identical to Lardil morphosyntax. When Ken Hale studied the language in 1967, only a few speakers of Damin remained; now, almost half a century later, Lardil itself is critically endangered on the UNESCO scale of endangerment, and Damin is almost certainly gone. Salish-Pend d’Oreille, with just a handful of remaining speakers, has at least two other highly unusual structural features, from a global perspective, in addition to its morphological distinction between more and less important actors in a story. One is an optional truncation rule, a grammatical process that deletes everything after the stressed (accented) vowel in a word. The truncation rule could be expressed as something like this: “Delete everything after the stressed vowel of any word if you want to – but you won’t want to if there are crucial grammatical suffixes after the stressed vowel.” Nouns are truncated more often than verbs are, because crucial information is contributed more often by verb suffixes than by noun suffixes. An example of a truncated form is x.w eëé ‘she hurried’, shortened from x.w eëéˇcst; in contrast, this verb’s singular imperative form x.w eëéˇcstš! ‘hurry!’ cannot be truncated, because the imperative suffix -š is necessary to the meaning of the verb. A truncated noun like st’šá ‘huckleberry’ (literally ‘sweet berry’), however, is permanently shortened from its original long form st’šáëq: the remaining fluent speakers are unfamiliar with the old long form. A few of the closest relatives of Salish-Pend d’Oreille have a similar optional truncation process, but truncation happens more frequently in SalishPend d’Oreille than in its sister languages. All the languages that have this process are gravely endangered; they could all have vanished before linguists documented the rule and its effects. As far as I know, no other language in the

Endangered languages and human cognition

world has a grammatical process quite like this; and if that is true, then the loss of these languages before they could be documented would have left linguists ignorant of this grammatical possibility. The second Salish-Pend d’Oreille example is an unusual sound-symbolic formation that is used to express – symbolize – sounds or (more rarely) sights. Most roots in the language have the shape Consonant-Vowel-Consonant (CVC); the sound-symbolic formation repeats the second root consonant three times. Compare, for instance, the root liq ‘rip’ with the sound-symbolic word liqqqq ‘the sound of a lot of threads breaking (e.g., a shirt coming unraveled when you tear it)’, or the root cikw ‘shine’ with cikw kw kw kw ‘little shiny things sparkling, like sequins on a dress or sparks flying up’. Like the truncation rule, this soundsymbolic formation is (as far as I know) unknown elsewhere in the world, except in at least one of the language’s closest relatives. And as with the other examples in this section, a slight adjustment in the course of history could easily have caused this language to lose all its speakers before linguists discovered the existence of the feature. If that had happened, scholars would never have known about its existence. Other examples come from different areas of language structure. For example, an influential article by Joseph Greenberg, published fifty years ago, identified two logically possible patterns of basic word order as nonexistent, or nearly so, in the world’s languages. Most languages (like English) have either SubjectVerb-Object (SVO) basic word order in a sentence or SOV word order (like Turkish), and VSO word order is also fairly common (e.g., Classical Arabic and Irish Gaelic). But Greenberg’s survey found that OSV basic word order was almost nonexistent and OVS basic word order completely unknown. At the time this was true; but since the 1960s, a few languages have been found to have these rare word order patterns. The best-known of these (to linguists) is Hixkaryana, an indigenous Brazilian language with only a few hundred speakers, which has OVS as its basic word order pattern. This pattern is illustrated in the sentence toto yonoye kamura ‘The jaguar ate the man’, literally ‘man ate jaguar’. If Hixkaryana and the few other OVS languages had never been studied by linguists, we might all still be looking for deep universal reasons for the absence of OVS languages. Another large group of examples can be found in the literature on the linguistic relativity hypothesis, which was mentioned in Chapter 4 (with references in that chapter’s Sources and Further Readings section). The psychologist and anthropologist John Lucy, in particular, has found effects in his experimental studies that suggest a linkage between certain kinds of linguistic features and cognition. An especially intriguing example (not one of Lucy’s) is the metaphorical reversal, in the South American language Aymara, of time and space orientation, in contrast to what most Westerners would think of as the “natural” correlation: for English speakers and other Westerners (and a great many other people all over the world), the future lies in front of us, as the prepositional phrase in front of indicates; we say things like going forward into

103

104

what science loses

the future. And the past is behind us, as the preposition behind indicates: we say things like back in the old days. But in Aymara, ‘the basic word for FRONT (nayra “eye/front/sight”) is also a basic expression meaning PAST, and the basic word for BACK (qhipa “back/behind”) is a basic expression for FUTURE meaning.’ Rafael Núñez and Eve Sweetser explore the cognitive implications of this contrast, taking into consideration both the Aymara words and Aymara speakers’ gestures in developing a ‘taxonomy of spatial metaphorical mappings of time’. Now, Aymara is not at present endangered – it has more than two million speakers in Bolivia and other Andean countries and has official status in Bolivia – but, like all indigenous languages of the Americas, it is under threat because many of its speakers are shifting to a colonial language, in this case Spanish. The examples given in this section are a tiny sample of the crosslinguistically unusual features that linguists have discovered in endangered languages: a complete list, supposing it were possible to compile one, would be very long indeed. There are certainly many, many more features in many languages still waiting to be discovered; and a high percentage of those languages are almost certainly endangered, given the fact that thousands of languages are known to be vulnerable or endangered. Michael Cahill pointed out recently that formal theories of language cannot attempt to account for linguistic phenomena that are unknown to the theorist. Moreover, the phenomena of greatest interest might not be confined to the endangered languages in which they are first discovered: as Lucy Thomason says with respect to the use of the obviative/proximate distinction in Algonquian narratives, ‘an understanding of how speakers of Algonquian languages judge and manipulate prominence relations may prove crucial to understanding how discourse prominence and reference tracking work in languages where both these things are less explicit.’ In short, minority languages around the world are likely to contain structural and lexical surprises of great interest to scholars who try to understand the interactions of language and cognition. It is not only individual features that make endangered languages important for understanding human cognition. A final example in this section is an entire language, the mixed language Mednyj Aleut that we met in Chapter 3. This language is, or was, a bilingual mixed language – that is, a language created in a social context where both component languages were spoken fluently by some or all community members. Only five or six bilingual mixed languages are fairly well understood by linguists; of these, Mednyj Aleut has the most surprising composition. Some of the others have a split between vocabulary and grammar (for instance Media Lengua of Ecuador, a mixture of Spanish lexicon and Quechua grammar). Still others appear to have arisen via codeswitching in bilingual conversation, with, for instance, noun phrases from one language embedded within verb phrases and sentence structure from another (the best-known example is Michif, a mixture of French noun phrases combined with Cree verb phrases and sentence structure). But Mednyj Aleut is different, because it consists of Aleut structure – vocabulary, noun structure, non-finite

Endangered languages and knowledge of the natural world

verb structure, sentence structure – combined with an entire finite inflectional morphology imported intact from Russian. As noted in Chapter 3, there are numerous Russian loanwords and other borrowings throughout Mednyj Aleut, but not to an extraordinary degree; Mednyj Aleut is basically Aleut, except for the finite verb morphology. Because Mednyj Aleut is unique, with a type of mixture unknown in any other language in the world, combining pieces of two languages in this way would almost certainly have been considered impossible if the language had disappeared without being documented. And, of course, we have no idea at present how many other endangered languages can provide different but also unique glimpses of the human linguistic capacity. The lesson to be drawn from this and all the other examples in this section is that the urgency of documenting endangered languages is acute, from the perspective of linguistics and of cognitive science more generally.

5.3

Endangered languages and knowledge of the natural world

As K. David Harrison has observed, ‘For many endangered languages that have never been put down in writing, entire domains of knowledge are likely to be lost when the language ceases to be spoken.’ Some of the lost knowledge will be specific to a particular culture and perhaps of limited interest to outsiders; some of it will provide insights into human culture and, especially in the areas of ethnobotany and ethnozoology, human cognition; and some of it, especially medicinal and nutritional knowledge, may be of considerable potential benefit to all humans. Nor are these categories mutually exclusive. Some cultural knowledge has implications both for understanding human cognition and for healing or feeding people. An example is in the quotation at the beginning of this chapter from a New York Times article about the Wayeyi people of Botswana, which reports Baagi Letsapa as saying ‘that the boiled roots of the sage plant help ease indigestion, that papyrus makes the best sleeping mats and that the ukayi plant produces a poison that kills fish, but not people’ – all potentially useful information for people of any culture who might live in the area. We have already seen examples, in Chapter 4, of actual and potential losses in elaborated lexical domains. Some Eskimoan words for snow and ice may be important for others trying to live and work in the Arctic; knowing which grasshoppers are good to eat may be useful information to others besides Dogon speakers. But lexical elaboration is not just a matter of lists of individual terms: there are also cognitive lessons to be learned from complex lexical systems in a wide variety of languages all over the world. These lessons arise from the classification of animals, plants, and other natural phenomena into culturespecific categories. For example, folk taxonomies often highlight the functions of organisms rather than their biological nature. A rather extreme example is the

105

106

what science loses

classification of organisms used by the ||Gana people of Botswana (the language name begins in a click sound; ||Gana is a dialect of Khoe, a Khoisan language). They divide organisms into three categories: things that they eat (kx’ooxo), things that are harmful to them (paaxo), and things that don’t fall into either of the first two categories (goõwahaxo). The language has no general term for ‘animal’ or ‘plant’, only these very broad categories. Another organizing principle that one finds in folk taxonomy involves ecological linkages that tend to be invisible to outsiders. In the Australian Aboriginal language Kunwinjku, spoken in the Northern Territory, a particular fish species (the spangled grunter fish) and a tree species (the native white apple tree) have the same name because the fish eat the fruits that fall into the water from this tree. Similarly, the Aboriginal language Mparntwe Arrernte names several kinds of edible grubs after the bushes they are found in; examples are the witchetty grub (tnyematye), named after the witchetty bush (tnyeme), and the utnerrengatye grub, named after the emu bush (utnerrenge). In Chapter 4 we explored some of the cultural implications of losing vocabulary for plants and other domains of the natural world, along with knowledge of their uses. Here we’ll also examine these domains of cultural knowledge, but focusing this time on the implications for science, and humanity at large, of lost words and the knowledge that goes with them, especially in the areas of food and medicine. Examples of practical folk knowledge are given by Nicholas Evans in his book Dying words. One is from Seri, a language with no known relatives (an isolate) that is spoken in Baja California, Mexico. Fieldworkers investigating the language found that it has a whole vocabulary for eelgrass – its harvesting, its products – and that the plant has considerable nutritional value. It is said to be ‘the only known case of a grain from the sea being harvested as a human food source’, and in the future it may turn out to be a potential general food source for humans. It would not have been known if fieldworkers had not studied Seri intensively. Evans also discusses a drug, prostarin, which is used to treat HIVtype 1 and which was discovered as a result of a conversation between a Samoan tribal healer and an ethnobotanist ‘about traditional medicinal uses of the stem of a particular tree, Homalanthus nutans’. Other folk medicines and traditional foods used by speakers of endangered languages could potentially also provide valuable practical benefits beyond their original communities, but only if they become known before the last speakers die and take the knowledge with them. It is impossible to know how many traditional remedies might be efficacious enough to be the basis of medicines useful to outsiders as well as to indigenous communities; given the numbers of plants that are used as medicines, the useful medicines could be numerous. This is a difficult and controversial area, however: endangered-language communities are likely to be interested in the potential “valuable practical benefits” only if their culture permits sharing their knowledge with outsiders, and certainly only if the shared knowledge provides benefits to them rather than simply putting massive

Sources and further readings

amounts of money into the pockets of multinational pharmaceutical companies while leaving the indigenous community bereft. In this context, consider again the Thompson (Salish) people of British Columbia, whose ethnobotanical riches were mentioned in Chapter 4. The Thompson people use at least 200 plant species as herbal medicines and have names for all of them – or rather, the language had all the names more than twenty years ago. Thompson medicines can be grouped into a number of different categories, depending on their uses, and some plant species were used for more than one purpose. So, for instance, there were fifty-four “tonics and general medicines”; thirteen “purgatives, laxatives, and emetics”; forty-nine medicines for “colds, coughs, tuberculosis, influenza, and other respiratory elements”; ninety-six medicines used as “poultices, salves, or washes for wounds, infections, burns, sores”; twenty-seven medicines for “arthritis and/or rheumatism and/or muscular aches and pains”; ten medicines for “kidney and urinary ailments”; twenty-one medicines for “venereal diseases”; twenty-one medicines for “eyes”; fifty-seven medicines for “stomach and/or digestive tract”; twenty-four medicines for “women (especially at childbirth)”; and so forth. In 1990, when Nancy Turner and her colleagues published their book Thompson ethnobotany, a few Thompson elders still used a few of the old medicines. The people who knew the names and the uses from first-hand experience are probably all gone by now, taking their priceless knowledge with them. But Turner and her colleagues reported that efforts were being made back then to establish educational programs that would revive that knowledge, and the elders who knew the names and uses were actively teaching them to younger tribal members. It is possible, therefore, that some or even much of the knowledge still lives, even beyond the 1990 book of almost 300 pages. It is hardly likely, though, that many other endangered-language communities will have the opportunity to try to rescue their inherited knowledge of the natural world. For one thing, there are not so many dedicated ethnobotanists who can afford the time and money to compile an authoritative catalog of plant names and uses on the scale of the Thompson ethnobotany book; and the same is true for other ecological domains. This means that a large proportion of the endangered knowledge held by the last speakers of gravely endangered languages will be lost when the last speakers are gone. In the meantime, while there is still time for hundreds of languages, the most vigorous efforts should be made, whenever possible, to document these vital aspects of a community’s ecology thoroughly.

5.4

Sources and further readings

The 1766 quotations from Samuel Johnson are on p. 321 of a 1955 edition of Boswell’s Life of Samuel Johnson L.L.D. (1791). Johnson was writing to William Drummond, an Edinburgh bookseller, to express his dismay at the fact

107

108

what science loses

that a Scottish ‘society ... for propagating Christian knowledge’ was opposing a plan to translate the Bible ‘into the Erse or Gaelick language’, for political reasons having to do with wanting to assimilate Gaelic-speaking Highlanders to English-speaking Lowlanders (p. 321). (Note that traduction in the first quotation means ‘derivation from ancestry, descent’ (Oxford English dictionary); it has nothing to do with traducing.) I am grateful to Peter Grey for bringing this passage to my attention (personal communication, 1995). The Jefferson quotation is on p. 510 of his Notes on the state of Virginia (originally published in Paris in 1782). He followed up on this interest by having vocabulary lists collected from as many Indian tribes and nations as possible; unfortunately, most of the collection was destroyed when a thief, rummaging through the trunk in which it was being transported, threw the trunk’s contents into the James River (American Philosophical Society ‘Treasures of the APS’ Web page, www.amphilsoc.org/exhibits/treasures/vocab.htm). I am grateful to Thomas Trautmann for showing me this passage from Jefferson’s writings (personal communication, 2010). The Ken Hale quotation at the beginning of the chapter is from p. 35 of his 1992 article ‘Language endangerment and the human value of linguistic diversity’. Rachel L. Swarns’s report of Baagi Letsapa’s insight into what is being lost in the Wayeyi community of Botswana is in her 2001 New York Times article ‘A sleepy river’s people fight to keep old ways’. The information about eastern Algonquian languages grouped together as a family, by Jonathan Edwards, Jr., is from pp. 29–30 of Lyle Campbell’s 1997 book American Indian languages: the historical linguistics of Native America. Edwards’s report, whose lengthy title begins Observations on the language of the Muhhekaneew Indians...", was first published in 1787 and reprinted in 1788; a more accessible edition, with added notes, was published in 1823. Information about early missionary materials on Algonquian languages, as well as about the documentation and status (extinct, endangered, vulnerable) of the languages, can be found on pp. 327–337 of Marianne Mithun’s book The languages of Native North America (1999). On p. 337 Mithun discusses the two so-called Ritwan languages, Wiyot and Yurok, of California (including their present status) and cites the scholars who helped establish their relatedness to the Algonquian family. Grouping Wiyot and Yurok into the same language family as Algonquian was a controversial proposal for years after Edward Sapir first proposed it in his 1913 article ‘Wiyot and Yurok, Algonkin languages of California’. In 1958 Mary Haas took the next step by adducing more impressive evidence in her article ‘Algonkian-Ritwan: the end of a controversy’ – but that did not end the controversy, because some specialists found her evidence promising but inconclusive. Not until Karl Teeter (especially in ‘Algonkian languages and genetic relationship’, 1964) and Ives Goddard (especially in ‘Algonquian, Wiyot, and Yurok: proving a distant genetic relationship’, 1975) presented more extensive evidence did the controversy vanish, along with doubts about the historical connection between Algonquian, Wiyot, and Yurok. William

Sources and further readings

J. Poser has given a history of the controversy in his online paper ‘On the end of the Ritwan controversy’ (www.billposer.org/Papers/oerc.pdf, accessed 28 May 2014). The reason for the adjective “so-called” before the Ritwan label is that Wiyot and Yurok do not seem to have a special relationship to each other; that is, they do not seem to be more closely related to each other than either is to the Algonquian branch of the Algic family. The catalog of Algonquian languages and their degree of endangerment is taken from information on the UNESCO Atlas of the world’s languages in danger website: www.unesco.org/culture/en/endangeredlanguages/atlas. The Wampanoag language reclamation project website is at www.wlrp.org (accessed 28 May 2014). A major part of the story is the work of Jessie Little Doe Baird, the prime mover in the project; to aid in her efforts, Baird earned an M.A. in linguistics at MIT (studying with Ken Hale). The west-of-the-Mississippi location of the original Algonquian homeland is discussed by Lyle Campbell on pp. 152–154 of his 1997 book, citing as his authority for the most likely location Ives Goddard’s 1994 argument in ‘The West-to-East cline in Algonquian dialectology’. The information about the Rhenish Fan, and most of the illustrative data, comes from pp. 342–345 of Leonard Bloomfield’s 1933 book Language; the quotation is from p. 345. Damin, the ritual language of the Lardil people of Australia, is discussed on pp. 36ff. of Ken Hale’s 1992 article ‘Language endangerment and the human value of linguistic diversity’. For more information on Salish-Pend d’Oreille truncation, see Lucy Thomason and Sarah Thomason’s 2004 article ‘Truncation in Montana Salish’ (an alternate name for the language); the quoted rule is on pp. 355–356. See also my 2006 blog post on the subject, ‘Trunca in Monta Sa’: http://itre.cis.upenn. edu/~myl/languagelog/archives/003356.html The sound-symbolic example from Salish-Pend d’Oreille is from my 1999 article ‘Crackle, plop, twinkle: sound-symbolic words in Montana Salish’. Saami length distinctions and Hixkaryana word order are discussed by Naomi Palosaari and Lyle Campbell on pp. 105–106 of their 2011 article ‘Structural aspects of language endangerment’. For the information about Saami they cite a 2008 colloquium presentation by Ante Aikio, ‘The Saami languages: history, present situation, and work on documentation’; the Hixkaryana information is from Desmond Derbyshire’s 1979 book Hixkaryana. The article by Joseph Greenberg that is cited in this paragraph – a famous article that is often credited with founding the linguistic field of typology – is ‘Some universals of grammar with particular reference to the order of meaningful elements’, published in 1963 and again in 1966. See the list of references at the end of this book for examples of John Lucy’s research on linguistic relativity. The quotations about Aymara time/space orientation, where back = future and front = past, are from p. 402 of Rafael Núñez and Eve Sweetser’s 2006 article ‘With the future behind them: convergent

109

110

what science loses

evidence from Aymara language and gesture in the crosslinguistic comparison of spatial construals of time’. Michael Cahill’s remarks about the importance of intensive study of endangered languages are in a 2003 letter to the editor of the journal Language, published in the journal in 2004. The quotation from Lucy Thomason about obviative and proximate is from p. 3 of her 2003 dissertation Proximate and obviative contrast in Meskwaki. Several bilingual mixed languages are discussed in a 1997 book that I edited, Contact languages: A wider perspective. Among them are ‘Mednyj Aleut’, written by me, pp. 449–468 (see also the sources listed there); ‘Media Lengua’, by Pieter Muysken, pp. 365–426; and ‘Michif: a mixed language based on Cree and French’, by Peter Bakker and Robert A. Papen, pp. 294–363. Another important source on Michif is Bakker’s 1996 book “A language of our own”: the genesis of Michif – the mixed Cree-French language of the Canadian Métis. The quotation from K. David Harrison is from p. 23 of his 2007 book When languages die: the extinction of the world’s languages and the erosion of human knowledge. The example of ||Gana folk taxonomy of living things is from p. 39 of Harrison’s book. Nicholas Evans describes the naming of the spangled grunter fish in the Australian Aboriginal language Kunwinjku and the several kinds of edible grub in the Aboriginal language Mparntwe Arrernte on p. 22 of his 2010 book Dying words: endangered languages and what they have to tell us, and he discusses the Seri use of eelgrass and the HIV drug that owes its discovery ultimately to a Samoan healer on pp. 20 and 21, respectively. The quotations about Thompson Salish ethnobotany are from p. 43 of Nancy J. Turner et al.’s 1990 book Thompson ethnobotany: knowledge and usage of plants by the Thompson Indians of British Columbia. The comments about efforts to preserve Thompson ethnobotanical knowledge are on p. 296 of the same book.

6

Field research on endangered languages

To what extent are endangered languages a priority in modern linguistics? Are graduate students encouraged to document moribund or endangered languages for their dissertations? How much encouragement is there to compile a dictionary of one? How many academic departments encourage applied linguistics in communities for the support of endangered languages? How many departments provide appropriate training for speakers of these languages who are most ideally suited to do the most needed work? Obviously we must do some serious rethinking of our priorities, lest linguistics go down in history as the only science that presided obliviously over the disappearance of 90% of the very field to which it is dedicated. (Michael Krauss, 1992:10) If it wasn’t for his work, we wouldn’t have our language. He single-handed preserved our language ... You can’t put a price on something like that. (Jim Pepper Henry, a Kaw tribal member, quoted in The Wichita Eagle’s obituary of Robert Rankin, 2 March 2014) Johnnie Ray McCauley, who was Kaw, told The Eagle in 1996: “I just want to hear it again. There has been no one else to talk it with.” McCauley died in his home a few months after receiving the CDs. When he was found, he was wearing headphones and the CD recordings of his late aunt speaking the family’s native Kaw Indian language were still playing, according to his obituary in The Eagle. (From The Wichita Eagle’s obituary of Robert Rankin, who had recorded Mr. McCauley’s aunt’s stories along with other narratives on reel-to-reel tapes in the 1970s, digitized and converted the tapes to CDs in 1996, and presented the Kaw community with a set of the CDs that same year.)

Michael Krauss’s warning, and others issued around the same time, did not fall on deaf ears. Since 1992 there has been an explosion of research on endangered languages by graduate students and older scholars too, many new funding opportunities have arisen to support this research all over the world, much digital ink has been spilled on the topic in the popular press as well as in scholarly publications, and many endangered-language communities have initiated ambitious revitalization programs. The worldwide decline of endangered languages has not been stopped or even appreciably slowed. But the picture is nevertheless brighter than it was in 1992, both because a great many endangered-language communities have come to understand what losing their language could mean 111

112

field research on endangered languages

to them, and because the community of linguists has finally recognized the importance of documenting endangered languages and working with speech communities to preserve and revitalize them. This chapter concerns one of the responses, the increasing amount of field research on endangered languages; Chapter 7, the final chapter, examines the topic of revitalization. Linguists who investigate endangered languages in the field face challenges that are less likely to trouble someone who works on non-endangered languages, especially languages with thousands of speakers in all age groups. In this chapter I discuss some of those challenges for fieldworkers, focusing primarily on languages with only a few remaining fluent speakers. The reader I have in mind here, in particular, is one who is considering embarking on a first linguistic fieldwork venture, as a lone fieldworker rather than as a member of a team. Although, as we will see, the ideal modern fieldwork context involves teamwork, it is still the case – because of funding limitations, because of a lack of relevant professional connections, or because of other personal factors – that many or most beginners set out alone. The number of endangered languages that remain to be investigated is far greater than the number of linguists available to investigate them, so restricting fieldwork to well-organized interdisciplinary teams that have all the desirable kinds of expertise would (it seems to me) be extremely counterproductive, even if it were possible. It is important for neophytes to know about up-to-date methods and strategies for full-scale documentation and description of an endangered language; but it is not vital for them to have all the pieces of such a project in place at the outset. And in any case, the initial stages of field research are likely to be the same regardless of whether one is working alone or as part of a team: you cannot, for instance, study ethnographic aspects of language usage until you have a basic understanding of the language’s vocabulary and structure. The first section of this chapter contrasts descriptive linguistics – the main topic of the chapter – with the broader area of documentary linguistics, an emerging subfield of linguistics that owes its existence to the upsurge of scholarly interest in endangered languages over the past two decades. The next section is a brief overview of general issues in linguistic fieldwork, together with discussion of three specific topics that have special relevance to fieldwork on endangered languages. We will then turn to the narrower topic of fieldwork on gravely endangered languages and dialects, with subsections on different aspects of a fieldwork project.

6.1

Descriptive linguistics and documentary linguistics

Language documentation, or documentary linguistics, has been defined as ‘the creation, annotation, preservation and dissemination of transparent records of a language’ – with the emphasis on developing a large corpus

Descriptive linguistics and documentary linguistics

of material that is accessible to a variety of users, including, crucially, nonlinguist members of the language’s speech community as well as linguists. Creating the records essentially means collecting data (in the field or from existing field notes or other sources). Annotation involves marking up the records with analytic notes and with metadata indicating the date, place, speakers, genre, context, and other aspects of each record. Preservation has to do with depositing the data in existing digital archiving systems or developing new ones, and dissemination means making the annotated records available to the world, or (in the case of records that can’t, for whatever reason, be made public) to an appropriate group of users. It is highly unlikely that any single person will have all the skills that are required for a major documentation project; linguists and archivists, for instance, have very different kinds of professional expertise, and relatively few linguists are trained as ethnographers or as database designers. And these are just a few of the kinds of skill sets that are needed. The relationship between language documentation and language description is a matter of some debate. At one extreme are scholars who seem to believe that one can and should do documentation without description; at the other extreme are linguists who still believe that describing a language without paying any attention to the primary concerns of documentary linguistics is a perfectly reasonable goal. Probably the majority view nowadays, however, is that both enterprises are of crucial importance to a research project that includes collecting primary data on an endangered language. It is obviously possible to some extent to do documentation without description and vice versa. You can collect large quantities of diverse genres of texts – conversations, traditional tales, songs, ethnographic texts (e.g., how a man traditionally asked a woman to marry him, or how to bake camas over heated rocks in a pit), and the like – and perhaps also experimental data on a language’s prosodic features, and you can archive all of that without preparing detailed, publishable analyses of the data in the form of dictionaries, grammars, and scholarly articles. Or, on the other hand, you can collect data on a language and publish analyses of the material without making any attempt to put your primary data in an archive in a form that will permit someone other than you to find it and understand your system of data transcription, your analytic notes, and your organization of the material; all your data might, for instance, reside permanently in your computer or even in old-fashioned card files. Some analysis is necessary for any documentation project, of course: what makes the records transparent is annotation of the material. Annotating an audio file, for example, might mean time-aligning transcriptions and translations of words and phrases with chunks of the audio; annotating a transcribed text might include breaking words into their component parts (for instance root and affixes) and aligning the gloss (meaning) of each part under the word or word part that it translates. No annotation can be done without analyzing and describing the parts of the language represented in the annotations. Those who argue in favor of documentation instead of description focus instead on the importance of

113

114

field research on endangered languages

collecting and making available as much material as possible, without taking time off (as it were) to write up and publish a dictionary or a grammar or scholarly articles. Several years ago Nicholas Evans expressed a view of this argument that would, I believe, find widespread agreement among linguists who conduct fieldwork on endangered languages: I think it is a mistake for documentarist linguists to argue that they should consecrate all their time and effort to pure documentary activities at the expense of preparing descriptive grammars or other reference materials. A much more apt strategy is Colette Grinevald’s vision (Craig 2001) of an eternal spiralling upwards through the elements of the classic Boasian trilogy – grammar, texts (now = documentary corpus), and dictionary – with each step forward producing advances and refinements in how the other steps proceed.

In support of this view, Evans stresses both the scientific aspects, notably the fact that (as his reference to Grinevald’s comment suggests) the analysis of data collected in the field proceeds incrementally, changing and improving each time it is set down in writing, and the human aspects, primarily the great value of published products such as stories and dictionaries in the eyes of members of the speech community. Language documentation is a highly technical topic – or rather set of topics – and this is not the place to enter into a detailed discussion of its methods and theories. Although it arose in response to the perceived need to record as much of endangered languages as possible as fast as possible, language documentation is in principle quite general, not specific to endangered languages. Its intricacies can and should be understood by anyone planning to start fieldwork on an endangered language, even if the beginner cannot satisfy the requisites of full-scale documentation, at least not at first. An excellent place to start is with the 2006 book Essentials of language documentation, edited by Jost Gippert, Nikolaus Himmelmann, and Ulrike Mosel. (The Evans quotation above is from his review of this book.) Some chapters in the book are devoted to issues that regularly arise in fieldwork projects, for instance fieldwork ethics and the incorporation of ethnography; other chapters concern documentation beyond fieldwork, for example, how to annotate digitally stored multimedia corpora and how to build and maintain a language archive. Two features of documentary linguistics should be highlighted before we turn to other topics. First, language documentation emphasizes collaborative research by teams of investigators, because no one person has the training to cover all the areas that need to be covered. This is true not only in the sense mentioned at the beginning of this section, where (for instance) collecting data in the field requires one set of skills while designing and maintaining an archive requires a quite different set of skills, but also in the sense that different kinds of training qualify a fieldworker to collect and interpret different kinds of data. Interdisciplinary teams are needed in the field in order to ensure that important specialized areas besides basic linguistic fieldwork are included, among them

On linguistic fieldwork in general (not just on endangered languages)

ethnography, ethnobotany and ethnozoology (the classification and naming of plants and animals), geography, and ecology. A lone linguistic fieldworker can certainly make a start in most of the important areas, but for truly comprehensive documentation, other specialists will have to be added to the team. Second, language documentation ‘requires active and collaborative work with community members both as producers of language materials and as co-researchers’. The ideal situation is one in which the fieldworkers are themselves members of the community, trained in linguistics (and perhaps also in ethnography and other areas closely related to the collection of language data). More commonly, a linguistic fieldworker will train one or more community members to participate actively in the research, perhaps by interviewing elders or by recording and translating stories and songs. As we will see later in this chapter, this ideal is often an impossible goal when the remaining speakers of the endangered language are elderly, few in number, and frail in health. The ideal should still be kept in mind as a goal, even if the goal cannot be approached in a given case. The fact remains that many or most linguists who set out to do fieldwork still work alone. They will fulfill as many of the requisites of description and documentation as possible within the time at their disposal. The main purpose of this chapter is to offer some guidance to prospective fieldworkers for their endangered-language projects, and we will now turn to that topic.

6.2

On linguistic fieldwork in general (not just on endangered languages)

In the United States, the problem of language endangerment became highly visible to linguists and then to the general public primarily as a direct and indirect result of the 1992 publication in Language, the journal of the Linguistic Society of America, of the multiauthored article set ‘Endangered languages’. In Europe, the emergence of widespread concern about endangered languages stemmed from a conference presentation a few years earlier: The trigger for a series of events that led to the establishment of language endangerment as a major topic of concern in mainstream linguistics was a short presentation by Johannes Bechert to the section on universals and typology at the fourteenth International Congress of Linguists in East Berlin in 1987. ... An immediate reaction to this presentation was a motion drafted by Christian Lehmann, which was presented to the business meeting of the Comité International Permanent des Linguistes (CIPL) at the same conference. This motion, signed by more than 200 of the linguists present, urged the committee to take action with the goal of bringing the issue of language endangerment to the attention of professional linguists and the general public.

115

116

field research on endangered languages

Other important efforts around the same time were made by Stephen A. Wurm and Eugenius M. Uhlenbeck, who urged UNESCO to become interested and involved, and Hans-Jürgen Sasse, who led a similar movement in Germany. One outcome of the new emphasis on documenting and preserving linguistic diversity has been the publication of a sizable number of books on field linguistics, among them Newman and Ratliff, eds., Linguistic fieldwork (2001); Crowley’s Field linguistics: a beginner’s guide (2007); Vaux et al.’s Linguistic field methods (2007); Bowern’s Linguistic fieldwork: a practical guide (2008); Grenoble and Furbee, eds., Language documentation: practice and values (2010); Chelliah and de Reuse’s Handbook of descriptive linguistic fieldwork (2010); Sakel and Everett’s Linguistic fieldwork: a student guide (2012); and Thieberger, ed., The Oxford handbook of linguistic fieldwork (2012). Most of the authors and editors of these books, as well as the authors of the chapters in the edited volumes, are experienced fieldworkers who have done extensive research on endangered languages; there is a close connection between a deep interest in endangered languages and a concern with appropriate fieldwork methodology. The authors and editors have as a major goal the education of young fieldworkers, so that more and more undescribed and under-studied languages can be recorded and analyzed. Some of the books discuss issues concerned specifically with endangered languages, though the primary focus of the textbooks, in particular, is on more general fieldwork issues. The present chapter cannot be, and is not intended to be, a substitute for a textbook on field linguistics: the scope of this chapter is considerably narrower, and readers who wish to engage in fieldwork on endangered languages should consult the book-length coverage of the general topic as well as keeping in mind the specific topics covered here. This section surveys typical textbook contents and adds examples to illustrate a few of the issues that arise in a fieldwork project. To give an idea of the kinds of things covered in a textbook devoted to linguistic fieldwork, here are two sample tables of contents. Terry Crowley’s 2007 textbook Field linguistics: a beginner’s guide comprises seven chapters: 1, ‘Field linguistics: why bother?’ (with section headers like ‘Threat to diversity’); 2, ‘Ethical issues’; 3, ‘Getting started’ (choosing a language, doing background reading, planning a trip, getting funding, getting permits, equipment & supplies, etc.); 4, ‘Gathering your data’ (on working with consultants at home and in the field, elicitation techniques, keeping track of data, archiving data); 5, ‘Beyond elicitation’ (on working with texts and participant observation); 6, ‘Problems and pitfalls’ (on analyzing data, learning to speak the language, and other topics, including a final section labeled ‘Linguists behaving badly’); and 7, ‘Salvage fieldwork’, about fieldwork on gravely endangered languages. The second sample table of contents is from Claire Bowern’s 2008 textbook Linguistic fieldwork: a practical guide: 1, ‘Introduction’ (some terminology and other preliminaries); 2, ‘Technology in the field’ (on recording and other equipment); 3, ‘Starting to work on a language’ (what to do at the first session, working out the language’s phonemic inventory, transcribing, etc.);

On linguistic fieldwork in general (not just on endangered languages)

4, ‘Data organization and archiving’ (software, metadata, etc.); 5, ‘Fieldwork on phonetics and phonology’; 6, ‘Eliciting: basic morphology and syntax’ (how to do morphosyntactic elicitation, and why); 7, ‘Further morphology and syntax’ (eliciting paradigms, productivity, etc.); 8, ‘Lexical and semantic data’; 9, ‘Discourse, pragmatics and narrative data’; 10, ‘Consultants and field locations’ (field methods classes and the field, choosing a field site, choosing a consultant, working with semi-speakers, etc.); 11, ‘Ethical field research’ (in recording, archiving, acknowledging speakers, paying consultants, etc.); 12, ‘Grant application writing’; 13, ‘Working with existing materials’ (published materials, other people’s fieldnotes, etc.); and 14, ‘Fieldwork results’ (orthography, pedagogical and reference grammars, training community members, web materials, etc.). Both Crowley’s and Bowern’s books, and other textbooks as well, cover the basics very well. They discuss the fundamentals of choosing a language and a field site, they discuss ethical issues in considerable detail, and they cover the fieldwork project all the way through, from advance preparation to analyzing and reanalyzing the data and the importance of keeping meticulous records in the form of metadata sheets – for each field session, for instance, one records the date, times, consultant(s), and information about what can be found where on the session’s audio (and, if relevant, video) recording. One general desideratum might benefit from greater emphasis: in preparing to do fieldwork, linguists should be sure to familiarize themselves with both older and modern ethnographic writings on the field site and its neighborhood. A solid understanding of the cultural setting of the language is vital for a successful project. Other authors, especially in the edited volumes listed earlier in this section, cover a variety of related topics, among them team-based fieldwork, monolingual fieldwork (when the fieldworker and the native speakers do not share a common language), political aspects of fieldwork, collecting ethnobotanical terms, toponymy (place names), and kinship systems. All these perspectives are important, but as they are not specific to endangered languages, I will not dwell on them here. I will comment, however, on three specific topics that, while they are relevant for all kinds of field research, have special resonance for fieldwork projects on endangered languages: a definition of fieldwork; deciding how much time a documentation project should take; and being prepared for emotional stresses in a remote field site. 6.2.1

What is fieldwork?

First, what counts as fieldwork? Does eliciting data from a native speaker in an office at the linguist’s home university count, or does fieldwork have to be undertaken in the area where the target language is spoken? Linguists’ answers to this question vary. Paul Newman and Martha Ratliff, on p. 1 of the Introduction to their coedited 2001 volume, characterize the task of linguistic fieldwork as ‘describing language as it is used by actual speakers in natural settings’. Later in the same book, Daniel Everett gives this definition:

117

118

field research on endangered languages

Fieldwork describes the activity of a researcher systematically analyzing parts of a language other than one’s native language (usually one the researcher did not speak prior to beginning fieldwork), within a community of speakers of that language, prototypically in their native land, living out their existence in the milieu and mental currency of their native culture.

To these characterizations one might add that fieldwork encompasses language use and cultural aspects of language as well as language structure more narrowly – and that these things are difficult or impossible to study effectively outside the consultant’s native milieu. The question remains as to whether it’s still fieldwork if the research is carried out with one or more native speakers but not on their traditional home ground – in the linguist’s campus office, for instance, or (in a situation that is closer to the more traditional setting) in a diaspora community: a sizable number of endangered languages are now spoken primarily or entirely far away from their traditional homelands, most notably in refugee and other immigrant communities in other countries. The answer to this question is not as important as understanding why the question arises. No linguist would argue (I hope!) that the process is exactly the same regardless of where the work is done; the chances of getting large quantities of reliable natural data, in particular data at the interface of language and culture, are much greater if the linguist spends time with the speakers in their home community. But that does not mean that the processes are totally different. They are in fact closely related, and working on campus with (for example) an engineering student who happens to speak the language of interest natively can produce excellent results. Ideally, for a full-scale research project, these will be preliminary results that will serve as preparation for a field trip to the engineering student’s home territory. I have done fieldwork in the field, within the target language speech community, on just two languages; but I have collected and analyzed data from native speakers (one each) of four other languages as well: Wu Chinese and Karo Batak in field methods classes that I took as a graduate student, and Zulu and Malayalam in field methods classes that I’ve taught. In each instance I learned a great deal about the language and – crucially – about how to work effectively with native speakers. The reason I’ve raised the general issue about whether fieldwork must include travel to “the field” is that the last speakers of a gravely endangered language might no longer live in ‘the milieu and mental currency of their native culture’. If they live in a distant city, uprooted physically and mentally from that culture, the only way to document their language is to interview them far from their heritage territory and unconnected to their heritage culture. In documenting an endangered language, the linguist will often be forced to deal with this and other less-than-ideal circumstances, as we will see in the next section. The connection between fieldwork proper and eliciting data gathered directly from native speakers on the linguist’s home ground explains why most field linguists teach and value field methods courses. (Not all of them would agree,

On linguistic fieldwork in general (not just on endangered languages)

however. One of the most skilled and experienced fieldworkers I have ever known refuses to teach field methods courses because he believes that the subject cannot be taught in a classroom. Prospective fieldworkers who wish to learn from him must do so at his field site. This is definitely a minority view.) A good field methods course can prepare students well for an independent fieldwork project: they learn how to elicit, manage, organize, and store linguistic data; how to analyze pieces of grammatical structure that emerge starting in the very first elicitation session; and how to work with and, if necessary, around the idiosyncrasies of the class’s consultant(s). And if all goes well, they also learn to appreciate the opportunity to get a glimpse of a culture that is entirely new to them. To get an idea of what happens in a field methods class, let’s look at an example, a class that was taught recently by Carmel O’Shannessy at the University of Michigan. Each session of the weekly three-hour class began with a discussion of the previous week’s data and analysis, and preparation of questions for elicitation – questions like ‘How do you say “mother” or “I went to the market” in your language?’, or ‘How do you say “I went, you went, he went, she went, we went, you all went, they went” in your language?’ The next hour was spent in eliciting data from the consultant, a native speaker of Mandinka (a Mande language spoken in Africa). For the rest of the class period, the instructor and students discussed the data, the readings, and documentation tools. Several times in the term the students “met” in a videoconference with fieldworkers from different field situations around the world, so that they could find out how experienced fieldworkers went about their fieldwork. Class participants worked collaboratively to learn how to use ‘tools for recording, transcribing, annotating, searching, compiling, and archiving audio-visual and text data’ (CLAN, ELAN, Fieldworks, etc.). Students were required to analyze each week’s data before the next class session and to be prepared to discuss their own and others’ analyses. In addition to this preparation and to required readings, they worked with other students to learn how to use documentation tools and to teach the tools to other class members. Finally, each student chose a question about the language’s structure to study independently in greater depth and wrote up a grammatical analysis on that topic as a term paper. I would guess that this course was rather typical of one-term field methods classes, including, for all but the most technophobic instructors, the emphasis on technological tools for aiding data management and analysis. The videoconferences with experienced fieldworkers were an innovative and especially valuable part of O’Shannessy’s course: this feature gave the students a broader perspective on the practicalities of fieldwork than any one instructor could offer. Other instructors might emphasize additional aspects of a beginning fieldwork project; in favorable circumstances, for instance, the students could visit a diaspora community where the language is used regularly, developing their language skills by interacting with native speakers in social settings (exercise classes, card-playing, or the like). And in some universities field methods is taught as a two-term sequence, which enables students to learn much more about

119

120

field research on endangered languages

designing and implementing a fieldwork project than is possible in a single term. One term, however, is ample for giving students the basic skills they’ll need to conduct fieldwork in a professional manner. 6.2.2

What is a reasonable time span for a fieldwork project?

A second issue that arises in fieldwork projects on both endangered and non-endangered languages is the question of how long such a project should last. Probably the most typical case is fieldwork undertaken as a Ph.D. dissertation project. In this kind of research the fieldworker may be in the field for about a year, though longer and shorter stays in the field are also common. It may never be possible to return to the field site for any substantial length of time: graduate students earn their degrees, find university jobs or other employment, and put down roots in a community; they then tend to be tied to one place most of the time. In such a case it is irrelevant to ask whether it might have been better if the linguist had been able to spend more time at the field site, collecting more data and learning more about the language. There is certainly great value in working hard under time pressure to produce (for example) a reference grammar of a previously undescribed language, a preliminary dictionary, and a modest collection of texts: this is an excellent start, and it is much, much better than nothing, especially if the language is endangered and unlikely to be studied this intensively by any other linguist. But I believe – though some linguists might disagree – that, after an initial stay of six months to a year, repeated visits to the same field site to gather more data are likely to be even more valuable for extending knowledge of the language. The reason is obvious: one can learn a great deal in six to twelve months, but there will be inevitable gaps in one’s knowledge and understanding of a language after a single visit, no matter how diligent and talented the fieldworker is. It’s true that the learning curve will flatten out to some extent after the first dramatic burst of incoming information at the beginning of a field project. But the more time linguists spend at a field site, the more insight they are likely to acquire into the workings of the language. Here are three examples to illustrate this point. Nancy Dorian began her dissertation fieldwork in East Sutherland, Scotland, in 1963, under the auspices of the Gaelic Division of the Linguistic Survey of Scotland. In the fifty years since then she has produced a continuing stream of important publications on the Gaelic of East Sutherland fisherfolk, from ‘Grammatical change in a dying dialect’ (1973) through ‘The problem of the semi-speaker in language death’ (1977) to Language death: the life cycle of a Scottish Gaelic dialect (1981) and Investigating variation: the effects of social organization and social setting (2010). She revisited her field site for years after her first trip there, and afterward, in addition to working through her accumulated data, she supplemented it by telephone calls and visits by her East Sutherland friends to her own home in Maine. Much of what she has published, including her 2010

On linguistic fieldwork in general (not just on endangered languages)

book, would not have been possible without continuing contacts with Gaelic speakers over the years. A second example is Carmel O’Shannessy’s long-term fieldwork in Lajamanu, an Aboriginal community in the Tanami Desert of northern Australia. O’Shannessy began working in Lajamanu in 1998 and has been there regularly since then, developing pedagogical materials and recording traditional songs as well as investigating the language situation for scientific purposes. If she had not made repeated trips to Lajamanu, she would have been much less likely to discover the emerging new language that she calls Light Warlpiri, a mixed language comprising components of the indigenous Warlpiri language and Aboriginal English/Kriol, together with elements unique to Light Warlpiri itself. The oldest speakers of Light Warlpiri are only about thirty-five years old, so her research has spanned most, or possibly all, of the language’s history. Her longitudinal study of the rise of Light Warlpiri, which was (or is being) created by children who hear code-switching from their parents and other caregivers, would have been impossible without her regular visits to the community. Because this is almost the only known case in which the emergence and development of a new mixed language has been observed, the value of this long-term field research is incalculable. A third example is Daniel Everett’s fieldwork on the Amazonian language Pirahã of Brazil – field research that lasted more than thirty years and included more than seven full years of residence in a Pirahã community. Everett’s 1983 Ph.D. dissertation, A lingua Pirahã e teoria de sintaxe (State University of Campinas, Brazil), was written in the theoretical framework of generative syntax, Noam Chomsky’s famous theory. After continuing to collect data and working to understand Pirahã even better, however, Everett eventually rejected Chomsky’s theory and argued that a completely different theoretical approach best explained the structure of the language. His new claims about Pirahã have received much publicity and caused much controversy among adherents of Chomsky’s theory and others. Some of the criticisms leveled against Everett’s current analysis are based on the premise that it is inconsistent with the analysis in his 1983 dissertation. Everett’s response is that he understands more about Pirahã now than he did in 1983. I do not believe that any experienced fieldworkers, regardless of whether they accept his current claims about Pirahã and its implications for syntactic theory, would disagree with Everett’s view that as one learns more and more about a language, one’s analyses of its structure will inevitably change. Examples of the advantages of long-term field study of a single language could be multiplied without difficulty. The fact remains that the vast majority of fieldworkers either pay one lengthy visit to a given field site or pay regular visits to field sites but add languages to their portfolio instead of continuing to work on a single language over a period of many years. In giving examples of sustained fieldwork projects, I do not mean to suggest that all fieldworkers should adopt this strategy: there is no “should” in this discussion. My purpose, rather, is to counteract the impression (which I believe is widespread) that a single one-year

121

122

field research on endangered languages

visit is enough to provide fieldworkers with all the data they need to understand a language fully. It isn’t. In undertaking fieldwork, then, a wise linguist will take into account both the desirable and the possible. For many fieldworkers, one long visit to a given field site is all they want or can manage. For others, the primary goal will be to document as many languages as possible in the time available, and that goal too will dictate spending limited time in any one field site. But some fieldworkers will be willing and able to return again and again to the same field site, and this strategy should also be applauded. 6.2.3

Dealing with psychological stress in the field

The final specific topic that deserves some discussion here is the psychological dimension of fieldwork – again, on the assumption that the fieldworker is a novice alone in the field and not (as in the ideal case described earlier) a member of a team. There is a considerable literature on this subject in anthropology, but much less in linguistics. Psychological problems can afflict a fieldworker in a remote field situation: culture shock, loneliness, emotional and sometimes physical stress, social awkwardness. Most linguists who raise this issue acknowledge the stresses while emphasizing that the rewards tend to outweigh them. Nancy Dorian, whose joy in her fieldwork among East Sutherland fisherfolk in Scotland shines through all her writings on Gaelic, presents this eloquent picture of both aspects of fieldwork, as she experienced it in her dissertation research: Entering an unfamiliar social world is guaranteed to plunge the novice researcher into something like a second adolescence: a constant succession of uncomfortable situations in which he or she has no clear idea how to behave and is very likely to behave inappropriately. There must be some substantial inducements to coax the student forth, as of course there are: the excitements of novelty and discovery, and the satisfactions of making a first real trial of professional skills. In retrospect, I wouldn’t wish the tensions or even the painful blunders away. They belong to the learning process of an immersion experience and are often the engine of discovery, casting linguistic and cultural differences into sharp relief. Some of the special insights of fieldwork may hinge on them.

But although most linguist-fieldworkers would agree, to judge by their comments on their experiences in their writings, that the stresses of fieldwork are counterbalanced by the good parts, this feeling is not universal. One linguist has described what sounds like severe distress lasting for weeks, and cites anthropological sources that say such distress is commonplace. I hadn’t previously experienced or heard, from either linguists or anthropologists, about stresses of such severity – awkwardness and loneliness, yes, but not to the point of repeated uncontrollable crying jags. A major fieldwork project is an intense experience that requires an enormous time investment and can take an emotional

On linguistic fieldwork in general (not just on endangered languages)

toll. But although warnings are needed, so that young scholars don’t set out alone for remote field sites unprepared to deal with the various problems that can arise, first-time fieldworkers should begin in a spirit of optimism. Unless one is very unlucky, the difficulties will be more than offset by the personal and intellectual benefits. I’ll give two examples from my own experience as illustrations. In the mid-1960s, as a graduate student, I spent a year in Yugoslavia doing dissertation research for a project on word formation in nonstandard dialects of the language then known as Serbo-Croatian. I spent much of the year in Novi Sad, studying with the preeminent Serbian dialectologist Pavle Ivi´c and reading everything I could find on Serbo-Croatian dialects. Then I traveled to rural villages to collect data from dialect speakers. Yugoslavia was then a communist country (though independent of Soviet power), ruled by the dictator Josip Broz Tito, who was President of Yugoslavia from 1953 until his death in 1980. I did suffer from culture shock and loneliness while in residence at the University of Novi Sad – the sort of thing that can happen to any foreign student in any country – and also, though to a lesser degree, while I was away from Novi Sad doing fieldwork in rural areas. Professor Ivi´c helped me choose villages for the fieldwork part of my project and told me what to do when I arrived in a village: go to the local communist official and announce my presence and my plans. In the villages I visited, the local officials were welcoming, and they put me in touch with villagers who were willing to talk to me. I had been terrified by horror stories I’d heard before arriving in Yugoslavia – of American graduate students being tossed out of the country for acting in a politically inappropriate way, and having their painstakingly collected field data confiscated – so I rigorously avoided engaging in any conversation or other activity that could possibly offend any communist official. But the officials I encountered were uniformly polite; the only people who were rude to me that entire year were some American diplomats’ wives at the American embassy in Belgrade. I had a questionnaire that I needed to fill out in each village, but this presented difficulties: it was not appropriate (as I soon discovered) for a woman to question a man. Men were happy to talk to me at length, but they wouldn’t help me fill out my questionnaire. So I needed to work with women only; but the women wouldn’t talk at all when their husbands were present, which meant that I had to find the women alone somehow. On the island Krk, for instance, I would leave the inn in town each morning and hike out to the house of a particularly helpful woman. I would then duck down behind the low stone wall across the road from her house and wait there in hiding until her husband (who was always eager to talk to me, in English, about his years in the American merchant marine) left to go into town. I felt like a complete fool hiding behind that stone wall for an hour or more each morning, but it was well worth the effort for the wonderful data I collected from the wife once the coast was clear. Another problem was that I spoke only Standard Serbo-Croatian, and my standard speech caused nonstandard-dialect speakers to switch from dialect to

123

124

field research on endangered languages

standard to the best of their ability during our interviews. It would not have been appropriate for me to try to speak in their dialects, even if I had been able to do so; it would have sounded as if I were mocking them for being uneducated. I dealt with this problem by finding consultants who had never been to school – that was still possible in the mid-1960s – so that their command of the standard dialect was minimal. My standard speech caused trouble of a different kind in a Montenegrin village near the Albanian border. The village turned out to be a popular tourist destination, and it was hard to find any speakers of the local dialect. Finally I was directed to an elderly woman who was herding sheep on a mountainside within sight of Albania. Albania was then a country closed to Westerners, with a Chinese-friendly communist regime. That was relevant because the elderly shepherd, I assume because of my nonlocal appearance and my Standard Serbo-Croatian, thought I was from the Yugoslav secret police and was extremely reluctant to tell me anything at all. Most of my consultants would chuckle when I asked (for instance) what they would call a man with a big nose if they didn’t like him. This one said, ‘God gave him a big nose; it is not for me to criticize.’ And then, at the end of the interview, she made me erase everything I’d written. It was not a successful stop on my travels around the country. Overall, however, I succeeded in my main goal in Yugoslavia (other than the dissertation) – I learned how to do fieldwork. Although the experience was not always enjoyable, it was never traumatic, and the benefits of learning about fieldwork and Serbo-Croatian dialects far outweighed the inevitable inconveniences of being so far out of my comfort zone. My second example is from my current fieldwork project, a long-term study of the Salish-Pend d’Oreille language with the goal of producing a grammar, a dictionary, and a text collection – the old-fashioned ideal for a major descriptive fieldwork project (but it was not yet old-fashioned when I started more than thirty years ago). Every year since 1981 I have spent some time on the Flathead Reservation in northwestern Montana, working with Salish and Pend d’Oreille elders to document their language. The usual pattern for the past twenty years or so – always in the summer – is a weekly one-day field trip to the Salish-Pend d’Oreille Culture Center in St. Ignatius, a 212-mile round-trip drive from my summer base in Condon, Montana. The drive takes more than two hours each way; it’s unsafe to drive fast at a time of day when suicidal deer are most likely to leap out of the woods in front of the car. It takes me one or two full days to prepare for each trip, and it takes two or more full days afterward to type up and organize my field notes from a one-day field session, including entering the data in the various computer files for the dictionary and the grammar. (One-day trips are the norm for me because consecutive days of work yield diminishing returns, due in part to the limited stamina of the elders and in part to my own limited stamina. And I do not join the elders in cultural activities such as berry-picking and camas-baking because I have not been invited to do so; in any case, English

Investigating gravely endangered languages in the field

is the language used during even the most traditional activities nowadays, so I would hear little if any Salish-Pend d’Oreille if I were able to attend.) Every time I leave to drive to the reservation, shortly before 7:00 AM, I feel a bit glum about the day to come: long drive there, uncertainty about which elders will come to the Culture Center to work with me, social anxiety (staying home is much less stressful), exhausting day’s work, and then a long drive home again when I’m very tired. I always feel like an outsider and an intruder, weighed down by the burden of past White government officials, clergymen, nuns, linguists, and anthropologists who have exploited and otherwise mistreated Native Americans. (This is not because the elders I work with display any distrust or hostility in our time together; they appear to accept me, their laughter at my mistakes is good-natured, and I always enjoy their company.) This feeling of I-don’t-wantto-do-this lasts until I enter the Culture Center. But by the end of each day’s work with the elders – and this is the main point of this paragraph – I am exhilarated as well as exhausted: I’ve learned so much, I’ve heard new words and gained new insights into cultural attitudes, names and uses of medicinal plants, hunting and gathering practices, and other aspects of the elders’ traditional culture. Usually I’ve had at least a few lightbulb moments, sudden flashes when I understood something about the language or the culture that had puzzled me before. Every time I leave the Culture Center to drive home in the evening, I know how extraordinarily lucky I am to have this priceless opportunity to learn about this beautiful language. I do not know to what extent my two main fieldwork experiences are typical; certainly the details of my current annual fieldwork differ from most fieldworkers’ schedules. I do know, however, that many, probably most, other linguists who have had substantial fieldwork experience have found it exciting and rewarding, and have dealt with the inevitable emotional and physical problems without letting them erode the positive impact of the experience – the opportunity to learn about a language in its own setting. Let’s turn now to an examination of some special problems that arise in fieldwork projects devoted to describing endangered languages. Of course not all of these difficulties will arise in every project that aims to document a gravely endangered language. But if a novice fieldworker knows about them in advance, they are less likely to derail an otherwise carefully planned project: forewarned is forearmed.

6.3

Investigating gravely endangered languages in the field

As mentioned at the beginning of this chapter, fieldworkers who set out to describe and otherwise document endangered languages often face challenges that differ in kind and degree from those encountered when the target

125

126

field research on endangered languages

language is not endangered. Prominent among those challenges are these. The choice of language consultants is likely to be severely limited, so that cultural barriers that would be trivial under other circumstances can become serious hindrances (for instance when men are disinclined to answer questions posed by a female fieldworker). It may be impossible to resolve issues of incompatible grammaticality judgments (answers to questions like “Is this sentence grammatical in your language?”) and conflicting data received from different consultants. The last fluent speakers may well speak a language that differs sharply from earlier generations’ speech, and the fieldworker must try to sort out the differences and to satisfy community members who, in some cases, will prefer to standardize and preserve the current state of their language rather than that of the earlier generations – although the earlier state, if it is recoverable, may be of intrinsically greater interest to linguists and other scientists. A related problem arises when linguists help community members’ efforts to revive a language that is not currently spoken. In such cases the only available evidence will be in libraries and/or fieldnotes (often unpublished). The rest of this chapter outlines the joys and frustrations of working with the last speakers of a language – salvage linguistics, in the most severe cases, when no fully fluent speakers of the language remain – and with the partial successes in documenting these languages. Almost all of the pieces of a fieldwork project on a non-endangered language will also be vital for a project on an endangered language: choosing a field site, paying close attention to ethical matters, selecting consultants, technological concerns, and the rest. But some options that are available for a non-endangered language may not be available to a fieldworker who is investigating a gravely endangered language. Here I focus on issues that arise in the latter situation but are not as likely to present problems in the former. 6.3.1

Working with consultants: access, collaboration, communication, selection, and data-collection techniques

First, the fieldworker needs to have access to fluent speakers of the language, assuming of course that there are still a few fluent speakers. Access is not as likely to be problematic with a non-endangered language, partly because the community is less likely to feel a need to protect its elders from exploitation and partly because, if the community is large, its members are less likely to be interested in what everyone else in the community is doing. In North America, and in some other parts of the world as well, indigenous communities have learned to be wary of outside linguists and anthropologists: often outsiders are viewed as stealing the community’s linguistic and cultural knowledge and using it to make themselves rich. (Academics, at least in the U.S., do not consider themselves rich. It’s all relative: compared to most Native Americans, Americans with tenure-track academic positions are definitely rich.)

Investigating gravely endangered languages in the field

To ensure that the work can proceed smoothly (or at all), it is therefore vital to have the project approved by the relevant authorities – the local communist official during my time in the former Yugoslavia, for instance, or community leaders of other kinds in less bureaucratic cultures: the tribal council, the elders’ council, whatever person or group has the authority to give approval. Approval can take the form of anything from a simple verbal OK to a formal written protocol that specifies precisely what the linguist can do, what payments are expected, who s/he can work with, and what s/he can do with the data once it has been collected. The Salish-Pend d’Oreille Culture Committee has never asked me to sign a written protocol, for example, perhaps because I began working there at their invitation: they wanted a linguist, back in 1981, to help them learn how to write their language in their new IPA-based alphabet. But I don’t submit a paper for formal publication until the Culture Committee has given me permission to submit it. In some parts of the world, it might be unnecessary to request and receive approval for a fieldwork project. But the fieldworker should make sure of that, and receive approval in all other cases, before approaching any speakers. Research universities in the United States, the context I’m most familiar with, now have Institutional Review Boards (IRBs) that must approve all research inolving human subjects, before the research begins. The purview of these boards includes both ethical issues and the requisite that the project not cause harm to subjects – two categories that overlap, of course. Definitions of “human subject” vary from university to university, however, and linguists must learn what their university’s definition is, and satisfy the IRB’s requirements, before setting out on a field trip. Many linguists, and often their consultants too, object to the designation “human subject” for fluent speakers of the language being investigated: to me, and I think to most of us, the consultants are language workers, colleagues, and not “subjects”. Still, in some parts of the world the IRB restrictions make excellent sense. Speech can kill, and fieldworkers’ highest priority should be to avoid causing harm to the speakers they work with. Denying the expert status of the speakers may seem offensive, but that feeling will not remove the necessity of having IRB approval for a fieldwork project. The U.S.-style IRB is merely the tip of the iceberg when one begins to consider ethical issues that are important in fieldwork. Most ethical issues (including those monitored by IRBs) arise in any fieldwork context and are not specific to fieldwork on endangered languages; these are covered by field-methods textbooks such as the ones listed early in this chapter. Ethical concerns that are specific to endangered-language fieldwork arise most frequently in contexts involving revitalization efforts. These are discussed in the next chapter. Once the fieldworker meets the people s/he plans to work with – perhaps the last fluent speakers of the language – the most important task by far is to get along with them. The wise linguist never argues with consultants, never disagrees, and always shows respect for their opinions as well as for their knowledge of their

127

128

field research on endangered languages

language; and s/he is sensitive to local cultural norms. As noted earlier, I found that I could not work with men in the Yugoslav villages I visited, and I also couldn’t work with their wives while the men were present – men would talk to me, but as a woman I shouldn’t have been asking them questions, and it wasn’t considered proper for their wives to talk to a visitor (me) while the men were present. Note well the point about showing respect for the speakers’ knowledge of their heritage language. This is imperative even when (or especially when) the linguist knows full well that a particular consultant is providing flawed data, either in general or in a particular context – because of an accidental speech error, or because more fluent speakers have provided data that conflicts with this consultant’s utterance. Even the most fluent speakers make actual mistakes when speaking their language, and paradoxically this often happens when they are thinking about what they’re saying, and analyzing it, instead of just speaking naturally. As we saw in Chapter 3, it is not always easy for an outsider to distinguish a fluent speaker from a semi-speaker of a dying language; and if the only remaining speakers are semi-speakers or even rememberers, the linguist must work with them. In such cases the linguist may well know better than a consultant what the “correct” forms are, because of reading done in preparation for setting out to do fieldwork: there may be earlier documentation on the language, and the language may also have close relatives that the linguist has read about. If a linguist is so foolish as to challenge a semi-speaker on some point of grammar, that consultant may disappear, and tell his or her friends about the visitor’s rudeness, and further fieldwork at that site may be impossible. As we also saw in Chapter 3, there may turn out to be no fluent speakers left in the speech community, or even semi-speakers, so that the fieldworker may be forced to conduct a salvage linguistics project. In this case the consultants will be rememberers, people who can at most provide some words and phrases of the dying language. This doesn’t mean that the investigation has failed; it does mean that the results of the analysis must be treated with extreme caution, because they cannot be as reliable as an analysis that is based on data collected from fully fluent speakers. In most cases, however, careful advance planning will ensure that a first-time fieldworker will not have to face a salvage situation. Salvage linguistics is not a project for the faint-hearted, or for beginners. The availability of a few fluent speakers does not, of course, lessen the need to get along with one’s consultants. If a fieldworker offends consultants, they won’t want to continue working on the project. Fieldworkers and consultants often develop close working relationships and even personal friendships, but the linguist should always be aware of one salient fact: the linguist needs the consultants more than they need the linguist. If their language is gravely endangered, the fieldworker must not risk doing anything that might make them skip scheduled sessions or, in the worst case, decide not to continue the work. There might be no other fluent speakers available and/or willing to carry on the work, especially as any offense given will probably damage the fieldworker’s

Investigating gravely endangered languages in the field

reputation throughout the community. Moreover, offending the community is liable to have repercussions for future fieldworkers as well, because community members might conclude from one bad experience that they don’t want to work with any outsiders at any time. As indicated in the section on documentary linguistics, the ideal fieldwork situation is one where a linguist collaborates with native speakers who are trained in language work – that is, speakers who are not only fluent but who also have a good grasp of the nature of the task, if possible including some training in linguistic analysis. But in a field situation where the linguist works with the last fluent speakers of a language, organizing this kind of collaborative project might be impossible. If only a few fluent speakers remain, they will very likely be elderly. There are barriers to teaching elderly people analytic techniques. No matter how passionate they are about helping document their language, elders may be deterred by health problems and lack of stamina, and in some instances by educational background as well, from learning what they would need to learn in order to be full scientific collaborators in a documentation project. This does not make them mere helpers instead of colleagues; they are still the experts, and the linguist is the learner. But their participation is more likely to involve providing data than helping to analyze the data they provide. Another model of collaboration that is highly recommended for major fieldwork projects – also described earlier in the section on documentary linguistics, where the team is interdisciplinary and comprised of scientists from outside the community – might or might not be feasible in work with an endangered language. That will depend on whether there are several fluent speakers or just one or two, and also, of course, on the community’s willingness to welcome a whole group of outsiders into their midst. It is easier to get permission for one person to work with elders than for a group. A team consisting of a linguist, an ethnologist, an ethnobotanist, an ethnozoologist, and perhaps other experts as well can’t easily work effectively with just one or two fluent speakers; elders can hardly be expected to work for many hours at a time, with little rest. A third collaborative model, one that can be highly effective under the right circumstances and that is feasible with endangered language research, has been used in various places (for instance by Russian fieldwork expeditions going back to the 1930s). One prominent instantiation of this model was designed and implemented by Terrence Kaufman. He calls it the parallel project approach to language documentation, and he employed it in field projects for many years, starting in 1970. Most recently, starting in 1993, he was co-director of the Proyecto para la Documentación de las Lenguas de Meso-América (PDLMA) in Mexico. The project functioned in the summer months, when most U.S. academics do not have classroom obligations, and its members worked on at least twenty-six different languages. He says that similarly organized projects ‘could be done anywhere there are roads, buildings for rent, doctors, markets, electric power, and no armed conflict in the neighborhood’.

129

130

field research on endangered languages

The PDLMA employed several linguists working on parallel research on a variety of languages, one linguist to each language. They all lived in the same quarters (for instance a local hotel), ate together, and met regularly to discuss analytic and other issues. Native speakers of the languages being documented also lived in the same place and ate and worked with the linguists. This arrangement produced a remarkably efficient operation. The linguists, most of whom were students, were extremely productive under these circumstances, and the project director was always at hand to provide guidance as needed. Kaufman noted another advantage to having the whole team together in a central location rather than living and working separately in the various speech communities: It can be disruptive and politically problematic (if not downright dangerous) for individual linguists or teams of linguists to spend long periods of time in minority communities whose languages are being documented. At the beginning stages of investigation, the linguist’s presence in the community should be highly circumscribed. Linguists should develop good working relationships with one or two members of the community who will serve as buffers and filters when the linguists make any serious appearance in the community.

Some linguists will disagree sharply with Kaufman on the advisability of doing one’s field linguistics outside the relevant speech community. Certainly there are drawbacks to his approach in many, perhaps most, fieldwork contexts: most obviously, the consultants are separated from the communities in which their language has its life, so that the possibilities for observing and recording the language in use are greatly diminished or even nonexistent. Using this method is only feasible, of course, if there are fluent speakers who are willing and able to be away from home for two or more summer months. The considerations that Kaufman raises, however, should be kept in mind in planning a fieldwork project. Sometimes, as he warns, going immediately to the speech community is not a good idea. For efficiency, at the beginning of a documentation project, Kaufman’s arrangement is excellent: the speakers are outside their community, sometimes quite far from home, living and working in an intense professional milieu that encourages diligence. This is also excellent if the primary goal is collecting a great deal of data rapidly. But it may be unworkable for languages with only a handful of remaining fluent speakers; they might not be willing or able to be sequestered far from home for nine weeks of the summer. (The elders I work with in Montana would certainly refuse to cooperate: they wouldn’t want to miss the season when the huckleberries are ripe, and that’s just one part of their lives that would be unacceptably disrupted.) Often Kaufman’s consultants were young people, always more likely to be mobile. Some PDLMA linguists did later spend long periods of time in the speech communities, typically after they had spent the necessary time – at least one long summer session – in the central project location to begin data collection and to establish good working relations with key consultants.

Investigating gravely endangered languages in the field

Another question that arises frequently is how fieldworkers should communicate with their consultants. To begin with, the fieldworker will almost certainly not know the language s/he is investigating, but fieldworker and consultants must somehow communicate with each other. The most obvious solution is to use a language that both of them already know. My fieldwork in Montana has never presented any communicative obstacles, because all the elders I’ve worked with have native-speaker fluency in English as well as in their heritage language. In fact, for all but two or possibly three of them, English is their dominant language; most of them have not spoken Salish-Pend d’Oreille regularly for decades. They have told me that our weekly sessions during the summer, plus monthly elders’ meetings during the rest of the year, are their main opportunities nowadays to use their language. (I did work for years with the last married couple who spoke the language daily at home; the wife, though fluent in English, did not like to speak it. She was the only elder I knew who clearly still had Salish-Pend d’Oreille as her dominant language.) The situation was different when I conducted fieldwork in the former Yugoslavia; there, I needed to speak Serbo-Croatian to the dialect speakers I interviewed, because that was the only language we had in common. But what if the consultants share no common language with the visiting linguist? Again, there is an obvious answer: one can use another community member as an interpreter, to translate questions from English or Spanish or French or some other language back and forth between linguist and consultant. This process is cumbersome, and it adds a layer of uncertainty (as well as expense) to the collection of data, because there may be no way to tell how accurate the interpreter’s translations are, in both directions. It can work very well, however, if the interpreter is a community member who is also a full collaborator on the research, trained in linguistics and in methods of collecting linguistic data. Another possibility is to eschew interpreters and do monolingual fieldwork. There are well-developed techniques for working one’s way into a language when the linguist and the consultants cannot (at first) understand each other. The linguist Kenneth Pike used to give monolingual demonstrations of these techniques at linguistics conferences and at different universities. I first heard one of his impressive performances when I was a student at the 1964 Linguistic Institute at Indiana University. Pike was joined on the stage of a large auditorium by a man he did not know, a native speaker of a language whose identity was also unknown to Pike, and the young man had been instructed not to say a word to Pike in any language other than his native tongue. Pike used a few natural props like sticks to illustrate dropping and setting things on surfaces, he demonstrated actions like walking and falling, he held up different numbers of fingers to indicate numbers, and he used various other techniques as well to elicit words and short sentences. Throughout, he used gestures to convey the meaning of his questions to the young man. He said all his questions aloud, not in English (which both he and the young man knew well) but in an indigenous Mexican

131

132

field research on endangered languages

language that he, Pike, was fluent in. Pike’s questioning methods were clear enough that the young man was able to provide answers confidently, and Pike wrote all of the utterances on a large blackboard. Eventually, after half an hour or so, Pike stopped asking questions and analyzed the data he had written on the blackboard, forming hypotheses about the language’s structure; and finally, as the peak of his performance, he guessed – correctly! – at the identity of the language. Pike was a master showman, and he made the dazzled audience members believe that this method could indeed be used successfully in the field. The goal, if one is obliged to start a fieldwork project monolingually, is to learn to speak the consultants’ language and then communicate in that. That takes quite a bit of time, but it’s the only way to proceed if there is no other route to direct communication. Learning to speak the language one is investigating is, for many mentors of novice fieldworkers, an obvious requirement for any major fieldwork project. Experienced fieldworkers routinely report that speaking the consultants’ language ‘contributed greatly to their fieldwork success (and that lack of speaking ability hindered their progress). ... If the linguist knows the language, he or she can learn a tremendous amount by merely listening.’ Learning the language can provide additional benefits in helping the linguist contribute to revitalization efforts, as Peter Sutton reports from Australia: ‘[L]inguistic competence in an Aboriginal language by a non-Aboriginal person [is taken to] imply not only cultural competence and understanding, but also the worth of Aboriginal culture itself.’ There is no doubt that this is a correct assessment; the benefits of speaking the language one is documenting are enormous. But there are circumstances, especially with fieldwork on an endangered language, when it is inadvisable for the field linguist to learn to speak the language, and quite possibly it would be impossible to achieve fluency even if it were appropriate to learn the language. The advantages of learning to speak the language are obvious; the disadvantages may become evident only gradually. I mentioned earlier that it would have been inappropriate for me, back in my dissertation-research year in Yugoslavia, to try to speak the dialects I was interested in: the speakers would have viewed any such effort as condescending, and as unsuitable for an outsider in any case (regardless of whether the outsider was a foreigner, like me, or a visitor from elsewhere in Yugoslavia). They themselves mostly did not speak Standard Serbo-Croatian well, but they knew it when they heard it, and they knew that their dialect was nonstandard. (They might also have felt that it would be intrusive and presumptuous for an outsider to try to speak their dialects; I don’t know, but I believe that this is a common feeling in minority speech communities.) In Montana, the few fluent elders rarely speak their native language nowadays, and they speak it only to each other – they are not accustomed to speaking it to younger family members or to the numerous elders who do not know it, much less to non-tribal members. They tell me that there used to be White people who lived on the reservation and spoke Salish-Pend d’Oreille (then called Flathead

Investigating gravely endangered languages in the field

in English), but that was before 1910, the year Congress opened the Flathead Reservation to White settlement and Whites poured in to snap up the cheap land. In my field sessions I always work with a group of elders rather than with one person at a time. Because the fluent elders do not expect anyone but themselves to speak their language, they also don’t expect to be understood by anyone else when they talk to each other in Salish-Pend d’Oreille. I have found that it makes them a bit nervous when I show that I understand parts of their conversation by (for instance) laughing at their jokes. This makes me try hard not to cause discomfort to the people whose good will I absolutely depend on to do my language work. When I mentioned the elders’ reaction in a recent conference presentation, an audience member responded with a similar but more poignant story: when he was working on the language of an indigenous community in Mexico, he got along well with the language’s speakers, who were happy to work with him – until, by devoting much time and effort to the task, he learned to speak their language. They then told him that they no longer trusted him, and his working relationship with them suffered accordingly. It is difficult to know how common it is for community members to react negatively to the prospect of a fieldworker learning their language. Certainly the opposite is frequently reported, when speakers of endangered languages are thrilled to find that their language and culture are taken seriously by an outsider (as exemplified earlier in the reaction of Australian Aboriginals cited by Peter Sutton). But the more endangered a language is, the more careful the field linguist must be to avoid violating community norms on this matter, because this is when community members may be most protective of their language and the rest of their fading cultural heritage – and the most reluctant to share them with outsiders. After getting permission to work on documenting an endangered language in the field, the linguist will turn his or her attention to the selection of consultants, assuming that there are choices to be made. In discussing the participation of community members in a beginning fieldwork project, I focus here exclusively on the consultant role for two reasons. First, a project begins with elicitation of data from fluent speakers, who will be consultants; other kinds of community participation come later, after the linguist and the community have become familiar with each other – unless, of course, earlier visiting linguists have already laid the groundwork by training interested local people to contribute in other ways, such as helping with transcription (where linguistic training is required) and translation, or operating video and/or audio recording equipment. And second, if only a few fluent speakers remain, there may be no community members who can or wish to participate in other ways. In most cases it will take time for the visitor to develop working relations with the community; in the meantime, collecting data from consultants should proceed. Fieldwork textbooks typically have advice for the novice field linguist on selecting consultants. Jeanette Sakel and Daniel Everett’s textbook, for

133

134

field research on endangered languages

instance, suggests the following selection criteria for a good consultant: a good story-teller, friendly, a speaker of the correct dialect (i.e., the dialect the linguist wishes to document), a native speaker of the language, lacking speech impediments, able to reflect on the language as a formal system, and well respected in the community (or at least not marginalized). This is a good set of criteria, but it won’t be of much use if there are only two or three remaining fluent speakers of the language. Carefully selecting one or more consultants who are close to the ideal will not be an option in that case; the linguist will work with whoever is available and be grateful for the opportunity. One selection criterion cannot be set aside: to be acceptable, a consultant must be willing and available to work on the project. A second obvious criterion, that the consultant be a native speaker of the language, also seems like a no-brainer until one considers the possibility that, in the least favorable case, there might no longer be any native speakers of the language. Then the project is a salvage operation, and the linguist will work with the next best consultant, perhaps someone who learned the language as an adult and is a fairly fluent speaker. In any case, even with a non-endangered language the ideal consultant probably does not exist (though many come close). I have found that different speakers of Salish-Pend d’Oreille have different skills, all of them valuable. One elder who worked with me until his death (at age ninety-two) in 2001 would happily work for hours at a time with hardly any breaks. He was invariably patient with my mistakes, he had a deep knowledge of and love for his native language, and his memory was phenomenal. He would tell stories, but he didn’t enjoy going over them sentence by sentence afterward so that I could transcribe them, and he wasn’t interested in listening to recordings of earlier storytellers and helping me transcribe those. Other elders have been willing to help me transcribe old and new recordings, but not to tell stories of their own. One elder’s pronunciation of the language’s four pharyngeal (back-of-the-throat) sounds is especially clear, but his glottalized stops and affricates are so soft that I have great trouble hearing them accurately. Another elder is especially talented at explaining nuances of meaning that separate near-synonyms, but he prefers to let others provide the actual data. Every consultant I’ve worked with has had both strengths and weaknesses, and I’m sure that other field linguists have had similar experiences. Before leaving the topic of choosing consultants, I will make one other point. As noted earlier, the linguist will choose the most suitable consultant available and make the best of the situation, drawing on the consultant’s strengths and avoiding any areas of weakness. But the linguist is not the only one who has to make difficult choices when few options are available. The endangeredlanguage community too will typically have to settle for a nonoptimal linguist to help document its endangered language. A community trying to preserve and revitalize its heritage language has an agenda; a linguist coming to the community for fieldwork also has an agenda, and the two agendas are unlikely to

Investigating gravely endangered languages in the field

match perfectly. To be successful, a fieldwork project on an endangered language must involve compromise on both parts – but the linguist is the one who should give way and permit community preferences to guide the process. I was lucky when I began fieldwork on Salish-Pend d’Oreille, because my research career was already established and I was therefore under no pressure to publish my analyses as scholarly articles. Anything I could learn about the language would be valuable for me, so I was eager to focus on whatever the tribes most wanted. It wasn’t easy to find out what they wanted, because they were reluctant to ask for anything; but finally, by trial and error, I discovered that they would most value a dictionary and a collection of analyzed texts, so those are the projects I have worked on. Many fieldworkers, especially those who are at the beginning of an academic career, do not have the luxury of accommodating completely to the community’s wishes, however. In most cases, some compromise will be necessary. 6.3.2

Field sessions

In this section I envision field sessions in which the fieldworker and the consultant(s) sit around a table or on the ground, all together and sedentary. Although this is hardly the only appropriate setting for fieldwork, it has great advantages in efficiency: one can gather a great deal of data quickly, with few distractions. It is the best setting at the beginning of a project. Later on, especially if it is both possible and culturally appropriate for the fieldworker to learn to speak the target language, excellent data can be collected by joining consultants in traditional cultural activities (if and only if the fieldworker is welcome on such occasions), or by taking them to culturally important sites that they might otherwise not be able to visit easily: these activities offer opportunities to hear the language spoken in its natural setting, provided that it is still spoken at all at the events. But in the beginning stages of a project, the novice fieldworker is not likely to derive much direct benefit from hearing the language spoken naturally. Each field session with one or more elderly speakers of an endangered language should have these three features: thorough preparation by the linguist for the day’s work (this is equally important in working on a non-endangered language, of course); a careful pacing of the work throughout the day, so that consultants do not become overtired; and the use of a variety of data-gathering techniques, to avoid boring the consultants and to increase the quantity and quality of the collected data. I’ll discuss each of these in turn. The linguist is also responsible for making sure that the work goes as smoothly as possible, that the consultants are comfortable and well fed, and that the day’s work is recorded – in transcriptions by the linguist as well as on recording devices. Earlier in this chapter I mentioned that it takes me one to two full days to prepare for each one-day field trip to the Flathead Reservation to work with Salish-Pend d’Oreille elders. It used to be a much quicker process. In the beginning stages of a fieldwork project, when the total amount of data is very small,

135

136

field research on endangered languages

thinking of new questions to ask is easy: basic vocabulary from a standard list, paradigms of verbs (‘I saw Mary’, ‘you saw Mary’, ‘he saw Mary’, etc.), simple sentences (‘Johnny killed a cow elk’, ‘Mary went into the store’), and so forth. Later in the course of a long-term project, when the database is much larger, it’s harder to come up with new things to ask. For well over a decade now, my focus has been on the dictionary part of the Salish-Pend d’Oreille project: I need more words, many more words, and I need prompts to stimulate the elders’ memories. I’ve already exploited all the easy sources (more about that later in this chapter), and it takes more and more time each week to put together a long enough list of questions – four to six pages of specific questions will be needed. Novices might expect that it would be easy just to think of things to ask on the spot, while the elders sit attentively waiting for the questions. It isn’t. Without a script, the fieldworker will inevitably waste time, and that is a serious mistake. With a gravely endangered language, time is always a factor because it is inherently limited by the age and infirmity of the last few speakers, even if the linguist has unlimited time. The consultants may free-associate and offer new words or sentences or forms or stories, and the linguist will welcome all offerings (even when the particular items are already in his or her database). But the fieldworker cannot afford to indulge in the luxury of random free-associating. S/he must be prepared with more questions than there will be time for, to ensure that the supply won’t run dry before the day ends. The linguist will therefore arrive at every session ready to fill all the time available with questions. But part of the preparation is a readiness to have the session canceled without warning. Things can’t be expected to happen on a rigid schedule: if the huckleberries are ripe, the most important consultants might be out in the mountains berry-picking; or they might have doctor appointments and skip part or all of a day’s session; or they might have visitors and skip a day’s session. Or they might have fallen ill and be unable to work. Working with a group is some insurance against a canceled session – if one consultant is missing, the others can carry on – but a change of plans might still be necessary, because the plans for the day might require the expertise of the missing consultant. So the linguist needs to be flexible enough, and sufficiently well prepared, to ask different kinds of questions if s/he is forced to abandon a particular line of questioning. Flexibility and a non-anxious frame of mind are also required for successful fieldwork, together with a time frame that is at least somewhat open-ended. A linguist might arrive in the field, having planned carefully, received all permissions, and scheduled field sessions, only to find that no consultant will be available for the next two or three weeks. In Montana, I learned long ago that no field sessions will occur during the week before the annual July powwow or during the days of the powwow itself, because the elders will be preparing for the powwow and then at the powwow grounds dancing, watching the dancing, eating frybread, and playing stickgame. This kind of scheduling restriction is not problematic in a long-term field project, but it can be devastating to a project

Investigating gravely endangered languages in the field

with a short, rigid time frame. The lesson for novice fieldworkers who plan to work on a seriously endangered language is to allow more time than they expect (or hope) to need. Pacing the work during a day’s field session is always important, but especially so when there are only a few elderly speakers of the language. Most elderly people, no matter how enthusiastic they are about the work, have limited reserves of energy. Long disuse of their language, which is often the case when the few remaining speakers live in different households with people who don’t speak their heritage language, means that it will take them time and effort to remember the words and forms they are asked about. Casual conversation may be easy; remembering words that they last heard from grandparents sixty years ago will be harder. Even for younger people, concentrating hard for six or seven hours is exhausting. So it is necessary, especially when working with elderly consultants, to give them regular five- or ten-minute breaks, at least one in the morning and one in the afternoon, plus an hour or more for lunch. One of the advantages of working with a group of elders rather than with one at a time is that a consultant can leave the room for a few minutes without stopping the work. The other main advantage, I have found, is that the linguist will have a chance to hear casual conversations among the elders – and the elders themselves may also welcome the opportunity to chat in their native language during a day when they are thinking constantly about that language. In my visits to the Flathead Reservation for fieldwork, the atmosphere around the big table where we work is always relaxed. The work may be interrupted briefly if an elder thinks of something to mention to the other consultants; short conversations in Salish-Pend d’Oreille occur frequently. Sometimes a visitor will come into the room to be introduced to the elders. We laugh a lot while we’re working (usually at my mistakes). Setting a leisurely pace serves three purposes: it helps the elders enjoy the day (so that they’ll continue to be willing to attend the sessions); it keeps them and me from getting too tired too fast; and it also helps them think of new words and other language material that will move the project forward. The official start of our one-day session is at 9:00 AM, and the official end is 4:00 PM. With delays in getting started, breaks, lunch, and miscellaneous interruptions, the actual time spent in gathering data is about five hours. Because fieldworkers routinely record all field sessions on audio and/or video devices, one might think that transcribing the data by hand during the session would be unnecessary. Not so: even if the recording devices work perfectly (and they often don’t), there are two excellent reasons to transcribe all the material by hand as the work proceeds. First, handwritten notes permit the linguist to make extra comments on setting, context, elders’ reactions, connections between a piece of data and something recorded weeks earlier, and even tentative analyses – all valuable information that won’t be part of the recording. And second, writing things down slows the pace of the work and helps to create a relaxed atmosphere. It is of course possible, and is in fact probably the most common fieldwork mode, to work with one consultant at a time. For some purposes this is preferable,

137

138

field research on endangered languages

but it may not always be ideal with a severely endangered language. It takes some time each day to establish a steady rhythm of the work, so switching from one consultant to another during the day would inevitably cause loss of working time; and working with a single consultant for an entire day might be impossible because of limited stamina – all the pressure to perform, to provide data, would rest on the one elderly person, and some speakers would be uncomfortable with all-day work. The last topic under the field sessions heading is the selection of data collection techniques. A successful fieldwork project must employ a variety of techniques for collecting data; no single method will work for all purposes. It is especially important to vary the techniques when the target language is endangered, because of the need to jog consultants’ memories of aspects of their language that they may not have used or even heard for a very long time. The primary method to use at the beginning of a project, and often throughout the entire course of the project, is direct elicitation of material – as mentioned earlier, words, paradigms, and simple sentences at first. The prompt for direct elicitation is always something like ‘How would you say X in your language?’ Elicitation provides the basics of the grammar and some vocabulary words to put into paradigms and sentences, and it enables the linguist to build on these to elicit more complicated structures. A list of basic vocabulary permits the discovery of the language’s phonemic system and the phonetic variants (allophones) of the phonemes, though some phonological features (especially prosodic features like sentence intonation) can only be established by examining connected speech. Elicitation is also the speediest way to collect a large amount of data. But elicitation can be problematic too, because (except in a monolingual field research project) it makes a bilingual consultant translate from the language in which the linguist and the consultant are communicating into the consultant’s native language. The danger of translation is that the consultant might accommodate his or her native speech, unconsciously or otherwise, to the language of communication. Not all consultants will do this, certainly, and it may be unusual for a consultant to match patterns in the language of communication so closely as to produce ungrammatical utterances in the native language. But given the inherent variability of linguistic structure, bilingual consultants who accommodate their speech to the language they are translating from may provide utterances that are quite unnatural in the context. My favorite example is from my Salish-Pend d’Oreille research, from a session when I was eliciting ditransitive sentences – sentences with two object nouns, for instance, ‘Johnny stole a (deer) hide (direct object) from Mary (indirect object)’, as in example (1) on the next page. On that particular occasion I was working with just one consultant; if I’d been working with the usual group, I might well have received different translations. But on this occasion I was repeatedly given Salish-Pend d’Oreille translations of the type in the example, where the top line has the Salish-Pend d’Oreille sentence and the second line has the literal English meaning for each word aligned under the Salish-Pend

Investigating gravely endangered languages in the field

d’Oreille word. (Readers may wonder why I use the names ‘Johnny’ and ‘Mary’ in this example, instead of indigenous names or names of mythical characters. The reason is that these are the Salish-Pend d’Oreille names; the people – at least the men – also have indigenous names, but they don’t seem to use them in public, and they don’t talk about them much. The vast majority of the personal names are of French origin – ‘Mary’ is from French Marie rather than from English Mary, as indicated by the fact that it is accented on the second syllable, as in French, and not on the first, as in English. ‘Johnny’ is an exception, as it is strictly English; when speaking Salish-Pend d’Oreille, the elders call most men named ‘John’ Šan, from French Jean ‘John’. And the reason I don’t use names of mythical characters is that I do my fieldwork in the summer, and it is not permitted to talk about myths or mythical characters in the summer.) ˇ (1) Coní naqw ’ t q’ett tl’ Malí Johnny steal obl hide from Mary Other than the oblique particle t and the vocabulary items, this sentence is very close to the structure of the English version: Subject-Verb-Object(s) word order, the verb in an intransitive bare root form – that is, with no suffixes – and a preposition tl’ ‘from’ marking the indirect object. (In this language, the particle t ‘oblique’ obligatorily marks the object of an intransitive verb form or the subject of a transitive verb form, and it has no English translation.) The sentence is completely grammatical in Salish-Pend d’Oreille; such a construction could occur in natural speech, but never outside a discourse. It is very far from what would be expected as a natural Salish sentence in a neutral context – as in the present case, where there is no context at all outside the sentence itself – where speakers would use Verb-Object-Subject word order and a transitive verb with all the suffixes that are required on a transitive verb form. Subjects normally appear at the beginning of a sentence only when they receive special contextual emphasis (comparable to the emphasis in an English sentence like It was Johnny who stole the hide from Mary); and an intransitive verb with an object (much less with both a direct object and an indirect object) is typically used only in particular limited discourse contexts. In isolation, these sentences were bizarre. When I finally asked Joe, ‘Aren’t these sentences a bit, um, English-y?’, he was surprised. ‘Yes, of course they are,’ he said. ‘You were asking in English, so I thought that’s what you wanted.’ ‘Ah,’ I said. ‘So how would you say these things if you were just talking normally in your language?’ ‘Oh, well, then I’d say it like this’: (2)

Naqw’-m-ë-t-s steal-der.trans-rel-trans-he

Malí Mary

q’ett-s hide-3sg.poss

t particle

ˇ Coní Johnny

This sentence has the expected complex ditransitive verb form (der.trans = ‘derived transitive’, a stem-forming suffix; rel = ‘relational = second object’) and VOS word order. The English translation is slightly different too: this version means ‘Johnny stole Mary’s hide’. This sentence is typical for ditransitive

139

140

field research on endangered languages

constructions with no special emphasis on subject or object. In offering examples like the odd sentence in (1), this elder was deliberately accommodating his speech to English structure and thereby producing Salish-Pend d’Oreille sentences which, though fully grammatical, were extremely strange outside the appropriate context. Admittedly, this is an extreme example. But the danger of accommodation is real when one is working with bilingual speakers, and less extreme examples can be hard to spot. Other reasons for using methods other than elicitation of isolated utterances are as important, or even more so. With a gravely endangered language, some grammatical features that might never have been particularly common are buried so deep in consultants’ memory that elicitation can’t pull them out. A problem that I once encountered in my Salish-Pend d’Oreille research will illustrate this point. I had great difficulty in eliciting sentences in which a first-person plural (‘we’) subject acted on a second-person plural (‘you, you all’) object, and vice versa. (I guessed that this might have been because it had been so long since enough speakers gathered together that the forms had fallen out of use; but that’s only a guess.) I finally gave up trying to elicit the forms in isolated sentences and constructed a humorous story about two young men who were courting two young women. I put dialogue into the story so that there would be first plural and second plural pronouns, and then elicited the story through sentence-bysentence translation. This was just one short step removed from eliciting isolated sentences: it was still elicitation, but with context to set the stage for the forms I needed. It worked well; I got the forms. The next step after eliciting stories sentence by sentence is to elicit stories by prompting the speakers with a simple ‘And then what happens next?’ I’ve found this most helpful with descriptions of traditional cultural practices, like preparing a deer hide or baking camas (an onion-like staple in the traditional Salish-Pend d’Oreille diet). In this way I could collect an entire text, and also a series of natural sentences – one by one, but in a fairly natural context – that informed the subsequent analysis and, when I was lucky, contributed structural features that I had not previously heard. The gold standard of data gathering is naturally occurring text, including stories (myths, history, etc.), oratory, negotiating, joking, conversations, and other genres. The last few speakers of an endangered language might not include any storytellers, and if there is a storyteller s/he might be out of practice and unwilling to perform. The conversations I’ve heard during my field sessions with the Salish-Pend d’Oreille elders are brief exchanges, not extended talk. For these reasons, the linguist might have to be content with texts elicited sentence by sentence, as just described. But in most situations, it will be possible to collect longer narratives, and until about fifteen years ago my consultants included elders who would tell stories. Narratives can be collected simply by asking the elders to tell stories, or by asking them to describe traditional matters like

Investigating gravely endangered languages in the field

marriage customs in their youth, or by such techniques as showing them picture books without words or silent films and asking them to tell the stories. It is important to understand, however, that naturally occurring text cannot be expected to replace elicitation entirely as a method of data collection. The main reason is that it would take too long, and require much too much textual material, to be confident that one has uncovered most of the language’s main structural features. Some features occur rarely in natural speech; full paradigms of most verbs, for instance, might never emerge if the linguist’s data is confined to natural speech. 6.3.3

Individual variation

Individual speakers’ differences are not likely to create major analytic difficulties in a field project on a non-endangered language, because in such cases it will (with luck) be fairly easy to distinguish a single speaker’s idiosyncrasies from structured variation within the speech community. The situation is quite different when a speech community is so diminished that only a few fluent speakers remain. Of course variation is universal in all human language, and its sources are diverse: besides individual differences, there are dialect differences of various kinds and also register (e.g., formal and informal) differences within a single person’s speech. A given speaker’s judgment might vary from one field session to the next about the acceptability of a particular utterance. But dialect differences and individual differences may be indistinguishable if the number of fluent speakers is very small, and finding patterns of variation is also likely to be much harder in a severely endangered language. In analyzing any language, it is not acceptable to deal with variation by sweeping it under the rug. But it is impossible to analyze a language’s structure without abstracting away from at least some of the variation in order to focus on the features shared by all the speakers one works with. Both of these problems are easier to identify than to solve. Asking speakers for grammaticality judgments is often a very useful tool in linguistic fieldwork, but what if the only two remaining speakers disagree about whether an utterance is grammatical or not? Ultimately, the solution is to keep the individual differences in one’s database and mention them, along with the speakers’ differing grammaticality judgments, when presenting a variable piece of structure in a linguistic description. The question of “correct” vs. “incorrect” utterances tends to be problematic in any situation involving linguistic variation, in both endangered and non-endangered languages; in endangered languages especially, it is sometimes best not to try to make the distinction. The problem of individual variation does raise an ethical issue that can be troublesome, at least for fieldworkers based in U.S. universities – namely, the anonymity that IRBs normally require prospective fieldworkers to guarantee to their “subjects”. This requirement is in direct conflict with the custom in some

141

142

field research on endangered languages

communities of personal ownership of stories and songs, which dictates that the owners must be identified if the story or song is published. Moreover, in the last stages of a language’s life, it may be vital to know which consultant provided which form. One consultant’s utterances may fit together in a fairly neat system but be incompatible with utterances provided by another consultant. Attaching names, or at least initials, to each utterance will enable the fieldworker and future scholars to determine which differences result from interspeaker variation and which from intraspeaker variation. And because some consultants are undeniably more reliable than others are as sources of linguistic information, being able to identify the most reliable speakers is also important. Using identifying initials that don’t actually reveal the consultant’s personal identity is a possibility, but concealing the identity can obscure family relationships that could also help linguists find patterns in the variation. Nor is the fieldworker’s IRB the only concerned observer: some consultants may be willing or even eager to be thanked by name for their contributions to a documentation project, but others may wish to remain anonymous. If a consultant has died before publication of the results of the research, surviving family members may have views about whether their deceased relative should be identified by name – and the views of different family members may conflict, raising a different set of issues. I have no solution to offer to any of this, however, so this paragraph is meant primarily as a warning to prospective fieldworkers to be careful about getting all necessary permissions before identifying consultants in print. 6.3.4

An extended example: dictionary-making

A vital part of a primary documentation project is the preparation of a dictionary of the language. Particular difficulties arise with an endangered language because of the need for speakers to remind themselves of words they haven’t heard for many years – in some cases, as the Salish and Pend d’Oreille elders I work with tell me, words that they themselves never used, but that they used to hear from their parents or grandparents. Some excellent techniques for expanding a dictionary file rapidly, such as asking groups of consultants to brainstorm together using lists of semantic domains, cannot be employed with a language so endangered that its remaining few speakers are elderly and no longer accustomed to speaking their language regularly. How many words are “enough” for a dictionary of an endangered (or other) language? There can of course be no definitive answer to this question, but here are two estimates. Michael Krauss has said that 6,000 words are sufficient for a moribund language – that is, a language that is no longer being learned as a first language by children – and 14,000 words for a non-moribund language; Terrence Kaufman suggests 4,000–6,000 words as a minimum dictionary size for a documentation project, and 10,000 words as a good place to stop. No matter how many words you collect, however, a truly complete dictionary is not a realistic goal.

Investigating gravely endangered languages in the field

In the very first stages of the fieldwork, the linguist will elicit short(ish) words from a list of basic vocabulary, typically a 100- or 200-word Swadesh list – one of the two lists of basic vocabulary compiled in the mid-twentieth century by the linguist Morris Swadesh, who used them in his efforts to establish genetic relationships among languages. Because these lists were designed to include only words that would be expected to occur in every language (and therefore less likely to be borrowed), the words on the lists tend to be relatively short. In later stages of the project, the linguist will be able to collect words from both elicited and, if there are any, naturally occurring texts. But it will also be necessary to use techniques that are specifically designed to produce additional words for the dictionary, including words for culturally relevant terms and concepts. (Here’s a strategy that will not be used: the linguist will not start working through a monolingual dictionary of English, which begins with words like a, aardvark, aardwolf, ....) A warning: one shouldn’t stick rigidly to a 200-word Swadesh list. Some of the words on it, like at, won’t have straightforward translations in the target language, and it’s important not to risk making consultants uncomfortable by asking for a word they don’t have or don’t know. Other English words might have more than one translation: in that case the linguist should naturally get all the words the consultants can think of. In Salish-Pend d’Oreille, for instance, asking for a translation of brother will produce two words: sínceP ‘younger brother’ and qéws ‘older brother’; as we saw in Chapter 4, the language has no generic term for ‘brother’. Many English words that aren’t on the Swadesh list will also make consultants think of different words, and they’ll be useful additions to the wordlist even if they aren’t exact synonyms. An example: Salish-Pend d’Oreille has several words for ‘horse’, among them snˇcëc’á(Psqa) (literally, roughly, ‘domestic elk’) and x.ň’cín (literally, roughly, ‘bite (grass, etc.)’). But when they start thinking about words for horses, consultants might also volunteer words like kw laqín ‘strawberry roan’, cˇ pí ‘palomino horse’, i kw íl ‘bay horse’, and so forth. All these will be welcome. Organizing lexical elicitation by semantic fields is also useful, for example body parts, kin terms, animal and plant names (pictures help with these), ways of walking (here the linguist demonstrates striding, staggering, waddling, etc.), color terms (using a color chart), traditional clothing items, and, in relevant cultures, parts of a tipi. But in an endangered language, some semantic domains are already likely to be partly lost, no matter how hard the consultants try to remember them; they may never have been part of the consultants’ experience, or the words may simply be beyond the reach of current memory. Another valuable technique is having the consultants define and explain words in their own language instead of simply providing English translations. This technique will produce monolingual dictionary entries, which will inevitably add more new words and additional cultural information; but the technique is difficult to use at the beginning of a fieldwork project, because the linguist will not yet

143

144

field research on endangered languages

know enough of the target language to understand the definitions well, and explanations would take up time unnecessarily. Direct elicitation can produce culturally and linguistically interesting items. Salish-Pend d’Oreille, for example, has several different sound-symbolic formations, and the most interesting one has triple reduplication of the second root consonant, as we saw in Chapter 5. Two additional examples are i p’átttt’ ‘sound of a cow-pie plopping’ (from the root p’át’) and i kw íˇccˇ cˇ cˇ ’ ‘the creaking sound a tree makes when it’s starting to fall’ (root: kw íˇc’). Asking for words for additional items in a particular category, in this case words for sounds, can help expand the dictionary (as in ‘Do you have a word for the sound of a ruffed grouse drumming?’). For eliciting words, open-ended follow-up questions are a good idea. If the topic is sound-symbolic words, for instance, one can ask whether there are other words for sounds or sights that haven’t yet been asked for; and with luck the elders will remember some. That’s how I collected i méllll, which the elders defined as ‘a bunch of things going in and out of vision, like when you see running horses through a picket fence’. As noted earlier, a dictionary project requires as much textual material as the consultants can provide in the time available. There will inevitably be many, many words that won’t emerge from direct elicitation, because the linguist won’t think to ask for them and the consultants won’t happen to think of them out of context. This means, of course, that one can’t compile an adequate dictionary just by eliciting words; it’s equally important to collect texts of as many kinds as possible. Waiting for words to occur in texts (especially if the consultants don’t readily provide texts) is not an efficient way of building a dictionary rapidly, however, so one should also make full use of published sources on the same language, on related languages, and even on nearby unrelated languages, because they are likely to share some of the same natural-world and cultural features. For my work on Salish-Pend d’Oreille, important sources have been dictionaries of Spokane (a dialect of the same unnamed language as Salish-Pend d’Oreille) and ColvilleOkanagan (a Southern Interior Salishan language that is closely related to SalishPend d’Oreille and Spokane). My most important published source by far is the nineteenth-century Dictionary of the Kalispel or Flat-head Indian Language, compiled by Gregory Mengarini and other Jesuit missionaries and published in 1877–1879. (Kalispel, spoken in eastern Washington state, is another dialect of the same language as Salish-Pend d’Oreille; Flathead is the name given to the Salish tribe of southwestern Montana by Whites, for mysterious reasons. This tribe, also called Bitterroot Salish after the traditional homeland south of Missoula, Montana, is half of the Salish-Pend d’Oreille tribal organization on the Flathead Reservation.) The monumental Jesuit dictionary comprises two volumes and a total of 1,000 pages, 644 in the “Kalispel-English” volume and 456 in the “English-Kalispel”

Summary and commentary

volume. The former contains more than 700 main entries, many of them with dozens of subentries. The Jesuits’ dictionary is hard to use, in part because of underdifferentiated orthography – the letter sequence ko, for instance, is used to represent the phonemes /kw ’/, /qw /, and /qw ’/. Some forms in the dictionary are now very rare or are not recognized at all by the current elders; and there are surely some outright errors as well, although the overall quality of the dictionary is impressive. These features make it desirable to re-elicit as much of the material in the dictionary as possible, and I’ve found that this is an excellent way of generating more words. Similar procedures (minus the orthographic difficulties, with luck) can help in dictionary-making with other endangered languages too. Here are a few examples of ways in which re-elicitation can add words to the dictionary. I might ask, ‘The Fathers’ dictionary has a word chin-chem-echst ‘I join my hands together to receive something’. Do you have that word? No? Then how would ˇ t’íx.cˇ stm, like, you open your hand to you say that?’ The elders’ response: ‘Cn receive something.’ Or I ask, ‘The Fathers have es-chs-chisti “he’s exploring”. Do you have that?’ Them: ‘Yes, esˇc’sˇc’ísti, but it means “he goes into enemy country to scout”.’ Whatever source one is using, the elders should be encouraged to free-associate. Me: ‘The Fathers have es-chet-us “eyes could be screened off”; do you have that one?’ Them: ‘No, but there’s nsˇc’etús “the opponent, like in a stick-game”.’ Or this: Me: ‘The Fathers have n-pe-us “narrow hole, as of a needle”.’ Them: ‘No, we say esnp’ePús “eye of a needle”. And then there’s esp’Púps “flat ass”.’ (And then they giggle, because this means that someone’s behind is flat rather than nicely rounded.) Creative use of published sources thus helps the fieldworker to expand dictionary files quickly, while enlarging his/her stock of knowledge of the culture that the language expresses.

6.4

Summary and commentary

This chapter began with a brief characterization of documentary and descriptive linguistics and an overview of linguistic fieldwork in general, together with a consideration of three particular questions that arise in any fieldwork project: what counts as fieldwork, how much time should a linguist spend in the field, and will the emotional stresses of fieldwork be greater than the emotional and intellectual benefits. None of these questions has a simple answer; all of them are things to think about before undertaking linguistic fieldwork. The bulk of the chapter then focused on fieldwork aimed specifically at describing and/or documenting extremely endangered languages – gaining permission to work with the last few fluent speakers, collaborating whenever

145

146

field research on endangered languages

possible, learning the target language (or not), working with a group of elders rather than with just one at a time, pacing sessions carefully in order to avoid tiring elderly speakers too much, using a variety of data-gathering techniques, and the problem of individual variation. The chapter concluded with a longer discussion of dictionary-making, a task that should be part of any primary description or documentation project. Much more could be said, of course, about how to investigate an endangered language in the field, on these and other topics. Describing an endangered language, and documenting it in the broader sense discussed at the beginning of this chapter, can preserve linguistic and cultural knowledge that would otherwise be lost forever. This fact alone makes such a project worthwhile – all the hard work, all the incidental frustrations that are an inevitable part of fieldwork. All fieldworkers must keep in mind the fact that there are two audiences for their research, the speech community and the world of academic scholarship. I won’t try to argue for the greater importance of one of these audiences over the other, but it is imperative – as everyone, I believe, now recognizes – for the fieldworker to make every effort to satisfy the speech community as well as academia (although it is sometimes much easier to understand what academia requires than to discover what the speech community wants). The research results must be accessible to the community, not just to other linguists; and what happens to the data that is gathered and the analyses that are produced by the linguist is ultimately the community’s decision, not the linguist’s.

6.5

Sources and further readings

The 1992 article cluster ‘Endangered languages’ – which, in many linguists’ view (as noted previously), brought endangered languages to the forefront of linguists’ attention, at least in the United States – is divided into essays written by Ken Hale, Michael Krauss, Lucille J. Watahomigie and Akira Y. Yamamoto, Colette Craig, Laverne Masayesva Jeanne, and Nora C. England. The quotation at the beginning of the chapter is from p. 10 of Michael Krauss’s contribution to this set of articles, ‘The world’s languages in crisis’. The two quotations from The Wichita Eagle’s obituary of the linguist Robert Rankin (1939–2014) illustrate the enormous potential importance of documenting an endangered language – not only, or primarily, for the sake of science, but for the community’s sake. In the mid-1970s, when Rankin began doing fieldwork on Kaw (a Siouan language, also known as Kansa – from which the state name Kansas is taken – or Kanza), there were, according to the obituary, only four speakers of the language left. As mentioned earlier, the 2006 book Essentials of language documentation, edited by Jost Gippert, Nikolaus Himmelmann, and Ulrike Mosel, gives a comprehensive view of what is involved in a documentation project. For

Sources and further readings

excellent brief overviews of what a full-scale documentation project encompasses, see Anthony Woodbury’s 2011 article ‘Language documentation’ – the definition of language documentation quoted above is Woodbury’s opening sentence (p.159) – and Nicholas Evans’s 2008 review of the Gippert et al. book. On pp. 342–343 of his review Evans warns that the growing number of specialists that one “‘needs to be” in order to be a documentary linguist’ is so daunting that the prospect of conducting linguistic fieldwork, especially alone, might seem too discouraging to attempt. He suggests that it would be wise to pay heed to ‘Voltaire’s famous advice that le mieux est l’ennemi du bien – the best is the enemy of the good – to avoid scaring too many readers away with the feeling that they could never manage to satisfy all the “need to bes” set out in this book.’ Evans’s argument in favor of preparing and publishing descriptions of endangered languages is on pp. 346–348 of his review of Gippert et al. His reference to Colette Grinevald Craig (now Colette Grinevald) is to her 2001 article ‘Encounters at the brink: linguistic fieldwork among speakers of endangered languages’; and, in the same quotation, ‘the classic Boasian trilogy’ refers to the set of materials advocated by Franz Boas, one of the founders of anthropological linguistics (and, some say, of modern anthropology as a whole), for a major study of a language. Other good sources on documentary linguistics are Nikolaus P. Himmelmann’s early article ‘Documentary and descriptive linguistics’ (1998) and Peter K. Austin’s pedagogically oriented 2008 article ‘Training for language documentation: experiences at the School of Oriental and African Studies’. The two highlighted features of documentary linguistics, an interdisciplinary team of investigators and collaborative work with community members, are listed in Himmelmann’s 2006 article ‘Language documentation: what is it and what is it good for?’ (cited on p. 26 of Austin’s 2008 article). The quotation about the origin of endangered-language research in Europe is from p. 339 of Nikolaus Himmelmann’s 2008 article ‘Reproduction and preservation of linguistic knowledge: linguistics’ response to language endangerment’. Alert readers will have noticed that, while Himmelmann credits Bechert’s conference paper as the stimulus for the establishment of endangered languages as a major topic of concern in mainstream linguistics, I credit Hale et al.’s 1992 set of articles as the main stimulus in (at least) the United States. Bechert’s paper was apparently more important in Europe, but Hale et al. had a greater impact in the U.S., in effect building on Nancy Dorian’s earlier research on language endangerment. It is possible, of course, that some American linguists signed Lehmann’s motion at the 1987 International Congress of Linguists and brought the news back to the U.S., and that that news helped inspire Ken Hale to organize his 1991 endangered-languages symposium at the Linguistic Society of America annual meeting – the symposium that led to the 1992 publication of Hale et al. But as far as I know, the symposium and subsequent publication of Hale et al. was independent of Bechert’s and Lehmann’s call for linguists to take up arms in defense of endangered languages. In any case, many linguists’ sense of urgency

147

148

field research on endangered languages

about the description and preservation of endangered languages became acute in a number of countries around the same period in the late 1980s and early 1990s. The list of general books about field linguistics in an early section of this chapter focuses on newer publications. But although it is outdated in some respects, William Samarin’s classic and pioneering 1967 textbook Field linguistics: a guide to linguistic fieldwork is still well worth reading. Another good source of information about linguistic fieldwork is Keren Rice’s annotated bibliography ‘Fieldwork’, published online by Oxford University Press in 2011. A source cited in the further readings section of Chapter 1 should be mentioned here too: the open-access electronic journal Language Documentation and Conservation (nflrc.hawaii.edu/ldc/) publishes articles on all aspects of language documentation, especially endangered languages. In addition, important articles have been devoted to the topic; a good recent article is Anthony Jukes’ ‘Researcher training and capacity development in language documentation’ – which, as the title suggests, focuses on the needs of documentary linguistics rather than solely on field methods. A 2008 documentary film about fieldwork on endangered languages, ‘The Linguists’, shows Greg Anderson and David Harrison doing fieldwork in several communities around the world. It’s very enjoyable and definitely worth watching, and you can see a trailer on the film’s official website, www.thelinguists.com (accessed 28 May 2014). The film paints an unrealistic picture of what linguistic fieldwork is like, however: it makes fieldwork look rather easy, and above all quick. But of course any documentary that attempted to depict the more difficult and slower aspects of fieldwork would be unwatchable, and this film has considerable value as a way of helping to educate the general public about the importance of documenting endangered languages while it is still possible to do so. Another good fieldwork film is the 2007 video ‘Khinalug 2007: creating a digital portrait of an endangered language’, about Aleksandr E. Kibrik’s fieldwork on Khinalug, a northeastern Caucasian language. This video is in Russian with English subtitles, and it can be found at www.youtube.com/watch?v=yZBUXNddzT8 (accessed 28 May 2014). There is now a sizable literature on the subject of linguistic fieldwork ethics. In addition to the coverage in the books already listed, see for example Keren Rice’s 2006 article ‘Ethical issues in linguistic fieldwork: an overview’; Arienne Dwyer’s 2006 article ‘Ethics and practicalities of cooperative fieldwork and analysis’ in the Gippert et al. volume; and Nancy Dorian’s 2010 article ‘Documentation and responsibility’. The point about preparing for fieldwork by becoming familiar with the relevant ethnographic literature is made eloquently by Lise M. Dobrin in her 2008 article ‘From linguistic elicitation to eliciting the linguist: lessons in community empowerment from Melanesia’. In the list of related topics covered by different authors writing about fieldwork, monolingual fieldwork is discussed by Daniel Everett in his article in the 2001 Newman and Ratliff book. Ethnobotany is the topic of Barry J. Conn’s article in the Thieberger volume, and toponymy and kinship systems are covered

Sources and further readings

in articles in the same book, by David Nash and Jane Simpson and by Laurent Dousset, respectively. Team-based fieldwork and political aspects of fieldwork are discussed by several authors each, in particular in the 2010 Grenoble and Furbee volume. Daniel Everett’s definition of fieldwork is on p. 168 of his 2001 article ‘Monolingual field research’ in the Newman and Ratliff book. I am grateful to Carmel O’Shannessy for permission to report the details of the syllabus for her Winter Term 2013 field methods course at the University of Michigan. See the glossary for brief descriptions of the software tools listed in the passage about her class; the quotation about technological tools is taken from her class syllabus. The suggestion about visiting a diaspora community where people regularly speak the language being studied in a field methods class comes from Peter Austin (personal communication, 2014). To sample Carmel O’Shannessy’s published work on Light Warlpiri, see her 2013 article ‘The role of multiple sources in the formation of an innovative auxiliary category in Light Warlpiri, a new Australian mixed language’. For non-technical descriptions of her research, google “Light Warlpiri” to find 2013 news accounts, one of which is here: www.nytimes.com/2013/07/16/science/ linguist-finds-a-language-in-its-infancy.html?_r=0. The only other case I know of where the emergence of a new mixed language has been observed as it happened is Gurindji Kriol (Northern Territory, Australia), which has been studied since the 1970s by Patrick McConvell and extensively documented in recent years by McConvell and Felicity Meakins, for instance in their 2005 article ‘Gurindji Kriol: a mixed language emerges from code-switching’, and in Felicity Meakins’s 2012 article ‘Which mix? Code-switching or a mixed language? – Gurindji Kriol’. Daniel Everett’s article that triggered the controversy over his anti-Chomsky conclusions is ‘Cultural constraints on grammar and cognition in Pirahã: another look at the design features of human language’ (2005). For one of the numerous media accounts of the controversy, see John Colapinto’s article ‘The interpreter: has a remote Amazonian tribe upended our understanding of language?’ in the April 16, 2007, issue of The New Yorker, available electronically here: www.newyorker.com/reporting/2007/04/16/070416fa_fact_colapinto. One exchange in the scholarly debate about Everett’s proposals appeared in the journal Language in 2009: see the criticisms of Everett’s analyses by Andrew Nevins et al., ‘Pirahã exceptionalism: a reassessment’, and Everett’s reply, ‘Cultural constraints on grammar in Pirahã: a response to some criticisms’. The two quotations about Nancy Dorian’s fieldwork experiences are from p. 133 and p. 149, respectively, of her article ‘Surprises in Sutherland: linguistic variability amidst social uniformity’, in the Newman and Ratliff volume. The author who describes an extreme emotional reaction to the (mostly nonphysical) difficulties of fieldwork is Monica Macaulay in her article ‘Training linguistics students for the realities of fieldwork’ in the 2012 volume edited by Thieberger. On p. 470 she quotes the anthropologist John Wengle’s comment that

149

150

field research on endangered languages

‘one characteristic does seem to stand out in all the accounts of fieldwork that I am familiar with. I am referring to the associated problems of disorientation, worry, depression, fatigue, loneliness, stress, and the like’ (from p. xviii of his 1988 book Ethnographers in the field: the psychology of research). Another article that discusses related issues (and others) is Paul Newman’s 2009 paper ‘Fieldwork and field methods in linguistics’. For a recent discussion of field linguists’ experiences with their Institutional Review Boards (IRBs), see Claire Bowern’s 2010 article ‘Fieldwork and the IRB: a snapshot’. The IPA is the International Phonetic Alphabet, which was used as the basis for the orthography adopted by the Salish-Pend d’Oreille Culture Committee (then called the Flathead Culture Committee) as the standard for writing their language. Some of the material in this section is drawn from my 2013 article ‘How to avoid pitfalls in documenting endangered languages’, in Elena Mihas et al., eds., Responses to language endangerment: in honor of Mickey Noonan. The subsection on dictionary-making is nearly identical to the corresponding section in the 2013 article, as is the ‘Summary’ section, and these sections are included here with the permission of the book’s publisher, John Benjamins Publishing Company; they are based on a 2010 talk I gave on ‘Recovering and preserving linguistic knowledge: making a dictionary of an endangered language’. To sample the growing literature on collaborative research that involves community members in the research itself, see for instance these four articles: Raquel Yamada’s ‘Collaborative linguistic fieldwork: practical application of the empowerment model’ (2007); Susan Penfield et al.’s ‘Community collaborations: best practices for North American indigenous language documentation’ (2008); Jocelyn Ahlers’s ‘The many meanings of collaboration: fieldwork with the Elem Pomo’ (2009); and Ewa Czaykowska-Higgins’s ‘Research models, community engagement, and linguistic fieldwork: reflections on working within Canadian indigenous communities’ (2009). For further reading about issues specific to documenting endangered languages, see two papers in Sakiyama and Endo, eds., Lectures on endangered languages: 2 (2001): Terrence Kaufman’s ‘Two highly effective models for large-scale documentation of endangered languages’ and Colette Grinevald’s ‘Encounters at the brink: linguistic fieldwork among speakers of endangered languages’. Kaufman’s article is the main source of the passage about the “parallel project” approach to language documentation; the quotations are from p. 271 and p. 273, respectively. The information about Russian fieldwork expeditions using a similar model comes from Peter Austin (personal communication, 2014). Nancy Dorian also describes the experience of watching Kenneth Pike give a monolingual demonstration, on p. 134 of her 2001 paper ‘Surprises in Sutherland: linguistic variability amidst social uniformity’. For an engaging account of learning a language monolingually, see Daniel Everett’s 2008 book Don’t sleep, there are snakes: life and language in the Amazonian jungle; Everett has also

Sources and further readings

done monolingual demonstrations, including one at the 2013 Linguistic Institute of the Linguistic Society of America at the University of Michigan – which can be viewed at www.youtube.com/watch?v=sYpWp7g7XWU&noredirect=1 The first two quotations about the importance of learning to speak the language one is investigating in the field are from p. 4 and p. 6 of Paul Newman and Martha Ratliff’s Introduction to their 2001 edited volume. The quotation from Peter Sutton, about additional benefits of speaking the target language, is from p. 461 of his 2001 article ‘Talking language’, as cited on p. 318 of Lise M. Dobrin’s 2008 article ‘From linguistic elicitation to eliciting the linguist: lessons in community empowerment from Melanesia’. I’m sorry to say that I can’t credit the audience member who reported, in a comment after a 2011 conference talk I’d given, how he lost the trust of an indigenous community in Mexico when he learned to speak their language, because I didn’t recognize him when he spoke and don’t know his name. I am grateful to Lucy Thomason for the insight that endangered-language speech communities usually have to settle for whatever linguist appears in their midst, with an agenda of her/his own, instead of seeking out the ideal linguist to serve the community’s purpose – a counterpoint to the perspective of the linguist who must work with the few remaining fluent speakers of a language because seeking the ideal consultant is not an option. On the value of varying data collection techniques, see (in addition to the field linguistics textbooks listed at the beginning of this chapter) Marianne Mithun’s 2001 article ‘Who shapes the record: the speakers and the linguist’, pp. 34-48, in the Newman and Ratliff volume. Two stimuli that have long been popular for eliciting narratives are the frog stories, a series of wordless children’s books by Mercer Mayer, and the pear story, a silent film involving a man picking pears and a boy who steals one of the pear-filled baskets while the man isn’t looking: www.youtube.com/watch?v=bRNSTxTpG7U (This film may be less than ideal for field settings that don’t have fruit trees.) Nancy Dorian’s work on individual variation, based on her work with Scottish Gaelic, is the best place to start exploring this topic. See, for instance, her 2010 book Investigating variation: the effects of social organization and social setting and her 1994 article ‘Varieties of variation in a very small place: social homogeneity, prestige norms, and linguistic variation’. I am grateful to Peter Austin for reminding me about situations in which individual owners of stories and songs expect to be identified with a story or song when it is published. The technique of eliciting lots of words rapidly by having consultants work together from lists of semantic domains is described by Ron Moe in his 2001 article ‘Lexicography and mass production’. Typical semantic domains might be bird and animal names, kinship terms, medicinal plants, daily activities, and the like. See the glossary entry Basic vocabulary for more details about Swadesh lists. I am grateful to Peter Austin (personal communication, 2014) for suggesting

151

152

field research on endangered languages

collecting dictionary entries by asking consultants to explain terms in their own language rather than simply translating them into English. Michael Krauss gave his estimates of the number of words needed for a dictionary of an endangered language in a personal communication to me in 2001; Terrence Kaufman’s estimates are from p. 275 of his 2001 article ‘Two highly effective models for large-scale documentation of endangered languages’. The dictionary of Spokane that I’ve made extensive use of for my Salish-Pend d’Oreille dictionary project is Carlson and Flett’s 1989 Spokane dictionary, and Mattina’s 1987 Colville-Okanagan dictionary is my source on that language. Peter Austin comments (personal communication, 2014) that satisfying the speech community is by no means always a straightforward task. Measuring community satisfaction is a controversial matter; for instance, what if some community members want X and others want Y or even Not-X? The fieldworker’s responsibility is to do his or her best to do right by the community. Omniscience is not required.

7

Revitalizing endangered languages

Ceeki aweeya iilaataweeciki kaakisiitootaawi! Let’s maintain everybody’s language! (Wesley Y. Leonard, Miami Tribe of Oklahoma Language Committee, 2010) The title of this volume is an answer to another publication, the UNESCO Red Book of Endangered Languages. . . . We hope this Green Book will be of use to everyone who wants it to be no longer necessary for their language to be listed in the Red Book. (Leanne Hinton and Ken Hale, 2001:xi) Preservation...is what we do to berries in jam jars and salmon in cans. Preserved foods are different from thriving berry patches and surging runs of salmon, and dictionaries are not the same as speech. Books and recordings can preserve languages, but only people and communities can keep them alive. (Nancy Lord, 1996:68, citing Tlingit oral historians Nora Marks Dauenhauer and Richard Dauenhauer)

Someone who reads only the first six chapters of this book is likely to come away with a picture of unrelieved gloom as far as the future prospects for endangered languages are concerned. Chapter 1 surveyed the general topic of language endangerment and introduced some key concepts, none of them tending to generate optimism about the fate of endangered languages. That chapter also introduced UNESCO’s Atlas of the world’s languages in danger, which (according to its home page as of 24 May 2014) classifies 2,471 languages as endangered to some degree – more than a third of the world’s languages, and almost certainly an unrealistically low estimate of the actual number. This is not a pretty picture. In Chapter 2 we examined causes and processes of language endangerment. Chapter 3 was about language death, or dormancy, illustrated with five case studies (Eyak, Cornish, Egyptian, Yaaku, and Mednyj Aleut) and covering the topics of “tip”, semi-speakers, attrition, and grammatical and lexical replacement as an unusual route to language death. Chapters 4 and 5 were devoted to showing what is lost when a language is lost – a culture and its intellectual riches (Chapter 4) and knowledge about human history, human cognition, and the natural world (Chapter 5). Chapter 6 explored the topic of linguistic fieldwork, focusing on aspects that are more challenging when the language being investigated is endangered. That chapter was not in itself gloomy, because documenting endangered languages is a good thing, certainly for science and potentially for the speech communities; and doing fieldwork brings joy and excitement to the fieldworker who discovers never-before-seen structures and, with luck, develops close working relationships with speakers who are continuing a fascinating culture. 153

154

revitalizing endangered languages

But all six chapters, taken together, appear to point to just one possible outcome for an endangered language: decreasing use by members of its speech community, often accompanied by reduced linguistic resources, until no one speaks the language any longer – that is, temporary or permanent language death. We have also seen in these six chapters a few indications of hope for the future of some endangered languages, however. The Yaaku people of Kenya, for instance, decided some years ago to revive their heritage language (Chapter 3 and below), and the Cornish people are also working to revive their language (Chapter 3 and below). Like many other endangered language communities, the Salish and Pend d’Oreille tribes of Montana want their language to be documented so that the documentation can aid the tribes’ revitalization efforts (Chapter 6). This seventh and final chapter cannot offset the sad impression left by the first six chapters: language endangerment is a fact of the modern globalizing world, and it is not going to go away. Still, there is genuine cause for cautious optimism for at least some endangered languages. The past twenty years have seen a dramatic sea change in attitudes toward endangered languages – attitudes of the speech communities, of the linguistics profession, and even of the general public. Endangered speech communities all over the world have launched vigorous efforts to revitalize their languages. Their efforts are reported regularly, with increasing frequency, in the popular press. And linguists have woken up to the danger of seeing Michael Krauss’s dire prediction come true – that the discipline of linguistics could go down in history as ‘the only science that presided obliviously over the disappearance of 90% of the very field to which it is dedicated’. To be sure, the optimism expressed in writings on language revitalization is tempered, as we can see, for instance, in the wording of Leane Hinton and Ken Hale’s dedication of their edited volume The green book of language revitalization (2001): ‘To the brave people who work against all odds to help their endangered heritage languages survive’. But optimism is nevertheless warranted. The result of all this attention to the tragedy of language loss has been an explosion of activity. Googling any of dozens or perhaps hundreds of endangered languages will lead to one or more websites devoted to revitalizing that language. As an illustration, consider the fact that all but one of the five languages described in the case studies of language death in Chapter 3 have become targets of revival efforts. First, although (as we saw in Chapter 2) Eyak may be said to have died when its last fluent native speaker, Marie Smith Jones, died on January 21, 2008, a revitalization program has begun; surviving relatives and other community members have organized the Eyak Language Project, funded by the Eyak Preservation Council, the Eyak Corporation of Alaska, and Alaska’s Humanities Forum. Its prime movers are Eyak tribal members, together with Michael Krauss – who, thanks to his earlier fieldwork on the language with Marie Smith Jones and several other fluent speakers, speaks Eyak fluently – and the French linguist Guillaume Leduey. Second, Cornish revitalization efforts are represented by the Cornish Language Partnership (CLP), whose mission is to encourage the use of the language; the CLP was established in 2005 ‘to oversee the implementation

Factors that contribute to successful revitalization efforts

of the Cornish Language Development Strategy’. As we saw in Chapter 3, this revitalization program has led to the reclassification of Cornish on the UNESCO endangerment scale from “extinct” to “critically endangered” (with an accompanying R to indicate revitalization), and to a formal acknowledgment by the government of the United Kingdom of its status: Chief Secretary to the Treasury, Danny Alexander, today (24 April 2014) announced that the proud history, unique culture, and distinctive language of Cornwall will be fully recognised under European rules for the protection of national minorities.

Third, when, in 2005, the Yaaku community regretted its decades-old decision to abandon their heritage language, they invited a team consisting of a linguist (Maarten Mous), a videographer (Matthijs Blonk), and a pastoral worker with the Maasai (Hans Stoks) to visit. The team’s mission was to find out whether ‘there is sufficient knowledge available in the community to revive the language’. And fourth, even long-dead Egyptian, in its late Coptic form, has been the subject of revival efforts: a Google search for “Coptic lessons” receives more than a thousand hits, some of them for lessons designed to teach people how to speak Coptic. (Admittedly, it seems likely that most of the enthusiasm for learning Coptic stems from its status as the liturgical language of the Coptic church – a status resembling that formerly enjoyed by Latin in the Catholic church – and not from a desire to reintroduce Coptic as the everyday spoken language of any community.) Of the five case studies in Chapter 3, only the bilingual mixed language Mednyj Aleut seems to lack an actual or potential revival movement. It may be surviving in an ideological sense, but apparently not in a plausible linguistic sense. Nikolai Vakhtin reports that this mixed language is “extremely viable” in that community members believe it to be an Aleut language even when whole sentences are more or less pure Russian with only an Aleut word or two thrown in. Because Russian with a few Aleut words is not a mixed language, Mednyj Aleut as it existed fifty years ago is effectively dead. As we saw in Chapter 5, this mixed language was unique in its particular combination of elements from its two source languages; if linguists had encountered it only in its present state, the possibility of such a combination might never have been discovered. The remainder of this chapter focuses on two sets of issues surrounding revitalization efforts: factors that contribute to success, and types of revitalization programs. The final section looks to the future of endangered-language revitalization.

7.1

Factors that contribute to successful revitalization efforts

Just as speakers and would-be speakers are trying to revive four of the five critically endangered or dead languages discussed in Chapter 3, so we see

155

156

revitalizing endangered languages

this pattern repeated over and over around the world. In an attempt to revitalize an endangered language, the need to know what works and why – and, almost equally important, what doesn’t work – is therefore crucial. What the record shows to date is one spectacularly successful case of language revival, a growing number of partial successes in revitalizing endangered languages, and a tiny handful of partial successes in reviving languages that had fallen completely out of use. (These are sometimes called “sleeping languages”, partly in an effort to avoid the stark finality of the term “dead language” and partly to convey the belief that any language can in principle be reawakened.) The one definitely successful case of revival is Modern Israeli Hebrew, as we saw in Chapter 1. It is risky, to put it mildly, to generalize over a set of one, but we can list some of the major factors that made the revival of Hebrew possible: its status as the sacred language of a major world religion; the desire for a language which (unlike Yiddish) was unconnected with the European exile, for the new nation of Israel; the fierce dedication of the lexicographer and editor Eliezer Ben-Yehuda and others who advocated for a revived Hebrew language; the early establishment of kibbutzim, collective communities in which children were raised in groups, speaking Hebrew; the fact that an accepted Hebrew orthography already existed, together with a large body of sacred and other writings; and the early establishment of Hebrew-medium schools. Some of the factors that helped Hebrew burst into vigorous renewed life cannot, of course, be replicated in modern endangered-language communities. Most obviously, almost no other endangered languages enjoy the prestige of being vehicles for major world religions. A well-established orthography is also fairly rare for an endangered language, and raising children in kibbutzim where the endangered language is the only community language is not a strategy that is readily available in the vast majority of cases. Still, it is possible to identify factors that contribute to success or failure and to use them in assessing the prospects for successful revitalization of an endangered language. Various authors have provided lists of such factors; in the following paragraphs we examine three of those lists to try to get a sense of how much agreement there is among authors. The three lists have somewhat different purposes, but comparing them will nevertheless be illuminating. In their 2006 book Saving languages, Lenore Grenoble and Lindsay Whaley note that ‘language revitalization involves counter-balancing the forces which have caused or are causing language shift’, and they name a number of variables that should be taken into account in planning a revitalization campaign. They divide the variables into two sets. Their macro-variables are international lingua francas (e.g., English); national and regional language policies, including educational policies; language attitudes toward multilingualism at the national level; regional autonomy with respect to language policy; national financial support for revitalization; and specifically regional variables (pressures from major regional languages, levels of multilingualism in neighboring languages, and the like). Their micro-variables, all at the local level, are language attitudes,

Factors that contribute to successful revitalization efforts

human resources available for revitalization efforts, religion and language, literacy, and financial resources. David Crystal, in Language death (2000), offers six postulates for predicting successful revitalization of an endangered language. ‘An endangered language will progress,’ he says, ‘if its speakers increase their prestige within the dominant community’; if they ‘increase their wealth relative to the dominant community’; if they ‘increase their legitimate power in the eyes of the dominant community’; if they ‘have a strong presence in the educational system’; if they ‘can write their language down’; and if they ‘can make use of electronic technology’. The most-cited list of all is Joshua Fishman’s, first in his 1991 book Reversing language shift and then, with slight modifications, in later publications. The version here is from his 2004 article ‘Language maintenance, language shift, and reversing language shift’. Fishman’s scale of ‘stages of reversing language shift’ is designed to ‘both describe the situation and prescribe the necessary ameliorative steps’; the ultimate goal, in his opinion, is to create a diglossic situation, one in which the dominant language is used to fulfill certain functions (including formal contexts) and the heritage language is used for others (especially less formal contexts). His eight stages are sequential – that is, each stage should be dealt with successfully before the next stage is addressed. Here are the steps Fishman recommends. First, adults in the community are to (re)learn the heritage language so that they can become the teachers of younger community members. Second, once that has been achieved, settings for interactions in the endangered language should be established, so that members of the older generation (who might otherwise be isolated from each other) will have an opportunity to speak the language to each other regularly. Third, the language should be transmitted from the older to the younger generations in homes and neighborhoods; informal intergenerational conversations in the language should be encouraged. Fourth, now that community members can converse easily with each other informally, literacy in the endangered language should be introduced, but in addition to, rather than in competition with, the official educational system. Fifth, schools should offer some instruction using the endangered language as the medium, but still with the dominant community in charge of the curriculum and staffing of the schools. Sixth, once all the previous stages are achieved, the endangered language should be used in the workplace, both between members of the endangered speech community and between them and members of the dominant-language speech community (which means that members of the dominant community must also learn to speak the endangered language to some extent). Seventh, the endangered language should be used in mass media and local governmental operations. The eighth and final stage is to have the endangered language used at the national level, in the educational system (including universities), at work, in the government, and in mass media. How well do these lists of factors fit together? They are not directly comparable, of course, because the three sources emphasize different aspects of the revitalization process. Grenoble and Whaley set out variables which, for

157

158

revitalizing endangered languages

any particular endangered language, must be analyzed in planning a revitalization project. Each situation will inevitably present unique problems; some variables will be relevant for one language but irrelevant for another. Crystal is making predictions about very broad circumstances that will, in his view, bring about revitalization, or at least make it possible. But only the last two of his six postulates – literacy in the endangered language and its use in electronic media – are clearly under the control of the speech communities (although in the U.S. a number of Native American tribes have acquired wealth and prestige by opening casinos that attract gamblers from non-Native communities, and in some cases the new prestige may have extended to their heritage languages). Fishman prescribes particular practical measures that should be taken sequentially to bring a language from “critically endangered” to “safe”. The lists therefore serve different purposes and, understandably, are not interchangeable. They do overlap, however. The two most consistent features are the need for literacy in the endangered language and the need for the endangered language to have (as Crystal puts it) ‘a strong presence in the educational system’; both of these appear in all three sources. This requirement may be somewhat controversial – assigning major responsibility for revitalization to the schools could turn out to be a recipe for failure – but teaching and using the language in education in addition to promoting its use in the community outside the school system should be beneficial. Grenoble and Whaley’s discussion of attitudes and Crystal’s listing of prestige cover overlapping ground; but Fishman’s stages concern tangible procedures rather than intangible features like attitudes, and he does not (in the particular passage cited earlier) include the modification of attitudes as a step in the revitalization process. Elsewhere, there is explicit agreement between Fishman and the other authors. For instance, Fishman’s stages in which adults (re)learn the heritage language and then create settings for using the language is very similar to Grenoble and Whaley’s observation, under the heading for the micro-variable “human resources”, that speakers ‘are critical for teaching the language and for helping create new domains for its use.’ Only Crystal specifically makes a major point of wealth and power relative to the dominant-language community, and only Crystal – somewhat surprisingly, given that his is the earliest of the three sources – highlights the importance of electronic technology. Today, far more than when his book was published, the impact of electronic technology is immense, and attention to it is often seen as crucial for revitalization. Its scope is evident from the very large number of websites devoted to promoting the learning and use of endangered languages, from the discussions about and use of endangered languages on social media, and from such innovations as iPhone apps for endangered languages. (I consider this topic in more detail later in this chapter.) The tight connection between economic factors and the fate of minority languages is underscored by Wayne Harbert, who says that the fundamental shapers of the fortunes of endangered languages are economic in the narrow sense: people change their linguistic behaviours,

Factors that contribute to successful revitalization efforts

including shifting from one language to another, most typically because of real or perceived or desired changes in their material circumstances. The status and viability of endangered languages can thus be affected (positively or negatively) by measures that affect the material circumstances of the speakers and the economies of their communities.

Harbert would presumably agree with Crystal, then, about the importance of increased wealth for the success of a revitalization program. Modern Hebrew, as noted earlier, has features that limit the usefulness of its revival process as a blueprint for revitalizing other endangered or extinct languages. Its strong religious connection, the urge to shed the main language(s) of exile, and the kibbutzim are all rare almost to the vanishing point. Hebrew does, however, display other features that recur in the lists of factors already discussed, including its established orthography, the preexistence of literacy in Hebrew (including things to read in the language), and the Hebrew-medium schools. The process of revitalizing Hebrew also had the human resources listed by Grenoble and Whaley as a micro-variable – not only Eliezer Ben-Yehuda, whose passion for reviving Hebrew is famous, but other committed champions of the revival process as well. Modern Hebrew did not start out with any fluent speakers who used the language in their everyday lives, but it certainly had, as Grenoble and Whaley put it, ‘committed, energetic people to implement’ the revival program. Still, in spite of this and other links to the lists of contributing factors in various works on language revitalization and revival, Modern Hebrew probably cannot be used as a general model for revival or revitalization of any other language. One further point needs to be made before we move on to a survey of some revitalization and revival programs. An astute reader might have noticed the conspicuous absence of two specific factors in the lists: none of these authors mentions any particular level of documentation that is required for successful language revitalization, and none of the lists includes contributions by linguists as a desirable part of the process, much less as a necessary part. These are not accidental omissions. Most authors do believe that linguists’ participation can be helpful, and these authors would agree. Grenoble and Whaley, for example, mention linguists among other ‘external human resources’ – their list is ‘linguists, professional pedagogues, teacher-trainers, and language planners’. But they add that the revitalization effort ‘needs to come from within the commmunity’ and that, while sometimes essential, ‘these external sources cannot provide the core of support necessary to create and sustain a revitalization program’. In other words, like other authors, and surely also like every fieldworker who has tried to help with a revitalization program, Grenoble and Whaley believe that no external person can revitalize a language, or even lead a revitalization effort. The impetus must come from the community, and the leadership must be provided by the community. The decision as to whether linguists or other outside experts should contribute to the effort rests with the community, not with the outside experts. Certain circumstances make a linguist’s participation useful, of course. Linguists (at least some of them) are professionally qualified to describe

159

160

revitalizing endangered languages

languages; nonlinguists rarely have the requisite expertise for this task. If an endangered language has not already been investigated and analyzed, that expertise will be needed. The language cannot be taught in schools, for instance, without pedagogical materials, and pedagogical materials should be based on a solid analysis of the language’s structure. An unfortunate fact should be admitted here, however: few fieldworkers engaged in linguistic description and/or documentation have any training whatsoever in applied linguistics. This means that although fieldworkers are capable of providing grammatical descriptions, dictionaries, and texts to aid in the development of pedagogical materials, they are not themselves professionally qualified to prepare textbooks, grammar lessons, and the like. Nor are they qualified to develop language curricula. Working with skilled local teachers, if any are available, can make up for this deficiency; but it would be better if the preparation for linguistic fieldwork were to include at least a modicum of training in applied linguistics so that fieldworkers could contribute directly to pedagogical aspects of a revitalization program. Ideally, as we saw in Chapter 6, the linguist doing the description and documentation will be a member of the endangered language community. Many authors who write about language revitalization urge that community members (not necessarily speakers of the endangered language) be trained in linguistic analysis so that they can both engage in documentation and implement the revitalization program. They should also be trained in relevant areas of applied linguistics, in particular curriculum development, language teaching methods, and (if there is not yet an established writing system for the target language) the development of orthographies. As we will see in the next section, new organizations are providing this training to community members, but progress to date has been slow: so far, only a few endangered-language communities have members who are trained in linguistic analysis and/or language pedagogy. If the need is for revival rather than revitalization – that is, if the language is not merely endangered but entirely lacking in speakers, even semi-speakers – and if in addition it is poorly documented, then only a linguist trained in both documentation and philology (the interpretation of texts, including old written publications and field notes) will be able to produce the linguistic material to support a revival program. In the majority of cases, linguistic fieldworkers are able and willing to help the communities they work with in implementing a revitalization program. But after making their willingness known to the community, they must (at least in North America, and in many other field sites too) wait to be asked for help. Even if they are not asked to participate directly in implementing the program, they may still be able to provide materials to support it, once they have the community’s permission to work with the elders, assuming there are still some elders who speak the language. In other words, outside linguists can do only what the commmunity wants them to do.

Factors that contribute to successful revitalization efforts

But although conforming to the community’s wishes is always the best practice, waiting to be asked for help is not the right move everywhere. In exploring this issue in the context of her field research in Papua New Guinea (PNG), Lise M. Dobrin found (not surprisingly) that the complexity of ethical and moral dimensions may result in dramatic differences in community attitudes from setting to setting. In PNG, she found that an outside linguist’s active participation in a local program might be not only welcome but actually vital for the success of the program – not primarily because of a need for the books and teaching materials that the linguist could provide, but because of the significance of the linguist’s participation for the community’s valuation of their heritage language. One community member told her, ‘Oh Lise, because you have come we speak our language. If you weren’t here we’d be speaking pidgin.’ (The irony is that this comment was itself made in Tok Pisin, a pidgin-turned-creole language that is, as noted in Chapter 2, the major lingua franca of PNG.) It is therefore not safe to make sweeping global assumptions about community expectations for an outsider’s potential contributions to a revitalization program. It is the linguist’s responsibility to find out what kind of help will be welcome. The general topic of factors that help make a revitalization program successful is enormous, and this section has barely scratched its surface. I close the section with two examples to illustrate the complexity of the issues that communities wrestle with in launching revitalization efforts. The first is Pite Saami, a critically endangered Saami language or dialect (there is some debate about its linguistic status) spoken in northern Sweden by a few elders. The youngest of these elders speaks the language at home to his two children; the children also attend classes in Northern Saami, a different Saami language (or dialect). Other children in the community have not (yet) been exposed systematically to Pite Saami. Pite Saami is still used, however, in a number of domains, always between adults: at home, on the telephone, and when herding reindeer or hunting or fishing. But all Pite Saami speakers are fluent in Swedish too, and Swedish is also spoken regularly in these domains. For Pite Saami speakers, Swedish is apparently the dominant language, and they will often code-switch from Pite Saami to Swedish in mid-conversation. Efforts are being made in Sweden to increase the presence of other varieties of Saami in the media, on radio and TV programming; but Pite Saami is not (yet) represented in the media, although one second-language speaker reports using it on the internet. The Pite Saami Documentation Project has a website that is funded by one of the most important funding organizations for endangered languages, the Hans Rausing Endangered Languages Project at the School of Oriental and African Studies, University of London. Pite Saami lacks an established orthography, a dictionary, and textbooks that could be used in classrooms, but materials prepared for a different Saami variety, Lule Saami, have been modified for use in Pite Saami classes (which, as of 2011, were rare). Older documentation materials exist, but they are not usable by most Pite Saami speakers. Speakers are now

161

162

revitalizing endangered languages

compiling wordlists, together with some grammatical information, and these are expected to facilitate progress toward other pedagogical materials – although there are disputes about appropriate spellings for certain of the language’s phonemes. Meanwhile, the lack of qualified teachers for prospective language classes remains a problem. Saami is an official minority language of Sweden, which means that Saami revitalization efforts (not just Pite Saami efforts) are eligible to receive financial and other support. The speakers’ attitudes toward their language are positive, a vital factor in calculating the chances for successful revitalization; but there are too few of them. Pite Saami study circles have been organized. Riitta-Liisa Valijärvi and Joshua Wilbur conclude their 2011 discussion of the Pite Saami situation on a pessimistic note: ‘The revitalization of Pite Saami may appear very difficult, if not impossible, because of the small number of people involved.’ They argue that the revitalization effort is nevertheless worthwhile, even if its goals must be modest. My second example is more upbeat. The future of the Myaamia Project looks more promising than that of Pite Saami, in spite of the fact that the starting point for the Myaamia Project was zero: this project aims to revive a dormant language, the Algonquian language Miami, sometimes called Miami-Illinois because the two tribes’ languages were linguistically very close. This is the heritage language of the Miami Tribe of Oklahoma, formerly of Indiana (where members of an eastern branch of the Miami Nation still live), western Ohio, and neighboring regions. The last speaker of Miami died in 1962. The impressive momentum of the revitalization project is due primarily to two men, Daryl Baldwin and David Costa, and to two institutions, the Miami Tribe of Oklahoma and Miami University (Ohio), both of which provide support for the Myaamia Project. The project is housed at Miami University, which is named for the Miami Nation. Baldwin is the Director of the Myaamia Project and a member of the Miami Tribe of Oklahoma. Determined to revive his heritage language, he earned an M.A. in linguistics from the University of Montana in 1999 and taught himself Miami as a second language, from written documents. He has directed the Myaamia Project at Miami University since its inception in 2001. He is raising his four children to speak Miami – he says they speak it more fluently than he does, since they were exposed to the language from birth. Costa is a linguist who received his Ph.D. in 1994 from the University of California, Berkeley, with a dissertation entitled The Miami-Illinois language, the first systematic description of the language’s structure. He now works for the Myaamia Project. In his dissertation he used descriptive and philological methods to reconstitute Miami-Illinois from written sources that started as far back as the early eighteenth century; he also uncovered a rich trove of old materials in the language, a major contribution to the revival project. Besides offering college classes in the language, the Myaamia Project organizes language camps for children and other cultural and educational activities. At least two other Miami families are also learning Miami, so that it is reasonable to speak of an emerging (though tiny so far) Miami speech community. Leanne Hinton observed in 2001 that ‘a single individual can produce

Some types of revitalization programs

miracles for even the tiniest of language communities’, and she went on to say that Baldwin ‘has done wonders’ with his program. The major factors that make the Miami revival project promising are the existence of a dedicated and effective project leader; two hundred years of extensive documentation; the introduction of Miami in the educational system (Baldwin and his wife are home-schooling their own children to ensure constant exposure to Miami); the support of the Miami Tribe of Oklahoma and Miami University; positive attitudes toward their heritage language among tribal members, including young members; and electronic savvy (see for instance the project website at www.myaamiacenter.org). National U.S. language policies are not particularly favorable to multilingualism or indigenous languages, but the project has solid funding sources in the tribe and Miami University. By contrast, the Pite Saami revitalization effort lacks most of these factors: money, a groundswell of support from the community, a solid place in the educational system, an established, noncontroversial spelling system, and extensive documentation. Reviving a no-longer-spoken language is hardly possible without written records or (if the last speakers died relatively recently) recorded speech. Miami has the written records; Pite Saami, which is still spoken, does not, and there may be too few speakers to make up for the absence of rich documentation.

7.2

Some types of revitalization programs

Let us turn now to a brief survey of different kinds of revitalization programs. Before beginning the survey, however, we need to consider what it means for a revitalization program to be successful. Some people might say that success means restoring the language to the community in its traditional form, so that the language as spoken by new speakers will be indistinguishable from the speech of the last few fluent native speakers. But this, in most cases, is an unrealistic goal if the community’s heritage language is severely endangered or (in the worst case) no longer spoken. In such cases there are too few fluent native speakers to serve as models for language learners, and they are likely to be at an age where they cannot devote themselves to intensive language instruction, or to the task of learning how to be effective language instructors. The elders might object (and sometimes do object) to a program that produces speakers whose speech is markedly deviant from the elders’ language; but as Leanne Hinton put it in 2001, ‘if a language is close to extinction, many people are willing to settle for what they can get’. She quotes two language learners who adopt this position: Terry Supahan, a learner and teacher of Karuk (an isolate, or possibly a Hokan language, of northern California), commented, ‘I’m interested in communication, not in preservation’; and Cody Pata, a learner of Nomlaki (a Wintuan language of northern California, without native speakers), said that ‘he would rather use “Pidgin Nomlaki” than simply not be able to communicate in Nomlaki at all’.

163

164

revitalizing endangered languages

Depending on how seriously endangered a language is when revitalization efforts begin, it may be completely impossible for learners to acquire the language as it is, or was, spoken by the last fluent native speakers. Obviously a line has to be drawn somewhere between speaking some version of the endangered language, as in these cases, and speaking some other language with a few words of the dying or dead language thrown in, as in the current ideological version of Mednyj Aleut discussed earlier in this chapter. The remnant of Mednyj Aleut, which appears to be Russian with the addition of some Aleut words, may well serve its community as an identity marker; but by an objective linguistic measure it cannot be considered to be a linguistic descendant of Mednyj Aleut. Successful revitalization programs may not produce speakers of the traditional language of their elders, but they can produce speakers of a language recognizable as an abruptly altered version of the ancestral language. In other words, closeness to the language of the ancestors is not, for many revitalization programs, an appropriate criterion for measuring success. It is not surprising, given the large number of different methods for teaching foreign languages in the classroom, that different endangered-language communities have chosen different methods for revitalizing their languages. All revitalization and revival programs, regardless of whether they involve classroom instruction or not, certainly involve language learning. As with lists of factors that contribute to the success or failure of revitalization programs, different authors have different lists of types of revitalization programs. I review three lists here, those of Ofelia Zepeda, Lenore Grenoble and Lindsay Whaley, and Leanne Hinton. Zepeda’s list comprises just three methods, or “best practices”, for language revitalization: the master/apprentice approach, language and culture camps, and immersion programs. Grenoble and Whaley have a longer list in their chapter on ‘Models for revitalization’: total-immersion programs; partial-immersion or bilingual programs; the local language as a second, “foreign” language; community-based programs; the master-apprentice program; language reclamation models; and, in a final section, a question: documentation or revitalization? Their last two categories are relevant to situations where the target language is spoken by few or no community members. The language needs to be reclaimed when there are no speakers – when it is sleeping (or silent, or dormant, or dead). Miami, as we have seen, is such a case: reviving it means relying on primary written documents and on David Costa’s dissertation, a detailed grammatical description based on the primary written documents. The question in Grenoble and Whaley’s final category heading is meant to underscore the fact that a language that has only a handful of elderly speakers must be documented as thoroughly as possible in order to make revitalization or (if the last speakers have died before the effort starts) revival possible. Hinton’s list divides programs into three categories: school-based programs, adult language programs, and family-based programs at home. Then there are subcategories. The school-based programs are (a) the endangered language as

Some types of revitalization programs

a subject, (b) bilingual education, (c) immersion schools and classrooms, (d) language and culture, and (e) children’s programs outside the school. The two subcategories under adult language programs are (a) documentation and materials development and (b) development of teaching tools. And the two subcategories under family-based programs at home are (a) raising bilingual children and (b) one parent, one language. These three lists overlap considerably, although the overlap is not always obvious from the labels. My survey is based on Grenoble and Whaley’s categories, because their clear and sharply focused list falls between Zepeda’s very short list and Hinton’s more diffuse list in length. If I were to guess which methods are most effective in endangered-language communities, I would split my vote: a total-immersion method if the number of fluent speakers is fairly robust, a master/apprentice method if there are only a very few elderly speakers. (If there are no speakers at all, the only practical choice is a reclamation method.) This section is a selective rather than an exhaustive survey of revitalization methods, concentrating on the most-cited methods. First, consider revitalization methods that involve total immersion. This is how children growing up in a monolingual environment learn the language of their surroundings: it’s the language they hear from their parents, from older siblings, in the neighborhood, everywhere. A child of three has already learned much of the structure of his or her native language and a great many words as well. Children are better language learners than adults are, but some adult readers of this book will have had the experience of living in a foreign country, surrounded by speakers of a language previously unknown to them, and picking up the language whose environment they are newly immersed in. The purpose of a total-immersion method of language revitalization is to recreate that kind of setting for learners of the endangered language. The language nest is by far the best-known type of immersion program; this was mentioned in Chapter 1 in connection with Maori (a Polynesian language of New Zealand), which is by far the best-known case in which language nests were the major revitalization method. Hawaiian (also a Polynesian language) is a fairly close second: it has seen some striking successes in creating a new speaker base of children who have grown to adulthood as fluent speakers of Hawaiian. A language nest targets preschool children. The idea is that fluent speakers spend each day in the schoolroom with the children, talking to them only in the community’s heritage language. The speakers, who are typically elderly, are not tasked with child care: the preschool’s regular teachers take care of the children’s nonlinguistic needs, speaking to them either in the heritage language or not at all. The children thus hear, and speak, the heritage language all day long. The Maori language was declining rapidly before language nests were established in New Zealand in 1985. It is difficult to assess the level of success that the language nests have achieved; by many accounts, the erosion of the language has slowed or (according to the most optimistic reports) stopped altogether. Some of the graduates from the language nests have been able to attend Maori-medium

165

166

revitalizing endangered languages

primary and secondary schools. As Grenoble and Whaley observe, the Maori language program has inspired numerous other endangered-language communities to employ the language nest model in their own revitalization programs, among them Ojibwe (Algonquian), Lakota (Siouan), and Halq’emeylem (Salishan) in North America, Miriwoong and Gooniyandi in Australia, and the Uralic languages Karelian and Saami in Europe and Forest Enets in Siberia. Grenoble and Whaley note that a Mohawk revitalization program started with a language nest and then followed the children into primary and secondary schools, establishing Mohawk-medium instruction incrementally. Grenoble and Whaley express doubts about the efficacy of partial immersion or bilingual programs, which are implemented in schools and which feature some classes taught in the dominant language and some taught in the endangered language. But although they do not recommend this type of program, they believe that it is ‘arguably the most frequently encountered model’. The next type of program, where the heritage language is learned by children or adults as a second or “foreign” language, is akin to (for instance) studying a major European language in an American classroom, except that the available teaching materials are likely to be limited when the target language is gravely endangered. Community-based programs take the revitalization process out of the classroom and into an informal setting where the learners can acquire cultural knowledge along with linguistic knowledge. This model, as Grenoble and Whaley observe, has the advantage of ‘creating a domain for use of the local language’, something a classroom-based program cannot easily achieve. The Master/Apprentice Language Learning Program was developed in 1992 by Leanne Hinton and three colleagues as a means to rescue the many dying indigenous languages of California from oblivion. This is the method of choice for revitalizing endangered languages that are so reduced as to have only a very few speakers. The plan is simple: the master, a fluent native speaker (in most or all cases an elder), is paired with an apprentice, a younger community member who is eager to learn the community’s heritage language. The teams work on their own, without expert supervision, and the master is unlikely to have any teaching experience, but intensive training is provided to the teams before they begin their work. Once they start their work, the teams spend ten to twenty hours together each week, speaking only the target language. The training, among other things, highlights the following “Eight Points of Language Learning” for both the teacher (the master) and the learner (the apprentice): Point 1, “Be an active teacher” (“Find things to talk about”), “Be an active learner” (“Ask about things”). Point 2, “Don’t use English, not even to translate” (teacher), “Don’t use English, not even when you can’t say it in the language” (apprentice). Point 3, “Use gestures, context, objects, actions to help the apprentice understand what you are saying” (teacher), “Use gestures, context, objects, actions to help in your communication when you don’t know the words” (apprentice). Point 4, “Rephrase for successful communication” (teacher, when the apprentice doesn’t understand), “Practice” (“Use new words and new

From the past to the future

sentences and grammar as much as possible”). Point 5, “Rephrase for added learning” (“rephrase things the apprentice says”, to show the correct forms or extend his/her knowledge), “Don’t be afraid of mistakes” (“If you don’t know how to say something right, say it wrong” – apprentice). Point 6, “Be willing to play with language” (both teacher and apprentice – make up word games together, make up stories, etc.). Point 7, “Understanding precedes speaking” (“Use various ways to increase and test understanding”, such as giving commands to the apprentice; for the apprentice, “Focus on understanding ... After you understand an utterance fully, learning to speak it will not take long”). And Point 8, “Be patient” (“Repeat words and phrases often” – teacher), “Be patient with yourself” (apprentice). Even more than the language nest approach, the master/apprentice method has spread widely. According to the (undated) Advocates for Indigenous California Language Survival website, for instance, more than seventy teams have graduated from their training program, and thirty teams are currently in training. The Resource Network for Linguistic Diversity in Australia held workshops in 2012 to train thirty-six Indigenous trainers, representing thirtyone Aboriginal languages, to train master/apprentice pairs all over the country. And at the Smithsonian Institution in Washington, DC, the National Breath of Life Institute for Indigenous Languages has run ten-day workshops at which ‘teams of participants (Native American heritage language learners, teachers and activists) paired with linguists (experts in linguistics who assist the participants in their research) to explore the language resources in the District of Columbia area’.

7.3

From the past to the future

In most of this book’s chapters we have looked backward – to causes and processes of language endangerment, to language dormancy and permanent language death, to cultural loss and the loss to science when a language disappears. In this final chapter we have glimpsed a brighter future for those lucky endangered-language communities, almost certainly a very small percentage of the total number, for which revitalization efforts will enjoy a substantial measure of success. In this closing section I comment on two more topics concerning the future of endangered languages: the significance of technological advances, and what communities gain from attempts, even failed attempts, to revitalize or revive their heritage language. 7.3.1

Language revitalization in the digital age

The use of technological resources in a campaign to revitalize an endangered language has barely been touched on in this chapter. I will not go deeply into the topic here either, because it quickly leads away from problems

167

168

revitalizing endangered languages

specific to revitalizing endangered languages and into areas of wider concern. But it is important to give some idea of what the present offers and what the future might offer in the way of digital potentialities with specific reference to language revitalization. First, let’s look briefly at a recent event of relevance – a symposium that was organized by two prominent specialists in endangered languages, David Harrison and Claire Bowern, and presented at the 2012 Annual Meeting of the American Association for the Advancement of Science, focusing in part on these issues. The symposium was entitled ‘Endangered and minority languages crossing the digital divide’, and its abstract is worth reproducing in full here: Speakers of endangered languages are leveraging new technologies to sustain and revitalize their mother tongues. The panel explores new uses of new digital tools and the practices and ideologies that underlie these innovations. What new possibilities are gained through social networking, video streaming, twitter, software interfaces, smartphones, machine translation, and digital talking dictionaries? How can crowd-sourced translation and localization projects protect intellectual property while providing a technology resource? The panelists present actual and imagined uses and impacts of new digital technologies for a variety of stakeholders: speakers, educators, archivists, linguists, language activists, and technology providers. There are also benefits to science when indigenous languages assume a prominent role in digital technologies. They can provide testing grounds for new media and technological delivery, presenting a level of data complexity often not found in major global languages and thus leading to new discoveries. And they lend greater prominence to traditional knowledge, thus expanding access to the human knowledge base.

The first part of the symposium abstract concerns ways in which endangered language communities can benefit from the use of new technologies in their efforts to revitalize their heritage languages. Technological advances have opened up exciting opportunities for developing new methods for promoting the revitalization of endangered languages. They have also, it must be said, increased the complexity of making data widely available while still treating members of the speech communities with the greatest respect. A well-known problem, in an age when it is so easy to put multimedia endangered-language materials on the internet and make them available to the world, is the issue of informed consent by the endangered-language commnities and their members. Many elders do not wish to have their words, much less their voices – or their deceased relatives’ words and voices – made available to the general public. Nor are these the only intellectual property issues that arise: in 2006, for example, the Mapuche community of Chile (whose language is not at present endangered) was planning to sue Microsoft for translating its Windows software package, with the support of the Chilean Education Ministry, into the Mapuches’ language Mapuzugun without the Mapuches’ permission. Microsoft’s response, in part, was that it ‘wanted to help Mapuches embrace the digital age’ and ‘open a window so that

From the past to the future

the rest of the world can access the cultural riches of this indigenous people’. It isn’t hard to see why the Mapuches might find Microsoft’s argument to be paternalistic and intrusive. Solving problems of these kinds, as well as the problem of safe permanent data storage, will take time and ingenuity. As David Nathan has warned, ‘factors at the core of language endangerment argue against across-the-board free access to data’. But beyond putting materials on the web, a crucially important development is the increasing use of endangered languages (along with English, Russian, Spanish, and all the other “safe” languages) in social media and other web-based resources. This makes social media, collectively, a powerful tool for revitalization. András Kornai has argued recently that the popular saying ‘If it’s not on the web, it doesn’t exist’ has an important application to the revitalization of endangered languages: a digital presence is vital for the prospects of successful revitalization, because that is where digital natives – young people who have never known a world without digital media (computers, cell phones) – experience much of their verbal communication. If an endangered language is to be revitalized, young people must be encouraged to use it in their digital lives – in tweeting, texting, blogging, and other kinds of informal written communication. Successful development of a robust online presence, which Kornai labels digital ascent, ‘requires use in a broad variety of digital contexts’. Kornai concludes his statistical study of languages with and without a digital presence on a gloomy note, however: ‘for ... 95% of the world’s languages ... there is very little hope of crossing the digital divide’. But modern language revitalization programs are heavily dependent on the internet. As already mentioned earlier in this chapter, revitalization websites can be found for dozens or even hundreds of endangered languages. It is hardly likely that most or all members of all those endangered-language communities have easy access to digital media, but the members who do have easy access are likely to be younger people, and many of those are likely to be digital natives. This would of course leave many, many endangered-language communities around the world with little or no access to the new electronic world – primarily for economic reasons and secondarily for related reasons having to do with technological support and opportunities for education in such matters as web-page design. And many communities whose members do have access will not be using their access either to set up websites for their endangered language or to communicate with their friends on social media in their endangered language. Nevertheless, successful initiatives have been undertaken to establish a digital presence for endangered languages. One example is a creative way of using social media for group action – an endangered-language analogue of the digitally-based collaborative activism that helped elect Barack Obama to the U.S. Presidency in 2008: Udmurt (an endangered Uralic language whose speakers, as we saw in Chapter 2, have suffered from feelings of linguistic inferiority by comparison to Russian) is the subject of ‘an internet-based contest’ to create Udmurt neologisms to replace the very numerous business and computer terms

169

170

revitalizing endangered languages

borrowed from Russian. Another example: a study of language use by Canadian Hungarians found that digital language use is ‘a highly important domain of heritage language use ... while only slightly more than half of the (typically older) first-generation ... members of this community used Hungarian on the internet, almost all of the (typically younger) second-generation speakers did so’. This is a surprising result in view of the usual North American pattern of rapid shift from the immigrants’ language to the new country’s language, which would predict the rapid demise of this variety of Canadian Hungarian. Another important aspect of using endangered languages (and dialects) for communication on social media is that the writing tends to be informal, and this is ideal for fostering use of the language among younger community members. As Anna Fenyvesi observes, [T]he various informal forms of writing required of indigenous language use in the digital domain are likely to be a good “playground” for language users in their progression of developing literacy in their indigenous/minority language.

7.3.2

The value of unsuccessful revitalization efforts

My final observation is that endangered-language communities whose language revitalization efforts do not succeed in establishing fluent speakers and regular domains of heritage-language use may still rejoice in the efforts and their results. Even if the language vanishes, the community will have the products of the revitalization program, including some or all of the following: a dictionary of the language, recorded narratives of personal and community history, myths, and grammar lessons, all of which will enable future generations of the community’s youth to learn the heritage language and recapture part of the old culture if they wish to; also culture-specific vocabulary that provides insights into traditional cultural values and customs, and – even more important – a newly invigorated sense of community and an intergenerational closeness arising from the community-wide revitalization efforts. The linguist William Poser has lived and worked for many years with the Carrier (Athabaskan) Nation in northern British Columbia. He has taught Carrier to community members, trained language teachers, developed curricula for language classes, produced multiple dictionaries and other materials, and worked with fluent elders. One of the elders he worked with most intensively was Mary John, Sr. (1913–2004). A few years ago one of Mary John’s granddaughters told Poser that she had just written a college essay on medicinal plants, for which she had consulted a dictionary compiled by Poser. She added, ‘Of course we knew how much time you spent with Granny, but until I did this I never realized how much you managed to learn. It is sooo precious.’ If Poser had not studied Carrier so thoroughly, the deep knowledge that Mary John possessed would have been lost to her community when she died. Thanks to his work, that knowledge has been preserved for her grandchildren and their grandchildren, into the future.

Sources and further readings

The appreciation of young community members for the preservation of the community’s traditional knowledge is one major reward of a revitalization program. Another is the possibility of trying again in the future, using the documentation assembled in the first attempt, if the first revitalization program does not succeed.

7.4

Sources and further readings

The literature on language revitalization is already very large and is growing fast. The most prominent books to date, in chronological order of publication, are Joshua Fishman’s 1991 book Reversing language shift: theoretical and empirical foundations of assistance to threatened languages (and see also his 2001 edited volume Can threatened languages be saved? Reversing language shift, revisited: a 21st century perspective); Leanne Hinton and Ken Hale’s coedited volume The green book of language revitalization in practice (2001); Hinton et al.’s How to keep your language alive (2002); Lenore Grenoble and Lindsay Whaley’s Saving languages: an introduction to language revitalization (2006); and Leanne Hinton’s edited volume Bringing our languages home: language revitalization for families (2013). There is also a 2005 textbook, Tasaku Tsunoda’s Language endangerment and language revitalization: an introduction. Among the most important articles on this subject are Hinton’s 1997 article ‘Survival of endangered languages: the California Master-apprentice program’; her 2001 article ‘Language revitalization: an overview’ (in the Hinton and Hale volume), her 2011 article ‘Revitalization of endangered languages’; and Fishman’s 2006 article ‘Language maintenance, language shift, and reversing language shift’. Two recent survey articles are Suzanne Romaine’s ‘Preserving endangered languages’ (2007) and Colin Baker’s ‘Endangered languages: planning and revitalization’ (2011). See also Andrea Wilhelm’s bibliography, ‘Language revitalization’, in the Oxford Bibliographies series (Oxford University Press). A much-admired documentary film on the Wampanoag language reclamation project was mentioned in the Sources and Further Readings section of Chapter 5: it depicts the work of Jessie Little Doe Baird and her community in reawakening their dormant language. Their project website is at www.wlrp.org (accessed 28 May 2014). Another good film is ‘Myaamiaki Eemamwiciki: The Miami Awakening’ (www.myaamiacenter.org/publications/261-2/, accessed 29 May 2014). The first quotation at the beginning of the chapter is from Wesley Y. Leonard’s 2010 talk ‘iilaataweeyankwi: Miami language reclamation as decolonization’. The second quotation is from the front matter of the Hinton and Hale’s The green book of language revitalization in practice (2001); since 2001 the UNESCO publication that was called The red book of endangered languages has turned into the UNESCO’s online publication Atlas of the world’s languages in danger.

171

172

revitalizing endangered languages

The third quotation at the beginning of the chapter is from Nancy Lord’s 1996 article ‘Native tongues’. The quotation from Michael Krauss is on p. 10 of his 1992 article ‘The world’s languages in crisis’. The Eyak Language Project is described here: https://sites.google.com/site/ eyaklanguageproject/ and here: www.eyakpeople.com (both accessed 18 December 2013). The home page of the Cornish Language Partnership is at www. magakernow.org.uk, and the source of the quotation about the status of Cornish as a newly recognized official minority language is at https://www.gov.uk/ government/news/cornish-granted-minority-status-within-the-uk. A description of nascent efforts to revive Yaaku can be found at www. matthijsblonk.nl/paginas/YaakuENG.htm, from which the quotation about Yaaku revival is taken. Nikolai Vakhtin writes about Russianized Mednyj Aleut in his 1998 article ‘Copper Island Aleut: a case of language “resurrection”’. For a description of Mednyj Aleut when it was still a bilingual mixed language, see for instance my 1997 article ‘Mednyj Aleut’ and references cited there. The first quotation from Grenoble and Whaley is on p. 21 of their 2006 book Saving languages. They discuss their macro-variables on pp. 22–38 and their micro-variables on pp. 38–45; the second quotation from their book, about human resources, and the third quotation, about external human resources, are both on p. 41. The quoted postulates for language revitalization from David Crystal’s book Language death (2000) are from pp. 130, 132, 133, 136, 138, and 141, respectively. The quotations from Joshua Fishman’s 2004 article ‘Language maintenance, language shift, and reversing language shift’ are on p. 426, and his scale of eight ‘stages of reversing language shift and severity of intergenerational dislocation’ is on p. 427. He intends his scale to be read from bottom (stage 8) to top (stage 1); I have reordered them in listing them. I am grateful to Peter Austin (personal communication, 2014) for the warning about the danger of expecting the educational system to do all the heavy lifting of revitalization. An early(ish) source on the uses of electronic media for language revitalization is Laura Buszard-Welcher’s 2001 article ‘Can the Web help save my language?’. The quotation about the importance of economic factors for a successful revitalization program is from p. 404 of Wayne Harbert’s 2011 article ‘Endangered languages and economic development’. Lise M. Dobrin’s thoughtful article on ethical and moral issues involved in outside linguists’ participation in revitalization efforts is ‘From linguistic elicitation to eliciting the linguist: lessons in community empowerment from Melanesia’ (2008); the comment made to her by a community member is from p. 310 of this article. For further discussion of these and related matters, see also Jane H. Hill’s much-cited 2002 article “‘Expert rhetorics” in advocacy for endangered languages: who is listening, and what do they hear?’ and the 2011 article coauthored by Dobrin and Josh Berson, ‘Speakers and language documentation’. Hinton and Hale’s coedited 2001 volume The green book of language revitalization has an entire section on training, with the following articles: ‘Training people to teach their language’, by Leanne Hinton, pp. 349–350; ‘Inuttut and

Sources and further readings

Innu-aimun’, by Ken Hale, pp. 351–352; ‘The role of the university in the training of native language teachers: Labrador’, by Alana Johns and Irene Mazurkewich, pp. 354–366; ‘Languages of Arizona, southern California, and Oklahoma’, by Leanne Hinton, pp. 367–369; ‘Indigenous educators as change agents: case studies of two language institutes’, by Teresa L. McCarty, Lucille J. Watahomigie, Akira Y. Yamamoto, and Ofelia Zepeda, pp. 371–383; ‘The Navajo language: III’, by Ken Hale, pp. 385–387; and ‘Promoting advanced Navajo language scholarship’, by Clay Slate, pp. 389–410. The description of Pite Saami is taken from a 2011 article by Riitta-Liisa Valijärvi and Joshua Wilbur, ‘The past, present and future of the Pite Saami language: sociological factors and revitalization efforts’; the quotation about the slim prospects for successful revitalization is from p. 322. The authors argue that Pite Saami is ‘a full-fledged language and not a dialect’ (p. 298), in spite of its very close linguistic similarity to other varieties of Saami. The website of the Pite Saami Documentation Project, maintained by Joshua Wilbur, is http:// saami.uni-freiburg.de/psdp/. The Myaamia Project has recently been renamed the Myaamia Center, but since most sources still refer to it as the Myaamia Project, I have used the original name in describing it here. A brief autobiographical note by Daryl Baldwin, the Director of the Myaamia Project, is at http://listserv.linguistlist.org/cgi-bin/ wa?A2=e-meld-queryroom-ab;3bOuUw;200302031626130500. The website of the Myaamia Project is at www.myaamiacenter.org. The quotation from Leanne Hinton about Daryl Baldwin is on p. 416 of her 2001 article ‘Sleeping languages: can they be awakened?’ in The green book of language revitalization in practice. See also the 2013 article ‘Miami: myaamiaataweenki oowaaha: Miami spoken here’, by Daryl, Karen, Jessie, and Jarrid Baldwin, in the edited volume Bringing our languages home: language revitalization for families. A revised version of David Costa’s 1994 Ph.D. dissertation on the Miami-Illinois language was published by the University of Nebraska Press in 2003. For an excellent description and analysis of the Baldwins’ learning and use of Miami, see Wesley Y. Leonard’s 2007 dissertation Miami language reclamation in the home: a case study; the abstract is on line at www.myaamiacenter.org/wp-content/uploads/ 2013/02/Wes_leonard_Abstract.pdf. The quotations from Leanne Hinton about individuals’ ability to produce miracles of language revitalization are from p. 16 of her 2001 article ‘Language revitalization: an overview’. Her quotations from Terry Supahan about Karuk and Cody Pata about Nomlaki are from the same page of that article. Ofelia Zepeda listed her three “best practices” methods for language revitalization in a 2005 lecture presented at the University of Michigan, ‘More than just words: stages of American Indian language revitalization’. Grenoble and Whaley devote chapter 3 of their 2006 book Saving languages to a discussion of types of revitalization programs, and Hinton presents her categories in ‘Language revitalization: an overview’ in The green book of language revitalization in practice (2001).

173

174

revitalizing endangered languages

Grenoble and Whaley, in their 2006 book, discuss language nests, citing Maori and Mohawk as their examples, on pp. 52–54. Their negative comment about partial-immersion revitalization methods is on p. 55, and their comment about an advantage of community-based programs is on p. 60. The Master/Apprentice Language Learning Program is described by Leanne Hinton in her 2001 article by the same name; the ‘Eight points of language learning’ come from pp. 243– 244 of Hinton’s essay ‘Rebuilding the fire’ in her 1994 book Flutes of fire. The Advocates for Indigenous California Language Survival website is here: www.aicls.org. The website of the Australian organization Resource Network for Linguistic Diversity is at www.rnld.org/MALLP. The website on which the Breath of Life Institute’s description can be found is www. endangeredlanguagefund.org/BOL_2013_home.php. An excellent starting point for learning about language revitalization in the digital age is Gary Holton’s 2011 article ‘The role of information technology in supporting minority and endangered languages’. Holton does not, however, address the vital and vexed question of ‘who gets access?’. The report of the Mapuches’ objection to Microsoft’s action is from the Sydney Morning Herald, 24 November 2006: www.smh.com.au/news/biztech/mapucheindians-to-bill-gates-hands-off-our-language/2006/11/24/1163871586715.html (accessed 27 May 2014). David Nathan’s warning that it is often or always inappropriate to make endangered-language data available to everyone, and his discussion of the reasons why, are in his 2011 article ‘Digital archiving’; the quotation is from p. 265. András Kornai’s analysis of the digital presence of the world’s languages is in his 2013 article ‘Digital language death’; the quotation is from p. 10 of the article. The emergence of digital natives (vs. digital immigrants, people who remember what life was like before the internet) was proclaimed by Marc Prensky in two 2001 articles, ‘Digital natives, digital immigrants’ and ‘Digital natives, digital immigrants, Part 2: do they really think differently?’. The last three paragraphs in the section on endangered-language usage in digital media owe much to Anna Fenyvesi’s 2014 article ‘Language endangerment and revitalization in the digital age: realities and responsibilities’, and the two quotations – about the use of Canadian Hungarian in digital media and about the value of digital media for encouraging endangered-language use among young community members – are from this article. For the Udmurt example, Fenyvesi cites ‘Veme berykton as a social technology’, a paper presented by Artyom Malykh et al. at the MinorEuRus Conference in Helsinki, Finland, 16–18 December 2013; and for the Canadian Hungarian example she cites Máté Huber’s 2013 University of Szeged M.A. thesis, Intergenerational transmission of Hungarian as a heritage language in Canada: the macrosociolinguistics of the Hungarian community in Hamilton, Ontario. Bill Poser’s Yinka-Déné Language Institute’s website is at www.ydli.org. The Wikipedia page for Carrier elder Mary John, Sr., is at https://en.wikipedia.org/ wiki/Mary_John,_Sr.

Glossary – languages and terms

Adjective, predicative: An adjective that forms (part of) the predicate of a sentence. For example, in the English sentence That dog is ugly, is ugly is the predicate and ugly is a predicate adjective. Adjectives that modify nouns directly are attributive adjectives, for instance English ugly in that ugly dog. Affix: A morpheme (unit of grammatical analysis) that is added to a stem to form a new word and/or a new stem. Most affixes are either prefixes (added to the beginning of a stem) or suffixes (added to the end of a stem). In English, the prefix un- is usually added to adjectives (as in, e.g., unhappy, where the stem is happy) or verbs (as in, e.g., unlock, where the stem is lock); the suffix -ness is added to adjectives to form nouns, as in happiness. Some English words have both a prefix and a suffix, for instance unhappiness, in which -ness has been added to the stem unhappy. And some words have more than one suffix, for example realities, which consists of the adjective root real, a noun-forming suffix -ity to form a noun reality (which is also a stem), and a plural suffix -es. Affricate: A complex speech sound that begins with a short oral stop component and ends with a short fricative component. Afro-Asiatic: A language family named for its range, which extends from westernmost Asia (ancient and modern Mesopotamia, the Arabian Peninsula, the eastern Mediterranean coast) eastward to northern Africa. The family consists of five or six branches: Semitic, Egyptian (an extinct one-language branch), Cushitic, Berber, Chadic, and – according to some, though not all, specialists – a sixth branch, Omotic. Agreement: Grammatical linkage (feature sharing) between words in a clause (or sentence). The most common type in the world’s languages is subject/verb agreement, where the verb is marked for (for example) person and number to agree with its subject, as in Englsh she walk-s, where the -s indicates, among other things, that the verb has a third-person singular subject. Other common types of agreement are object/verb and adjective/noun, for example, in Swahili ni-li-m-penda ‘I liked him’ (literally ‘I-past-him-like’) vs. ni-li-wapenda ‘I liked them’ (literally, ‘I-past-them-like’), where the prefixes m- and wa- indicate, respectively, members of the singular and plural noun classes that include humans; and Russian staraja kn’iga ‘an old book’ vs. starij dom ‘an old house’, where the adjective ‘old’ agrees in gender, number, and case with

175

176

Glossary

its head noun (nominative singular feminine for ‘book’, nominative singular masculine for ‘house’). Algic: A language family comprising the large Algonquian family and two distantly related Californian languages, Wiyot and Yurok. Algonquian: A language family with about twenty living members spoken primarily in eastern and midwestern Canada and the U.S., but also in some western regions of both countries. Almost all Algonquian languages are gravely endangered; some now have only a handful of fluent speakers. Among the Algonquian languages mentioned in this book are Abenaki (Eastern and Western), Arapaho, Atsina (Gros Ventre), Blackfoot, Cheyenne, Cree, Delaware, Mahican, Massachusett (including Wampanoag), Meskwaki (Fox), MiamiIllinois, Ojibwe, Potawatomi, Powhatan, and Shawnee. Allophone: One of the set of speech sounds that are the phonetic realizations of a phoneme. In English, for instance, there are two allophones of the phoneme /l/: a so-called “dark l”, which occurs after the vowel in its syllable, and a so-called “light l”, which occurs before the vowel in a syllable. If you’re a native speaker of English, you can hear both of these allophones if you listen carefully to your pronunciation of the word little, which begins with the light l and ends with the dark l. Most allophones of a single phoneme are positionally determined – they occur in mutually exclusive phonetic environments – but sometimes allophones can occur in the same environment: in an English word like skip, for instance, the final p phoneme can be pronounced with or without opening the lips; each variant is a different allophone, but it’s the same word skip regardless of which allophone is chosen. Analogic leveling: A process of morphological change in which variant forms of a single morpheme generalize so that the total number of variants is reduced. In English, for instance, the verb seem once had a past tense with the same vowel as Modern English kept and slept; but the alternation between seem and what would have become Modern English sem- was leveled analogically – that is, the past-tense form of the verb root, sem-, changed to seem- by analogy to the form of the present-tense verb seem – so that today seem is a regular verb. In addition, the past-tense suffix changed from irregular -t to regular (-ed; this was also an instance of analogic leveling, as the suffix changed by analogy to the regular past-tense suffix. Aspect: Temporal information about processes and events other than their location in time relative to the time of utterance, that is, other than tense. Often, though not always, aspect is grammatically marked on the verb. In English, for instance, the distinction between I was walking and I walked is aspectual – durative or continuous vs. nondurative action. Perhaps the most common temporal distinction in the world’s languages is the aspectual distinction between completed and noncompleted action (often called perfective vs. imperfective). Other common aspectual categories are durative (or continuous), iterative

Glossary

(repeated action), and inchoative (the beginning of an action). See tense for the other major type of time indication. Assimilation (in speech sounds): A process, in a language’s phonology, by which one sound becomes more similar to a nearby sound. Most assimilations affect neighboring consonants, but distant assimilation also occurs. English examples are the variation, in one of the negative prefixes, between in- and im-: im- occurs before other labial sounds, as in impossible and imbalance, while in- occurs elsewhere, as in intolerant or inability; thus the consonant in im- has assimilated from the basic tongue-tip n to a labial m because of the following labial consonant p or b. Athabaskan: The major branch of the Athabaskan-Eyak language family. (Eyak is a single language and is now extinct; see Chapter 3 and Chapter 7 for discussion.) Subgrouping within the family and boundaries between various pairs of languages are hard to determine, but there are about twenty-four northern Athabaskan languages in Alaska and western Canada, nine languages in Oregon and California, and two southern (Apachean) languages, Navajo and Apache, spoken primarily in New Mexico and Arizona. There are six major dialects of Apache, some of them quite divergent. Atlas of the world’s languages in danger (UNESCO): This website, which is said to be updated regularly, aims to list all of the world’s endangered languages; see www.unesco.org/culture/languages-atlas/. Its scale of language vitality and endangerment has six categories: safe, vulnerable, definitely endangered, severely endangered, critically endangered, and extinct (see Chapter 1 for details). The list includes “vulnerable” languages, that is, languages that are not yet endangered but that might soon become endangered. It also includes “extinct” languages, but only those that have died recently; readers will not find such ancient languages as Latin, Sumerian, or Akkadian on the site. Attestation of a language: Documentation – actual recorded data – of a language. For dead languages in particular, the date and nature of the attestations limits our knowledge of the language’s lexicon and structure. See also unattested. Attrition in a dying language: The loss of vocabulary and structure, without any compensating additions in the form of borrowings or new creations. Often, maybe most frequently, the loss is not universal in a dying language, but is confined to the speech of semi-speakers – who may well be living among fully fluent speakers who still know and use the structures that the semi-speakers lack. Austronesian: An enormous language family comprising hundreds of languages and covering most of the islands in the Pacific Ocean and large parts of the neighboring Asian mainland. Among the languages in the family are Malagasy (the westernmost Austronesian language, spoken on the island

177

178

Glossary

of Madagascar off the coast of Africa), Malay, Indonesian and many other languages of Indonesia, the languages of Polynesia (Hawaiian, Maori, Tahitian, etc.), and the indigenous (i.e., non-Chinese) languages of Formosa. Bantu: Languages belonging to the Bantu subgroup of the Benue-Congo branch of the Niger-Congo language family; one of the most numerous and geographically widespread language groups of Africa. Among the Bantu languages mentioned in this book are Shambala and Swahili. Basic vocabulary: Vocabulary items that are likely to exist in every language in the world and are therefore less likely to be borrowed than nonbasic vocabulary items. (The term “core vocabulary” is used sometimes to mean basic vocabulary, though for some authors “core vocabulary” has a different meaning.) This is a vague concept – there is no solid theoretical basis to support it – but it has proved useful as one kind of evidence for establishing genetic relationships among languages and also, more commonly, as a tool for eliciting words at the beginning of a language documentation and description project. Basic vocabulary items tend to be shorter than words for nonbasic concepts and objects. As has often been demonstrated, not all items on standard lists of basic vocabulary do in fact occur in every language, and all the items on all the lists can be shown to have been borrowed somewhere, sometime. But in general the expectation that basic vocabulary is less often borrowed than nonbasic vocabulary is valid. Examples of basic vocabulary items are mother, father, dog, water, cloud, sun, walk, run, swim, one, two, white, and black. Typical nonbasic vocabulary items in English are airplane, camera, philosophy, and church. Probably the most-used standard lists of basic vocabulary items are the 100- and 200-word lists compiled in the mid-twentieth century by the linguist Morris Swadesh, who used them in his efforts to establish genetic relationships among languages. Bilingual mixed language: A language created by bilinguals, with major components drawn from each of the two languages in a contact situation. It is distinguished from the other category of mixed languages – pidgins and creoles – by the fact that its creators are bilingual. As with pidgins and creoles, the major lexical and structural components of bilingual mixed languages can’t all be traced back primarily to a single source language; in contrast to pidgins and creoles, each major component of a bilingual mixed language can be traced back to a particular source language. The term “bilingual mixed language” is misleading in that such a language might be composed of subsystems from more than two languages. Examples of bilingual mixed languages are Michif, Mednyj Aleut, and Ma’a. Case: A means of expressing syntactic relations between nouns and other words in a sentence, especially by morphological marking on noun phrases. Compare, for instance, German Ich sah den Mann ‘I saw the man’ (literally ‘I.nominative saw the.masc.sg.accusative man’ and Der Mann sah mich ‘The man saw me’ (literally ‘the.masc.sg.nominative saw me.accusative’), where the accusative

Glossary

case marks the object noun phrase and the nominative case is used for the subject noun phrase. Other common cases are the genitive, typically the case of a possessor (as in Russian kn’iga Ivan-a ‘John’s book’, literally ‘book.nom.sg John-gen.sg’), instrumental (as in Serbian putujem avtobus-om ‘I travel by bus’, literally ‘I.travel bus-inst.sg’), and various locative cases (‘in’, ‘onto’, ‘out of’, etc.). Causative: A verb construction that involves x causing y to do something. Typically (though by no means always), a causative construction has an affix added to the verb, adding an extra actor to the verb’s set of arguments (subject/actor, object/patient, etc.). For instance, the English predicate is thick in The soup is thick is non-causative, but She thickened the soup (with flour) has a causative verb thicken; compare also non-causative sit and causative set (‘cause to sit’), or fall and fell as in He felled the tree (‘caused the tree to fall’). Celtic: A branch of the Indo-European language family. Celtic was originally spoken in continental Europe, but after ancient Gaulish became extinct (eliminated by the spread of Latin and its descendent Romance languages), all the remaining Celtic languages were spoken in the British Isles: Irish Gaelic, Scottish Gaelic, Welsh, Cornish (extinct since the nineteenth or very early twentieth century, but now being revived), and Manx (now extinct). The one Celtic language that is now spoken on mainland Europe is Breton (spoken in Brittany, a region of France); the ancestors of Breton speakers migrated to what is now Brittany in the fifth century CE from the Cornish-speaking region of southern England. CLAN: Programs for editing files, data analysis, and other tasks connected with description and documentation projects. See http://childes.psy.cmu.edu/clan/ (accessed 28 December 2013). Clicks: Speech sounds formed in the mouth with a closure at the back of the mouth and a second closure farther forward; the forward closure is released with a loud popping sound (a click) on an ingoing airstream. Clicks are found as regular speech sounds primarily in southern Africa, in Khoisan languages and – as borrowed features – in Zulu and a few other southern Bantu languages. Although they hardly occur as regular speech sounds elsewhere, clicks are very common all over the world as “para-linguistic” vocal sounds. Examples in English-speaking environments are the kissing sound (made by suction from a bilabial stop closure), the scolding tongue-tip sound often spelled tsk-tsk, and the clucking sound that is sometimes used to urge a horse to start moving. Code-switching: The use of material from two (or more) languages by a single speaker with the same people in the same conversation. A famous 1980 article by Shana Poplack on this topic has a code-switch in its title: ‘Sometimes I’ll start a sentence in Spanish y termino en español’. (This title is correct as it stands – it’s what one of Poplack’s consultants said. Some sources “correct” the first part to

179

180

Glossary

‘Sometimes I’ll start a sentence in English’, but the actual title has both Spanish and español.) Comparative Method: The most valuable method in the field of historical linguistics; the main methodology by means of which genetic linguistic relationships are established and unattested parent languages (proto-languages) of language families are reconstructed. The Comparative Method – which involves systematic comparison of the vocabularies, sounds, word structures, and syntax of a group of related languages – dates from the late nineteenth century in essentially its modern form. It provides us with a window on prehistory by enabling us to reconstruct sizable portions of protolanguage lexicon, phonology, and morphology, and to a lesser extent syntax as well; in this way it greatly expands our ability to examine language changes over considerable time depths. Contact-induced language change: Any linguistic change that would have been less likely to occur outside a particular contact situation – that is, a linguistic change that was caused at least in part by contact with another language. The most common types of contact-induced changes are words and (less frequently) structures borrowed by bilingual speakers from other languages, on the one hand, and on the other hand changes introduced into a target language by people learning it as a second (or third, or fourth, etc.) language. Coptic: The final stage of the Egyptian language, usually dated from the fourth century to sometime between the fourteenth century and 1700 CE. Coptic is known from mainly religious texts written after the Egyptians were converted to Christianity in the third and fourth centuries CE. It is still used as the liturgical language of the Coptic Church, but it is no longer spoken as an everyday language, although a few revitalization efforts have been made. Creole language: A mixed language that is the native language of a speech community. Like pidgins, creoles develop in contact situations that typically involve more than two languages; also like pidgins, they typically draw their lexicon, but not their grammar, primarily from a single language, the lexifier language. Some creoles arise as nativized pidgins, some arise abruptly with no pidgin stage, and others arise gradually, with or without a pidgin stage. Crucially, the creators of a creole (unless it’s a nativized pidgin) are not bilingual in their interlocutors’ languages. Cushitic: A branch of the Afro-Asiatic language family. Cushitic languages are spoken in Ethiopia, Kenya, and other African countries. Daughter language: Any of the descendants of a single parent language. For example, the Romance languages are all daughters of Latin, and Modern English is the daughter of Old English. Dead language: A language that has no remaining speakers. Most languages that lose all their speakers are gone forever, truly dead. But some dead

Glossary

languages are targets of language revival, or reclamation, programs – and those speaker-less languages can be called dormant or sleeping languages, since they are being awakened to life again. Degemination: A phonological process in which a geminate sound sequence is shortened, or reduced, to a single sound. English once had geminate consonants between vowels, but in Middle English times, they were all degeminated. Thus, for instance, a word like kiss, which used to have a phonetically long (double) ss and a final suffix -an as well (it was spelled cyssan in Old English), now has just one phonetic s. The double-s spelling is a relic of its former pronunciation. See also geminate. Devoice: Change from a voiced segment to a voiceless segment. In German, for instance, all voiced stops devoice at the end of a word, so that the word Bund ‘union, alliance’ is pronounced [bunt], but when a suffix like -es is added it is pronounced [bund]-es. Digital native: A person who has grown up using digital media (computers, cell phones) and has never known a world without digital media. Diglossia: Traditionally, a situation in which two dialects of the same language are used in different domains in the same community. In the prototypical case, one dialect is used for “high” purposes such as formal writing and formal speech, while the other dialect is used for “low” purposes, typically informal contexts such as conversations within a family or between friends and written communications on social media. The concept of diglossia has been extended to include situations in which two different languages, rather than two dialects of the same language, are used in different domains. Ditransitive construction: A grammatical construction in which a verb has a subject and two objects. The English sentence John gave Mary the book is an example of a ditransitive construction, as opposed to John read the book, which is a simple transitive construction. Documentary linguistics: ‘the creation, annotation, preservation and dissemination of transparent records of a language’ (Woodbury 2011:102) – with the emphasis on developing a large corpus of material that is accessible to a variety of users, including, crucially, non-linguist members of the language’s speech community as well as linguists. This new subfield of linguistics arose as a response to the perceived urgent need to investigate endangered languages all over the world. Dual number: A morphological and syntactic category that refers to two entities. Most languages that have a dual have three number categories in all, singular, dual, and plural; singular means one of anything, dual means two things, and plural means more than two things. The dual is not a superrare category, but it’s not especially common in the world’s languages. Most languages distinguish only singular and plural, and some languages don’t even

181

182

Glossary

distinguish those very prominently in their morphology and syntax. A few of the world’s languages, mainly languages belonging to the Austronesian family, have more complicated systems of number categories, adding a trial (‘three things’) or a paucal (‘a few things’) category. For instance, Tok Pisin, an English-vocabulary pidgin created in large part by speakers of Austronesian (specifically Melanesian) languages, has the following second-person pronouns: yu ‘you’ (singular), yutupela ‘you (dual)’, yutripela ‘you (trial)’, and yupela ‘you (plural, i.e., more than three of you)’. The English sources of the morphemes in these pronouns are, respectively, you, two, fella (fellow), and three. Egyptian: An ancient language belonging to the Afro-Asiatic language family. Its chronological stages are Old Egyptian (3000–2000 BCE), Middle Egyptian (2000–1300 BCE), Late Egyptian (1300–700 BCE), Demotic (seventh century BCE to the fourth century CE), and finally Coptic, which lived from the fourth century to sometime between the fourteenth century and 1700 CE. ELAN: A ‘tool for the creation of complex annotations on video and audio resources’ – tla.mpi.nl/tools/tla-tools/elan/ (accessed 28 December 2013). Endangered language: A language that is declining precipitously in numbers of speakers and/or in domains of usage, especially if it is no longer being learned as a first language by children. The UNESCO scale identifies three degrees of endangerment (in addition to “safe”, “vulnerable”, and “extinct”; see Chapter 1 for these): “definitely endangered”, “severely endangered”, and “critically endangered”. (Note that pidgin languages are not necessarily endangered: because they have no native speakers by definition, they cannot be said to have lost native speakers.) English Only movement: A political movement in the United States that aims to establish English as the sole official language of government. The movement has not succeeded at the federal level, but numerous states have made English their official state language. The movement is closely related to the Official English movement, which has the same aim. Wikipedia has a good general article, with good references, on the English Only movement in the United States: en.wikipedia.org/wiki/English-only_movement (accessed 22 May 2014). See also English Plus. English Plus movement: In response to the English Only movement, the English Plus movement was launched in the United States to promote an environment in which minority-language speakers would learn English while still maintaining their heritage languages. That is, the English Only monolingual movement was countered by the English Plus bilingual movement. Wikipedia provides good references for English Plus: en.wikipedia.org/wiki/English_Plus (accessed 8 June 2014).

Glossary

Ethnologue: languages of the world (www.ethnologue.com): The most comprehensive source for speaker numbers of all known languages, a product of the missionary organization SIL International. Some of the numbers are outdated, but this is a very important resource. Exclusive ‘we’: See Inclusive vs. exclusive ‘we’. Exogamy: Out-marriage, marrying someone from outside your own ethnic group. In some cultures this is a requirement, with varying criteria for determining who counts as an outsider: depending on the culture, you are not allowed to marry a member of your own clan, or your tribe, or a speaker of the same language. Fieldworks: ‘Software tools that help you manage linguistic and cultural data. Fieldworks supports tasks ranging from the initial entry of collected data through to the preparation of data for publication’, for example, interlinearization of texts, dictionary development, bulk editing of many fields, and morphological analysis. – fieldworks.sil.org (accessed 28 December 2013). Finite verb: A verb that can (but needn’t) be the main verb of a sentence and that typically (but not always) has inflectional morphology marking the person and number of the verb’s subject and also the verb’s tense and/or aspect. As the hedges in this definition suggest, defining finiteness precisely is difficult, and in some languages there seems to be no distinction between finite and non-finite verbs. But in other languages the distinction is important because non-finite verbal forms, such as participles, gerunds, and infinitives, appear in subordinate constructions. See Chapter 5 for discussion of the formation of Mednyj Aleut, in which the finite verb morphology comes from Russian but the non-finite verb morphology, together with most of the rest of the language’s structure, is Aleut in origin. Finnic: A subbranch of the Uralic language family. Among its members are Finnish, Estonian, and Veps. Fricative: A speech sound in which two articulators are brought close enough together that the airstream is thrown into turbulence on its way out from the lungs; the turbulence in turn results in audible friction as the air passes through the narrowed passage. Examples of fricatives in English are the phonemes /f/, /z/, /š/, and /D/. Gaelic, Scottish: An endangered Celtic language spoken primarily in the Scottish highlands. It is very closely related to Irish Gaelic: both are descended from Old Irish. Geminate sounds: Speech sounds (usually consonants) pronounced as double sounds, two in a row – which are usually realized phonetically as extra-long. English has phonetic geminates only at morpheme boundaries, as in the word

183

184

Glossary

bookkeeper, with a long k sound; within morphemes, doubled English letters are pronounced as single sounds, as in gaggle or supper. See also degemination. Gender, grammatical: Grammatical gender is a system of noun classification that has grammatical relevance, usually in the morphology and/or the syntax. In most Indo-European languages, gender is biologically based – hence the term “gender”. In French, for instance, masculine creatures have masculine gender and feminine creatures have feminine gender; but since all nouns belong to one of these two gender categories, each category also includes a great many inanimates as well as gendered animates. The most common type of grammatical relevance is agreement. Here are a few French examples in which the definite article ‘the’ agrees in gender with the following noun: le chien ‘the male dog’, la chienne ‘the female dog’, le livre ‘the book’, la table ‘the table’. Some noun classification systems have different semantic bases. In Swahili and other Bantu languages, for instance, the numerous noun classes include such categories as “humans and animals”, “trees and plants”, and “abstract concepts”. Genetic relationship of languages: Genetic relationship in linguistics refers to language families that arise through descent with modification from a parent language. Genetically related languages, or sister languages, are changed later forms of a single common parent language, which is called a proto-language if (as is almost invariably the case) it is unattested. Genetic relationship is established by means of the Comparative Method. (The term “genetic” is metaphorical in this context: it does not imply a claim about inheritance of languages through biological genes.) Glottalized stops (and affricates): Oral stopped consonants pronounced on a glottalic airstream, that is, with a closure of the glottis – the passage through which the air stream moves from the lungs through the larynx – that is maintained until after the release of an oral closure. Grammaticality judgment: A speaker’s assessment of whether a given utterance is acceptable (grammatical) in his or her language. All native speakers of English would agree, for instance, that a sentence like The horse ate an apple is grammatical, while an utterance like Ate apple an horse the is ungrammatical. Individual grammaticality judgments often disagree, however: some English speakers would find a sentence like Everybody should turn off their cell phone grammatical, but others wouldn’t. Hebrew: A Semitic language, one of the two official languages of Israel (the country’s other official language is Arabic). Hebrew is by far the most successful example of language revival; for two thousand years it was not learned as a first language by children and was used only in religious ceremonies and in formal discussion. See Chapters 1 and 7. Hittite: The earliest attested (documented) language of the Indo-European family, known from documents dating from the sixteenth to the thirteenth century

Glossary

BCE. Hittite was spoken in the ancient Near East in what is now Turkey, where it was the language of a powerful empire for several hundred years (1750–1200 BCE). Inclusive vs. exclusive ‘we’: A distinction between including and excluding the hearer in an utterance like We’re going to a party tomorrow night – that is, between the meanings ‘we (including you) are going to a party tomorrow night’ and ‘we (not including you) are going to a party tomorrow night’. Many languages that have this distinction express it in pronouns, for instance mit ‘we (inclusive)’ vs. bu ‘we (exclusive)’ in Evenki (a Tungusic language of Siberia), abu ‘we (inclusive)’ vs. ale ‘we (exclusive)’ in Mundari (a Munda language of India), or yumitupela ‘we two (inclusive)’ vs. mitupela ‘we two (exclusive)’ in the English-lexicon pidgin language Tok Pisin. Indo-European (IE): The name of a very large and widespread language family that includes (among many other languages) English and the other Germanic languages, Latin and its daughters the Romance languages, Greek, Russian and the other Slavic languages, Gaelic and the other Celtic languages, Albanian, Armenian, Hindi and other Indic languages together with Persian and other Iranian languages, and the long-extinct ancient languages Hittite and Tocharian. Inflection: Affixation or other morphological processes that alter words, typically to express syntactic relations between words in a sentence. Inflection never changes the part of speech of the stem it’s added to, it is highly productive and in fact usually obligatory in the appropriate contexts, and it usually occurs farther from the root than derivational morphology does. Among the most common inflectional categories are case and number on nouns and, on verbs, tense and aspect and person and number of subjects and/or objects. Agreement morphology is inflectional. Examples of inflection in English (there aren’t many) are the past tense suffix -ed in walked, the comparative and superlative suffixes -er and -est in the adjectives safer and safest, the plural suffix -es in the noun horses, and the third-person singular subject present tense suffix -s in the verb walks. An example of a more complex inflectional process is the Ancient Greek first-person singular perfect aspect form le-loip-a ‘I have left’ from the verb leip- ‘leave’, with a reduplicative prefix le- that repeats the first root consonant and a vowel change in the root from e to o, both marking the perfect aspect, plus the special perfect aspect first person singular subject suffix -a. Institutional Review Board (IRB): A university committee whose approval is required for any research project that involves human subjects. IRB approval is mandatory for such research in most or all U.S. universities and for major U.S. funding agencies. IRB policies vary from university to university with respect to linguistic fieldwork, however: some IRBs require approval of descriptive fieldwork projects, others don’t. Intransitive verb: A verb that has no object. Two English examples are Sam slept soundly and Sam already ate. The English verb sleep usually does not

185

186

Glossary

take any object, while the English verb eat may be used either intransitively or transitively (as in Sam ate chicken). IPA: the International Phonetic Alphabet, a system of phonetic transcription based on the Latin alphabet. Iranian: A sub-branch of the Indo-European language family; together with Indic, Iranian forms the Indo-Iranian branch of IE. Iranian languages are spoken primarily in Pakistan, Iran, Afghanistan, and parts of the former Soviet Union. The most important Iranian language politically is Farsi, or Standard Persian, the official language of Iran. Other Iranian languages mentioned in this book are Balochi and Tadzhik. IRB: See Institutional Review Board. Isogloss: A line on a map, typically a dialect map, that surrounds the area in which a particular feature is found. The notion is sometimes extended to a map of a linguistic area, so that the term can also be used for a line that surrounds the languages that share a particular feature due to contact-induced change. When two or more dialect features have nearly the same distribution in space, their isoglosses are said to bundle (i.e., they are bundled isoglosses). Isolate: A language that has no known relatives; thus, a one-language family. Basque is an isolate, for instance; so is the ancient language Sumerian. Khoisan languages: A group of languages which were once believed to constitute a single language family, but which are now considered to form several different language families. They share the phonological property of click phonemes and are therefore sometimes called click languages. Most Khoisan languages are spoken in southern Africa, but two of them, Sandawe and Hadza, are spoken farther north, in Tanzania. Labialized consonant: A consonant, most often a velar or uvular consonant, pronounced with lip protrusion; to an English speaker’s ear, a labialized consonant sounds like a short w added to a k or other consonant pronounced at the back of the oral cavity. Examples of labialized consonants, from the SalishPend d’Oreille language of Montana, are the velar stop /kw / and the uvular fricative /x.w /. Language family: One or more languages that are descended from – that is, changed later forms of – a single parent language. The notion of descent with modification, as applied to languages, does not mean that the entire lexicon and structure of a language must come from a single parent language; rather, it means that at least most of the basic vocabulary and most of the structure of the language come from a single parent language. (Since “most of” is vague, it isn’t surprising that there are some borderline cases.) One large and well-known language family is Indo-European; another example is the one-language family Basque (a linguistic isolate). Among the other language families

Glossary

that are mentioned in this book are Algonquian, Austronesian, Eskimoan, Salishan, and Uralic. Language nest: An immersion method for language revitalization in which preschool children in day-care centers speak and are spoken to only in the target language; it was developed for the Maori revitalization program in New Zealand and has been used in other endangered language communities around the world. See Jeanette King’s 2001 article ‘Te Kohanga Reo: Maori language revitalization’ for a description of the Kohanga Reo (Language Nest) program, and see Chapter 7 for discussion of language nests. Language shift: The shift, by a person or a group, from the native language to a second language. Bilingualism isn’t language shift, although shifts usually involve a period in which individuals or whole groups are bilingual; a shift occurs when people give up their native language and start speaking another group’s language instead. Lateral consonant: A consonant pronounced with the tongue blocking passage through the center of the mouth, but not at both sides, so that the air passes out along the side(s) of the tongue. The most common lateral consonant by far in the world’s languages is an /l/ like the English sound in like, but there are others as well. Most of the languages of the Pacific Northwest linguistic area, for instance, also have two or more lateral consonants in addition to an ordinary /l/, including a voiceless lateral fricative. Lateral fricative: A fricative pronounced with the air coming around one or both sides of the tongue, which touches the roof of the mouth but not both sides of the mouth. Lateral fricatives (which are most often voiceless) are not found in English, but there is one in Welsh, and the spellings of the names Lloyd and Floyd are orthographic attempts to represent the lateral fricative – as are the two double l’s in the name Llewellyn. Lingua franca: A language of wider communication – that is, a language that is used for communication between groups who don’t speak each other’s languages, as well as between native speakers (if any) of the lingua franca and other groups. A lingua franca is by definition learned as a second language by at least some of its speakers. Some lingua francas are also learned as a first language by some speakers; the most obvious example is English, a (or the) worldwide modern lingua franca. Others are spoken solely as second languages; these are pidgin languages. The term “lingua franca” literally means “Frankish language”, that is, the language of the Franks – a German-speaking people who also gave their name to France after invading and conquering parts of modern-day France in the early Middle Ages. The original Lingua Franca was spoken in the eastern Mediterranean at the time of the Crusades. Linguistic relativity hypothesis: The proposal that the language one speaks determines (strong version) or influences (weak version) the way one thinks. This

187

188

Glossary

is often called the Sapir-Whorf hypothesis because the idea was first proposed by Benjamin Lee Whorf; Edward Sapir’s name is attached to the label primarily because Whorf was his student. The hypothesis, especially in its strong version, has been controversial since it was first proposed. The most vigorous modern proponent of the weaker version is John A. Lucy, who has tested the hypothesis experimentally in numerous studies (see for instance Lucy 1992, 1996, 1997). Maasai: A language belonging to the Nilotic sub-branch of the proposed NiloSaharan language family. Maasai is spoken by more than a million people (according to the Ethnologue), most of them in Kenya but some in Tanzania. Maori: A language belonging to the Polynesian branch of the Austronesian language family. Maori is one of the two official languages of New Zealand (the other official language is English). Language nests were first developed and used in community members’ efforts to revitalize Maori. Markedness: A distinction between linguistic features that derives from universal tendencies of occurrence. The basic claim of markedness theory is that unmarked features are easier to learn than marked features (or even, in some versions of the theory, that unmarked features are innate). According to the theory, this difference in learnability is reflected primarily in typological distribution (unmarked features are more common than marked features in languages of the world) and age of learning in first-language acquisition (children learn unmarked features before marked features); other ways of determining markedness have also been explored. So, for instance, the presence of nasal consonants in a phonological system is considered unmarked and the absence of nasals is considered marked, because almost all languages have at least one nasal stop; a sibilant fricative phoneme /s/ is considered unmarked by comparison to an interdental fricative phoneme /T/ because /s/ is much more common cross-linguistically than /T/ and is usually learned earlier in those languages that have both (e.g., English). Master/Apprentice Language Learning Program: Developed in 1992 by Leanne Hinton and three colleagues, this program is widely acclaimed as the best method for revitalizing endangered languages that have only a very few remaining speakers. The master, a fluent native speaker, is paired with an apprentice, a community member who wishes to learn the community’s heritage language. See Chapter 7 for further information about how the program works. Mednyj Aleut: A bilingual mixed language that arose on Mednyj (‘Copper’) Island off the east coast of Russia. Its lexicon and structure are primarily Aleut, with some lexical and structural borrowing from Russian, but the entire finite verb morphology is from Russian. Melting-pot ideology: The belief, especially in the United States, that people from a wide range of cultural and linguistic backgrounds come together to form

Glossary

a single united population, all shifting to English and to mainstream American culture and abandoning their home country’s language and customs. Mixed language: A language that did not arise primarily through descent with modification from a single earlier language. There are two types of mixed languages. On the one hand there are pidgins and creoles, which typically (though not always) derive their lexicon from one language but their grammar from no single language; on the other hand there are bilingual mixed languages, which derive one or more grammatical subsystems (including their lexicon) from one language and the other grammatical subsystem(s) from another language. By definition, mixed languages are not members of any language family and thus belong in no family tree – except perhaps as the ancestor of a language family: a mixed language has no single parent language in the historical linguist’s usual sense, but it may have descendants. This of course does not mean that mixed languages have no historical links to their source languages; the definition derives from the technical historical linguistic analysis of language history, as developed over the past century and a half and elaborated on the basis of a rich body of data from language families all over the world (but most importantly Indo-European, the focus of most of the research). Montana Salish: See Salish-Pend d’Oreille. Moribund language: A language that is no longer being learned by children as a first language. Morpheme: The smallest unit of grammatical analysis. Some morphemes are lexical, with independent semantic content, for example the noun dog or the verb run. Others are grammatical, with little or no independent semantic content; examples are the -s of dogs, the un- of unhappy, and the English infinitive marker to in to run. Morphology: Word structure – roots and affixes (prefixes before the root, suffixes after the root, and sometimes infixes within the root), shape-altering morphological processes such as reduplication (repeating part of a root or stem), and also compounds (combinations of two or more roots). Prosodic features are sometimes also morphological, as when an English word like present is a noun if the first syllable is stressed and a verb if the second syllable is stressed. Morphosyntax: The combined morphology (word structure) and syntax (sentence structure) of a language. In contact-induced change, one language’s syntax may influence another language’s morphology, or vice versa, so it’s often useful to consider these two grammatical subsystems as a unified set of morphosyntactic structures. Navajo: A southern Athabaskan language spoken in Arizona, New Mexico, and Utah. It is classified as “vulnerable” on the UNESCO scale of Language Vitality

189

190

Glossary

and Endangerment, but the number of Navajo children who are monolingual in English has risen sharply in the last thirty years (see Chapter 1). Nez Perce: A Sahaptian language spoken in northern Idaho. It is a gravely endangered language (“critically endangered” on the UNESCO scale of Language Vitality and Endangerment). Obviative: The less prominent third person in an Algonquian sentence containing two third-person forms. Sometimes called the “fourth person”, the obviative is used to refer to a third person (he, she, it, they) that is less important in the particular context than another third person; the more important third person is called proximate. So, for instance, imagine a story about someone named Mildred. In an Algonquian translation of the sentence Mildred saw Bill on the street, and she greeted him, Bill and him would be marked as obviative, because the main character is Mildred, not Bill. Paradigm: In grammar, a set of forms for a particular inflectional category or an individual word; the different forms in a paradigm typically serve different syntactic functions in a sentence. The Latin verb meaning ‘to love’ has a famous present-tense paradigm: amo: ‘I love’, ama:s ‘you (singular) love’, amat ‘he, she, it loves’, ama:mus ‘we love’, ama:tis ‘you (plural) love’, amant ‘they love’. In this paradigm, the verb root is am- and the words comprise its present-tense paradigm. This is actually only one small part of the entire paradigm of this root; together with other roots in its verb class, its full paradigm has other sets of forms as well, for the future tense, several past-tense categories, and so on. Parallel project approach to language documentation: An approach employed by Terrence Kaufman in Guatemala (1971–1979) and Mexico (starting in 1993), in which several linguists and native speakers of (mostly endangered) languages in a linguistically diverse area live and work in the same central location for weeks or months. Each linguist is assigned a language to document; s/he works with two or three native speakers to produce a preliminary dictionary, grammar, and text collection. The linguists consult regularly with each other and with the project director(s) about analytic and other issues that arise. Pharyngeal sounds: Speech sounds produced in the pharynx, down below the oral cavity. Learning how to pronounce these sounds involves pulling the tongue so far back and down into the throat that a novice tends to gag. Pharyngeals are extremely rare in the world’s languages, but Arabic famously has two pharyngeal consonants, a voiceless fricative and a voiced pharyngeal. Pharyngeals are also found in several indigenous languages in the Pacific Northwest of North America (primarily Washington, Oregon, and neighboring parts of British Columbia, but also in outlier languages as far east as western Montana). The Salish-Pend d’Oreille language of northwestern Montana, for instance, has four pharyngeal consonant phonemes: /Q, Qw , Q’, Qw ’/ (two plain, two labialized; two nonglottalized, two glottalized).

Glossary

Phoneme: A distinctive speech sound – that is, a sound that is capable of distinguishing one word from another. In English, for instance, /p/ and /b/ are separate phonemes, as indicated by the fact that pit and bit are different words. Phonology: The sound structure of a language, or the theory of languages’ sound structures. Pidgin language: A mixed language that arises in a contact situation involving (typically) more than two linguistic groups; the groups have no shared language – that is, no single language is widely known among the groups in contact – and they need to communicate regularly, but for limited purposes, such as trade; they do not learn each other’s languages. By definition, a pidgin language has no native speakers: it is learned and spoken solely as a second (or third, or fourth, etc.) language. A new pidgin’s vocabulary typically (but not always) comes from one of the languages in contact; the grammar does not come from that language or from any other single language. New (non-expanded) pidgins have less linguistic material than non-pidgin languages: fewer words, fewer stylistic resources, less morphology than the source languages, and so forth. Some pidgins are endangered, but their endangerment status is independent of the fact that they lack native speakers. Pirahã: A language spoken in the Brazilian Amazon, along the Maici River, by a few hunded people. Its only known relative is Múra. The language is classified as “vulnerable” on the UNESCO scale of Language Vitality and Endangerment; the Ethnologue classifies it as “vigorous”. Pragmatics: The subfield of linguistics that focuses on language in use, specifically the different meanings of utterances in different contexts. Prefix: An affix that is added to the beginning of a stem to form a new word and/or a new stem. In English, for instance, the prefix un- is usually added to adjectives (as in unhappy) or verbs (as in unlock). Prosodic features: Phonetic and phonological aspects of language that are not segmental (i.e., not consonants or vowels) but that instead have to do with such features as stress, for instance what syllable a word’s accent falls on, and sentence intonation, as in rising (on certain questions) vs. falling (on statements) intonation at the end of a sentence. See also Tone, a prosodic feature that distinguishes words from each other in many languages. Proto-language: The parent language of a language family, that is, the language from which all the other languages in the family are descended. By definition, proto-languages are unattested – they are not documented at all, but were rather prehistoric in the literal sense: before recorded history. The descendants of a proto-language are genetically related; genetically related languages, or sister languages, are changed later forms of their single common parent language (see also the Comparative Method).

191

192

Glossary

Proximate: The more prominent third person in an Algonquian sentence containing two third-person (‘he/she/it/they’) forms. The less important third person is marked as obviative. So, for instance, imagine a story about someone named Mildred. In an Algonquian translation of the sentence Mildred saw Bill on the street, and she greeted him, Mildred is the proximate third person and Bill and him will be marked as obviative. Proyecto para la Documentación de las Lenguas de Meso-América (PDLMA): A major documentation project in Mexico, organized and codirected by Terrence Kaufman (see his 2001 article ‘Two highly effective models for large-scale documentation of endangered languages’) in Mexico, starting in 1993. The PDLMA is an example of Kaufman’s parallel project approach to language documentation in a linguistically diverse region, where several linguists and their consultants live in the same place and eat and work together on different languages in a nine-week period of the summer. Among the languages that have been described on this project are San Miguel Chimalapa Western/Oaxaca Zoque, Ayapa Gulf Zoquean, Sayula Mixean, Eastern Mixe: SanJuan Guichicovi, Isthmus Zapotec: Juchitán, Coapan Northern Zapotec, Lachixío Mixtepec Zapotec, and Zenzontepec Chatino. Quechua: The most widely spoken indigenous language in the Americas, with up to 8 million speakers in Peru (2.5 million), Bolivia (3 million), Argentina (1 million), Ecuador (1.5 million), and Chile. There are numerous different Quechua dialects – some of them different enough that they should be considered separate languages (thus Quechuan languages), though the simple designation “Quechua” is usually retained for the entire set. Quechua was the language of the Inca empire that was destroyed by the Spanish invaders. It is sometimes thought to be related to Aymara, the second most widely spoken indigenous language in the Americas, but this genetic grouping remains highly controversial. Red book of endangered languages (UNESCO): See the UNESCO Atlas of the world’s languages in danger, which has now replaced the Red book. Regularity hypothesis of sound change: If, in one word in Language A, a sound x changes to y in Language A", a changed later form of A, then all the occurrences of x in similar environments will become y in A", not considering outside factors. The hedge “not considering outside factors” is needed because various linguistic and even nonlinguistic factors (such as social relations leading to dialect borrowing) can disturb the regularity of sound change. Nevertheless, in the great majority of sound changes the hypothesis makes the correct predictions. Rememberer: A member of an endangered-language speech community who can provide some words and phrases of the dying language, but who cannot use it in conversation. Romance languages: The languages descended from (daughter languages of) Latin, among them Catalan, French, Galician, Italian, Portuguese, Provençal, Romansh, Romanian, and Spanish.

Glossary

Root: The most basic, simplest form of an item in a major lexical class (usually a noun, verb, or adjective), to which no affixes of any kind have been added. In English, happy is an adjective root, and is also a word; happiness and unhappy are words but not roots because they each have an affix – a suffix on happiness, a prefix on unhappy. English pack is either a noun root or a verb root, depending on how it is used in a sentence. In some languages, at least some roots are never words because they must acquire one or more affixes before they can be used in sentences. An example is the Salish-Pend d’Oreille root laq’, which is used only in the word sˇclaq’mn ‘sneezeroot, Indian hellebore’ (a plant used in medicines); the word has three affixes, all untranslatable in this word: the prefixes s- and cˇ - and the suffix -mn. Safe language: On the UNESCO scale of Language Vitality and Endangerment, a language that meets this criterion: “spoken by all generations; intergenerational transmission is uninterrupted”. Salishan languages: A family of twenty-three Native American languages spoken in Washington, British Columbia, Oregon, Idaho, and Montana. SalishPend d’Oreille, spoken in Montana, is the westernmost Salishan language. All the living languages in the family are gravely endangered. The family comprises five branches: Bella Coola; Coast Salish (Comox, Pentlatch, Sechelt, Squamish, Halkomelem, Nooksack, Straits Salish [two languages, Northern Straits and Clallam], Twana, Lushootseed); Tillamook; Tsamosan (Upper Chehalis, Lower Chehalis, Cowlitz, Quinault); and Interior Salish, with two subbranches, Northern (Lillooet, Thompson, Shuswap) and Southern (Columbian, Colville-Okanagan, Coeur d’Alene, and Montana Salish-Kalispel-Spokane). Salish-Pend d’Oreille (also called Montana Salish and – mainly by Whites – Flathead): A Native American language belonging to the Southern Interior branch of the Salishan language family. The name comes from the names of two tribes, Bitterroot Salish and Pend d’Oreille, whose members speak the same language with (at present) undetectable dialect differences. The language is gravely endangered, with perhaps twenty fluent native speakers remaining as of 2015. Salvage linguistics: Linguistic fieldwork undertaken after the last fully fluent speakers are gone. If working with the remaining almost-fluent speakers, the fieldworker will have to jog consultants’ memory to help them think of words and structures that they haven’t used for decades. If the consultants have not used their language at all for decades, the need for retrieving old memories will be even more acute – in this case the consultants may be rememberers. If there are no remaining almost-fluent speakers, the linguist must glean whatever information is available from people whose knowledge of the language is entirely passive, not active. The goal in all these situations is to document as much as possible as fast as possible. The term “salvage linguistics” is considered pejorative by some community members, so that some linguists prefer to avoid it.

193

194

Glossary

Semantic domains: Groups of words in a language that fit into well-defined categories of meaning. Examples are birds, beasts, body parts, kinship terms, medicinal plants, placenames, and daily activities. Semi-speaker: Someone who doesn’t speak a particular language with full fluency because s/he never learned it fully. Typically, semi-speakers are the last speakers of a language that is undergoing attrition as it gradually dies – the children of some of the last fully fluent speakers. (By fluency, in this context, I mean full control of the language’s traditional structure, rather than the ability to speak the language quickly and confidently: some semi-speakers are able to carry on conversations with ease and confidence.) Semitic: A branch of the Afro-Asiatic language family. Semitic languages are spoken on the Arabian Peninsula, in North Africa, and in parts of sub-Saharan Africa, and they were prominent among ancient Near Eastern languages. The best-known Semitic languages are Arabic and Hebrew, but there are numerous other modern and ancient Semitic languages. Akkadian, for example, was one of the earliest known written languages (but it has been extinct for several thousand years, and it left no direct descendants). Slavic: One of the two subbranches of the Balto-Slavic branch of the IndoEuropean language family. Slavic languages, spoken primarily in eastern Europe, are further subdivided into East Slavic (Russian and its closest relatives), South Slavic (Serbian, Croatian, Macedonian, Bulgarian, and Slovenian), and West Slavic (Czech, Slovak, Polish, and their closest relatives). The oldest widely attested Slavic language is known as Old Church Slavic because of its primary use in Gospel translations and other religious writings. Sound-symbolic words: Words that imitate sights or (more often) sounds. Sound imitation is also called onomatopoeia; sound symbolism is a more general term. English examples of sound-symbolic words are smash, pow, and whoosh. Examples of sound-symbolic words in Salish-Pend d’Oreille are liqqqq ‘the sound of a lot of threads breaking (e.g., a shirt coming unraveled when you tear it)’ and cikw kw kw kw ‘little shiny things sparkling, like sequins on a dress or sparks flying up’. See Chapter 5 for a bit more discussion of this four-consonant formation. SOV word order: A sentential word order pattern in which the subject (S) precedes the object (O), which is then followed by the verb (V). Standard dialect: The official or unofficially acknowledged “correct” dialect of a language, typically the most prestigious dialect. A standard dialect is established as the literary and/or spoken form that is considered “correct”, taught in schools, and used instead of nonstandard local dialects in broader public discourse. English and other major languages around the world have long had well-established standard forms. But a great many minority languages, including

Glossary

(or in particular) those that are endangered, have no standard form: all dialects are roughly equal in prestige and spheres of usage, there is typically no written form, and the language is not taught in schools. Stem: A linguistic form containing one or more morphemes to which one or more additional morphemes can be added. In English, for instance, roots are also stems; but some stems are more complex. The verb root lock is a stem that can take a prefix un- to form the word unlock; and unlock is also a stem, because the suffix -able can be added to it, to form unlockable, meaning ‘able to be unlocked’. But if we first add -able to the root lock, we get the word lockable ‘capable of being locked’, which is also a stem; and if we then add the prefix un- to lockable, we get unlockable, which means ‘incapable of being locked’. Unlockable, in either meaning, is a word but not a stem, because no further affixes can be added to it. Stop consonant: A speech sound in which the air stream is completely blocked in the oral cavity until the articulation is released. All human languages have at least two or three stops; English, for instance, has six: the voiceless stop phonemes /p t k/ and the voiced stop phonemes /b d g/. Many languages (unlike English) have only voiceless stops; few or no languages have only voiced stops. Subordinate clause: A clause that is part of a sentence and is an adjunct to some part of speech elsewhere in the sentence. English examples of subordinate clauses are when we were Elvis fans and because you didn’t go to that movie in the sentences We used to go to Graceland when we were Elvis fans and You missed seeing Alice because you didn’t go to that movie. These two clauses are marked as subordinate by their subordinate conjunctions (when, because); comparable sentences with coordination rather than subordination would be We were Elvis fans and we used to go to Graceland and You didn’t go to that movie and so you missed seeing Alice. Many languages, like English, have subordinate clauses with finite verbs and subordinate conjunctions; but in many other languages subordination is expressed by non-finite verb forms (e.g., participles like English going or having gone). This distinction is a major typological split among languages, and a common locus of contact-induced change. Suffix: An affix that is added to the end of a stem to form a new word and/or a new stem. In English, for instance, the suffix -ness is added to adjectives to form nouns, as in happiness. Some English words have more than one suffix, e.g. realities, which consists of the adjective root real, a noun-forming suffix -ity, and a plural suffix -es. Sumerian: The language of the ancient Near Eastern kingdom of Sumer; it is now long extinct and is not known to have been related to any other language. SVO word order: A sentential word order pattern in which the subject (S) precedes the verb (V), which is then followed by the object (O).

195

196

Glossary

Swadesh list of basic vocabulary: See Basic vocabulary. Syntax: The structure of sentences. The syntactic features that are most commonly affected by contact-induced change (to judge by the examples in the literature) are sentential word order and other word order features, the syntax of subordination, and relative clause formation. Tense: Temporal information about processes and events that locates them in time relative to the time of utterance – present (occurring right now), past, and future. In English, for instance, I’m walking is present tense (and continuous aspect), I walked is past, and I will walk is future. See aspect for the other major type of time indication. Three-generation rule for language shift by immigrants: A predicted sequence in which adult immigrants speak only their heritage language and do not learn the language of their new country well; their children are bilingual in both languages; and their grandchildren are monolingual speakers of the language of the new country. Tip: A situation in which a minority language appears to be demographically stable within its speech community but then suddenly and rapidly declines in response to sociopolitical change, and then vanishes: the tip is from apparent robust life to rather sudden death. The concept of tip was introduced by Nancy Dorian (1981). For further discussion of tip, including examples, see Chapter 3. Tone: Distinctive pitch variation at the word or syllable level – that is, pitch differences that distinguish word meanings. Many, possibly most, of the world’s languages are tone languages; examples are Chinese, Zulu and most other Bantu languages, Thai, and Vietnamese. Tone contrasts with intonation, which usually refers to meaningful patterns of pitch variation that distinguish clause or sentence meanings. Toponymy: Place names, ‘the most direct link between a language and its territory, current or ancestral’ (Nash & Simpson 2012:392). Transcriber: ‘A tool for segmenting, labeling and transcribing speech’. trans.sourceforge.net/en/presentation.php (accessed 28 December 2013). Transitive verb: A verb that takes an object. In English, the verb find is transitive, as evidenced by the fact that it requires an object (as in I found the book). Typology: A subfield of linguistics that focuses on the distribution of structural linguistic features in languages of the world. It is typological investigation that provides evidence for such statements as these: only a very few languages of the world lack nasal consonants; suffixes are more numerous than prefixes

Glossary

crosslinguistically; and SOV word order is more common crosslinguistically than VSO word order. Unattested language: A language that once existed but is not documented in any way – that is, there are no written records of the language at all. The existence of the vast majority of unattested languages is surely completely unknown, but under two circumstances scholars know that an unattested language existed: first, if the language is mentioned by name in old documents (but without any sample words or other data from the language); and second, if the language was the parent of a language family – the proto-language – in which case information about its vocabulary, sound system, and word structure can be reconstructed by means of the Comparative Method, comparing the vocabularies and structures of its descendent languages. UNESCO: The United Nations Educational, Scientific, and Cultural Organization. UNESCO’s scale of language endangerment: See Atlas of the world’s languages in danger. Uralic: A family of languages spoken around the Baltic Sea, in the interior of Russia, and in Hungary; the family is named for the Ural Mountains of Russia, the region that is thought to have been the original homeland of the parent language of this family. Three Uralic languages are official national languages: Finnish in Finland, Estonian in Estonia, and Hungarian in Hungary. Among the other Uralic languages mentioned in this book are Saami, Veps, Votic, Udmurt, and Forest Enets. uvular consonant: A consonant pronounced with the back of the tongue toward or against the uvula (the tip of the soft palate). Uvular sounds are pronounced farther back in the mouth than velar sounds. English has no uvular sounds; they are fairly uncommon in the world’s languages. velar consonant: A consonant pronounced with the back of the tongue toward or against the velum (the soft palate). English examples are /k/ and /g/; Russian and many other languages also have a velar fricative /x/ (the sound pronounced by some English speakers at the end of the name Bach). Veps: An endangered language belonging to the Finnic subbranch of the Uralic language family. It is spoken by a few hundred people in Russia in the vicinity of Lake Ladoga and is classified as “severely endangered” on the UNESCO scale of Language Vitality and Endangerment. Voiced sounds: Speech sounds produced with vibration of the vocal cords. All English vowel phonemes are voiced, as are such consonants as /m d v g l/.

197

198

Glossary

Voiceless sounds: Speech sounds, usually consonants, produced with no vibration of the vocal cords. All of the world’s languages have at least a few voiceless consonants; some examples of voiceless English consonants are /p t k s f/. Vowel harmony: A phonological process in which the vowels of one or more syllables assimilate phonetically to some other vowel(s) in the same word. The assimilation need not be, and usually is not, complete; more often, vowels assimilate (become more similar) in just one or two features, such as front/back tongue position or rounded/unrounded lips or high/non-high tongue position. So, for instance, the Turkish plural suffix -ler/-lar has a front vowel e when it is added to a root that has a front vowel, and a back vowel a when it is added to a root that has a back vowel: compare ev ‘house’ and evler ‘houses’ with yol ‘road’ and yollar ‘roads’. VSO word order: A sentential word order pattern in which the verb (V) precedes the subject (S), which is then followed by the object (O). Vulnerable language: On the UNESCO scale of Language Vitality and Endangerment, a language that meets this criterion: ‘most children speak the language, but it may be restricted to certain domains (e.g., the home)’. Yaaku language: A member of the Afro-Asiatic language family, spoken in Kenya. Yaaku is gravely endangered – it is classified as “extinct” in the UNESCO scale of Language Vitality and Endangerment – but in 2005 a visiting linguist found three fluent speakers. Efforts to revitalize the language are under way. See Chapters 3 and 7 for discussion of Yaaku.

Bibliography

Ahlers, Jocelyn C. 2009. The many meanings of collaboration: fieldwork with the Elem Pomo. Language and Communication 29:230–243. Aikio, Ante. 2008. The Saami languages: history, present situation, and work on documentation. Paper presented at a University of Utah colloquium. Aitchison, Jean. 1981. Language change: progress or decay? London: Fontana. Amery, Rob. 2000. Warrabarna Kaurna! Reclaiming an Australian language. Lisse: Swets & Zeitlinger. Ariste, Paul. 1970. Die Wege des Aussterbens zweier finnisch-ugrischer Sprachen [The routes to death of two Finno-Ugric languages]. Monda Lingvo-Problemo 2:77–82. Austin, Peter K. 2006. Gamilaraay. In Keith Brown, ed., Encyclopedia of language and linguistics, 2nd edn., vol. 4:724–727. Oxford: Elsevier. Austin, Peter K. 2008a. One thousand languages: living, endangered, and lost. Berkeley: University of California Press. Austin, Peter K. 2008b. Training for language documentation: experiences at the School of Oriental and African Studies. In Margaret Florey and Victoria Rau, eds., Documenting and revitalising Austronesian languages, 25-41. (Language Documentation and Convservation Special Publication 1.) Honolulu: University of Hawai‘i Press. http://hdl.handle.net/10125/1372 (accessed 12 May 2014). Austin, Peter K., and Julia Sallabank, eds. 2011. The Cambridge handbook of endangered languages. Cambridge: Cambridge University Press. Baker, Colin. 2011. Endangered languages: planning and revitalization. In Colin Baker, ed., Foundations of bilingual education and bilingualism, 5th edn., 40–63. Bristol, UK: Multilingual Matters. Bakker, Peter. 1996. “A language of our own”: the genesis of Michif—the mixed CreeFrench language of the Canadian Métis. Oxford: Oxford University Press. Bakker, Peter, and Robert A. Papen. 1997. Michif: a mixed language based on Cree and French. In Thomason, ed., 294–363. Baldwin, Daryl, Karen Baldwin, Jessie Baldwin, and Jarrid Baldwin. 2013. Miami: myaamiaataweenki oowaaha: Miami spoken here. In Leanne Hinton, ed., Bringing our languages home: language revitalization for families (Berkeles CA: Heyday Books). Baron, Dennis. 2009. Supporters of official English in the United States can learn from Slovakia. Web of Language website, 13 July 2009 (http://illinois.edu/db/view/25, accessed 12 September 2009). Bavin, Edith L. 1989. Some lexical and morphological changes in Warlpiri. In Dorian, ed., 267–286. 199

200

Bibliography

Bernard, H. Russell. 1996. Language preservation and publishing. In Nancy H. Hornberger, ed., Indigenous literacies in the Americas: language planning from the bottom up, 139–156. Berlin: Mouton de Gruyter. Bialystok, Ellen. 2009. Effects of bilingualism on cognitive and linguistic performance across the lifespan. In Ingrid Gogolin and Ursula Neumann, eds., Streitfall Zweisprachigkeit—The bilingualism controversy, 53–67. Wiesbaden: VS Verlag für Sozialwissenschaften. Blonk, Matthijs, Maarten Mous, and Hans Stoks. The last speakers—Yaaku language saved from extinction. http://wwww.matthijsblonk.nl/paginas/YaakuENG. htm, accessed 16 October 2009. Bloomfield, Leonard. 1927 [1970]. Literate and illiterate speech. American Speech 2/10:432–439. Reprinted in Charles F. Hockett, ed., A Leonard Bloomfield anthology, 147–156 (Bloomington: Indiana University Press, 1970). Bloomfield, Leonard. 1933. Language. New York: Holt, Rinehart and Winston. Boswell, James. 1955 (1791). The life of Samuel Johnson L.L.D. New York: The Modern Library. Bowern, Claire. 2008. Linguistic fieldwork: a practical guide. New York: Palgrave Macmillan. Bowern, Claire. 2010. Fieldwork and the IRB: a snapshot. Language 86:897–905. Bradley, David, and Maya Bradley, eds. 2002. Language endangerment and language maintenance. London: Routledge Curzon. Brenzinger, Matthias. 1992a. Lexical retention in language shift: Yaaku/MukogodoMaasai and Elmolo/Elmolo-Samburu. In Brenzinger, ed., 213–254. Berlin: Mouton de Gruyter. Brenzinger, Matthias, ed. 1992b. Language death: factual and theoretical explorations with special reference to East Africa. Berlin: Mouton de Gruyter. Brenzinger, Matthias, ed. 1998. Endangered languages in Africa. Köln: Rüdiger Köppe Verlag. Brenzinger, Matthias, Arienne M. Dwyer, Tjeerd de Graaf, Colette Grinevald, Michael Krauss, Osahito Miyaoka, Nicholas Ostler, Osamu Sakiyama, María E. Villalón, Akira Y. Yamamoto, and Ofelia Zepeda—UNESCO Ad Hoc Expert Group on Endangered Languages. 2003. Language vitality and endangerment. Document submitted to the International Expert Meeting on UNESCO Programme Safeguarding of Endangered Languages, Paris, 10–12 March 2003. Brown, Thomas. 2006. Did the U.S. army distribute smallpox blankets to Indians? Fabrication and falsification in Ward Churchill’s genocide rhetoric. Ann Arbor, MI: MPublishing, University of Michigan Library (http://quod.lib.umich.edu/p/plag/5240451.0001.009/–did-the-us-armydistribute-smallpox-blankets-to-indians?rgn=main;view=fulltext#N2, accessed 29 May 2014). Buszard-Welcher, Laura. 2001. Can the Web help save my language? In Hinton and Hale, eds., 331–345. Cahill, Michael. 2004. An appeal regarding endangered languages: letter to the editor. Language 80:1–2. Campbell, Lyle. 1994. Language death. In R.E. Asher, ed, The encyclopedia of language and linguistics, vol. 4:1960–1968. Oxford: Pergamon Press.

Bibliography

Campbell, Lyle. 1996. Defense strategies for endangered languages. In Hans Boebl, Peter H. Nelde, Zdenˇek Starý, and Wolfgang Wölck, eds., Kontaktlinguistik/Contact linguistics/Linguistique de contact, 659–666. Berlin: Walter de Gruyter. Campbell, Lyle. 1997. American Indian languages: the historical linguistics of Native America. New York and Oxford: Oxford University Press. Campbell, Lyle, and Martha C. Muntzel. 1989. The structural consequences of language death. In Dorian, ed., 181–196. Campbell, Lyle, and Verónica Grondona. 2010. Who speaks what to whom? Multilingualism and language choice in Misión La Paz. International Journal of American Linguistics 39:617–648. Campbell, Lyle, Raina Heaton, Eve Okura, Sean Simpson, Alex Smith, and John Van Way. 2014. New knowledge: findings from the Catalogue of endangered languages (“ELCat”), presented at the Annual Meeting of the Linguistic Society of America, January 2014. Carlson, Barry F., and Pauline Flett. 1989. Spokane dictionary. Missoula: University of Montana Occasional Papers in Linguistics, 6. Chelliah, Shobhana L., and Willem J. de Reuse. 2010. Handbook of descriptive linguistic fieldwork. Dordrecht and New York: Springer. Colapinto, John. 2007. The interpreter: has a remote Amazonian tribe upended our understanding of language? The New Yorker, 16 April 2007. Collins, James T. 1980. Laha, a language of the Central Moluccas. Indonesia Circle 23:3–19. Conn, Barry J. 2012. Botanical collecting. In Thieberger, ed., 250–280. Costa, David. 2003. The Miami-Illinois language. Lincoln: University of Nebraska Press. Crowley, Terry. 2007. Field linguistics: a beginner’s guide. Oxford: Oxford University Press. Crystal, David. 2000. Language death. Cambridge: Cambridge University Press. Cultural Diversity: Newsletter: The Diversity of Cultural Expressions (Quebec), vol. 8, no. 21 (www.diversite-culturelle.qc.ca/index.php?L=3D1#article5849). Reprinted on LINGUIST List, vol. 19-1967 (21 June 2008). Cunningham, Graham. 2006. Sumerian. In Keith Brown, ed., Encyclopedia of Language and Linguistics, 2nd edn., vol. 12:271–274. Oxford: Elsevier. Czaykowska-Higgins, Ewa. 2009. Research models, community engagement, and linguistic fieldwork: reflections on working within Canadian indigenous communities. Language Documentation and Conservation 3:15–50. Dauenhauer, Nora Marks, and Richard Dauenhauer. 1998. Technical, emotional, and ideological issues in reversing language shift: examples from Southeast Alaska. In Lenore A. Grenoble and Lindsay J. Whaley, eds., Endangered languages: current issues and future prospects, 57–98. Cambridge: Cambridge University Press. Dawkins, R.M. 1916. Modern Greek in Asia Minor. Cambridge: Cambridge University Press. Dench, Alan Charles. 1995. Martuthunira: a language of the Pilbara region of Western Australia. Canberra: Pacific Linguistics, Department of Linguistics, Research School of Pacific and Asian Studies, The Australian National University. Dench, Alan Charles. 1998. Yingkarta. Munich: LINCOLM Europa.

201

202

Bibliography

Dench, Alan. 2001. Descent and diffusion: the complexity of the Pilbara situation. In Alexandra Y. Aikhenvald and R.M.W. Dixon, eds., Areal diffusion and genetic inheritance: problems in comparative linguistics, 105–133. Oxford: Oxford University Press. Derbyshire, Desmond C. 1979. Hixkaryana. Amsterdam: North-Holland. Derhemi, Eda. 2014. The puzzle of assessing linguistic vitality: Arbëresh (Sicily) and Arvanitika (Greece). Paper presented at the 36th International LAUD Symposium: Endangerment of Languages across the Planet: The Dynamics of Linguistic Diversity and Globalization, University of Koblenz-Landau, Germany. Diaz, Rafael M. 1983. Thought and two languages: the impact of bilingualism on cognitive development. Review of Research in Education 10:23–54. Dobrin, Lise M. 2008. From linguistic elicitation to eliciting the linguist: lessons in community empowerment from Melanesia. Language 84:300–324. Dobrin, Lise M., and Josh Berson. 2011. Speakers and language documentation. In Austin and Sallabank, eds., 187–211. Dorian, Nancy C. 1973. Grammatical change in a dying dialect. Language 49:413–438. Dorian, Nancy C. 1977. The problem of the semi-speaker in language death. In Dressler and Wodak-Leodolter, eds., 23–32. Dorian, Nancy C. 1978. The fate of morphological complexity in language death: evidence from East Sutherland Gaelic. Language 54:590–609. Dorian, Nancy C. 1981. Language death: the life cycle of a Scottish Gaelic dialect. Philadelphia: University of Pennsylvania Press. Dorian, Nancy C. 1986. Abrupt transmission failure in obsolescing languages: how sudden the “tip” to the dominant language in communities and families? In Vassiliki Nikiforou, Mary VanClay, Mary Niepokuj, and Deborah Feder, eds., Proceedings of the 12th annual meeting of the Berkeley Linguistics Society, Feb. 15–17, 1986, 72–81. Berkeley: Berkeley Linguistics Society. Dorian, Nancy C., ed. 1989. Investigating obsolescence: studies in language contraction and death. Cambridge: Cambridge University Press. Dorian, Nancy C. 1993a. Internally and externally motivated change in language contact settings: doubts about dichotomy. In Charles Jones, ed., Historical linguistics: problems and perspectives, 131–155. London: Longman. Dorian, Nancy C. 1993b. A response to Ladefoged’s other view of endangered languages. Language 69:575–579. Dorian, Nancy C. 1994a. Varieties of variation in a very small place: social homogeneity, prestige norms, and linguistic variation. Language 70:631–696. Dorian, Nancy C. 1994b. Review of Robins and Uhlenbeck, eds. Language 70:797–802. Dorian, Nancy C. 1999. Linguistic and ethnographic fieldwork. In Joshua A. Fishman, ed., Handbook of language and ethnic identity, 25–41. Oxford: Oxford University Press. Dorian, Nancy C. 2001. Surprises in Sutherland: linguistic variability amidst social uniformity. In Newman and Ratliff, eds., 133–151. Dorian, Nancy C. 2010a. Investigating variation: the effects of social organization and social setting. Oxford: Oxford University Press. Dorian, Nancy C. 2010b. Documentation and responsibility. Language and Communication 30:179–185.

Bibliography

Dorian, Nancy C. 2014. Small-language fates and prospects: lessons of persistence and change from endangered languages: collected essays. Leiden: Brill. Dousset, Laurent. 2012. Understanding human relations (kinship systems). In Thieberger, ed., 209–234. Dressler, Wolfgang, and Ruth Wodak-Leodolter, eds. 1977. International Journal of the Sociology of Language 12: Language death. Dwyer, Arienne M. 2006. Ethics and practicalities of cooperative fieldwork and analysis. In Gippert et al., eds., 31–66. Edwards, Jonathan, Jr. 1823 [1788, 1787]. Observations on the language of the Muhhekaneew Indians: in which the extent of that language is shewn; its genius is grammatically traced; some of its peculiarities, and some instances of analogy between that and the Hebrew are pointed out. (Communicated to the Connecticut Society of Arts and Sciences, and published at the request of the Society.) Boston: Phelps and Farnham. [Origin edn. 1787, New Haven: Josiah Meigs. Reprinted 1788, London: W. Justins, Shoemaker-Row, Blackfriars.] Emlen, Nicholas Q. 2014. Language and coffee in a trilingual Matsigenka-QuechuaSpanish frontier community on the Andean-Amazonian borderland of southern Peru. Ann Arbor: University of Michigan dissertation. d’Errico, Peter. n.d. Jeffrey Amherst and smallpox blankets. http://www.nativeweb.org/ pages/legal/amherst/lord$_$jeff.html (accessed 29 May 2014). Ethnologue: languages of the world (www.ethnologue.com). Dallas: SIL International. Evans, Nicholas. 2008. Review of Gippert et al. 2006. Language Documentation and Conservation 2:340–350. Evans, Nicholas. 2010. Dying words: endangered languages and what they have to tell us. Oxford: Wiley-Blackwell. Everett, Daniel L. 2001. Monolingual field research. In Newman and Ratliff, eds., 166–188. Everett, Daniel L. 2005. Cultural constraints on grammar and cognition in Pirahã: another look at the design features of human language. Current Anthropology 46:621–646. Everett, Daniel L. 2009a. Don’t sleep, there are snakes: life and language in the Amazonian jungle. New York: Random House. Everett, Daniel L. 2009b. Cultural constraints on grammar in Pirahã: a response to some criticisms. Language 85:405–442. Farfán, Flores, José Antonio, and Fernando Ramallo, eds. 2010. New perspectives on endangered languages. Amsterdam: John Benjamins. Fenyvesi, Anna. 1995. Language contact and language death in an immigrant language: the case of Hungarian. Pittsburgh: University of Pittsburgh M.A. thesis (published in University of Pittsburgh Working Papers in Linguistics 3.1–117). Fenyvesi, Anna. 2014. Language endangerment and revitalization in the digital age: realities and responsibilities. In press in FUM, Finnisch-Ugrischen Mitteilungen. Fishman, Joshua A. 1965. Who speaks what language to whom and when. La Linguistique 2:67–88. Fishman, Joshua A. 1991. Reversing language shift: theoretical and empirical foundations of assistance to threatened languages. Clevedon: Multilingual Matters.

203

204

Bibliography

Fishman, Joshua A., ed. 2001. Can threatened languages be saved? Clevedon: Multilingual Matters. Fishman, Joshua. 2004. Language maintenance, language shift, and reversing language shift. In Tej K. Bhatia and William C. Ritchie, eds., The handbook of bilingualism, 406–436. Oxford: Blackwell. Fortescue, Michael, Steven Jacobson, and Lawrence Kaplan. 1994. Comparative Eskimo dictionary with Aleut cognates. Fairbanks: Alaska Native Language Center, University of Alaska Fairbanks. Gal, Susan. 1979. Language shift: social determinants of linguistic change in bilingual Austria. New York: Academic Press. Gippert, Jost, Nikolaus P. Himmelmann, and Ulrike Mosel, eds. 2006. Essentials of language documentation. Berlin: Mouton de Gruyter. Goddard, Ives. 1975. Algonquian, Wiyot, and Yurok: proving a distant genetic relationship. In M. Dale Kinkade, Kenneth L. Hale, and Oswald Werner, eds., Linguistics and anthropology: in honor of C. F. Voegelin, 249–262. Lisse: Peter de Ridder Press. Goddard, Ives. 1994. The West-to-East cline in Algonquian dialectology. In William Cowan, ed., Actes du vingt-cinquième congrès des algonquinistes, 187–211. Ottawa: Carleton University. Golla, Victor. 2007. North America. In Moseley, ed., 1–96. Golluscio, Lucía A., and Hebe González. 2008. Contact, attrition and shift in two Chaco languages: the cases of Tapiete and Vilela. In K. David Harrison, David S. Rood, and Arienne Dwyer, eds., Lessons from documented endangered languages, 195–242. Amsterdam: John Benjamins. Greenberg, Joseph H. 1966 [1963]. Some universals of grammar with particular reference to the order of meaningful elements. In Joseph H. Greenberg, ed., Universals of language, 2nd edn., 73–113. Cambridge, MA: MIT Press. Grenoble, Lenore A., and N. Louanna Furbee, eds. 2010. Language documentation: practice and values. Amsterdam: John Benjamins. Grenoble, Lenore A., and Lindsay J. Whaley, eds. 1998. Endangered languages: current issues and future prospects. Cambridge: Cambridge University Press. Grenoble, Lenore A., and Lindsay J. Whaley. 2006. Saving languages: an introduction to language revitalization. Cambridge: Cambridge University Press. Grinevald, Colette. 2001. Encounters at the brink: linguistic fieldwork among speakers of endangered languages. In Osamu Sakiyama and Fubito Endo, eds., Lectures on endangered languages:2 – from Kyoto Conference 2000, 285– 313. Kyoto: Endangered Languages of the Pacific Rim. Haas, Mary R. 1958. Algonkian-Ritwan: the end of a controversy. International Journal of American Linguistics 24:159–173. Halberstam, David. 2007. The coldest winter: America and the Korean War. New York: Hyperion. Hale, Ken. 1992. Language endangerment and the human value of linguistic diversity. In Hale et al., 35–41. Hale, Ken. 2001a. Inuttut and Innu-aimun. In Hinton and Hale, eds., 351–352. Hale, Ken. 2001b. The Navajo language: III. In Hinton and Hale, eds., 385–387. Hale, Ken, Michael Krauss, Lucille J. Watahomigie & Akira Y. Yamamoto, Colette Craig, Laverne Masayesva Jeanne, and Nora C. England. 1992. Endangered languages. Language 68:1–42.

Bibliography

Harbert, Wayne. 2011. Endangered languages and economic development. In Austin and Sallabank, eds., 403–422. Harrison, K. David. 2007. When languages die: the extinction of the world’s languages and the erosion of human knowledge. Oxford: Oxford University Press. Haugen, Einar. 1953 [1969]. The Norwegian language in America: a study in bilingual behavior. Philadelphia: University of Pennsylvania Press [1969: Bloomington: Indian University Press]. Haugen, Einar. 1989. The rise and fall of an immigrant language: Norwegian in America. In Dorian, ed., 61–73. Hedblom, Folke. 1980. Swedish dialects in the Midwest: notes from field research. In Paul Schach, ed., Languages in conflict: linguistic acculturation on the Great Plains, 29–47. Lincoln: University of Nebraska Press. Heine, Bernd. 1975. Notes on the Yaaku language (Kenya). Afrika und Übersee 58:26– 61, 119–138. Henley, Jon. 1999. Chirac defends pure French tongue against regional Tower of Babel. The Observer, 27 June 1999 (www.guardian.co.uk/world/1999/jun/ 27/jonhenley.theobserver). Hill, Jane H. 1978. Language death, language contact, and language evolution. In William C. McCormack and Stephen A. Wurm, eds., Approaches to language: anthropological issues, 45–78. The Hague: Mouton. Hill, Jane H. 2002. ‘Expert rhetorics’ in advocacy for endangered languages: who is listening, and what do they hear? Journal of Linguistic Anthropology 12/2:119–133. Hill, Jane, and Kenneth Hill. 1977. Language death and relexification in Tlaxcalan Nahuatl. In Dressler and Wodak-Leodolter, eds., 55–69. Himmelmann, Nikolaus P. 1998. Documentary and descriptive linguistics. Linguistics 36:161–195. Himmelmann, Nikolaus P. 2006. Language documentation: what is it and what is it good for? In Gippert et al., eds., 1–30. Himmelmann, Nikolaus P. 2008. Reproduction and preservation of linguistic knowledge: linguistics’ response to language endangerment. Annual Review of Anthropology 37:337–350. Hinton, Leanne. 1994. Flutes of fire. Berkeley, CA: Heyday Books. Hinton, Leanne. 1997. Survival of endangered languages: the California MasterApprentice Program. International Journal of the Sociology of Language 123:177–191. Hinton, Leanne. 2001a. Language revitalization: an overview. In Hinton and Hale, eds., 3–18. Hinton, Leanne. 2001b. The Master-Apprentice Language Learning Program. In Hinton and Hale, eds., 217–235. Hinton, Leanne. 2001c. Training people to teach their language. In Hinton and Hale, eds., 349–350. Hinton, Leanne. 2001d. Languages of Arizona, southern California, and Oklahoma. In Hinton and Hale, eds., 367–369. Hinton, Leanne. 2001e. Sleeping languages: can they be awakened? In Hinton and Hale, eds., 413–417. Hinton, Leanne. 2011. Revitalization of endangered languages. In Austin and Sallabank, eds., 291–311.

205

206

Bibliography

Hinton, Leanne, ed. 2013. Bringing our languages home: language revitalization for families. Berkeley, CA: Heyday Books. Hinton, Leanne, and Ken Hale, eds. 2001. The green book of language revitalization in practice. San Diego: Academic Press. Hinton, Leanne, with Matt Vera and Nancy Steele. 2002. How to keep your language alive: a commonsense approach to one-on-one language learning. Berkeley, CA: Heyday Books. Holton, Gary. 2011. The role of information technology in supporting minority and endangered languages. In Austin and Sallabank, eds., 370–399. Huber, Máté. 2013. Intergenerational transmission of Hungarian as a heritage language in Canada: the macrosociolinguistics of the Hungarian community in Hamilton, Ontario. Szeged: University of Szeged M.A. thesis. Huntington, Samuel P. 2004a. The Hispanic challenge. Foreign Policy, March/April 2004:30–45. Huntington, Samuel P. 2004b. Who are we? The challenges to America’s national identity. New York: Simon & Schuster. Hurst, Charles E., and David L. McConnell. 2010. An Amish paradox: diversity and change in the world’s largest Amish community. Baltimore: Johns Hopkins University Press. Hymes, Dell. 1979. How to talk like a bear in Takelma. International Journal of American Linguistics 45:101–106. Ingoldsby, Pat. 2005. i’mouthere. Dublin: Willow Publications. Jackson, Jean. 1974. Language identity of the Colombian Vaupés Indians. In Richard Bauman and Joel Sherzer, eds., Explorations in the ethnography of speaking, 50–64. Cambridge: Cambridge University Press. Jackson, Jean. 1983. The Fish People: linguistic exogamy and Tukanoan identity in Northwest Amazonia. Cambridge: Cambridge University Press. Jansson, Johanna. 2014. Saving threatened languages – findings from an Old Order Amish setting. Paper presented at the 36th International LAUD Symposium: Endangerment of Languages across the Planet: The Dynamics of Linguistic Diversity and Globalization, University of Koblenz-Landau. Jefferson, Thomas. 1904 (1782). Notes on the state of Virginia. The works of Thomas Jefferson, ed. by Paul Leister Ford, vol. 3:510. New York and London: Knickerbocker Press and G.P. Putnam’s Sons. Johansen, Inger. 2009. Changes in a small language community. Paper presented at the 12th International Conference on Minority Languages, Tartu, Estonia. Johns, Alana, and Irene Mazurkewich. 2001. The role of the university in the training of native language teachers: Labrador. In Hinton and Hale, eds., 354–366. Jukes, Anthony. 2011. Researcher training and capacity development in language documentation. In Austin and Sallabank, eds., 423–445. Kaufman, Terrence. 2001. Two highly effective models for large-scale documentation of endangered languages. In Osamu Sakiyama and Fubito Endo, eds., Lectures on endangered languages:2 – from Kyoto Conference 2000, 269– 284. Kyoto: Endangered Languages of the Pacific Rim. Keiser, Steven Hartman. 2012. Pennsylvania German in the American Midwest. Durham, NC: Duke University Press. King, Jeanette. 2001. Te Kohanga Reo: Maori language revitalization. In Hinton and Hale, eds., 119–131.

Bibliography

King, Jeanette. 2014. Revitalizing the M¯aori language? Proceedings of the British Academy 199:215–230. Kolbert, Elizabeth. 2005. Last words. The New Yorker, 6 June 2005. Kornai, András. 2013. Digital language death. PLoS ONE 8/10: e77056. doi:10.1371/journal.pone.007056. Koven, Michèle. 2007. Selves in two languages: bilinguals’ verbal enactments of identity in French and Portuguese. Amsterdam: John Benjamins. Krauss, Michael. 1992. The world’s languages in crisis. Language 68:4–10. Kulick, Don. 1992. Language shift and cultural reproduction: socialization, self, and syncretism in a Papua New Guinean village. Cambridge: Cambridge University Press. Kushala, S. 2005. This village speaks gods language. The Times of India, 13 August 2005. Ladefoged, Peter. 1992. Another view of endangered languages. Language 68:809–811. Lee, Tiffany R., and Daniel McLaughlin. 2001. Reversing Navajo language shift, revisited. In Joshua Fishman, ed., Can threatened languages be saved? 23–43. Clevedon, UK: Multilingual Matters. Leonard, Wesley Y. 2007. Miami language reclamation in the home: a case study. Berkeley: University of California dissertation. Leonard, Wesley Y. 2010. iilaataweeyankwi: Miami language reclamation as decolonization. Martin Luther King, Jr., Linguistics Colloquium presentation, University of Michigan, 15 January 2010. Linguistic Society of America. 1994. The need for the documentation of linguistic diversity. www.linguisticsociety.org/files/lsa-stmt-documentationlinguistic-diversity.pdf Lizarralde, Manuel. 2001. Biodiversity and loss of indigenous knowledge in South America. In Luisa Maffi, ed., On biocultural diversity: linking language, knowledge, and the environment, 265–281. Washington, DC: Smithsonian Institution Press. Loh, Jonathan, and David Harmon. 2014. Comparing status and trends in linguistic and biological diversity. Paper presented at the 36th International LAUD Symposium: Endangerment of Languages across the Planet: The Dynamics of Linguistic Diversity and Globalization, University of Koblenz-Landau, Germany. Loprieno, Antonio. 1995. Ancient Egyptian: a linguistic introduction. Cambridge: Cambridge University Press. Lord, Nancy. 1996. Native tongues. Sierra Magazine 81/6:46-69. www.sierraclub.org/ sierra/199611/lord.asp (accessed 11 May 2014). Lozano, Elena. 1970. Textos vilelas. La Plata: CEILP. Lucy, John A. 1992. Language diversity and thought: a reformulation of the linguistic relativity hypothesis. Cambridge: Cambridge University Press. Lucy, John A. 1996. Grammatical categories and cognition: a case study of the linguistic relativity hypothesis. Cambridge: Cambridge University Press. Lucy, John A. 1997. Linguistic relativity. Annual review of anthropology 26:291–312. Macaulay, Monica. 2012. Training linguistics students for the realities of fieldwork. In Thieberger, ed., 457–472 (reprinted from Anthropological Linguistics 46:194–209, 2004). Magga, Ole Henrik. 2006. Diversity in Saami terminology for reindeer, snow, and ice. International Social Science Journal 58/187:25–34.

207

208

Bibliography

Mair, Victor. 2009. Shandong dialect intelligibility. Language Log post (on www. languagelog.org), 24 May 2009. Malykh, Artyom, Aleksey Shklyaev, and Olga Urasinova. Veme berykton as a social technology. Paper presented at the MinorEuRus Conference: Language Revitalization in a Russian and European Context: Exploring Solutions for Minority Language Maintenance, 16–18 December 2013, Helsinki. Mannheim, Bruce. 1991. The language of the Inka since the European invasion. Austin: University of Texas Press. Martin, Laura. 1986. Eskimo words for snow: a case study in the genesis and decay of an anthropological example. American Anthropologist 88/2:418. Mattina, Anthony, compiler. 1987. Colville-Okanagan dictionary. Missoula: University of Montana Occasional Papers in Linguistics, 5. McCarty, Teresa L., Lucille J. Watahomigie, Akira Y. Yamamoto, and Ofelia Zepeda. 2001. Indigenous educators as change agents: case studies of two language institutes. In Hinton and Hale, eds., 371–383. McConvell, Patrick, and Felicity Meakins. 2005. Gurindji Kriol: a mixed language emerges from code-switching. Australian Journal of Linguistics 25:9–30. McConvell, Patrick, and Nicholas Thieberger. 2001. State of Indigenous languages in Australia – 2001. Australia State of the Environment Second Technical Paper Series (Natural and Cultural Heritage), Department of the Environment and Heritage, Canberra. www.ea.gov.au/soe/techpapers/index.html (accessed 8 May 2014). McKay, Graham. 2011. Policy and indigenous languages in Australia. Australian Review of Applied Linguistics 34:297–319. McMahon, April M.S. 1994. Understanding language change. Cambridge: Cambridge University Press. Meakins, Felicity. 2012. Which mix? Code-switching or a mixed language? – Gurindji Kriol. Journal of Pidgin and Creole Languages 27:105–140. Mengarini, Gregory, S.J., Joseph Giorda, Leopold van Gorp, Joseph Bandini, and Joseph Guidi. 1877–79. A dictionary of the Kalispel or Flat-head Indian language. St. Ignatius, MT: St. Ignatius Print. [This work is usually referred to as Giorda 1877–79.] Menovšˇcikov, G.A. 1969. O nekotoryx social’nyx aspektax èvoljucii jazyka [On some social aspects of the evolution of language]. In Voprosy social’noj lingvistiki, 110–134. Leningrad: Nauka. Mertz, Elizabeth. 1989. Sociolinguistic creativity: Cape Breton Gaelic’s linguistic “tip”. In Dorian, ed., 103–116. Mitchell, Blackhorse, and Anthony K. Webster. 2011. “We don’t know what we become”: Navajo ethnopoetics and an expressive feature in a poem by Rex Lee Jim. Anthropological Linguistics 53:259–286. Mithun, Marianne. 1981. Stalking the Susquehannocks. International Journal of American Linguistics 47:1–26. Mithun, Marianne. 1999. The languages of Native North America. Cambridge: Cambridge University Press. Mithun, Marianne. 2001. Who shapes the record: the speakers and the linguist. In Newman and Ratliff, eds., 34–54. Moe, Ron. 2001. Lexicography and mass production. SIL: Notes on Linguistics 4, pp. 150–156.

Bibliography

Moran, Steven, Daniel McCloy, and Richard Wright. 2012. Revisiting population size vs. phoneme inventory size. Language 88:877–893. Moseley, Christopher, ed. 2007. Encyclopedia of the world’s endangered languages. London and New York: Routledge. Mous, Maarten. 2003. The making of a mixed language: the case of Ma’a/Mbugu. Amsterdam: John Benjamins. Muysken, Pieter. 1997. Media Lengua. In Thomason, ed., 365–426. Nash, David, and Jane Simpson. 2012. Toponymy: recording and analysing placenames in a language area. In Thieberger, ed., 392–404. Nathan, David. 2011. Digital archiving. In Austin and Sallabank, eds., 255–273. Nettle, Daniel, and Suzanne Romaine. 2000. Vanishing voices: the extinction of the world’s languages. Oxford: Oxford University Press. Nevins, Andrew, David Pesetsky, and Celine Rodrigues. 2009. Pirahã exceptionalism: a reassessment. Language 85:355–404. Newman, Paul. 2009. Fieldwork and field methods in linguistics. Language Documentation and Conservation 3:113–125. Newman, Paul, and Martha Ratliff. 2001. Introduction. In Newman and Ratliff, eds., 1–14. Newman, Paul, and Martha Ratliff, eds. 2001. Linguistic fieldwork. Cambridge: Cambridge University Press. Nicholson, Rangi. 1997. Marketing the Maori language. In Jon Reyhner, ed., Teaching indigenous languages, 206–213. Flagstaff: Northern Arizona University. Noonan, Michael. 1999. The fall and rise and fall of the Chantyal language. Milwaukee Studies on Language 9:248–260. Norman, Jerry. 1988. Chinese. Cambridge: Cambridge University Press. Núñez, Rafael, and Eve Sweetser. 2006. With the future behind them: convergent evidence from Aymara language and gesture in the crosslinguistic comparison of spatial construals of time. Cognitive Science 30:401–450. O’Shannessy, Carmel. 2013. The role of multiple sources in the formation of an innovative auxiliary category in Light Warlpiri, a new Australian mixed language. Language 89:328–353. Palosaari, Naomi, and Lyle Campbell. 2011. Structural aspects of language endangerment. In Austin and Sallabank, eds., 100–119. Paulston, Christina Bratt. 1994. Linguistic minorities in multilingual settings: implications for language policies. Amsterdam: John Benjamins. Payton, Philip J. 2000. Cornish. In Glanville Price, ed., Languages of Britain and Ireland, 109–119. Oxford: Blackwell. Payton, Philip J. 2006. Cornish. In Keith Brown, ed., Encyclopedia of Language and Linguistics, 2nd edn. Oxford: Elsevier, vol. 3:205–206. Peacock, John Hunt. 2010. The good news and the bad news about reviving my Spirit Lake Dakota Nation’s endangered language. In Hywel Glyn Lewis and Nicholas Ostler, eds., Reversing language shift: how to re-awaken a language tradition, 132–139. Bath: Foundation for Endangered Languages. Pemberton, Mary. 2008. Last native Eyak speaker dead at 89. The Associated Press: AP Online, 23 January 2008. Penfield, Susan D., Angelina Serratos, Benjamin V. Tucker, Amelia Flores, Gilford Harper, Johnny Hill, Jr., and Nora Vasquez. 2008. Community collaborations: best practices for North American indigenous language documentation. International Journal of the Sociology of Language 191:187-202.

209

210

Bibliography

Perley, Bernard C. 2011. Defying Maliseet language death: emergent vitalities of language, culture, and identity in eastern Canada. Lincoln: University of Nebraska Press. Perley, Bernard C. 2012. Zombie linguistics: experts, endangered languages and the curse of undead voices. Anthropological Forum: A Journal of Social Anthropology and Comparative Sociology 22/2:133–149. Pite Saami Documentation Project. http://saami.uni-freiburg.de/psdp/ (accessed 10 January 2014). Platero, Paul R. 1992. Navajo Head Start language study. Window Rock, Navajo Nation, AZ: Division of Diné Education. Poplack, Shana. 1980. Sometimes I’ll start a sentence in Spanish y termino en español. Linguistics 18:581–618. Poser, William J. n.d. On the end of the Ritwan controversy. www.billposer.org/Papers/ oerc.pdf (accessed 28 May 2014). Prensky, Marc. 2001a. Digital natives, digital immigrants. On the horizon 9/5:1–6. Prensky, Marc. 2001b. Digital natives, digital immigrants, Part 2: do they really think differently? On the horizon 9/6:1–6. Pullum, Geoffrey K. 1991. The great Eskimo vocabulary hoax and other irreverent essays on the study of language. Chicago: University of Chicago Press. Puura, Ulrikka. 2009. Searching for aspectuality in Finnic: how to interpret morphosyntactic change in Vepsian? Paper presented at the Twelfth International Conference on Minority Languages, Tartu, Estonia. Rao, Subha J. 2008. Keeping Sanskrit alive. The Hindu, 2 March 2008. Raymond, Joan. 1998. Say what? Preserving endangered languages. Newsweek, 14 September 1998. Reichard, Gladys. 1951. Navaho grammar. New York: J.J Augustin Publisher. Reintges, Chris H. 2004. Coptic Egyptian (Sahidic dialect): a learner’s grammar. Cologne: Rüdiger Köppe Verlag. Rice, Keren. 2006. Ethical issues in linguistic fieldwork: an overview. Journal of Academic Ethics 4:123–155. Rice, Keren. 2011. Fieldwork. Oxford: Oxford University Press: Oxford Bibliographies (www.oxfordbibliographies.com). Robins, Robert H., and Eugenius M. Uhlenbeck, eds. 1991. Endangered languages. Oxford: Berg. Rogers, Chris, and Lyle Campbell. 2012. Endangered languages. Oxford: Oxford University Press: Oxford Bibliographies (www.oxfordbibliographies.com). Romaine, Suzanne. 2007. Preserving endangered languages. Language and Linguistics Compass 1.1–2:115–132. Romaine, Suzanne. 2010. Contact and language death. In Raymond Hickey, ed., The handbook of language contact, 320–339. Oxford: Wiley-Blackwell. Romanova, Evgenia. 2007. The process of revitalization of culture and indigenous ethnic identity: the case of the Vepsian people in Karelia. Tromsø: University of Tromsø M.A. thesis, www.ub.uit.no/munin/bitstream/10037/1156/4/thesis. pdf (accessed (7 September 2009)). Sakel, Jeanette, and Daniel L. Everett. 2012. Linguistic fieldwork: a student guide. Cambridge: Cambridge University Press, 2012. Samarin, William J. 1967. Field linguistics: a guide to linguistic fieldwork. Cambridge: Cambridge University Press.

Bibliography

Sapir, Edward. 1913. Wiyot and Yurok, Algonkin languages of California. American Anthropologist n.s. 15:617–646. Sasse, Hans-Jürgen. 1991. Arvanitika: die albanischen Sprachreste in Griechenland, part 1. Wiesbaden: Harrassowitz. Schmidt, Annette. 1990. The loss of Australia’s Aboriginal language heritage. Canberra: Australian Institute of Aboriginal and Torres Strait Islander Studies. Shaw, Patricia A. 2004. Negotiating against loss: responsibility, reciprocity, and respect in endangered language research. In Osamu Sakiyama, Fubito Endo, Honore Watanabe, and Fumiko Sasama, eds., Lectures on endangered languages 4 – from Kyoto Conference 2001, 181–194. Osaka: Osaka Gakuin University. Shirobokova, Larisa. 2009. The role of sociolinguistics in progress and revitalization of minority Finno-Ugric languages. Paper presented at the 12th International Conference on Minority Languages, Tartu, Estonia. Slate, Clay. Promoting advanced Navajo language scholarship. In Hinton and Hale, eds., 389–410. Sosa, Lionel. 1998. The Americano dream: how Latinos can achieve success in business and in life. New York: Dutton Adult. Spicer, Edward. 1961. Perspectives in American Indian culture change. Chicago: University of Chicago Press. Splawn, A.J. 1944 (1917). Ka-Mi-Akin: last hero of the Yakimas, 2nd edn. Yakima, WA, Caldwell, Idaho: The Caxton Printers. Spolsky, Bernard. 1969. The Navajo reading study. Albuquerque: Linguistics Department, University of New Mexico. Spolsky, Bernard. 2009a. Language management for endangered languages: the case of Navajo. In Peter K. Austin, ed., Language documentation and description, vol. 6:117–131. London: SOAS. Spolsky, Bernard. 2009b. Rescuing Maori: the last 40 years. In Peter K. Austin, ed., Language documentation and description, vol. 6:11–36. London: SOAS. Sutton, Peter. 2001. Talking language. In Jane Simpson, David Nash, Mary Laughren, Peter Austin, and Barry Alpher, eds., Forty years on: Ken Hale and Australian languages, 453–464. Canberra: Pacific Linguistics. Swarns, Rachel L. 2001. A sleepy river’s people fight to keep old ways. New York Times, 29 December 2001, A4. Teeter, Karl V. 1964. Algonkian languages and genetic relationship. In Horace G. Lunt, ed., Proceedings of the Ninth International Congress of Linguists, 1026– 1034. The Hague: Mouton. Thieberger, Nicholas, ed. 2012. The Oxford handbook of linguistic fieldwork. Oxford: Oxford University Press. Thomason, Lucy. 1995. The assignment of proximate and obviative in informal Fox narrative. In David Pentland, ed., Papers of the Twenty-Sixth Algonquian Conference, 462–496. Winnipeg: University of Manitoba. Thomason, Lucy. 2003. The proximate and obviative contrast in Meskwaki. Austin: University of Texas dissertation. Thomason, Lucy, and Sarah Thomason. 2004. Truncation in Montana Salish. In Donna B. Gerdts and Lisa Matthewson, eds., Studies in Salish linguistics in honor of M. Dale Kinkade, 354–376. Missoula: UMOPL-Linguistics Laboratory, University of Montana. Thomason, Sarah G. 1997. Mednyj Aleut. In Thomason, ed., 449–468.

211

212

Bibliography

Thomason, Sarah G., ed. 1997. Contact languages: A wider perspective. Amsterdam: John Benjamins. Thomason, Sarah G. 1999. Crackle, plop, twinkle: sound-symbolic words in Montana Salish. In Mercedes Q. Hinkson, ed., Papers for the 34th International Conference on Salish and Neighboring Languages, 251–266. Kamloops, BC: Secwepemc Cultural Education Society and Simon Fraser University. Thomason, Sarah G. 2000. On the unpredictability of contact effects. Estudios de Sociolingüística 1/1:173–182. Thomason, Sarah G. 2001. Language contact: an introduction. Edinburgh and Washington, DC: Edinburgh University Press and Georgetown University Press. Thomason, Sarah G. 2007–2008. At a loss for words. Natural History December 2007/January 2008:24–29. Thomason, Sarah G. 2010. Recovering and preserving linguistic knowledge: making a dictionary of an endangered language. Colloquium talk at the Smithsonian Institution, 11 January 2010. Thomason, Sarah G. 2013. How to avoid pitfalls in documenting endangered languages. In Elena Mihas, Bernard Perley, Gabriel Rei-Doval, and Kathleen Wheatley, eds., Responses to language endangerment: in honor of Mickey Noonan: new directions in language documentation and language revitalization, 79– 93. Amsterdam: John Benjamins. Thomason, Sarah G., and Terrence Kaufman. 1988. Language contact, creolization, and genetic linguistics. Berkeley: University of California Press. Timberg, Craig. 2005. A culture vanishes in the Kalahari Dust. Washington Post, 2 June 2005. Tsunoda, Tasaku. 2005. Language endangerment and language revitalization: an introduction. Berlin: Mouton de Gruyter. Turner, Nancy J., Laurence C. Thompson, M. Terry Thompson, and Annie Z. York. 1990. Thompson ethnobotany: knowledge and usage of plants by the Thompson Indians of British Columbia. Victoria, BC: Royal British Columbia Museum. Underwood, Emily. 2009. The lost art of listening. High Country News, November 23, 2009, 10–13. www.hcn.org/issues/41.20/the-lost-art-of-listening/article$_ $view?b$_$start:int=0&-C= (accessed 26 May 2010). UNESCO Atlas of the world’s languages in danger. www.unesco.org/culture/ languages-atlas (accessed 10 January 2014). UNESCO Red book of endangered languages. [This has now been replaced by the UNESCO Atlas of the world’s languages in danger.] Vakhtin, Nikolai. 1998. Copper Island Aleut: a case of language “resurrection”. In Grenoble and Whaley, eds., 317–327. Valijärvi, Riitta-Liisa, and Joshua Wilbur. 2011. The past, present and future of the Pite Saami language: sociological factors and revitalization efforts. Nordic Journal of Linguistics 34:295–329. Vaux, Bert, Justin Cooper, and Emily Tucker. 2007. Linguistic field methods. Eugene, OR: Wipf & Stock. Vidal, John. 2014. As forests are cleared and species vanish, there’s one other loss: a world of languages. The Observer, 7 June 2014. www.theguardian. com/environment/2014/jun/08/why-we-are-losing-a-world-of-languages (accessed 23 June 2014).

Bibliography

Watson, Seosamh. 1989. Scottish and Irish Gaelic: the giant’s bedfellows. In Dorian, ed., 41–59. Webster, Donald H., and Wilfried Zibell. 1970. Iñupiat Eskimo dictionary. Fairbanks, AK: Summer Institute of Linguistics. Wengle, John O. 1988. Ethnographers in the field: the psychology of research. Tuscaloosa: University of Alabama Press. Wilhelm, Andrea. 2014. Language revitalization. Oxford: Oxford University Press: Oxford Bibliographies (www.oxfordbibliographies.com). Will, Vanessa. 2004. The semiotics of socialization: Gaelic-medium education and language revival in Scotland. Paper presented at Michicago 2004 at the University of Chicago. Woodbury, Anthony C. 1994. In defense of the proposition, “When a language dies, a culture dies”. Austin: Texas Linguistic Forum 33:102–130. Woodbury, Anthony C. 2011. Language documentation. In Austin and Sallabank, eds., 159–186. Woolard, Kathryn A. 1989. Language convergence and language death as social processes. In Dorian, ed., 355–367. Yamada, Raquel M. 2007. Collaborative linguistic fieldwork: practical application of the empowerment model. Documentation and Conservation 1:257–282. Zepeda, Ofelia. 2005. More than just words: stages of American Indian language revitalization. Martin Luther King, Jr., Lecture, University of Michigan, 14 January 2005. Zuckermann, Ghil’ad, Shiori Shakuto-Neoh, and Giovanni Matteo Quer. 2014. Native tongue title: compensation for the loss of Aboriginal languages. Australian Aboriginal Studies (AAS) 24/1.

213

Language index

||Gana, 106, 110 Abenaki, Eastern, 96, 176; Western, 96, 176 Aboriginal English (Australia), 121 Aboriginal languages (Australia), 9, 17, 31, 56, 58, 62, 106, 110, 132, 133 Afro-Asiatic languages, 49, 175, 180, 194, 198 Ainu, 10 Akkadian, 1, 42, 43, 177, 194 Albanian, 28, 40, 66, 185; Arbëresh Albanian, 28, 40; Arvanitika Albanian, 28, 40, 66, 72 Aleut, 50–52, 104, 155, 183, 188; see also Bering Aleut, Mednyj Aleut Algic languages, 97, 109, 176; see also Algonquian languages, Ritwan languages. Algonquian languages, 57, 62, 95–97, 101, 104, 108, 109, 162, 166, 176, 187, 191 Ambonese Malay, see Malay. Anatolian languages, 43 Apache, 177 Apachean languages, 177 Arabic, 1, 19, 22, 37, 43, 49, 63, 65, 101, 194; Classical Arabic, 8, 103 Arapaho, 78, 96, 176; Northern Arapaho, 73, 90, 92 Arawakan languages, 81 Arbëresh, see Albanian. Armenian, 185 Arrernte, 9; see also Mparntwe Arrernte Arvanitika, see Albanian. Athabaskan languages, 27, 31, 45, 177 Athabaskan-Eyak languages, 27, 177 Atsina (Gros Ventre), 96, 176 Austronesian languages, 7, 64, 177, 178, 182, 187, 188 Aymara, 101, 103–104, 110, 192 Balto-Slavic languages, 194 Baluchi, 63 Bantu languages, 63, 65, 102, 178, 179, 196 Barí, 84, 92 Basque, 186 Bella Coola, 10, 193 Berber languages, 22, 175 Bering Aleut, 52

214

Bininj, 31, 40, 78 Bitterroot Salish, see Salish-Pend d’Oreille. Blackfoot, 74, 78, 90, 96, 176 Bosnian, 87 Brahui, 63 Breton, 25, 39, 179 Bulgarian, 194 Buryat, 19 Bushman, see San. Cacaopera, 45 Carrier, 31, 41, 170, 174 Catalan, 192 Caucasian languages, 148 Celtic languages, 25, 46, 179, 183, 185 Central Torres Strait, 9 Chadic languages, 175 Chantyal, 31, 40 Chatino, Zenzontepec, 192 Chehalis, Lower, see Lower Chehalis. Chehalis, Upper, see Upper Chehalis. Cheyenne, 96, 176 Chinese languages (“dialects”), 22, 35, 37, 41, 48, 69, 101, 196; Mandarin, 35; Wu, 118; see also Putonghua, Chinook, Clackamas, see Clackamas Chinook. Chinook, Lower (Shoalwater), 20 Chinook, Shoalwater, see Chinook, Lower. Chinookan languages, 29, 31; see also Chinook, Lower; Clackamas Chinook; Wasco; Wishram. Chorote, 32 Chukchi, 19 Clackamas Chinook, 32 Clallam (Klallam), 10, 193 Coast Salish languages, 193 Coeur d’Alene, 10, 193 Columbian, 10, 193 Colville-Okanagan, 10, 144, 193 Comox-Sliammon, 10, 193 Copper Island Aleut, see Mednyj Aleut. Coptic, 1, 8, 49, 66, 69, 155, 180, 182; see also Egyptian.

Language index Cornish, 46–48, 49, 52, 53, 57, 66, 69, 153, 154–155, 172, 179 Cowlitz, 10, 193 Cree, 96, 104, 110, 176 Croatian, 87, 194 Cupeño, 63 Cushitic languages, 49, 63, 64, 65, 88, 175, 180 Czech, 22, 63, 101, 194 Dahalo, 88 Dakota, 28, 40 Damin, 102, 109 Danish, 101 Delaware, 7, 32, 96, 176; Munsee, 96; Unami, 96 Dogon languages, 83, 84, 91, 101, 105 Dravidian languages, 63 Dutch, 22 Dyirbal, 62, 71 East Slavic languages, 194 East Sutherland Gaelic, see Gaelic, Scottish. Egyptian, 1, 8, 19, 42, 48–49, 52, 66, 69, 153, 155, 175, 180, 182; Demotic, 48; Old Egyptian, 48; Middle Egyptian, 48; Late Egyptian, 48; see also Coptic. Enets, Forest, 166, 197 English, 4, 12, 16, 20, 21, 22, 29, 32, 33, 34, 37, 46, 47, 48, 53, 54, 56, 58, 59, 62, 63, 64, 67, 74, 75, 77, 78, 83, 87, 90, 94, 101, 103, 108, 124, 131, 139–140, 143, 156, 169, 176, 180, 188, 194; see also Aboriginal English, Middle English, Old English, Standard English. Eskimoan languages, 82–83, 84, 91, 101, 105, 187; see also Yupik Estonian, 18, 26, 183, 197 Etchemin, 96 Ethiopic Semitic languages, 63 Even, 19 Evenki, 185 Eyak, 27, 40, 45–46, 49, 52, 66, 69, 78, 153, 154, 172, 177 Finnic languages, 8, 16, 65, 183, 197 Finnish, 101, 183, 197 Flathead, see Salish-Pend d’Oreille, Forest Enets, see Enets, Forest, Fox, see Meskwaki, French, 25, 37, 39, 43, 63, 75, 90, 101, 104, 110, 131, 139, 192 Gaelic, Irish, 85, 92, 103, 183 Gaelic, Scottish, 13, 34–35, 41, 53–54, 70, 85, 92, 108, 151, 179, 183; East Sutherland Gaelic, 53–54, 55, 66, 70, 71, 120–121 Galician, 192 Gamilaraay, 57, 71

Gaulish, 179 German, 22, 37, 63, 75, 98–100, 101, 181; High German, 98–100; Low German, 98–100; Standard German, 100; see also Pennsylvania German Gooniyandi, 166 Greek, 22, 24, 28, 42, 43, 49, 185; Asia Minor Greek, 22, 38; Attic Greek, 43; Modern Greek, 43; Standard Greek, 24 Gros Ventre, see Atsina. Gujarati, 101 Gurindji Kriol, 149 Hadza, 186 Halkomelem (Halq’emeylem), 10, 166, 193 Hawaiian, 165, 178 Hebrew, 6, 8, 43, 88, 101, 184, 194; Ephraimite Hebrew, 44, 67; Gileadite Hebrew, 44; Modern Israeli Hebrew, 6, 8, 156, 159 Hindi, 87, 101, 185 Hittite, 11, 42, 43, 184–185 Hixkaryana, 103, 109 Hokan languages, 163 Hungarian, 6, 22, 25, 101, 197; American Hungarian, 6, 61, 71; Canadian Hungarian, 170, 174; McKeesport Hungarian, 61, 62, 67 Illinois, see Miami-Illinois. Indic languages, 185 Indo-European languages, 43, 179, 184, 185, 186, 194 Indonesian, 101, 178 Ingrian, see Ižora. Innu-aimun, 173 Interior Salish, 193 Inuttut, 172 Iranian languages, 63, 185, 186 Irish Gaelic, see Gaelic, Irish. Iroquoian languages, 20, 95 Italian, 22, 24, 28, 43, 192 Itonama, 10 Ižora (Ingrian), 65, 72 Japanese, 37, 76–77, 101 Jumaytepeque Xinca, see Xinca. Kalispel, 10, 144, 193 Kannad.a, 101 Kansa (Kanza), see Kaw. Karelian, 166 Karo Batak, 118 Karuk, 163, 173 Kaw, 111, 146 Khinalug, 148 Khoe, 106 Khoisan languages, 102, 106, 179, 186 Klallam, see Clallam.

215

216

Language index

Korean, 76–77, 90, 101 Kriol (Australia), 121 Kunwinjku, 106, 110 Kurrama, 58 Kutenai, 10 Laha, 64, 65, 67, 72 Lakota, 166 Lao, 22 Lapp, see Saami. Lardil, 102, 109 Latin, 8, 42, 43, 48, 155, 177, 179, 180, 185 Latvian, 37 Leco, 10 Lenape, see Delaware. Lenca, 10, 45 Light Warlpiri, 121, 149 Lillooet, 10, 193 Loup, 96 Lower Chehalis, 193 Luiseño, 63 Lule, 60, 71 Lule-Vilela languages, 60 Lushootseed, 193 Ma’a (Mbugu), 65, 68, 72, 178 Maasai, 49–50, 53, 70, 155, 188 Macedonian, 194 Mahican, 96, 176 Malay, 64, 101, 178; Ambonese Malay, 64, 72 Malayalam, 118 Maliseet-Passamaquoddy (Maliseet), 3, 70, 96 Mam, 61; Tuxtla Chico Mam, 61, 71 Mande languages, 119 Mandinka, 119 Manx, 179 Maori, 6, 16, 165–166, 174, 178, 187, 188 Mapuzugun, 168 Marathi, 101 Martuthunira, 1, 15, 58 Massachusett, 96, 176; see also Wampanoag. Matsigenka, 81–82, 91 Mayan languages, 61, 101 Media Lengua, 104, 110 Mednyj Aleut (Copper Island Aleut), 50–52, 53, 66, 70, 104–105, 110, 153, 155, 164, 172, 178, 183, 188 Menomini, 62, 96 Meskwaki (Fox), 81, 82, 91, 96, 101, 110, 176; see also Sauk-Fox. Miami-Illinois (Miami), 57, 96, 162–163, 164, 173, 176 Michif, 104, 110, 178 Micmac, 96 Middle English, 181 Miriwoong, 166

Mixe, Eastern, 192 Mixean, Sayula, 192 Mohawk, 166, 174 Mohegan, 95, 96, 108 Montana Salish, see Salish-Pend d’Oreille. Montana Salish-Kalispel-Spokane, 10; see also Salish-Pend d’Oreille. Mparntwe Arrernte, 106, 110 Muhhekaneew, see Mohegan. Munda languages, 185 Mundari, 185 Muniche, 10 Munsee Delaware, see Delaware. Múra, 191 Nahuatl, 19, 55; Tlaxcalan Nahuatl, 55, 70 Nanticoke, 96 Narragansett, 96 Navajo, 5, 16, 36, 79–80 Nez Perce, 7, 16, 32, 190 Niger-Congo languages, 178 Nilo-Saharan languages, 49, 188 Nilotic languages, 49, 188 Nivaclé, 32 Nomlaki, 163, 173 Nooksack, 10, 193 Norse, 63 Norwegian, 17, 23, 28, 38, 101 Ojibwe, 166, 176; Central, 96; Eastern, 96; Northwestern, 96 Okanagan, see Colville-Okanagan. Old Church Slavic, 42, 43, 194 Old English, 42, 43, 64, 180 Old Irish, 42, 43, 183 Panyjima, 1, 58 Passamaquoddy, see Maliseet-Passamaquoddy Pend d’Oreille, see Salish-Pend d’Oreille. Pennsylvania Dutch, see Pennsylvania German. Pennsylvania German, 22–23, 26, 36, 39 Pentlatch, 193 Persian, 185 “Pidgin Nomlaki”, 163 Pipil, 61; Teotepeque Pipil, 61, 71 Pirahã, 5, 16, 121, 149, 191 Pitjantjatjara, 9 Polish, 22, 24, 37, 101, 194 Polynesian languages, 6, 165, 178 Pomo, 77, 90; Elem Pomo, 150 Portuguese, 5, 37, 43, 75, 90, 101, 192 Potawatomi, 96, 176 Powhatan, 96, 176 Proto-Algonquian, 97 Proto-Semitic, 64 Proto-Xincan, 62 Provençal, 192

Language index Puelche, 10 Putonghua (Standard Chinese), 35, 37 Quechua, 19, 38, 81–82, 91, 104, 192 Quechuan languages, 101, 192 Quinault, 10, 193 Quiripi, 96 Ritwan languages, 96, 108, 109; see also Wiyot, Yurok. Romance languages, 43, 179, 185, 192 Romanian, 43, 192 Romansh, 192 Russian, 18, 26, 27, 51–52, 66, 101, 105, 155, 166, 169, 170, 172, 183, 185, 188, 194 Rusyn (Ruthenian), 25 Ruthenian, see Rusyn. Saami (Sámi; Lapp), 39, 91, 101, 102, 109, 161, 166, 197; Lule Saami, 161; Northern Saami, 83, 84; Pite Saami, 161–162, 173; Southern Saami, 23, 26, 28, 36; Standard Saami, 28 Sahaptian languages, 190 Sahaptin, 32 Salishan languages, 10, 144, 187, 193 Salish-Pend d’Oreille, 20, 21, 37, 42, 55–56, 58–89, 66, 77, 80–81, 82, 89, 90, 92, 101, 102–103, 109, 124–125, 131, 132–133, 134, 135, 136, 138–140, 142–145, 154, 186, 190, 193, 194 Sámi, see Saami. Samoan, 106, 110 San, 26, 36, 39 San Miguel Chimalapa, 192 Sanskrit, 8, 16, 42, 43 Sauk-Fox, 96; see also Meskwaki. Scottish Gaelic see Gaelic, Scottish. Sechelt, 10, 193 Semitic languages, 1, 43, 63, 64, 175, 184, 194; see also Ethiopic Semitic. Serbian, 87, 194 Serbo-Croatian, 87, 123–124, 131, 132; Standard Serbo-Croatian, 123, 132 Seri, 106, 110 Seto, 18, 26 Shambala, 65, 178 Shawnee, 96, 176 Shuswap, 10, 193 Siouan languages, 28, 146, 166 Slavic languages, 194 Slovak, 22, 25, 194 Slovenian, 194 Snohomish, 10 South Slavic languages, 194 Spanish, 12, 21, 22, 30, 37, 43, 55, 61, 62, 81–82, 87, 101, 104, 131, 169, 192 Spokane, 10, 144, 152

Squamish, 10, 193 Standard (Mandarin) Chinese, see Putonghua. Standard English, 33, 34 Straits Salish, 193 Sumerian, 1, 42, 43, 44, 48, 69, 177, 195 Susquehannock, 20, 38, 67 Swahili, 37, 63, 88, 178 Swedish, 23, 28, 40, 161; American Swedish, 29, 34 Tadzhik see Tajik. Tahitian, 178 Taiap, 19 Tajik, 63 Takelma, 79, 80, 90, 101 Tamboran, 45, 67, 69 Tamil, 101 Taushiro, 10 Teotepeque Pipil, see Pipil. Thai, 37, 101, 196 Thompson Salish, 10, 84, 92, 107, 110, 193 Tillamook, 10, 193 Tlaxcalan Nahuatl, see Nahuatl. Tlingit, 27, 28, 39–40, 45, 46, 52 Tocharian, 185 Toda, 88 Tok Pisin, 7, 19, 161, 182, 185 Toro Tegu, 91, 161 Torres Strait Islander languages, 9 Tsamosan languages (Upper Chehalis, Lower Chehalis, Cowlitz, Quinault), 10, 193 Tukano (Tucano), 31 Tungusic languages, 185 Turkic languages, 63 Turkish, 22, 101, 103, 198 Tuxtla Chico Mam, see Mam. Twana, 10, 193 Udmurt (a. k. a., Votyak), 27, 40, 169, 197 Unami Delaware, see Delaware. Upper Chehalis, 10, 193 Uralic languages, 27, 65, 83, 166, 187, 197 Urdu, 87, 101 Uto-Aztecan languages, 61 Uzbek, 63 Veps, 8, 16, 27, 40, 183, 197 Vietnamese, 22, 37, 101, 196 Vilela, 60–61, 62, 71 Votic, 65, 67, 72, 197 Votyak, see Udmurt. Wampanoag, 96, 109, 171, 176; see also Massachusett. Warlpiri, 9; see also Light Warlpiri. Wasco (Chinook), 31, 32; see also Chinookan languages.

217

218

Language index

Wayeyi, 94, 105, 108 Welsh, 48, 179 West Slavic languages, 194 Wichí, 32 Wintuan languages, 163 Wishram (Chinook), 31, 32; see also Chinookan languages. Wiyot, 96, 97, 108, 109; see also Ritwan languages. Xinca, 61, 62; Jumaytepeque Xinca, 61–62, 71 Yaaku, 49–50, 52, 53, 66, 67, 69–70, 153, 154, 155, 172, 198

Yámana, 10 Yiddish, 156 Yindjibarndi, 1, 58 Yingkarta, 56, 71 Yuchi, 10 Yupik, 19 Yurok, 96, 97, 108, 109 see also Ritwan languages Zapotecan languages, 30, 40; Coapan Northern Zapotec, 192; Juchitán (Isthmus Zapotec), 30, 40, 192; Lachixío Zapotec, 192 Zoque, Western/Oaxaca, 192 Zoquean, Ayapa Gulf, 192 Zulu, 118, 179, 196

Names index

Ahlers, Jocelyn, 150 Aikio, Ante, 109 Aitchison, Jean, 72 Alexander, Danny, 155 Allen, Elsie, 90 Amherst, Lord Jeffrey, 38 Anderson, Greg, 148 Ariste, Paul, 72 Aristophanes, 43 Aristotle, 43 Austin, Peter K., 14, 41, 71, 147, 149, 150, 151, 152, 172 Baird, Jessie Little Doe, 109, 171 Baker, Colin, 171 Bakker, Peter, 110 Baldwin, Daryl, 162–163, 173 Baldwin, Jarrid, 173 Baldwin, Jessie, 173 Baldwin, Karen, 173 Baron, Dennis, 25, 39 Bavin, Edith, 71 Beaverhead, Pete, 80 Bechert, Johannes, 115, 147 Ben-Yehuda, Eliezer, 156, 159 Bernard, H. Russell, 1, 15 Berson, Josh, 172 Bialystok, Ellen, 92–93 Blonk, Matthijs, 70, 155 Bloomfield, Leonard, 62, 71, 99, 100, 109 Boas, Franz, 114, 147 Bonaparte, Prince Louis, 47 Boswell, James, 107 Bowern, Claire, 116, 117, 150, 168 Bradley, David, 14 Bradley, Maya, 14 Brenzinger, Matthias, 14, 15, 69–70 Brown, Thomas, 38 Buszard-Welcher, Laura, 172 Cahill, Michael, 104, 110 Campbell, Lyle, 14, 15, 32, 41, 66, 67, 68, 69, 71, 72, 108, 109 Carlson, Barry F., 152 Charles I. King of England, 48

Chelliah, Shobhana L., 116 Chirac, Jacques, 24–25, 39 Chomsky, Noam, 121 Churchill, Ward, 38 Colapinto, John, 149 Collins, James, 72 Conn, Barry J., 148 Cooper, James Fenimore, 96 Costa, David, 162, 164, 173 Craig, Colette, 13, 114, 146, 147; see also Colette Grinevald, Crowley, Terry, 116, 117 Crystal, David, 14, 68, 72, 157, 158, 172 Cunningham, Graham, 69 Czaykowska-Higgins, Ewa, 150 Dauenhauer, Nora Marks, 39, 153 Dauenhauer, Richard, 39, 153 Dawkins, R. M., 38 Dench, Alan, 1, 15, 56, 58, 71 de Reuse, Willem J., 116 Derbyshire, Desmond, 109 Derhemi, Eda, 28, 40 d’Errico, Peter, 38 Diaz, Rafael, 92 Dobrin, Lise M., 148, 151, 161, 172 Dorian, Nancy, 3, 13, 15, 40, 42, 53, 54, 55, 66, 68, 69, 70, 71, 72, 73, 88–89, 90, 93, 120, 122, 147, 148, 149, 150, 151, 196 Dousset, Laurent, 149 Drummond, William, 107 Dwyer, Arienne, 148 Edwards, Jonathan, 95 Edwards, Jonathan, Jr., 95, 108 Emlen, Nicholas Q. (Nick), 38, 40, 91 Endo, Fubito, 150 England, Nora C., 13, 146 Evans, Nicholas, 14, 106, 110, 114, 147 Everett, Daniel L, , 16, 116, 117, 121, 133, 148, 149 Farfán, Flores, 14 Fenyvesi, Anna, 67, 71, 72, 170, 174 Fishman, Joshua, 66, 72, 157, 158, 171, 172

219

220

Names index

Flett, Pauline, 152 Fortescue, Michael, 91 Furbee, N. Louanna, 116 Gal, Susan, 38 Garrett, Rev, John, 47 Gippert, Jost, 114, 146, 147, 148 Goddard, Ives, 108, 109 Golla, Victor, 16 Golluscio, Lucía, 71 Golovko, Evgenij V., 70 González, Hebe, 71 Greenberg, Joseph, 103, 109 Grenoble, Lenore A., 4, 14, 16, 116, 149, 156, 157, 158, 159, 164, 165, 166, 171, 172, 173, 174 Grinevald, Colette, 114, 147, 150; see also Colette Craig. Grondona, Verónica, 32, 41 Grey, Peter, 108 Haas, Mary, 108 Halberstam, David, 76, 77, 90 Hale, Ken (Kenneth L, ), 13, 14, 73, 79, 90, 93, 94, 101, 102, 108, 109, 146, 147, 153, 154, 171, 172, 173 Harbert, Wayne, 38, 158 Harmon, David, 2, 15 Harrison, K. David, 14, 68, 71, 105, 110, 148, 168 Haugen, Einar, 17, 38 Heath, Jeffrey, 91 Hedblom, Folke, 40 Heine, Bernd, 69–70 Henley, Jon, 39 Henry VIII, King of England, 48 Henry, Jim Pepper, 111 Herodotus, 43 Hill, Jane H., 15, 55, 70, 71, 172 Hill, Kenneth, 55, 70 Himmelmann, Nikolaus P., 114, 146, 147 Hinton, Leanne, 90, 153, 154, 162, 163, 164, 165, 166, 171, 172, 173, 174, 188 Holton, Gary, 174 Homer, 43 Howard, Victoria, 32, 41 Huber, Máté, 174 Huntington, Samuel P., 87, 92 Hurst, Charles, E., 39 Hymes, Dell, 79, 90 Ingoldsby, Pat, 73, 75, 77, 90 Ivi´c, Pavle, 123 Jackson, Jean, 40 Jacobs, Melville, 32, 41 Jansson, Johanna, 39

Jeanne, Laverne Masayesva, 13, 146 Jefferson, Thomas, 94, 95, 96, 108 Jim, Rex Lee, 90 Johansen, Inger, 39 John, Mary, Sr., 170, 174 Johns, Alana, 173 Johnson, Frances, 80, 90 Johnson, Samuel, 94, 95, 107 Jones, Marie Smith, 27, 31, 40, 45–46 Jukes, Anthony, 148 Kaufman, Terrence, 72, 129–130, 142, 150, 152, 190, 191 Keiser, Steven Hartman, 39 Kibrik, Aleksandr E., 148 Kilpatrick, Paul, 40 King, Jeanette, 16, 187 Kolbert, Elizabeth, 69 Kornai, András, 169, 174 Koven, Michèle, 75, 90 Krauss, Michael, 13, 46, 69, 79, 111, 142, 146, 152, 154, 172 Kulick, Don, 19, 38, 70 Kushala, S., 16 Ladefoged, Peter, 13, 88–89, 93 Leduey, Guillaume, 154 Lee, Tiffany R., 16 Lehmann, Christian, 115, 147 Leonard, Wesley Y., 153, 171, 173 Letsapa, Baagi, 94, 105, 108 Lizarralde, Manuel, 92 Loh, Jonathan, 2, 15 Loprieno, Antonio, 69 Lord, Nancy, 153, 172 Lozano, Elena, 71 Lucy, John, 91–92, 103, 109, 188 Macaulay, Monica, 149 Maffi, Luisa, 91 Magga, Ole Henrik, 91 Mair, Victor, 35, 41 Maldonaldo, Sergio, 73, 90 Malykh, Artyom, 174 Mandela, Nelson, 74, 75, 90 Mann, John, 47 Mannheim, Bruce, 38 Martin, Laura, 91 Mattina, Anthony, 152 Mayer, Mercer, 151 Mazurkewich, Irene, 173 McCarty, Teresa L., 173 McCauley, Johnny Ray, 111 McConnell, David L., 39 McConvell, Patrick, 10, 17, 149 McKay, Graham, 17 McLaughlin, Daniel, 16

Names index McMahon, April, 67, 72 Meakins, Felicity, 149 Mengarini, Gregory, S. J., 144 Menovšˇcikov, G. A., 70 Mertz, Elizabeth, 70 Mihas, Elena, 150 Mitchell, Blackhorse, 90 Mithun, Marianne, 38, 90, 108, 151 Moe, Ron, 151 Moran, Steven, 91 Mosel, Ulrike, 114, 146 Mous, Maarten, 70, 72, 155 Muntzel, Martha, 66, 69, 71, 72 Muysken, Pieter, 110 Nash, David, 149, 196 Nathan, David, 169, 174 Nettle, Daniel, 14 Nevins, Andrew, 149 Newman, Paul, 116, 117, 148, 149, 150, 151 Nicholson, Rangi, 16 Noonan, Michael (Mickey), 40, 150 Norman, Jerry, 69 Nuñez, Rafael, 104, 109 Obama, Barack, 169 O’Shannessy, Carmel, 119, 121, 149 Palosaari, Naomi, 109 Papen, Robert A., 110 Pata, Cody, 163, 173 Paulston, Christina Bratt, 39 Payton, Philip J., 69 Peacock, John Hunt, 28, 40 Pemberton, Mary, 69 Penfield, Susan, 150 Pentreath, Dolly, 46–48, 69 Perley, Bernard, 3, 15, 16, 70 Pike, Kenneth, 131–132, 150 Platero, Paul R., 16 Plato, 43 Poplack, Shana, 179 Poser, William J., 41, 109, 170, 174 Prensky, Marc, 174 Pullum, Geoffrey, 91 Puura, Ulrikka, 16 Rankin, Robert, 111, 146 Rao, Subha J., 16 Ratliff, Martha, 116, 117, 148, 149, 151 Raymond, Joan, 40 Reichard, Gladys, 91 Reintges, Chris, 69 Rice, Keren, 148 Robertson, David, 41 Robins, R. H., 13 Rogers, Chris, 14

Romaine, Suzanne, 14, 68, 69, 84, 92, 171 Romanova, Evgenia, 17, 40 Sakel, Jeanette, 116, 133 Sakiyama, Osamu, 150 Sallabank, Julia, 14 Salminen, Tapani, 39 Samarin, William J., 148 Samediggi, Rolf Olsen, 39 Sapir, Edward, 91, 108, 188 Sappho, 43 Sasse, Hans-Jürgen, 66, 72, 116 Schmidt, Annette, 17, 71 Shaw, Patricia A., 89 Shirobokova, Larisa, 40 Simpson, Jane, 149, 196 Sims, Carolyn, 92 Slate, Clay, 173 Sophocles, 43 Sosa, Lionel, 87, 92 Spicer, Edward, 41 Splawn, A. J., 16 Spolsky, Bernard, 6, 16 Stevens, Minnie, 46 Stoks, Hans, 70, 155 Stoyle, Mark, 69 Supahan, Terry, 163, 173 Sutton, Peter, 132, 133, 151 Swadesh, Morris, 143, 178 Swarns, Rachel L., 94, 108 Sweetser, Eve, 104, 109 Teeter, Karl, 108 Thieberger, Nicholas, 10, 17, 116, 148, 149 Thomason, Lucy, 91, 104, 109, 110, 151 Thompson, Laurence C., 92 Thompson, M. Terry, 92 Timberg, Craig, 39 Tito, Josip Broz, 123 Trautmann, Thomas, 108 Tsunoda, Tasaku, 171 Turner, Nancy J., 84, 92, 107, 110 Uhlenbeck, Eugenius M., 13, 116 Underwood, Emily, 90, 92 Vakhtin, Nikolai B., 70, 155, 172 Valijärvi, Riitta-Liisa, 173 Vaux, Bert, 116 Vidal, John, 15 Voltaire, 147 Watahomigie, Lucille J., 13, 146, 173 Watson, Seosamh, 92 Webster, Anthony K., 90 Webster, Donald H., 91 Wengle, John, 149

221

222

Names index

Whaley, Lindsay J., 4, 14, 16, 149, 156, 157, 158, 159, 164, 165, 166, 171, 172, 173, 174 White Thunder, 62, 71 Whitworth, Harriet (Alyé), 90 Whorf, Benjamin Lee, 91, 188 Wilbur, Joshua, 173 Wilhelm, Andrea, 171 Will, Vanessa, 35, 41 Woodbury, Anthony, 76, 90, 147

Woolard, Kathryn, 70 Wurm, Stephen A., 57, 116 Yamada, Raquel, 150 Yamamoto, Akira Y., 13, 146, 173 York, Annie Z., 92 Zenk, Henry, 41 Zepeda, Ofelia, 164, 165, 173 Zibell, Wilfried, 91 Zuckermann, Ghil’ad, 90

Subject index

accommodation to English syntax (by language consultants), 138–140 Ad Hoc Expert Group on Endangered Languages (UNESCO), 1, 15 adjective, predicative, 62, 175 adult language revitalization programs, 164, 165 advantages of learning to speak a consultant’s language, 132 Advocates for Indigenous California Language Survival, 167, 174 affix, 51, 62, 90, 113, 175, 185, 189, 193, 195 affixes, borrowed, 51 affricate, 98, 134 Africa, 9, 11, 14, 26, 31, 41, 52, 69, 97, 119, 179, 186, 194 Age of Exploration, 7 age of speakers as a factor in endangerment, 5, 7, 8, 18 agreement, 62, 175, 184, 185 agreement morphology, lost in attrition, 62 Alaska, purchased from Russia, 45 Albania, 28, 124 allophone, 61, 138, 176 American Association for the Advancement of Science, 168 analogic leveling of morphological paradigms in attrition, 62, 176 Andes, 91, 104 annotation of linguistic records, 112–113 applied linguistics and revitalization, 111, 160, 170 approval, local, for fieldwork projects, 127 archiving linguistic data, 113, 114, 117 Argentina, 10, 32, 41, 60, 192 Asia, 9, 97 aspect, 62, 176, 183, 185 aspect (grammatical category), lost in attrition, 62 assimilation (in speech sounds), 177 assimilation and endangerment, 18, 20, 64; see also resistance to complete cultural assimilation. assimilation phonological rules, lost in attrition, 61, 67 asymmetrical bilingualism/multilingualism, 12, 32, 41

Atlas of the World’s Languages in Danger (UNESCO), 2, 4, 14, 16, 34, 45, 47, 48, 49, 96, 153, 171, 177, 192, 197 attitudes, as a factor in endangerment, 20, 26–32, 36, 37, 52, 53, 54, 109; important for revitalization efforts, 30, 156, 158, 162, 163 attrition in a dying language, 42, 44, 57–64, 66, 67, 71, 153, 177 Australia, 9, 10, 17, 18, 19, 26, 31, 32, 56, 78, 102, 106, 109, 121, 132, 167, 174 Aztec Empire, 19 basic vocabulary, 138, 143, 151, 178 Bering Island, 51, 52; see also Commander Islands. bilingualism and language endangerment, 11–12 bilingualism, 11–12, 22, 88, 92–93; stable, 12, 32 bilingual mixed languages, 52, 65, 70, 104–105, 110, 155, 178, 188, 189 bilingual revitalization programs, 164, 166, 174 bilinguals, 21, 30, 52 biological diversity, 2, 15 blogging and revitalization, 169 boarding schools as a factor in endangerment, see residential schools. Bolivia, 10, 104, 192 Book of Common Prayer, English, 48, 69 borrowing, massive, 64–65, 66, 72 Botswana assimilation policy (targeting the San people), 26, 36, 39 Brazil, 5, 30, 31, 103, 121, 191 Breath of Life Institute for Indigenous Languages, National, 167, 174 British Columbia, 84, 107, 110, 170, 193 camps, language and culture, 164 Canada, 10, 18, 26, 95 Cape Breton Island (Nova Scotia), 54 Cascade Indians, 31 case (grammatical category), 62, 178, 185; lost in attrition, 62, 67 casinos, gambling, 158 Catalogue of Endangered Languages (EL-Cat), 2, 15 Catholic church, 24

223

224

Subject index

cattle-herding economy, 50, 53, 69 causative (grammatical category), 64, 179; changed as a result of language contact, 64 Chaco (Argentina), 32, 60, 71 Chile, 10, 168, 174, 192 Chilean Education Ministry, 168 Chirac Foundation (France), 25, 34 Civil War, U.S., 23, 87 CLAN (software), 119, 179 classroom-based revitalization programs, 166 click languages, 11, 102 click sounds, 101, 106, 179, 186 CLP, see Cornish Language Partnership code-switching, 104, 121, 149, 161, 179–180 collaborative fieldwork, 114, 129, 131, 147, 150 colonialism as a factor in endangerment, 19 Comité International Permanent des Linguistes (CIPL), 115 Commander Islands (Russia), 51; see also Bering Island, Mednyj Island. Committee on Endangered Languages and Their Preservation (LSA), 14 community as the prime mover in revitalization, 159–160 community-based revitalization programs, 164, 166 Comparative Method, the (in historical linguistics), 95, 97, 180, 184, 192, 197 complex sentence types, reduced in attrition, 62 conquest as a factor in endangerment, 19–21, 52 consultants, selection of (for a fieldwork project), 126, 133–134 contact-induced language change, 63–64, 180, 186, 195; in phonology, 63; in morphosyntax, 63–64 Copper Island, see Mednyj Island. Cornish Language Development Strategy, 155 Cornish Language Partnership (CLP), 154–155, 172 creole language, 7, 70, 178, 180, 189 creole (mixed-blood) population, 51, 52, 70 critically endangered languages (on UNESCO’s scale of endangerment), 4, 48, 52, 62, 96, 102, 155, 158 Croatia, 25 culture loss, 82, 84, 85, 88, 91, 167 culture shift, 74, 75 curriculum development for endangered languages, 160, 170 data collection techniques for dictionary-making, 142–145 data, naturally occurring, 140–141 daughter language, 43, 180 dead language, 42, 180–181 death, language, 3, 15 death by borrowing, 44

decreolization, 72 definitely endangered languages (on UNESCO’s scale of endangerment), 4, 96 degemination, 61, 67, 181, 184 descriptive linguistics (vs, documentary linguistics), 112–115, 145 devoicing phonological rule, 61, 181; lost in attrition, 61 dialect shift, 35 dialects and endangerment, 23 dialects, endangered, 6, 7, 24 diaspora fieldwork, 118, 119 dictionaries, talking, and revitalization, 168 Dictionary of the Kalispel or Flat-Head Language, 59, 144–145 dictionary-making, 111, 136, 142–145 digital archiving, 113, 114 digital ascent, 169 digital immigrant, 174 digital native, 169, 174, 181 digital technologies and revitalization, 157, 158, 163, 167–170, 172; see also blogging; digital archiving; digital ascent; machine translation and revitalization; smartphones and revitalization; social media and revitalization; dictionaries, talking, and revitalization; texting; twitter and revitalization; video streaming and revitalization. diglossia, 157, 181 diminutive marking, 79, 101 disadvantages of learning to speak a consultant’s language, 132 disaster, natural, as a cause of language death, 45 discourse prominence, 104 disease as a factor in endangerment, 2, 18, 20, 37, 45–46, 51, 53, 94, 95, 116 ditransitive construction, 138–140, 181 diversity, linguistic, 14, 18, 25, 74; a good thing, 25, 37, 74, 76, 85–89, 101; a bad thing, 74, 85–89; loss because of standardization, 32–35 documentary linguistics, 112–115, 145, 146–147, 148, 159, 178, 181 Documentation of Endangered Languages (DOBES), 15 dominant language of a bilingual, 21, 29, 56, 77 dormancy, language, 3, 13, 42, 46, 153, 167 dormant language, 57, 60, 162, 181 Dorobo (Kenya), 50 dual number (grammatical category), 181; lost as a result of language contact, 63 ecological linkages in folk taxonomy, 106 economic factors in revitalization, 158–159 economic power as a factor in saving languages, 18, 36, 38, 157, 158

Subject index economic pressures as a factor in endangerment, 18, 21–23, 26, 38, 46, 47, 52, 58, 83, 85, 88 Ecuador, 104, 192 education, and endangerment, 34, 36, 38, 41, 54, 56; in endangered languages, 34, 156, 157, 158, 159, 163, 166 ELAN (software), 119, 182 EL-Cat, see Catalogue of Endangered Languages. electronic technology and revitalization, see digital technologies and revitalization. elicitation, of word-lists, 136, 138; of paradigms, 116, 117, 119, 136, 138 El Salvador, 10, 44, 61, 69 Endangered Languages Project, 15 English Only movement (monolingual), 21, 38, 87, 88, 182 English Plus movement (bilingual), 23, 88, 182 Ephraimites, 44, 67, 88 Eskimo words for ‘snow’, 82–83, 91 Estonian Bureau for Lesser-Used Languages, 18, 37 ethics, fieldwork, 116, 117, 126, 127, 141–142, 148 Ethiopia, 63 ethnobotany, 58, 73, 84–85, 92, 105, 106–107, 110, 115, 117, 129, 148, 183; see also folk taxonomy ethnography and linguistic fieldwork, 113, 115, 148 ethnolect, 28 Ethnologue: Languages of the World, 5, 8–9, 10, 45, 52, 60, 62, 90, 191 ethnozoology, 73, 105, 115, 129; see also folk taxonomy European Charter on Regional and Minority Languages, 25 exclusive ‘we’, see inclusive/exclusive ‘we’ distinction. exogamy, 30, 40, 183 expressive devices in verbal art, 79–80 extinct languages (on UNESCO’s scale of endangerment), 4, 60, 95, 108, 177 Eyak Corporation of Alaska, 154 Eyak Language Project, 154, 172 Eyak Preservation Council, 154 failed revitalization efforts, value of, 170–171 family-based revitalization programs, 164, 165 farming economy, 58 field methods classes, 118–120, 149 fieldwork, linguistic, 54, 56, 58, 112, 116ff., 153 fieldwork textbooks, 116–117, 148 Fieldworks (software), 119, 183 financial resources in revitalization efforts, 157 finite verb inflection, 51, 183, 188, 195; borrowed, 51

Flathead Culture Committee, see Salish-Pend d’Oreille Culture Committee, Flathead Reservation (Montana), 20, 37, 124, 135, 137, 144; opened to White settlement, 20, 37, 56, 133 folk taxonomy, 105–106; see also ethnobotany, ethnozoology Foreign Policy magazine, 87, 92 France, 24, 25, 39, 90, 179, 187; minority language status in, 24–25, 39 fricative, 61, 79, 98, 175, 183, 187, 188, 197 frog stories (stimuli for text collection), 151 geminate consonants, 61, 183; degeminated in attrition, 61 gender (grammatical category), 63, 184; lost as a result of language contact, 63 genetic relationship of languages, 108, 143, 178, 184, 192 genocide as a cause of language death, 18, 37, 44 genres of texts, 113 Gileadites, 44, 88 globalization and endangerment, 154 glottalization of consonants, 62, 134, 184, 190; extended in attrition, 62, 71 government support as a factor in saving languages, 6, 18, 36, 156, 162 grammaticality judgment, 126, 141, 184 grasshoppers, 83, 105 Great Plains (U.S.), 28, 95, 96 Greece, 22, 24, 28, 40, 66 Greek Orthodox church, 24 Green Book of Endangered Languages, 153, 154, 171, 172, 173 Hangul, 76 Hans Rausing Endangered Languages Project (SOAS), 14, 15, 161 Hell Gate Treaty (1855), 20 High German Consonant Shift, 98–100 Hindu religion, 8 Hindus, 87 historical linguistics, 94, 95 Hittite Empire, 43 Holocaust, the, 68 homogeneity, linguistic, 85–87; as a promoter of peace, 85–87 Humanities Forum (Alaska), 154 Hungary, 197 hunter-gatherers, 20, 26, 50, 58, 125 identity, cultural, 34, 40, 41 identity, ethnic, 31, 40, 50 identity, language as a symbol of, 3, 31 immersion revitalization programs, 6, 164, 165 immigrants’ languages, loss of, 22, 40 Inca Empire, 19, 38

225

226

Subject index

inclusive/exclusive “we” distinction, 62, 63, 185; lost in attrition, 62, 71; lost as a result of language contact, 62, 63 India, 88, 185 Indian Peace Commission (U, S, , 1868), 86–87, 92 inflectional morphology, 51, 105, 183, 185 innate linguistic features, 100 Institutional Review Board (IRB), 127, 141, 142, 150, 185 International Committee for the Defense of the Breton Language, 39 International Congress of Linguists, 14th, 115, 147 International Phonetic Alphabet (IPA), 127, 150, 185 intonation, 138, 191 intransitive verb, 139, 186 IPA, see International Phonetic Alphabet. iPhone apps for endangered languages, 158 IRB, see Institutional Review Board. irregularities, grammatical, lost in semi-speakers, 55; see also regularization of irregular paradigms in attrition. isogloss, 100, 186 isolate (a language with no relatives), 10, 43, 45, 106, 163, 186 isolation, geographical, as a factor in saving languages, 22–23, 26, 36, 47

language shift, 2, 18, 22, 25, 35, 45, 47, 50, 52, 64, 74–75, 84, 156, 159, 172, 187, 196; see also dialect shift. “language suicide”, 44, 67–68, 72 language suppression policies, 18, 24, 25, 37, 76–77 Language Vitality and Endangerment Framework (UNESCO), 4, 19, 27, 189–90, 193, 198 lateral fricative, 61, 187; as an expressive device meaning ‘small size, condescension’, 79; lost in attrition, 61 lexical elaboration, 83, 101, 105 lexicography, see dictionary-making. lexifier language (of a pidgin or creole language), 180 lingua franca, 19, 156, 161, 187 “linguicide”, 67 linguistic fieldwork, see fieldwork, linguistic. linguistic relativity hypothesis (Sapir-Whorf hypothesis), 83–84, 91–92, 103–104, 109, 187–188 linguistics, descriptive, see descriptive linguistics. Linguistic Society of America (LSA), 14 Linguistic Survey of Scotland, Gaelic Division, 120 loanwords, 53, 64, 105 loss of artistic expression, 79–82 loss of cultural knowledge, 58, 106, 107 loss of vocabulary in endangered languages, 57–59, 71, 106

Jesuit missionaries, 144–145 Kansas, 146 Karelia, 27 Kenneth L, Hale Award (LSA), 14 Kenya, 49, 88, 154, 180, 188, 198 ‘Khinalug 2007: Creating a Digital Portrait of an Endangered Language’ (documentary film), 148 kibbutzim and the revival of Hebrew, 156, 159 kinship systems, 10, 77–78, 117, 148 kinship terms, Salish-Pend d’Oreille, 77–78, 90 Kohanga Reo (Language Nest) program (New Zealand), 187 Korean War, 76 labialized velar consonants, 63, 186, 190; acquired as a result of language contact, 63–64 Lajamanu (Australia), 121 “Lament for Eyak”, 46 language family, 10, 14, 27, 43, 45, 49, 60, 62, 65, 95–97, 175, 177, 184, 186, 189 “language genocide”, 67–68 “language murder”, 67–68, 72 language nests, 6, 165–166, 167, 187, 188 language, sacred (of a religion), 8, 156

machine translation and revitalization, 168 macro-historical factors in endangerment, 2, 18, 38 macrovariables in revitalization efforts, 156, 172 Mali, 83 Maliseet First Nation, 3 Mapuche community (Chile), 168–169, 174 markedness, 61–62, 188; marked sounds, lost in attrition, 60, 61 Master/Apprentice Language Learning Program, 164, 165, 166–167, 171, 174, 188; Eight Points of Learning in, 166–167 media, social, and revitalization, 158, 168 medicinal plants, knowledge of, 84–85, 94, 105, 106–107, 125, 170 Mednyj Island (Copper Island), 50–52; see also Commander Islands. Melanesia, 148, 151, 172 melting-pot ideology, 23–24, 36, 37, 188–189 Mesopotamia, 43 metadata, 113, 117 Mexico, 12, 21, 30, 40, 84, 87, 129, 131, 133, 151, 190, 191 Miami Nation, 162 Miami Tribe of Oklahoma, 153, 162–163; Language Committee, 153

Subject index Miami University (Ohio), 162–163 Microsoft, 168–169, 174 microvariables in revitalization efforts, 156, 172 MinorEuRus Conference (Helsinki, 2013), 174 minority language rights, 23, 24, 34 Misión La Paz (Argentina), 32, 41 mixed languages, see bilingual mixed languages. models, theoretical, of language death, 66–67 monolingual demonstration, 131–132, 150–151 monolingual fieldwork, 117, 131–132, 138, 148, 149, 150 monolingualism, 54 moribund language, 4, 6, 111, 142, 189 morpheme, 61, 183, 189 morphology, 51, 59, 83, 102, 188, 189; attrition in, 62, 79; see also inflectional morphology. morphosyntax, 62, 80, 117, 189 multilingualism, 11, 58, 156; non-prestigious, 32; prestigious, 31–32, 41 multimedia linguistic corpora, 114, 168 multinational pharmaceutical companies, 107 Muslim conquest of Egypt, 19, 49 Muslims, 19, 49, 87 Myaamia Project/Myaamia Center, 162–163, 173 ‘Myaamiaki Eemamwiciki: The Miami Awakening’ (documentary film), 171 names, personal, borrowed from French, 139 nation-states, 52 New Guinea, 97 New Zealand, 6, 165–166, 188 nonstandard dialects, 34, 100 Norway, 23, 26, 36, 39 number of speakers, 42; as a factor in saving languages, 36 obviative (in Algonquian morphosyntax), 81, 101, 104, 110, 190, 191 Official English laws, 21, 39, 182 official language laws, 25 official languages in nations, 25, 36, 104 OLAC (Open Languages Archives Community), 15 Old Order Amish, 23, 39 Old Order Mennonite, 23 onomatopoeia, see sound symbolism. orthography, 33, 34, 117, 127, 145, 156, 157, 161, 162; as a factor in revitalization, 34, 159, 160 ownership of stories and songs, 142, 151 Pacific Northwest linguistic area, 31, 41, 187, 190 Pan-American Highway, 30 Papua New Guinea, 7, 19, 38, 161 paradigm (set of grammatical forms), 62, 117, 136, 138, 141, 190

parallel project approach to language documentation, 129–130, 150, 190, 191 pear stories (stimuli for text collection), 151 pedagogical materials for endangered languages, 160 pedagogy, language, 117 person (grammatical category), 51, 62, 80, 81, 140 personality transplant, 75, 90 Peru, 10, 81, 91, 192 pharyngeal consonants, 134, 190 philology and revitalization, 160, 162 phoneme, 60, 138, 176, 191 phonological attrition, 60–62, 79 phonology, 60, 177, 191 pidgin language, 7, 8, 16, 178, 180, 182, 187, 189, 191–192 Pilbara, 1, 15, 71 Pite Saami Documentation Project, 161, 173 Pittsburgh, PA, 24, 38, 61 politics as a factor in endangerment, 18, 24–26 possessive morphology, lost in attrition, 62, 67 power and revitalization, 157, 158 pragmatics, 117, 191 prefix, 64, 175, 189, 191, 195 preparation for a field session, 119, 124, 135–136 prestige, as a factor in language endangerment, 21, 50; as a factor in revitalization, 36, 157, 158 Principles for Revitalization of First Nations Languages, Canada, 1990, 73 property, intellectual, 168 prosodic features, 113, 138, 189, 191 proto-language, 97, 180, 184, 192, 197 Proyecto para la Documentación de las Lenguas de Meso-América (PDLMA; Mexico), 129–130, 192 psychological stress during fieldwork, 122–125 quantifiers, 62 reciprocal kin terms, 78 reclamation revival programs, 164, 165, 181 reconstruction of unattested ancestral words, 95, 97, 180 Red Book of Endangered Languages (UNESCO), 14, 39, 153, 171, 192 reduplication, 144, 185, 189 Reformation, English, 48 regularity hypothesis (of sound change), 97, 192 regularization of irregular paradigms in attrition, 55, 62 reindeer economy, 26, 83, 84, 91, 101 religion as a factor in endangerment, 8, 22, 24, 28 rememberer, 56–57, 66, 128, 192 replacement, grammatical and lexical, as a route to language death, 64–65, 153

227

228

Subject index

Report to the President by the Indian Peace Commission, January 9, 1868, 86–87, 92 residential schools as a factor in endangerment, 12, 20, 26, 27, 37, 52, 56, 77, 90 resistance to complete cultural assimilation, 64, 65, 77; Irish resistance, 77; Native American resistance, 77 Resource Network for Linguistic Diversity in Australia, 167, 174 reversing language shift, 40, 157, 172 revitalization programs, 29, 56, 155; see also adult language revitalization programs, bilingual revitalization programs, classroom-based revitalization programs, community-based revitalization programs, family-based revitalization programs, immersion revitalization programs. revival programs, 48, 50, 126; see also reclamation revival program. Revolutionary War, U.S., 87 Rhenish Fan, 98–100, 109 Rhineland, the, 98–100 root (of a word), 82, 83, 103, 113, 139, 144, 175, 189, 193, 195 Russia, 8, 9, 45, 188, 197 Russian-American Company, 51 Russian in Siberia, 19 “safe” language, 4, 36–37, 101, 158, 169, 193 Salish-Pend d’Oreille Culture Committee (St, Ignatius, MT), 124, 127, 150 salvage linguistics, 3, 56, 58, 116, 126, 128, 134, 193 Sámi University College, 28, 29 Sanskrit villages, 8, 16 Sapir-Whorf hypothesis, see linguistic relativity hypothesis. school-based revitalization programs, 164–165 School of Oriental and African Studies (SOAS; London), 14, 147, 161 semantic domains, 57, 82, 83, 101, 105, 106, 142, 143, 151, 194 semi-speaker, 28, 50, 53, 54–56, 57, 59, 66, 67, 70, 71, 128, 153, 160, 177, 194 Serbia, 25 severely endangered languages (on UNESCO’s scale of endangerment), 4, 96 shibboleth, 44, 69 shift, see language shift. Siberia, 19, 166, 185 simplification of linguistic structure in attrition, 54, 60–63 sleeping languages, 3, 68, 156, 181 Slovak Republic language policy, 25 Slovakia, 25, 39 smallpox-infected blankets, 20, 38

smallpox as a factor in language endangerment, 20, 45, 51, 52, 53 smartphones and revitalization, 158, 168 Smithsonian Institution, 167 social media, see media, social. sound symbolism, 81, 103, 109, 144, 194 South America, 97 SOV word order, see Subject-Object-Verb word order. standard dialect, 6, 33, 194 standardization as a factor in endangerment, 18, 32–35, 37; and endangerment of nonstandard dialects), 18, 32–35, 100; and power, 33–34 status of minority languages, 6, 88 stem (of a word), 51, 139, 175, 185, 189, 191, 195 stigmatization of nonstandard forms, 33 stop consonant, 98–100, 195 structural features, borrowed, 63–64 Subject-Object-Verb (SOV) word order, 63, 103, 194; acquired as a result of language contact, 63 subordinate clauses, 63, 183, 195; reduced in attrition, 63 suffix, 51, 80, 175, 189, 195 suffixes, borrowed, 51 Sumbawa (Indonesia), 45 support, official, as a factor in saving languages, 6, 18, 36, 156, 162 Swadesh list, 143, 151 Sweden, 23, 26, 29, 36, 39, 161–162 syntax, 62, 189, 196; attrition in, 62–63, 79 talking dictionaries and revitalization, see dictionaries, talking, and revitalization. techniques for collecting linguistic data, 135, 138–141, 142–145, 151 technology and fieldwork, 116, 119, 126 technology and revitalization, see digital technologies and revitalization. tense (grammatical category), 51, 183, 185, 196 texting and revitalization, 169 ‘The Linguists’ (2008 documentary film), 148 three-generation rule (shift pattern for immigrant languages), 12, 24, 196 time and space orientation, 103–104 tip, 42, 44, 53–54, 66, 69, 70, 153, 196 tone distinctions, 63, 191, 196; lost as a result of language contact, 63 toponymy, 117, 148, 196 tracking characters in stories, grammatical devices for, 80–81 Transcriber (software), 196 transitive verb, 80–81, 139, 181, 196 truncation, 102, 103, 109

Subject index twitter and revitalization, 168, 169 typology, 109, 188, 195, 196–197 unattested language, 95, 97, 177, 180, 184, 192, 197 UNESCO, 1, 2, 4, 14, 16, 19, 27, 34, 36, 45, 47, 48, 49, 52, 60, 62, 96, 102, 109, 116, 153, 155, 171, 177, 189, 191, 192, 193, 197, 198 United States, 5, 10, 12, 18, 20, 21, 22, 23, 26, 39, 45, 77, 86–88, 95, 115, 127, 158 universal human learning strategies, 100–101 universal linguistic features, 10, 11, 100–101 uvular consonants, 60–61, 186, 197 variation, individual, 141–142, 151 variation, interspeaker, 142 variation, intraspeaker, 141, 142 Vaupès (Brazil), 30, 32, 40 velar consonants, 60–61, 186, 197 velar fricative as an expressive device meaning ‘large size’, 79–80, 90 velar/uvular phonological distinction, lost in attrition, 60–61, 71 Venezuela, 84

Vepsän Seura-Vepsian Society, 40 verbal art, 73, 79–82 Verb-Object-Subject word order (VOS), 139 video streaming and revitalization, 168 vocabulary, basic, see basic vocabulary voiced sounds, 61, 71, 181, 197 voiceless sounds, 61, 71, 181, 198 vowel harmony, 61, 198; lost in attrition, 61; lost as a result of language contact, 63 VSO word order, 103, 198 vulnerable languages (on UNESCO’s scale of endangerment), 4, 95, 97, 108, 177, 189, 191 Warradjan Aboriginal Culture Centre, Kakadu, 31 ‘We Still Live Here – Âs Nutayumeân’ (documentary film), 96, 171 World War II, 76 writing, ancient, 1 Yinka-Déné Language Institute, 174 Yokiri (Peru), 81–82, 91 Yugoslavia (former), 87, 123–124, 127, 128, 131, 132

229